suckerPUNCH will be posting interviews in a new section – in the ring
the first five interviews will also be featured in the upcoming publication of TARP, coding parameters. TARP is the architecture manual published by PRATT.
image: swarm matter by kokkugia
images clockwise from top left: taipei performing arts center, fibrous tower, taipei performing arts center, taipei performing arts center all by kokkugia
suckerPUNCH: Could you tell us a bit about starting your practice, Kokkugia?
roland SNOOKS: The practice is a collaboration with, Jono Podborsek and Rob Stuart-Smith, which started out in Melbourne in 2004. Although the three of us had worked on projects together previously, the practice needed a critical point to congeal into a formalized entity and for us this was an invitation to an exhibition. Our background in generative design and algorithmic work developed predominately while we were students and subsequently instructors at RMIT. The practice evolved in part through our common interests as well as through the studio space that we shared. RMIT is curious in that it had no dedicated studio space and consequently at that point in time students would band together and rent space in the city. After we graduated no one wanted to leave the space so series of small practices emerged. It’s an interesting model.
sP: What work and research are you currently exploring in the office?
rS: One strand of our work over the last 7 or 8 years has been focused on agent-based methodologies which is central to my research. This started as an interest in generative design, not necessarily as a specific interest in computational, algorithmic or scripted work, but as an interest in understanding the emergent nature of public spaces. At least that is how the agent-based work began. We were looking at how the most successful public spaces in Melbourne are emergent as opposed to ones that are prescribed. For example all of the public squares in Melbourne fail and are continually redesigned, while the successful public spaces are small laneways which originally operated as service access and have since been appropriated by bars and cafes. We became interested in the emergent public spaces of Melbourne and how we could develop emergent methodologies. That led us to develop swarm systems and multi-agent models. This offers a way that design intent operates at the micro level through encoding a series of entities that interact to give rise to complex order at the macro scale. This interest keeps rearing its head at different moments within our work. Some of the earlier projects were agent based and then in the last 2 years we have started to revisit that work more intensively. Consequently, a lot of our current work is agent based. We are working on collaboration with Tom Wiscombe for a pavilion in Korea which is using swarm systems and we are also conducting ongoing research work with Cecil Balmond.
sP: How do the consultancy and academic research overlap with your own work?
rS: Originally Kokkugia was set up as a research practice however our interest is shifting toward building. At the time we realized research was the primary way we could continue to develop new ideas and design strategies. That research now partly happens within institutions. I am involved with the NSO, Non-linear Systems Organization which is headed by Cecil Balmond at the University of Pennsylvania School of Design, so some of the research operates there. Some of the research operates through consultancy with people like Cecil. We also run a computational consultancy where we collaborate with architects such as Michel Rojkind as well as working on projects as diverse as car design with Audi/VW. The consultancy is a way to support some of our more indulgent research interests. Ultimately we want to develop the practice to a point where it can build complex projects.
sP: Could you talk about what you are currently exploring in your academic seminars and studios? What is the relationship between the academic work and that of Kokkugia?
rS: The academic work increasingly is closely following the work we are doing in the office. It has evolved into a very tight relationship between the two. There has been a series of seminars I have been running under the title Swarm Intelligence which is a continuation of this work with multi-agent systems. This started its life at Columbia University as a series of seminars and more recently has expanded to UPenn and USC as seminars and studios. The other teaching work is the collaboration I am doing with Cecil Balmond. We teach a seminar in the fall and a studio in the spring at UPenn. This partly involves the multi-agent methodologies but also has broader interest in nonlinear design and non-hierarchical systems.
sP: Do you test problems and incubate new ideas that have begun in the office in your academic work or do ideas begin more in the academic realm? Do the two strains of research parallel one another or do they remain autonomous?
rS: It is a bit of both. When we run a seminar on multi-agent systems we bring 8 years of experience and development from Kokkugia. But whenever someone else tests it there is always a new way of imagining it, rethinking it, or applying it, so things emerge from the teaching. So teaching becomes a form of research, another environment where ideas are tested. Consequently some aspects of the research are very closely aligned and others diverge.
sP: What do you see as the differences in algorithmic research and practice between the United States, Europe and Australia?
rS: I do not know if geographically it makes a significant difference. There has been a long history in Melbourne of algorithmic work that dates back at RMIT to the late 80s and early 90s. Mostly this work was never published outside Australia. Paul Minifie is a leader in this group, he developed some very sophisticated computational work in the 1990’s and is responsible for teaching a generation of young architects algorithmic design – this is perhaps the most significant influence on the early development of my algorithmic design research.
sP: Do you see a shift in the teaching and practicing of algorithmic work form when you were studying to now that you are teaching?
