Download Kinetic Facades...
Designing Kinetics for Architectural Facades Architecture has typically resisted kinetics, but with advances in dynamic screens and animated surface there are new opportunities for designers. This book examines the architectural facade, to identify a latent aesthetic within contemporary practice. How might precedent from architecture and the arts inform this new field? What are the design parameters, when composition shifts from stasis to a state of constant flux? And given this liquidity, what are the distinctive contours of this new aesthetic? These questions of precedent, design ontology and kinetic morphology are explored here in a strategic mix of theory and experiment. Analysis of architectural praxis intersects with discourse from kinetic art, to provide the conceptual spark for the study of ‘movement itself’. The potential is for the realization of indeterminate states, where parts coalesce, forming clusters and sublime patterns that resonate over time. This is the first book to articulate a framework for the design of kinetics developed from first principles. Located between theory and practice, the analytical diagrams and time-lapse images provide a unique and timely resource for designers, theorists and students interested in the potential of kinetics to enliven the public face of architecture. Jules Moloney is Associate Professor in Interdisciplinary Digital Design at the Victoria University of Wellington, New Zealand. His research, creative practice and teaching span the fields of architecture, kinetic art and virtual environments.
Designing Kinetics for Architectural Facades State change
First published 2011 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Simultaneously published in the USA and Canada by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business This edition published in the Taylor & Francis e-Library, 2011. To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk. © 2011 Jules Moloney The right of Jules Moloney to be identified as author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data Moloney, Jules. Designing kinetics for architectural facades : state change / Jules Moloney. p. cm. Includes bibliographical references and index. 1. Facades--Designs and plans. 2. Motion in architecture. 3. Architectural design. I. Title. II. Title: State change. NA2941.M65 2011 729’.1--dc22
ISBN 0-203-81470-3 Master e-book ISBN
ISBN: 978-0-415-61033-9 (hbk) ISBN: 978-0-415-61034-6 (pbk) ISBN: 978-0-203-81470-3 (ebk)
Illustration credits List of illustrations Foreword Acknowledgements
vii viii xii xv
PART I 1
Movement at the periphery A morphology of pattern for kinetic facades The potential of kinetics
3 3 8
Kinetic precedent Contemporary practice Contemporary discourse Kinetic theory The challenge of kinetics
13 13 24 30 33
Systems, fields and reflexivity A wider perspective Compositional systems Field thinking Cybernetics Some implications for kinetics
39 39 40 46 51 55
Kinetic art The temporal arts Popper: kinetic procedures Lye: figures of motion Rickey: the ship at sea Dorin: taxonomy of process
57 57 61 64 66 70
PART II 5
Decision planes Rewind < 1 Animated variables Towards a framework
Experiments with kinetic pattern Index and intuition Variables Visualizing pattern Stages
91 91 93 99 101
All at sea: a provisional taxonomy Taxonomy as heuristic device Overview of animations From ship to sea A first cut
105 105 106 120 123
State change Non-ascribable Taking stock Towards state change From theory to practice
135 135 137 143 148
Notes Bibliography Index
77 77 79 80
151 168 174
All illustrations are original drawings and images produced solely by the author, or in collaboration with John Bleaney.
1.1 2.1 2.2 2.3 2.4
2.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12 2.13 2.14 2.15 2.16 2.17
Definition of kinetics as three spatial transformations and material deformation Analytical drawings of pneumatic structure prototypes developed by the Hyberbody research group, TU Delft Analytical drawing of adaptive shading for the Ciudad de Justicia, Madrid, 2006–2011, by Hobermann Associates Analytical drawings of screen translations based on a student project undertaken at the California Polytechnic, 2002 Analytical drawings of screen translations based on Kiefer Technic showroom designed by Ernst Giselbrecht and Partner, Bad Gleichenberg, Austria, 2010 Analytical drawings of perimeter wall to Nordic Embassies, Berlin, designed by Berger and Parkkinen, 1999 Analytical drawings of Malvern Hills Science Park, UK, designed by Rubicon Design, 2008 Analytical drawings of student project, by Andreas Chadzis, 2005 Analytical drawings of LIGO Science Education Center, Livingston, Louisiana, designed by Eskew, Dumez and Ripple, 2006 Analytical drawings of kinetic wall sculpture Battleship, by Anthony Howe, 2006 Analytical drawing of project by Ho Sun for a pneumatic ‘quilted’ facade, University of Melbourne, 2007 Analytical drawing of Institut du Monde Arabe, Paris, by Jean Nouvel, 1987 Analytical drawing of Aegis Hyposurface by dECOi, Birmingham, UK, 1999–2001 Analytical drawing of Dynamic Terrain, by Janis Pönisch, Amsterdam, 2006 Analytical drawing of Flare facade prototype by WHITEvoid, Berlin, 2008 Analytical drawing of responsive awning by MIT Kinetic Design Group, Boston, 2000–2002 Analytical drawing of responsive timber surface by Ocean North, London, 2008 Analytical drawing based on proposal for robotic ‘edge monkeys’, by Stephen Gage, London, 2005
7 14 15 16
16 17 17 18 18 18 19 19 20 21 21 22 22 23
2.18 Analytical drawing based on Digital Water Pavilion by MIT Media Lab and Carlo Ratti Associati, Zaragoza, Spain, 2008 24 2.19 Analytical drawing based on Blur Building by Diller and Scofidio, Swiss national expo, Yverdon les-Bains, Switzerland, 2002 24 3.1 Diagram of catastrophe surface that shows control space, event space, fold, and its projection as a cusp (the catastrophe set) 47 3.2 Diagrammatic drawing of an extension to Kaufmann’s three compositional systems: (A) typical ‘ancient’ proportional system; (B) proportional system in tension with Baroque hierarchy of parts; (C) repetition and reverberation of eighteenth-century French rationalism; (D) neutral grid of the modernist curtain wall; (E) graduated surface of contemporary field-field composition 50 4.1 The ship at sea (redrawn from George Rickey, 1963) 67 5.1 Diagram of design variables conceived as a planar continuum between two extremes. Location on the plane identifies the zone of a design instance, but when mapped to time, multiple outcomes are possible from the same combination of variables 82 5.2 Global design variables of temporal structure overlaid on a decision plane 88 5.3 Summative diagram of sampling, control and tectonic decision planes. Specification of variable continuum in combination with periodic structure and temporal scale produces design multiplicity over time 89 6.1 Summative diagram of variables to be used for design experiments 98 6.2 Trials of different geometry undertaken as a pilot study. Hexagonal parts provided the best mix of edge detection and shading depth for motion detection 100 6.3 Summative diagram of 19 control scripts used for stage 2 102 7.1 Stage 1 animation study to determine singular and compound kinetics. The seven kinetic types selected for the experiment are annotated 107 7.2 Stage 2. TRANSLATION × 19 control types 108 7.3 Stage 2. ROTATION × 19 control types 109 7.4 Stage 2. SCALING × 19 control types 110 7.5 Stage 2. TWIST × 19 control types 111 7.6 Stage 2. ROLL × 19 control types 112 7.7 Stage 2. YAW × 19 control types 113 7.8 Stage 2. SPRING × 19 control types 114 7.9 Stage 3-A. Selected animations from intuitive experimentation with one representative kinetic type (rotation) and various amplitude- and period-based arithmetic and geometric progressions 115 7.10 Stage 3-B. Selected animations from intuitive experimentation with one representative kinetic type (rotation) and various amplitude- and period-based arithmetic and geometric progressions 116 7.11 Stage 3-C. Selected animations from intuitive experimentation with
7.14 7.15 7.16 7.17 7.18 7.19 7.20 7.21 7.22 7.23 7.24 7.25 7.26 7.27 7.28 7.29 7.30 7.31 7.32 8.1 8.2 8.3
one representative kinetic type (rotation) and various periodic- and noise-based scripts Stage 3-D. Selected animations from intuitive experimentation with one representative kinetic type (rotation) and various periodic- and noise-based scripts Stage 3-E. Selected animations from intuitive experimentation with one representative kinetic type (rotation) and various noise-, cellular automata- and flocking-based scripts Illustration of range of patterns ascribed within sea pattern taxonomy SWELL pattern generated by a radial geometric progression SWELL pattern generated by a linear geometric progression EDDY pattern generated by a geometric progression WAVE pattern generated by a sine equation WAVE pattern generated by a radial displacement CHOP pattern generated by a prime number sequence SWELL-EDDY pattern generated by a geometric progression SWELL-WAVE pattern generated by a Perlin noise algorithm SWELL-CHOP pattern generated by geometric progressions SWELL-PEAK pattern generated by a life-like cellular automata WAVE-EDDY pattern generated by geometric progressions EDDY-CHOP pattern generated by geometric progressions WAVE-CHOP pattern generated by geometric progressions CHOP-PEAK pattern generated by a lattice noise algorithm NON-ASCRIBED pattern generated by a Perlin noise algorithm NON-ASCRIBED pattern generated by a day-night cellular automata NON-ASCRIBED pattern generated by a life-like cellular automata NON-ASCRIBED pattern generated by a flocking algorithm Cloud formation photographed by the author; Metcalfe, Australia 2009 State change, identified by distinctive shape and dynamic WAVE STATE. The typical simple wave state is characterized by a linear or curvilinear ridge of movement with a uniform and consistent dynamic. The example illustrates the case of a ridge with a regular diagonal dynamic from bottom left to top right FOLD STATE. The typical simple fold state is characterized by adjacent patches of movement, with a constant reconfiguration of boundaries that produces a typical interweaving or expansion/ contraction FIELD STATE. The typical simple field state is characterized by fragmented movement of singular units, or small groups. The dynamic is inconsistent, irregular and multidirectional. The fragments of movement, as captured by the time-lapse image, are highlighted STATE CHANGE. Illustration of kinetic pattern as a dynamic morphology where there are three simple states of wave, fold and field and typical intermediate state transitions – swell/stratify,
119 124 125 125 125 126 126 126 127 127 127 128 128 128 129 129 130 130 130 131 139 141
aggregate/disintegrate, atomize/ribbon. The compound state turbulence occurs when all simple states are present WAVE-FIELD. The intermediate state between a wave and a field typically has a balance of a ridge shapes and pockets of small irregular movement. The example illustrates the dissipation of a wave ridge and atomization along the edges WAVE-FOLD. As was regularly evidenced in the motion studies, the intermediate state between a wave and a fold is typically a swelling of a wave ridge shape. The case illustrated shows the intersection of two wave ridges and the forming of two adjacent patches of movement typical of a fold state FOLD-FIELD. The state between a fold and a field is the most complex of the intermediate states. The example illustrates a movement pattern based on a flocking algorithm, where field fragments are forming into curvilinear patches
Every age has its own architectural tropes: ‘national identity’ in the nineteenth century, ‘functionalism’ in the 1920s, ‘systems’ in the 1960s, etc. There are many tropes around today, of which ‘intelligent building’ is surely one of the most important and persistent. How can architecture become part of an increasingly responsive and changing social environment characterized by digital technology, global nomadism and ubiquitous consumption? Is it possible for buildings themselves to move and adapt in relation to natural or man-made parameters, and if so, how does this improve their performance or relevance? The problem however is that most of the rhetoric about ‘intelligent building’ has been confused and tendentious, at times almost deliberately empty-headed in its technological fixation. What is needed is someone to examine the subject with a cool head and a subtler approach. That is where this fascinating new book by Jules Moloney comes in. What he offers is a highly original and ambitious exploration of kinetic facades beyond the usual enabling technology, in order to focus on the potential for a literal poetics of movement for architecture. As such, the study uses the ‘decision plane’ of the facade – as Moloney states, a crucial manifestation of architectural design since the Renaissance – as its anchor. There are of course many ways in which architecture can be said to be kinetic, whether through material properties, patterns of use, mechanical elements, etc., and hence the book’s specific focus on facades – which can be interpreted either as intelligent environmental screens or as public communication interfaces – provides a clear research agenda. In terms of research methodology, the book is manifestly a hybrid form of research for design; in other words, a study that can help others to think more clearly about the kinetic facades they are designing. Moloney is neither an outand-out theoretician nor an out-and-out designer, preferring to act as a negotiator between the two approaches. In this sense, his book can be analogized to one of the transition states between categories of kinetic movement that he discusses. The wide historical retrospective of theory and practice provides a scholarly overview, from which a highly original and productive tracing of ideas from kinetic art is undertaken. In particular the analysis of 1960s artist and theorist George Rickey offers an extremely convincing case for the uniqueness of designing ‘movement itself’, not least because the nautical metaphors used by Rickey tally with those of early cyberneticians (why there was an obsession with the transport mechanisms of early imperialism isn’t explored in this book, but might make for a fascinating post-colonial analysis). Also of great interest are the links to meteorological analysis, especially
cloud formation; this too can be related back to maritime trade, and again suggests an intriguing research subject. The second part of Moloney’s book explores a number of different ways of conceiving kinetic design for architectural facades. It distils the hitherto complex discussion down to three distinct kinetic forms – wave, fold and field – and this design insight really comes to life when looking specifically at the aforementioned transitions, or ‘state changes’, which develop as interfaces between these three conditions. Anyone who is interested in the theoretical aspects of architectural technology is bound to be stimulated by Moloney’s insights, which reveal him to be one of the most interesting figures in the global field of digital architecture – and indeed possibly the only person as yet to make sense of the thorny area of kinetic design. Professor Murray Fraser Bartlett School of Architecture, UCL
The substantial research for this book was undertaken as part of doctoral studies at the faculty of Architecture, Building and Planning at the University of Melbourne, Australia between 2006 and 2009. I am indebted to the supportive research culture there, in particular the mentorship of professors Bharat Dave and Tom Kvan. For the wider historical perspective that informs this study, I wish to acknowledge the impact and friendship of Dr Ross Jenner at the School of Architecture, University of Auckland. Ross provided the initial spark for the research into Italian Futurism, setting the trajectory towards the poetry of kinetics. I am also indebted to Professor Mark Burry at the Spatial Information Architecture Laboratory at RMIT University, for his continued friendship and support. Professor Murray Fraser at the Bartlett School of Architecture is acknowledged for his insightful comments on the doctoral submission that led to this book, and is warmly thanked for providing the foreword. There is also a generation of brilliant design students at both the University of Auckland and the University of Melbourne, who through design studios based around the temporal aspects of architecture, have allowed some of the ideas to fly. The computer programming that underpins the design experiments was undertaken in collaboration with Richard Penman, who showed much fortitude in coping with the many changes that occurred during the process. Technical insight was also gratefully received from Dr Alan Dorin at Monash University. David Fairservice is thanked for his invaluable proof reading. The illustrations were undertaken in collaboration with John Bleaney, whose elegant eye has contributed significantly to the production. Ultimately, it is the love and humour shared with those closest that make everything possible. Thank you, Helen and Scarlett.
Movement at the periphery A morphology of pattern for kinetic facades Architecture has typically resisted kinetics, yet there is a poetics of movement emerging at the periphery. This book examines the zone between environment and interior, the architectural facade, in search of a latent aesthetic enabled by kinetics. What precedent might inform theory and practice? What actually is being designed, when the outcome is in constant flux? These questions reveal the relative development of kinetic design within architecture. The design of motion is typically outside architecture’s domain, and while there are many engaging technology prototypes, there is minimal ‘content’. By content I mean, for lack of a better phrase, kinetic composition. Composition is used here as a broad and open-ended term, allowing for directed and indeterminate approaches to the design of kinetics. The lack of content and the step outside the traditions of static form provide a challenge for this new field of design research, as there is no coherent body of theory to reference, nor are there sufficient designs to critique. This sparse landscape has led to the trajectory of this study, which undertakes a sectional slice through relevant theory, and uses this to inform the generation of ‘content’ in the abstract. The motivation is that of a designer, examining the lay of the land before traversing it. Through critique and experiment, the aim is to locate the contours of this new design space. While there is an underlying poetic agenda, the inquiry has a succinct focus, which leads to a mode of design research bordering on the scientific. The approach is to examine the design of kinetics from the bottom up, to locate the various parameters that influence kinetic form, and through methodical indexing and intuitive experimentation, generate abstract studies of form in motion. This focus on the underlying structure of kinetics is accepted as being reductive, but a necessary first move that floats above the contingency of technology, site and brief. The ideas here are developed outside the world of matter, to provide an abstract point of reference for those interested in the poetry of movement. The scope of the search for kinetic form is encapsulated in the phrase morphology of pattern for kinetic facades. Morphology is aligned with the manner in which Philip Steadman speaks of essential forms1 and its use by George Rickey in his Morphology of Movement.2 Kinetics is defined as spatial transformation, with a clear distinction being made in relation to several traditions of movement in architecture.
Facade is positioned alongside other terminology such as envelope and skin, before distinguishing this research from kinetic structure and operable interiors. The fourth term, pattern, when considered in relation to kinetics for the particular context of a facade, is more difficult to define at the onset. This emerges, through critique of precedent in the kinetic arts and when developing design variables, which are argued to be the most influential in its formation. Multiple permutations of kinetic pattern, as evident through close examination of animation experiments, provide further insight. The distinctive qualities of kinetic pattern for architectural facades, and the compositional potential these afford, are central themes that shape the inquiry.
Morphology Morphology is typically associated with the field of biology, and refers to the outward appearance and physical structure of an organism, as opposed to physiology, which primarily deals with functional processes.3 The term has been used in architecture in reference to urban morphology, and in design research on the geometry of plans. According to Batty, urban morphology developed around the establishment of the journal Environment and Planning B in 1973.4 From the onset urban morphology has concentrated on the underlying structure of urban form, primarily around the issue of accessibility. As proposed by Batty, the emphasis of contemporary research has shifted from the modelling of static structures to understanding the process by which they come about. The second strand of research within architecture which explicitly uses the term morphology is the analysis of building plans. As introduced by Philip Steadman in his book Architectural Morphology, the emphasis is on exploring the possible range of plan forms within geometric limits. It is primarily concerned with the limits which geometry places on the possible forms and shapes which building and their plans may take. The use of the term ‘morphology’ alludes then to Goethe’s original notion, of a general science of possible forms, covering not just forms in nature, but forms in art, and especially the forms of architecture.5 Morphology for this design research is more aligned with Steadman’s exploration of possible forms than with urban morphology. Just as he explored the limits of planar configuration, this research will undertake a study of kinetic composition. However, while there is similar intent, caution needs to be exercised in making a direct comparison, particularly in relation to methodology. Inspired by the projective drawing techniques of D’Arcy Thompson, Steadman uses a mathematical approach to define rectilinear plan configurations. The possible range of plans is limited by ‘the underlying symmetry lattice or grid by which the pattern is organized’.6 Mathematics is also used to calculate room relationships using graph theory.7 Steadman explores morphology as a form of design science, literally calculating possible combinations of rooms within geometric limits. These techniques are not considered to be directly applicable to the orientation of this inquiry, which embraces a more liquid approach to morphogenesis. The emphasis here is on locating the underlying parameters that determine kinetics, and using these in an open-ended design experiment to produce 4
Movement at the periphery
a series of animation studies. Close scrutiny of the animations identifies consistent types of resonance that occur when manipulating design variables. This location of typical thickenings in the multiplicity of possible form, provides a slice through the new space of possibility afforded by kinetics. While this research does not utilize Steadman’s mathematical analysis and is deliberately open ended rather then definitive in its aspirations, one technical aspect of his approach is adopted. In a careful introduction to his book, he makes the case for the dimensionless representation of plans.8 His argument is that configurations of possible types are best considered in terms of abstract geometric relationships, rather than the dimensions of the spaces. Nor is the physical thickness or materiality of walls considered. For Steadman, morphology is a study of geometric relationships independent of scale or materiality. In a similar manner, the dimensions of kinetic parts or the overall size of a facade are not crucial to the morphology of kinetic pattern. The focus of this study of morphology is on the configuration of geometric transformation in space, the underlying structure of kinetic formation, independent of physical scale or materiality. A second relevant source for morphology is found in the kinetic arts. Towards the end of an essay titled The Morphology of Movement, artist and theorist George Rickey discusses the situation of kinetics as a new genre, in which there was a lack of significant forms, around which practice could reference and develop. Writing in 1963, Rickey articulates the need for an understanding of the range of forms for the new field of kinetic art. He argues that form ‘is without immediate aesthetic or quality implications’,9 and it is in this context that Rickey attempts to outline the essential forms of movement, or the term he uses for his essay, morphology. The question posed by Rickey for kinetic art can be repeated for the emerging practice of kinetic facades. What are the significant forms, independent of value appraisals or production context, and what is the theoretical range around which practice can reference and develop? Steadman shares a similar aspiration for the theoretical range of geometric plan types, independent of production context or value appraisal. ’Morphology’ is the word which Goethe coined to signify a universal science of spatial form and structure. Goethe’s method in botany, where his first morphological interests lay, was intended not just to provide abstract representations, and a classification, of the variety of existing plants, but to extrapolate beyond these and to show how recombination of the basic elements of plant form could create theoretical species unknown to nature.10 It is with a similar intent that morphology underpins this research. Through abstract representations and the recombination of basic elements of kinetics, the aspiration is to locate the theoretical range of kinetic form, extrapolating beyond contemporary practice and historical precedent.
Kinetics A clear distinction needs to be made between kinetics and other approaches to designing for movement and time. Typically, architectural theory and practice have engaged with movement in terms of: • • • • • •
transformation through the event of occupation physical movement of the occupant a sense of movement due to the optical effects of changes in light or the presence of moisture the weathering of materials and effects of decay the representation of movement through form and surfaces that appear dynamic design methods that use geometric transformations or other animation techniques.
Each of these modes is outlined below, with the observation that many are inherent capacities of architecture, or constitute design approaches that have been exploited throughout history. The section concludes with a definition of kinetics, which focuses the scope of this research to areas outside these typical approaches to considering movement in architecture. The first mode – change due to the event of occupation – is relatively self-evident, but has been articulated most clearly in Bernard Tschumi’s thesis that architecture acts as a frame for ‘constructed situations’.11 For example, his 1989 Bibliothèque de France competition entry crossed sports and library programmes to alter the architectural experience. Here the building itself is typically inert, but with architectural ‘movement’ occurring due to indeterminate programmatic encounters. The building is transformed over time by the event of occupation, to create both literal movement in terms of occupancy and activity, but also movement in terms of the perception of the architecture – the stadium empty, for example, versus the stadium heaving with spectators. This capacity is inherent in all architecture but, as articulated by Tschumi, can be deliberately exploited to place architecture in a constant state of occupational flux. The second tradition of movement is also inherent: architecture is experienced by the body in motion and through vision that is constantly shifting focus. The seminal essay, ‘A picturesque stroll around Clara-Clara’, traces a genealogy of the peripatetic view, from the Greek revival theories of Leroy, the multiple perspective of Piranesi, Boulee’s understanding of the effect of movement, to the Villa Savoye where architecture is best appreciated, according to Le Corbusier, ‘on the move’.12 This type of movement is reliant on the mobility of the surveyor in relation to typically inert form. The third mode occurs where perception of static surface, form and space is altered by changing environmental conditions. In this case, buildings can be designed to accentuate visual transformation in response to different light intensity and direction, the presence of moisture, and wind conditions.13 At a completely different order of time is the ageing of materials. Patina is typically resisted by contemporary forms of construction, but as explored in On Weathering, there is a long tradition of exploiting the properties of materials to design 6
Movement at the periphery
in deformation over time.14 A fifth mode of engagement with the theme of movement has its origins in the early twentieth century – the representation of motion through dynamic form. With links to Italian Futurism and German Expressionism, buildings such as Mendelsohn’s Einstein Tower appear as if they are in motion, with streamlined profiles and smooth uninterrupted surfaces.15 A more recent digital engagement in relation to movement and time in architecture has developed around tactics of geometric transformation, as used in design process. As explored by Terzidis, this has origins in constructivism.16 He argues that the computer extends this agenda by facilitating the animation of geometry, such as the sweeping of sectional profiles along paths.17 While his study of kinetics utilizes similar tactics of geometric transformation to those explored here, there is a fundamental difference. For the design of static architecture, geometric transformation is a design method, with the ultimate goal of locating one frozen moment. For kinetic facades, there is no singular moment in time. The design outcome is shifting patterns of geometry in a constant state of flux. While acknowledging the ongoing relevance of the above approaches for architecture, the focus here is on the implications for design when kinetics is defined in spatial terms. As illustrated in Figure 1.1, this includes movement through three geometric transformations in space – translation, rotation, scaling – and movement via material deformation. Translation describes movement of a component in a consistent planar direction; rotation allows movement of an object around any axis; while scaling describes expansion or contraction in size. These are the basic building blocks of kinetics, which are combined to produce more complex motion, such as a directional twist or roll. The fourth aspect of the definition considers the micro-scale, where manipulation of material properties, such as mass or elasticity, allows incremental deformation. This concentration on transformation and deformation in spatial terms allows a distinction between kinetics and what have become known as media facades.18 The incorporation of motion graphics via projection screens, LEDs or visual effects generated by lighting a facade, are not included in the scope of this research.
Figure 1.1 Definition of kinetics as three spatial transformations and material deformation 7
Facade There are a number of terms to describe the zone between architectural exterior and interior: envelope is a generic term that describes the total enclosure of a building; wall was traditionally used to describe a vertical load-bearing construction; the term curtain wall appears in the early twentieth century to distinguish a non-load-bearing construction;19 skin was initially coined to continue the distinction between cladding and the structural ‘bones’ of a wall but has more recently been associated with conceiving the envelope as an intelligent environmental system;20 while facade still retains an association with wall as the site of urban composition. As observed by Neumeyer, the urban facade has, since the Renaissance, been considered in terms of a vertical plane of composition. Since the discovery of central perspective in the Renaissance the surface has been seen as a transparent plane cut by the optical pyramid. From then onwards the urban façade also worked on a projection principle that sees the surface as the spatial breeding-ground for its art of layering. This explains or transfigures the depth links and creates a play of figure and ground that brings living and three dimensional form into life on the surface.21 It is this context, and in relation to the previous discussion of morphology, that facade is utilized for this research. Facade defines a generally vertical plane of abstract composition, as observed externally. As well as articulating the research limits in terms of a more or less vertical plane, facade allows a distinction between kinetics that operate at a larger scale such as kinetic structures, or the kinetic reconfiguration of internal spaces. As will be evident in the discussion of kinetics in architecture undertaken in Chapter 2, there is a large body of work concerned with kinetic structure. For this study a facade as a whole is considered to be static, within which parts are in motion. This scalar differentiation excludes proposals for re-locatable buildings such as those envisioned by the Archigram group, or the genre of revolving architecture.22 A second distinction that facade allows is between kinetics operating on the external perimeter, and the large body of work on reconfigurable spaces and interactive rooms. This excludes, for example, projects such as the Fun Palace proposed by Cedric Price,23 or contemporary research on intelligent rooms.24
The potential of kinetics The careful scope articulated above provides a necessary focus on the emergent field of kinetic design. Typically, the compositional potential of kinetics is obliquely referred to within discussion of interactive or reflexive architecture. For many in the architectural mainstream, kinetics is dismissed as irrelevant or a distracting novelty. According to Ingraham, the tension between stasis and movement is even thought to be the origin of a certain lament within architectural history.25 There is clearly some kinetic resonance within architecture’s oeuvre and this informs some of the propositions developed here. However, it is argued that kinetics opens up a specific field of design research and, as a way to engage with this specificity, the approach taken here is to examine the underlying design parameters. This is, in part, contingent, as 8
Movement at the periphery
there are not enough examples of realized kinetic facades. The reduction of design to abstract diagram is also a deliberate tactic, enabling a focus on morphology independent of scale or materiality. It is anticipated that by undertaking a close study of kinetic diagrams, some progress can be made towards understanding the potential for a kinetic enlivening of the public face of architecture. Within the scope of this study there are two primary areas of activity: intelligent and media facades. As evidenced by intelligent facades, the possibilities are for a responsive membrane that adapts to changing environmental conditions and user occupancy, continuing the trajectory of functionalism.26 Media facades, by contrast, use technology to realize facades as information screens or artworks at an urban scale. Facades are being recast as a zone of interactivity, with the potential to engage users with dynamic information displays or to embed abstract artworks. Regardless of the design intent, the emerging field of kinetic facades offers the challenge of developing a sophisticated approach to the design of movement. Through the lens of morphology, this book explores the possibilities of kinetic composition afforded by facades in motion. As evidenced by contemporary practice, while there is technical innovation there is minimal discussion of the design of ‘movement itself’.27 This term, which has its origins in the 1920 realist manifesto and subsequent development of kinetic art as a specific practice, highlights the opportunity and challenge for architecture. There is the capacity to develop new compositional approaches based on the design of movement, but minimal work on which to build understanding and trigger the exploration of kinetic form. This research engages with this potential through studying the morphology of kinetic pattern. Three interrelated questions shape the trajectory of the inquiry: • • •
How may design variables that influence pattern be conceived? What is the theoretical range of kinetic form? What nomenclature robustly describes the morphology of pattern?
Aspirations Kinetic design emerges from the trajectory of the modernist free facade and the following chapters trace the escape from traditions of part to whole composition, to fields of pattern enabled by the curtain wall. These field tactics are shown to intersect with the digital granularity of contemporary parametric design, with the glimpses of kinetics afforded by contemporary experiments, suggesting the liquid potential of the free facade in motion. This new design space is explored from the position of morphology. A focus on the underlying parameters, which interact to produce shifting intensities of kinetic pattern, emphasizes the shift that is required when designing. The outcome of kinetic design is not a singular form, but a process from which a range of forms manifest over time. This requires designers to consider the design of control system and data input, as well as the design of the physical components. It is anticipated that the conceptual model developed in the following pages can assist designers and theorists to locate the multiple variables that influence kinetic composition. Through methodical indexing and intuitive exploration, an instance of 9
these variables is used here to shape design experiments. These experiments result in abstract patterns, visualized orthogonally through the abstraction of animation software, deliberately presented as low-resolution images that map out a range of design patterns to document an initial set of visual references. The animations clearly do not capture the phenomenology of architectural experience. The register of the mobile surveyor and experiential affect is, for now, put to one side. The experiments are more usefully considered diagrams of possibility, where design variables are mapped to generate animation patterns. Having produced numerous permutations, their analysis allows identification of difference by degree and kind, enabling the development of a nomenclature. It is anticipated that other designers will adapt and refine the approach developed here, as suits their specific agenda. For a design language to be productive, it should be accurate but also evocative and extendable. In part this research is inspired by the work of George Rickey, who, in 1963, articulated a vocabulary of motion for the emerging field of kinetic art. Building on this precedent, the basic forms of kinetics are developed for the particular case of kinetic facades. The aspiration is to provide a reference for other design researchers to adopt, critique, adapt, or extend in relation to personal design agendas and the particular context in which they are operating. Any particular instance of movement takes place in its own time and becomes, for the artist, what a colour or a shape is to the painter. The basic movements are surprisingly few and surprisingly simple. Western music has twelve tones. Kinetic art has scarcely more. Its gamut of movements must, of course, be within the range of human perception, just as the painter is limited to the visible spectrum, and they must be within the artist’s capacity to control. Few though they be, they offer themselves, just as visible colours do, for an almost infinite range of variation, permutation, and combination.28
Tactics The book is organized in two parts. In the first, a critique of theory and practice is undertaken, locating and clarifying the research agenda. From the articulation of the research scope in this opening chapter, the following two chapters are a reflective account of contemporary activity, and sources from architectural history. These provide snapshots of relevant projects and discourse, with the pragmatic objective of locating precedent that may inform this research into kinetic morphology. Chapter 4 extends the search to what is argued to be the most relevant source outside architecture, kinetic art. This is examined through seminal sources from theorists and practitioners who explore the art of movement. The overall aim of Part I is to ground the inquiry in a specific field of knowledge. The shift from current experimentation to the wider context of architecture, and outside to the aligned practice of kinetic art, locates strategic moments. What flickers into focus are glimpses of sublime form and lines of thought that resonate with kinetic facades. From these, the distinction between kinetic art and the specificity of architectural facades enables a working definition of kinetic pattern to be developed. 10
Movement at the periphery
The second part of the book builds on the critique of Part I to inform a more direct mode of design research. In Chapter 5, a model for conceiving the variables that influence kinetic pattern is developed. Using insight from theory and practice, design variables are visualized as three interrelated planes of activity. Each is considered, and through argument and example, the variables that have a direct influence on morphology are articulated. In Chapter 6, an instance of this generic framework is examined at a closer grain, to frame a series of design experiments. Not all variables are considered to be equally influential on morphology, and a select set, or instance of these, is articulated. These variables are used to specify a comprehensive algorithmic code that enables a series of animations to be quickly generated and reviewed. Through subsequent variable ‘tweaking’, the subtleties of kinetic formation are explored in a more intuitive manner. The final two chapters analyze these animations and develop a nomenclature for describing the morphology of kinetic pattern. This is undertaken by the proposition of a provisional taxonomy, a fine-grained classification developed from precedent in kinetic art. This is used speculatively, as a heuristic device to undertake the identification of pattern range. In turn, the robustness of this classical approach to classification is evaluated, and found to be problematic. From discussion of the outcomes and using insight from On the Modification of Clouds, a more liquid set of terms is developed to conclude the book.
Kinetic precedent Contemporary practice The number of publications, blogs and projects indicate an increasing contemporary interest in architectural kinetics, which follows – after a significant gap – a similar flurry of design experimentation undertaken in the 1970s. How has the current generation of designers considered the opportunities of kinetics? Despite the large body of material, the emphasis here on composition and facade enables the undertaking of a directed slice through current activity to address this question. As evidenced by a recent survey, Interactive Architecture by Fox and Kemp, the majority of activity is generally concerned with the functional possibilities and enabling technology, rather than investigation of kinetics per se. Fox and Kemp present a comprehensive overview of current interactive architecture (including kinetic facades), through a simple distinction between ‘ways and means’. That is, the multiple ways in which kinetics are manifest, ‘folding, sliding, expanding, shrinking and transforming’ and the means by which kinetics is realized – the apparatus, ranging from mechanical to chemical technology.1 The questions that drive this study are more to do with affordance and potential. What range of kinetic composition do the kinetic types afford? How have designers exploited this potential? As a framework to examine contemporary activity, an approach developed from a recent study of contemporary facades, The Function of Ornament, will be used. In that publication, case studies were considered in terms of sectional depth, differentiating between structure, screen and surface.2 This approach is extended here by overlaying spatial kinetic as defined in the previous chapter. That is, structure, screen and surface are subdivided according to translation, rotation and scaling, and, where appropriate, material deformation. The scan through contemporary activity will be selective, identifying key examples and examining the compositional aspects evident or afforded by the project. This starts with the largest scale, that of kinetic structure. It then proceeds to examine in turn the intermediate scale of kinetic screens, the fine granularity of surface relief and concludes with other examples at the limits of the research scope.
Operable structure Fox provides a taxonomy of kinetic structure, which distinguishes between embedded, deployable and dynamic configurations.3 The emphasis here on facade excludes embedded structure (structural kinetics at the scale of the complete building, such as earthquake dampening), and deployable structures (where a building can be
physically relocated). Dynamic structures that occur when a large opening is required in a building envelope are within the scope of this discussion. The classic examples are sports stadiums that incorporate an operable roof, with other common occurrences being kinetic wall sections, ranging in scale from aircraft hangers to shop fronts. Güçyeter has undertaken research on types of kinetic structures,4 while in a similar vein Korkmaz explores the possibilities of umbrella-like structures.5 The kinetic of such works is usually a monolithic translation or rotational movement, as sections slide or fold back within the main structure. While they provide a good analysis of the various kinetic operations, neither Güçyeter nor Korkmaz considers the compositional opportunity they afford. This focus on the mechanics reflects typical design research in this field. In most cases, dynamic structures are conceived and undertaken as engineering solutions, and the compositional aspects of the design are less explicit. Even when high-profile designers are involved, there appears to be little interest in the design of the kinetic operation, as opposed to the design of the components. Take, for example, Shigeru Ban, an architect who consistently uses roller panels to create kinetic walls, often incorporated with large free-hanging curtains. Similar to the traditions of Japanese internal screens, the external facade is activated through kinetics at the scale of a bay. This approach has been utilized for single dwellings, apartments, an art museum and commercial buildings. The use of kinetic shutters is primarily to allow a continuity of space that can flexibly adjust to the seasons or specific occasions. There is no indication in the project descriptions or interviews that the design of the kinetic operation is considered.6 A second example of high-profile designers using kinetics is the work of the Hyperbody research group. Director Kas Oosterhuis and his research team are designers with an ambitious agenda that embraces kinetic interactivity. Among the first to engage experimentally with digital technology, Hyperbody are realizing prototypes of a responsive kinetic architecture based on pneumatic structures. These are described as ‘programmable pro-active structures’ reacting in real time ‘based on input values from both the users of the building and from environmental forces acting upon the structure’.7 Their vision of a truly reactive architectural machine has been simulated using interactive virtual environments. The aptly named ‘Muscle’ series of projects
Figure 2.1 Analytical drawings of pneumatic structure prototypes developed by the Hyberbody research group, TU Delft 14
represents a first step towards the actualization of a tensile architecture activated by pneumatic ribs.8 At a larger scale to these working prototypes is the design entry for the 9/11 competition in New York, which proposes data-driven hydraulic cylinders to allow a tower to contract and reconfigure its form by up to 50 per cent. The emphasis of designs from Hyperbody is on flexible membranes – designs that have a strong correlation with the double curvature aesthetic, ubiquitous with users of contemporary NURBS software. On the evidence of the prototypes or their documentation, there does not appear to be an examination of the kinetics from a compositional point of view. The final example to be considered within this section on kinetic structure is Hoberman Associates, arguably the leading international design and construction consultancy in kinetics. Typically, their projects utilize scissor joints, which produce singular, incremental motion. Director Chuck Hoberman has patented three-dimensional scissor joints, which have been realized as consumer products and at the scale of exhibition works. Typically, Hoberman structures expand and contract uniformly upon themselves in captivating slow motion, ‘blooming successively like fireworks’.9 While the overall motion is singular and is executed uniformly across the structure, the three-dimensional complexity is such that the eye is drawn to individual parts of the sophisticated assembly. The work is precision engineered, resulting in a silent, successive folding and unfolding of structural components. A proposed sun-shading project in Madrid, undertaken as a consultant to Norman Foster and Partners, suggests alternative compositional approaches to the singular motion typically associated with these dynamic structures. As visualized in Figure 2.2, each hexagonal sunscreen contracts independently into the overall triangulated grid, and, while the animation shows a regular centre to periphery composition, the multiple orientation of the hexagonal forms adds an additional complexity that is not readily predictable. The regular centre to periphery transitional movement is enhanced by the contrasting orientation of the individual parts. Moreover, the independent motion of each part potentially allows a range of non-uniform compositions beyond centre to periphery.
Figure 2.2 Analytical drawing of adaptive shading for the Ciudad de Justicia, Madrid, 2006–2011, by Hobermann Associates 15
Kinetic screens Moving to a finer grain than structure reveals a myriad of activity in kinetic screens. This review of contemporary activity is organized in terms of kinetic type – translation, rotation and scaling – with projects again being selected on the basis of the potential for kinetic composition. While sash and roller mechanisms have been available for centuries to activate external screens, it has been difficult to locate contemporary examples of translational movement, considered in terms of kinetic composition. One example from academia is a student project that allows horizontal and vertical translation, illustrated in Figure 2.3.10 Rectangular panels are intended to be operated by a wire and pulley system, allowing translation in two axes. This allows consistent horizontal or vertical movement, or a sequential stacking kinetic. Figure 2.4 illustrates an alternate stacking composition as evident in the design for a small showroom, Kiefer Technic. The kinetic is one of vertical translation incorporated with a folding joint that also enables a scaling effect. When activated along the facade, this allows a range of vertical compositional patterns of translation and scaling. The system is computer controlled, allowing multiple permutations of the vertical ‘stacking’ motion. In contrast to translation, there are a large number of projects that use rotation, in particular those that use adjustable louvers to provide dynamic
Figure 2.3 Analytical drawings of screen translations based on a student project undertaken at the California Polytechnic, 2002
Figure 2.4 Analytical drawings of screen translations based on Kiefer Technic showroom designed by Ernst Giselbrecht and Partner, Bad Gleichenberg, Austria, 2010 16
Figure 2.5 Analytical drawings of perimeter wall to Nordic Embassies, Berlin, designed by Berger and Parkkinen, 1999
Figure 2.6 Analytical drawings of Malvern Hills Science Park, UK, designed by Rubicon Design, 2008
sunscreening. However, these are generally conceived as functional panels within the overall facade, with minimal evidence of kinetic composition.11 The kinetics is typically a uniform and regular adjustment of each bay in relation to sun position. Examples of different approaches to rotational screens include: horizontal orientation as in the case of the Nordic Embassies at Berlin, where each panel is individually controlled and able to be rotated through 90 degrees (Figure 2.5); vertically, as in the example of the Malvern Hills Science Park in the UK, where large fins rotate slowly through the day to track the movement of the sun using thermo-hydraulic drives12 (Figure 2.6); and a student competition entry that explores rotational movement in all three axes, which enables two-dimensional patterns, oblique compositions, or can be folded back into horizontal or vertical planes minimizing impact on views out (Figure 2.7).13 Figure 2.8 illustrates the ‘wave wall’, a unique project designed for the Laser Interferometer Gravitational-Wave Observatory (LIGO) in Pasadena, California. Rectilinear aluminum sections are suspended on low-friction bearings at their centre of gravity, with each section having an electromagnet embedded in the ends, so that movement of the singular is transferred to adjacent members. Motion is dependent on wind but can also be instigated or dampened by controlling the strength of the magnets. This project was conceived within the practice of science museums commissioning installations to demonstrate physical behaviour, in this case wave 17
Figure 2.7 Analytical drawings of student project, by Andreas Chadzis, 2005
Figure 2.8 Analytical drawings of LIGO Science Education Center, Livingston, Louisiana, designed by Eskew, Dumez and Ripple, 2006
Figure 2.9 Analytical drawings of kinetic wall sculpture Battleship, by Anthony Howe, 2006
motion. The educational context has enabled the deliberate investigation of patterns related to electromagnetic fields, and as realized provides a glimpse of the aesthetic potential of large-scale architectural screens. Completing this sampling of rotational screens is a unique example of a double rotation, the kinetic wall sculpture Battleship, by Anthony Howe. As illustrated in Figure 2.9, circular discs are able to rotate in the x and y axes simultaneously. 18
Compared to rotation, there are relatively few examples based on a scaling transformation. Most examples of expansion and contraction are found in elastic membranes. Typically, these operate at the scale of pneumatic structures, such as the previous Hyperbody example, or at the scale of a kinetic relief. As an example of a pneumatic facade, a student project undertaken at the University of Melbourne is illustrated in Figure 2.10. Pneumatic ellipses are individually controlled and inflate to create a variable quilted sunscreen, which potentially allows a wide range of kinetic patterns based on expansion and contraction. The Institut du Monde Arabe is perhaps the most famous example of a kinetic facade, and represents a particular scaling kinetic. The south facade is composed as a 24 × 10 grid of square bays. Each bay consists of a central circular shutter set within a grid of smaller shutters, referencing the geometry of traditional Arab screens. In this example, the kinetic definition becomes somewhat ambiguous, as the actual movement is one of rotation of flat sheets over each other, similar to the mechanism of a camera lense. But as the planar rotation is perpendicular to the facade, the kinetic is perceived as a radial scaling kinetic. The expanse of the facade allows for multiple kinetic reading: the kinetic within each bay is one of simple multiple contraction and expansion; while as each bay is individually controlled, the overall composition allows a rich tapestry of kinetic oscillation between bays.
Figure 2.10 Analytical drawing of project by Ho Sun for a pneumatic ‘quilted’ facade, University of Melbourne, 2007
Figure 2.11 Analytical drawing of Institut du Monde Arabe, Paris, by Jean Nouvel, 1987 19
Surface The operable surface has the longest kinetic pedigree in architecture, with perhaps the first instance being that of a tent flap, which through the most minimal of means allows a viewing function, physical access, air movement and the penetration of light. Operable surface has been theorized in Surface Architecture by Leatherbarrow and Mostafavi, which traces the development of what the authors term the ‘temporal operation’ of the building facade.14 Any building with operable windows or doors can be considered in this light, but the emphasis here is on locating contemporary examples that go beyond the commonplace. The design of Auroa Place by Renzo Piano is cited as an explicit example of an operable surface.15 In this commercial high-rise project there is a typical planar curtain wall glazing, but unusually for such a building type, there are automated operable windows. These have horizontal proportions and are operated as vertical groups of three, with opening mechanism, drive gear and rods articulated on the external skin. The kinetic operation is purely functional, but the incremental movement enabled by the finely calibrated mechanisms goes beyond the typical engineering solutions to enable a subtle kinetic interplay along the facade. Another example of an operable surface that goes beyond the pragmatic is the storefront for Art and Architecture, an early work by Stephen Holl. It was conceived when he was particularly interested in proportional systems, and the project has been described in these terms.16 The explicit transformation of external wall using a kinetic composition has not been repeated in subsequent projects, although the use of asymmetrically stepped openings is a Holl signature. This small project has become a New York landmark, with the configuration of the openings being mapped to the temporal scale of daily changes in weather and the longer scale of exhibition turnover.17 A second area of kinetic surface is that operating as relief, a term generally used in relation to sculpture, where a three-dimensional form is contiguous with a surface.18 The most well-known kinetic relief in architecture is the iconic Aegis Hyposurface, designed by dECOi architects led by Mark Goulthorpe.19 The project was initially conceived in relation to a specific site, but has since been developed as a ten by three metre prototype. The original computer visualization presented the relief as a smoothly undulating surface, but the prototype was eventually realized
Figure 2.12 Analytical drawing of Aegis Hyposurface by dECOi, Birmingham, UK, 1999–2001 20
as a triangulated mechanism. Capable of producing abstract or figurative relief, the primary constraint is the dimensions of the metal plates, which determine the level of resolution and degree of curvature possible. As will be examined in the next section, the Aegis was informed by a particular approach to composition that speculated on abstract ‘alphabets’ of movement pattern. Unfortunately, this potential has not being fully realized, but the project has stimulated a succession of architectural projects that explore similar approaches to generating kinetic relief. There are a number of elastic membranes that enable smooth undulation, producing the most contiguous relief surfaces. For example, as illustrated in Figure 2.13, Dynamic Terrain consists of a thick cast-rubber membrane that is pushed and pulled by mechanical pistons to produce undulating form. Designed as a free-standing art piece, the compositional effect is determined by the scale of the actuators. In this case a furniture scale deformation occurs, but in principle the systems can be scaled up or down to produce a range of undulating surface patterns. The majority of the relief prototypes have been developed and fabricated in academic research institutions, but recently a commercial product has become available. As illustrated in Figure 2.14, Flare is a three-dimensional relief based on an efficient geometric design, in which an obliquely faceted ‘flake’ is rotated from one fixed edge. Differing combination of the oblique angles of adjacent flakes produces a remarkable range of effects, given the actual movement is in only one axis.