rS: There has been a shift in terms of enthusiasm or popularity of computational design in the last 5 or so years. When I studied at Columbia there was only a handful of students who took algorithmic design seriously. However in recent semesters from an anecdotal standpoint I can see an increased enthusiasm for the work that is probably partly due to fashion and partly due to a larger shift toward engaging complexity and computational techniques. It is also becoming more accessible. There is a whole set of tools that make it easier for people to engage with computational design. However this dissemination of tools is also blurring the reasons for why much of this work and research began. Our enthusiasm for this work came from an interest in the emergence of public space and has evolved into a more complex set of interests. Our interest was never focused on technique, or the digital, or a fetishisation of scripting, instead our focus has been on methodologies that generate complex systems and emergent outcomes. I think in recent years an obsession within architecture has emerged with the tools rather than an interest in why those tools might be deployed. This a concern to us. In some ways I would describe our interest as developing nonlinear methodologies of design. Of course design has always been nonlinear. In any manual process there has always been feedback to a certain degree between different influences on the design project. With the shift towards digital tools that is getting eroded and designers talk about the nonlinear but actually design is becoming more linear. A lot of this is due to the obsession with parametric tools. These are of course incredibly useful but at the same time force a hierarchy within the design process that is not the way designers are accustom to working. While they are useful in design development, I do not consider these to be useful tools in sketch design or early stages of a project. There is often a conflation of terms between generative, parametric, algorithmic, and scripting, which are too often seen as interchangeable. I consider parametric and emergent as polar opposites. Within parametric hierarchical tools all possibility is given within the starting condition, while emergent conditions arise from non-linear systems such as multi-agent models. In the last few years we have witnessed a dissolution of generative intent within the popularity of what are increasingly constraining techniques.
sP: How do you see the role of open source networks or your wiki which shares codes effecting accessibility to computational design?
rS: Open source is critical for this part of the discipline to advance. So we have always had an interest in posting a lot of code on our wiki and making it available. At the same time we realize there are problems with this, in particular the misunderstanding of the intention behind our methodologies. Many algorithmic tools are being seen as black boxed techniques in the way that Maya or Rhino or any other commercially available software has a set of black box techniques. In some respects the scripts we produce are not as robust or sophisticated as the code that underlies something like Maya, nor as generalized. Their operation is specific with specific design intention. A dominant reason for writing the code is to be liberated from black box software and to develop the underlying behavior of an algorithm. If the algorithm is taken as found technique we fall back into same problem we always had with reliance on software and consequently a deterministic relationship between the tool and that which is generated.
sP: We have recently been noticing a departure from technique driven work and a return to
architectural ambition as a generator. How do you see the role of architectural ambition in your work?
rS: One thing we strongly advocate with our students is that the algorithm should emerge from the architectural problem rather than simply the architecture emerging from the algorithm. We believe the behavior of the algorithm has tight relationship to the formal and organizational characteristics that are generated. If design operates through the selection of algorithms there is already an a priori understanding of the type of formal characteristics you are interested in, where as if you are interested in developing emergent characteristics, then it is the behavior of a procedure must be designed. Another aspect of this is how tightly related the algorithm is with the architectural problem. That really is an issue of intent. Algorithms often work in a deterministic way where there is a linear relationship between the input and output and often this does not have the sophistication to enable you to embed any architectural concern within the process. My interest in this work is how the intention of the designer is embedded in the procedural or algorithmic process and how this intent self-organizes as opposed to simply critiquing the output of these processes. For us the ability to embed intent is critical to the design process.
sP: At what point does aesthetic criteria enter into your design process? Is there a defined line between explicit design vs generative design in your work? How do you negotiate the explicit versus the generative in order to challenge normative architectural conditions?
rS: This negotiation is something that happens the whole way through a project. Design operates through designing the algorithm, which is not something that happens before designing the project. As you begin to design the procedure it produces things that you critique, this in turn informs the design the algorithm. There is this constant feedback within the process. Part of the criteria for that is always aesthetics. We are beginning to talk about the design process as being the organization of matter. We can think about form simply as organization. In some ways I think a lot of algorithmic work is seen as a formal exercise. Even in our earliest work the interest was in programmatic self-organization rather than formal self-organization. We are beginning to blur what those distinctions are. To get back to your point about what is explicit versus what is generative, when we started the practice there was a certain desire to be able to see the whole of architecture as a generative concern. As we progressed we realized that parts of architecture would certainly benefit from working through complex systems, such as resolving a complex problem or the generation of certain types of characteristics but not all problems are best suited to self-organization. So consequently we are increasingly looking at where we use these strategies. This is not a comment to sideline that work and to suggest that generative design is incapable of solving many problems, but simply that some problems are more efficiently solved through explicit techniques.