Figure 2.13 Analytical drawing of Dynamic Terrain, by Janis Pönisch, Amsterdam, 2006
Figure 2.14 Analytical drawing of Flare facade prototype by WHITEVoid, Berlin, 2008 21
Figure 2.15 Analytical drawing of responsive awning by MIT Kinetic Design Group, Boston, 2000–2002
Another elegant example of animated relief is the work of Ned Kahn, who has been working with a system of small metallic discs, hinged in a wire grid to produce a kinetic relief responsive to wind. Reminiscent of 1950s advertising displays, Kahn’s wind walls have been implemented at extremely large scale, and from a distance the wind-activated disks produce fluid multidirectional patterns of movement similar to a water surface.20 An alternative approach to a continuous surface is to use an array of vertical rods or fibres to create ‘hair-like’ relief. For example, Mitchell Joachim has developed the Super Cilia Surface in collaboration with the MIT tangible interface research group; while the Kinetic Design Group,21 also at MIT, have produced a prototype that, as illustrated in Figure 2.15, uses a similar tactic based on sparsely distributed flexible rods. In addition to triangulated, pneumatic and fibrous approaches, there are examples of spatial deformation through controllable variance in material property. There has been much speculation on the possibilities for nanotechnology in architecture, but at present, apart from self-cleaning surfaces, there have been minimal applications.22 One type of material change currently available is that enabled by shape memory alloys in conjunction with tensile skins.23 There are a number of approaches being researched. Benjamin and Yang embed shape memory alloys in a flexible skin to achieve gill-like apertures. In this case, shape memory alloy wire embedded in flexible silicon expands to create the apertures. In another example Pavel Hladik uses a three-dimensional structural frame that expands and contracts
Figure 2.16 Analytical drawing of responsive timber surface by Ocean North, London, 2008 22
to create an undulating surface.24 Shape memory alloy is formed into a space frame when in a ‘hot’ state, and, when cool, collapses to a relatively flat mesh. The design group Ocean North has undertaken an innovative project that exploits the material properties of wood to expand and contract in relation to humidity.25 Key parameters of wood fibre orientation, geometry of the component and fixing position, determine kinetic direction and curvature. As illustrated in Figure 2.16, the subtle material differences in grain and density results in a regular overall movement, but with individual variation between the degree of curling movement, producing a slow motion, variegated kinetic.
Other kinetics This section of activity covers a grey zone, on the edges of the definition of spatial kinetics established in Chapter 1. There are a number of self-powered, independent kinetic assemblies that can operate as part of a building facade. These include robotic systems such as wall-climbing window-cleaning robots 26 and turbines that have been developed for generating electricity via wind.27 An intriguing example of an independent kinetic assembly is the proposal by Stephen Gage for wall-climbing robots, designed to undertake environmental and maintenance tasks. The proposal for wall robots, or what Gage terms ‘edge monkeys’, is based on a critique of centrally controlled environmental systems.28 Rather than develop fully automated systems, Gage argues for the use of simple mechanical window latches and taps that can be activated by mobile robots. As visualized, independently controlled figures roam up and across the building facade. While these are primarily considered in functional terms, Gage proposes they could also have an educative or performative role when not engaged in maintenance tasks. As illustrated in Figure 2.17, the potential (if there are sufficient numbers in the troupe) is for any number of choreographed patterns. A second group of kinetic examples on the edge of the research scope revolve around water as an integral part of a facade composition. The use of water to enliven architecture has a long history. Apart from fountains incorporated into building designs, the development of water walls as part of building facades is reasonably common in twentieth-century architecture, perhaps the most sophisticated being Grimshaw’s design for the 1992 Expo in Seville.29 One project that goes beyond that of a simple water wall is the Zaragoza Digital Water Pavilion, which utilizes
Figure 2.17 Analytical drawing based on proposal for robotic ‘edge monkeys’, by Stephen Gage, London, 2005 23
Figure 2.18 Analytical drawing based on Digital Water Pavilion by MIT Media Lab and Carlo Ratti Associati, Zaragoza, Spain, 2008
Figure 2.19 Analytical drawing based on Blur Building by Diller and Scofidio, Swiss national expo, Yverdon-les-Bains, Switzerland, 2002
computer-controlled water to generate falling text and patterns. The water walls act in a similar way to large-scale interactive media screens. The control system can process images or patterns based on input from motion sensors, or create openings based on proximity of surveyors.30 Equally innovative is the Blur Building by Diller and Scofidio, which used water mist to generate a dynamic volume. The project was commissioned as a temporary structure for a world media expo at Lake Neuchatel in Switzerland and has been analyzed in relation to consistent themes in their work – the breaking down of boundaries between ‘the human and the technological’ and the ‘interweaving of the organic and the inorganic, the ”natural” and the ”artificial”’.31 Finally, no discussion of contemporary kinetics can be undertaken without mentioning the extraordinary projects of Philip Beesley.32 In particular his later works such as his 2004 Reflexive Membranes and Hylozoic Soil, initially exhibited in 2007 and subsequently selected for the Canadian Pavilion at the 2010 Venice Biennale. While not a facade, the lattice-like installation of delicate acrylic armatures and feathery laser-cut Mylar is evidence of the subtle poetry of motion achievable with a fine density of kinetic parts.
Contemporary discourse The slice through activity reveals the range of compositional opportunities afforded by contemporary kinetics. Facades are being designed at various scales, tested via physical prototypes and in some cases realized. Moreover, the projects locate a 24
number of designers and research groups theorizing architecture variously described as responsive, reactive or interactive. This section briefly outlines the literature generated in relation to contemporary practice, in terms of four recurrent themes: indeterminacy, functional expression, intelligence and dynamic structure.
Indeterminacy Within media facades there is a consistent line of thinking that contrasts traditional composition with the indeterminacy of kinetic form. For example, Mark Goulthorpe, the driving force beyond the Aegis Hyposurface, describes the project in relation to the concept of ‘trauma’, in the positive sense of ‘an intense sampling of experience as the mind deploys its full cognitive capacity to account for unfamiliar pattern’.33 From this position, he makes a case for indeterminacy over prescribed composition. In its creation as in its reception it suggests an alloplastic rather than autoplastic logic, the designer’s role becoming that of editor or sampler of a proliferating range of effects. Aegis is perhaps therefore not a ‘form’ at all, since it escapes ‘design’ ideology, conceived much rather as a matrix of the possibility of form: is, in fact, the becoming/absenting of form-in-pattern.34 From this stance, the idea of formal composition in the traditional design sense appears redundant. The range of kinetic pattern for the Aegis was intended to be an indeterminate play between embedded text and its deformation through local interaction, in effect a mutable hieroglyph. The designer retains some editorial control of this interaction, through the qualitative filtering of the input by the computer control system. While a full-scale prototype was implemented, the project now appears stalled, with the legacy being a number of photographs and videos that typically show image or sound being mapped to produce wave-like kinetic pattern. As noted in the previous section, the Aegis has inspired a number of other prototypes. Skins are being stretched, set in motion with intriguing use of selfactuating materials, and developed in terms of biological analogies such as Joachim’s ‘cilia’ skin. There are a number of projects that pursue similar natural analogies in the context of art installations. For example, the archeological trigger for Philip Beesley’s work initially led to large-scale landscape installations. The later Hylozoic Ground is the culmination of a shift to the vertical plane and the creation of internal kinetic environments.35 In a similar vein to dECOi, the kinetic within Beesley’s works is to a degree indeterminate, being reactive to local interaction but modulated to produce consistent patterns. These have been described as ‘empathetic motions’ that ‘ripple out from hives of kinetic valves and pores in peristaltic waves’.36 Payne has undertaken an extended discussion of Beesley’s projects in terms of contemporary art theory. One point relevant to this discussion is his reference to a ‘neo-constructivist conception of art’. This, according to Payne, enables the organization of novel forms, ‘conceived according to game-like criteria’.37 Terms that are often used in relation to works such as Beesley’s Hylozoic Soil or the Aegis, are that of reflexive or responsive. As articulated in the introduction 25
to Responsive Architectures, the term is used in a very general sense to describe an interaction between a kinetic system and environment. Responsive is used throughout this book to speak of how natural and artificial systems can interact and adapt. Speaking of evolution, we might think of how environments act via natural selection on diverse populations. While that traditional definition is included here, we also want to include conscious action.38 Responsive, according to the above description, includes the facility for human interaction to generate a direct response, in addition to ‘natural’ reaction to an environmental force. In contrast to this direct intervention, the reflexive approach of dECOi suggests a more nuanced approach. In the Aegis Hyposurface interaction is translated within the editorial or ‘game rules’ that inflect and modulate outcomes. External forces, either environmental or by human intervention, do not produce a correspondingly direct reaction. Rather, these inputs are processed according to the logic of the control system to produce a reflexive kinetic. In contrast to responsive or reflexive kinetics there are simple environmentally reactive surfaces. Consider, for example, the wind walls of Ned Kahn,39 where nets of hinged discs based on 1950s advertising signage techniques produce stunning visual effects of wind on a reactive surface. No doubt every major city will get a wind wall incorporated into a building skin, but the simple reactive kinetic effect of wind on metal disc allows little more than a vertical version of wind rippling on water. The distinction between such reactive kinetics with responsive and reflexive approaches provides a continuum of types of interaction, all of which produce differing levels of indeterminacy: at one level of indeterminacy is the un-mediated environmental reaction of Ned Kahn’s wind walls; responsive interaction adds another layer of indeterminacy based on direct human intervention; while sophisticated reflexive systems interpret environmental and human interaction, to produce highly modulated indeterminacy. These distinctions locate a fine-grained nuance of indeterminate form, which may prove insightful for understanding the design variables for kinetics. The context of the art installation, or the experimental prototype such as the Aegis, provides a fertile test bed for indeterminate kinetics. Within the commercial sphere we located Flare, a kinetic relief that uses a tiled geometric figure to maximize effects from a simple one-dimensional rotation. The computer animations on the marketing site reveal the kinetics arranged in horizontal bands, with wave-like patterns being propagated in a simple linear manner. Given that each flake is individually controlled, there is the potential for a wide range of kinetic indeterminacy, but at present there is no evidence of this being explored. Flare is typical of most current activity within kinetic media facades. There is a range of plausible kinetic media facades of various configurations, but there is a paucity of content that exploit the potential of the technology. By way of summary, we can observe that the overriding agenda for kinetic media facades appears to be a loosely defined aspiration for indeterminacy – what dECOi term an ‘alloplastic logic’. This is a deliberate attempt to avoid 26
design ideology, through the tactic of a locally responsive, event-driven interaction. Haeusler, in his proposition for voxel facades, uses dECOi’s concept of alloplastic indeterminacy and folds in discourse on virtual worlds. His ambition is to combine image and form, ‘mutually creating a new architectural zone that is in constant flux’.40 The content in his case is anticipated to be both online and local data, such as that obtained by social networking sites or local mobile phone users. The recurring interest in indeterminacy potentially challenges this inquiry into kinetic morphology. Arguably, however, the projects still require the parameters of this interaction to be designed. There may be indeterminacy at the level of data sampling, but the resultant kinetic is moderated through control systems and tectonics that typically produce a consistency (note the repeated reference to wave and ripple motion, for example). The open remit of this discussion is on identifying the parameters that are available to the designer, and locating the full range of kinetic forms that result. It might be that the tactic of indeterminacy merely suggests an alternate way of conceiving composition. The projects have been carefully and skillfully constructed by designers, fully in control of the parameters by which users interact, and the mechanisms by which data is translated into kinetics. This compositional tactic parallels artists such as Brian Eno, who have shifted their activity from the discrete object, to specifying the parameters for multiple iterations:41 families, or species of form, rather than singular works. None the less, the subtle variables that control how interaction is translated to kinetics still have to be designed. More often than not, the framework for such indeterminacy produces distinctive outcomes that can be analyzed in terms of morphology.
Functional expression While media facades may be considered to provide a social function, there is a separate and longstanding interest in kinetics to enhance the environmental performance of architectural facades. The area of most activity for environmental control is active sunshading systems, which builds on the legacy of the modernist brise-soleil. The majority of the literature is concerned with the pros and cons of active sunscreens in terms of functional effectiveness.42 It is, however, acknowledged that these systems have an aesthetic agenda, as evidenced by the title of one of the more thorough reviews, George Baird’s The Architectural Expression of Environmental Control Systems.43 In terms of material form, most environmental control systems are based on mechanically operated louvers or fins of varying materials, profiles and proportions. There are two general design approaches: either to embed the kinetic component or fin within the composition of the facade, thus minimizing aesthetic impact; or to articulate the shading or ventilation elements as a separate prosthetic device which serves the building. While Baird notes successful expression and integration of environmental control systems within the architectural aesthetic, discussion of the actual kinetics is seldom referred to. Perhaps the most telling comment is a reference to users finding the abrupt on/off motion of an external roller blind distracting. It would appear the design of the kinetics is generally not considered as part of the agenda of functional expressionism. From a historical perspective, the seminal project for kinetic facades is 27
the Institut du Monde Arabe. Conceived as a modern interpretation of traditional Arabic screens, the intent of the kinetics was to control light, and an intricate mechanism was developed to achieve this. Given this hybrid role as semiotic reference and light manipulator, this project could have been included in the documentation of discourse around media facades. Regardless of how it is interpreted, the Monde Arabe is an important project that has received much acclaim, but minimal discussion as to the potential of its kinetics. The designer, Jean Nouvel, appears to have no interest in the kinetic effects of the 25,000 shutters, and defends the operational failure of the system by saying the movement is so slow that most people thought it was not working.44 Nouvel has not designed kinetics into any subsequent work, instead producing a series of static skins that explore transparent and etched surface. Neither his comments, nor subsequent critique of the project by others, discuss in any depth the potential of the kinetics in terms of range of pattern or strategies of kinetic composition.
Intelligence In addition to discussion of the indeterminate media surface and the previous section on functional expressionism, there is a track of discourse on intelligent facades. A recent overview defines the basic criteria by which a building can be considered intelligent. These are: 1 2 3 4 5
input system a processing system which analyzes this input an output system which reacts to the analysis of the input this response occurs with a consideration of time learning ability.45
However, a survey of intelligent facades resulted in the declaration that a truly responsive, adaptive and controllable intelligent facade has yet to be found.46 Most systems, while they may be able to consider a range of factors simultaneously and interpolate a response, fail on the criterion that they can learn from previous data. In terms of the focus of this research, there is minimal discussion on the compositional potential of kinetics within contemporary discourse on intelligent facades. The kinetic is typically considered in terms of a sun-tracking motion, or an on/off approach, where the motion defaults to the operational speed of the mechanical system. The development of intelligent facades is driven by an environmental performance agenda, which arguably continues the functionally determinative trajectory of architecture as a built form of design science.47 According to recent critique, the continuing dominance of this agenda results in such facades being socially inert, despite the potential inherent in such kinetic systems for engagement.48 Perhaps the best example of the potential of an intelligent screening system is the Nordic Embassies complex in Berlin. Over 200 metres in length, what could have been a rather heavy-handed urban design gesture to give disparate buildings a common identity, is enlivened by the kinetics. Each louver is independently adjustable, varying over time according to local preferences, resulting in unpredictable motion over the course of day and seasonal cycles. 28
Dynamic structure At a scale beyond media or environmental screens and surfaces, is discourse on dynamic structures. Within the scope of this inquiry, there are three significant researchers: Chuck Hoberman, Michael Fox and Kas Oosterhuis. While Hoberman is a leading practitioner, who started from a background as a sculptor, he appears comparatively reticent to engage in discourse. His aesthetic is dominated by a distinctive approach to engineering, which produces a singular, minimal motion as the structural component folds in on itself. In a rare record of his design approach, his comments confirm this minimalist approach: ‘the beauty you see in the product isn’t something that I’ve imposed on it. Rather it emerges from the function’.49 Hobermann’s work focuses on producing ingenious engineering solutions based on elegant scissor joints. The outcome is an eerily silent motion of structural components collapsing upon themselves, an engineering feat that typically produces a singular incremental motion. Perhaps of more interest are the animations of the proposed large-scale retractable sunshade Hoberman has developed for the courtyard roof of the Campus of Justice in Madrid. Rather then a singular motion, the shading system is broken into independently controlled components, which, as visualized in the animation, produce an elegant part to whole kinetic as the total shading area unfolds. Michael Fox founded the MIT Kinetic Design Group, and, as well as his recent book with Miles Kemp on Interactive Architecture, he has published a taxonomy of kinetic structure. This was useful in helping to determine the scope of the project survey, locating embedded structures as the most appropriate for this research. The other major contribution Fox makes in terms of theory is a taxonomy of control systems for kinetics. This identifies six general types, ordered by level of complexity: 1 ‘internal controls’, such as a mechanical hinge, which do not have any direct control or mechanism 2 ‘direct control’, where movement is actuated directly by an energy source external to the apparatus 3 ‘indirect control’, based on a sensor feedback system 4 ‘responsive indirect control’; optimization of multiple feedback sensors 5 ‘ubiquitous responsive indirect control’; a network of controls use predictive algorithms 6 ‘heuristic, responsive indirect control’; algorithmically mediated networks that have a learning capacity.50 Fox goes on to make a distinction by which he divides the control systems into two groups. The first three systems are actuated via hand, motor or sensor switch directly, while in systems 4–6 the kinetics are mediated by use of a computer. While these are thorough classifications, they are all variations on the standard input-control-output (ICO) paradigm. This taxonomy of controls, while useful in understanding approaches to controlling kinetics, is not explored in terms of the range of kinetic patterns that will manifest. This is in line with Fox’s concentration on the ‘ways and means’ of kinetics. His taxonomy of control systems outlines the various 29
ways kinetic structure may be controlled, with the majority of his work exploring the means by which these can be technically implemented. The kinetic consequences, the potential poetry of movement that results from these approaches to control and their technical implementation, receives minimal attention, although it is implicit to varying degrees in the examples he uses to illustrate his survey of activity. Despite an introductory statement in Interactive Architecture alluding to kinetic technology allowing a revisiting of ‘traditional kinetic aesthetics’, this agenda is left implicit in the diagrams and photographs of various technical solutions.51 The third major figure in relation to dynamic kinetic structures is Kas Oosterhuis, who leads the Hyperbody research group at The Technical University of Delft, as well as being a director of architectural and design consultancy ONL. Oosterhuis was an early adopter of the computer as a design tool and has published widely in relation to digital production methods. Of particular significance for this research is his version of interactive architecture. Interactive Architecture (iA) … is defined as the art of building relationships between built components in the first place, and building relations between people and built components in the second place. iA is the art of building bi-directional relationships. In the approach of my Hyperbody Research group at the TU Delft all building components are in essence seen as input-processing-output [IPO] devices.52 Oosterhuis goes on to outline an approach to creating interactivity using Hyperbody’s iA/Protospace software, which is based on flocking algorithms.53 Projects are generally conceived as a singular double-curved surface, with the kinetics based on deformation of the surface by pneumatic structure. In general, the projects focus on the mechanics of realization, with minimal exploration of the possible range of kinetics enabled by the technology. In terms of a theoretical contribution, the use of an agent-based approach to the control mechanism used by Oosterhuis has synergies with the complex ‘heuristic responsive indirect’ type identified in Fox’s taxonomy. But, again, the concept of a control mechanism based on self-organizing behaviour, such as that achieved with flocking algorithms, while potentially useful in indicating one approach to control systems, is not explored in terms of conceiving the opportunity for kinetic composition.
Kinetic theory The reviews of contemporary practice and discourse have revealed sophisticated taxonomy of control systems and a range of approaches to interactivity, but provides minimal insight for this study of kinetic morphology. This section extends the discussion, through an examination of key texts that address kinetics in a wider historical context. While there are useful backgrounds to movement in architecture within books on Santiago Calatrava54 and an excellent study of transportable architecture by Robert Kronenburg,55 two sources are particularly relevant. The most direct precedent is Kinetic Architecture, by Zuk and Clarke, which summarizes architectural research in kinetics during the 1960s.56 The second is Flying Dutchmen: Motion in 30
Architecture, which places Dutch architecture of the 1990s within a potentially useful philosophical context.57
The legacy of the 1960s The first book to comprehensively articulate definitions, design philosophy, and architectural applications of kinetics is Zuk and Clarke’s Kinetic Architecture, published in 1970. The authors have an academic background and the book collates research and teaching agendas undertaken during the 1960s at the Universities of Virginia and North Carolina. The emphasis of the book is on adaptable spaces via kinetic structural systems, with no documentation of kinetic screens, relief or materials. The majority of the examples are structural systems that operate at a scale outside the scope of this research. Zuk and Clarke’s thorough classification does, however, reveal one possibility not considered in contemporary discourse. A wider definition of kinetics might be considered around the category of ‘disposable architecture’. Potentially, this may be incorporated into the agenda of kinetic facades if, for example, materials that rapidly decompose are zoned and deformation controlled in some manner. This hovers on the margins of the inquiry within the tradition of weathering, and calls into question the issue of temporal scale. Kinetics as defined for the purposes of this study is physical movement in space via the three geometric transformations and their composites, or deformation due to controlling material properties. Zuk and Clarke’s idea of material change at a longer temporal scale suggests speed of transformation needs to be considered. As discussed in Chapter 1, the concentration on facades locates a generally vertical orientation, observable from a fixed point of view. Observable, by extension, infers that the timescale is such that the detection of movement is within the limits of human vision. These limits will obviously vary dependent on individual and context, but the lower threshold is between two and three seconds.58 That is, for a kinetic to be based on material deformation as suggested by Zuk and Clarke, recognizable change needs to occur within three seconds. While the focus of Zuk and Clarke is firmly on kinetic structures at a scale generally outside the research scope, they do include a taxonomy of machines that can be seen as prefiguring the taxonomy of controls by Fox. The four categories are based on the degree of adaptation: level 1 machines perform a singular function and include lever devices such as water clocks, or rotating machines such as water mills; while level 2 machines can perform multiple functions, for example a bicycle or a steamboat; level 3 machines are distinguished by automatically adaptive control systems; level 4 machines link the adaptive control system to a computer – this, as conceived in 1970, offered the optimistic promise of ‘machines which will construct whole buildings completely and automatically, to machines that will automatically repair and perhaps reproduce themselves’.59 This machine taxonomy would appear to provide precedent for contemporary control systems, such as the taxonomy of Fox discussed in the previous section. After methodically documenting their categories of architectural kinetics, in the final section on future implications for design, Zuk and Clarke introduce the issue of kinetic aesthetics. This would appear to be the first direct discussion 31
of kinetics in architecture that goes beyond function or technology. Their emphasis is on the need to develop a temporal understanding of aesthetics. This idea occurs throughout their examples, as designers conceive kinetics in terms of multiple overlapping rates of change, from daily to yearly temporal scales. Since time is the basic measure of motion, it becomes an important factor in design. This suggests that kinetic architecture must be considered as a continuum. The movement unfolds, but what the form has just been or what it will be, are a matter for recollection or conjecture. This architecture can never be confronted whole. A definition of form which is time-dependent must be recognized …. The sense of motion, itself, then, can be a visual aesthetic much as has traditionally being the case with basic elements like colour, texture and pattern.60 Zuk and Clarke identify the potential for a new aesthetic based on kinetics, and suggest ways in which this might be studied, such as time lapse or stroboscope. However, there is no further development of this idea, either through subsequent publications or design studies. The book ends with the evocative idea of an architectural aesthetic based on the ‘sense of motion itself’, suggesting that the kinetic arts may provide further insight. Their intuition that the kinetic arts provide a valuable source for understanding aesthetics will be taken up in some detail in Chapter 4.
Philosophical insight The second source that provides some insight is Kari Jormakka’s Flying Dutchmen: Motion in Architecture. Jormakka is a professor in architectural theory, and his critique of motion in architecture is undertaken by positioning architectural projects within a historical framework.61 It is in a section titled ‘living Architecture’ that Jormakka considers actual movement, as opposed to the representation of dynamic lines of force. His references include: Aldo Rossi’s floating theatre, the Teatro del Mondo; Ron Herron of Archigram and his walking city; Bruno Taut’s expressionist vision of a city on wheels; and John Ruskin’s ‘living architecture’.62 This thread is then followed through to the 1935 Casa Girasole, ‘the classic version of the rotating house’, and Herman Hertzberger’s water villas, which achieve a similar effect by floating on water.63 The tendency to actual movement in contemporary architecture is discussed in relation to Kas Oosterhuis and the Hyperbody research group. Jormakka is critical of the effectiveness of Oosterhuis’s approach, describing his Trans-Port project as ‘a strong form that does not appear to change very much: it can never assume identities other then what it already has as a large sculpture, albeit a kinetic one’.64 From actual movement, Jormakka then explores other ways in which the theme of movement can be utilized in architecture. A chapter on ‘Framing Movement’ considers architects UN Studio’s, ‘creative use of diagrams,’65 while a chapter on ‘Architectural Promenades’ discusses the Japanese concept of architectural experience as ‘topological chains of discontinuous space’, and compares this with Le Corbusier’s Villa Savoye and other contemporary examples, in which movement through space ‘unfolds’ in time.66 Both these examples consider the movement 32
of the surveyor rather than the focus here on kinetics. Perhaps of more potential for this inquiry is a philosophical discussion, which Jormakka argues underpins the theme of movement in architecture. This provides some references that offer insight for this study of kinetic morphology. In particular, duration as proposed by philosopher Henri Bergson, suggests that a study of kinetics might be informed by the ideas of continuity and difference. Bergsonian duration is introduced by Jormakka through the example of a melody. It is argued that melody resides in the memory of past notes and in anticipation of a complete phrase rather then the real time hearing of discrete notes. The concept of duration, as will be developed at the end of this chapter, also provides a basis for considering the temporal form of kinetic facades.
The challenge of kinetics The greatest challenge of all is not scientific (creating increasingly mature mathematical models), nor technological (creating the physical and electronic systems to enable levels of interactivity and sensibility in building and settings). And neither is it even functional. No the true challenge is, as always, of an aesthetic nature.67 As wryly noted by Antonino Sagio in a discussion of interactive architecture, the aesthetic challenge of imbuing technology and functional logic with poetry ultimately distinguishes competent buildings from inspirational architecture. The particular challenge addressed here, is that kinetics requires the consideration of (to use WIlliam Zuk’s phrase) ‘a sense of motion, itself’. To advance beyond obvious compositional approaches (such as the proliferation of wave forms), some basic research needs to be undertaken. This includes an understanding of the variables that determine kinetics, iterative design studies to explore the possible range of kinetic forms, and a shared set of terms to enable considered critique. The aim of this book is to provide this basic framework. As a first step towards this goal, this chapter has scanned contemporary practice and discourse and examined the wider historical context. What has been revealed? Contemporary designers tend not to engage in discourse on aesthetics when discussing projects, and this is the case for most working with kinetic facades. The review of practice and discourse was undertaken to examine precedent for an articulated design theory, or reflection on practice by designers in terms of kinetic composition. Some useful information has been uncovered, in particular, research on control systems for kinetics in architecture provides precedent for some of the variables that determine kinetics. There is, however, minimal discussion that engages with the aspiration of this book. There is little evidence of designers stepping back from their design prototypes, and considering the possible range of compositional approaches afforded by kinetics. Why is this so? Environmentally responsive building facades have been possible for many years, such as Buckminster Fuller’s automated sunscreens for the US pavilion at Expo ’67, or Jean Nouvel’s Institut du Monde Arabe.68 Considering these two high-profile examples, it would seem they were ahead of their time, with the available technology proving to be inadequate. They are singular moments in architectural history, which have become 33
as well known for the technical problems as for the innovative compositional potential they signaled. The contemporary emphasis within architectural design on environmental performance has led to a renewal of interest in the capacity for facades to be kinetically responsive. A range of environmental sunscreens have been constructed, and new systems are continually being developed. At present the kinetics is basic, but there is no doubt that the current trajectory of research into responsive building facades will continue, with the result that more sophisticated technical solutions will be available. Within these environmental systems there is a growing recognition that the aesthetics of the technology needs to be considered, alongside the technical performance. The drivers are twofold: first, there is a formal agenda to express the technology architecturally; second, the feedback from occupants has been negative towards unconsidered kinetics, variously described as distracting or disturbing. The technology is constantly improving, with more robust mechanics and refined controls to coordinate the kinetics of individual components. This provides an opportunity for designers to integrate kinetics in a manner that enhances performance, user experience and provides new avenues for the architectural expression of technology. The potential for kinetic environmental control systems is one clear area of activity, but, understandably, composition is not at the forefront of this research. The gap in design knowledge is more immediately apparent outside environmental control, where designers are contemplating the use of kinetic relief as an engaging, interactive form of cladding. There are commercial systems on the market, informed by the high profile of dECOi’s Aegis Hyposurface. The overview of activity also revealed alternative technology, which explores kinetic membranes and other approaches to producing animated relief. That research is being driven by an interest in the recasting of architectural surface as a zone of interactivity, with the potential to engage users with public artworks or to embed socio-cultural information. At present, the focus is on the technical and functional aspects of the kinetic systems, but implicit in the discourse is an interest in composition. Within the contemporary mainstream, an alternative approach to the articulation of static facades has been developing for some time. The Function of Ornament articulates a revived interest in urban facades beyond the legacy of functional expressionism and post-modern composition: Architecture needs new mechanisms that allow it to become connected to culture. It achieves this by continually capturing the forces that shape society as material to work with. Architecture’s materiality is therefore a composite one, made up of visible as well as invisible forces. Progress in architecture occurs through new concepts by which it becomes connected with this material, and it manifests itself in new aesthetic compositions and affects.69 From this position on architecture’s materiality, kinetics provides new opportunities, a sophisticated approach to the incorporation of kinetics, be they driven by 34
environmental or socio-cultural agendas. The use of the term ‘composition’ is problematic for some designers, who still appear to be reacting to the excess of post-modernist formalism. Take, for example, the Aegis, where the declared interest in kinetics is an aspiration for an indeterminate zone of interaction, in avoidance of ‘composition’. But as the designers reveal, the control systems that process the interactions provide a ‘qualitative filtering’, a facility to inflect outcomes, from which the designers conceived ‘alphabets of patterning’.70 Is this not just a higher level of composition, where the activity is directed towards manipulating the variables from which form will evolve? To be fair, the declared target of dECOi was Gombrich’s The Sense of Order, which argues for representative certitude in art. The filtering and control for the Aegis, by contrast, aspired to shape a precise order of indeterminacy – to suspend the composition in an undulating zone of mutable surface where text, image and sampling of local interaction momentary coalesce as fleeting pattern on pattern. Unfortunately, as the Aegis appears to have been mothballed, it is not possible to gauge how far their qualitative filtering progressed. The photographs and stills reveal a predilection for waves and eddies of various configurations, which resonates with many of the other examples examined. For the purposes of this study, the aspiration is that a morphology of kinetic pattern will enable such precise control over levels of indeterminacy, or if it is the design intent, the converse – a predictably ordered composition. That is, kinetic composition as used here, considers design as a continuum between explicit predetermined form and degrees of indeterminacy – reactive, responsive and reflexive. Beyond contemporary activity, the historical widening of the scan through kinetics revealed overviews of technology and a philosophical perspective. In terms of the former, Zuk and Clarke provide a taxonomy of machine controls, culminating in a heuristic system linked to a cybernetic device. How might this taxonomy be utilized in relation to a morphology of kinetic pattern? A framework for considering the design variables that influence kinetic pattern, if developed from these precedents, would emphasize composition as a direct outcome of technology. The classifications may allow a greater understanding of technical approaches, but as presented, there is no connection made with the kinetic composition that results. The emphasis on technology as the primary factor in determining kinetic pattern does not address the wider context, evidenced by contemporary examples – in particular those systems that are not directly related to an environmental agenda, such as the range of projects that explore kinetic relief. In contrast to Zuk and Clarke, the discussion of contemporary Dutch architecture by Jormakka considers kinetics primarily in relation to theory. Of some potential relevance is his discussion of the philosophical legacy of Henri Bergson and Gilles Deleuze. Two concepts – diagram and duration – appear to be used productively by some of the architects cited in his discussion. The use of the ‘diagram’ as a creative representational device could potentially be useful for kinetic facades. Rather than using new technology to repeat existing conventions, kinetic composition conceived through diagrammatic relationships may be a tactic to avoid what Jormakka describes as the problem of ‘topological fixation’.71 What diagrams might enable an open approach to composition, in avoidance of kinetics being merely an 35
outcome of the enabling technology? The intuition here is that the answers to this question reside in an articulation of the full range of variables that determine kinetics. Diagram as used by Jormakka, can be seen to equate to the focus on morphology that underpins this research. As will be evident in Part 2, a diagrammatic or morphological approach enables a systematic exploration of the compositional potential of kinetic facades. The second concept introduced by Jormakka is Bergsonian duration. A recent discussion of Cedric Price’s Fun Palace (an often-quoted precedent for interactive architecture) provides an introduction to this concept in an architectural context. Mathews draws a comparison between Bergson’s conception of matter as ‘time and duration, continuous flux and infinite succession without distinct states’, and Price’s conception of architecture: ‘events rather than of object, events not as static snapshots but as a continuous evolution of phenomena unfolding in time’.72 Kinetic facades provide a similar context in which ‘phenomena’ unfold in time. Can Bergson’s concept of duration be useful in theorizing an approach to kinetic facade composition? Duration has been developed by Gilles Deleuze, through his highlighting of the distinction between continuous and qualitative multiplicity. It would be a serious mistake to think that duration was simply the indivisible, although for convenience, Bergson often expresses himself in this way. In reality, duration divides up and does so constantly: That is why it is a multiplicity. But it does not divide up without changes in kind, it changes in kind in the process of dividing up.73 The distinction is between two forms of duration: continuous multiplicity is a ‘quantitative differentiation, of difference in degree’; while a second form involves ‘qualitative discrimination, or of difference in kind’.74 The concept of qualitative and continuous multiplicity, provides one approach to conceiving a morphology of kinetic form. A facade which operates as a continuous multiplicity would be capable of potentially infinite variation, but this would be ‘already visible in the image of the object’.75 Consider one of Ned Kahn’s surfaces, rippling under the influence of wind. While there are infinite combinations of wind eddies, once a number are comprehended, then all possibilities are ‘already visible’. The surveyor may continue to observe infinite variation by degree and, while the register of subtle variations may be engaging, once the initial discovery has passed, this is a relatively neutral state. There is no qualitative change – the variations, while infinite, do not achieve any state change and hence do not heighten perceptual awareness. In contrast, a kinetic facade which operates as a qualitative multiplicity would, within the Bergsonian framework, undergo changes in kind as well as in degree. Returning to the example of Kahn’s wind veil, what would constitute a change in kind? For example, would differential resistance to rotation suffice? Consider if the small discs that make up the surface were ferrous, rather than aluminum. An electromagnetic field controlling the facade could enable a dampening of the wind patterns. If the control was in turn linked to another non-linear data source, an unpredictable mix of wind force and dampening effect would result. Potentially, there would be the capacity for two 36
types of pattern formation, and permutations between states. That is, the capacity for difference in degree and kind. The concept of duration articulates that, as phenomena unfold in time, there is the capacity for changes in degree, that in turn trigger changes in kind. This may provide a useful insight for understanding the influence of design variables on kinetic pattern, and conceiving a robust morphology that can cope with the potentially infinite variation possible. The context for understanding the potential of kinetics has a wider background than that being considered by many designers. Contemporary activity is evidence of a growing field of design, but one that still requires some fundamental work to be undertaken. Designers are focusing on technical implementation, and are producing innovative and inspirational prototypes to meet the requirements for robust, environmentally efficient systems. The ideas being explored here are no less challenging – that is, the potential for a subtle aesthetics of movement.
Systems, fields and reflexivity A wider perspective The scan of contemporary theory and practice undertaken in the previous chapter revealed innovative prototypes and associated critique, taxonomy of control systems, and philosophical perspectives on morphology. The motivation was to review how theory and practice has engaged with the liquid potential of kinetics. While the examples provide evocative glimpses, there is minimal evidence of designers realizing the full potential of kinetics to extend the agenda of architectural composition. Nor are there theoretical models that locate design variables, or a nomenclature that captures the poetry of motion. Putting direct precedent for kinetics to one side, can paradigms from the wider perspective of static facades be usefully adapted? In order to deal with the wealth of potential sources, insight from the discussion of qualitative and continuous multiplicity informs the review. The distinction between difference by degree and difference by kind that concluded the previous chapter, focuses the attention on theorists and historians who take a wider view than stylistic iteration. In this respect, the work of historian Emil Kaufmann locates generic approaches to design composition prior to the modern movement. He has developed an approach to historical analysis, through what he describes as three ‘systems’ of composition. The mechanism he uses to locate difference is the reciprocity between component parts and the composition as a whole. From Kaufmann, the contribution of David Leatherbarrow provides a contemporary view of the legacy of modernism. Leatherbarrow has examined material weathering, but the most relevant text is his Surface Architecture and subsequent discourse on performative architecture. He traces the development of the freestanding facade and the neutralizing impact of mass-produced and standardized cladding. The idea of the performative surface is developed, as a way to re-engage with what he refers to as the lost project of representation. In order to place the idea of the performative surface in context, key historians of the modern movement are examined. In particular, Sigfried Giedion and Reyner Banham provide a link to the origins of the shift in compositional emphasis, that continues to this day. Giedion and Banham both allocate a significant amount of their accounts of the early modern movement to the agenda of Italian Futurism. Subsequent analysis of this influential spark for twentieth-century art and architecture has been
undertaken by Sanford Kwinter. His book Architectures of Time examines in detail the futurist concept of plastic dynamism, relating this pivotal moment in the twentiethcentury avant-garde, to nineteenth-century developments in the physical sciences. Kwinter in turn provides the link back to contemporary practice, in particular his activity within the ANY group of theorists and designers. This thread between futurist experimentation and contemporary ‘field theory’ provides a second relevant vein of discourse for kinetic facades. A third area of discussion follows the identification in the previous chapter, that control systems are central to any discussion of kinetics. The taxonomy of Michael Fox and the categories of machine controls in the earlier work of Zuk and Clarke, belong to a wider sphere of activity, that of systems theory and cybernetics. This is introduced via the critique of new media theorist Katherine N. Hayles, followed by three contemporary figures who consider cybernetics from an architectural viewpoint: Ranulph Glanville is an architectural design theorist who has written extensively on cybernetics; John Frazer, a pioneer in generative computational approaches to design, provides an essay on the architectural relevance; while Stephen Gage critiques the analogy of the boat/helmsman to locate reflexive indeterminacy as a key to designing interactive architecture.
Compositional systems We speak of an architectural system as long as one ideal of configuration is valid. Beyond this basic and, perhaps, somewhat nebulous idea there is nothing permanent, there is only change. The relationship of the whole and the parts, and of the parts to each other, are dictated by the ruling idea of the system, but the variations are infinite. The artist searches for ever new solutions. This is the essence of artistic development.1 In this view of history, Emil Kaufmann privileges what he terms systems over historical demarcation through stylistic features. When considering architecture up to the twentieth century, he proposes there have been three systems, each differentiated by the reciprocity between component parts and the composition as a whole. The first is the ‘ancient system’ of the Graeco-Roman period as epitomized by the theories of Vitruvius. Kaufmann proposes that while there are proportional relationships between parts and in relation to the whole, there is no differentiation or hierarchy between parts in ancient Graeco-Roman architecture.2 The second distinctive system is located from the fourteenth-century Italian Renaissance through to the architecture of eighteenth-century France. Kaufmann refers to this entire period as the ‘Baroque system’. His proposition is that the Baroque maintained the ideals of harmony and proportion from the Graeco-Roman system, but introduced a new compositional principle. The parts now should be presented not only in aesthetically satisfying relationships of size and in mathematical reciprocity, but they should be differentiated as superior and inferior components.3 40
Systems, fields and reflexivity
The phrase ‘mathematical reciprocity’ is a reference to the key theoretical text underpinning the Renaissance, Leon Alberti’s Ten Books of Architecture. The principle of part-to-whole harmony is captured by the term concinnitas, which appears multiple times in Alberti’s treatise. According to Wittkower, this relates to the aspiration for a uniform system of proportion throughout all parts of the building, based on the ‘Pythagorean system of musical harmony’.4 Fellow historian George Hersey clarifies the principles by which concinnitas can be achieved. According to Alberti’s doctrine concinnitas is the primary principle of nature. It is based on number in the Pythagorean sense. When concinnitas is present in a building the observer is challenged to work out the values involved in its siting, planes, or elevations (collocatio), and in the volumes (finitio). When the totals and their components equal particular values (certum numerum), the building’s concinnitas or beauty has been revealed.5 While there seems to be consensus between Wittkower and Hersey that composition is reliant on the application of Pythagorean number, there are few direct references within Alberti as to the optimal proportions. Wittkower cites a discussion of musical harmony, located prior to one reference for types of plan sizes, as evidence that, ‘for Alberti, harmonic ratios inherent in nature are revealed in music’.6 Alberti’s recommendations for plan types are the square (1:1), one to one and a half (2:3) and one to one and a third (3:4), which, as Wittkower points out, relate to simple musical consonance. He concludes that the key to understanding Renaissance proportion, as interpreted through Alberti’s writing, is the principle of number generation using compound ratios. For example, the series 4:6:9 is generated by sequentially applying the ratio of 1:1.5. The use of number series to create proportional relationships was developed alongside the overriding principle of bilateral mirror symmetry.7 The outcome being harmonic proportional relationships in the respective parts, replicated across the central viewing axis to produce an ideal configuration. For Kaufmann, what stylistically is referred to as the high Baroque, is the extreme manifestation of hierarchy between parts. The attempt to produce balanced symmetrical harmony based on proportional relationships is increasingly at odds with the Baroque articulation of a competing hierarchy of features. The goal of concinnitas becomes increasingly harder to maintain when the second requirement for differentiation between parts is accentuated. Sixteenth century architects began to realize there was no solution: that it was impossible to bring the individual parts into a perfect union and at the same time to endow some of them with power. One can strive for the reconciliation of graduation and concatenation, but one can never reach it. Gradation in particular is the natural foe of integration.8 In relation to facade design, this dilemma resulted in the following characteristics: 41
exaggerated scale of vertical and horizontal components; the exaggeration of volutes and introduction of spiral towers in an attempt to resolve the vertical and horizontal; exaggeration of details; the treating of stone as soft and flexible matter (as a consequence of the attempt to express binding forces between parts); and the anthropomorphic or animistic transformation of components. In Italy, the tension between unification and hierarchy led to extreme manifestations, such as the fluid facades of Borromini and Guarini. Perhaps the most inventive outcome in relation to the Borromini facades was the introduction of concave and convex planes, which bind vertical and horizontal surface behind highly articulated columns and cornice. An original solution to the Baroque dilemma of the competing demands for unification and articulation, the curved planes and other swirling details produce sensual facades that literally ripple with tension. Kaufmann, after some discussion of the drawings of Boullée and the architecture of Ledoux, lists the formal characteristics of what he considers to be the third system of composition. The ancient system was based on harmonic, non-hierarchical relations between parts, while the Baroque introduced to this the competing idea of hierarchical differentiation. In the third system parts are not related, but are what Kaufmann terms independent or individual. He proposes two general categories by which independence is achieved – repetition and antithesis. While there was repetition in the ancient and Baroque systems, the difference Kaufmann identifies is that in the third system, parts are repeated as individual units. These are not proportionally related using harmonic number series, nor are they part of a hierarchical system of competing features. Identical components are duplicated side by side, arrayed without any alteration in size, or the same component is scaled to produce what he terms ‘reverberation’. These tactics of simple repetition and scaling provide relatively neutral facade compositions, prefiguring to some degree the arrival of twentieth-century modernism.
Modernist surface Once the skin of the building became independent of its structure, it could just as well hang like a curtain or clothing.9 Surface Architecture traces what the authors term the ‘project of representation’ in Western architecture. Leatherbarrow and Mostafav identify the development of the free facade, with what they consider to be a tragic break with traditions of composition and the semiotic role of facades. As the public face of architecture, the facade was traditionally the site of visually engaging composition, or took on an associational character. In the context of this ‘project of representation’, architectural surface was considered a form of public art or, as in the tradition of ‘architectural parlante’, elicited emotional response. The theme developed in Surface Architecture is that the technical development of the free facade enabled a fundamental shift in facade composition. Rather than designing openings in terms of part to whole, the composition occurs in relation to a ‘repetitive unbroken covering or uniform or continuous wrapping’.10 The ultimate outcome of the free facade being the curtain wall, 42
Systems, fields and reflexivity
where the identity of the part is consumed in the uniform texture of industrialized surface. Compared to Kaufmann’s systems, there is no Baroque-like accentuation, and harmonic relationships between parts or part-to-part antithesis, are at a completely different order of detail. While it is possible to trace the legacy of traditional compositional devices in the curtain wall, the dominating aesthetic is the neutral rectilinear grid.11 Surface Architecture identifies the tension between the traditional project of representation and the shift to industrialized modes of architectural production. The free facade developed quickly into the curtain wall, with the relatively neutralizing impact of mass-produced and standardized cladding. This is seen to produce a dilemma for architecture, which, according to Leatherbarrow, is still to be resolved. How then is architectural cladding to be understood: as aesthetic impulse or technological imperative? Is it, perhaps, both?12 As a way of rethinking this question, the authors introduce the idea that the surface can be re-conceived as an active zone of adjustment. The task of disclosure in architecture is not limited to that of representation in the traditional sense of the word. An alternative strategy could involve seeing the building’s external cladding as elements that structure both the building’s skin and its temporal operations.13 By temporal operations, the window is transformed from a purely visual function to one in which, ‘it is also an instrument of adjustment’, not only an ‘eye’ but a ‘hand’. In these terms, they cite Gerrit Rietveld’s Schroeder House and the placement of operable windows in the corners, as a tactic to develop alternate relationships between interior and exterior. The architectural surface, shed of its load-bearing function, is free to allow experimentation with new technologies of adjustment. Leatherbarrow continues the theme of surface as an active interstitial zone in a subsequent essay on performative architecture. In this he proposes that contemporary practice and criticism concentrate on either the functional aspects or the aesthetic experience, both of which treat architecture as object, independent of the temporal context in which they operate.14 The theme of ‘adjustment’ is developed as a way of avoiding ‘the old debate between works that are useful or beautiful’.15 Citing Renzo Piano’s Aurora Place tower, he suggests that architecture has an autonomy of performance, which transcends modernist ideals of prescribed functionalism, or architecture as pure aesthetic object. Approximate movements can be intended, but settings can also yield, respond or react to unforeseen events. The architectural drama, then, comes alive through the building’s performances. The first step in the development of a performative architecture is to outline strategies of adjustment.16
This description of performance in temporal terms would seem relevant to this study of kinetics. Unfortunately, Leatherbarrow does not expand on his call for ‘strategies of adjustment’. Instead, he continues to discuss the idea of performance in relation to static elements of architecture and on the manner in which traces of structural and environmental dynamics are manifest. While functional performance may be scripted, the suggestion is that once a facade becomes kinetically responsive to oscillating functional demands, a level of indeterminacy is introduced. The implication is that this enables a re-engagement with the traditional project of representation, albeit this resides in the play of functional kinetics. However, we are left wondering as to the range of ‘strategies of adjustment’ necessary for this re-casting of the facade as performative surface. In an evocative end to the essay, architectural performance is seen in a wider perspective, unfolding within an indeterminate ‘milieu’. Performance in architecture unfolds within a milieu that is not of the building’s making. A name for this milieu is topography, indicating neither the built nor the un-built worlds, but both.17 Developed from a historical critique of architectural surface, Leatherbarrow locates the shift to a temporal form of composition. At one extreme would be the weathering of materials, as explored in his earlier writing on architectural surface. The operable surface presents a more immediate temporality. Yet for Leatherbarrow, this still appears to be located within the functional agenda of modernist composition: performative movement, rather than kinetic design per se. However, implicit in the idea of ‘strategies of adjustment’ is the potential for kinetics to be considered compositionally, rather than as a mere byproduct of functional imperative. Despite the alignment of kinetics with functionalism, the manner and frequency of adjustment requires explicit design for the potential of the unfolding ‘milieu’ to be fully realized.
Space, time, machine Leatherbarrow’s reference to early modernism and the conception of the wall as a new zone of spatiality, suggest examination of some of the key texts of modernism may be productive. To this end, two early histories of modern architecture will be considered: Sigfried Giedion’s Space, Time and Architecture and Reyner Banham’s Theory and Design in the First Machine Age. The architectural revitalization of the wall, like Ronchamps can be dangerous. It has already thrown a beguiling cloak over the playboy fashions of the 1960s. One can observe everywhere a tendency to degrade the wall with new decorative elements. This is not the basic purpose. Architecture is fundamentally concerned with the revitalization of the wall from within.18 Given the passage of time and the context of contemporary ‘playboy fashions’, the tone of this condemnation of Le Corbusier’s design for Ronchamps appears rather quaint. For Giedion, the wall is not composed, but is a result of a modernist 44
Systems, fields and reflexivity
conception of space. In a similar strategy to Kaufmann’s three part-to-whole systems, architecture is considered in terms of a triptych of ‘spatial conceptions’. In the first, space is concerned in terms of external relationships between buildings as distinct figures. He then describes a second conception of space in which ‘space became synonymous with hollowed out buildings’.19 The third stage is the integration of these external and internal space conceptions, which he proposes distinguishes architecture from the turn of the twentieth century onwards. From this synthesis of external and internal space, a functionally derivative view of facade design is developed. Part to whole composition became subordinate to architecture of interpenetrating space and the play of light on smooth surface. In this context the sculptural wall composition and asymmetrical play of Ronchamp is only barely acceptable, if attributed to internal spatial and lighting requirements. Further still, any patterning or wall relief was rejected as mere ornament, unless it was integral with the surface construction.20 As revealed in Chapter 2, this modernist justification of composition in functional terms is still prescient within much contemporary kinetic facade design. Perhaps of more relevance to our inquiry is the section titled ‘The Research Into Movement: Futurism’. Here, Giedion describes aspects of Italian Futurism and their relationship to developments in physics, which, according to him, instigated the re-conception of architectural space and time.21 From this brief reference to developments in science, Giedion summarizes the research by futurist painters and sculptors into the representation of motion. In painting, the technique adopted was to capture movement as superimposed frames, while in sculpture Umberto Boccioni developed his practice around the theory of plastic dynamism.22 In relating this research to architecture, Giedion considers this to have minimal relevance. The problems facing architecture in the early twentieth century are, according to Giedion, best answered through the architectural projects of Antonio Sant’Elia. He discerns a new viewpoint present in the evocative drawings, ‘an architecture imbued with elasticity and lightness’, behind which Giedion locates an artistic aim of ‘mobility and change’.23 However, this clearly did not include kinetics (presumably this would be cast in the same light as ‘playboy fashions’). Rather, according to Giedion, the interest in movement was explored in pursuit of a ‘more basic purpose’: designing for mobility in the form of separate circulation systems for vehicles and pedestrians; and, second, by considering buildings as non-permanent, opening up an agenda of lightweight materials and construction techniques. The legacy of futurist research into movement is also closely examined in Banham’s Theory and Design in the First Machine Age. After setting the scene through introductory chapters on theory between 1900 and 1914, Banham devotes a group of chapters to present the futurist manifestos and projects. The key figure for Banham is Umberto Boccioni, who articulates a ‘field theory of aesthetic space, a space which exists as a field of force or influence radiating from the geometrical centre of the objects’. 24 When discussing Sant’Elia’s contribution, Banham describes his projects in terms of compositional ‘knots’ and ‘networks’, equating this dynamism to Boccioni’s field theory.25 This connection is subsequently traced within early modern architecture, such as aspects of the De Stijl movement where ‘the
technology of tensions, invisible motions, action at-a-distance’ are interpreted as ‘Futurism reworded’.26 This vein of discourse on the impact of Italian Futurism within modernism suggests potential for kinetics. For Banham in particular, futurist-inspired movement is interpreted as invisible networks, setting architecture into compositional knots and implied motion. The legacy of the futurist field perpetuates in contemporary discourse and is omnipresent in the architecture of many practitioners. How might this idea of the field align with kinetics?
Field thinking The field describes a space of propagation, of effects. It contains no matter or material points, rather functions, vectors and speeds.27 A key source for a contemporary analysis of Italian Futurism in architecture is Sanford Kwinter, who examines in some detail the work of key figures Boccioni and Sant’Elia. His analysis will be introduced, and then related to the parametric design tactics of some contemporary practitioners. Kwinter locates Italian Futurism alongside the development of electromagnetic field theory. While no direct connection is established, he argues for an affinity between field theory as it was developed in the physical sciences, and Umberto Boccioni’s writing on plastic dynamism, contained in his Technical Manifesto of Modern Sculpture.28 According to Kwinter, Boccioni develops a theory that considers form as ‘arrangements’ of forces within ‘plastic zones’.29 These zones of force are not considered to be fixed but are dynamic, in a qualitative sense. These two principles, that the boundaries of form are a result of force fields and that these fields also determine material qualities, are captured in the term ‘plastic dynamism’. According to Kwinter, this results in a third principle – that form, space and time are coexistent and interdependent. Having articulated the hypothesis of plastic dynamism, Kwinter then undertakes an analysis of the projects of Sant’Elia. Like Banham he focuses on the Musei Civici drawings, and argues that this conception of urban architecture can be interpreted in relation to Boccioni’s theories. For Kwinter the designs are evidence of a dynamic system, in which the architecture ceases to be considered an object, but rather is considered a zone subject to internal and external ‘fields of pressures’. While there is minimal kinetics, apart from the animation of the building surface with circulation devices (lifts, stairs, ramps and bridges), what is evoked is an architecture of ‘kinematic plasticity’.30 Kwinter extends his analysis of Futurism in a subsequent essay on Boccioni’s painting Stati d’Ianimo. This essay starts by introducing mathematician Henri Poincaré’s work on the three-body problem, which laid the foundations for chaos theory.31 Poincaré’s realization was that three interrelating objects within a gravitational force field produced behaviour beyond calculation using Newtonian physics. Kwinter locates this as a seminal moment for rethinking theories of form. The shift was from a spatial conception of form to a temporal one where ‘transformation and novelty were the irreducible qualities that any theory of form would need to confront’. The influence of this discovery is, according to Kwinter, explicit 46
Systems, fields and reflexivity
in the aesthetic experimentation of the futurist movement and the terms under which this was undertaken – ‘waves, fields, fronts’.32 Kwinter embeds a number of photographs that evoke similar kinetic process: patterns in sand caused by the ocean; the self-aggregation of larvae around a food source; the surface geometry of water at the moment a sphere is dropped. The images represent moments in time of a transformational process, or what are referred to as singularities or ‘points in a continuous process’.33 As a means to conceptualize the conditions that give rise to such singularities, Kwinter cites visualization of topological flow on a two-dimensional plane. He argues that the axial location of parameters and their interaction over time allow an intuitive understanding of the emergence of form, which in turn can be extrapolated to more complex systems. Now clearly, a plane is a very simple, even rudimentary space. A flow in the plane can essentially be described by two parameters, or two degrees of variability or ‘freedom’. Most systems in the real world, that is, most forms or morphogenetic fields, are clearly more complex than this. Yet it is enough to understand how forms emerge and evolve in simple ‘2-space’ to gain an appreciation of how more complex forms evolve in more complex spaces.34 Could this insight be usefully adapted for conceiving the kinetic form of architectural facades? The visualization of complex process as parameter interaction has resonance with kinetics, perhaps the most explicit design activity requiring conception in terms of fields of force. The visualization of parameters for dynamic form as planar interaction in relation to a vertical temporal axis, suggests a potential adaptation for conceiving parameters for kinetic facades. The complexity of kinetics is likely to involve more than the interaction of two parameters, but the planar model provides an interesting basis from which to start. As revealed in Part II of this book, the visualization of complex interaction between parameters over time provides a potent idea for conceiving the variables which determine kinetic pattern. Figure 3.1 Diagram of catastrophe surface that shows control space, event space, fold, and its projection as a cusp (the catastrophe set). Figure redrawn from the original by Joseph Macdonald in Kwinter, S. ‘Landscapes of Change: Boccioni’s “Stati d’animo” as a General Theory of Models’, Assemblage, 19, 1992, p. 60.
Control space 47
Animate form/novel tectonics The concept of a force field that underpins plastic dynamism and an emphasis on temporal dynamics as developed by Kwinter, can be traced within the writing and projects of contemporary designer and fellow ANY group member Greg Lynn.35 In Animate Form, Lynn articulates an approach to architectural design that has strong resonance with Kwinter’s analysis of Italian Futurism. Animate design is defined by the co-presence of motion and force at the moment of formal conception. Force is an initial condition, the cause of both motion and the particular inflections of a form.36 Digital technology is pivotal to Lynn’s practice. He utilizes the capacity of design software not to draw, but rather sets up flexible models that can be animated over time. The dynamic model is controlled by adjusting parameters, which propagate change through networks of associated geometry. By mapping geometry to attractor points, changes can be inflected through the linked geometry as if they were reacting to field forces. In Animate Form, precedent for this approach to design is examined and includes D’Arcy Thompson, Étienne-Jules Marey and Hans Jenny.37 These are followed by documentation of Lynn’s design experimentation with animated parametric techniques.38 The designs explore various tactics for generating form through the dynamic interaction of geometric parameters, within the computer simulation of gradient force fields. As we shall discuss later, there are close synergies with such time-based parametric design and the conception of the design variables that determine the temporal form of kinetic facades. Lynn was a pioneer in exploiting the capacity of software from the motion graphics industry for architecture. While animation had been used for some time for presentation purposes, he inverted the approach, to set form in motion. Rather than animate a camera through a design, kinematic chains of geometry were animated to allow fine-grained manipulation of the composition. Published in 1999, Animate Form has stimulated a generation of designers. Prominent within this digital wave are Reiser and Umemoto who have recently documented their own take on parametric design, the Atlas of Novel Tectonics. According to Kwinter’s preface to the book, there is a direct correlation between their granular parameterization of architecture, and the field thinking that stimulated Italian Futurism. What follows is the first design manual that reflects the foundational shift that first took place seventy years ago in physics, when life was first understood to represent a pattern in time that could no longer be amenable to explanation in purely physical and chemical terms.39 While the designers distance themselves from the literally animate, their ‘manual’ may provide some insight into the design of kinetics. The book is organized as a series of practice-orientated strategies for design, grouped into sections with cryptic titles – ‘geometry’, ‘matter’, ‘operating’ and ‘common areas to avoid’. Unlike Lynn, there is an avoidance of generalized position statements or models for design. There 48
Systems, fields and reflexivity
are, however, recurring themes that may be instructive for conceiving kinetics. One concerns their rejection of composition in terms of part-to-whole relationships and its replacement by considering form in terms of multiple global fields, or ‘wholewhole relationships’. … the emergence of new organizations and new architectural effects out of wholes that are not reducible to their parts. These emergent organizations become legible not as parts to a whole but as whole-whole relationships.40 The rejection of part-to-whole composition in favour of continuity is explored further via the design tactic of repetition. Modularity is embraced, but, rather than repetition of identical units, they advocate minimal transformation at the level of a part. These are argued to accumulate to produce a qualitative difference, ‘transformation is a quality perceived through deployment in quantity’.41
Five systems? The idea of composition as ‘whole-whole relationships’ can be considered in relation to Kaufmann’s three systems of part-to-whole composition. In addition to the modernist neutrality of the curtain wall grid, whole–whole graduation might be considered one of two additional ‘systems’. It is arguable that the modernist free facade allowed architects of the twentieth century to obscure part-to-whole composition through the mechanism of the neutral grid. The grid suggests a fourth system that, to a degree, transcends Kaufmann’s differentiation between ancient, Baroque and the independent part–part compositional tactics of nineteeth-century rationalism. The twenty-first-century embracement of digital technology has facilitated a ‘whole– whole’ fineness, which, as illustrated in Figure 3.2, might constitute a fifth approach to sit alongside the modernist grid. Reiser and Umemoto’s ideas of whole–whole exploit the potential implicit in the free facade and the capacity of design software for fine-grained modulation. They locate a compositional principle of ‘fineness’, the use of repetitive and interconnected parts, each minimally transformed. This produces graduations or intensities within a continuous whole, in contrast to the periodic uniformity of the modernist curtain wall.42 Kwinter’s theoretical work has evolved with Reiser and Umemoto, Greg Lynn and a generation of architectural theorists and designers who, enabled by digital technology, have embraced time-based parametric design. Together with the legacy of part-to-whole systems, these precedents provide a wider perspective in which to consider kinetic design. These invite further development in terms of possible compositional approaches for the design of kinetic facades and are taken up in some detail when planning the design experiments undertaken in Part II of this book. Before then, pointers towards the legacy of cybernetics implicit in Leatherbarrow’s call for a strategy of adjustment, provides an equally relevant source for kinetics.
Figure 3.2 Diagrammatic drawing of an extension to Kaufmann’s three compositional systems: (A) typical ‘ancient’ proportional system; (B) proportional system in tension with Baroque hierarchy of parts; (C) repetition and reverberation of eighteenth-century French rationalism; (D) neutral grid of the modernist curtain wall; (E) graduated surface of contemporary fieldfield composition
Systems, fields and reflexivity
Cybernetics The building’s elements do nothing more then sketch out the guidelines of a performance, allowing for spontaneous qualifications.43 How might Leatherbarrow’s strategy for ‘spontaneous qualifications’ be carried out? Arguably, the aspiration for performative composition requires an understanding of control systems and the legacy of ‘cybernetic devices’, located in the discussion of 1970s kinetics undertaken in Chapter 2. The reference to cybernetics is clearly pertinent to kinetics, which typically requires control systems to set facades in motion. The research field of cybernetics is explicitly concerned with theoretical models for control systems, but unfortunately has a plethora of rather opaque terminology. As a shorthand introduction, Katherine N. Hayles provides a useful summary, while the research of Ranulphe Glanville amplifies the distinction between what are referred to as first- and second-order cybernetics. As a way of locating how cybernetics has influenced design practice, the work of Stephen Gage and John Frazer is briefly discussed. What results is a rich variety of approaches to kinetic controls and, by implication, design composition.
Homeostasis, reflexivity and emergence Cybernetics has its origins in self-regulating systems, where a device intervenes in a process to provide corrective feedback. The ubiquitous mechanical example is a float valve that maintains water at a constant level. Invented in Alexandria, circa 270 BC,44 this simple device illustrates the central principle of feedback: a low water level opens the valve, the rising water then provides feedback into the system, closing the valve when the required level is reached. This then reoccurs in a circular fashion as the water level fluctuates – circularity being another important principle of cybernetics. The definition of cybernetics as a modern field of research is credited to Norbert Weiner, as developed in his influential book Cybernetics, or Control and Communication in the Animal and the Machine.45 Explicit in the title is the idea that all self-regulating systems – biological, cultural and mechanical – can be understood through the same principles. Cybernetics was conceived as an abstract study of process and has been compared to mathematics, in that it enables understanding in a variety of fields.46 Weiner was a key participant in the Macy Conferences, which brought together researchers and resulted in further refinement and development of cybernetic theory. Katherine N. Hayles has critically examined the Macy conferences between 1946 and 1952. She traces the debate that surfaced during these discussions, which much later developed into what has become known as second-order cybernetics. She then contrasts these two approaches to cybernetics with contemporary research on self-organizing systems, which explore the principal of emergent behaviour. Hayles considers emergence to be the third stage in the development of cybernetic theory. The central idea of first-order cybernetics is homeostasis, ‘defined as the ability of an organism to maintain itself in a steady state’.47 Second-order cybernetics 51
is distinguished by the additional principle of reflexivity, where the rules that maintain the steady state are recursive, enabling more complex behaviour. A reflexive system is based on the idea that the feedback device is not neutral, but has properties or ‘motivations’ that condition the feedback. Hayles traces what she terms a third wave, which extends the cybernetic concept of reflexivity to contemporary research on selforganizing systems that generate novel outcomes. There is a shift away from the legacy of homeostasis, which Hayles maintains is still implicit in a second-order cybernetics, to focus ‘not on how systems maintain themselves intact, but rather on how they evolve in unpredictable and often highly complex ways through emergent processes’.48
Cybernetics and architecture In the context of 1960s England, where architecture had been declared a design science, the idea of built environments as self-regulating systems gained some traction. The centre of design-led cybernetics was the Architectural Association, London, with the key figure being Gordon Pask. A pioneer in cybernetics, Pask was involved in the design of the Cedric Price’s Fun Palace project, and has had a strong influence on a generation of UK designers.49 Ranulph Glanville completed his doctorate under Pask, and is a champion of second-order cybernetics and its role in the creative act of design. He provides a detailed argument for second-order cybernetics, which arguably also provides valuable insight for conceiving kinetic controls. The key principle, as articulated by Glanville, is the distinction between the circularity of cybernetic systems compared to the linearity of what he terms traditional science,50 the distinction being that the trajectory of traditional science is towards repeatability and consistency, whereas the feedback principle of cybernetics propagates a constant state of variability. From the general feature of circularity, Glanville presents Gordon Pask’s conversation theory as the essential paradigm of second-order cybernetics. In conversation theory, information is reconstructed by constant interaction between entities. Compared to first-order cybernetics (where there is an identification of a control and output), in second-order cybernetics, which element is recognized as the controller, and which is controlled, is essentially arbitrary. For kinetic facades these principles of circularity and the interchangeability of control and output potentially provide alternate models for conceiving composition. There is a resonance with the aspiration for indeterminacy located in Chapter 2, suggesting second-order cybernetics may provide insight into how this may be achieved. Another aspect of cybernetics which Glanville highlights, and one which is clearly important for conceiving the variables that may determine kinetic behaviour, is Ashby’s Law of Requisite Variety. This states that the most effective cybernetic control requires ‘the controlling system has at least as much variety as the system to be controlled’.51 Relating this to controls for kinetic facades, we might conceptualize this in terms of the requisite grain, or density of control to moving part. Ashby’s law would suggest the granularity of both moving parts and controls will influence kinetic outcomes (a line of thinking that will be developed as part of the definition of design variables undertaken in Chapter 5). While Glanville usefully introduces the general features of cybernetics, Stephen Gage provides a more directed discussion on interactive architecture. His 52
Systems, fields and reflexivity
design for robotic maintenance devices, what he describes as ‘edge monkeys’, is an intriguing example of kinetics that pushes the boundary of what might be considered a facade. In his writing, Gage revisits the central metaphor of cybernetics – the helmsman and the boat. He argues the typical use of the image of a helmsman controlling a craft, observing deviation and adjusting the wheel in a feedback mechanism, deserves closer attention. His proposition is that the helmsman as a metaphor underplays the significance of the boat and the environment; that the trajectory and behaviour of the boat is as much a function of the dynamic environment as the helmsman’s attempt to control direction. His writing is also informed by his observations tutoring interactive architecture projects. The theme of environmental forces is revisited, with an emphasis on the shift required by designers to realize the essential time-based quality of interactivity design. He describes what he terms ‘picturesque’ machines that create wonder and delight through kinetics. A wax cylinder greenhouse ventilator opener is a good example. When the weather is warm the wax in the cylinder expands, driving a piston that opens the ventilator. Devices that are not much more complex can operate shading devices. Can simple devices like this be seen with wonder and delight?52 The capacity for intriguing movement, embedded in environmental control systems, suggests empathy with the thrust of this research. Unfortunately, Gage provides no more detail on the qualities of such ‘picturesque machines’. A suggestion of a minimalist aesthetic can be read in his observation that complex behaviours can be generated with minimal control systems, ‘where the signal from one object affects the behaviour and signal from another’.53 It is to cybernetics that he turns, albeit with a critical frame of mind, when considering the complexity of designing such interactivity. The two cybernetic theorists he considers are, not surprisingly, Pask and Glanville. However, he detects significant difference between the two in regard to the transparency of the interaction. He is distrustful of the ‘black box’ approach of Pask, while in comparison Glanville is considered the only cybernetic theorist to embrace ambiguity. According to Gage, ambiguity is an essential and desirable quality of interactive architecture. The third key figure in relation to architecture and cybernetics is John Frazer. A pioneering figure in the use of computers for design, Frazer undertook a series of research projects at the Architecture Association, which culminated in the exhibition Evolutionary Architecture. Pask participated in the projects and wrote the introduction to the book documenting the exhibition, in which he stresses the cybernetic feature of the research.54 Frazer describes the design research more in terms of a biological analogy. The evolutionary model requires an architectural concept to be described in a form of ‘genetic code’. This code is mutated and developed by computer program into a series of models in response to a simulated environment. The models are then evaluated in that environment and the 53
code of successful models used to reiterate the cycle until a particular stage of development is selected for prototyping in the real world.55 The legacy of cybernetics is most pronounced in the concept of feedback mechanisms, or what Frazer terms ‘iterative adaptation’.56 Frazer succinctly defines the difference between his research and traditional approaches to architectural composition. Simply stated, what we are evolving are the rules for generating form, rather then the forms themselves. We are describing processes, not components; ours is the packet-of-seeds as opposed to the bag-of-bricks approach.57 Frazer’s ‘iterative adaptation’ corresponds to Hayles’ identification of a ‘third order’ of cybernetics, based around the idea of emergent form. As a way to bring this discussion of cybernetics to a close, we can sum up the different orders of cybernetics, and briefly outline the implications for kinetic control systems. The distinction between homeostasis of first-order and the reflexivity enabled by second-order cybernetics, suggests ways to conceive alternate control systems for kinetic facades. According to Hayles, both are still orientated towards maintaining a steady state, i.e. they are goal orientated. From the perspective of design, homeostasis and reflexivity provide a model by which kinetics might be directed towards recurring steady state patterns. There would be capacity for variation over time, but this would be by degree, rather than producing qualitatively different outcomes. The differing approaches to control can be related to the compositional intent of the designer. First-order approaches would be the most top-down in terms of compositional intent. The designer would have a clear design in mind, and, while the kinetics would be reactive, the feedback loop would, as in the metaphor of the boat, be constantly adjusting the kinetic to maintain a prescribed outcome. Second-order cybernetics accepts that the rules governing the feedback loop can be altered. Moreover, interaction between input and output is constant, and information can flow in both directions. In this case the design goal is more indeterminate and open to variance in response to changes in the environment. However, the legacy of homeostasis is present in this reflexivity – the route may change, but there is still a coherent map to locate the destination. The development of the concept of emergence, from the legacy of cybernetics in the research of John Frazer, provides an alternative model of interaction. Rather than preconceiving kinetic pattern, the outcomes would emerge from interaction between variables. In this third order, the design journey starts without a boat or a helmsman. There are fitness criteria for their design, but the shape evolves from interaction with the environment. While this can be orientated towards general goals, the bottom-up structure of emergence enables differences in kind, as well as by degree. Continuing the nautical analogy, the vessel might morph to submarine or flying boat. In summary, we can posit that the cybernetic models of control ranging from homeostasis, reflexivity and emergence, provide a potential way of locating multiple approaches to kinetic controls, which in turn would appear to be a major factor in determining the range of kinetic form. 54
Systems, fields and reflexivity
Some implications for kinetics Apart from cybernetics providing alternate models for conceiving control systems, how else has this chapter informed the inquiry? The work of Kaufmann provides an analysis of static composition in terms of part-to-whole. Could this be adapted for considering the design of kinetics? Two built architectural examples suggest how this might be considered. Rietveld’s Schroeder house was cited by Leatherbarrow as an example of a new spatiality, with the corner placement reconfiguring internal and external spatial relationships. But the Schroeder house window ‘operations’ were also designed in terms of the impact they would have on the composition as a whole. As explored by Mulder et al., the degree of opening of windows and shutters (including the impact of colour) was carefully considered as part of the asymmetrical composition.58 In a similar manner, Mies Van der Rohe constrained the movement of the internal blinds of the Seagram Building to conform to the overall composition.59 By restricting the movement to three positions, the movement always occurs in relation to the proportions of the whole. The result is a dynamic enlivening of the modernist facade, as individual blinds are moved over the course of the day within geometric constraints designed in relation to the whole composition. As these two examples indicate, the dynamics of simple window openings might be considered in terms of Kaufmann’s three systems. While the discussion of cybernetics opened up the possibility of emergent form, the legacy of compositional part-to-whole composition operates at the other end of the spectrum. Graeco-Roman, Baroque and the reverberation of Kaufmann’s third system provide various approaches to relationships between parts, which potentially can be mapped to kinetics. The relative movement of parts and their relationship to the whole could be programmed as slow series, or subtle loops. For example, Wittkower’s compound number ratios, when used in temporal sequence, would presumably produce a measured harmonic kinetic. Similarly, Baroque dissonance could be introduced through counterposing discrete areas of movement within a facade. The third system, based on repetition and scaling, could also be explored to produce the kinetic equivalent of reverberation. The extension of Kaufmann’s three systems to include the uniform texture of the modernist curtain wall suggests alternative tactics. Rather than clear demarcation between parts, subtle deviation and inflection could produce a minimalist aesthetic of movement. This approach could be extended through the compositional tactics of ‘fineness’ articulated by Reiser and Umemoto. Oscillation via graduation and superimposition suggest possibilities for kinetic versions of their field–field approach. These precedents, along with the various cybernetic approaches to feedback and control, would appear to provide a basis for kinetic compositional ‘systems’. Potentially, they provide a continuum of possibilities, from directed proportional modulation to recursive, indeterminate kinetics. Alongside compositional systems and cybernetics, Kwinter’s work on Italian Futurism provides a link to artistic practice that, while not involving kinetics, invites development in this regard. A primary theme is the idea of form being the dynamic outcome of non-linear systems: force fields in the case of plastic dynamism; and catastrophe theory in relation to the Stati d’animo paintings. While direct application of either of these may not be appropriate, one key concept that he uses to 55
explain a non-linear system may be usefully developed. As a means to conceptualize catastrophe theory, Kwinter cites the example of flow on a two-dimensional plane. The interaction of simple two-dimensional parameters allows insight on the complexity of three-dimensional fluid dynamics. The intuition here is that the visualization of kinetic form as sets of parameters mapped against temporal flux, may provide a productive design strategy. The idea of manipulating such parameter variables to develop architectural form is central to the practice of Greg Lynn and others, providing an obvious precedent for kinetics. There is no intention here to literally translate particular design techniques. Rather, the idea of interaction between parameters can be adapted for conceiving the variables that determine the motion of kinetic facades. The concept of a plane, and the mapping of parameters against time as a vertical axis, provide a framework for layering the various design variables that determine kinetics. Before exploring these insights from the wider perspective of architecture in more detail, there is one further area of precedence to be traversed: that of kinetic art. While traditions of facade composition, insight from cybernetics, Kwinter’s planar model and parametric practice may provide a conceptual basis for design, architecture has no ‘language’ for kinetic composition. Is there an established kinetic morphology, or language of form that can be usefully adapted for architecture? One would expect that the field of kinetic art, now approaching a century of development and discourse, may provide the basis for locating a design nomenclature for the aesthetics of motion.
Kinetic art The temporal arts Kinetic art is manifest in a rich variety of ways from floating mobiles to thumping mechanics, each sharing a common focus, an aesthetic of movement. For architecture the current motivation for kinetic facades tends to be utilitarian, but the instinct that drives this enquiry is the potential for poetic composition. The previous chapters have identified designers and theorists developing the ways and means to realize kinetics. However, despite the promise of the technology, there is minimal discussion of the compositional opportunity afforded by kinetics. Architectural facades are typically functionally orientated, for example to provide shade, ventilate a space or communicate information. Yet, regardless of purpose, the capacity for movement presents an opportunity. Those examples where composition has being explored are, given the infancy of practice, tentative. The previous chapter concluded with some promising ideas of how to conceive design variables. What might be achieved by manipulating such parameters – can the legacy of kinetic art provide an agenda for architectural facades? Frank Popper’s Origins and Development of Kinetic Art provides a comprehensive introduction to artists experimenting directly with form in space, and will be examined for its relevance. From Popper’s insightful review we turn to leading artists, in particular those who articulate a theoretical stance to their practice. While there are numerous artists who make isolated reflections on their work, two figures come into focus – George Rickey and Len Lye. Rickey is known as a theorist as well as a significant practitioner, through his book on the origins of constructivism.1 The theoretical contribution of Lye was less recognized in his lifetime, but through biographer and theorist Roger Horrocks, we are able to trace the evolution of Lye’s distinctive kinetic figures.2 From the discussion of Rickey in particular, and through comparing the planar aspect and vertical orientation of a facade, the specificity of architectural kinetics is developed. Towards the end of this chapter it is proposed that facade composition is more appropriately thought of in terms of pattern, as opposed to form or figure. The examination of kinetic art enables a working definition of kinetic pattern to be articulated, which will underpin the design experiments undertaken in the second part of this book. Rickey and Lye were working in an analogue era and they share a passion for precision engineering. The current generation of kinetic artists, in the main, work computationally in one form or another. Alan Dorin is an artist and theorist who has developed a classification of process, which he argues underpins what he terms kinetic-virtual art. Dorin’s taxonomy provides
a closing contribution for kinetic pattern, from the contemporary perspective of virtual-kinetic art. It is anticipated that examination of kinetic arts practice may provide some insight, but before doing so, a short outline of architecture’s relationship with other temporal forms – music and dance – is undertaken. In a similar manner to Chapter 1, this provides a focus to the inquiry. It is argued that while there are some interesting music and dance interactions with architecture, they are not of central relevance for this enquiry into spatial movement.
Frozen music Architecture has traditionally been classified as a spatial or plastic art, as opposed to the arts of time.3 Kinetic facades are inherently temporal. They are in flux from one moment to the next, and while surveyed in space, kinetic composition is manifest over time. This raises the issue of precedent in the temporal arts, such as the welltravelled analogy of music and architecture. The speculations about the relation between music and architecture are probably as old as both arts themselves. Generally speaking, they occur on two levels: the intellectual and the phenomenological.4 The phenomenological tradition in architecture suggests a connection between the experience of architecture as temporal sequence and the serial progression of music. The movement of the occupant, through the design of circulation and the manipulation of space, have been referenced to musical composition.5 The focus of the inquiry here on kinetic morphology bypasses the phenomenological, negating this thread of discourse. Of more relevance may be the intellectual relationship suggested in the above reference. In the Renaissance, systems based on the principles of musical harmony were theorized for most arts. The belief was that harmonic number provided a universal approach to the production of aesthetically pleasing outcomes in art, architecture and music. This was primarily achieved by the use of number series to create ideal proportional relationships. Alberti’s principle of concinnitas, discussed in the previous chapter, locates the use of harmonic number as the guiding principle for what Kaufmann referred to as the ancient system of part-to-whole composition. This tradition is examined closely by Robin Evans, who critiques the accepted historical view. Through a review of drawings and remaining buildings, he argues there are very few examples where there is direct correlation. According to Evans, the proportions were seldom explicitly implemented in the Renaissance.6 Rather they were an ideal that informed underlying compositional approaches, often distorted by the use of perspective drawing techniques.7 According to Evan’s analysis, the correlation between musical harmony and architecture is more complex than it might appear. In a section discussing the role of architect and composer Iannis Xenakis in Le Corbusier’s design studio, Evans locates what he calls a ‘good closing shot’ on the relationship between music and architecture.8 Xenakis borrowed Le Corbusier’s modular proportional system to develop a musical Fibonacci series. He also undertook an extraordinary musical breakthrough through the use of the drawing 58
technique of a ruled surface. Originally developed in engineering practice, Xenakis used the ruled surface in Le Corbusier’s studio where it was used to rationalize the curvilinear geometry of the Philips Pavilion.9 Xenakis adapted the drawing method to compose music, where the horizontal axis represented time and the vertical axis represented pitch. The ‘closing shot’ Evans locates, is the inversion of the relationship between music and architecture. Xenakis uses architectural techniques – Le Corbusier’s proportional system and the drawing technique of ruled surface – to generate a new form of music composition. Rather than architecture as frozen music, the compositions of Xenakis in the 1950s might be considered as a case of music as temporal architecture. The idea that compositional techniques in music can be used to generate a geometric sequence has continued with the uptake of digital design strategies.10 Rather then transposing musical harmony via proportional series, there have been some isolated experiments with digital mapping. Typically this involves translating the tempo, dynamics and rhythm of music to geometric parameters in order to generate form.11 Generally, the mapping is to overall form rather than facade, with no development of an overall conceptual framework that can be adapted to meet the research aims here. Moreover, the typical approach is to map composition derived from music as linear series of elements ‘frozen’ in static architecture form. Phrases from a piece of music might be used to inform composition, say an undulation in a facade from left to right. The implication is that the surveyor, casting an eye across a facade or traversing a building in space, can read or sense the musical phrase. Perhaps some informed individuals can do this, but the impact of architecture on music evident in Xenakis is arguably more direct. As opposed to the intellectual reading of architecture as a frozen musical phrase, the drawing informed spatialization from architecture has a direct and continuing place in contemporary music.12
Dance notation While the dance/architecture analogy does not have as long a history as music, the correlation covers similar territory. The tactic of mapping dance figures as the means to generate architectural form has been explored in the same manner as mapping music.13 However, most speculation has involved discussion of the use of dance notation as a way to describe movement through architectural space.14 The method potentially most applicable for kinetics is that of Eshkol-Wachman.15 In this approach, movement is abstracted as two-dimensional geometric forms, and, while it was developed for dance, it is proposed that the technique can be used to describe any movement in space. The method has been translated to the fine arts as shorthand for describing and conceiving visual art.16 Other references to dance occur in the work of Paul Virilio and his proposition of the dancer on an oblique plane, as a counter to Le Corbusier’s Modulor.17 However, despite this range of activity there is little that can advance this inquiry. The correlations are similar to music. Dance is interpreted phenomenologically, as a way of mapping movement of the occupant in relation to static form. Or, as in the mapping of musical phrases, it is ultimately realized as static architecture that requires the surveyor to make an intellectual association. 59
Kinetic art While aesthetics generated by movement can be traced back to ancient wind chimes, the beginning of kinetic art is associated with avant-garde experimentation of the early twentieth century.18 The generally acknowledged starting point is Naum Gabo’s 1920 publication of the realist manifesto and his exhibition of Virtual Kinetic Volume in the same year. Kinetic art explicitly introduced the temporal dimension into art, and movement was incorporated into works hung and framed as conventional paintings, freestanding sculpture both in and outside the gallery, machine works, and installations at a range of scales. With Anthony Calder’s exhibition of mobiles in Paris and New York in 1932, the genre received heightened exposure.19 He dominated the pre-war period with a series of developments on the mobile theme, while the most prolific period for kinetic art was during the 1950s and 1960s. In addition to the continuing popularity of Calder, prominent artists include Schoeffer, Takis, Len Lye and George Rickey. As illustrated by Kepes’ The Nature and Art of Motion, kinetic art was part of a larger movement, which explored the potential of motion to transform practice in a range of art forms.20 Within architecture there is a legacy of this larger movement within the Bauhaus pedagogy.21 From the 1960s onwards, kinetics began to be incorporated with early computers, and the associated genre of cybernetic art took centre stage.22 Despite this breadth of activity, there are relatively few artists, critics or historians who have developed significant theoretical discourse on kinetic art. From the 1960s to the 1970s we can locate: George Rickey, Morphology of Kinetic Art;23 Frank Popper, Origins and Development of kinetic Art;24 Guy Brett, Kinetic Art: The Language of Movement;25 and Frank J. Malina, Kinetic Art: Theory and Practice.26 This flurry of activity was followed by two decades of relative silence. As remarked by Yves-Alain Bois, ‘kinetic art suffered the unhappy fate of a flash in the pan’, and became associated with mechanical gadgets and domestic kitsch such as the ubiquitous 1970s lava lamp.27 More recently, there has been some resurgence in interest, with Guy Brett editing Force Fields: Phases of the Kinetic in 2000 28 and Art that Moves: The Work of Len Lye, published in 2009 by Roger Horrocks.29 The contribution of Brett and Frank Malina will be briefly considered, before looking closely at Popper’s encyclopaedic survey and the distinctive contributions of Len Lye and George Rickey. Malina compiled selections from the journal Leonardo as a book on kinetic art theory and practice. This provides a mix of intriguing artworks, which are categorized as pictorial (optical, mechanical, chemical), audio-kinetic or electronic. The majority are descriptions of technique or individual outcomes, with Malina’s paper being the exception. In describing his Lumidyne system (modulating light via mechanical filtering) he speculates on the broader issues of the range of a motion aesthetic. He briefly articulates range in terms of natural motion (flames, water, wind) and machine enabled, such as the lateral translation and pivoting enabled by a crane. In a similarly brief discussion of aesthetics he considers the problem of attention span of the surveyor, arguing against imposing ‘a well defined plot in time’ in favour of constant variation. He concludes by stating the field is too new to establish a full aesthetic understanding, leaving the artist to ‘search empirically for satisfying visual 60
experiences’.30 This description certainly fits his fellow contributors, resulting in the book being a compendium of experiments with a range of techniques. Guy Brett, some five years later, undertakes a similar review of practicing artists. His approach, by comparison to Malina, attempts to place kinetic art in a wider context. He locates kinetics as part of a wider arts language where ‘movement presents the possibility of a work of art whose form is a process of growth’.31 Brett is somewhat scathing of the tendency for ‘stylistic connotations, most of them technical’.32 For Brett, kinetics is representative of a wider artistic shift from art object to art as process. He highlights the grappling of twentieth-century avant-garde painting and sculpture against the Renaissance conception of space. Kinetic art is considered, by comparison, a natural outcome of the ‘relativity of things’.33 Rather than seeing kinetics as moving objects, these are conceived as sampling a new space, that of matter and energy. Brett’s short book is beautifully evocative, but unfortunately not particularly useful for this comparatively prosaic inquiry. Given the intent here is to map the boundaries of kinetic form within a very specific context, we move on to perhaps the complete opposite of Brett: the compendium of Frank Popper.
Popper: kinetic procedures Origins and Development of Kinetic Art is considered the primary reference text on kinetic art. It provides a comprehensive overview, from the early experiments of Futurism to the high point of kinetic art in the 1950s and 1960s. Considerable attention is paid to distinguishing types of kinetic art, with the primary classifications being: • • •
virtual/real spatial/non-spatial predictable/non-predictable.
This provides an overall classification within which Popper undertakes two detailed analyses: a taxonomy based on what he terms procedures; and second a summary of aesthetic categories. After rejecting three categorizations – geometric type, semantics, plastic element (e.g. line, volume, texture) – Popper describes the rationale for a classification of kinetic art in terms of procedure. It consists in examining the procedures used by artists to convey, represent, suggest or introduce movement into the plastic arts. The notion of procedure is sufficiently general to allow us to deal with the technical, semantic and plastic aspects of movement. At the same time it is sufficiently particular to allow us to make distinctions between the various ways in which artists have approached the problem.34 Popper discusses 31 procedures that are grouped into 8 categories: figuration, representation, formal suggestions, precise perceptual suggestions, photographic procedures, filmic procedures, movement itself, and various. Given the scope of this study, our interest resides primarily in the category of ‘movement itself’. The four groups of procedures listed under this category are as follows: 61
• • • •
simple mechanical electro-mechanical, electronic, thermal and magnetic mobiles projections, reflections, refractions of light.
Apart from the last procedure involving the manipulation of light, these map reasonably well to the first three aspects of the definition of kinetics informing this research. The fourth aspect of the definition – controllable transformation of material properties – is partially covered by a procedure in Popper’s ‘various’ category described as ‘movement through the growth or deterioration of the material’.35 It would appear that kinetics as defined for this study of facades might be productively aligned with selected aspects of these procedures.
Three-dimensional movement Of particular relevance to this research is Popper’s discussion of three-dimensional works of movement. The first work of this type is attributed to Naum Gabo, who exhibited Virtual Kinetic Volume in 1920. This consisted of a vertical strip of metal attached to a motor which vibrated to produce the profile of a curvilinear volume. Gabo’s artwork was developed in conjunction with ‘The Realistic Manifesto’, also published in 1920 which represented ‘profound reflections on the problem of movement in art’.36 According to Popper, the manifesto denounces the Italian futurists’ representation of movement through superimposition of multiple moments in time, such as the painting of Balla.37 ‘The Realistic Manifesto’ identifies five principles for art: colour is renounced in favour of tone; lines are to be used ‘only as a direction of the static forces and their rhythm in objects’; volume is renounced with ‘depth the only pictorial and plastic form of space’; mass is renounced as a sculptural element in favour of planar constructions; while static rhythm in the plastic arts is to be replaced by ‘kinetic rhythms as the basic forms of our perception of real time’.38 According to Popper, Gabo’s reticence to follow up his pioneering artwork into actual movement was due to his dissatisfaction with available mechanics. He proposes Gabo’s theory on movement was developed further by Laszlo Moholy-Nagy. His artworks emphasized relationships between light, space and movement, with his most influential work being his Light Machine, developed between 1923 and 1930. According to Popper, although this work comprised many moving parts, it was the embedded lights and resultant shadow play projected into the space of the exhibition that was the most powerful aspect of the work.39 Popper continues to describe the most significant artists within the threedimensional category, making a primary distinction between mechanical works and those where the kinetics is the result of natural forces. The former include Pol Burry’s ‘mobile planes’, the Dadaist-inspired machines of Jean Tinguely, Nicholas Schoeffer’s cybernetic inspired works, the use of magnetic force by Takis, and the ‘movement choreography’ of Len Lye. These are discussed in relation to the degree by which mechanics are revealed and the tactics by which unpredictability can be introduced. The theme of unpredictability is also behind Popper’s distinction between mechanical works and those powered by natural forces. The origin of natural movement 62
is traced to the suspended reliefs of Tatlin and Rochenko, which are described as being part of ‘a prolonged investigation of the interplay between interior and exterior space’. However, it was Anthony Calder who developed the tactic of suspension, with influential exhibitions in 1932 of what became known as mobiles. Popper cites Jean Paul Sartre’s catalogue description of these works. His fundamental aim is to express movement in terms of movement itself, and to avoid at all times the problem of imitation. Most of the time, Sartre writes, ‘he is not imitating anything. I know of no art that is less mendacious than his.’40 Two other artists discussed by Popper who explore the unpredictability of suspended forms reacting to air currents are Kenneth Martin and George Rickey. Martin, according to Popper, brings an architectural background to provide an alternate palette of materials and a mathematical precision.41 Rickey brings a similar precision, pivoting long counterweighted rods, typically in pairs or small groups, to create what Popper describes as ‘the subtle possibilities of movement prolonged into space’.42 Besides the emphasis on the value of unpredictability, Popper articulates another recurring theme – the most successful kinetic works are those that, like the slow and graceful movement of Rickey’s counterweights, operate at micro time scales. For example, Schoeffer’s Anamorphoses series of small sculptures are discussed in terms of ‘the infinitesimal units of time’.43 Similarly Pol Burry’s works, while using totally different procedures, are discussed at length in relation to an alternative time scale. His true aim is to present an image of pure movement. This might be called movement for movement’s sake, so subtle that it can be indicated simply by relationships of material or texture, as in the case of smooth balls on rough surface, or vice versa. By this insistence on the slowness and irregularity of movement, he aims to achieve a new region of existence – a universe which leaves behind the world of forms and concentrates our attention upon relative ‘tempi’.44 Popper concludes his book of kinetic art with a chapter titled ‘Sketch for an aesthetic of movement’. This is an assemblage of 27 aesthetic terms taken from the theory and history of art, organized into categories: the intellect (e.g. surprise, humour); the environment (e.g. nature, life, machine aesthetics); sensibility (e.g. hypnosis, anguish); action (e.g. grace, acrobatic); transcendence (e.g. evolution, sublime). This bewildering compilation of terms from aesthetic theory and art criticism are of minimal use for this inquiry, except where Popper qualifies these specifically in relation to kinetics.