sP: We are seeing a lot of young designers shifting from the all or nothing attitude in terms of scripting and algorithmic design. There seems to be more of a strategic deployment of explicit and generative strategies in order to best solve architectural problems. How do you integrate both into your work?
rS: Part of it is about authorship as well. Within some designer and students we see a desire which is ideological or even religious that everything must be somehow scripted, however we have never thought of it this way. We are not purists. We started writing code to try to link obscure pieces of software together to get them to talk to each other and the further down that road we went we realized it was easier to write our own algorithms than try to link other software. We have always been interested in design and not in writing code. Throughout the development of our practice we have developed more expertise in coding but it has never been an end in itself.
sP: As part of this shift in the attitude away from one of all or nothing, designers are pinning their hopes less on one technique, software or process and bringing bigger design intent to current projects. How do you see the role of technique in your work?
rS: We have seen a shift from the work coming out of Columbia in the mid 1990s which has often been labeled as Greg Lynn’s project or animate form project which was clearly an obsession with the possibilities of procedural design and certain software and algorithms. I think the generation, or half generation after that are primarily interested in the affects these techniques generated rather than the generative methodologies themselves. So we have seen a shift from an interest in generative work to that of affect and form and now this seems to be shifting back in the next half generation who are looking more seriously again at the possibility of generative design. I think part of that is because the tools have become more sophisticated enabling more design intention to be encoded in the system and therefore more carefully curated architectural concerns can be dealt with.
sP: What do you think of the shift from pure formal based design and shiny surface affects to a more sophisticated curated ambition of materiality and atmosphere and effect? Does this play a role in your work?
rS: We are interested in that shift and less interested in shiny surfaces of 90s. Inherently the processes we work with develop a topological intensity as well as an intensity of articulation that resists the single surface. We are making things that are topologically very complex and this leads to a whole set of other affects. When I talk about topology it is not necessarily spatial topology but also micro topology operating at the level of articulation. We have always been interested in the formal characteristics that are generated by algorithmic design. Increasingly our interest is shifting toward higher population algorithmic work that leads to a fibrous aesthetic. More crucial to us than an interest in sensation and affects is an interest in the way this type of design might affect tectonics. Architecture has always been highly tectonically articulated and partitioned into discrete system and this has been particularly emphasized within classical modernism. We are interested in how non-linear algorithmic work is able to break down some of these hierarchies. Instead of the a priori assumptions of form follows function, or function follows form, structure follows surface and ornament follows structure, we are looking at how these relationships can begin to interact in a non-hierarchical way. How does structure negotiate its relationship with ornament for example. In doing this we are interested in the way normative hierarchies begin to dissolve and an entirely different reading of form and articulation emerges. One that has a certain resonance in terms of the affects it generates. So in a way it is an interest in affects but not simply looking at surface affects but instead the whole way tectonics begin to re-negotiate.
sP: in your Taipei Performing Arts Center there are several strategies and scales of interventions and a layer of styling through color and ornamentation. Could you talk about how the approach to this project ties into those ideas?
rS: It is probably the most fragmented of all our projects and pluralist in some ways. Part of this is a design concern where we are actually interested in pluralism. We have a postmodern upbringing, a strange form of post modernity that existed in Melbourne that is less about historical pastiche and more about a serious and literal concern for complexity and contradiction. At moments within our work this pluralism comes out. But it is also a logistical issue. The project was designed between London and New York. A lot of our projects are designed in one of those offices and led by one partner. This project was a relatively large project so we tried to work on a way of splitting the project up which is why it almost has a collaged relationship. This is not a trait that we would typically identify as something we are trying to explore. It also has a classic hierarchical diagram underlying the project that is something in our more recent work we are trying to undermine. It is in some ways an anomaly in our work.
sP: Is the Fibrous Tower more in line with this research?
rS: It is closer in terms of our interest in trying to dissolve hierarchies or exploring the continuity of systems within architecture and also I would argue that the fibrous tower has more emergent affects and fewer a priori inputs. I would argue that it is more of an emergent project.
sP: You have referenced your postmodern background. What is relationship to history and precedent in your work?
rS: Everything we do comes out of our understanding of architecture. Our work is about architectural design first and contemporary techniques second. Our background comes from a combination of postmodern and modernist training. All of our work emerges from that understanding but increasingly we are interested in looking at architectural problems through non-linear operations. In part this is manifest through our interest in tectonics and the relationship between systems such as structure and ornament. Essentially it is an attempt to deal with some of the fundamental questions architects have always dealt with. The seminars that I teach are increasingly focusing on historical precedent in an attempt to challenge codified relationship and hierarchies. Often when people are dealing with new techniques there is a desire to reject history and claim a break from precedent, this is certainly not something we indulge in.