Unpredictability and the agogic Popper’s book on kinetic art is a thorough survey up to 1968. It is all-inclusive, tracing the tendency to movement in impressionism and the abstract art of surrealism. Unfortunately, this encyclopaedic breadth, and the proliferation of categories it 63
produces, is generally not insightful for kinetic facades. However, we can extract two potentially useful observations. Popper’s emphasis on what he terms ‘procedures’ is used to categorize activity. These are discussed in terms of an aesthetic of movement, and from this two themes – unpredictability and the agogic – are informative. The use of unpredictability in the most effective and engaging works is consistently noted by Popper. This suggests that the key variables for designers of kinetic facades may reside in the approach to control. Popper notes that the work of Calder, in the delicacy of response to environmental forces, enables an inherent unpredictability. He also detects a contrasting approach in the cybernetic works of Schoeffer, ‘a mechanical element of ”indifference” in his most elaborate work’.45 Elements of indeterminacy were located in some of the contemporary kinetic facades examined in Chapter 2, but, in the main, a directed approach was taken to control systems. The re-evaluation of design variables to allow for degrees of unpredictability would appear to be a crucial factor if an engaging range of kinetic patterns is to be produced. The second recurring theme was the careful consideration of temporal scale, which Popper describes in relation to the musical term agogic. In music this refers to the use of accents to prolong the duration of a note.46 Popper refers to the work of theorist Étienne Souriau who differentiates the agogic from rhythm. In Souriau’s view the agogic or tempo (with its nuances of andante, adagio, presto, etc.) must be firmly distinguished from the rhythm, which only refers to ‘a certain structural organization of time, generally referred to as a cyclic figure’. The agogic also refers to nuances of acceleration (allegro and andante in music), and the category is technically almost impossible to explain – fastness and slowness being simply the response to an impression. Two groups can nevertheless be singled out: values concerned with excitation or incitation, and values connected with appeasement and calm.47 Souriau developed his idea of the agogic as an explicit reaction to the ‘rather banal description arts of space in contrast to the phonetic and cinematic arts’.48 Architecture is included in his proposition, although this is based on the reading of static form in terms of perception – for example the eye being drawn up a tower by articulation of openings and detail. Of more interest is Popper’s use of the term to describe the quality of temporal pattern that he identifies in a range of works. At one extreme is the elegant slowness and smooth acceleration of George Rickey’s counterweight suspensions. At the other is the velocity and dramatic choreography of a Len Lye installation. The term agogic conflates speed, acceleration and duration and would appear to be a significant aspect of kinetic form.
Lye: figures of motion As I was looking at those clouds I was thinking, wasn’t it Constable who sketched clouds to try to convey their emotions? Well I thought, why clouds, why not motion? Why pretend they are moving, why not just 64
move something? All of a sudden it hit me – if there is such a thing as composing music, there could be such a thing as composing motion. After all there are melodic figures, why can’t there be figures of motion?49 What is a figure of motion? After the above epiphany, which is dated about the same time as Gabo’s ‘Realistic Manifesto’, Len Lye, a young New Zealand art student, would spend a lifetime addressing this question. Roger Horrock’s Art That Moves documents Lye’s contribution as an experimental filmmaker, kinetic artist and writer. Of more significance for this study, he analyzes the body of work to develop ‘one artist’s perspective on the art of motion’.50 After confirming a similar distinction made in this study between spatial kinetics and other approaches to movement, Lye’s term ‘figures of motion’ is discussed in some depth. ‘By ”figure” Lye meant ”form or shape”, but what interested him was form as something performed, something that involved a process.’51 According to Horrocks, this locates the distinctive aspect of the work. Lye studied particular motion figures that interested him and meticulously composed these as repeatable performances. Initially, he developed this figural approach through experimental film, made by scratching directly into the film surface to produce ‘performances’ of generally four- to five-minutes’ duration. Typically, figures of abstract body shapes or rhythmic lines would dance across the screen in a series of repeating motion phrases. Lye shifted to kinetic sculpture in the 1970s, repeating similar figures to the experimental animation, and developing others through ‘doodling’ on paper, and with handheld steel sheet and wire rods. The kinetic sculptures, like the animations, are also typically a five-minute performance, distinguishing Lye’s work from the continuous works of most other kinetic artists. Lye undertakes an insightful investigation into a particular range of figures. In a list titled ‘tangible motion forms’, ten are methodically diagrammed and annotated.52 All but two of these are combinations of rods, aligned vertically and set in motion by force applied at the base: swaying ‘grass’, revolving ‘fountains’, reciprocating harmonic motion, double and multiple overlapping harmonics. The exceptions are a singular vertical blade struck with a ball on a rod, and a horizontal ‘wingtip’ vibrated into oscillation from a central pivot point. Arguably, the consistent vertical orientation and singular point of force is significant, related to his embodied approach to practice. Lye would seize a sheet of metal or bunch of rods and experiment by hand, shaking, flicking and twisting in a form of physical doodling. Once figures he liked were observed, he would then start the challenging job of reproducing the motion mechanically and composing the four- to five-minute performance. Experiencing a Len Lye work is, like the working practice, physically engaging. He aimed to stimulate what he referred to as the old brain, triggering deep memories of motion.53 The figures of movement are typically accentuated by dramatic lighting, which in conjunction with the sound of the whispering or crashing metal, heighten the physical and emotive experience of the figures. ‘Tangible motion forms’ appear to be a list of completed or proposed works, rather than an open-ended exploration. It documents a personal vocabulary of figures developed through an embodied working practice. The works are clearly not 65
intended to be comprehensive, but are a valuable example of one artist’s considered and refined approach to composing motion. What insight can we gain? One challenging aspect for this study is the emotive charge of Len Lye’s works. Is it possible to describe movement, independent of the ‘full body’ experience? Probably not, but this is not the intent of the inquiry. As declared in the opening chapter, this study is deliberately reductive, but with a positive aspiration. Using Lye’s term, if a rich range of underlying figures can be located, these provide a potential morphology of motion for aspiring designers of kinetics. These should be described as simply as possible and be open ended to allow multiple permutation and combination. Such a kinetic morphology may be akin to a form of notation, allowing composition at an abstract level. The example of Lye’s ten ‘tangible motion forms’ demonstrates a notation used to document particular kinetic sculptures. These appear to be an artist reflecting on a body of work, identifying the underlying structure of each composition. Horrocks documents other forms of ‘notation’ – the expressive drawings Lye refers to as doodling. These may appear to be spontaneous marks generated by hand movements, but according to Horrocks, Lye would spend weeks refining and identifying particular figures.54 These carefully conceived kinetic phrases would ultimately be manifest as a bodily engaging performance. Lye presents as an artist able to conceive kinetics through abstract drawing technique, which identifies the essential structure, or, in Lye’s terminology, ‘figure’. The paradox is that these analytical moments on a page transform to a four-minute ‘energy construct’ (a term Horrocks adopts from the poet Charles Olson).55 Lye’s working method and artworks are a personal, individual search for particular figures. They are not intended to be all encompassing, but they suggest the fertile use of abstraction be it words, drawing, or physical experimentation. The figures do not cover a particularly wide range of kinetic forms, being addressed to particular series that interested Lye. They do provide some reassurance though, that even the most embodied kinetics can be conceived through an abstract morphology.
Rickey: the ship at sea While George Rickey receives minimal attention within Popper’s book, he has written what I consider to be the most significant theoretical text by a kinetic artist, his Morphology of Movement. Rickey trained as a painter in Paris before returning to America in 1949, where he started producing steel sculpture based on a system of meticulously engineered counterweights and bearings, activated by air currents and the pull of gravity. He would continue to refine his work for the next 53 years, while at the same time teaching and writing in the United States. Rickey’s essay is one of the few attempts at a formal discussion by a leading artist and provides a succinct overview of general directions for the period up until 1963. These include: optical phenomena, such as moiré effects; transformation due to the motion of the observer; machines where mechanization causes ‘orchestrated’ movement; light play; and ‘movement itself’.56 These general directions are reasonably self-evident, apart from the expression ‘movement itself’. The term comes from ‘The Realistic Manifesto’, in which Gabo observes the limits of Italian Futurism: ‘It is now obvious
to every one of us that by the simple graphic registration of a row of momentarily arrested movements one cannot re-create movement itself’.57 In a similar approach to Popper, Rickey argues that the ontology of kinetic art is best addressed by dealing directly with actual movement, rather than optical effect. He has a particular dislike of the use of machinery, where repetitive motion generates, for him, ‘a more emphatic stasis’ than lack of motion. The argument is that true kinetic works are those where the capacity for motion is designed and is intrinsic, allowing an experience of ‘movement itself’, without the distractions of mechanics, form, relief, colour or figurative associations.58 Kinetic art is discussed in relation to other abstract arts, with the argument that there is immediacy to kinetics, which facilitates an intuitive engagement with the surveyor. Motion is measured by time, of which we all have some rather precise perception. We can compute it, sometimes with uncanny precision, witness catching a ball, passing a car on the highway, or riding a surfboard. We can measure slow-fast, long-short, pause, interval, beats per second, period of swing, coming towards us, going away from us, acceleration, vibrations separated, vibrations as a tone – these are all measurable without comparison with other objects or recollections of past experience or relation to other events in time; they have a kind of immediate measure, which, in spite of the abstractness, can give a sense of scale.59 Rickey continues his assertion that the essence of kinetic art is the design of movement, by articulating a range of examples: ‘the classic movements of a ship at sea (pitch, roll, fall, rise, yaw, shear)’, ‘vibrating springs’, ‘the non periodic movement of a pendulum’. For Rickey these are evidence of a vocabulary of form in motion, small in number and surprisingly simple, ‘scarcely more’ than the twelve tones of Western music.60 Continuing the analogy to language, this vocabulary is arranged over time, which is described by Rickey in terms of a possible syntax for kinetic art. There is besides the ‘vocabulary’ a sort of ‘syntax’ of motion. One aspect of this is sequence, such as the merging of traffic lines at a tunnel
Figure 4.1 The ship at sea (redrawn from George Rickey, 1963)
entrance and their dispersal, after a strict sequential episode, at the exit, or the knocking down of a long line of dominos.61 A possible syntax of motion is distinguished from spatial sequence in terms of the irreversibility of time. One can postulate plus or minus length, breadth and height: there is no negative time. So sequence must always involve subsequence. Becoming irreversible, sequence is unlike the other three dimensions, impossible to arrange symmetrically. A reversed sequence is really a later sequence in opposite order.62 The use of sequence to produce repeating movement is discussed, with the observation that if these reoccur without deviation, they quickly become tedious. This leads Rickey to propose a central factor in alleviating the problem of mechanical kinetic art, that is, the incorporation of chance. This is discussed in terms of analogy with movement in nature. However, while natural movement provides an infinite catalogue, Rickey warns of the dangers of verisimilitude. He argues that, paradoxically, it is the abstraction of the movement of a marionette, for example, that permits artistic interpretation.63 Having introduced the concept of a vocabulary and syntax of kinetic art, and having discounted direct mimesis, Rickey discusses at length the state of kinetic art in the mid 1960s. While he acknowledges the value of experimentation, it is argued that true art evolves from the careful and conscious exploration of particular types of movement. The idea of ‘type’ is subsequently discussed in relation to form, or what Rickey interprets as the morphology of kinetic art. When Roger Fry wrote of ‘significant form’ he was trying to find a universal equivalent of the distinguishable or identifiable forms of different epochs and cultures – of what Giotto, El Greco, Cézanne and Negro art had in common. His was an idea of ‘form’ as a discernible factor in a value appraisal. The ‘form’ referred to in these paragraphs is without immediate aesthetic or quality implications and is closer to ‘style’ or to the morphology of modern art.64 The question for Rickey then becomes what is the morphology of kinetic art? He argues that, despite the popularity of Calder or the novel appeal of much kinetic art, these have not been developed by talented artists into ‘monumental’ artworks. He argues that the essential form or morphology of kinetic art is not the mobile, but still resides in the undeveloped ‘Realistic Manifesto’ of Gabo and Pevsner.65
Vocabulary of kinetics? In contrast to Popper’s proliferating sets of procedures, Rickey is focused on what he considers to be the essential and differentiating characteristics of kinetic art. His morphology of movement provides a direct link to the origins of kinetics, through his referencing of the realist manifesto. Rickey summarizes a range of activity that 68
is considered under the umbrella term kinetic art, and then rejects all but those that focus on ‘movement itself’. This coincides with the trajectory of this inquiry. The focus here is on locating the specificity of kinetic facade design. While the design of a facade involves traditional activity of weathering, environmental control and the nuts and bolts of fabrication, the defining aspect is the capacity for kinetics. Designing kinetic facades adds a new agenda to architectural design, the design of ‘movement itself’. What does this phrase actually mean for architecture? Can Rickey’s articulation of a vocabulary of movement and his suggestion of a syntax of composition be directly applied? These are questions that can be productively explored. Rickey’s example of a ship at sea is simultaneously accurate, in terms of indicating axis of rotation or translations in space, and poetically evokes an aesthetic particular to the circumstance. The ship is an object suspended between fluid and air, with movement by sail or engine dampened by the viscosity of water. The hull is floating, able to simultaneously slide or pivot on any axis; movement is directional, but with the inertia of water and wind mediating the propulsion of screw or sail. There is a particular tempo and the sublime evocation of mass and force resolving in slow motion, an example of ‘movement itself’. Does this provide a ready-made vocabulary of movement for the design of kinetic facades? The analogy of the ship at sea will have to be examined further, but the precedent of an evocative language that accurately describes types of movement is inspirational. The challenge is to develop this precedent to address the particular context of kinetic facades. The other aspects of his morphology of movement that may be usefully explored are the observation of the irreversibility of time. His argument is that this implies movement cannot be symmetrical. This is a challenging concept for architecture, which has a long tradition of symmetrical and asymmetrical composition. Rickey’s assertion is based on his concentration on the temporal aspect of movement. Symmetry, from his logic, is a spatial quality that cannot translate to kinetics. This is a questionable assertion and perhaps is related to his particular approach to practice. Rickey’s artworks are consistently the free movement in space of a singular or doubled object, or small clusters of similar objects. They produce a constant variation in movement, dependent on interaction with wind currents and the counterbalancing weights embedded in the construction. The chance of a repetitive sequence or the symmetrical interaction of the minimal number components is extremely unlikely. However, one could conceive of works composed of multiple adjacent parts that did produce simultaneous identical movement to produce temporal forms of symmetry or asymmetry. Rickey’s morphology references his freestanding sculpture, while the kinetics of facades typically involves multiple moving parts, orientated in a generally vertical plane. Facades involve patterns of movement between multiple parts, as opposed to the movement of a singular object in space. This contrast between singular or small numbers of parts, with the planar expanse of the multiple, suggests a way to define the specificity of kinetic form for facades.
Movement itself: from part to pattern It is proposed that the distinguishing aspect of kinetic facades lies in patterns of motion, which arise from the movement of multiple parts. These parts have an 69
individual vocabulary – translation, rotation, scaling and their combinations – which is analogous to Rickey’s vocabulary. But it is the manner in which movement is synchronized or offset in time, the part-to-part clustering or dispersal of parts as patterns of movement, which is the equivalent of movement itself for kinetic facades. It is these abstract patterns of movement that come into focus as the specific object of this study. Rickey’s essay is driven by an artist’s passion for his craft. It is simultaneously a critique of activity and the evocation of an ephemeral art of motion via a range of analogies. The writing is poetic, then bombastic or at times critical of artists who pursue the novel and spectacular, rather than his version of focused practice. The morphology is an engaging piece of writing that evokes more than it declares. Rickey’s analogies and critiques of kinetic artworks are not codified in terms of an essential diagram or taxonomy of approaches, such as the all-inclusive lists of Popper. Therein lies the charm, and a certain frustration, in reading Rickey’s text. At some disservice to Rickey’s prose, his morphology of kinetic art can be summarized as four linked principles: 1 There are a small number of basic movement types. 2 Basic movements combine to produce composite movement. 3 Basic and composite movement, in combination with temporal variables produces sequences of movement. 4 Sequence may have a spatial dimension, i.e. the sequence may occur in the same location and/or propagate through space. These principles provide a basis for describing movement at the scale of a singular part. They reflect Rickey’s art practice which typically consists of singular or small groups of geometric forms, pivoted to allow sequences of composite movement within a fixed spatial location. However, kinetic facades operate at a different granularity than Rickey’s sculpture. It is proposed that the scale and configuration of facades require these principles to be developed in terms of kinetic pattern. Rickey’s morphology of kinetic art provides general principles for conceiving the formal structure of kinetics at the scale of a part. This has direct relevance to kinetic facades, when considering geometric transformation at a part level. Moreover, the principles of movement sequence, based either on variation within the same spatial location or sequence based on propagation through space, provide valuable insight for considering the formation of movement patterns within architectural facades. It is proposed that kinetic pattern results from relative movement of individual parts, which produce differentiated clusters of similar movement or propagation of similar kinetic resonance across a facade. This working definition will be explored further in the design experiments framed in subsequent chapters.
Dorin: taxonomy of process There has been a resurgence of interest in systems theory and cybernetics by contemporary artists and critics. Kinetics is not necessarily at the forefront of this, but there are some potentially interesting connections between artists working 70
with kinetic process. Before examining the contribution of Alan Dorin, the current theoretical landscape of what has been termed ‘systems aesthetics’ will be briefly sketched. The basic distinction between homeostasis, second-order cybernetics and emergence was introduced through Katherine N. Hayles in the previous chapter. Her synopsis was chosen for its relative clarity, and, conveniently for this section on contemporary approaches to kinetic art, she is a media art theorist. Although exact distinctions between cybernetics and systems theory are a confused and at times contested arena, the general principles espoused in Chapter 3 double as an appropriate introduction here. As noted by Edward Shanken, ‘the terms cybernetics and systems theory are often used interchangeably’.66 Shanken has recently published a valuable summary of current discourse, in which he identifies the central references for contemporary critique as being Jack Burnham’s 1968 essay Systems Esthetics, and his book Beyond Modern Sculpture published in the same year. Systems Esthetics, is a wide-ranging attack on formalist art. The emphasis for Burnham (paraphrasing 1960s systems theory in the sciences) is on process rather than art objects. Rather than a novel way of rearranging surfaces and spaces, it is fundamentally concerned with the implementation of the art impulse in technological society. As a culture producer, man has traditionally claimed the title, Homor Faber : man the maker (of tools and images). With continued advances in the industrial revolution, he assumes a new and more critical position. As Homor Arbiter Formae his prime role becomes that of man the maker of esthetic decisions.67 This manifesto for an art of process places the emphasis on aesthetic decisions embedded in the specification of the system. Burnham’s ideas have been used by Shanken and others to counter two tendencies in contemporary art.68 The first argues for an alignment of the arts of technology (in particular those using computational process) and the now accepted legacy of other process art, such as earth art and video. The theoretical and political objective is to legitimize contemporary ‘Information Arts’. The second use of Burnham’s ideas is to turn these against artists and curators who focus on the objects that result from systems. The provocation is to resist the normalizing tendency of formalist art traditions and their emphasis on an object, and exhibit the artwork as (to paraphrase the realist manifesto) the system ‘itself’. This body of discourse provides a theoretical backdrop for this inquiry, but the more prosaic aims here are to understand the mechanisms of process from the point of view of a designer. As in the previous generation of 1970s kinetic artists, artists who practice and articulate a coherent theory of potential relevance are rare individuals.
Process for virtual and kinetic art Alan Dorin is a contemporary artist whose systems aesthetic is based on the computational generation of what has become known as ‘artificial life’.69 His artworks are typically mutating three-dimensional forms, or an evolving soundscape generated 71
algorithmically. The artworks are kinetic in character, where forms grow and decay in a manner analogous to cell growth. The kinetics is virtual, manifest on media screens or as electronic sound. His most significant theoretical discussion for this enquiry is an essay that proposes that such virtual kinetics can be analyzed in relation to a small number of distinctive process. He articulates a classification of physical process, which, he argues, provides a basis for creative practice and a conceptual tool to undertake analysis. Just as a painter manipulates and coordinates colours and a composer combinations of sound, kinetic artists coordinate processes. The computer is a powerful tool for the kinetic artist willing to explore the utility of algorithms represented as program code. Yet the kinds of processes which may be modeled need to be studied, just as a painter studies colour.70 The use of a computer to model kinetic process defines an activity that Dorin terms ‘virtual-kinetic’ art, where the artist uses ‘simulations to create representations of movement’.71 He cites a range of precedent, but views Wolfgram’s categories of complex dynamic systems and cellular automata as the most relevant.72 These are, according to Dorin, the following: 1 2 3 4
move into a homogeneous state (limit-point) move into simple, separated, periodic structures (limit-cycle) produce chaotic aperiodic patterns (strange attractors) produce complex patterns of localized structures.73
Dorin proposes Wolfgram’s taxonomy is not complete, as it was based on describing outcomes rather than the full set of processes that determine them. His proposition is introduced by clarifying that the emphasis on process requires perception over time, i.e. that the temporal dimension is central to any discussion of kinetic art. For Dorin, kinetics is the outcome of a process which can be reduced to five actions: pulse, stream, increase, decrease, complex. The example Dorin cites as a pulse is the regular pumping of a heart. He equates the concept of stream to Wolfgram’s category of a ‘limit-cycle state’.74 The spacing between events in a pulse can be of such a scale that it is perceived as uniform stream. These occur at the limits of visual perception – a revolving sphere may be rotate so fast that it appears motionless, or so slow that process is not apparent. This category, according to Dorin, is not part of Wolfgram’s taxonomy. He then introduces two more processes that are implicit – there will be either an increase or a decrease in kinetics. These are characterized by forever higher or lower intensity in which the nature of the change is constant. The final category is complex process, which is equated with Wolgram’s identical term. Complex process forever change into new forms without reiteration. Hence a particular state of a system undergoing a complex process will 72
be different to all future and past states of that system. The changes the system undergoes may occur in a predictable (but infinite) sequence or they may be random and unpredictable (noisy).75 The taxonomy is demonstrated by using it to analyze traditional garden fountains and water-based artworks. He then describes its use in relation to conceiving new works. For the programmer employing physical simulation for artistic purposes, the taxonomy outlines the fundamental techniques required to devise virtual-kinetic works. Each physical process to be implemented may be seen as a mixture of the five process types. Hence at a basic level, a process will be related to many others which might also be implemented. The programmer can begin, instead of with a blank slate, with a process category from which to develop a satisfactory simulation.76
Process as syntax? Dorin’s taxonomy of physical process, for what he describes as virtual-kinetic art, appears a useful precedent. Dorin and his colleague Jon McCormack have developed significant artworks based on the application of computational process.77 Their works are typically presented using computer display monitors or video projection, so they do not meet the definition of spatial kinetics that shapes this inquiry. However, while the taxonomy is clearly orientated towards virtual-kinetic arts practice, it is based on the description of physical process conceived in the abstract terms of periodic structure. Dorin argues that kinetics can be described and conceived as variations and combinations of a small number of processes. Given the discussion of Rickey’s argument for a minimal set of movement types, which he describes as a vocabulary, it is tempting to consider these periodic structures as providing a type of syntax. Could the variables that influence kinetic pattern be conceived in these terms? That is, can pattern be conceived in terms of Rickey’s movement types – roll, yaw, pitch, etc. – that are controlled by combinations of Dorin’s taxonomy – pulse, stream, acceleration and complex? These are possibilities that need to be evaluated within the specific context of architectural facades. The previous discussion of Rickey’s morphology has already revealed that direct translation of theory from the kinetic arts to the context of this research can be problematic. While Rickey’s ideas were accurate for his form of kinetic art, it was proposed that the particular scale and granularity of facades required thinking in terms of pattern. Dorin’s description of a kinetic process in terms of a limited set of periodic structures will be examined further in the next chapter, where a generic framework for conceiving the variables that influence kinetic pattern will be developed.
Decision planes Rewind < 1 This is a book of two parts. The previous section articulated the scope and intent of the inquiry, scanning precedent in architecture and kinetic art to locate relevant theory and practice. For Part II, the inquiry shifts to a more active mode. In the following chapters, aspects of the scan through precedent are used to propose a general model for design. An instance of this model is then developed to explore kinetic range through a series of design experiments. The penultimate chapter attempts a classification of the outcomes, as a first pass at reflecting on the possible extent of this new design space, the aim being to locate the general bounds of form, and articulate a useful set of terms – a nomenclature for designers and critics. To this end, a provisional taxonomy is developed from kinetic art, based on the surface patterning of the sea. As will be revealed, this approach to locating difference is found to be problematic. In the final chapter, a more open set of terms is proposed, through the concept of state change. This final nomenclature embraces the capacity for continuous and qualitative multiplicity – that kinetic pattern can slide from difference by degree to difference in kind. Given this shift to research through design experimentation, it may be useful to briefly restate the scope of this book and summarize the previous discussion. The focus is on actual physical movement of a kinetic facade. Facade is loosely defined as a zone between the outside and inside of architecture, generally oriented towards the vertical. The interest here is not in the enabling technology, nor the design of the kinetic components – the focus is on the kinetic form, that is, ‘movement itself’. The motivation is to map out the edges of this new design space and to this end three questions were articulated: What are the design variables that determine kinetic form? What is the potential range? And what nomenclature can usefully communicate design intent? These queries are pursued through the abstraction of morphology. This enables experimentation with the underlying design parameters, which, when manipulated, produce shifting intensities of kinetic pattern. Part I located relevant precedent in contemporary architecture, the wider traditions of facade composition and insight from kinetic art. Discussion of contemporary theory and practice revealed a number of built examples and innovative prototypes. The emphasis of current architectural activity is on the ways and means – the techniques and technology of kinetics – with minimal discussion of the formal possibilities. In Chapter 3 the survey of kinetics was extended to consider approaches to the static composition of architectural facades. Strategically chosen
sources on traditional and contemporary approaches to facades were examined. Kaufmann’s analysis of part-to-whole composition as three generic systems provided an example of a generic, rather than stylistic interpretation of form. This was extended by including the field composition enabled by the modernist curtain wall, with the idea of the field examined further, from its roots in Italian Futurism to contemporary designers. From this, the use of cybernetics in architecture was discussed as a potential model for conceiving kinetic control systems. The discussion extended beyond architecture in Chapter 4, to consider the kinetic arts, starting with the standard reference text by Frank Popper. His encyclopaedic approach and proliferation of categories were challenging, but two points that recur were insightful: the use of elements of unpredictability in kinetic art; and the importance of temporal scale, as explored through the agogic. Two significant artists, who also have contributed to theory, were subsequently examined. Len Lye conceived kinetic form as ‘figures of movement’. The figures he explored were particular to his practice, but demonstrated a mode of working that encapsulated ideas through drawing and physical miniatures. A subsequent analysis of George Rickey proved very useful, in particular his Morphology of Movement, where he identified a minimal kinetic vocabulary. From a discussion of these ideas it was proposed that for architectural facades, kinetic form is best conceived in terms of pattern. A working definition was proposed: kinetic pattern is the relative movement of individual parts, which produce differentiated clusters of similar movement, or propagation of similar kinetic resonance across a facade. The discussion of the kinetic arts concluded with contemporary use of Burnham’s ‘systems aesthetic’. In particular the work of Alan Dorin, who has proposed taxonomy of kinetic process for what he terms virtual-kinetic art. Part II builds on these ideas from architecture and the kinetic arts to develop a framework, or model for conceiving design. Kinetics shift architectural focus from the design of objects and spaces to the articulation of a process. The outcome of kinetics is not the design of the component parts, nor is it the enabling technology. The ‘thing’ that is designed is the process that determines a multiplicity of kinetic pattern. There are strong parallels with the systems aesthetic of contemporary art practice; the subtle specification of systems to produce environments in flux, as opposed to discrete art objects. The current generation of architectural designers will see a relationship to parametric design techniques. For kinetics the approach is similar, but the ends radically different. Rather than manipulating parameters to locate an optimal (singular) design, for kinetics, the parametric system is the outcome. The aim of this chapter is to build on the ideas developed in Part 1 to locate the variables that determine kinetic pattern. To this end, a conceptual model for design is developed. This does not represent a theory for design or an argument for a particular kinetic outcome. The modest goal is to articulate the interrelationships between sets of design variables, in terms of potential influence on kinetics. By bringing together disparate sources from architecture and kinetic art, a general model is developed. Through argument and example, this is populated by what are considered to be the most influential design variables. Kinetics requires a shift in 78
focus, from the optimization of a singular design solution to the specification of the ‘variable space’, from which multiple designs will emerge. As a way to visualize this design space, the idea of decision planes is developed. Design is conceived in section, as three planes of activity. Linking these in the vertical axis are the omnipresent parameters of temporal scale and periodic structure.
Animated variables The contemporary use of parametric techniques in architecture introduces a mode of designing applicable to kinetics. Parametric design has been enabled by realizing the power of the computer as a processor of information. Currently, there is a shift away from replicating analogue techniques, towards working with three-dimensional form generated by the manipulation of global geometric parameters.1 The approach is based on conceiving a three-dimensional computer model as a series of linked assemblies, so that changes in parts are propagated throughout the whole. The geometry of the computer model can be controlled at the ‘meta level’ by parameters, which propagate changes throughout the various assemblies. By manipulating these higher-level variables, a wide range of geometry can be generated in relation to factors such as site conditions, surface-to-volume building efficiencies, or as the means to experiment with novel form. The biological morphology of D’Arcy Thompson and the gradient fields of artist Hans Jenny have been cited as a precedent for parametric design.2 A more direct architectural reference is William Mitchell’s concept of design vectors. Mitchell, in a section from his influential text Digital Design Media, briefly outlined the potential of animation for design. He anticipated that data structures of geometry and materials could be animated between data states or what are known in animation as key frames, allowing exploration of a design solution space. The Cartesian product of the ranges of variables in the data structure is the multidimensional solution space that the designer explores. A keyframe corresponds to a point in that solution space. Linear interpolation between keyframes, then, yields a vector in the solution space.3 Written at a time in which the capacity of hardware to animate large data sets in real time was not available in design practice, Mitchell could see the potential of computer animation to provide the means to explore design variation. Availability of sufficient computing power to allow rapid interpolation and playback of these sequences opens up the exciting possibility of extensive high-speed exploration by vectoring through a solution space.4 One of the first to realize this approach was Greg Lynn, who as discussed in Chapter 3, utilizes animation techniques to produce kinetic design assemblies. Using a range of techniques, including animated particle systems, Lynn’s design process consists of manipulating the kinetic assembly to explore a design solution space. However, the kinetics is a means to an end – the refinement of a singular optimized design. Parametric design may allow an animated design process, but for Lynn and many 79
other contemporary designers, the design objective remains the same – to identify and refine a singular static artefact. By contrast, for kinetic facades the outcome is the parametric system of interlinked data sources, controls and kinetic assemblies. Kinetic facades are realized as a parametric process-in-action. The design of kinetic facades sets a different agenda for designers, who have traditionally worked towards finding the best static mix of performance and elegance. Rather than a fixed tectonic form, the outcome is a kinetic process that interacts with users and performs in response to changing environmental and socio-cultural context. The emphasis is on the design of the system, or to use Mitchell’s terminology, the design of the ‘solution space’. This requires creative manipulation of three interrelated sets of variables: what input and how this is ‘sampled’; the logic of the control system that processes this data; the parameters of the building components or materials that will move in response to the control system. Typically, architects focus on the final stage – the design of the physical components or specification of materials. However, if the opportunities offered by kinetic facades are to be realized, designers need to be involved in the design of the input and control systems as well as the components. The design of kinetic facades is similar to parametric design where the designer animates a kinetic assembly to generate a range of outcomes. Except in this case of parametric design, there is no final form; rather, the design outcome is the specification of variable limits, from which multiple forms will occur over the life cycle of the building. The focus of parametric design in architecture is on geometry, with the kinetic assembly being ‘frozen’ when an optimized state is found. From this point, the selected form goes through detailed design, which involves a construction-orientated parametric model. Ultimately, the objective is to produce a singular, static design. For kinetic facades the construction is of a parametric process-in-action, or variable space. This shift in emphasis towards variable space as an outcome requires the consideration of data capture and control logic as significant variables, alongside the geometric design of the kinetic assemblies.
Towards a framework Throughout the reviews of precedent, the input-control-output paradigm, which has its origins in systems and cybernetic theory, has been recurrent. The pioneering text on kinetic architecture by Zuk and Clarke, considers kinetics in terms of taxonomy of machine controls that range between human, mediated and ‘cybernetic devices’.5 Contemporary researchers Fox and Yeh update this analysis to propose six types of control systems, which add a degree of complexity but are essentially variations on the input-control-output model.6 This is clearly a useful structure for considering kinetic facades, but it is proposed here that the manner in which it has been used is dominated by a functionalist emphasis. There is little development beyond the abstract terms of input-control-output. What is required for architecture is a more particular model, developed in more detail to address the range of variables that occur when designing kinetic facades. The input-control-output paradigm is typically represented as a diagram of boxes and arrows, with the content of these nodes not declared. They are conceived as neutral containers, representing the 80
legacy of cybernetics, where all biological and man-made processes are conceptually described by the same system. What would a particularized model look like for kinetic facades? The approach taken here is to adapt the representation of complex systems as sets of parameters on a plane, interacting over time. As introduced in Chapter 3, Kwinter identified its use to visualize the complex interactions of catastrophe theory. It is a mathematical model, where x and y variables are conceived as ‘repellors’ and ‘attractors’ in a constant state of flux. Interaction via feedback mechanisms produces catastrophic changes, which cause the system to ‘flip’ and produce a completely new form. The well-known illustration is the behaviour of a stressed dog that oscillates between cowering and aggression.7 At moderate stress levels there may be a smooth transition between these states, while at high stress a cusp may be quickly reached, triggering a sudden change of behaviour. The adaptation of the planar model for the purposes of this study is independent of this legacy. There is no intention to consider kinetic facades in terms of catastrophe theory. Rather, it provides an idea for a spatialized framework to visualize interaction between multiple dynamic variables. Putting mathematical catastrophe theory to one side, the planar model allows the positioning of two variables. The values of these might remain constant, or move in a constant zone, but this does not necessarily produce a regular and consistent result. The third variable, time, impacts on the resultant form and, if this is non-linear, there may be extreme variation at different periods. Nevertheless, the range of outcomes of the dynamic system can typically be anticipated as a range between the planar limits.
Decision planes: from discrete measure to continuum There is an obvious criticism to the adaptation of the planar model – the limitation of the number of parameters. A plane articulates two parameters, equating to the x and y axes, plus time represented as the vertical axis. Superimposing this on to the three phases of input, control and output would allow six axes, plus the temporal parameters, which would likely be too reductive. There is, however, a way of alleviating this. Rather than conceiving the plane in terms of two discrete variables, literally calibrated to a measure such as temperature, each axis can be considered a continuum between extremes. As an example, consider the variables of kinetic type. In Chapter 1, kinetics was defined as three geometric transformations and their compounds, together with movement based on material deformation, such as elasticity. A restriction to two discrete types, for example, measures of translation and rotation, would give a finely graduated range of compounds between the two. However, in this case, the parameter range would exclude scaling and material-based transformations. Consider instead a continuum between simple geometric transformations at one extreme and simple material transformation at the other. The axis between geometric and material transformations would not be based on discrete individual types, but would locate range between extremes. At one end would be kinetics based solely on either translation, rotation and scaling, while at the other would be complex movement based on material change, such as mass or elasticity. In the middle would be the range of hybrid geometric/material 81
combinations, with zones of compound geometric and hybrid material transformation either side. The adaptation of the planar model to map design variables against time is referred to here as ‘decision planes’. Explicit in the term is that decisions taken in relation to variable range and their periodic structure, determine the morphology of kinetic pattern. A decision plane, as conceived here, is an abstract mechanism to locate broad design decisions in relation to design variables. The essential feature is that it embeds time as a central design variable, highlighting the distinctive aspect of the field of kinetic facades. Variables are considered in terms of a continuum between extremes, allowing the potential for multiple intermediate hybrid conditions. As illustrated in Figure 5.1, these are represented as zones rather than discrete measures against x and y axes. A plane allows the mapping of two continuums, with the zonal location of their intersection indicating the variable space for a design. This mapping might be fixed, indicating minimal deviation; oscillate over time within a zonal range; or be randomly indeterminate. These planar variables are mapped against the vertical temporal axis to locate the impact of temporal scale and periodic structure. The planar model, when overlaid on the input-control-output paradigm, allows the conception of three planes, each having two variables that articulate a broad continuum between extremes. The model maps design variables, where zonal position on the plane represents the variable range. The dynamic of these variables is continually being reassessed over time, represented by the vertical axis. The inputcontrol-output terminology is adapted for the specifics of architectural facades: data input, or what will be termed here as sampling; the processing of this by the control system; and the way in which this is manifest, which is captured by the architectural term tectonic. Specification of sampling, control and tectonic, against periodic structures of time, articulates the variable space that shapes kinetic pattern. As a mechanism to visualize the design of kinetic facades, the planes provide a generic template which can be developed to consider the range of variables that influence kinetics. Specification by designers locates the variable space, within which kinetic pattern will manifest over time. How might identification of variables be undertaken? The approach taken
Figure 5.1 Diagram of design variables conceived as a planar continuum between two extremes. Location on the plane identifies the zone of a design instance, but when mapped to time, multiple outcomes are possible from the same combination of variables
here is to develop the ideas located in the scan through architecture and kinetic art undertaken in Part I. This example and precedent enables the specification of the variables for each plane: the definition of kinetics and a review of contemporary and historical examples enable the articulation of tectonic variables; cybernetics considered as a range between homeostasis and emergence provides one continuum for the control plane; the particular context of facades locates the second as spatial configuration; while the sampling plane is articulated in terms of data source and density. The sectional decision plane model allows time to be considered in terms of a vertical axis, indicating its significance for all design planes. This requires identifying variables in terms of periodic structure, and to this end Dorin’s taxonomy of kinetic process provides a clear guide. From these examples and precedent, variables that impact on kinetics are identified as a means to progress the inquiry. These are not definitive but represent the most significant within the scope outlined here. The variables should be seen in the wider context of the research aims: they provide a provisional set; in the next chapter they are examined in more detail for impact on morphology; a selection of these, an instance of the developed decision plane, is used to frame design experiments; these provide material for analysis and critique, ultimately leading to the proposition of a morphology and nomenclature for kinetic facades.
Sampling source and density What are the design variables that affect the sampling of information? Sampling is a term borrowed from electronic music and refers to the practice of using small loops of music to generate a composition.8 Considering this stage of the process in terms of a range of possibilities makes it explicit that input specification is a design variable. Specification determines what information is to be sampled and, as such, excludes or includes opportunities. The scan of contemporary practice revealed the sampling of either environmental control data or socio-cultural sources. Is there any advantage to be gained from mixing data types? In the context of kinetic facades, most approaches to data sampling are tuned to particular uses, predominantly that of environmental control. It would seem self-evident that data sampling will depend on the objectives of the design. Typically, in an environmental context a wind gauge would not be used for temperature control, while motion sensors, for example, are regularly used for media facades but seldom considered for environmental facades. However, multiple data sources can be combined to give a more accurate representation of either environmental or socio-cultural context. For example, combining information on wind and temperature allows incorporation of windchill factor. While for some climates, the percentage humidity is as important as temperature. The range of potential samples could also include motion sensors to determine occupancy, adding to the range of variables to be processed by the control system. In terms of socio-cultural sampling, the majority of examples occur in media facades. There is an extensive range of precedent for devices developed for interactive media, which sample data from both local and remote sources. Architectural facades have a long tradition as a form of communication and there exists an opportunity to utilize the
precedent of such installations to revitalize the agenda of facades in a world of digital information.9 How would sampling influence kinetic pattern? From the above discussion, arguably one key variable would be the range of data sources sampled. Take, for example, a kinetic screening system that is designed to moderate direct sunlight on a facade. Light sensors that monitor the presence of direct sunlight could be installed. If this was the only data sampled and the shades were repositioned to allow particular views out when not needed, the resulting pattern of movement would be a map of daily cloud activity and sun position. There would be a regular daily pattern, based on position of sun and the screening of adjacent urban or natural form, interspersed with periods of cloud cover. This daily pattern would gradually shift over the course of the year. However, if other data were sampled, such as allowing users to override the system for an individual location, this pattern would be interspersed with isolated and generally unpredictable events. Another scenario might be to utilize the screening system in its ‘down time’ – cloudy days or at night – as a low-resolution tangible interface. For example, graphic forms could be embedded that communicated an aspect of local interest, or a programme of artworks could be commissioned.10 The added value of using such multiple data sources is outside the scope of this research, but clearly multiple data sources potentially influence the complexity of kinetic pattern formation. It is argued that data source is a key variable. However, rather than use a distinction between environmental and socio-cultural source, it is proposed to use the generic distinction between quantitative and qualitative data. This more robust terminology allows the incorporation of qualitative input into environmental systems and vice versa. A second aspect that could be considered in terms of articulating data type could be the distinction between discrete and continuous data.11 We could anticipate that, if the type of data were highly discrete, the pattern formation would be less ‘smooth’ than continuous data. This would be highly significant, if the sampled data was directly interfaced with the tectonic output, by-passing the control plane. However, as evidenced by the survey of contemporary practice, this is not typical. For the purposes of this study of morphology, a more generally influential aspect of data sampling would be more appropriate. Beyond type of data, it would be reasonable to consider density of data sources. Arguably, the quantity and spatial distribution of samples would have a more consistent and direct impact on pattern formation. Referring to the previous hypothetical example of a sunscreening system, it can be posited that the number of sensors across a facade will affect the resolution of the outcome. The greater the number of sensors, the more fine-grained the sampled information and, potentially, the more complex pattern formation would result. In summary, it is proposed that the two broad continuums that have a consistent influence of pattern formation are data source, with a range between quantitative and qualitative sources, and density of data sources, with a range between a sparse and a dense number of sampling devices.
Degrees of control: reflexivity and spatial differentiation What are the key variables for a control system in terms of pattern formation? An examination of the contemporary examples provides some initial insight. First, there are a few examples that do not utilize any mediation between data and kinetic, which illustrate the impact of a lack of control: The Malvern Hills project uses a thermohydraulic fluid, and, as the fluid heats and cools, the expansion and contraction drive the position of the sunscreen; Ned Kahn’s wind wall designs are based on wind pressure on hinged plates, which produces a wide range of patterns. The resultant kinetic patterns map reaction of a kinetic assemble to data: temperature in the case of Malvern Hills, and wind pressure in the case of Kahn. The pattern is a result of the variability of the data source and the speed of reaction of the kinetic mechanism. At the other end of the spectrum are examples that are highly controlled, typically using a computer. These include the Aegis Hyposurface by dECOi, and the commercially available Flare wall relief by WHITEvoid, both of which can produce patterns through algorithmic functions or image sampling. There is a range of approaches in between. For example, the LIGO wind wall is reactive to wind, but this can be moderated by controlling the power of electromagnets embedded in the tips of the pivoted slats. Similarly, a number of projects use shape memory alloys that have the capacity to react to natural fluctuation in temperature, or in addition, external heat can be introduced allowing a degree of control. The earlier chapters disclosed several sources for considering types of control. Zuk and Clarke provided taxonomy of machine controls, before the wide availability of computers. The four levels of machine control involve two types of human interaction (singular or multiple controls) and two types of automatic controls (machine and computer). More recently, Fox has proposed a similar taxonomy of control systems for kinetic structure. He distinguishes between degrees of control: simple feedback mechanisms through mechanical actuators or sensors; digital controls that can process multiple data sources; and those able to integrate a heuristic or learning capacity. As discussed in Chapter 4, the type of feedback is central to the development of cybernetics: homeostasis where the intent is to maintain a steady state according to pre-defined criteria; second-order cybernetics based on reflexivity; and autonomous systems where there is no predetermined performance criteria, with the control being emergent. All of the above provide precedent which could be considered for articulating key control variables that influence pattern. Given the context of morphology, which seeks broad underlying structures, it is proposed to adopt the precedent from cybernetics. Cybernetics allows the intent of the designer, in terms of degree of control, to be highlighted. The terminology provides two distinct positions. At one extreme there may be a defined goal to produce a particular pattern of movement. In this scenario the control would be classed as homeostatic, evaluating the sampled data and maintaining a steady state to reproduce the desired pattern. At the other extreme would be controls where there is no preconception of outcome, where pattern formation would evolve through emergent processes. In between these two extremes would be degrees of predefined pattern formation and the facility for pattern rules to be reconfigured to produce degrees of indeterminacy. The general term 85
that will be used to describe the first variable for the control plane is reflexivity: that is, the degree to which the control system is open to reconfiguration. Put another way, from the perspective of a designer, the intent may be to produce a steady state, introduce varying degrees of indeterminacy, or allow the control system be reflexive to the point where pattern is emergent. What would be another key control variable, complementing reflexivity, that would influence pattern? Patterns are emergent over time but also have a spatial dimension. The spatial configuration of controls – how a system may control discrete areas of a facade – would have an impact. Take, for instance, the configuration of controls for a system of horizontal louvers. These are often controlled as linked assemblies, resulting in the segmentation of the facade into bays. Alternatively, there are some projects, such as the Nordic Embassy project, where each louver is individually controlled. Rather than pattern being based on large spatial areas, if the spatial configuration of the controls is at the level of an individual component, there is the capacity for more complex pattern formation. The spatial configuration of controls can be considered a variable continuum. At one extreme, the facade may be controlled in such a way that each component is simultaneously activated by one control. At the other end of the spectrum are control systems that allow each component to be individually controlled. This 1:1 configuration of control and component equates to Ashby’s Law of Requisite Variety, located in the discussion of cybernetics in Chapter 3. Between these two extremes are continuums of different spatial configurations, which provide different configurations of components and discrete or linked controls.
Tectonics: kinetic type and granularity The final decision plane concerns the output, or what is physically manifest. In this case, the term tectonics is used. This is often heard in the architectural design studio and refers to an aspiration for a visual quality that is inherent in construction assemblies.12 The intent of the decision plane framework is to focus on kinetic pattern and, while the design details and construction of the facade will be observed on close inspection, the focus here on morphology is independent of materiality. For this study, it is proposed that the ‘tectonic’ resides in the kinetics of a part. As previously discussed in the development of the decision plane structure, the definition of kinetics provides a range of movement types. At one extreme are simple geometric translation, rotation and scaling, and at the other, more complex and incremental movement based on change in material properties. In between these is a range of composite kinetic types, where simple geometric transformations combine to produce hybrid kinetics. Or the tectonic might involve a combination of material deformation and geometric transformation. Take, for example, a pneumatic system such as the net of inflatable pillows discussed in Chapter 2. The kinetics is a scaling transformation, which produces a consistent ellipsoid. However, if the materiality was controllable, allowing differential elasticity in the pneumatic surface, there could be a range of deformations. The consistent symmetrical scaling motion could be moderated by controlling elasticity so as to produce hybrid kinetic patterns. 86
Besides the kinetic type – be it a simple geometric transformation, a compound, or kinetics based on activating a material property – what other aspect of the implementation of kinetics would have an influential impact on pattern formation? Pattern for kinetic facades is reliant on parts or groups of parts producing areas of movement. If the granularity were coarse, for example the ratio of part to whole being in the region of 1:20, then pattern formation would be constrained. We might equate this to computer graphics resolution where 640 × 480 pixels constrain the fidelity of an image, or to playback speed where less then 24 frames per second can generate a loss of smooth motion. The degree of granularity – the number of parts in relation to the whole – will impact on the amount of detail in the pattern and, as a consequence, influence pattern range.
Temporal scale and periodic structure Integral to all decision planes is a consideration of temporal variables – periodic structures such as acceleration and rhythm. There is minimal information that can be gleaned from contemporary examples, beyond observations on relative temporal scale. At one extreme is the (unperceivable) motion at which the timber veneers of Ocean North’s responsive surface prototype operate, or the slow motion mechanical unfolding of Hoberman’s Iris Dome. At the other are the 60 km/hour pneumatic pistons of the Aegis Hyposurface. This is potentially only surpassed by the windresponsive facades of Ned Kahn. Each example has a temporal range at which it operates, but the impact which this has on pattern is unclear. One reference to speed from the kinetic arts is the agogic, where comparison was made between the slow motion of Rickey’s wind-responsive sculpture and the high-speed operation of some of Lye’s machine kinetics. Both Rickey and Popper consider the issue of speed an essential issue for the kinetic artist, with a tendency towards favouring either extreme. They argue that for kinetic art, engagement of the surveyor is best engendered by both slow motion and high velocity. Poignantly evidenced by one of the first acknowledged kinetic artworks, these limits have been well travelled in kinetic art: Gabo’s 1920 Virtual Kinetic Volume oscillates above the upper limit of perception so that it appears to produce a static volume.13 Detection of speed is one of the fundamental aspects of vision, but how does speed affect pattern formation? One example from vision research suggests the significance of contrast: laboratory tests have shown that against a background of random motion, a single dot moving at a consistent speed is detectable, even when its speed is at the lowest level of perception.14 This evidence from science had already being exploited by some kinetic artists, for example, the localized point disturbances of Pol’s kinetic fields.15 These examples from science and art suggest that speed, or what might be termed temporal scale, may impact on pattern formation. As discussed in Chapter 2, there are lower and upper limits to human perception of temporal scale. If no change is detected in a part of a visual field after more than 2–3 seconds, this is no longer perceived as being in motion. The upper limits are between 24 and 30 frames per second. That is, the temporal scale is a range between approximately 0.3 and 30 frames per second (with slight variation dependent on individual visual acuity). 87
increase / decrease
Figure 5.2 Global design variables of temporal structure overlaid on a decision plane
In addition to temporal scale, kinetics may have particular rhythms of movement, which can be considered in terms of an underlying periodic structure. Dorin’s taxonomy of process provides a useful way to understand periodic structure as a limited set of temporal process. He argues that virtual kinetic art is reliant on five processes: pulse – a regular oscillation of events over time; stream – events are evenly spaced; increase and decrease – events accelerate or slow down; and complex, where new forms occur over time without reiteration – this may be simple and predictive sequence or this may be random and unpredictable. As illustrated in Figure 5.2, this provides a range of periodic structures. The motivation for Dorin’s taxonomy is to relate temporal process in the physical world, to computational process used to generate digital kinetic works of art. There would appear to be the capacity for application to the decision plane framework. In effect, the categories provide a simple range of periodic structures which, when combined with a planar variable, have the capacity for a wide range of designs. For example, a tectonic variable, such as a simple translation, can pulse, move as a uniform stream, the frequency can be increasing or decreasing, or there may be a complex irregular periodic structure.
Continuum As illustrated in Figure 5.3, the tri-planer model has resulted in six planar continuums. For the sampling plane these are sampling source, with a continuum between quantitative and qualitative sources, and sampling density, with a range between a sparse and a dense number of sampling devices. The control plane has one continuum based on control reflexivity, where there is a range between homeostasis (steady state) and emergence. The second continuum on the control plane is the control configuration with a range between single and multiple configurations. The continuum for the tectonic plane resides in kinetic type and the tectonic granularity. Type is in effect the definition of spatial kinetic that underpins this research. It ranges from simple transformation, such as translation, through compound translations, to the fine-grained multidirectional kinetics enabled by material deformation. Granularity, like sampling density and control configuration, is informed by Ashby’s Law of Requisite Variety. The continuum between coarse tectonic and fine granularity determines to a degree the capacity for pattern formation. The vertical axis through 88
all planes represents the temporal variables. As illustrated earlier in Figure 5.2, these are periodic structure (pulse, stream, acceleration, complex) and temporal scale. The decision plane model allows the conceptual mapping of design variables that have a direct influence on kinetic pattern. In terms of a required shift in design focus, the model highlights that approaches to data sampling and control systems are key design parameters, alongside the typical architectural emphasis on tectonics. Each decision plane impacts on the overall variable space which determines multiplicity – the range of pattern that occurs over time. The model as illustrated here is generic and open to interpretation in relation to specific design contexts. In the next chapter each variable will be considered and developed for the particular context of this research. Morphology is an abstraction of physical form, enabling a study of theoretical range of formal types. As will be developed in detail, not all the planar continuums are directly relevant for the morphology design experiment.
multiplicity over time
( c o
Figure 5.3 Summative diagram of sampling, control and tectonic decision planes. Specification of variable continuum in combination with periodic structure and temporal scale produces design multiplicity over time
TECTONICS [ kinetic type ] ( granularity )
( s ing u
CONTROL [ reflexivity ] ( configuration )
( s pa
SAMPLING [ source ] ( density )
Experiments with kinetic pattern Index and intuition Kinetic pattern is ephemeral – clusters, thickenings and gradients of movement. The ambition here is to explore the bounds of such patterns, through a series of experiments that use the decision plane model from the previous chapter. What range will emerge from iterating through the variable space of sampling, control and tectonics? The number of possible combinations explicit in the multiple permutations between variables (in themselves continuums) is daunting. Hence the recourse to morphology, which provides methods that have provided insight when confronted with complexity. Goethe’s systemic intuition of botanical range through drawing provides one example of analogue computation through experimentation with morphological forces. By locating the underlying structures and their potential superimposition, he systematically imagined possible form beyond that observable in nature. In architecture, Philip Steadman, inspired by Goethe and D’Arcy Thompson, undertook morphology of plan relationships using mathematical calculation. Given the relatively small set of variables, Steadman used adjacency graphs to calculate a large, but ultimately finite, number of plan types. There are, after all, only so many ways a finite number of orthogonal rooms can be related. Given the range of variables for kinetic pattern and the ephemeral quality of movement, this level of precision is not likely for these experiments. Nor would it be useful, as the aim is to locate the general shape of possibility, as a guide for kinetic designers. The approach developed here is to iterate through variable space, in a manner akin to the experimental intuition of Goethe. His was a study of natural forms, and while this scientific analogy is well travelled, artistic practice has its own pedigree. Henri Focillon provides a useful articulation of ‘experiment’ in his mediation on autonomous formal development in The Life of Forms in Art. By experiment I mean an investigation that is supported by prior knowledge, based on a hypothesis, carried out with intelligent reason and carried out in the realm of technique.1 This description appears in the context of a discussion of architecture, where he was making the case for understanding diverse forms of the Gothic. Here it guides experiments that algorithmically explore kinetic pattern. The prior knowledge we
have gleaned from precedent in architecture and kinetic art is embedded in the decision plane model. My hypothesis is that the multiplicity of paths through this variable space should generate a sectional slice that locates the theoretical range of kinetic pattern. Through a mix of reason and intuition, variables can be combined to locate distinctive thickenings within the expanse of possible kinetic form. When considering the ‘realm of technique’, the tactics for these experiments are less obvious. Focillon proposes form is related to technique and matter, but, for him, technical or material primacy is but one aspect of artistic form. His proposition is that arts practice has an internalized logic or sets of rules. These are inflected by technique and material, but ultimately the ontology of arts practice is located ‘in the regions of the mind’.2 Through insight from artists such as Lye and Rickey, we have observed that the material form of kinetics is not the tangible mechanism, but movement per se. Arguably, technique and ‘matter’ reside between the finite range of kinetics at the scale of the part – translation, rotation, scaling and their composites – and the infinite ways in which multiple parts can be combined over time, as clusters and shifting gradients of pattern. In the following sections the detailed tactics for such an experiment are developed. The shift is to a mode of production, which carries forward the trajectory of the previous chapter, where a model for conceiving design as sets of interlinked variables was proposed. This tri-planar model provides a general framework that positions a multiplicity of design parameters as zones of influence. Here, an instance of this multiplicity is developed algorithmically to structure a series of design experiments. By referring to these as an instance, these experiments are flagged as but one of many possible ways to use the model. The realm of kinetic form possible through the three planar sets and their continual re-calibration against temporal structure is accepted as infinite. Any instance of the model is incomplete. Accepting incompleteness enables what Mennan refers to as a ‘light formalism’, where computational method meets intuitive experimentation.3 Shifting to a mode of research through design, a rigorous indexing of variables supplemented by intuitive experiment enables the hypothesis to be tested. That is, through index and intuition, the theoretical range of kinetic pattern can be explored. The aim is to locate distinctive clustering in the variable space, as an initial guide for designers to produce their own (incomplete) design instance.
Designing the experiment The decision plane framework was conceived as a way to visualize the relationship between sampling, control and tectonic variables. Using the definition of kinetic pattern developed in Chapter 4, a focused review of the seven variable continuums is undertaken. These are examined closely, with the objective of locating the most significant variables for the experiment. For this instance of the planar model, patterns will be evaluated in terms of morphology, where physical scale and the practical constraints of materials are not considered. Once the most significant parameter planar zones are located, the variables are subsequently developed in more detail, tuned to known algorithmic approaches to generating pattern. This facilitates a systematic approach to the production of animations that are generated to visualize 92
Experiments with kinetic pattern
the resultant pattern. The generation of animations are undertaken in three stages: [S1] kinetics at the scale of a part are combined and distinctive compound motion selected; [S2] these are then indexed to the continuum of control variables; [S3] from sifting through this methodically generated batch, latent forms are located as starting positions for more intuitive exploration. The ephemeral nature of kinetics provides a challenge to visualization of these experiments. The approach taken here is to build on the tradition of facade study drawings. This method of design study has been widely used to inform compositional approaches for static architectural designs. Typically, drawings delineate significant proportional relationships, indicate window fenestration, and sectional profiles, with quick shading techniques employed to suggest three-dimensional relief.4 They are not intended to represent the experience of the built facade, but operate as abstractions, diagrams of relationships between parts, profiles and surface effects. In a similar vein to such drawings, the proposed animations that result from these experiments are diagrammatic, as compared to a more instrumental approach to digital visualization.5 The deliberate adoption of a non-realistic mode is intended to focus attention on kinetic pattern, independent of physical scale, materiality or figurative associations. Included in the framing of the experiments below is a discussion of viewing angle, aspect ratio, the rationale for the geometry of the part and numerical scale (the number of parts).
Design Variables The tectonic plane The tectonic plane was conceived in terms of kinetic type (geometric to material) and granularity (coarse to fine). The geometric transformations – translation, rotation and scaling – are the three basic kinetic types. Combinations of these, such as translation and rotation producing a rolling motion, give rise to composite movement. The definition of kinetics included material deformation as an additional type of complex movement. However, for the purposes of a study of morphology, which considers pattern at an abstract level and does not take into account physical scale, the deformation of materials can be accommodated as combinations of geometric transformations. For example, at the micro scale, melting wax can be considered composites of translation and scaling. For the purposes of these experiments, kinetic type will be limited to the three basic geometric transformations and their composites. This can result from the combination of the same base movement, such as a rotation around the x axis and a simultaneous rotation in the z axis, which produces a composite twist. Alternatively, different base movements can be combined, such as a translation along the z axis and rotation around the x axis that produces a pitching movement. The second variable on the tectonic plane is granularity. Kinetic pattern results from relative movement of individual parts, which produce differentiated clusters or propagation across a facade. Clearly, the granularity of the proposed animation will need to be as high as usefully discernible (given the resolution of the display technology). However, for this instance of the variable space, there is no intent to evaluate the impact of granularity on morphology. For the purposes of 93
the animation experiment, a constant fine granularity (at the limits of the display resolution) will be adopted.
Control plane and kinetic variables The control variables identified in the decision plane model are reflexivity (from homeostasis to emergence) and spatiality (from singular to multiple controls). A control system based on homeostasis would, in the context of this research, suggest that the pattern formation would be predetermined. For example, simple wave-like movement patterns can be produced using regular control of period, typically based on proportional number series. Controls could be designed to interpolate sampled data, resulting in consistent wave patterns. By comparison, an emergent control would be relatively indeterminate. The control would be highly reflexive, with an extreme position being the autonomous interaction between singular facade parts. Typically, this is achieved algorithmically by the assignment of simple rules for part behaviour in response to the proximity of neighbouring parts. Patterns result from the part-to-part interaction, as defined by the control. There is obviously a range of hybrid controls between homeostasis and emergence: for example, indeterminacy constrained within certain thresholds; or emergent outcomes can be studied and subsequently manipulated to produce some consistence in pattern. The second variable identified for the control plane was the spatial configuration. The movement of a facade could be initiated as one spatial whole, which, in an extreme case, would simultaneously move all parts to produce a singular monolithic pattern. Alternatively, there could be multiple controls, each affecting separate areas of a facade. These separate controls could produce coordinated movement between areas, or be autonomous, where each zone has different types or reflexivity. Both control plane variables will directly influence the range of clustering or propagation of similar kinetics and, consequently, pattern range. These need to be considered in more detail. Starting with the case of a singular spatial configuration and an extreme homeostasis position on the decision plane, the simplest and most predictable outcome would be a control that produces simultaneous movement of all facade parts, with the same amount of force or amplitude. A variant of this simultaneous control would be to spatially differentiate amplitude. While each part making up the facade would be moving at the same time, the differential force would result in differing acceleration. This amount of deviation in acceleration could be predicted by proportionally altering the amplitude. Alternatively, a non-proportional spread of amplitude across the facade would generate less uniform outcomes. While there would be a potentially wide range of variation, controls based on proportional or non-proportional amplitude would still remain reasonably predictable. The introduction of multiple and spatially discrete controls introduces another level of complexity. Rather than treating the facade as a whole, it can be divided into areas of movement. Each could have a separate control producing coordinated movement between areas, or each area could be autonomous. For example, the temporal spacing or period between activation of areas could, in a similar manner to the amplitude variation, be proportional. For instance, rows of facade parts could be activated progressively over time, producing bands of vertical 94
Experiments with kinetic pattern
acceleration. If the multiple controls were synchronized, these bands of movement would be reasonably predictable. Alternatively, if non-proportional periods were utilized, the resulting patterns would, similar to non-proportional amplitude, generate more complex outcomes. Further complexity could be introduced if irregular-shaped spatial configurations with overlapping boundaries were used. This might enable the potential for unforeseen pattern outcomes. The controls, while predictable as singular events, when combined and overlapped may produce complex outcomes that may not have been anticipated. This locates multiple controls / irregular spatial configuration towards the middle of the continuum between homeostasis and emergent. As a way of further increasing complexity, unpredictability could be deliberately introduced into the controls. One approach would be to utilize random number sequences to generate irregular amplitude or period. A homeostatic control where, for example, rows of facade parts are coordinated by a simple periodic relationship, could be disturbed by the use of random amplitude values. The facade would have an overall pattern structure, with variance as a result of the random degree to which an individual part moves. An extension to unmediated random numbers has been developed in computer graphics in order to generate controlled variation between limits. Known as noise algorithms, these have been used to generate controlled variation across a three-dimensional surface to simulate naturally occurring relief such as wood bark. These same algorithms are used in motion graphics to add naturalistic variation to fluids or gas simulations.6 There is a number of differing noise algorithms and it would be possible to generate a range of movement patterns that would exhibit degrees of variation. It would be appropriate to consider noise algorithms as a separate group of controls placed within the emergent side of the reflexivity continuum, where variation and a degree of novelty can be produced within set limits. More emergent outcomes can be developed by allowing parts to operate independently. For computer-controlled animation, a prevalent technique for generating independent interaction is an approach known as cellular automata (CA).7 As indicated by the name, individual parts or cells operate independently according to an autonomous set of rules. These rules are typically simple instructions for interaction with neighbouring cells. With many interacting cells, complex indeterminate patterns of behaviour result. A commonly used approach is ‘life-like’ CA, which have been used to simulate naturally occurring phenomena, such as that exhibited by ant colonies.8 These typically exhibit a diversity of behaviour over time, alternating between apparent randomness and the formation of regularly ordered patterns. A second category of cellular automata, which produces behavioural patterns distinct from life-like automata, is cyclic cellular automation.9 Rather than a simple ‘live– propagate–die’ interaction based on proximity, values of cells are compared and change when a threshold is reached. Different threshold values produce different interactions and distinctive patterns. A third commonly used approach, which has similarities with the two types of CA, are flocking algorithms as originally developed by Reynolds.10 These were developed to simulate the behaviour that occurs when bird flocks or schools of fish move as a group. As with CA there is no centralized control. Rather, each 95
member of the group evaluates its position in relation to neighbours and moves according to three simple rules: collision avoidance, velocity matching and flock centring (attempting to stay close to nearby flock members).11 The combination of short-range repulsion, long-range attraction and alignment of direction produces distinctive but highly variable behavioural patterns. The above discussion has identified several strategies for the design of controls, through a close examination of the variables and by considering them in terms of algorithmic implementation.12 The hypothesis is that variance in the control will produce variance in the clustering or propagation of similar kinetics, resulting in a range of patterns. The aim was to identify typical approaches that can be used algorithmically to produce the animation experiments. In summary, these can be distinguished as five groups: the simple synchronous activation of all parts with the same amplitude; synchronous activation but with spatially differentiated amplitude; the third is reliant on sectioning the facade into areas and controlling the period of activation of one area relative to another; the fourth distinctive group is the use of noise algorithms to produce controlled variation, while the fifth is based on cellular automata and flocking algorithms.
Sampling plane and kinetic variables The sampling plane variables are data source (quantitative-qualitative) and density of samples (sparse–dense). The issue to be evaluated for these experiments is the potential influence of data sampling variables at the level of morphology. In the formation of the generic decision plane model, it was observed that data type and number of samples would have the greatest impact when there is no intermediary control system. The kinetic patterns would be a direct mapping of data, within the constraints of the technology driving the tectonics. However, once data is processed in some way, then arguably the influence of data type diminishes. Even for data sampled as discrete values recorded at long intervals, control systems could interpolate between discrete values to produce the effect of a continuous source. The intent of the animation experiments is to generate patterns at the level of abstract and non-scalar morphology. These studies will be generated algorithmically and, while it might be possible to link real data sources such as temperature, pedestrian movement or economic trends, this is considered unnecessary for these experiments. For kinetic facades it is the relationship between sequences of kinetic parts, the clustering or propagation of these along a facade, which is anticipated to be the major influence on pattern formation. For this experiment, data type is considered to be of minimal significance. Similarly, the number of data sources may impact in a real-world scenario but, in the context of morphology, the number of data sources is theoretically infinite.
Temporal scale and periodic structure The discussion of temporal variables (visualized as the vertical axis intersecting all planes) identified temporal scale and periodic structure as the most significant. Scale was considered in terms of relative speed, slow motion to high velocity within the constraints of human vision. Periodic structure was based on Dorin’s taxonomy of 96
Experiments with kinetic pattern
kinetic process: pulse, stream, acceleration and complex. Arguably, these periodic structures are implicit in the identification of control variables undertaken earlier. For example, synchronous activation of parts would produce a regular stream while the proportional differentiation by amplitude or period would produce a stream, an alternating pulse or an accelerating sequence, dependent on type of proportional series. A regular spacing, such as changing relative amplitude or period by a simple regular addition (2, 4, 6, 8 …), would likely produce a regular stream. These types of number progressions, where change is constant throughout the series, are known as arithmetic progressions (AP).13 A second type of proportional series is based on the multiplication of the preceding number by a common ratio, such as the series (1, 3, 9, 27 …). As discussed in Chapter 3, there is a long history within traditional composition of using common ratios to divide facades into proportional related parts. These types of compound number sequence are known as geometric progressions (GP).14 If a GP were used to control relative amplitude or period, we could anticipate an accelerating sequence or, if the sequence were reversed, a deceleration. By combining AP and GP, we could anticipate a movement sequence that would meet the description of a pulse. The last periodic structure identified through the taxonomy is that of complex. Dorin defines a complex process as those which ‘forever change into new forms without reiteration’ and identifies two types: ‘predictable but infinite’, or ‘random and unpredictable’.15 The use of control variables that incorporate noise algorithms, cellular automata and flocking algorithms, could be expected to produce the predictable but infinite or random and unpredictable variation as defined by Dorin. In summary, the proposed control variables would seem to incorporate a full range of periodic structure. How then might temporal scale be incorporated? We might expect observation of pattern to be dependent on speed, i.e. some patterns may be easily discernible at a certain velocity but are less coherent at others. Nevertheless, it can be argued that in morphological terms, the clustering or propagation pattern will be present regardless of the temporal scale. The analogy is with Steadman’s identification of ‘essential configuration’ of floor plan types, which exists regardless of relative size or feature thickness.16 A kinetic pattern configuration based on changing clusters of kinetics at a scale of a part, or propagation between parts, will be present, dependent on the variables of the controlling script. Within animation software, the speed at which it is viewed can be adjusted to emphasize the distinctive characteristics of the pattern. Temporal scale is another example of a variable that will have an impact on pattern formation, when real-world constraints of technical performance are considered, but, in terms of a study of morphology, it is less significant.
Summary of design variables Through precedent in kinetic art, the principles for structuring a study of movement pattern have been developed. For kinetic art, basic and compound movement combine to produce a movement sequence such as the choreographed figures of Len Lye, or the more indeterminate (but equally distinctive) free play of George Rickey’s wind-activated works. It was argued that for facades, composition is best considered in terms of kinetic pattern rather than figure or sequence. Facades have 97
Figure 6.1 Summative diagram of variables to be used for design experiments
TECTONICS [ kinetic type ] ( granularity ) ( c oa
on at i
o sl mp ra n co [T e sit co
CONTROL [ reflexivity ] ( configuration ) ( s ing
od -lik eri l i fe P s e ou tud ari
v li m mp or A f e s f re ou
Experiments with kinetic pattern
a particular expansive scale and vertical orientation, where pattern is based on relative movement between relatively large spatial zones. It has been proposed that the key variables for pattern at an architectural scale would be those that have the most influence on the spatial clustering or propagation of similar kinetics. In summary, it is proposed that, for the purposes of generating pattern range, the type of kinetics at the scale of a part needs to be considered – translation, rotation and scaling transformation in space. When considering pattern as the non-scale-specific relations between parts, kinetics based on material deformation is less significant. In terms of granularity, it was proposed that a constant fine granularity will be adopted, i.e. a study of range will be best served by as fine a granularity as is possible, within the limitations of the display technology. The control plane was anticipated to have the greatest influence on pattern morphology. This was examined in detail and a full range of reflexivity and spatial differentiation will be utilized. Moreover, these control variables embed the full range of periodic structures. It was argued, however, that temporal scale is less significant in terms of the structure of pattern formation. Finally, for the purposes of a morphological study it is neither essential nor appropriate to map real data sources or consider number of sources. As visualized in Figure 6.1, the variables constitute a specific instance of the decision plane, intended to explore range of kinetic pattern.
Visualizing pattern The objective of the experiments is to algorithmically generate a wide range of movement patterns for kinetic facades. These will be visualized as computer animations in a deliberately non-photorealistic mode of representation. Clearly, there are many nuances that would determine perception of a facade when built, and, while current technology enables photorealism, this approach is not considered appropriate. The emphasis is on the essential features of patterns rather than the nuance of surface detail. In effect, the approach is similar to traditional techniques used for the design of static facades, where typically there is a series of sketches that explore compositional ideas. They are usually line drawings undertaken by hand or composed on a drawing board, sometimes partially rendered to suggest three-dimensional relief. They do not represent the facade in an instrumental way but capture relative design ideas by way of abstraction. They are diagrams of intent, not representations of constructional reality. The proposal is to represent movement patterns in a similar manner, with no attempt to relate the patterns of movement to constructional systems or materials. The definition of kinetics that guides the scope of this experiment excludes optical effect such as transparency and reflection. In line with these constraints, the colour and lighting will be neutral, with no shadow projection. The focus of the animations is on visualizing patterns of movement, with the diagrammatic greyscale approach intended to locate significant kinetic pattern, independent of physical scale or materiality.
Aspect ratio and viewing angle Building facades come in range of proportions, from the extreme verticality of towers to the expansive horizontality of low-rise strip developments. In line with the 99
emphasis on abstraction, it is proposed to use a more neutral facade proportion for the purposes of this experiment. The animations will be viewed on a standard computer monitor and a 4:3 aspect ratio will be used, enabling the animation to fill the borders of the screen to provide a neutral viewing context without the distraction of a frame within a frame. The viewing angle is informed by the precedent of traditional facade study drawings. Typically, these are elevation drawings, with facade relief indicated by line weight or neutral shading techniques. The animation set-up will use an orthogonal camera to generate an elevation view, with simple graduated shading to indicate three-dimensional depth. The scope of the inquiry (as articulated in Chapter 1) excludes the movement of the viewer and correspondingly the camera will be from a fixed position centred on the facade.17
Geometry of part and numerical scale Given that the movement patterns will be the result of multiple moving parts, the geometry and the number of parts need to be carefully considered. Rather than speculate, a pilot study was undertaken to test the impact of part geometry and numerical scale. Research on the visual mechanisms of human perception has informed the selection. While there is still debate on how human vision processes motion, the consensus is that the primary factor is feature recognition.18 In the context of a greyscale image on a computer monitor, controlled experiments by Derrington et al. indicate feature recognition will be determined by edge detection and relative shading.19 Figure 6.2 illustrates the final outcome of the pilot study. Each of the five shapes has variation in number of edges and relative shading. A sphere, when shaded, produces the most accurate depth perception, but because there is a circular profile, edge detection is difficult. Edge detection is improved with the case of a circular disc, with the front and back edges being distinct. Rectangular shapes increase the number of edges but result in less distinctive relative shading. Further shapes involving more edge detection and less ambiguous shading were generated. The triangular shape provided good edge detection and reasonable shading detection, but when combined and rotated to provide a closely packed area, distracting horizontal bands occur. The hexagon shape gave a good combination of edge detection and contrast between shaded areas. The trials showed that increasing the number of edges beyond six proved counterproductive, as edge differentiation became harder. Moreover, the hexagon provided a relatively neutral orientation when a large number were combined in an offset, closely packed configuration. There was no horizontal Figure 6.2 Trials of different geometry undertaken as a pilot study. Hexagonal parts provided the best mix of edge detection and shading depth for motion detection 100
Experiments with kinetic pattern
or vertical emphasis, and lateral movements were more easily recognized, as compared to the orthogonal grid that is the outcome of multiple rectilinear shapes. In summary, the trials demonstrated that a hexagon provided the best combination of edge detection and relative shading contrast, and can be closely packed, without privileging rectilinear pattern formation. The aspect ratio, viewing angle and geometry of parts have been determined. The remaining issue to be resolved is the number of parts utilized to represent movement patterns. When composing groups of objects, the generally agreed threshold for distinguishing individual parts is between five and seven.20 At seven and below, dependent on individual visual acuity, each entity can be tracked as a separate event. Above seven the parts are tracked as groups. For example, eight objects are tracked as groups of four, or as asymmetrical groups dependent on relations between entities. This sets a threshold at a lower limit to the number of parts at eight. Is there a logic for determining the upper limit? If we consider the control variables outlined previously, the proposal to use methods such as cellular automata (CA) has implications for numerical scale. The patterns that result from CA and flocking are reliant on interaction between multiple parts, but there is no established threshold of parts, below which pattern emergence is unlikely. Viewing published examples of the life-like and cyclic CA, typically the number of cells is in the region of 300–1000.21 The other factor affecting the upper limit of numerical scale is the fidelity of the computer monitor. Animation trials were undertaken, and there is a threshold above which the fine scale of the animations produces moiré-like patterns that interfere with the visualization. The trials revealed that, once the numerical scale got into the 500+ range, display interference was significant. It is proposed to undertake the experiment using a close packed 4:3 array of hexagon shapes of 21 × 19 which gives a numerical scale of 399.
Stages A staged approach to design is undertaken, where the results of the preceding are reviewed and determine the emphasis of subsequent exploration. The aim is to produce a wide range of animations through a mix of methodical and intuitive tactics. The strategy is to review and select the most distinctive examples from each stage for further iterations. In this way, it is anticipated that redundancy can be minimized, as repetitive outcomes can be foreseen and avoided. As important as avoiding repetition, is the intuitive identification of latency within individual patterns, which can guide more detailed experimentation. In the first stage the objective is to produce variations of base and compound kinetics at the scale of a part, with the objective being to select distinctive compound kinetic types. The number of kinetic types selected in this first stage will have a significant impact on the total number of animations produced, and hence the most distinctive of the various twist, roll and yaw combinations will be selected. For the second stage of the experiment, a matrix will be used to produce an indexing of kinetic type and control variable. Stage 3 is reliant on a close review of this methodically produced index of type to control. From observation of the impact type has on pattern formation, one type will be selected to allow concentration on control variables. By intuitive manipulation of combinations 101
of control variables, vague clustering may be thickened or gradient pattern disintegrated. This third stage accepts the (desirable) human factor in this experiment and warps the initial methodical objectivity into the realms of intuitive design speculation.
[S1] Compound kinetics The number of compound kinetic types can be reduced by considering the impact of orientation in three-dimensional space. The decision to select a fixed orthographic viewing position means it can be anticipated that one of the coordinate variations will be redundant. The facade study will be viewed from an orthographic camera view perpendicular to the x,y plane, which negates the detection of movement in the Z coordinate. When the duplicate combinations are removed, this results in 18 possible compound types, many of which, however, are likely to be very similar. Through observation and intuitive experimentation, only the most distinctive kinetic types will be selected.
[S2] Indexing control and kinetic The distinctive kinetic types selected from Stage 1 will be combined with the control variables. Using a matrix these are indexed against the full range of controls – synchronous, amplitude, period, noise, cellular automata (CA) and flocking.
Synchronous [script #1]
Linear [#2] Radial [#3]
Linear/inverse [#4/5] Radial/inverse [#6/7]
Arithmetic progression [#8] Geometric progression [#9] Position [#10] Sine [#11] Prime [#12]
Perlin [#13] Brownian [#14] Lattice [#15] Turbulence [#16]
Life-like [#17] Cyclic [#18]
Figure 6.3 Summative diagram of 19 control scripts used for Stage 2
Experiments with kinetic pattern
These six broad approaches to controls can be further developed in relation to the previous discussion. Amplitude-based controls consider the simultaneous activation of parts, but by differing amounts, i.e. the degree of force applied to each part can be differentiated. The part-to-part differentiation in amplitude can be controlled by proportional relationships based on arithmetic (AP) and geometric progressions (GP), of which two sub-types are identified, linear and radial. That is, the differentiation in amplitude will progress in a linear or radial spread across, up or at an oblique angle. Period-based controls are those where parts are activated at differing times. In a similar manner to amplitude, the part-to-part differential in activation can be based on AP or GP. In addition, other algorithmic relationships based on sine values and prime numbers are utilized, as well the calculation of period based on values summing X and Y coordinates. The fourth general type, noise, produces controlled variation using random values. The animation scripts are based on commonly used noise variants – Perlin, Brownian, lattice and turbulence.22 The final control type is based on cellular automata (CA) and flocking algorithms. Two distinct types of CA will be utilized for Stage 2 – life-like and cyclic. Completing the range of scripts is a flocking algorithm, based on the precedent of Reynolds. In summary, Figure 6.3 documents the 19 different control scripts that will be indexed against the selected kinetic types from Stage 1.
[S3] Intuitive control Stage 3 is reliant on a close review of the Stage 2 outcomes. Through the above discussion, and on the basis of observation of the pilot studies undertaken to determine the geometry and numeric scale, it would appear that control variables should have more influence than kinetic type. If this is confirmed by the Stage 2 animations, one kinetic type will be selected to enable the control variables to be explored in greater depth. The intent will be to explore the boundary conditions between typical outcomes from Stage 2, and to extend the possibilities indicated by one-off patterns. This will be undertaken by manipulating the values of the variables, and by combining different control scripts. Stage 3 involves a more intuitive, trial-by-error manipulation of script variables and values, with the deliberate aim of extending possibilities identified from the Stage 2 index.
All at sea A provisional taxonomy Taxonomy as heuristic device The objective of the experiments was to produce variety, and the three-stage method described in the previous chapter has resulted in the selection of over 200 distinctive patterns, selected from approximately 1200 animations. As noted in the design of the experiment, these outcomes are accepted as being incomplete. The variables used to generate the animations were instances along a continuum of possibility, and there is obviously a multitude of permutations. It is proposed, however, that the tactic of methodical indexing of variables and subsequent intuitive experimentation provide a useful snapshot of kinetic range. The challenge addressed in this chapter is the systematic analysis of this range in terms of difference by degree and kind. This requires an ordering mechanism, a taxonomy of some kind, which provides a basis for comparison. Given the emphasis of this research, a ‘classical’ form of taxonomy will be provisionally adopted. The term as used by Michel Foucault represents an approach which privileges external appearance over function. This taxonomy was constructed entirely on the basis of the four variables of description (forms, number, arrangement, magnitude), which could be scanned, as it were in one and the same movement, by language and by the eye.1 As is well exposed by Foucault, and in a more recent work by Bowker and Star, all approaches to classification have a strategic dimension, which valorizes some point of view and silences another.2 The emphasis of this design study is on morphology, deliberately privileging kinetic pattern over other technical, material or functional priorities. The working definition, developed in the earlier chapters, describes pattern in terms of differentiated clusters of similar moving parts or spatial propagation of similar kinetic resonance. A scientific approach to identifying clustering or propagation would likely involve the exact measure of ‘number, arrangement, magnitude’, but this would not likely suffice as a useful set of terms for design. What is required is an approach that accurately describes kinetic pattern, but one that is also evocative – something akin to Lye’s ‘figures of motion’ such as ‘flaring’ or ‘swaying’, which capture a particular trajectory and rhythm in a poetic, open-ended term.3 In the following
sections, a provisional terminology and rationale for locating pattern is developed from George Rickey’s Morphology of Movement. As discussed in earlier chapters, Rickey proposed the movement of a ship as a vocabulary of movement for kinetic art. Arguably, the motion of the sea surface is more appropriate for kinetic facades. Movement patterns are aggregates of individual moving parts and the descriptions used to describe sea patterns provide a rich source of terms. From these, six are chosen and quantitative thresholds are located in terms of spatial scale, relative proportion and orientation. These six classes, when combined as pairs, produce 15 hybrid patterns, providing a total nomenclature of 21. It is important to note that the development of this provisional taxonomy is a research strategy. It is used speculatively, as a heuristic device to undertake analysis and as a step towards the identification of pattern range. If the classification system proves robust, it may provide the basis for describing and conceiving a full range of movement patterns. Thus there are dual aspects to positing a provisional taxonomy: to use the system to enable methodical analysis of the animations, and to test the robustness of the nomenclature.
Overview of animations The following images document the three stages of the design experiments. Stage 1 was undertaken to identify the most distinctive compound kinetics that result from pairing the base kinetic transformations of translation, rotation and scaling. These produced similar compound types, of which four were selected – twist, roll, yaw and spring – as being the most distinctive. Twist, roll and yaw are well-known and precise types of compound transformations used in flight dynamics.4 Spring is a term introduced here to capture the distinctive quality of a compound translation that results from translation in the y axis combined with negative scaling in the x axis. These illustrations of singular kinetic types are followed by documentation of the Stage 2 animations, which were generated by indexing the three base and four compound kinetic types against 19 controls. The control algorithms are points along the variable continuum of reflexivity (from homeostasis to emergence). The animations are documented as groups of kinetic types, resulting in seven sets of 19 image sequences. Despite the different kinetic types, each of the sets displays a similar progression of patterns. The first image sequence of each group results from simple synchronous movement. Sequences 2–7 differentiate amplitude while sequences 8-11 differentiate period, each set of control scripts producing clearly identifiable and recurrent rhythms. These are followed by experiments with different types of noise algorithms (sequences 13–16) that introduce random motion within control limits. The final image sequences (17–19) correspond to the other extreme of the reflexivity continuum, where the kinetic is generated by cellular automata and flocking algorithms. As is apparent, despite the distinctive kinetic type of the part, there is a general consistency between the seven sets of image sequences. From this methodically generated set, Stage 3 involved more intuitive experimentation with one representative kinetic type (rotation) and alternate variable combinations. These progress from simple strategies of mixing amplitude to more 106
All at sea
Figure 7.1 Stage 1 animation study to determine singular and compound kinetics. The seven kinetic types selected for the experiment are annotated
Figure 7.2 Stage 2. TRANSLATION × 19 control types ranging from simple proportional movement in the top left to controls based on cellular automata and flocking algorithms in the bottom right
All at sea
Figure 7.3 Stage 2. ROTATION × 19 control types ranging from simple proportional movement in the top left to controls based on cellular automata and flocking algorithms in the bottom right
Figure 7.4 Stage 2. SCALING × 19 control types ranging from simple proportional movement in the top left to controls based on cellular automata and flocking algorithms in the bottom right
All at sea
Figure 7.5 Stage 2. TWIST × 19 control types ranging from simple proportional movement in the top left to controls based on cellular automata and flocking algorithms in the bottom right
Figure 7.6 Stage 2. ROLL × 19 control types ranging from simple proportional movement in the top left to controls based on cellular automata and flocking algorithms in the bottom right
All at sea
Figure 7.7 Stage 2. YAW × 19 control types ranging from simple proportional movement in the top left to controls based on cellular automata and flocking algorithms in the bottom right
Figure 7.8 Stage 2. SPRING × 19 control types ranging from simple proportional movement in the top left to controls based on cellular automata and flocking algorithms in the bottom right
All at sea
Figure 7.9 Stage 3-A. Selected animations from intuitive experimentation with one representative kinetic type (rotation) and various amplitudeand period-based arithmetic and geometric progressions
Figure 7.10 Stage 3-B. Selected animations from intuitive experimentation with one representative kinetic type (rotation) and various amplitudeand period-based arithmetic and geometric progressions
All at sea
Figure 7.11 Stage 3-C. Selected animations from intuitive experimentation with one representative kinetic type (rotation) and various periodicand noise-based scripts
Figure 7.12 Stage 3-D. Selected animations from intuitive experimentation with one representative kinetic type (rotation) and various periodicand noise-based scripts
All at sea
Figure 7.13 Stage 3-E. Selected animations from intuitive experimentation with one representative kinetic type (rotation) and various noise-, cellular automataand flocking-based scripts
complex experimentation with noise and CA control variables. Over a thousand animations were evaluated and the most distinctive are documented in Figures 7.9 to 7.13.
From ship to sea In 1963 George Rickey proposed the motion of a ship at sea encapsulates a full vocabulary of movement for kinetic art. As discussed previously, terms such as roll and pitch provide accurate descriptions of movement that also have an evocative resonance. In this sense, they meet the context of this design research and may provide a useful set of terms to undertake a discussion of the animations. However, there is an essential difference between the movement of a singular object and the patterns that result from multiple objects in motion. The set of terms proposed by Rickey succinctly capture the elegance of singular motion, or the subtle interplay between two interweaving parts. When a third, fourth or fifth part is introduced, the motion of each part may also be followed simultaneously. However, it would appear that the limits of human perception of simultaneous events occur between five and seven individual parts.5 Beyond seven, the tendency is to describe multiple objects by groups or to describe the whole as an overall pattern. Facades are typically conceived and described in this way, rather than as a collection of separate descriptions of individual parts. As discussed in Chapter 3, designers have typically designed in terms of part-to-whole relationships, as exemplified in the compound number ratios of Alberti, the dynamism of Guarani or the repetitive adjacency of Ledoux. Following on from traditional part-to-whole composition, the development of the free facade allowed designers to compose facades in terms of continuous patterns, independent of the structural frame. The patterns were based on the interweaving horizontal and vertical lines of cladding systems, often punctuated with asymmetrical zones. By layering external sunscreens, weathering systems, glazing and internal blinds, a range of facade patterns were able to be composed. Contemporary architects are developing this approach to produce what have been described as field effects.6 Kinetic facades might be considered a logical extension of this shift from part-towhole to field. A succinct and poetically evocative nomenclature, which captures breadth and nuance of this ephemeral form of design, is one of the overall aims of this book. Rickey’s vocabulary of motion, derived from the observations of a singular object, is inadequate in the case of architectural facades. If the vocabulary were to be extended, the sea, rather than the ship, might be considered a more apt reference for patterns of movement. A body of water, be it lake, sea or ocean, is a large-scale dynamic surface that is patterned with multiple surface events. While appearing monolithic when there is no tidal, wind or other force present, the sea rapidly transforms into a wide range of movement patterns when these forces come into play. Flat sea becomes undulating rhythms of swells, or sets of accelerating waves. Over time, multiple swells may form crossing patterns or break up into uneven choppy patches and whirlpools, or descend into the multiplicity of subtly interweaving ripples. It is proposed to define a taxonomy for kinetic pattern based on what will be termed a ‘sea nomenclature’. This will be used provisionally, as a way to undertake 120
All at sea
an analysis of the animations. Moreover, the nomenclature provides a structure to propose a full range of possibilities within the scope of the terminology, independent of the design of the animation experiments. That is, the theoretical number of pattern types that result from combining terms can be articulated in advance of the analysis. By utilizing the set of terms and their theoretical combinations, one analysis of the animations can be undertaken. This, in turn, enables an evaluation of the robustness of a taxonomy based on the surface of the sea.
Nomenclature When we recognize or identify something as a mouse, a token is ascribed to a type. In the process, we pass from the particular to the general. Only under these conditions can I use language and talk about a mouse. It has been seen that in the language of modern cognitive psychology this procedure is indicated (in a historically debatable way) as a phenomenon of categorization, and I have resigned myself pro bono communication is to go along with this usage.7 Umberto Eco, despite his misgiving, accepts that a shared set of terms is essential to language. Inevitably, this requires shared categories, a basis to determine one type from another (the mouse from the cat). In the spirit of the pragmatics of communication and as a speculative heuristic device, we need to delve into the murky business of categorization. It was proposed that, building on Rickey’s nautical analogy, the myriad of terms for describing the surface of the sea may provide the basis for a nomenclature for kinetic pattern. The Beaufort scale provides one systematic terminology to describe the sea surface, but its primary emphasis is on determining wind speed through observation of a range of phenomena.8 While there is no ‘ready-made’ taxonomy for describing the patterning of the sea surface, there are a number of terms in common use. These include: ripple, eddy, surge, swell, wave, roller, breaker, comber, dumper, chop, wash, current, peak, trough and face. Some would appear more apt than others. For example, dumper is a term used to describe a wave that ‘dumps’ on a swimmer or surfer and would appear to be too specifically associated with the experience of a beach wave. Among surfers, an even more specific set of terms is in use to describe both types of waves and the component parts. A wave has a face and lip that might be glassy or crumbly; the shape of the face may form a half-barrel, or form a tube (the ‘green room’). In contrast to this very specific anatomy of wave form, generic terms will be selected. For example, ‘ripple’ is a word in common usage that communicates in general terms multiple small-scale undulations. Given that the objective is to describe range, a set of terms that allow such general distinctions to be made between significantly different surface patterning is developed. In addition, given that the specificity of the exercise is to locate pattern within an existing set of animations, distinctions are articulated in terms of spatial scale, relative proportion and orientation. Below are the proposed terms and outline specifications that will be used to guide the analysis:
• • • • • •
swell: large-scale undulation or distortion; constant location or with a gradual incremental direction eddy: small-scale undulation or distortion; oscillating within a constant spatial zone; radial or linear wave: large-scale directional ridge of movement; defined front; moving in a linear or radial direction ripple: small-scale directional ridge of movement; defined front but with individual deviations in direction chop: multiple distortions occurring within a spatial zone; non-uniform multidirectional movement peak: isolated distortion with no apparent relation to neighbouring events; no linear or radial orientation.
The use of spatial scale, relative proportion and orientation to distinguish between types requires elaboration. A swell and an eddy, for the purposes of this study, have a similar structure but are differentiated by spatial scale. Typically a swell will be present as a singular spatial entity across the majority of the 21 × 19 hexagonal grid developed for the experiment. The same phenomena operating at smaller comparative scale and operating in less then approximately 25 per cent of the available grid would be termed an eddy. While a swell and an eddy are differentiated in terms of spatial scale, they have a similar non-directional orientation, typically undulating in the same location or moving in a gradual manner with incremental gradual changes in orientation. In addition, the width-to-height proportion (or the inverse) of a swell or eddy is within the range of 1:1 to approximately 1:4. Correspondingly, a wave or a ripple is differentiated from a swell or an eddy by a width-to-height proportion of greater than approximately 1:4. That is, they have a defined ridge rather than a general distortion. Moreover, they have a discernible orientation, with the ridge changing spatial position as a linear or radial front. The distinction between a wave and a ripple is made on a similar basis of spatial scale to that of a swell to an eddy. A wave will typically be present across the whole of the hexagonal grid, while a ripple will be located within approximately 25 per cent of the available area. The terms chop and peak refer to patterns that are the result of movement operating at a smaller scale than either of wave, ripple, swell or eddy. A peak is an isolated event, occurring either singularly or as a small group consisting of several parts. Once the event gets above this scale it is considered a ripple or an eddy. The components of a chop occur at a similar scale as a peak but are differentiated by granularity. There may be peaks occurring simultaneously, but if the proportion of peaks to the total area they cover is less than approximately 7:1, they are considered a number of individual peaks. Where the proportion goes beyond approximately 7:1, they become a chop. That is, a chop is a density of peaks co-located in an area. Both peaks and chops are either spatially stable or move in a non-uniform orientation.