sP: The Fibrous Tower seems to take on issues beyond surface articulation such as poche, aperture, and enclosure in the thickened skin. Rather than just being surface application or styling exercise, it takes on fundamental questions and problems of architecture at the scale of a building rather than stand alone formal exercise. How does your work intend to deal with the conventional problems of architecture?
rS: It depends on the project and how we frame the research. The first iteration of the Fibrous Tower emerged out of collaboration with Michel Rojkind. The project was set for construction in Guiyang China, however it stopped in sketch design at the time of the global economic crisis. However we decided to continue developing the research on structural skins. This project emerged from a concern for building complex structures. Our interest was in how far we could push normative construction techniques in terms of the complexity of geometry they are capable of creating.
sP: We are seeing a continuing interest in fabrication and materiality as a final output rather then simply the production of an image or renderings. Where does materiality and fabrication enter into your work?
rS: We are very interested in building. We come from a background where we all worked in practices for numerous years that are involved in the construction of large building projects so we have an understanding of construction and fabrication. However within the practice we do not have an explicit research agenda in terms of fabrication. We are seeing many young practices and schools developing specific fabrication agendas, which of course is important when the fabrication agenda has a close relationship to the design methodology. Some of the tectonic assemblages we are developing at Kokkugia with non-linear methodologies are only feasible if they are related to certain emerging automated fabrication techniques. However I think a lot of the fabrication work developed by architects is lagging a long way behind the modes of construction that exist in the progressive end of the building industry. The interest, which seems prevalent within architecture schools for laser cutting or milling, is an obsession with the conversion from the digital model and how things can be broken down into components, labeled, unrolled, fabricated, and assembled. There is a certain excitement about this and we appreciate the desire for students working computationally to actually build things, rather than simply produce digital representations. At the same time there is a certain naivety to this, which comes from lack of understanding of what actually happens in the construction industry. Some sectors within the construction industry are incredibly advanced. However the idea that we make a series of components, label them, and then reassemble them on site is something that has been happening since at least the Abbey at Mont-Saint-Michel. The Abbey was entirely fabricated off site and assembled based on a numbering system. I think the most progressive work in fabrication and construction by architects is the research and development of custom robotic and direct depositing machines. Our interest is not so much in small-scale fabrication but instead in the possibility of entirely new fabrication techniques that have a close relationship with design methodologies that could lead to a rethinking of tectonics and form.
sP: Do you approach your projects that deal with the scale of urbanism differently then you do your projects for a single building?
rS: Conceptually no, because conceptually what we are interested in is looking at design from its smallest element and the way that generates complex order at the macro level. At the urban scale we are trying to work out what is the urban monad in a way, what is the smallest component. Intention definitely shifts. At the urban scale these systems need to operate more as an ecology. There is an ability for systems within architecture to operate more discreetly even though discrete articulation of these systems is something we are trying to erode and develop non-linear modes of feedback, these of course are essential for urban design.
sP: Designers have a long history of borrowing tools and techniques from other disciplines. How do you use this appropriation in your practice and research?
rS: We are as interested, or as opportunistic, or as guilty, as anybody in trying to import ideas, techniques, or methodologies. We certainly look at computational biology. There is a whole series of theoretical interests we borrow from cultural theory or philosophy. What I am more interested in is how the discipline of architecture can begin to generate ideas or techniques that can feed into other disciplines. Which I think to a certain extent has happened but we are net thieves, we thieve more than we give back. Within the current methodological research, we are interested in how can we develop algorithms that are not simply borrowed from elsewhere but have their own characteristics and ability to solve certain problems. I am interested to see if these algorithms have the robustness to proliferate into other fields.
sP: In terms of cultural resonance, do you see swarm based work as being accessible to a larger audience?
rS: Architecture always emerges out of wider intellectual and cultural milieu. Culturally, society has an interest in complexity that has been growing in momentum since the 1960s in the way we understand formation. We are drawing on the same sources that many fields are drawing on so consequently I think what we are doing is a product of its time and has a resonance across a whole series of other fields. I see certain resonances, even if they are not direct relationships, between the methodologies that we are exploring and the underlying concepts of Antonio Negri’s multitude or social networking or formation in biology. I think there certainly is a resonance in agent-based design in the popular consciousness and in fields outside of architecture but it is not because what we do is necessarily being absorbed it is simply because it has some relationship to the wider cultural milieu.
sP: Whose work is on your radar?
rS: I am looking more closely at highly ornamental work such as art nouveau and baroque that relates to the research we are doing with our students. I have also been spending more time revisiting Mies.
sP: What are you reading/listening/watching?
rS: I am rereading Multitude as well as a book on self organization in biological systems.