All at sea
Theoretical number of hybrid patterns While there would appear to be animations that clearly meet these definitions, it is likely that there are animations that exhibit two patterns. Such patterns are termed here as hybrid. These may be present simultaneously but spatially differentiated, or the animation may transform over time from one characteristic pattern to another. Prior to an examination of the animations, the range of possible hybrids can be theorized. There are fifteen combinations. [swell] – [eddy/wave/ripple/chop/peak] [eddy] – [wave/ripple/chop/peak] [wave] – [ripple/chop/peak] [ripple] – [chop/peak] [chop] – [peak]
A first cut The primary objective of this initial attempt at locating the range is to use the terms to identify difference by kind and degree. Each of the Stage 2 and 3 animations has been reviewed and identified (where possible) as a swell, eddy, wave, ripple or peak. Where the animations appear to have dual characteristics (termed here as hybrids), a hyphenated description is utilized, e.g. swell-wave. A second objective is an evaluation of the suitability of the taxonomy to describe the full range of outcomes. To this end a ‘non-ascribable’ group collates those not adequately described by the six general types or by the fifteen hybrids. By applying the terms, a first cut of the animations can be undertaken. This enables a first analysis of the spread of pattern types and as important, tests the capacity of the taxonomy to describe kinetic pattern. From Figure 7.14 it can be observed that the total number of singular patterns is 109, with ripple and peak not being ascribed to any of the animations. The total number of hybrid patterns is 61. Not all theoretically possible combinations are evident. The missing patterns are confined to the four ripple combinations, where there are no examples, and the peak combinations, where there are two of a possible five examples evident. That is, of the fifteen possible hybrid patterns there are eight that have been identified in the sets of animations. In the subsequent sections, typical examples of the singular and hybrid patterns are presented and described, together with examples of the 27 non-ascribable animations that are not accommodated by the classification.
19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4
3 2 1
STAGE 2 HYBRIDS
SWELL EDDY WAVE RIPPLE CHOP PEAK
Figure 7.14 Illustration of range of patterns ascribed within sea pattern taxonomy
All at sea
Figure 7.15 SWELL pattern generated by a radial geometric progression
Typical singular patterns Approximately a third of the animations can be adequately described as a swell pattern. Two examples are presented. Figure 7.15 shows a swell pattern over the total area as a result of simultaneous movement of all parts, but with differential amplitude based on a radial displacement map. The movement type is yaw, a composite translation and rotation. The differential amplitude produces a radial ‘bulge’ that shifts from left to right, to produce an asymmetrical moving swell.
Figure 7.16 SWELL pattern generated by a linear geometric progression
This second example illustrates a swell pattern as a result of differential rotation based on two vertical displacement maps. The outcome is a shifting swell pattern left to right. Both Figure 7.16 and Figure 7.15 meet the requirements for the swell definition: large-scale undulation or distortion; either constant location or with a gradual incremental direction; coverage of more than 25 per cent of available area; proportional ratio of 1:4 or less.
Figure 7.17 EDDY pattern generated by a geometric progression
There was a single animation that correlated with the definition of an eddy pattern: small-scale undulation or distortion (area less then 25 per cent of available); oscillating or moving within a constant spatial zone. This resulted from the use of multiple displacement maps with abstract patterns overlapping over time. The outcome was shifting zones of movement, morphing from one zone to another, creating an undulating series of eddy patterns.
Figure 7.18 WAVE pattern generated by a sine equation
Animations that met the criteria for wave patterns were, after swell patterns, the second most common (37 animations). These were typically of two types. In the first, as illustrated in figure 7.18, a linear wave is captured progressing from left to right, as ‘columns’ of hexagonal parts sequentially undertaking a rolling movement (translation/perpendicular rotation). This produces a clearly defined linear ridge of movement, which in this case increases and decreases in speed due to the use of a sine equation.
All at sea
Figure 7.19 WAVE pattern generated by a radial displacement
A second set of wave patterns was those produced by ridges of movement extending in a radial direction. While these radial patterns are very different in appearance, they meet the criteria of being less than 25 per cent of the available area and having a proportion of greater then 1:4 (for radial wave patterns this is the proportion of the width of the ridge to the circumference).
Figure 7.20 CHOP pattern generated by a prime number sequence
Three of the animations were identified as being a chop pattern. One was a scaling movement controlled by a turbulence noise algorithm. The other two were the outcome of a number sequence controlling a twisting movement, as in the prime number example illustrated in Figure 7.20. These meet the definition for this type of pattern – zones of multiple non-uniform directional movement with no clear linear or radial orientation.
Figure 7.21 SWELL-EDDY pattern generated by a geometric progression
Typical hybrid patterns There were four animations that displayed hybrid swell and eddy patterns. As illustrated in Figure 7.21, these occurred as a transition state between overlapping swell patterns. As discussed in the definition of these patterns, an eddy is a swell pattern occurring within an area less than 25 per cent of the total available area. All the examples identified with this pattern had a radial orientation.
Figure 7.22 SWELL-WAVE pattern generated by a Perlin noise algorithm
As clearly shown in Figure 7.22, a linear ridge of rotation movement with a proportion above 1:4 has formed. In the second timelapse frame this ridge has moved from left to right and in the process expanded to cover more then 25 per cent of the available area. Moreover, the proportion has gone to approximately 1:3 with the linear ridge dissipating to an asymmetrical form.
All at sea
Figure 7.23 SWELL-CHOP pattern generated by geometric progressions
The largest number of hybrid animations identified were those exhibiting swell and chop patterns. In contrast to the previous swell-wave example, these patterns were typically co-present throughout the animation. In the example illustrated in Figure 7.23, there is an area of movement covering over 25 per cent of the available area. This is shifting incrementally from top to bottom with multiple non-uniform peaks of movement occurring at a density greater then 7:1.
Figure 7.24 SWELL-PEAK pattern generated by a life-like cellular automata
There was only one animation that exhibited hybrid swell and peak patterns. As illustrated in Figure 7.24, this was a roll movement controlled by life-like cellular automata. The result is an overall pattern of movement that meets the swell definition. This is interspersed with peaks of acceleration occurring randomly and with inconsistent acceleration or orientation.
Figure 7.25 WAVE-EDDY pattern generated by geometric progressions
There were two animations that exhibited hybrid wave and eddy patterns. As illustrated in Figure 7.25, these occurred when discrete areas of movement of less than 25 per cent of the available area transformed over time to form an identifiable ridge of movement. This meets the requirements for a wave to be an identifiable front with a proportion of approximately 1:4 or greater.
Figure 7.26 EDDY-CHOP pattern generated by geometric progressions
Eddy-chop was the second most common hybrid pattern, with 15 examples being identified. These typically occurred simultaneously, as illustrated in Figure 7.26. There are discrete areas of movement of an area of less than 25 per cent. This area of movement is ill-defined, with no clear orientation and with a proportion of less than 1:4. Each of these eddies is interspersed with multiple individual peaks occurring non-uniformly at a density of more than 1:7.
All at sea
Figure 7.27 WAVE -CHOP pattern generated by geometric progressions
One animation was considered to have wave and chop patterns occurring simultaneously. As illustrated in Figure 7.27, a linear movement pattern is progressing from left to right within a zone of less than 25 per cent and with a proportion of approximately 1:4. This is interspersed with clusters of peaks occurring at a density of 1:7 or more.
Figure 7.28 CHOP-PEAK pattern generated by a lattice noise algorithm
There were four animations that exhibited both chop and peak patterns. The differentiation between chop and peak is based on density, respectively above or below a density of 7:1. Figure 7.28 illustrates both clusters of high density and relatively isolated peak movements.
Figure 7.29 NON-ASCRIBED pattern generated by a Perlin noise algorithm
Typical non-ascribed animations Of the 197 animations, there are 27 that have not been ascribed a pattern within the sea pattern taxonomy. With two exceptions, these are all among those controlled with noise algorithms or cellular automata. A typical example of a noise animation that is unsubscribed is illustrated in Figure 7.29. The closest pattern description would be swell-chop, but this is complicated by what is best described as a ‘counter swell’, which produces a unique pattern. Figure 7.30 NON-ASCRIBED pattern generated by a day-night cellular automata
In the first frame of Figure 7.30 there are multiple areas of motion that could be described as an eddy-chop hybrid. However, as revealed by the second frame, a rapid change has occurred where former zones have disappeared or are in a transition state. This might be termed a ‘zonal state change’ and is a pattern that repeats in other examples of life-like CA that do not sit well within the pattern definitions.
All at sea
Figure 7.31 NON-ASCRIBED pattern generated by a life-like cellular automata
The second example of a non-ascribed animation generated by cellular automata rules is from what is known as a ‘demon’ rule set. In the two frames illustrated in Figure 7.31, it might appear that there are three or four eddy patterns overlaid with chop or peak patterns. However, in a similar manner to the previous example, these eddy zones are under continual revision. The patterns are continually reforming in a interweaving or ‘rippling’ pattern formation that is distinctive to this set of CA rules.
Figure 7.32 NON-ASCRIBED pattern generated by a flocking algorithm
The final example, Figure 7.32, is an outcome from the application of a flocking rule set. The first frame would appear to reveal two eddies formed in the two left hand corners, with the second frame suggesting these have combined to form a larger swell pattern. However, as in the two previous CA examples, there is a continual redefinition of the eddy or swell area. Typically, the swarming examples exhibit a mixture of the two previous examples: there are both large-scale shifts in zones of movement, and incremental interweaving along the edges.
State change A morphology of kinetic pattern Non-ascribable Writing at the height of activity in kinetic art, George Rickey proposed a vocabulary using various analogies, of which the motion of a ship at sea was most prominent. In the previous chapter this was extended to define a set of terms, by which the hundreds of animations generated to explore pattern range could be categorized. It was proposed that the sea surface, rather than the ship, is a more appropriate reference for designing kinetics for architectural facades. By selecting a range of terms and defining these quantitatively and qualitatively, categorization of the movement patterns evident in the animations was undertaken. Six terms were defined: swell, eddy, wave, ripple, chop and peak. These were developed on the basis of the capacity to describe a full range. From these, 15 hybrid patterns were theorized. For example, a swell-chop describes a pattern that evidenced two characteristics. In theory, this taxonomy may have enabled a way to locate the general characteristics of all the animations. However, despite this methodical approach, there were a significant number that could not be ascribed. The use of the sea pattern nomenclature was a provisional tactic, a first pass at describing the animations and theorizing the potential range of kinetic pattern. Clearly, there are deficiencies in the taxonomy, as there were a significant number of animations not ascribed within the range of classes. Despite the use of a rigorous approach where six basic types were defined, from which the theoretical number of hybrids could be generated, a significant number were considered to be outside the taxonomy. As discussed at the end of the previous chapter, the non-ascribable animations fell into three groups. The first was located in relation to those animations generated by noise algorithms, where what was described as a ‘counter movement’ occurred. Two patterns moving in opposite directions overlap, and rather than appearing as a hybrid, the two are distinct, creating a unique self-cancelling outcome. The other two typical non-ascribable outcomes occurred in relation to the series generated by cellular automata (CA) scripts. The variants of CA used produce patterns that are under continual reconfiguration: patterns form in seemingly random locations; their relative proportions shift; they may be relatively static; incrementally changing direction or dramatically re-orienting the direction of the pattern shift. However, on close analysis, two distinct transformations recur: life-like CA, as implemented in
this study, produce large-scale spatial shifts over rapid time periods; those based on cyclic CA were also continually reforming, but these occurred as an incremental interweaving along the edges of spatial zones; while the flocking script exhibited both these characteristics. When confronted with such a gap, one approach is to use the logic of the taxonomy to define new types to cater for the anomalies. As articulated in the previous chapter, the sea-pattern taxonomy was based on spatial scale, relative proportion and orientation. From this we could propose that three additional patterns would need to be defined. Following the logic of the previous definitions, where swell and eddy have a similar structure but are differentiated on their granularity, there could potentially be a variant for each.
Limits of classification Adding additional types to cover deficiencies in the original set of definitions may address the non-ascribable animations, but it raises a question on the limits of the approach for design. Classical taxonomy defines boundaries between objects, events or phenomena based on a reference to fine-grained and quantifiable observations. They are extendable, producing expanding trees of hybrids. Cases that do not meet existing definitions give rise to a new type or hybrid. This may be useful in circumstances where the population under study is relatively fixed, such as fauna or flora. The approach will eventually describe all found examples and, once established, the taxonomy will be stable within typical time frames. New species are unlikely to mutate to produce new forms that challenge the system. Taxonomy is less stable when describing temporal phenomena. For example, classical music, before the invention of the metronome, in effect developed a taxonomy that communicated the speed and rhythm by which a composition should be performed.1 Terms such as adagio and presto communicated tempo, but also conveyed the character of the music. These terms are meaningful within the relatively fixed formal structures of the music of the period. However, they no longer capture the diversity of musical forms in contemporary culture. New forms of music have developed that are outside the conventions of the taxonomy. We can posit that taxonomy based on fixed definitions may be useful for such fields as botany, but are less useful for arts such as music, where new forms are continually being designed. Kinetic facades allow experimentation with a new form of composition and the animations undertaken here are evident of a rich array of possibilities. While these have been carefully conceived, to deliberately explore as wide a range of outcomes as possible, others are likely to be added. The taxonomy could, of course, be extended to describe additional forms, but where would this end? The essence of kinetic facades is the capacity for change. The nonascribable animations, identified in the first attempt at classification, represent one pole of this flux. They are in a constant state of change, reforming into a seemingly endless array of movement patterns. A taxonomy based on a rationalization of the sea nomenclature has led to fixed boundaries, identified by relative proportion or granularity expressed as ratios and percentages. This has been useful to quantify the range of animations, but the approach is inflexible and will lead to an escalating 136
set of definitions to cope with the inherent dynamism of kinetic facades. What is being designed is not a fixed object or a static relief, which can be allocated a fixed place in a taxonomy. The approach used to analyze the animations has enabled some understanding of the potential of kinetic design, but an approach based on numerous distinct types and hybrids would appear to be flawed.
Taking stock The focus of this research has been on kinetic pattern, defined as the relative movement of multiple individual kinetic events in time and space. That is, the characteristic way in which singular events cluster or propagate across a facade over time. Through insight gleaned from the kinetic arts, the distinctive challenge of kinetic facades was located – the design of ‘movement itself’. Using the precedent of Rickey’s vocabulary of movement for kinetic art, and extending it to movement pattern has been, to a degree, productive. The rationale for the taxonomy derived from characteristic sea patterns was to identify an approach by which the space of this new composition could be explored. While the taxonomy used a simple rationale for describing kinetics in terms of spatial scale, relative proportion and orientation, the approach has led to a proliferation of types and hybrids. What is needed to progress this new field of architectural composition is a more robust approach, which maps out range in relation to a limited number of clearly communicable and recognizable pattern types. The ambition is for an engaging nomenclature that maps out the basic contours of this new design space – a set of terms that provide some structure, yet are suitably abstract, latent with possibility and applicable to a range of tectonics. Designers have developed various ‘languages’ of composition for architectural facades. Is there precedent within this legacy that can be used to inform a more strategic, design-orientated approach to describing kinetic pattern? The following sections re-examine sources identified in the previous chapters, to look for an alternative to a nomenclature based on fine-grained classifications. How might the range of formal structures, or what are termed here ‘movement patterns’, be expressed in a robust manner that encourages creative practice?
Formal systems Approaches to composition, from the Renaissance to the turn of the twentieth century, were reviewed earlier through the work of Emil Kaufmann. According to his view of history, facade composition up to the early twentieth century can be considered as three part-to-whole systems: proportional number relationships, intended to create a harmony between part and whole; hierarchical differentiation between parts creating tension and dissonance; independence between parts achieved through non-hierarchical repetition and scaling. Can this approach be adapted to provide a simple and robust way of conceiving and critiquing kinetic facades? Proportional relationships were included in the control scripts used to generate the animations. Some classical proportional approaches, such as the golden section and Fibonacci series, were adapted to trigger a range of kinetic patterns. Conceivably, these animations could be analyzed in Kaufmann’s terms. The proportionally generated scripts typically produced rhythmic wave patterns that, when overlapped, produced tension 137
and dissonance between parts. When repetitive short sequences were used, outcomes similar to Kaufmann’s third system – non-hierarchical repetition and scaling – were evident. However, there were many animations beyond these. For those animations based on erratic arithmetic or geometric progressions relationships, or those produced through noise or cellular automata algorithms, the three systems would seem inadequate. There is no harmonious relationship between parts and whole, and, while there is certainly tension and dissonance, this is not based on a hierarchical structure. Kaufmann’s third category – independence between parts based on non-hierarchical relationships – would seem to be applicable. However, while the movement patterns of many of the animations exhibit these characteristics, they are not based on the relatively simple and clearly defined boundaries which Kaufmann describes. Many of the kinetic patterns produced were under continual reconfiguration, with interweaving boundary conditions, irregular points of activity, and amorphous transitions in pattern formation. While direct adoption of Kaufmann’s three systems is not appropriate, it does suggest the value of a simple, formal structure. His concentration on abstract structures, such as the organizing lines and number theory of Alberti or the repetition and non-proportional scaling of Ledoux, allows recognition of general approaches over and above stylistic detail. His three formal systems allow comparison between designs and a multiplicity of variation within the three part-to-whole systems. A similarly coherent and adaptable structure would be a useful reference for the design and critique of kinetic pattern.
Modern surface Kaufmann’s historical formalism fell out of favour during the development of twentieth-century modernism. For theorists such as Giedion and Banham, the emphasis was on the spatial interplay between exterior and interior, in contrast to the external emphasis of nineteenth-century historicism. The concentration on the appearance of the facade in terms of formal strategies of pictorial composition, was supplanted by volumetric analysis. For many twentieth-century designers there was a general rejection of part-to-whole composition in favour of structural or functional expressionism. From such a viewpoint, walls were subordinate to volumes and to the sculptural play of light on smooth surface. This early period of modernism does not appear to offer much in terms of a formal system for conceiving pattern range. More recent views of the modern movement offer alternative analysis, with historians such as David Leatherbarrow tracing technological innovations and the impact these had on compositional outcomes. The development of the free facade, in particular, is given close attention by Leatherbarrow. As discussed in Chapter 3, his description of designs as ‘repetitive unbroken coverings or continuous wrapping’ potentially adds a fourth system to Kaufmann’s analysis.2 Harmonious, hierarchical or repetitive part-to-whole relationships are superseded by the continuity of the free facade. The overlaying of multiple horizontal and vertical grids, typical of curtain wall composition, might be considered an additional ‘system’, complimenting the three proposed by Kaufmann. However, while this might add a fourth category to Kaufmann’s framework, it does not address the irregularity and dynamism of a large number of the animations under 138
review. In the non-ascribable examples identified in the previous chapter, there are fleeting moments where parts and wholes coalesce, or where there is repetitive interweaving. But these designs fragment, reform and atomize, generating patterns outside either the three part-to-whole systems or the layering of textures enabled by the free facade.
Contemporary facades Contemporary approaches to describing and conceiving the formal appearance of facades are best described as pluralist. The vestiges of traditional part-to-whole composition, explicit in the neo-rationalism and post-modernism of the late twentieth century, continue in regional pockets. For some architects, modernist strategies still suffice: transparency and superimposition are being revisited with new materials, ‘hi-tech’ articulation of joints, the layering of environmental screen; opaque, transparent and patterned glazing systems, and internal blinds.3 For others, the play of light on sculptural form still fascinates.4 And the expression of internal function on the exterior of the building continues, in an age where adaptable vacuous interiors are required by the property market. Awareness of this contemporary functional ubiquity and the proliferation of large internalized complexes, such as shopping centres and business parks, are cited in the introduction to Functional Ornament.5 The authors document a contemporary emphasis on external appearance, rather than the modernist interplay of internal space and external constructional logic. The book provides analysis of contemporary facades exhibiting this trend, organized according to what they term ‘affect’. Despite the promise, the differentiation between different types of affect does not appear illuminating for the purposes of this inquiry. The reader is presented with a collection of analytical technical drawings, annotated with descriptions of affect organized in a smorgasbord of construction, programme and pattern. The logic is one of material and resultant affect. However, the definition of material includes programme, construction, light, cladding, pattern, branding, image, colour and reflection. The list of affects is similarly challenging. Thirty-nine terms are used, with ‘differentiated’ being the sole repetition in discussing the case studies: geometric terms such as spiral and vertical are used alongside terms usually associated with materials; general terms, such as complex and diverse, are followed by specific terminology such as tartan and quilted. While this pluralism has its merits and is reflective of contemporary design, it does not provide an adaptable model for this study of kinetic morphology.
Fields revisited The section on ‘field thinking’ undertaken in Chapter 3 has been productively used to underpin the decision plane framework on which the animations were structured. However, Sanford Kwinter takes a position that would appear to be an antithesis of formal classification. His analysis of Italian Futurism as dynamic arrangements of forces, and subsequent research on non-linear systems, such as Poincaré’s mathematics, does not align with classification based on definitive classes. Kwinter provided the foreword to The Atlas of Novel Techniques, in which contemporary designers Reiser and Umemoto outlined some compositional tactics. Of particular 139
relevance for the questions raised here is the idea of composition based on wholewhole relationships. Developed as a critique of a reductive use of typology, which echoes Kwinter’s essay on ‘true formalism’,6 they propose that emergent organizations become legible not as parts to a whole, but as ‘whole-whole relationships’. Following on from this critique, two design strategies are outlined as the means to generate qualitative change. After first rejecting the juxtapositional techniques of collage as accumulations of the merely different, we posit either an unchanging unit deployed along a variable trajectory or the simple repetition of a variable unit. In both cases transformation is a quality perceived through deployment in quantity.7 This approach to design involves a strategy of composition based on repetitive and interconnected parts, each minimally transformed, to produce graduations or intensities within a continuous whole. This strategy has strong resonance with the non-ascribable outcomes from the animation studies. Graduation and intensity within a continuity could be a description of the quality of many of the cellular automata transformation patterns. Moreover, when subjecting the range of their design output to analysis, Reiser and Umemoto address a similar problem to that considered here. How might they describe their designs and how might description be used productively to extend design range? Their response is to return to the architectural concept of typology, emphasizing how type has a use beyond classification. Typology plays a significant role within material practice. It allows for a clear selection of architectural organization from among the almost limitless possibilities available today … Typology is not only useful as a form of classifying something at the end of a process but also as a crude device for use in the design process.8 An overview of their identification of types, and the way in which this is used to conceive a range of variation in design, may be insightful for describing and conceiving kinetic pattern. Care must be taken, however, as there are explicit differences between their design focus and the ambitions here. Reiser and Umemoto are architects who typically work by conceiving design as volumes, in contrast to the focus here on external skins. Moreover, although they use dynamic systems in their design process, they appear to have no interest in kinetics as an outcome. The types they identify for their practice are planar, linear and punctual – abstractions that reference constructional systems. By way of illustration of these types, each is represented as a geometric unit that has been repeated along the trajectory of a doubly curved path. When conceiving a design, a type is chosen and variations are developed, based on the transformative possibilities of the planar, linear or punctual unit. Choice of type locates a set of parameters, which can be manipulated to produce variation. The three types are geometric abstractions that enable variation, within the constraints of the constructional system they reference. Tellingly, there 140
are no examples of mixed types or transformation from one to another. This reveals an essential difference between such parametric design practice and the opportunities suggested by kinetics. Reiser and Umemoto consider their typology in terms of an abstract geometric reference to a constructional system. There is no attempt to mix geometries and thus go outside the boundaries of the type. Kinetic facades, by contrast, offer the opportunity for such transformation. As evidenced by the hybrid animations from the study, the potential for range is as much in the transformations between types, as in the large degree of variation within type definitions. This suggests that a framework for describing range would, in the case of kinetic facades, need to embrace multiplicity: difference by kind, as well as difference by degree.
Clouds versus geometry Side by side with Reiser and Umemoto’s illustration of their geometric types are photographs of the three primary cloud formations – cumulus, stratus and cirrus.9 There is no reference to these images in the text. It would seem that the amorphous quality of clouds is intended as a metaphor for the variability of the geometric types. However, while there is synergy with the curvilinear geometry favored by the designers, there is an important difference – the taxonomy of clouds includes intermediate and compound types in addition to the three primary formations. The original paper by the inventor of the taxonomy, Luke Howard, was titled On the Modification of Clouds.10 Howard recognized in his title what we observe daily – that there are minimal periods where the pattern of clouds remains fixed in location or form. They are in a constant process of what Howard calls modification. Can the idea of modification be adapted for kinetic pattern? If a simple system can describe the infinite modifications of clouds, a similar approach might be possible for considering the dynamism of this new form of design composition.
Figure 8.1 Cloud formation photographed by the author; Metcalfe, Australia 2009 141
The naming of clouds The eighteenth century had been the great age of naming and fixing, with Linnaeus and Samuel Johnson enthroned as its ruling and rule giving lexicographers. Both the language of nature and the nature of language were being drawn up and standardized for the benefit of compatible usage. But the naming of clouds was a different kind of gesture. Here was the naming not of a solid, stable thing but of a series of self-cancelling evanescence.11 Describing and explaining the formation of weather patterns and of clouds, their most striking visual manifestation, have preoccupied cultures throughout history. In the West, by the sixth century BC, nephology had been established as a field of enquiry in Socratic Greece.12 Perhaps the most famous example is Aristotle’s Meteorologica, where he challenged the Platonic view that clouds were imperfect, changeable forms and hence merited little attention.13 Aristotle used the example of clouds to illustrate his argument for the world as being in a constant state of flux, with his deduction that clouds were the result of a sub-lunar void where air and water combine. Such a view held sway until the sixteenth century. From the Enlightenment onwards, attempts were made to classify the formation of clouds, all of which failed to reach consensus. Robert Hooke, a founder of the Royal Society, and the early French scientist Lamarck each produced a proliferation of terms that attempted to describe a full range of cloud formations.14 These attempts at classification failed due to an ill-defined set of terms and the futility of trying to describe the large number of possible cloud formations. Howard’s approach was to describe cloud patterns in terms of three basic forms that were under continual modification. According to Hamblyn, the breakthrough was this emphasis on modification: forms were under constant change but their modification could be described in relation to three basic forms. Moreover, Howard recognized that, not only was there variation within the three modifications, but also that there were a number of characteristic transitional states between them: ‘Clouds could change both generically and specifically by effecting transitions between their forms. They could pass, one into the other, not only between individual modifications but between entire families of forms.’15 Howard provided a full set of names together with definitions that described typical formations: the three simple modifications, cirrus, cumulus, stratus; the two intermediate modifications between these, cirrocumulus and cirrostratus; and two compound modifications, the cumulostratus and the cumulocirrostratus, which was given the shortened term of nimbus. Howard’s emphasis on what he termed modifications and the recognition that, in the case of clouds, there could be transition within and between states, was seminal.16 The taxonomy allowed for transitions between what he termed the three simple types, as well as internal difference within these. An approach based on articulating clearly definable boundaries between things had been superseded to provide a simple nomenclature for the dynamics of cloud formation.
Towards state change From outside architecture, the elegance of Howard’s taxonomy of clouds to describe range provides the scaffold to propose an approach for describing patterns for kinetic facades. The taxonomy of clouds allows for internal variation and transition across three types. For Howard there is scope for internal variation but also the articulation of transformations between the simple types, which he terms intermediate and compound modifications. The challenge of conceiving a robust set of terms for describing the range of kinetic facade patterns is that they offer a similar transitional capacity to that of clouds. That is, kinetic patterns can pass from one state to another. Movement patterns are formed by multiple singular movements. However, as evidenced by the first stages of the animation experiments, pattern is independent of the kinetics of the part. The dominant influence on pattern was found to be the control script. Regardless of whether the singular kinetic was translation, rotation or scaling, there was strong correlation between control and pattern. These animations were dominated by what were termed waves and swells, with a number of in-between hybrids. The objective of the animation studies was to map a wide range of patterns, and to this end a deliberate strategy of generating more hybrids was undertaken by intuitive experimentation with the animation scripts. Distinctive types of hybrids became evident, in a similar manner to the distinctive intermediate and compound modifications accommodated by Howard’s cloud taxonomy. Over time, a facade may stay within the bounds of a singular pattern or be in transition between patterns. It is proposed that an approach for describing movement patterns for kinetic facades can be developed along similar lines to Howard’s taxonomy of clouds. The phrase that is used to articulate the taxonomy of kinetic pattern is state change.
Three states For the purposes of describing movement patterns for kinetic facades, state is used as an alternative word for type. State has the inherent connotation of a dynamic condition, open to change, and hence is considered more appropriate for describing
Figure 8.2 State change, identified by distinctive shape and dynamic
linear or radial
constant non-uniform intertwining/ expansive
and conceiving kinetics. Its adoption makes explicit that movement patterns are snapshots of form in motion. That state change is the distinguishing feature of this fledgling design practice. In developing state change, two issues need to be addressed. On what basis is state defined, and how many will capture the range of movement patterns? As illustrated in Figure 8.2, it is proposed that kinetic pattern can be distinguished by a characteristic spatial form or shape, and also in terms of temporal behaviour or dynamic. Kinetic shape and dynamic enable the identification of three states that can be used, in a similar manner to Howard’s three cloud types, to describe the full range of kinetic pattern. In terms of a nomenclature, wave is carried over from the previous classification approach, as it clearly describes a significant number of animations. The second term is fold, which describes a particular state of re-forming or, to use Howard’s term, modification, clearly distinguishable from a wave. The third state name is field, which appropriately describes a set of animations that present as a continuous aggregate of non-uniform movement. States are differentiated on the basis of the geometric shape of the movement pattern and the characteristic manner in which the pattern forms and changes, which is termed here as the state dynamic. Wave, fold and field are equivalent to Howard’s three simple modifications. They are generic terms that describe recurring movement patterns. While there are many variations of the three states, clear distinctions can be made on the basis of difference in shape and dynamic, as described and illustrated in the following sections.
State shape Shape describes the geometric area that is in motion. That is, a shape is a spatially contiguous number of facade parts that move simultaneously as an identifiable group. Wave, fold and field can be distinguished by characteristic shapes, respectively ridge, patch and fragment. A ridge is an elongated shape, with width to length proportions typically above 1:4. Ridges may be linear or curvilinear, have a constant proportion, or be tapered. They may occur singularly, or as sets of ridges that are moving in a similar direction. The shape of the ridge may undergo small amounts of change, but typically the shape is consistent over time. In contrast to a wave, the shape corresponding to a fold is less elongated, with width to length proportions typically below 1:4. The area of movement of a fold is typically an irregular geometry and is termed here as a patch. When compared to a ridge, the boundaries of a patch are dynamic, subject to constant change in geometry, or expansion/contraction. A patch may occur as one entity within a facade, or as a number of patches of intertwining movement. In contrast to either a wave or fold, which are geometric areas, a field movement pattern consists of multiple ‘point’ shapes, or what are termed fragments. A simple field pattern would typically be one contiguous area of irregular moving fragments.
State dynamic The second criterion for distinguishing state is the movement pattern dynamic – the manner in which the shape is changing spatial location relative to the overall facade. The dynamic of a wave is a consistent linear or curvilinear movement. Typically, the 144
Simple state examples
Figure 8.3 WAVE STATE. The typical simple wave state is characterized by a linear or curvilinear ridge of movement with a uniform and consistent dynamic. The example illustrates the case of a ridge with a regular diagonal dynamic from bottom left to top right
Figure 8.4 FOLD STATE. The typical simple fold state is characterized by adjacent patches of movement, with a constant reconfiguration of boundaries that produces a typical interweaving or expansion/ contraction
Figure 8.5 FIELD STATE. The typical simple field state is characterized by fragmented movement of singular units, or small groups. The dynamic is inconsistent, irregular and multidirectional. The fragments of movement, as captured by the time-lapse image, are highlighted 145
Figure 8.6 STATE CHANGE. Illustration of kinetic pattern as a dynamic morphology where there are three simple states of wave, fold and field and typical intermediate state transitions – swell/ stratify, aggregate/ disintegrate, atomize/ribbon. The compound state turbulence occurs when all simple states are present 146
dynamic of a wave ridge will be at a discernible velocity in a constant orientation. A special case is a standing wave, where the parts making up the ridge shape are simultaneously moving to produce the characteristic ridge shape, but the spatial location remains constant. In contrast to the uniform linear or curvilinear dynamic of a wave, a fold dynamic is typically an interweaving movement along edges, or an expansion and contraction between boundaries. Rather than being a clearly distinguishable movement across a facade, as is the typical case for a wave, the dynamic of a fold occurs in a relatively constant location. Occasionally, the characteristic folding and interweaving dynamic may rapidly expand or contract across a facade, but this is typically incremental and non-linear. A fold may appear within a similar time frame as a wave, appearing then dissipating, but is typically a longer-term pattern. The dynamic of a field state is one of irregular multidirectional movement of the individual fragments. There may be a general overall direction of the fragments of the whole, but typically this is disturbed by individual directional change, or change in velocity of individual fragments. The dynamic of a field is one of constant minor variation punctuated by inconsistent behaviour of singular or small groups of fragments.
Intermediate and compound states As illustrated in Figure 8.6, the simple states are supplemented by six intermediate and one compound state. The combination of features may be relatively stable over time, or these may be constantly reforming as the pattern undergoes state change. As annotated on the diagram, change from one simple state to another will have a characteristic shape and dynamic. Transition from a wave to a fold will typically involve a swelling of the wave shape, while the reciprocal state change from fold to wave typically involves a stratifying of the fold shape. As a wave changes state from a wave to a field there is a distinctive atomizing of the ridge shape. The reverse, from field to wave, produces a ribboning shape as fragments form into strands of similar movement. A field to fold state change involves a similar process of aggregation, but the forming shapes are less elongated and develop in a slow interweaving pattern. The last of the intermediate state changes, from fold to field, involves the disintegration of folds into smaller and smaller shapes and a shift to irregular and non-uniform movement. The characteristics of the compound state turbulence is atypical, with the three states present to varying degrees.
From theory to practice During the writing of this book, interest generated by the Aegis Hyposurface resulted in an ever-expanding range of kinetic prototypes. Ned Kahn’s wind walls have proliferated, and examples such as the dynamic screens implemented by Kiefer Technic demonstrate the viability of individually controlled environmental systems. New forms of kinetics have emerged, such as the water pavilion at Zaragoza. These developments reinforce the intuition that set the project trajectory; that there is an opportunity to develop a poetic approach to kinetics, which embraces the capacity of state change. 147
Intermediate state examples
Figure 8.7 WAVE-FIELD. The intermediate state between a wave and a field typically has a balance of a ridge shapes and pockets of small irregular movement. The example illustrates the dissipation of a wave ridge and atomization along the edges
Figure 8.8 WAVE-FOLD. As was regularly evidenced in the experiments, the intermediate state between a wave and a fold is typically a swelling of a wave ridge shape. The case illustrated shows the intersection of two wave ridges and the forming of two adjacent patches of movement typical of a fold state
Figure 8.9 FOLD-FIELD. The state between a fold and a field is the most complex of the intermediate states. The example illustrates a movement pattern based on a flocking algorithm, where field fragments are forming into curvilinear patches 148
The morphology of kinetic pattern articulated in this final chapter provides a theoretical basis for progressing design activity in this distinct field. It is hoped that the concepts proposed here stimulate further exploration within the pragmatics of technology and materiality.17 The animations used to develop the morphology are not all inclusive, but they map a significant number of design possibilities. Moreover, the approach is based on the proposition of three simple states and a theoretically infinite number of state transitions. That is, it provides a robust structure, but one which is extendable to enable further design research. The individual requirements of programme and site will inform the degree to which the morphology can be adapted and developed. Clearly, not all design briefs will have composition to the forefront. Nonetheless, it is anticipated that any implementation of kinetics may consider the liquid potential of kinetics, and, in so doing, aspire to a poetry of movement.
Chapter 1 1 See Steadman, P. Architectural Morphology: An Introduction to the Geometry of Building Plans. London: Pion. 1983. 2 See Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; vol. 22. 3 The origin of morphology is accredited to Johann Goethe, who according to Sharpe made a clear distinction in zoology between function and form. See Sharpe, L. The Cambridge Companion to Goethe. Cambridge: Cambridge University Press. 2002. p. 168. 4 Michael Batty describes in 1999 a shift in emphasis for research in urban morphology. ‘Current approaches are still largely focused upon the representation of static structures, based on measuring morphologies at one cross section in time and determining important relationships such as those associated with accessibility. In other areas of urban modelling, a massive shift has taken place as we have realized that the only way to satisfactorily explain outcomes – structures – is through the processes that give rise to them.’ Batty, M. ‘A Research Programme for Urban Morphology’. Environment and Planning B: Planning and Design. 1999; vol. 26: 475. 5 See Steadman, P. Architectural Morphology: An Introduction to the Geometry of Building Plans. London: Pion. 1983. Preface. 6 Steadman, P. Architectural Morphology: An Introduction to the Geometry of Building Plans. London: Pion. 1983. p. 11. 7 Steadman, P. Architectural Morphology: An Introduction to the Geometry of Building Plans. London: Pion. 1983. p. 61. 8 Steadman, P. Architectural Morphology: An Introduction to the Geometry of Building Plans. London: Pion. 1983. pp. 6–12. 9 See Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; vol. 22: 230. 10 See Steadman, P. Architectural Morphology: An Introduction to the Geometry of Building Plans. London: Pion. 1983. p. 248. 11 See Tschumi, B. Architecture and Disjunction. Cambridge, MA: MIT Press. 1994. 12 For a seminal overview of the development of the architecture in relation to the movement of the surveyor, see Bois, Y.-A. and Shepley, J. ‘A Picturesque Stroll around Clara-Clara’. October. 1984; vol. 29. 13 The play of light and shadow on architectural surface and through the reflection and refraction of glass has been manipulated in architecture from the Middle Ages onwards. See Marks, R. Stained Glass in England During the Middle Ages. London: Routledge. 1993. 14 See Mostafavi M. and Leatherbarrow D. On Weathering. Boston, MA: MIT Press. 1993. 15 The status of Mendelsohn’s design has been subject to debate, with Kenneth Frampton linking the project to expressionism. See Frampton, K. Modern Architecture: A Critical History. London: Oxford University Press. 1981. p. 120. Reyner Banham considered Mendelsohn was representative of Italian Futurism in terms of a technological agenda, but Whiteley argues this was more symbolic
then actual. See Whiteley, N. Reyner Banham: Historian of the Immediate Future. Boston, MA: MIT Press. 2003. p. 47. 16 See Terzidis, K. Expressive Form. New York: Spon Press. 2003. p. 43. 17 Terzidis, K. Expressive Form. New York: Spon Press. 2003. pp. 33–43. 18 For a review of contemporary media facades, see Ag4: Media Facades 2000–2006. Cologne: Daab. 2006. 19 The trajectory of the facade from load-bearing construction to curtain wall is examined in Leatherbarrow, D. and Mostafavi, M. Surface Architecture. Cambridge, MA: MIT Press. 2002. 20 See Wigginton, M. and Harris, J. Intelligent Skins. Oxford: Butterworth-Heinemann. 2002. 21 See Neumeyer, F. ‘Head First through the Wall: An Approach to the Non-Word Facade’. Journal of Architecture. 1999; vol. 4: 257. 22 For a thorough documentation of revolving structure, see Randl, C. Revolving Architecture: A History of Buildings That Rotate, Swivel, and Pivot. New York: Princeton Architectural Press. 2008. 23 For a recent discussion of Price and the Fun Palace, see Anstey, T. ‘Where Is the Project? Cedric Price on Architectural Action’, in Rendell (ed.) Critical Architecture. London: Routledge. 2007. 24 For a recent review of intelligent rooms in architecture see Maher, M. L., Merrick, K. and Saunders, R. ‘From Passive to Proactive Design Elements’, in Dong, A., Vande Moere, A. and Gero, J. S. (eds) CAAD Futures. Sydney: Springer. 2007. For a general review of responsive spaces and Human Computer Interaction, see Anshuman, S. and Kumar, B. ‘Architecture and HCI: A Review of Trends Towards an Integrative Approach to Designing Responsive Space’. International Journal of IT in Architecture, Engineering and Construction. 2004; vol. 2. 25 Catherine Ingraham discusses what she describes as ‘the lament for an architecture-of-motion’ in book 10 of Vitruvius. See Ingraham, C. Architecture and the Burdens of Linearity. New Haven: Yale University Press. 1998. pp. 133–42. 26 See Anshuman, S. ‘Responsiveness and Social Expression: Seeking Human Embodiment in Intelligent Façades’. ACADIA 05. Savannah (Georgia). 2005. 27 ’Movement itself’ is a term that originates from the 1920 ‘Realistic Manifesto’ by Naum Gabo and his brother Antoine Pevsner. In a critique of Italian Futurism, the term was used to distinguish the act of movement from the representation of movement as multiple superimposed frames. It subsequently was utilized by kinetic art theorists such as Frank Popper and George Rickey. This legacy will be examined in Chapter 4, where it informs a definition of kinetic pattern for architectural facades. See Gabo, N. and Pevsner, A. ‘The Realistic Manifesto’, in Bowlt, J. (ed.) Russian Art of the Avant Garde: Theory and Criticism, 1902–1934. New York: Thames and Hudson. 1920. 28 See Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; vol. 22: 225.
Chapter 2 1 See Fox, M. A. and Kemp, M. Interactive Architecture. New York: Princeton Architectural Press. 2009. p. 46. 2 See Moussavi, F. and Kubo, M. The Function of Ornament. New York: Actar. 2006. p. 9. The authors utilize a classification by depth of facade and sub-categories based on a loose definition of materiality that includes lighting and branding. The classes are further complicated by overlaying a third category effect. The first category depth is useful and is adopted here. 3 Michael Fox developed a taxonomy of control systems in conjunction with B. Yeh while teaching at MIT. See Fox, M. A. and Yeh, B. P. ‘Intelligent Kinetic Systems in Architecture’, in Nixon, Lacey and Dobson (eds) Managing Interactions in Smart Environments: First International Workshop. London: Springer. 2000. 4 See Güçyeter, B. ‘A Comparative Examination of Structural Characteristics of Retractable Structures’, MSc thesis, Dokuz Eylül University. 2004. 152
5 See Korkmaz, K. ‘An Analytical Study of the Design Potentials in Kinetic Architecture’. PhD thesis, İzmir Institute of Technology. 2004. 6 The emphasis of the monograph on Shigeru Ban is on his unique use of materials. See Ban, S., Ambasz, E., Bell, E. and Wood, D. Shigeru Ban. Princeton, NJ: Princeton University Press. 2001. The theme of materiality also dominates critical reviews. See, for example, Webb, M. ‘Tradition Stood on End’. Architectural Review. 2005; February 1: 82–5. 7 See Oosterhuis, K. ‘A New Kind of Building’. 2005. Available online at http://www.haecceityinc. com/homepage.html. Transcript of a lecture originally titled ‘Programming the Point Cloud’ presented at the Royal College of Art, London, 2005. 8 See Oosterhuis, K. and Biloria, N. ‘Interactions with Proactive Architectural Spaces: The Muscle Projects’. Communications of the ACM-Organic user Interfaces. 2008; vol. 51. 9 See Craig, D. J. ‘The Future Tents: Kinetic Sculptor Chuck Hoberman Expands the Boundaries of Design’. Columbia Magazine. 2006; Spring. 10 This project was supervised by Michael A. Fox in a design studio undertaken at the California Polytechnic State University and can be accessed at http://www.mafox.net/ 11 For an overview of projects that incorporate kinetic sunscreens in architectural design, see Baird, G. The Architectural Expression of Environmental Control Systems. London: Taylor and Francis. 2001. 12 The project uses a commercial system, Girasol. Technical specification available from http://www. coltinfo.co.uk/products-and-systems/list-of-publications/ 13 The project for a kinetic facade with panels with three degrees of rotation was originally sourced online. Some documentation of the competition is available at http://www.faenza.com/Concorsi/ Verbale.htm 14 By ‘temporal operations’, Leatherbarrow and Mostafavi refer to the development of the window from a purely visual function ‘an instrument of seeing, or establishing the look of a building’ to that in which ‘it is also an instrument of adjustment’ not only an eye but also a hand. Leatherbarrow, D. and Mostafavi, M. Surface Architecture. Cambridge, MA: MIT Press. 2002. p. 62. 15 See Leatherbarrow, D. ‘Architecture’s Unscripted Performance’, in Kolarevic and Malkawi (eds) Performance Architecture: Beyond Instrumentality. New York: Spon Press. 2005. p. 12. 16 See Hill, J. ‘Storefront for Art and Architecture in New York’, 1999, A Weekly Dose of Architecture. 31 May. 17 See Acconi, V., Holl, S. and Ritter, A. Storefront for Art and Architecture. New York: Hatje Cantz. 2000. 18 Traditional architectural relief is typically considered in terms of the sculptural categories of basrelief, high relief where the figure is undercut to reveal 50 per cent of the form, or sunken relief where a shallow incision is made on a flat surface. For a general overview of sculptural relief see Rogers, L. R. Relief Sculpture. London: Oxford University Press. 1974. 19 The Aegis is described as one of a trilogy of projects that include the Pallas House (CNC surface) and the Paramorph (tessellated aluminium double-curved surface) that explore what dECOi Architects term an ‘alloplastic’ approach – ‘a self-determinate operative strategy’: dECOi Architects. ‘Technological Latency: From Autoplastic to Alloplastic’. Digital Creativity. 2000; vol. 11: 135. 20 For an overview of Ned Kahn’s kinetic art see Mather, D. ‘An Aesthetic of Turbulence: The Works of Ned Kahn’, in Narula, Senupta, Sundaram, Sharen and Lovink (eds) Sarai Reader 6: Turbulence. Delhi: Centre for the Study of Developing Societies. 2006. 21 The MIT Kinetic Design Group was founded by Michael Fox, who has since gone on to form a kinetic design consultancy, Fox Lin Inc. 22 The potential impact of nanotechnology for architecture is sketched in Spiller, N. ‘Nanotechnology – the Liberation of Architecture’, in Hill, J. (ed.) Architecture: The Subject Is Matter. London: Routledge. 2001. For a technical overview of nanotechnology and other approaches that have become known as smart materials, see Addington, M. and Schodek, D. Smart Materials and Technologies for Architecture and Design Professionals. Oxford: Architectural Press. 2005. 153
23 Shape memory alloys have been utilized in a range of prototypes through the commercial availability of alloys and actuators. An overview of nanomaterials can be found in Addington, M. and Schodek, D. Smart Materials and Technologies for Architecture and Design Professionals. Oxford: Architectural Press. 2005. 24 See Hladik, P. ‘Moving Structure’, in Beesley, Hirosue, Buxton, Trankle and Turner (eds) Responsive Architectures Subtle Technologies. Toronto: Riverside Architectural Press. 2006. pp. 126–9. 25 See Hensel, M. and Sungurog, D. ‘Material Performance’. Architectural Design. 2008; vol. 78. 26 Window-cleaning robots have been under development for some time. For example, see Miyake, T. and Ishihara, H. ‘Development of Small Size Window Cleaning Robot by Wall Climbing Mechanism’. Paper presented at the International Symposium on Automation and Robotics in Construction. Tokyo. 2006. 27 For examples of small-scale wind turbines incorporated with a building skin, see the architectural project page of AeroVironment at http://www.avinc.com/wind 28 See Gage, S. ‘Edge Monkeys – the Design of Habitat Specific Robots in Buildings’. Technoetic Arts. 2005; vol. 3. 29 The pavilion incorporated a water wall 65 metres long and 18 metres high. Information on the project can be accessed online at http://www.grimshaw-architects.com/ 30 For a review of the Zaragoza pavilion, see Fortmeyer, R. ‘Control Freaks’. Architectural Record. 2010; March. 31 See Wolfe, C. ‘Lose the Building: Systems Theory, Architecture, and Diller+Scofidio’s Blur’. Postmodern Culture. 2006; 16(3). 32 See Beesley, P. Kinetic Architectures and Geotextile Installations. Cambridge, ON: Riverside Architectural Press. 2010. 33 See dECOi Architects. ‘Technological Latency: From Autoplastic to Alloplastic’. Digital Creativity. 2000; vol. 11: 134. 34 dECOi Architects, ‘Technological Latency: From Autoplastic to Alloplastic’. Digital Creativity. 2000; vol. 11: 135. 35 Beesley describes the development of his design approach in Beesley, P. (ed.) Kinetic Architectures and Geotextile Installations, Cambridge, ON: Riverside Press, 2009. 36 See Beesley, P. ‘Hylozoic Ground’. 2010. Available online at http://www.hylozoicground.com/press/ index.html. 37
See Payne, A. ‘Surface: Between Structure and Sense’, in Beesley, P. (ed.) Kinetic Architectures and Geotextile Installations. Cambridge, ON: Riverside Press. 2009. p. 57.
38 See Beesley, P., Hirosue, S. and Ruxton, J. ‘Towards Responsive Architecture’, in Beesley, Hirosue and Ruxton (eds) Responsive Architectures Subtle Technologies. Cambridge, ON: Riverside Architectural Press. 2006. p. 3. 39 For an overview of Ned Kahn’s kinetic facades, see Kahn, N. ‘Wind Veil’, 2000. Available online at http://nedkahn.com/wind.html. 40
See Haeusler, M. H. Chromatophoric Architecture: Designing for 3d Media Facades. Berlin: Jovis. 2010. p. 25.
41 Brian Eno has been exploring generative approaches within a range of art genres since the 1970s. For an early articulation of his approach to indeterminate forms of art, see Eno, B. ‘Generating and Organizing Variety in the Arts’, in Battock, G. (ed.) Breaking the Sound Barrier: A Critical Anthology of the New Music. New York: Elsevier-Dutton Publishing. 1981. 42 For an overview of intelligent facades, see Wigginton, M. and Harris, J. Intelligent Skins. Oxford: Butterworth-Heinemann. 2002. For a selection of case studies in an Asian context, see Harrison, A. Loe, E. and Read, J. Intelligent Building in South East Asia. London: Taylor and Francis. 1998. 43 See Baird, G. The Architectural Expression of Environmental Control Systems. London: Taylor and Francis. 2001. 44 See Bosoni, B. Jean Nouvel: Architecture and Design 1976–1995, Milan: Skira Editore, 1997. 45 See Wigginton, M. and Harris, J. Intelligent Skins. Oxford: Butterworth-Heinemann, 2002. 154
46 Anshuman, S. ‘Responsiveness and Social Expression: Seeking Human Embodiment in Intelligent Façades’. Paper presented at ACADIA 05, Savannah (Georgia). 2005. 47 The trajectory of design science in education can be traced to the 1958 RIBA Conference on Architectural Education. The theme of design science in terms of literal and metaphorical impact on composition is explored in Moloney, J. ‘Architectural Science: Literal and Notional Force Fields’, in Proceeding of New Constellations: Art, Science and Society. Sydney. 2006. 48 Anshuman undertakes a review of intelligent facades and observes that most ‘invite little bodily participation both from the occupants’ perspective within the building and larger urban participation from the outside’. Anshuman, S. ‘Responsiveness and Social Expression: Seeking Human Embodiment in Intelligent Façades’. Paper presented at ACADIA 05, Savannah (Georgia). 2005. p. 14. 49 See Craig, D. J. ‘The Future Tents: Kinetic Sculptor Chuck Hoberman Expands the Boundaries of Design’. Columbia Magazine. 2006; Spring. 50 See Fox, M. A. and Yeh, B. P. ‘Intelligent Kinetic Systems in Architecture’, in Nixon, P., Lacey, G. and Dobson, S. (eds) Managing Interactions in Smart Environments: First International Workshop, London: Springer, 2000; p. 99. 51 See Fox, M. A. and Kemp, M. Interactive Architecture. New York: Princeton Architectural Press. 2009. p. 29. 52 See Oosterhuis, K. and Xia, X. (eds) Interactive Architecture. Rotterdam: Episode Publishers. 2007. 53 See Reynolds, C. W. ‘Flocks, Herds and Schools: A Distributed Behavioral Model’. Computer Graphics. 1987; 21. 54 See Tizonis, A. Movement, Structure and the Work of Santiago Calatrava. Berlin: Birkhauser, 1995. 55 See Kronenburg, R. Flexible: Architecture That Responds to Change. London: Lawrence King Publishing. 2007. 56 Zuk, W. and Clark, R. H. Kinetic Architecture. New York: Van Nostrand Reinhold. 1970. 57 See Jormakka, K. Flying Dutchmen: Motion in Architecture. Basel: Birkhausser. 2002. 58 The generally accepted scientific approach to measuring human perception of speed is to measure the time taken for an object or event to cross an arc of vision. That is, the eye orientation is assumed to be fixed giving a total visual field of approximately 180 degrees. Anything that crosses this field in less then five seconds starts to blur, therefore the upper threshold of accurate motion detection is approximately 45 degrees per second. The lower threshold is based on the stimulus of the eye. If no change in perception of an element within the visual field occurs after more than two or three seconds, the object is perceived as stationary. Vision research is an extremely large and progressive field, but for the purposes of motion detection a key reference is: Borst, A. and Egelhaaf, M. ‘Principles of Visual Motion Detection’. Trends in Neurosciences. 1989; vol. 12. 59 See Zuk, W. and Clark, R. H. Kinetic Architecture. New York: Van Nostrand Reinhold. 1970. p. 24. 60 Zuk, W. and Clark, R. H. Kinetic Architecture. New York: Van Nostrand Reinhold. 1970. p. 154. 61 The emphasis is on the wider context at the beginning of the twentieth century, where Jormakka proposes that motion was closely studied in painting and sculpture. Paul Klee’s rejection of G. E. Lessing’s distinction between the arts of time and space is referenced, as are the Italian futurist painters and sculptors. The Futurist agenda is explored through Umberto Boccioni’s double conception of form – ‘form in movement (relative movement) and movement in form (absolute movement)’. The continuation of the theme of motion in the fine arts is traced to the Bauhaus, in particular Wassily Kandinsky. In terms of architecture, the first reference is to Jugendstil designer Henry van de Velde, who is associated with Kandinsky’s painting. The positioning of architecture within the context of the twentieth century avant-garde is continued in a section titled ‘The Language of Speed’, in which the correlation of Erich Mendelson’s ‘fluid forms’ to the force lines of electromagnetic fields is cited. See Jormakka, K. Flying Dutchmen: Motion in Architecture, Basel: Birkhausser. 2002. 62 Jormakka, K. Flying Dutchmen: Motion in Architecture. Basel: Birkhausser. 2002. p. 17. 63 Jormakka, K. Flying Dutchmen: Motion in Architecture. Basel: Birkhausser. 2002. p. 20. 155
64 Jormakka, K. Flying Dutchmen: Motion in Architecture. Basel: Birkhausser. 2002. p. 21. 65 Jormakka, K. Flying Dutchmen: Motion in Architecture. Basel: Birkhausser. 2002. p. 26. 66 Jormakka, K. Flying Dutchmen: Motion in Architecture. Basel: Birkhausser. 2002. p. 32. 67 Saggio, A. ‘Interactivity at the Centre of Architectural Research’, Architectural Design. 2005; 75(1): 23–29. 68 As discussed earlier in Chapter 2, the kinetic facade of the Institut du Monde Arabe has a history of mechanical failure. Buckminster Fuller’s US pavilion for Expo’ 67 consisted of a 200-foot-tall geodesic dome clad in acrylic panels. A sunshading system based on automated blinds was implemented, but the ambitious goal of tracking sun position through a computer control system was not implemented. The blind motor mechanisms were constantly failing during its life. A recent account of the project can be found in Massey, J. ‘Buckminster Fuller’s Cybernetic Pastoral: The United States Pavilion at Expo 67’. Journal of Architecture. 2006; vol. 11. 69 See Moussavi, F. and Kubo, M. The Function of Ornament, New York: Actar. 2006. Foreword. 70 See dECOi Architects, ‘Technological Latency: From Autoplastic to Alloplastic’. p. 138. 71 See Jormakka, K. Flying Dutchmen: Motion in Architecture. Basel: Birkhausser. 2002. p. 47. 72 See Mathews, S. ‘The Fun Palace as Virtual Architecture: Cedric Price and the Practices of Indeterminancy’. Journal of Architectural Education. 2006; vol. 59: 42. 73 See Deleuze, G. Bergsonism. New York: Zone Books. 1988. p. 42. 74 See Deleuze, G. Bergsonism. New York: Zone Books. 1988. p. 43. 75 See Deleuze, G. Bergsonism. New York: Zone Books. 1988. p. 43.
Chapter 3 1 Kaufmann, E. Architecture in the Age of Reason: Baroque and Post-Baroque in England, Italy and France. New York: Dover Publications. 1955. p. 82. 2 Kaufmann, E. Architecture in the Age of Reason: Baroque and Post-Baroque in England, Italy and France. New York: Dover Publications. 1955. p. 78. 3 Kaufmann, E. Architecture in the Age of Reason: Baroque and Post-Baroque in England, Italy and France. New York: Dover Publications. 1955. p. 78. 4 See Wittkower, R. Architectural Principles in the Age of Humanism. Chichester: John Wiley and Sons. 1999. p. 41. 5 See Hersey, G. ‘The Renaissance Matrix’. The Journal of the Society of Architectural Historians. 1977; vol. 36: 257. 6 See Wittkower, R. Architectural Principles in the Age of Humanism. Chichester: John Wiley and Sons. 1999. p. 109. 7 While there is prior evidence of symmetry in architecture, Alberti is credited as the first theoretician to define and recommend strict symmetry around a central vertical axis. See Hearn, F. Ideas That Shaped Buildings. Cambridge, MA: MIT Press. 2003. p. 84. 8 See Kaufmann, E. Architecture in the Age of Reason: Baroque and Post-Baroque in England, Italy and France. New York: Dover Publications. 1955. p. 82. 9 See Leatherbarrow, D. and Mostafavi, M. Surface Architecture. Cambridge, MA: MIT Press, 2002. Back cover. 10 See Leatherbarrow, D. and Mostafavi, M. Surface Architecture. Cambridge, MA: MIT Press, 2002. Back cover. p. 28. 11 The continuation of proportional composition beneath the veil of modernism was highlighted by Colin Rowe’s analysis in Rowe, C. The Mathematics of the Ideal Villa and Other Essays. Cambridge, MA: MIT Press. 1976. 12 See Leatherbarrow, D. and Mostafavi, M. Surface Architecture. Cambridge, MA: MIT Press, 2002. p. 25. 13 See Leatherbarrow, D. and Mostafavi, M. Surface Architecture. Cambridge, MA: MIT Press, 2002. p. 7. 156
14 See Leatherbarrow, D. ‘Architecture’s Unscripted Performance’, in Kolarevic, B. and Malkawi, A. (eds) Performance Architecture: Beyond Instrumentality. New York: Spon Press, 2005. p. 11. 15 See Leatherbarrow, D. ‘Architecture’s Unscripted Performance’, in Kolarevic, B. and Malkawi, A. (eds) Performance Architecture: Beyond Instrumentality. New York: Spon Press, 2005. p. 13. 16 See Leatherbarrow, D. ‘Architecture’s Unscripted Performance’, in Kolarevic, B. and Malkawi, A. (eds) Performance Architecture: Beyond Instrumentality. New York: Spon Press, 2005. p. 12. 17 See Leatherbarrow, D. ‘Architecture’s Unscripted Performance’, in Kolarevic, B. and Malkawi, A. (eds) Performance Architecture: Beyond Instrumentality. New York: Spon Press, 2005. p. 16. 18 See Giedion, S. Space, Time and Architecture: The Growth of a New Tradition. Cambridge, MA: Harvard University Press. 1941. p. Iiii. 19 See Giedion, S. Space, Time and Architecture: The Growth of a New Tradition. Cambridge, MA: Harvard University Press. 1941. p. Iv. 20 See Giedion, S. Space, Time and Architecture: The Growth of a New Tradition. Cambridge, MA: Harvard University Press. 1941. p. Ii. 21 As a way of introducing ‘space-time’, Giedion cites the mathematician Herman Minkowski: ‘[P]roclaimed for the first time with full certainty and precision this fundamental change of conception. ”Henceforth,” he said, ”space alone or time alone is doomed to fade into a mere shadow; only a kind of union of both will preserve their existence.”’ See Giedion, S. Space, Time and Architecture: the Growth of a New Tradition. Cambridge, MA: Harvard University Press. 1941. p. 443. 22 See Giedion, S. Space, Time and Architecture: The Growth of a New Tradition. Cambridge, MA: Harvard University Press. 1941. p. 443. 23 See Giedion, S. Space, Time and Architecture: The Growth of a New Tradition. Cambridge, MA: Harvard University Press. 1941. p. 447. 24 See Banham, R. Theory and Design in the First Machine Age. London: Architectural Press. 1960. p. 112. 25 See Banham, R. Theory and Design in the First Machine Age. London: Architectural Press. 1960. p. 133. The full quote by Banham is: ‘The resemblance to Boccioni’s ”field” concept of space, with bodies connected by geometrical fields of force is very striking, as in the appearance of purely superficial Boccionisms like the illuminated advertising that appears on the roof of some of these projects.’ 26 See Banham, R. Theory and Design in the First Machine Age. London: Architectural Press. 1960. p. 193. Banham discusses the writing of Werner Graeff of the De Stijl ‘G’ group. 27 Kwinter, S. Architectures of Time: Toward a Theory of the Event in Modernist Culture. Cambridge, MA: MIT Press. 2001. pp. 59–60. 28 Kwinter, S. Architectures of Time: Toward a Theory of the Event in Modernist Culture. Cambridge, MA: MIT Press. 2001. p. 61. 29 Kwinter, S. Architectures of Time: Toward a Theory of the Event in Modernist Culture. Cambridge, MA: MIT Press. 2001. pp. 65–6. 30 Kwinter, S. Architectures of Time: Toward a Theory of the Event in Modernist Culture. Cambridge, MA: MIT Press. 2001. p. 74. In using the term ‘kinematic plasticity’, Kwinter inserts a detailed footnote made in reference to writing on kinetic sculpture by Laszlo Moholy-Nagy, suggesting the same ideas were taken up by kinetic artists of the 1920s. Kwinter references Moholy-Nagy’s phrase ‘material is energy’ from Moholy-Nagy, L. Vision in Motion. Chicago: P. Theobald. 1947. 31 In 1687 Isaac Newton published his equation for universal gravitation. This proposed the force of gravity between two objects decreases in inverse proportion to the square of the distance between the objects. Solving the mathematics for a case where a third object is added has confounded mathematicians ever since. Poincaré produced a solution for a restricted case where the mass of one of the three bodies is negligible. In the process he discovered chaotic behaviour, which provided the foundation for chaos theory. See Barrow-Green, J. Poincaré and the Three Body Problem. New York: American Mathematical Society. 1997. 157
32 See Kwinter, S. ‘Landscapes of Change: Boccioni’s ”Stati d’animo” as a General Theory of Models’. Assemblage. 1992; p. 53. 33 See Kwinter, S. ‘Landscapes of Change: Boccioni’s ”Stati d’animo” as a General Theory of Models’. Assemblage. 1992. p. 58. 34 See Kwinter, S. ‘Landscapes of Change: Boccioni’s ”Stati d’animo” as a General Theory of Models’. Assemblage. 1992. p. 58. 35 Sanford Kwinter and Greg Lynn were regular participants of the annual conferences sponsored by the Anyone corporation, that occurred between 1991 and 2000. In the conference of 1998, Kwinter presented a call for an informal grouping or ‘line of resistance’ against the ‘cult of objects’ in favour of a ‘new rationality that sees space and form as indistinguishable, as active mediums shaped by both embedded and remote events and the patterns they form’. Lynn acknowledged and extended the agenda in his subsequent presentation at the same conference. See Kwinter, S. ‘Leap in the Void: A New Organon’, in Davidson, C. (ed.) Anyhow. Cambridge, MA: MIT Press. 1998; p. 25. 36 See Lynn, G. Animate Form. New York: Princeton Architectural Press. 1999. p. 11. 37 See Lynn, G. Animate Form. New York: Princeton Architectural Press. 1999. pp. 26–35. D’Arcy Thompson is presented as a pioneer of ‘deformation as an index of contextual forces acting on an organism’; Étienne-Jules Marey is referenced as one of the first to ‘move from a study of form in inert Cartesian space, devoid of force and motion, to the study of rhythms, movements, pulses, and flows and their effects on form’. 38 These projects include the Cardiff Bay Opera House Competition entry that is based on site topography translated to a parametric model, inflected by a functional programme and views out from the site; the Port Authority gateway scheme translated pedestrian and vehicular movement as gradient fields that inflect on a particle system, which change position and shape according to the forces; with the Yokohama ferry terminal competition entry, the Long Island House prototype, and the Kunstenter installation being variations on these techniques. 39 See Kwinter, S. ‘The Judo of Cold Combustion’, in Reiser, J. (ed.) Atlas of Novel Tectonics. New York: Princeton University Press. 2006. p. 15. 40 See Reiser, J. Atlas of Novel Tectonics. New York: Princeton Architectural Press. 2006. p. 50. 41 See Reiser, J. Atlas of Novel Tectonics. New York: Princeton Architectural Press. 2006. p. 53. 42 See Reiser, J. Atlas of Novel Tectonics. New York: Princeton Architectural Press. 2006. p. 177. 43 See Leatherbarrow, D. ‘Architecture’s Unscripted Performance’, in Kolarevic, B. and Malkawi, A. (eds) Performance Architecture: Beyond Instrumentality. New York: Spon Press. 2005. p. 11. 44 Ktesbios is considered the inventor of the first hydraulic servo mechanism based on a float valve that opened and closed proportionally. See Mayri, O. Origins of Feedback Control. Cambridge, MA: MIT Press. 1970. p.11. 45 See Wiener, N. Cybernetics: Or Control and Communication in the Animal and the Machine. Cambridge, MA: MIT Press. 1965. 46 The place of cybernetics as a ‘meta-subject’ is explored at length in Glanville, R. ‘Try Again. Fail Again.’ Kybernetes. 2007; vol. 36: 1183. 47 See Hayles, N. K. ‘Boundary Disputes: Homeostasis, Reflexivity, and the Foundations of Cybernetics’, in Markley, M. (ed.) Virtual Realities and Their Discontents. Baltimore: Johns Hopkins University Press. 1996. 48 See Hayles, N. K. ‘Boundary Disputes: Homeostasis, Reflexivity, and the Foundations of Cybernetics’, in Markley, M. (ed.) Virtual Realities and Their Discontents. Baltimore: Johns Hopkins University Press. 1996. p. 33. 49 Gordon Pask is a key figure in the development of cybernetics, and was influential at the Architectural Association London in the 1970s. According to Glanville, he supervised at least eight doctorates by architects in London during the early 1970s. He is perhaps best known for his collaboration with Cedric Price on the Fun Palace project. The contribution of Pask to architecture is commemorated in Frazer, J. ‘The Cybernetics of Architecture: A Tribute to the Contribution of Gordon Pask’. Kybernetes. 2001; 20. 158
50 Glanville, R. ‘Try Again. Fail Again.’ Kybernetes. 2007; vol. 36: 1184. 51 See Ashby, W. R. ‘Requisite Variety and Its Implications for the Control of Complex Systems’. Cybernetica, 1958; vol. 1. 52 See Gage, S. ‘How to Design a Black and White Box’. Kybernetes. 2007; vol. 36: 1335–6. 53 See Gage, S. ‘How to Design a Black and White Box’. Kybernetes. 2007; vol. 36: 1336. 54 See Frazer, J. An Evolutionary Architecture. London: Architectural Association. 1995. Preface. 55 See Frazer, J. An Evolutionary Architecture. London: Architectural Association. 1995. p. 65. 56 Frazer states, ‘We are proposing an alternative methodology whereby the model is adapted iteratively in the computer in response to feedback from the evaluation’. See Frazer, J. An Evolutionary Architecture. London: Architectural Association. 1995. p. 67. 57 See Frazer, J. An Evolutionary Architecture. London: Architectural Association. 1995. p. 99. 58 See Mulder, B., Van Zijl, I. and Rietveld, G. T. Rietveld Schroeder House. New York: Princeton Architectural Press. 1999. 59 See Mertins, D. The Presence of Mies. New York: Princeton Architectural Press. 1994.
Chapter 4 1 Rickey took it upon himself to champion the work of the early constructivists in the 1960s, well before the recognition of the importance of this legacy for art and architecture. ‘The thesis of this book is that the pioneer works of the early Constructivists established a base from which many of the diverse and non-objective tendencies of the decade 1957–67 have sprung’. See Rickey, G. Constructivism: Origins and Evolution. New York: George Braziller. 1967. Preface. 2 Roger Horrocks briefly worked with Len Lye and subsequently published a biography and an anthology of Lye’s writing. In a more recent work he develops Lye’s conception of kinetic art ‘as a theory not only relevant to Lye but any artist who works with motion’. See Horrocks, R. Art That Moves: The Work of Len Lye. Auckland: Auckland University Press. 2009. p. 6. 3 The taxonomy of art understandably has an extremely long history. A concise summary of development from Greek, Medieval and Renaissance theory on the distinction between art forms is undertaken in Kristeller, P. O. ‘The Modern System of the Arts: A Study in the History of Aesthetics Part 1’. Journal of the History of Ideas. 1951; vol. 12. Contemporary views are pluralist, but the legacy of Dessoirs’s classification of architecture as a spatial art is still prevalent. See Dessoir, M. Aesthetics and Theory of Art. Detroit: Wayne State University Press. 1970. 4 Hanoch-Roe, G. ‘Scoring the Path: Linear Sequences in Music and Space’, in Muecke and Zach (eds) Resonance: Essays on the Intersection of Music and Architecture. Lulu Publishing. 2007. p. 21. 5 The collection of essays by Muecke and Zach provides a wide overview of contemporary approaches to architecture and music. 6 Evans argues that musical harmony was never intended to be explicit, but acted as an ordering strategy that was ‘sensed’. See Evans, R. The Projective Cast: Architecture and Its Three Geometries. Cambridge, MA: MIT Press. 1995. p. 249. 7 See Evans, R. The Projective Cast: Architecture and Its Three Geometries. Cambridge, MA: MIT Press. 1995. p. 254. 8 See Evans, R. The Projective Cast: Architecture and Its Three Geometries. Cambridge, MA: MIT Press. 1995. p. 296. 9 A ruled surface is a curved surface generated by multiple straight lines. Evans spends a considerable amount of space tracing the introduction and use of ruled surface technique in Le Corbusier’s office. In relation to the Philip’s Pavilion he notes the key role played by Xenakis. ‘On several occasions Xenakis has made it clear that his inspiration for the ruled surface was not Le Corbusier, whom he accused of ignorance in technical matters, but Bernard Lafaille, an engineer …’ Evans, R. The Projective Cast: Architecture and Its Three Geometries. Cambridge, MA: MIT Press. 1995. p. 301. 159
10 See Levy, A. J. ‘Real and Virtual Spaces Generated by Music’. International Journal of Architectural Computing. 2003; vol. 1. 11 See Ham, J. ‘Music and Architecture: From Digital Composition to Physical Artefact’. The Quest for New Paradigms. Lisbon: eCAADe. 2005. 12 The legacy of Xenakis is still evident in contemporary musical composition, particularly in relation to the interplay of mathematics and composition. See Xenakis, I. Formalized Music: Thought and Mathematics in Composition. New York: Pendragon Press. 2001. 13 See Gavrilou, E. ‘Inscribing Structures of Dance into Architecture’. Proceedings of the 4th International Space Syntax Symposium, London, 2003. 14 See Guest, A. H. ‘Dance Notation‘. Perspecta. 1990; vol. 26. 15 See Eshkol, N. and Wachman, A. Movement Notation. London: Weidenfeld and Nicolson. 1958. 16 See Harries, J. G. ‘A Proposed Notation for Visual Fine Art’. Leonardo. 1975; vol. 8. 17 See Virilio, P., Lotringer, S. and Taormina, M. ‘After Architecture: A Conversation’. Grey Room. 2001; vol. 3. 18 According to Popper, Russian artists such as Archipenko and Burliuk introduced movement as a consequence of a general interest in natural phenomena; while Duchamp’s artwork Bicycle Wheel is a continuation of his concept of the ‘ready-made’. Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. p. 123. 19 Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. p. 145. 20 See Kepes, G. The Nature and Art of Motion. New York: G. Braziller. 1965. 21 Exercises in motion were part of the foundation programme at the Bauhaus, with the strongest link to kinetics being Laszlo Moholy-Nagy, who replaced Johannes Itten as the instructor of the foundation course at the Bauhaus in 1923. For nine years he would continue to work on his seminal kinetic sculpture, the Light-Space Modulator. Moholy-Nagy started the new Bauhaus in Chicago in 1937, where motion was still included as one of the plastic elements in the structure of the preliminary course. See Findeli, A. ‘Moholy-Nagy’s Design Pedagogy in Chicago (1937–46)’. Design Issues. 1990; vol. 7. 22 Burnham provides insight into the adaption of cybernetic principles for art in the 1960s. See Burnham, J. Beyond Modern Sculpture: The Effects of Science and Technology on the Sculpture of This Century. New York: G. Braziller. 1968. 23 Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; vol. 22: 220–31. 24 Popper, F. Origins and Development of Kinetic Art, London: Studio Vista. 1968. 25 Brett, G. Kinetic Art: The Language of Movement. London: Studio Vista. 1968 26 Malina, F. J. Kinetic Art: Theory and Practice: Selections from the Journal Leonardo. New York: Dover Publications. 1974. 27 Bois, Y.-A., ‘Force Fields: Phases of the Kinetic’. Art Forum. 2000; vol. 39. Online at http:// findarticles.com/p/articles/mi_m0268/is_3_39/ai_67935450/ 28 Brett, G. Museu d’Art Contemporani (Barcelona Spain) and Hayward Gallery, Force Fields: Phases of the Kinetic. Barcelona: Museu d’Art Contemporani. 2000. 29 See Horrocks, R. Art That Moves: The Work of Len Lye. Auckland: Auckland University Press. 2009. 30 Malina, F. J. ‘Kinetic Painting: The Lumidyne System’, in Malina (ed.) Kinetic Art: Theory and Practice. New York: Dover Publications. 1974. p. 44. 31 Brett, G. Kinetic Art: The Language of Movement. London: Studio Vista. 1968. p. 8. 32 Brett, G. Kinetic Art: The Language of Movement. London: Studio Vista. 1968. p. 9. 33 Brett, G. Kinetic Art: The Language of Movement. London: Studio Vista. 1968. p. 12. 34 Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. p. 215. 35 Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. p. 217. 36 Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. p. 124.
37 The realist manifesto is reproduced in Gabo, N. Gabo: Constructions, Sculpture, Paintings, Drawings and Engravings. London: Lund Humphries. 1957. p. 151. 38 Gabo, N. Gabo: Constructions, Sculpture, Paintings, Drawings and Engravings. London: Lund Humphries. 1957. p. 152. 39 Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. p. 127. 40 Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. p. 146. 41 Kenneth Martin provides one direct crossover between architecture and kinetic art. Popper, in addition to commenting on his architectural material palette, notes his continued use of spiral and cross forms. Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. p. 149. 42 Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. p. 149. 43 Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. p. 136. 44 Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. p. 130. 45 Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. p. 233. 46 The Oxford English Dictionary’s definition of the agogic for music is ‘relating to or denoting accentuation within musical phrase by slight lengthening of notes’. Pearsall, J. The Concise Oxford English Dictionary. Oxford: Oxford University Press. 2002. 47 Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. p. 228. 48 Souriau, E. ‘Time in the Plastic Arts’. The Journal of Aesthetics and Art Criticism. 1949; vol. 7: 228. 49 See Horrocks, R. Art That Moves: The Work of Len Lye. Auckland: Auckland University Press, 2009. p. 40 50 Horrocks, R. Art That Moves: The Work of Len Lye. Auckland: Auckland University Press. 2009. p. 88 51 Horrocks, R. Art That Moves: The Work of Len Lye. Auckland: Auckland University Press. 2009. p. 91. 52 Horrocks, R. Art That Moves: The Work of Len Lye. Auckland: Auckland University Press. 2009. p. 78. 53 Horrocks, R. Art That Moves: The Work of Len Lye. Auckland: Auckland University Press. 2009. pp. 107–9. 54 Horrocks, R. Art That Moves: The Work of Len Lye. Auckland: Auckland University Press. 2009. pp. 54–5. 55 Horrocks, R. Art That Moves: The Work of Len Lye. Auckland: Auckland University Press. 2009. p. 127. 56 Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; vol. 22: 223–4. 57 Gabo, N. Gabo: Constructions, Sculpture, Paintings, Drawings and Engravings. London: Lund Humphries. 1957. p. 151. 58 Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; vol. 22. 59 Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; vol. 22: 224. 60 Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; vol. 22: 225. 61 Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; vol. 22: 226. 62 Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; vol. 22: 226. 63 Rickey locates his rejection of mimesis in relation to constructivism. ‘Thus it is not in the imitation of appearances that kinetic art is served by ”nature” but in recognition of its laws, awareness of analogies, and response to the vast repertory of movement in the environment. Kinetic art 161
has remained close to the constructivist principle that ”we construct our work as the universe constructs its own”, and that ”efficacious existence is the highest beauty.’ Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; vol. 22: 223. 64 Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; vol. 22: 230. 65 Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; vol. 22: 231. 66 Shanken, E. A. ‘Reprogramming Systems Aesthetics: A Strategic Historiography’ in Proceedings of the Digital Arts and Culture Conference, 2009. Irvine: University of California. 2009. 67 Burnham, J. ‘Systems Esthetics’. Art Forum. 1968; vol. 7: 35. 68 Shanken’s paper summarizes the use made of Burnham’s ‘systems esthetics’ by himself and a number of fellow media art theorists and practitioners. See Shanken, E. A. ‘Reprogramming Systems Aesthetics: A Strategic Historiography’, in Proceedings of the Digital Arts and Culture Conference 2009. Irvine: University of California, 2009. 69 For an overview on digital art and the computational techniques of artificial life, see Whitelaw, M. Metacreation: Art and Artificial Life. Cambridge, MA: MIT Press. 2004. 70 See Dorin, A. ‘Classification of Physical Process for Virtual-Kinetic Art’, in Proceeding of First Iteration. Melbourne. 1999. p. 68. 71 See Wolfram, S. ‘Universality and Complexity in Cellular Automata’. Nature. 1984; vol. 311. 72 See Dorin, A. ‘Classification of Physical Process for Virtual-Kinetic Art’, in Proceeding of First Iteration. Melbourne. 1999. p. 71. 73 Dorin, A. ‘Classification of Physical Process for Virtual-Kinetic Art’, in Proceeding of First Iteration. Melbourne. 1999. p. 71. 74 Dorin, A. ‘Classification of Physical Process for Virtual-Kinetic Art’, in Proceeding of First Iteration. Melbourne. 1999. p. 72. 75 Dorin, A. ‘Classification of Physical Process for Virtual-Kinetic Art’, in Proceeding of First Iteration. Melbourne. 1999. p. 73. 76 Dorin, A. ‘Classification of Physical Process for Virtual-Kinetic Art’, in Proceeding of First Iteration. Melbourne. 1999. pp. 77–8. 77 Jon McCormack is an established figure in the generative arts. A collection of his writing and documentation of his artworks can be found in McCormack, J. Impossible Nature: The Art of Jon McCormack. Melbourne: Australian Centre For The Moving Image. 2004. Both Dorin and McCormack contribute chapters to Romero, J. and Machado, P. The Art of Artificial Evolution: A Handbook of Evolutionary Art and Music. New York: Springer. 2007.
Chapter 5 1 There are now a number of books on computational techniques for architectural design, within the rubric of parametric, algorithmic or programming. These include: Burry, M. Scripting Cultures: Architectural Design and Programming. London: John Wiley and Sons. 2011; Coates, P. Programming Architecture. London and New York: Routledge. 2010; Sakamoto, T. and Ferre, A. From Control to Design: Parametric/Algorithmic Architecture. New York: Actar. 2008; Terzidis, K. Algorithmic Architecture. Boston: Architectural Press. 2006; Woodbury, R. Elements of Parametric Design. New York: Routledge. 2010. 2 See Lynn, G. Animate Form. New York: Princeton Architectural Press. 1999. p. 272. 3 See Mitchell, W. J. and McCullough, M. Digital Design Media: A Handbook for Architects and Design Professionals. New York: Van Nostrand Reinhold. 1991. p. 300. 4 Mitchell, W. J. and McCullough, M. Digital Design Media: A Handbook for Architects and Design Professionals. New York: Van Nostrand Reinhold. 1991. p. 300. 5 As previously discussed in Chapter 2 see Zuk and Clark, Kinetic Architecture, 1970. 6 As previously discussed in Chapter 2, see Fox and Yeh, ‘Intelligent Kinetic Systems in Architecture’, 2000. 162
7 Saunders discusses this example at length, drawing a series of diagrams explaining the phenomenon. See Saunders, P. T. An Introduction to Catastrophe Theory. Cambridge: Cambridge University Press. 1980. 8 For an overview of sampling technology and applications for music, see Braun, H. (ed.) Music and Technology in the Twentieth Century. Baltimore: Johns Hopkins Press. 2002. 9 An excellent example of an artwork that utilizes remote data input to create an engaging local outcome is the D-Tower designed by NOX. Commissioned by the city of Doetinchem in the Netherlands, the D-Tower consists of three parts: a website (accessible to everybody), a questionnaire (accessible to a hundred different people each year) and a 12-metre tower reminiscent of a glowing jellyfish. All three parts are interactively related to each other, with the tower being internally lit with a mix of red, green and blue light. Updated each night, the mix of colour reflects responses to the questionnaires, which are intended to gauge the mood of the town in relation to a variety of issues. Information and photographs of the D-Tower by NOX architects are available from http://www.arcspace.com/architects/nox/d_tower2/d_tower2.html. 10 The potential for collaboration between fine artists and architects in developing kinetic facades is outlined in Moloney, J. ‘Kinetic Architectural Skins and the Computational Sublime’. Leonardo. 2009; vol. 42. 11 Data is considered to be continuous when measurement could conceivably take on any value, such as the walking speed of a pedestrian. Discrete data can only take on integral values, for example a number based on counting the number of pedestrians. For an overview of data structures in relation to digital design media, see Mitchell, W. J. and McCullough, M. Digital Design Media: A Handbook for Architects and Design Professionals. New York: Van Nostrand Reinhold. 1991. 12 For an overview of the role of tectonics in architectural design, see Frampton, K. Studies in Tectonic Culture: The Poetics of Construction in Nineteenth and Twentieth Century Architecture. Cambridge, MA: MIT Press. 1995. 13 See Gabo, N. Gabo: Constructions, Sculpture, Paintings, Drawings and Engravings. London: Lund Humphries. 1957. 14 Derrington, A., Allen, H. and Delicato, L. ‘Visual Mechanisms of Motion Analysis and Motion Perception’. Annual Review of Phycology. 2004; vol. 55. The relation of visual mechanisms to kinetic art from a scientific perspective is also discussed in Zeki, S. and Lamb, M. ‘The Neurology of Kinetic Art’. Brain. 1994; vol. 117: 607–36. 15 The theme of slowness is examined in the essay Bury, P. ‘Time Dilated’, in René, (ed.) The Movement. Dusseldorf: Editions Denise René. 1955.
Chapter 6 1 See Focillon, H. The Life of Forms in Art. Cambridge, MA: MIT Press. 1989. p. 50. 2 Focillon, H. The Life of Forms in Art. Cambridge, MA: MIT Press. 1989. p. 52. 3 See Mennan, Z. ‘The Question of Non Standard Form’. METU JFA. 2008; vol. 25: 181. 4 The term instrumental representation is used in a negative tone by Dalibor Vesely, in relation to the use of realistic three-dimensional computer modelling for design. The use of an abstract approach for this research is considered appropriate here, given the emphasis on pattern formation at a morphological level. See Vesely, D. Architecture in the Age of Divided Representation: The Question of Creativity in the Shadow of Production. Cambridge, MA: MIT Press. 2004. 5 An overview of the tradition of architectural designers using line drawings and simple shading techniques is considered in detail by Herbert. A range of drawing types are considered – plan, section elevation and perspective. See Herbert, D. Architectural Study Drawings. New York: John Wiley and Sons. 1993. Herbert’s focus is on architects from twentieth-century modernism, with an emphasis on plan and section studies. Elevation studies were extensively used in the post-modern period. As an example, see Krier, R. Architectural Composition. London: Academy Editions. 1988. 163
6 A technical overview and examples of a range of uses of noise in computer graphics and modelling is available in Elbert. D., Musgrave, K., Peachey, D., Perlin, K. and Worley, S. Texturing and Modeling: A Procedural Approach. San Francisco: Morgan Kaufmann. 2003. 7 As described by Chopard and Droz, cellular automata (CA) ‘are an idealization of a physical system in which space and time are discrete, and the physical quantities take only a finite set of values’. The authors trace the origin of CA to John Von Neumann who, in the 1940s, conceived the concept in terms of architecture for massive parallel computing. Contemporary approaches are described as a simplification of the Von Neumann approach. ‘This simplification was made possible by giving up the property of computational universality, while still conserving the idea of having a spatially distributed sequence of instructions (a kind of cellular DNA) which is executed to create a new structure.’ Chopard, B. and Droz, M. Cellular Automata Modelling of Physical Systems. Cambridge: Cambridge University Press. 1998. pp. 1–3. 8 Life-like CA have their origin in John Conway’s Game of Life, which was originally conceived as a board game. See Gardner, M. ‘Mathematical Games: The Fantastic Combinations of John Conway’s New Solitaire Game ”Life”’. Scientific American. 1970. Wolfram pioneered the computer applications of CA. See Wolfram, S. ‘Universality and Complexity in Cellular Automata’. Nature. 1984; vol. 311: 419–24. There are now numerous versions of life-like CA, each of which produce emergent patterns that correspond to life-like animal behaviour. See, for example, Cole, B. and Cheshire, D. ‘Mobile Cellular Models of Ant Behavior’. The American Naturalist. 1996. 9 A mathematical description of cyclic cellular automata can be found in Fisch, R. ‘Cyclic Cellular Automata and Related Processes’, in Gutowitz, H. (ed.) Cellular Automata: Theory and Experiment. Boston, MA: MIT Press. 1991. 10 While flocking algorithms are computationally more related to particle systems then CA, they share a similar ‘local rule’ approach. In the case of flocking, the behaviour of individual ‘flockmates’ is reliant on proximity and behaviour of nearby flockmates. Reynolds’ original paper on flocking is Reynolds, C. W. ‘Flocks, Herds and Schools: A Distributed Behavioral Model’. Computer Graphics. 1987; vol. 21: 25–34. The approach has been implemented into computer animation to generate realistic behavioural interaction. For an overview from a content generation perspective see Parent, R. Computer Animation. San Francisco: Morgan Kaufmann. 2002. 11 Reynolds, C. W. ‘Flocks, Herds and Schools: A Distributed Behavioral Model’. Computer Graphics. 1987; vol. 21: 31. 12 Computer animation software typically incorporates a high-level programming language, which is usually referred to as a scripting language; hence the use of the phrase animation script. 13 Barratt provides an excellent introduction to arithmetic progressions in the context of design patterns. See Barratt, K. Logic and Design in Art, Science and Mathematics. London: Herbert Press. 1980. pp. 73–4. 14 Barratt, K. Logic and Design in Art, Science and Mathematics. London: Herbert Press. 1980. pp. 76–8. 15 Dorin, A. ‘Classification of Physical Process for Virtual-Kinetic Art’, in Proceeding of First Iteration. Melbourne. 1999. p. 73. 16 The phrase ‘essential configuration’ is taken from Steadman’s discussion of the D’Arcy Thompson and Albert Durer drawing techniques, where profiles are drawn on a grid which is then subject to scaling and shearing transformations. ‘Speaking rather loosely, we might say that D’Arcy Thompson’s and Durer’s method separates out the essential configuration or Gestalt of the shapes of faces, bodies, or other figures, from the particular relative sizes which the separate part of the figure may assume.’ Steadman, P. Architectural Morphology: An Introduction to the Geometry of Building Plans. London: Pion. 1983. p. 9. 17 It should be emphasized again that there is no attempt to replicate environmental visual perception, such as the ecological approach as developed by James Gibson. Gibson, J. J. The Perception of the Visual World. Boston: Houghton Mifflin. 1950.
18 The key reference for current knowledge on human vision used for this research is a journal paper that collates the current state of theory on motion recognition. See Derrington, A., Allen, H. and Delicato, L. ‘Visual Mechanisms of Motion Analysis and Motion Perception’. Annual Review of Phycology. 2004; vol. 55: 181–205. The authors summarize current debate on theories of motion. Feature tracking is one of two main approaches and refers to ‘inferring motion from changes in the retinal position of objects, or their features over time (p. 183). The second approach considers the anatomy of the eye, and is known as ‘first order’ motion filtering. The first approach has been studied in laboratory trials, and edge detection and relative shading were found to be primary factors in motion detection. 19 The tests were undertaken using controlled experiments, where greyscale patterns were displayed on computer monitors and trial participants were required to detect the motion of a variety of patterns that varied in number of edges and relative shading. Derrington, A., Allen, H. and Delicato, L. ‘Visual Mechanisms of Motion Analysis and Motion Perception’. Annual Review of Phycology. 2004; vol. 55: 187. 20 This well-known principle is discussed in relation to design by Barratt. ‘The basic principles of numeracy and grouping are simple. Difficulties usually arise from failure to follow flowpaths, and thus the hierarchy and sequence of units to be read. The limited human response to the numbers beyond seven is balanced by an intense awareness of the smallest integers.’ See Barratt, K. Logic and Design in Art, Science and Mathematics. London: Herbert Press. 1980. p. 30. 21 Conway’s original Game of Life was based on a 5 × 5 grid, i.e. 25 cells. A large number of CA examples available online were considered, many of which were at the upper end of the 300–1000 observed threshold. Email correspondence with an expert in the field (Dr Alan Dorin, Monash University) confirmed there is no optimum number of cells. Trials using the 21 × 17 grid using a range of CA demonstrated adequate pattern emergence. 22 For an overview of types of noise algorithms, see Elbert, D., Musgrave, K., Peachey, D., Perlin, K. and Worley, S. Texturing and Modeling: A Procedural Approach. San Francisco: Morgan Kaufmann. 2003.
Chapter 7 1 See Foucault, M. The Order of Things. New York: Random House. 1970. p. 268. Focault’s ‘archeology of the human sciences’ reveals a general shift in approaches to classification, which occurs in the early nineteeth century. As evident in the approach of French naturalist Georges Cuvier, there is a transition from taxonomy of external appearance to one of internal function, i.e. from morphology to physiology. ‘From Curvier onward, function, defined according to its nonperceptible form as an affect to be obtained, is to serve as a constant middle term and to make it possible to relate together totalities of elements without the slightest visible identity’. See Foucault, The Order of Things, p. 265. 2 See Bowker, G. C. and Star, S. L. Sorting Things Out: Classification and Its Consequences. Boston, MA: MIT Press. 2000. 3 ’Flaring’ and ‘swaying’ are two of the terms that occur in Len Lye’s table of ‘some suggested tangible motion’ sculptures. See Horrocks, R. Art That Moves: The Work of Len Lye. Auckland: Auckland University Press. 2009. p. 78. 4 Flight dynamics has developed as a highly sophisticated science, but the conventions of pitch, yaw and roll remain fundamental. For a contemporary introduction see Stengel, R. F. Flight Dynamics. Princeton, NJ: Princeton University Press. 2004. 5 In his discussion of visual rhythm, Barratt observes the role of five as an optimum number of parts. ‘Midway between our rhythmic limits at three and seven, it becomes an optimum number for most human beings.’ Barratt, K. Logic and Design in Art, Science and Mathematics. London: Herbert Press. 1980. p. 46. 6 As well as the references developed in Chapter 3 on field thinking, the concept has been used frequently by Stan Allen where he uses the term ‘field compositions’. See Allen, S. and Agrest, 165
D. Practice: Architecture, Technique, and Representation. Australia: G+B Arts International. 2000. pp. 155–8. Another design researcher who has explicitly developed the field is Kevin Rhowbotham, who conflates field and functional event to develop an innovative teaching pedagogy. See Rhowbotham, K. Field Event/Field Space. London: Black Dog. 1999. 7 See Eco, U. Kant and the Platypus. London: Vintage. 1999. p. 179. 8 The development of the Beaufort scale in terms of observation is eloquently traced in Huler, S. Defining the Wind: The Beaufort Scale, and How a 19th-Century Admiral Turned Science into Poetry. New York: Random House. 2005.
Chapter 8 1 The role of tempo in music and the problems of universal interpretation of temporal notation before the metronome is, not surprisingly, a wide and controversial field of discourse. The tempo categories and their application are part of a larger debate on the reciprocity between metre and rhythm. For example, Sachs has argued that pure accent and pure metre are extremes of the same class. Sachs, C. Rhythm and Tempo: A Study in Music History. New York: W. W. Norton. 1953. A contrasting (and controversial) view on musical time is the proposition by Epstein of a universal tempo based on the neurobiology of human perception. Epstein, D. Shaping Time: Music, the Brain, and Performance. New York: Schirmer Books. 1995. Both approaches articulate the problem of classification when applied to music and, arguably, all temporal forms of art. 2 See Leatherbarrow, D. and Mostafavi, M. Surface Architecture. Cambridge, MA: MIT Press: 2002. p. 28. 3 Jean Nouvel, for example, continues the modernist fascination with the reflection and refraction of light through glass, while Rowe and Slutzky’s theory of ‘phenomenal transparency’ and the interplay of volumes as a compositional tactic continues to be developed by a generation of architects informed by the work of Peter Eisenman. See Rowe, C. and Slutzky, R. ‘Transparency: Literal and Phenomenal’. Perspecta. 1963; vol. 8. 4 The contemporary preoccupation with continuous multi-curved geometry can be interpreted as a version of sculptural modelling reliant on the play of light on form, in a similar manner to Louis Kahn’s famous text on architectural form revealed through light. See Twombly, R. C. Louis I Kahn: Essential Texts. New York: W. W. Norton. 2003. 5 See Moussavi, F. and Kubo, M. The Function of Ornament. New York: Actar. 2006. 6 Kwinter articulates his interest in formalism in a similar manner to the use of morphology that underlies the design experimentation undertaken in this book. ‘What I call true formalism refers to any method that diagrams the proliferation of fundamental resonances and demonstrates how these accumulate into figures of order and shape.’ Kwinter, S. ‘Who’s Afraid of Formalism’, in Davidson (ed.) Far from Equilibrium: Essays on Technology and Design Culture. New York: Actar. 2008. p. 146. 7 See Reiser, J. Atlas of Novel Tectonics. New York: Princeton Architectural Press. 2006. p. 52. 8 Reiser, J. Atlas of Novel Tectonics. New York: Princeton Architectural Press. 2006. p. 66. 9 Reiser, J. Atlas of Novel Tectonics. New York: Princeton Architectural Press. 2006. p. 66. 10 See Howard, L. Essay on the Modification of Clouds. London: John Churchill and Sons. 1803. 11 See Hamblyn, R. The Invention of Clouds. New York: Farrar, Straus and Giroux. 2001. p. 124. 12 See Hamblyn, R. The Invention of Clouds. New York: Farrar, Straus and Giroux. 2001. p. 21. 13 See Hamblyn, R. The Invention of Clouds. New York: Farrar, Straus and Giroux. 2001. p. 25. 14 Robert Hooke attempted to gather information from a team of observers, and, according to Hamblyn, was the first to consider inventing terms specific to cloud formation. See Hamblyn, R. The Invention of Clouds. New York: Farrar, Straus and Giroux. 2001. p. 97. Larmarck’s approach was close to Howard’s but was flawed in its execution. ‘But though he had made the crucial recognition that every cloud could be described through a limited number of basic forms, he preferred, following Buffon, to view clouds in terms of individual entities rather than as members 166
of Linnean species. See Hamblyn, R. The Invention of Clouds. New York: Farrar, Straus and Giroux, 2001. p. 103. 15 See Hamblyn, R. The Invention of Clouds. New York: Farrar, Straus and Giroux. 2001. p. 123. 16 Howard’s idea of modification provides an example of an alternate approach to studying morphology. Compare, for example, the pioneering work of D’Arcy Thompson, an often-quoted reference in architecture. Significantly, one fundamental principle of Thompson’s approach to morphology was that comparison should only occur within variation of the same class. ‘We are limited, both by our method and by the whole nature of the case, to the comparison or organisms such as are manifestly related to one another and belong to the same zoological class. For it is a grave sophism, in natural history as in logic, to make a transition into another kind’. See Thompson, D. A. On Growth and Form. Cambridge: Cambridge University Press. 1961. p. 273. The transformative capacity of Howard’s taxonomy of clouds provides an example of an alternate approach. Modification is based on the principle of internal variance within simple cloud types and intermediate states that allow, to reference Thompson, ‘a transition into another kind’. 17 A key issue that needs to be addressed to enable transition from theory to practice is development in simulation software. As a way forward, an approach to conceiving software calibrated to the physical performance of kinetic systems, has been outlined. See Moloney, J. ‘A Framework for the Design of Kinetic Facades’, paper presented at CAAD Futures. Sydney. 2007.
Acconi, V., Holl, S. and Ritter, A. Storefront for Art and Architecture. New York: Hatje Cantz. 2000. Addington, M. and Schodek, D. Smart Materials and Technologies for Architecture and Design Professionals. Oxford: Architectural Press. 2005. Ag4 Media Facades 2000–2006. Cologne: Daab. 2006. Allen, S. and Agrest, D. Practice: Architecture, Technique, and Representation. Australia: G+B Arts International. 2000. Anshuman, S. ‘Responsiveness and Social Expression: Seeking Human Embodiment in Intelligent Façades’. Paper presented at ACADIA 05, Savannah (Georgia). 2005, 12–23. Anshuman, S. and Kumar, B. ‘Architecture and HCI: A Review of Trends Towards an Integrative Approach to Designing Responsive Space’. International Journal of IT in Architecture, Engineering and Construction. 2004; 2: 273–83. Anstey, T. ‘Where Is the Project? Cedric Price on Architectural Action’, in Rendell, J. (ed.) Critical Architecture. London: Routledge. 2007. Ashby, W. R. ‘Requisite Variety and Its Implications for the Control of Complex Systems’. Cybernetica. 1958; 1: 83–99. Baird, G. The Architectural Expression of Environmental Control Systems. New York: Spon Press, 2001. Ban, S., Ambasz, E., Bell, E. and Wood, D. Shigeru Ban. New York: Princeton University Press, 2001. Banham, R. Theory and Design in the First Machine Age. London: Architectural Press. 1960. Barratt, K. Logic and Design in Art, Science and Mathematics. London: Herbert Press. 1980. Barrow-Green, J. Poincaré and the Three Body Problem. New York: American Mathematical Society. 1997. Batty, M. ‘A Research Programme for Urban Morphology’. Environment and Planning B: Planning and Design. 1999; 26: 475–6. Beesley, P. ‘Hylozoic Ground’. 2010. Online reference available at http://www.hylozoicground.com/. Accessed 10 November 2010. Beesley, P. Kinetic Architectures and Geotextile Installations. Cambridge, ON: Riverside Architectural Press. 2010. Beesley, P., Hirosue, S. and Ruxton, J. ‘Towards Responsive Architecture’, in Beesley, P., Hirosue, S. and Ruxton, J. (eds) Responsive Architectures Subtle Technologies. Cambridge, ON: Riverside Architectural Press. 2006. Bois,Y.-A. ‘Force Fields: Phases of the Kinetic’. Art Forum. 2009; 39. Bois, Y.-A. and Shepley, J. ‘A Picturesque Stroll around Clara-Clara’. October. 1984; 29: 32–6. Borst, A. and Egelhaaf, M. ‘Principles of Visual Motion Detection’. Trends in Neurosciences. 1989; 12: 297–306. Bowker, G. C. and Star, S. L. Sorting Things Out: Classification and Its Consequences. Boston, MA: MIT Press. 2000. Braun, H. (ed.) Music and Technology in the Twentieth Century. Baltimore: Johns Hopkins Press. 2002. Brett, G. Kinetic Art: The Language of Movement. London: Studio Vista. 1968. Brett, G., Museu D’art Contemporani (Barcelona Spain) and Hayward Gallery. Force Fields: Phases of the Kinetic. Barcelona: Museu d’Art Contemporani. 2000. Burnham, J. Beyond Modern Sculpture: The Effects of Science and Technology on the Sculpture of
This Century. New York: G. Braziller. 1968. Burnham, J. ‘Systems Esthetics’. Art Forum. 1968; 7: 30–5. Burry, M. Scripting Cultures: Architectural Design and Programming. London: John Wiley and Sons. 2011. Bury, P. ‘Time Dilated’, in René, D. (ed.) The Movement. Dusseldorf: Editions Denise René. 1955. Chopard, B. and Droz, M. Cellular Automata Modelling of Physical Systems. Cambridge: Cambridge University Press. 1998. Coates, P. Programming Architecture. London and New York: Routledge. 2010. Cole, B. and Cheshire, D. ‘Mobile Cellular Models of Ant Behavior’. The American Naturalist. 1996; 148: 1–15. Craig, D. J. ‘The Future Tents: Kinetic Sculptor Chuck Hoberman Expands the Boundaries of Design’. Columbia Magazine. 2006; Spring. dECOi Architects. ‘Technological Latency: From Autoplastic to Alloplastic’, Digital Creativity. 2000; 11: 131–43. Deleuze, G. Bergsonism. New York: Zone Books. 1988. Derrington, A., Allen, H. and Delicato, L. ‘Visual Mechanisms of Motion Analysis and Motion Perception’. Annual Review of Phycology. 2004; 55: 181–205. Dessoir, M. Aesthetics and Theory of Art. Detroit: Wayne State University Press. 1970. Dorin, A. ‘Classification of Physical Process for Virtual-Kinetic Art’, Proceeding of First Iteration, Melbourne, 1999, 68–79. Eco, U. Kant and the Platypus. London: Vintage. 1999. Elbert, D., Musgrave, K., Peachey, D., Perlin, K. and Worley, S. Texturing and Modeling: A Procedural Approach. San Francisco: Morgan Kaufmann. 2003. Eno, B. ‘Generating and Organizing Variety in the Arts’, in Battock, G. (ed.) Breaking the Sound Barrier: A Critical Anthology of the New Music. New York: Elsevier-Dutton Publishing. 1981. Epstein. D. Shaping Time: Music, the Brain, and Performance. New York: Schirmer Books. 1995. Eshkol, N. and Wachman, A. Movement Notation. London: Weidenfeld and Nicolson. 1958. Evans, R. The Projective Cast: Architecture and Its Three Geometries. Cambridge, MA: MIT Press. 1995. Findeli, A. ‘Moholy-Nagy’s Design Pedagogy in Chicago (1937–46)’. Design Issues. 1990; 7: 4–19. Fisch, R. ‘Cyclic Cellular Automata and Related Processes’, in Gutowitz, H. (ed.) Cellular Automata: Theory and Experiment. Boston, MA: MIT Press. 1991. Focillon, H. The Life of Forms in Art. Cambridge, MA: MIT Press. 1989. Fortmeyer, R. ‘Control Freaks’. Architectural Record. 2010; March: 102–3. Foucault, M. The Order of Things. New York: Random House. 1970. Fox, M. A. and Kemp, M. Interactive Architecture. New York: Princeton Architectural Press. 2009. Fox, M. A. and Yeh, B. P. ‘Intelligent Kinetic Systems in Architecture’, in Nixon, P., Lacey, G. and Dobson, S. (eds) Managing Interactions in Smart Environments: First International Workshop. London: Springer. 2000. Frampton, K. Modern Architecture: A Critical History. London: Oxford University Press. 1981. Frampton, K. Studies in Tectonic Culture: The Poetics of Construction in Nineteenth and Twentieth Century Architecture. Cambridge, MA: MIT Press. 1995. Frazer, J. An Evolutionary Architecture. London: Architectural Association. 1995. Frazer, J. ‘The Cybernetics of Architecture: A Tribute to the Contribution of Gordon Pask’. Kybernetes. 2001; 20: 641–51. Gabo, N. Gabo: Constructions, Sculpture, Paintings, Drawings and Engravings. London: Lund Humphries. 1957. Gabo, N. and Pevsner, A. ‘The Realistic Manifesto’, in Bowlt, J. (ed.) Russian Art of the Avant Garde: Theory and Criticism (Documents of Twentieth Century Art), 1902–1934, New York: Thames and Hudson, 1988. (revised edition). Gage, S. ‘Edge Monkeys – the Design of Habitat Specific Robots in Buildings’. Technoetic Arts. 2005; 3: 169–79. Gage, S. ‘How to Design a Black and White Box’. Kybernetes. 2007; 36: 1129–39. 169
Gardner, M. ‘Mathematical Games: The Fantastic Combinations of John Conway’s New Solitaire Game ”Life”.’ Scientific American. 1970; 223: 120–3. Gavrilou, E. ‘Inscribing Structures of Dance into Architecture’, in Proceedings of the 4th International Space Syntax Symposium, London, 2003, 32.1–32.16. Gibson, J. J. The Perception of the Visual World. Boston: Houghton Mifflin. 1950. Giedion, S. Space, Time and Architecture: The Growth of a New Tradition. Cambridge, MA: Harvard University Press. 1941. Glanville, R. ‘Try Again. Fail Again.’ Kybernetes. 2007; 36: 1173–206. Güçyeter, B. ‘A Comparative Examination of Structural Characteristics of Retractable Structures’. MSc thesis. Dokuz Eylül University. 2004. Guest, A. H. ‘Dance Notation‘. Perspecta. 1990; 26: 203–14. Haeusler, M. H. Chromatophoric Architecture: Designing for 3D Media Facades. Berlin: Jovis, 2010. Ham, J. ‘Music and Architecture: From Digital Composition to Physical Artifact’, paper presented at The Quest for New Paradigms, eCAADe, Lisbon, 2005, 21–4. Hamblyn, R. The Invention of Clouds. New York: Farrar, Straus and Giroux. 2001. Hanoch-Roe, G. ‘Scoring the Path: Linear Sequences in Music and Space’, in Muecke, M. and Zach, M. (eds) Resonance: Essays on the Intersection of Music and Architecture. Lulu Publishing. 2007. Harries, J. G. ‘A Proposed Notation for Visual Fine Art’. Leonardo. 1975; 8: 295–300. Harrison, A., Loe, E. and Read, J. Intelligent Building in South East Asia. London: Taylor and Francis. 1998. Hayles, N. K. ‘Boundary Disputes: Homeostasis, Reflexivity, and the Foundations of Cybernetics’, in Markley, M. (ed.) Virtual Realities and Their Discontents. Baltimore: Johns Hopkins University Press. 1996. Hearn, F. Ideas That Shaped Buildings. Cambridge, MA: MIT Press. 2003. Hensel, M. and Sungurog, D. ‘Material Performance’. Architectural Design. 2008; 78: 34–41. Herbert, D. Architectural Study Drawings. New York: John Wiley and Sons. 1993. Hersey, G. ‘The Renaissance Matrix’. Journal of the Society of Architectural Historians. 1977; 36: 256–8. Hill, J. ‘Storefront for Art and Architecture in New York.’ A Weekly Dose of Architecture. 31 May 1999. Online reference available at http://www.archidose.org/wp/1999/05/31/storefront-for-art-andarchitecture/ Accessed 10 September 2010. Hladik, P. ‘Moving Structure’, in Beesley, P., Hirosue, S., Buxton, J., Trankle, M. and Turner, C. (eds) Responsive Architectures Subtle Technologies. Toronto: Riverside Architectural Press. 2006. Horrocks, R. Art That Moves: The Work of Len Lye. Auckland: Auckland University Press. 2009. Howard, L. Essay on the Modification of Clouds. London: John Churchill and Sons. 1803. Huler, S. Defining the Wind: The Beaufort Scale, and How a 19th-Century Admiral Turned Science into Poetry. New York: Random House. 2005. Ingraham, C. Architecture and the Burdens of Linearity. New Haven: Yale University Press. 1998. Jormakka, K. Flying Dutchmen: Motion in Architecture. Basel: Birkhausser. 2002. Kahn, N. ‘Wind Veil’. 2000. Online reference available at http://nedkahn.com/wind.html. Kaufmann, E. Architecture in the Age of Reason: Baroque and Post-Baroque in England, Italy, and France. New York: Dover Publications. 1955. Kepes, G. The Nature and Art of Motion. New York: G. Braziller. 1965. Korkmaz, K. ‘An Analytical Study of the Design Potentials in Kinetic Architecture’. PhD thesis. İzmir Institute of Technology. 2004. Krier, R. Architectural Composition. London: Academy Editions. 1988. Kristeller, P. O. ‘The Modern System of the Arts: A Study in the History of Aesthetics Part 1’. Journal of the History of Ideas. 1951; 12: 496–527. Kronenburg, R. Flexible: Architecture That Responds to Change. London: Lawrence King Publishing. 2007. Kwinter, S. ‘Landscapes of Change: Boccioni’s “Stati d’animo” as a General Theory of Models’. Assemblage. 1993; 19: 50–65. Kwinter, S. ‘Leap in the Void: A New Organon’, in Davidson, C. (ed.) Anyhow. Cambridge, MA: MIT Press. 1998. 170
Kwinter, S. Architectures of Time: Toward a Theory of the Event in Modernist Culture. Cambridge, MA: MIT Press. 2001. Kwinter, S. ‘The Judo of Cold Combustion’, in Reiser, J. (ed.) Atlas of Novel Tectonics. New York: Princeton University Press. 2006. Kwinter, S. ‘Who’s Afraid of Formalism’, in Davidson, C. (ed.) Far from Equibrium: Essays on Technology and Design Culture. New York: Actar. 2008. Leatherbarrow, D. ‘Architecture’s Unscripted Performance’, in Kolarevic, B. and Malkawi, A. (eds) Performance Architecture: Beyond Instrumentality. New York: Spon Press. 2005. Leatherbarrow, D. and Mostafavi, M. Surface Architecture. Cambridge, MA: MIT Press. 2002. Levy, A. J. ‘Real and Virtual Spaces Generated by Music’. International Journal of Architectural Computing. 2003; 1. Lynn, G. Animate Form. New York: Princeton Architectural Press. 1999. Maher, M. L., Merrick, K. and Saunders, R. ‘From Passive to Proactive Design Elements’, in Dong, A., Vande Moere, A. & Gero, J. S. (eds) CAAD Futures, 2007, Sydney: Springer, 2007. Malina, F. J. Kinetic Art: Theory and Practice: Selections from the Journal Leonardo. New York: Dover Publications. 1974. Malina, F. J. ‘Kinetic Painting: The Lumidyne System’, in Malina, F. J. (ed.) Kinetic Art: Theory and Practice. New York: Dover Publications. 1974. Marks, R. Stained Glass in England During the Middle Ages. London: Routledge. 1993. Massey, J. ‘Buckminster Fuller’s Cybernetic Pastoral: The United States Pavilion at Expo 67’. Journal of Architecture. 2006; 11: 463–83. Mather, D. ‘An Aesthetic of Turbulence: The Works of Ned Kahn’, in Narula, M., Senupta, S., Sundaram, R., Sharen, A. and Lovink, G. (eds) Sarai Reader 6: Turbulence. Delhi: Centre for the Study of Developing Societies. 2006. Mathews, S. ‘The Fun Palace as Virtual Architecture: Cedric Price and the Practices of Indeterminancy’. Journal of Architectural Education. 2006; 59: 39–48. Mayri, O. Origins of Feedback Control. Cambridge MA: MIT Press. 1970. McCormack, J. Impossible Nature: The Art of Jon McCormack, Melbourne: Australian Centre for the Moving Image, 2004. Mennan, Z. ‘The Question of Non Standard Form’. METU JFA. 2008; 25: 171–83. Mertins, D. (ed.) The Presence of Mies. New York: Princeton Architectural Press. 1994. Mitchell, W. J. and McCullough, M. Digital Design Media: A Handbook for Architects and Design Professionals. New York: Van Nostrand Reinhold. 1991. Miyake, T. and Ishihara, H. ‘Development of Small Size Window Cleaning Robot by Wall Climbing Mechanism’, paper presented at the International Symposium on Automation and Robotics in Construction. Tokyo. 2006. Moholy-Nagy, L. Vision in Motion. Chicago: P. Theobald. 1947. Moloney, J. ‘Architectural Science: Literal and Notional Force Fields’, in Proceedings of New Constellations: Art, Science and Society, Sydney: Museum of Contemporary Art. 2006, 31–4. Moloney, J. ‘A Framework for the Design of Kinetic Facades’ in Dong, A., Vande Moere, A. and Gero, J. S. (eds) paper presented at CAAD Futures, 2007, Sydney: Springer, 2007. 461–4. Moloney, J. ‘Kinetic Architectural Skins and the Computational Sublime’. Leonardo. 2009; 42: 65–70. Mostafavi, M. and Leatherbarrow, D. On Weathering. Boston, MA: MIT Press. 1993. Moussavi, F. and Kubo, M. The Function of Ornament. New York: Actar. 2006. Mulder, B., Van Zijl, I. and Rietveld, G. T. Rietveld Schroeder House. New York: Princeton Architectural Press. 1999. Neumeyer, F. ‘Head First through the Wall: An Approach to the Non-Word Facade’. Journal of Architecture. 1999; 4: 245–59. Nouvel, J. Architecture and Design 1976–1995. Milan: Skira Editore. 1997. Oosterhuis, K. ‘A New Kind of Building’. 2005. Online reference available at http://www.haecceityinc. com/homepage.html. Accessed 15 November 2010. 171
Oosterhuis, K. and Biloria, N. ‘Interactions with Proactive Architectural Spaces: The Muscle Projects.’ Communications of the ACM-Organic user Interfaces. 2008; 51: 70–8. Oosterhuis, K. and Xia, X. (eds) Interactive Architecture. Rotterdam: Episode Publishers. 2007. Parent, R. Computer Animation. San Francisco: Morgan Kaufmann. 2002. Payne, A. ‘Surface: Between Structure and Sense’, in Beesley, P. (ed.) Kinetic Architectures and Geotextile Installations. Cambridge, ON: Riverside Press. 2009. Pearsall, J. (ed.) The Concise Oxford English Dictionary. Oxford: Oxford University Press. 2002. Popper, F. Origins and Development of Kinetic Art. London: Studio Vista. 1968. Randl, C. Revolving Architecture: A History of Buildings That Rotate, Swivel, and Pivot. New York: Princeton Architectural Press. 2008. Reiser, J. Atlas of Novel Tectonics. New York: Princeton Architectural Press. 2006. Reynolds, C. W. ‘Flocks, Herds and Schools: A Distributed Behavioral Model’. Computer Graphics. 1987; 21: 25–34. Rickey, G. Constructivism: Origins and Evolution. New York: George Braziller. 1967. Rickey, G. W. ‘The Morphology of Movement: A Study of Kinetic Art’. Arts Journal. 1963; 22: 220–31. Rogers, L. R. Relief Sculpture. London: Oxford University Press. 1974. Romero, J. and Machado, P. (eds) The Art of Artificial Evolution: A Handbook of Evolutionary Art and Music. New York: Springer. 2007. Rhowbotham, K. Field Event/Field Space. London: Black Dog. 1999. Rowe, C. The Mathematics of the Ideal Villa and Other Essays. Cambridge, MA: MIT Press. 1976. Rowe, C. and Slutzky, R. ‘Transparency: Literal and Phenomenal’. Perspecta. 1963; 8: 45–54. Sachs, C. Rhythm and Tempo: A Study in Music History. New York: W. W. Norton. 1953. Saggio, A. ‘Interactivity at the Centre of Architectural Research’, Architectural Design. 2005; 75(1): 23–9. Sakamoto, T. and Ferre, A. (eds) From Control to Design: Parametric/Algorithmic Architecture. New York: Actar. 2008. Saunders, P. T. An Introduction to Catastrophe Theory. Cambridge: Cambridge University Press. 1980. Shanken, E. A. ‘Reprogramming Systems Aesthetics: A Strategic Historiography’, in Proceedings of the Digital Arts and Culture Conference, 2009. Irvine: University of California, Irvine. 2009. Sharpe, L. The Cambridge Companion to Goethe. Cambridge: Cambridge University Press. 2002. Souriau, E. ‘Time in the Plastic Arts’. The Journal of Aesthetics and Art Criticism. 1949; 7: 294–307. Spiller, N. ‘Nanotechnology – the Liberation of Architecture’, in Hill, J. (ed.) Architecture: The Subject Is Matter. London: Routledge. 2001. Steadman, P. Architectural Morphology: An Introduction to the Geometry of Building Plans. London: Pion. 1983. Stengel, R. F. Flight Dynamics. Princeton, NJ: Princeton University Press. 2004. Terzidis, K. Expressive Form. New York: Spon Press. 2003. Terzidis, K. Algorithmic Architecture. Boston: Architectural Press. 2006. Thompson, D. A. On Growth and Form. Cambridge: Cambridge University Press. 1961. Tizonis, A. Movement, Structure and the Work of Santiago Calatrava. Berlin: Birkhauser. 1995. Tschumi, B. Architecture and Disjunction. Cambridge: MIT Press. 1994. Twombly, R. C. (ed.) Louis I. Kahn: Essential Texts. New York: W. W. Norton. 2003. Vesely, D. Architecture in the Age of Divided Representation: The Question of Creativity in the Shadow of Production. Cambridge, MA: MIT Press. 2004. Virilio, P., Lotringer, S. and Taormina, M. ‘After Architecture: A Conversation’. Grey Room. 2001; 3: 32–53. Webb, M. ‘Tradition Stood on End’. Architectural Review. 2005; February 1: 82–5. Whitelaw, M. Metacreation: Art and Artificial Life. Cambridge, MA: MIT Press. 2004. Whiteley, N. Reyner Banham: Historian of the Immediate Future. Boston, MA: MIT Press. 2003. Wiener, N. Cybernetics: Or Control and Communication in the Animal and the Machine. Cambridge, MA: MIT Press. 1965. 172
Wigginton, M. and Harris, J. Intelligent Skins. Oxford: Butterworth-Heinemann. 2002. Wittkower, R. Architectural Principles in the Age of Humanism. Chichester: John Wiley and Sons. 1999. Wolfe, C. ‘Lose the Building: Systems Theory, Architecture, and Diller+Scofidio’s Blur’. Postmodern Culture. 2006; 16(3). Online at http://muse.uq.edu.au.ezproxy.lib.unimelb.edu.au/journals/pmc/ v016/16.3wolfe.html. Accessed May 2008. Wolfram, S. ‘Universality and Complexity in Cellular Automata’. Nature. 1984; 311: 419–24. Woodbury, R. Elements of Parametric Design. New York: Routledge. 2010. Xenakis, I. Formalized Music: Thought and Mathematics in Composition. New York: Pendragon Press. 2001. Zeki, S. and Lamb, M. ‘The Neurology of Kinetic Art’. Brain. 1994; 117: 607–36. Zuk, W. and Clark, R. H. Kinetic Architecture. New York: Van Nostrand Reinhold. 1970.
Note: page numbers in bold refer to figures and tables.
acceleration 64, 67, 73, 87, 90, 94, 128 adaptation, iterative see feedback mechanisms adjustment 43–4, 49 Aegis Hyposurface 20–1, 20, 25, 34–5, 85, 87, 147 agogic 63–4, 78, 87 Alberti, Leon 41, 58, 120, 138 alloplastic logic 25, 27 amplitude 94–7, 102–3, 121, 123 analytical drawings 14, 16–19, 24 animation software 10, 97 animation studies 5, 140 animations: computer 26, 79, 95, 99, 164n12; hybrid 127, 141; stages of generation 92–3 antithesis 42–3, 139 architectural design: and environmental performance 34; and field theory 48 architectural surface 34, 42–4 architectural systems 40–2, 49, 50, 55 architecture: cybernetics in 78; and dance 59; disclosure in 43; disposable 31; limit of geometry on 4; and music 58–9, 67; performative 39, 43; prior knowledge from 92; static 7, 59, 93; temporal pattern in 60, 64; transportable 30 arithmetic progressions (AP) 97, 103, 115–16 artificial life 71 Ashby’s Law of Requisite Variety 52, 86, 88 Aurora Place 43 awnings, responsive 22 Ban, Shigeru 14 Bauhaus 60, 155n61, 160n20 Beesley, Philip 24–5 Bergson, Henri 33, 35–6 Blur building 24 Boccioni, Umberto 45–6 boundary conditions 103, 138 Brett, Guy 60–1 building plans: dimensionless representation of 5; rectilinear configuration of 4 Burnham, Jack 71, 78
Calder, Anthony 60, 63–4, 68 Campus of Justice see Ciudad de Justicia catastrophe surface 47 catastrophe theory 55–6, 81 categorization 61, 64, 121, 133–4, 142 cellular automata (CA): definition of 164n7; in animation experiments 108–14; categories of 72; and flocking algorithms 103, 106; and hybrid patterns 128; and non-ascribed patterns 130, 131, 133; and numerical scale 95 chop: definition of 122; examples of 126; in hybrid patterns 128–9 Ciudad de Justicia 15, 29 cladding 8, 34, 39, 43, 120, 139 classification see categorization clouds 64, 141–3, 141 clustering 70, 92, 94, 96–7, 102, 105 complex systems 47, 81 composition: design see design composition; and indeterminacy 25; kinetic see kinetic composition; part-to-whole 39, 49, 55, 58, 78, 120, 138–9; static 55, 77; temporal form of 44; use of term 3, 35; and whole-whole relationships 49 compositional systems see architectural systems compound states 146–7 concinnitas 41, 58 constructed situations 6 constructivism 7, 57 contrast, significance of 87 control, degrees of 85 control plane 83–4, 86, 88, 94, 99 control scripts 103, 137, 141 control systems: in Aegis 35; design of 9; and indeterminacy 26–7, 95; key variables for 85; spatial configuration of 86; taxonomy of 29–31, 35, 39–40, 80, 85; theory of 51 see also cybernetics; of water walls 24 control variables 85–6, 93–4, 97, 99, 101–3 conversation theory 52 curtain walls 9, 42–3, 49, 50, 55, 78, 138 cybernetic devices 35, 51, 80
cybernetics: and adjustment 49; in artworks 64; central metaphor of 53; contemporary interest in 70; and control systems 40, 83; and design 85–6; and feedback 52, 54; firstorder 51–2; second-order 51–2, 54, 71, 85; and systems theory 71; third-order 54 D-tower 163n9 dance notation 59 D’Arcy Thompson 4, 48, 79, 91, 164n16 data: input 9, 82; sampling see sampling; sources 36, 80, 83–5, 96, 99; types of 84 De Stijl 45 decision planes 79, 82–3, 86, 89, 90–2, 94, 96 dECOi 20, 25–6, 34–5, 85 deformation: incremental 7; material 7, 13, 31, 81, 86, 88, 93, 99; by pneumatic structure 30; of scale 21; of text 25 design, tri-planar model of 88, 92 design composition see composition design experiments, of 1970s 13 design language 10 design research 3–4, 8, 14, 53, 120, 149 design science 4, 28, 52 design software 48–9 design solution space 79 design variables: combination of 92; as continuum 82; developing 4; for experiments 98; identification of 82; influence of 37; manipulating 5; mapping of 88, 90; models for considering 35, 39, 56–7; and pattern 9–10; relationships between 78; of temporal structure see temporal variables; and unpredictability 64 design vectors 79 designers, role of 25 diagram 35 digital controls 85 digital mapping 59 digital technology 14, 48–9 Digital Water Pavilion 23–4, 24, 147 Dorin, Alan 57–8, 71–3, 78, 97 duration 33, 35–6, 64–5 dynamic structures 14–15, 25, 29 dynamic systems 46, 72, 81, 140 Dynamic Terrain 21 dynamism, plastic 40, 45–6, 48, 55 Eco, Umberto 121 eddy: definition of 122; example of 126; granularity of 134; in hybrid patterns 127–8 elasticity 7, 45, 81, 86 embedded structures 13, 29 emergence 51, 54–5, 83, 94 energy constructs 66 Eno, Brian 27
envelope 8 environmental control 69, 83 environmental control systems 27, 34, 53 see also sunshading environmental performance 27, 34 essential configuration 97, 164n16 experiment, definition of 91 experimentation, intuitive 3, 92, 102, 105–6, 115–19, 143 facades: definition of 77; and communication 83; composition of 23, 56–7, 77, 97, 99, 137; contemporary approaches to 13, 139; evolution of 41–2; free 9, 42–3, 49, 120, 138–9; freestanding 39; intelligent 9, 28; and kinetic art 10; media see media facade; patterns in see kinetic facades, patterns of movement in; pneumatic 19; proportions of 99–100; responsive 34, 44; static 34, 39, 99; study drawings of 93; temporal operations of 20; urban 8, 34; use of term 4, 8 feedback mechanisms 51–4, 81, 85 Fibonacci series 58, 137 field 144, 145, 147 field thinking 46, 48, 139 fineness 49, 55 flare 21, 26 flaring 105, 165n3 flight dynamics 106, 165n4 flocking algorithms 30, 95–7, 103, 106, 108–14, 130, 134, 148, 164n10 flockmates 164n10 Focillon, Henri 91–2 fold 27, 47, 144, 145, 147–8 form, conceptions of 46 Foucault, Michel 105, 165n1 Fox, Michael 29, 31, 40 fragments 144, 147–6 Frazer, John 40, 51, 53–4 Fuller, Buckminster 33, 156n68 Fun Palace 8, 36 The Function of Ornament 13, 34, 139 functionalism 9, 44 Gabo, Naum 60, 62, 65, 68, 87 Gage, Stephen 23, 40, 51–3 Gaulthorpe, Mark 20, 25 geometric progressions (GP) 97, 103, 115–16, 125–9 geometric transformations 5–7, 31, 70, 81, 86–7, 92–3 Glanville, Ranulph 40, 51–3 Goethe, J. W. von 4–5, 91 graduations 41, 49, 55, 140 granularity: of controls 52; digital 9, 87; of facades 70, 73; of sea patterns 122, 134; tectonic 88, 93
Haeusler, M. Hank 27 Hayles, Katherine 51, 54, 71 hierarchical differentiation 42, 137 Hoberman, Chuck 15, 29 homeostasis: and control systems 85, 94; in cybernetics 51–2, 85, 94; and kinetics 54 Howard, Luke 141–4 human perception 10, 87, 100, 120 humidity 23, 83 hybrid patterns 106, 123, 127, 141 Hylozoic Ground 25 Hyperbody research group 14–15, 30, 32 ICO (input-control-output) 29, 80, 82 indeterminacy: and cybernetics 52, 94; degrees of 35, 85–6; and facades 25–6 indexing 3, 9, 92, 101, 103, 105–6 input-control-output (ICO) 29, 80, 82 input-processing-output (IPO) 30 Institut du Monde Arabe, Paris 28, 33 intelligence, in buildings 28 intensities 49, 72, 138 Interactive Architecture (iA): definition of 30; Gage on 52–3; overview of 13 interactive media 83 interactivity 30, 33, 53 intermediate states 148 Italian Futurism: and dynamic form 7, 39–40; and field theory 46–7; Gabo on 66; and movement 45–6, 62 Joachim, Mitchell 22, 25 Jormakka, Kari 31–3, 35–6 Kahn, Ned 22, 26, 36, 85, 87, 147 Kaufmann, Emil 40–1, 45, 55, 58, 135–6 Kiefer Technic 16, 147 kinetic aesthetics 31 Kinetic Architecture 31, 32, 80 kinetic art: characteristics of 68–9; classification of 57, 61–2; forms in 5, 10; history of 60; movement sequences in 97; prior knowledge from 92; process for see kinetic process; sculpture as 65–6, 69; theoretical contributions to 60–1, 67, 70, 78; threedimensional 62–3; and time 72 see also virtual-kinetic art kinetic composition: and adjustment 44; and diagram 35; and indeterminacy 35; range of 13; systems of 55–6 kinetic design: and digital technology 49; origins of 9–10; and structures 14 kinetic facades: agenda for 26; and change 136–7; compositional potential of 36; and cybernetics 52, 54, 64; and data sampling 83; design of 69, 80; and
disposable architecture 31; examples of 19, 27; and field theory 47, 56, 120; motivation for 57; movement patterns in 69–70, 77, 87, 99, 141; multiplicity in 36; as processin-action 80; range of 141; scarcity of 9, 33; spatial configuration of 94, 96; and time 58 kinetic form: design variables in 77; indeterminacy of 25; morphology of 36; and planar model 47; range of 5, 9, 27, 33, 92; theoretical range of 5; visualization of 56 kinetic formation 5, 11 kinetic morphology see kinetic pattern, morphology of kinetic pattern: in animations 93; and control systems 29; distinctive qualities of 4; dynamic of 144, 147; factors influencing 11, 35, 37, 73; and field 144; formation of 86–7, 143; intensity of 77; and mediation of data 85, 96; morphology of 5, 9–11, 35, 56, 66, 82, 149; range of 11, 77, 91, 99, 105, 133, 144; sampling and 84; study of 97; taxonomy of 120 (see also kinetic types); and temporal scale 97; in variable space 82 kinetic phrases 66 kinetic process: artists working with 97; in digital art 72; in photography 47; taxonomy of 57–8, 72–3, 83, 88, 96 kinetic range see kinetic pattern, range of kinetic reliefs 19–22, 26, 34–5 kinetic resonance 8, 70, 78, 105 see also resonance kinetic rhythms 62 kinetic screens 13, 16, 84 kinetic structures: and adaptable spaces 31; taxonomy of 13–14, 29 kinetic types: basic 93; compound 86, 101–2, 106; range of 107, 137 kinetics: definition of 3, 31, 86, 99; in 1960s 30; building blocks of see geometric transformations; in contemporary architecture 13; and control systems 40, 51; ephemeral nature of 91, 93; experiments in 9–10; and fields 46; and immediacy 67; implementation of 87, 149; indeterminate 26–7, 55; natural 62–3, 68; as new genre 5; planar model of 47, 56, 81–2, 92; potential of 8–9; spatial 13, 23, 65, 73, 88; technology of 30, 77; theory of 25, 61; virtual 72 Kwinter, Sanford 40, 46–9, 55–6, 81, 139–40, 157n30, 158n35 latency 101 Le Corbusier 6, 32, 58–9, 159 Leatherbarrow, David: and adjustment 49; and free facades 138; and performative architecture 39, 43–4; and spontaneous qualifications 51
light: control of 28 (see also sunshading); in kinetic art 62; as material 139 light sensors 84 LIGO Science Education Centre 17, 18 linked assemblies 79, 86 louvers 16, 27–8, 86 Lumidyne system 60 Lye, Len: notation of movement 62, 66, 92; tangible motion forms of 64–5, 97, 105; as theorist 57 Lynn, Greg 48–9, 56, 79, 158n35 machines: picturesque 53; taxonomy of 31 Malina, Frank 60–1 Malvern Hills Science Park 17, 85 Martin, Kenneth 63 mass 7, 62, 69, 81 materiality 5, 9, 86, 93, 99, 149 materials, weathering of see weathering media facades 7, 9, 25–8, 83 MIT (Massachussets Institute of Technology) 29 mobiles 60, 62–3, 68 modernism 9, 39, 42, 44–6, 49, 138 modification 141–4, 167n16 modularity 49 moiré 66, 101 morphology: of kinetic pattern see kinetic pattern, morphology of; study of 90–1; urban 4, 151n4; use of term 3–5; variables influencing 11 motion see movement motion graphics 7, 48, 95 motion sensors 24, 83 movement: aesthetic of 55, 57, 63–4; in architectural theory 6–7; basic types of 10, 70, 73, 86, 93, 106, 123; capturing 45; design of 9, 67, 69; Jormakka on 32; language of 67–9, 120; measure of 32; of multiple objects 120, 127, 143; notation of 59, 66; perception of 31, 87, 100–2, 100, 155n58, 165n18; poetry of 3, 30, 39, 149; relative 55, 70, 78, 93, 137; representation of 6–7, 45, 62, 152n27; sense of 33; and stasis 8; three-dimensional 63; vocabulary of see movement, basic types of movement sequence see sequence, of movement multiplicity: continuous 36, 39; qualitative 36, 77 music: taxonomy of 134; temporal notation in 166n1 musical harmony 41, 58–9, 159n6 nanotechnology 22 noise algorithms 95–7, 103, 106, 129–30, 131, 133 non-ascribed patterns 123, 130, 131, 133–4 non-hierarchical relationships 42, 137–8 non-linear systems 55–6, 139 Nordic Embassies, Berlin 17, 28, 86
Nouvel, Jean 28 numerical scale 93, 100–1, 103 occupancy 6, 58–9, 83 Ocean North 22, 87 Oosterhuis, Kas 14, 29–30, 32 operable surfaces 20–1, 44 parameters, number of 81 parametric design 9, 46, 48–9, 79–80 Pask, Gordon 52–3, 158n49 patch 144, 148 patina 6 pattern: definition of 105; abstract 10, 70, 125; and facade composition 57; hybrid see hybrid patterns; kinetic see kinetic pattern; morphology of 3, 9, 99; spatial dimension of 86; temporal 64; use of term 4 peak: definition of 122; in hybrid patterns 128–9 performative surfaces 39, 44 periodic structure 72–3, 79, 82–3, 87–8, 88–9, 92, 96–7, 99 pitch 67, 73, 120 planes, flow on 47, 56 plastic zones 46 pneumatic structures 14, 19, 30, 86 Poincare, Henri 46 Popper, Frank 57, 60–3, 78 process: art of 71; complex 47, 72, 97; kinetic see kinetic process propagation 46, 70, 78, 93–4, 96–7, 99, 105 proportional relationships 40–1, 58, 93, 103, 125, 137 pulse 72–3, 88, 90, 97 qualifications, spontaneous 51 qualitative discrimination 36 random numbers 95 reactive surfaces 26 Realist manifesto 9, 60, 62, 65, 68, 71 reciprocity 39–41 reflexive systems 26, 52 reflexivity: in cybernetics 52, 85–6, 94; and kinetics 54 Renaissance 8, 40–1, 58, 137 repetition 42, 49, 55, 101, 137–9 representation, project of 39, 42–4 resonance 5, 47–8, 52, 70, 120, 138 see also kinetic resonance responsive systems 26 reverberation 42, 55 rhythm 59, 62, 64, 87–8, 105–6, 120, 134 Rickey, George: kinetic art of 5, 10, 60, 63–4, 87, 92, 97; as theorist 57, 66–70, 78, 120, 133; and constructivism 159n1, 161–2n63; and morphology 3
ridges 122, 125, 127–8, 144, 145, 148 ripple 121 robotic systems 23, 53 roll 7, 67, 73, 101, 106, 112, 120 Ronchamps 44–5 rotation: definition of 7; in animation experiments 106, 109, 115–19; and dynamic structures 14; and screens 16–18; and swell 123 sampling 35, 61, 82–4, 91 sampling plane 83, 88, 96 Sant’Elia, Antonio 45–6 Sartre, Jean-Paul 63 scaling: definition of 7; and modernism 55; and operable surfaces 21; and pneumatics 86; and screens 19 Schroeder house 43, 55 scissor joints 15, 29 sea nomenclature 77, 106, 120–1, 124, 131, 133–5 sectional profiles 7, 93 self-regulating systems 51–2 sequence: of movement 70, 97; in syntax of motion 67–8 shading, adaptive 15 shape memory alloys 22–3, 85, 154n23 ships, movement of 67, 69, 73, 106, 133 Shop Front for Art and Architecture 20 significant form 68 skins 4, 8, 22, 25, 28, 42–3, 140 smooth surface 7, 45, 138 socio-cultural context 80, 83 space, discontinuous 32 spatial conceptions 45 spatial scale 106, 121–2, 134–5 spatiality 44, 55, 94 speed 46, 87; perception of 155n58 (see also temporal scale) spring 106, 114 state change 36, 77, 131, 143–5, 146, 147 Steadman, Philip 3–5, 91, 97, 164n16 stream 72–3, 88, 90, 97 structural components 15, 29 sun-tracking 28 sunshading 17, 19, 27, 33–4, 84–5, 120 see also shading Super Cilia Surface 22 superimposition 55, 62, 91, 139 swaying 105, 165n3 swell: definition of 122; in animations 123, 125, 143; counter 131; granularity of 134; in hybrid patterns 127–8 symmetry 4, 41, 69 synchronous activation 96–7 systems aesthetics 71, 78 systems theory 40, 70–1 see also cybernetics
tangible motion forms 65–6 taxonomy see categorization tectonic plane 88, 93 tectonic variables 83, 88, 92 tectonics: and indeterminacy 27; use of term 82, 86 temperature 81, 83, 85, 96 temporal arts 57–8 temporal dynamics 48, 144 temporal phenomena, taxonomy of 136 temporal scale: as design variable 87, 89, 90, 96; and kinetic art 64; and operable surfaces 20; and pattern formation 87, 97, 99; and weathering 31 temporal structure see periodic structure temporal variables 70, 87, 88, 90, 96 three body problem 46, 157n30 time: as design variable 81–2; irreversibility of 69; and motion 32; perception of 67 transformation: perception of 140; speed of 31 translation: definition of 7; and dynamic structures 14; and screens 16 transparency 53, 99, 139 Tschumi, Bernard 6 twist 7, 93, 101, 106, 111 typology 140–1 unpredictability 62–4, 78, 95 Van der Rohe, Mies 55 variable space 79–80, 82, 90–3 virtual environments 14, 27 virtual-kinetic art 58, 72–3, 78, 88 voxel facades 27 wall robots 23 walls 5, 8, 14, 20, 44, 136 see also curtain walls; water walls; wave walls; wind walls water walls 23–4 wave: definition of 122; in animations 125, 126, 143; dynamic of 144, 147; in hybrid patterns 127–9; shape of 144, 145; transition from 147, 148 wave walls 17 weathering 6–7, 31, 39, 44, 69, 120 Weiner, Norbert 51 whole-whole relationships 49, 140 wind, data on 83 wind walls 22, 26, 36, 85, 87, 147 windows: in drawings 93; operable 20, 43, 55 wood 23 Xenakis, Iannis 58–9 yaw 67, 73, 101, 106, 113, 123 Zaragoza pavilion see Digital Water Pavilion