A taxonomy of taxonomy.

8/8/202313 min read

Posterior taxonomy (= cartography)

When I first introduced the concept of urban cartography on the basis of flow to a group of architecture students, their initial reaction was provoked by the logic behind this thinking. To them it made perfect sense to chart a given urban complexity in terms of its capacities rather than its properties. In this type of mapping, shopping centres become conductors of human activity: the shoppers arrive by car (medium-speed intensive task) and transform into window-gazers (very slow speed, passive, rewarding activity). They would typically stay in such a place for an hour and a half, and then leave to become medium-speed motorists again. When they make purchases they transform into money outlets, while the store becomes a conductor of cash that accumulates into a bigger stream, which is later taken away by a security van. The van arrives empty (low-priority) and leaves filled with the day’s turnover (high-priority).

An airport is a transducer of even greater capacity: it transforms medium-speed elements (motorists parking their cars in the parking lot) into high-speed elements (passengers). The biggest transformation is that it reduces the number of elements (many people and goods in one plane) and transversely increases another type of element (one plane carrying many different types of goods and people). It is an exchange portal of intercontinental magnitude. Analyses such as these are potentially infinite and work through all the four domains that I have called ‘scapes’. When traversing a certain area, all the familiar constructions, infrastructures and places (especially what Marc Augé calls non-places) can be cartographically charted in a more abstract form.[i] The taxonomy of the urban fabric on the basis of intensive thinking proves to be a very useful tool for transforming a practical task into an abstract mental exercise, and vice versa. All urban elements, big or small, known or unknown, assume a completely different significance when judged on their capacities rather than their properties. This type of analysis provides a solid basis for the further development of concepts and interventions, as has been demonstrated in many student projects I have guided over the years.

The second stage of the students’ analysis was more problematic. After the initial taxonomy, the process came to a standstill. Classification at a more refined level proved to be more demanding. How, for instance, should one deal with different types of hybrid allowances (eating/driving, flying/reading and so forth)? On closer inspection, many ‘targets’ in the mapping-process became more and more difficult to categorise because of the discrepancy between their functionality and their actual usage. If these ‘targets’ were to be described precisely, it became clear that each element would require its own specific category, which for taxonomic purposes does not work, and is, in fact, a contradiction in terms: taxonomy needs a certain level of abstraction to be effective. Automatically, the students turned to familiar and proven methods to fill the gaps created by the classification. Programme, function, demographics and narrative are familiar (representational) tools for architects, so it is understandable that students automatically revert to them when they are baffled. In order to overcome this, I shared examples from my own urban explorations in order to shift the thinking to another plane. My main purpose in doing so was to make a case for selecting on the basis of flow rather than representation. I will describe one of these examples in detail.

Fieldwork socius

When working for Professor Arie Graafland in the theory section of Delft University of Technology Faculty of Architecture, I was asked to make three short films to accompany his book The Socius of Architecture.[ii] This publication deals with analyses of three specific sites (in Tokyo, Amsterdam and New York) through a series of socio-architectural interventions – or intervention proposals to be more accurate, since they were never realised. The short films were not meant to illustrate the chapters in the book (there was nothing to show of the projects in an actualised state), but were meant to clarify what the book was not able to do: they needed to show the affective capacities of the specific sites. Since these sites were especially chosen for their status in the urban fabric, it was not easy to convey these capacities through ‘traditional’ aesthetic means. The first site was the Tsukiji fish-market in Tokyo, which is the largest trading place of its kind in the world. The trading, shipping and processing activities operate twenty-four hours a day, providing work for over sixty thousand people. The organisation of all these activities is not easy to overview; much of the actual trading and processing is done by small (family) businesses and involves much manual labour. The impression one receives from this place is the reverse of a highly efficient industrial processing plant, yet the effectiveness is certainly comparable. The biggest problem I faced in capturing the atmosphere and forces that propel the economy in this seemingly chaotic environment was the lack of any centralised mode of thinking. Its entire agency is distributed in the assemblage that exists during market hours. Rather than having one core building which hosts all activities and enables processing in a serial, linear way, all the activities are accommodated in smaller structures that make up portions of the chain. The organisation is parallel-serial in a decentralised way.[iii] This rhizomatic type of organisation makes it especially hard to demonstrate the enormity of the scale.

The second location was the meatpacking district in Manhattan, New York. This site is known for its several connotative layers. Until the end of the 1970s, the railway connection with the hinterland ended on an elevated track running straight into the heart of the district, ensuring an almost uninterrupted supply of meat. The area was known for its rough character: it has been whispered that a major reason for its removal as a significant hub in the food-supply chain was driven by the wish to eradicate mafia practices. During the decline of its meatpacking functions it became a centre for sex clubs and other venues specialising in sadomasochism and other hard-core, gay male practices, at the same time attracting a string of illegal and dangerous activities (drugs, prostitution, violence) to the area. Building on this reputation, many top fashion designers found the district to be the ultimate spot for contrasting their high-end designs, which fuelled the most recent wave of gentrification and led to the establishment of flagship fashion stores. This in turn attracted another type of nightlife: the hippest bars and clubs are now centred in the meatpacking district. When the area became known for its role as the backdrop for the Sex In The City series and movies, its reputation as the ‘hippest place in Manhattan’ was definitively affirmed. My problem was clear: Graafland’s book had been written in the period just before the latest transition and reflected the problems of that era. His design intervention was situated on a part of the remains of the elevated railway track that was a token of a past era and served as a shelter for homeless people who, at that time, were of major concern to the municipality.

By the time I began making my film, this situation had already changed, and in order to capture some of the meatpacking activities, I really had to look hard for a place that was still working in the meat industry, and had to wait several days until a shipment arrived (the only one in a whole week). In the very early hours of the morning the area contained a strange blend of hard-core fetishists and bondage aficionados, who looked with disdain upon the young hipsters leaving the newer clubs and getting into the pimped-up limousines. The supply trucks for both the meat and fashion industry were queuing up in the adjacent streets, where the cleaning activities that are part of the mise en place for daytime shopping in the area were taking place in full swing. The question was obvious: which aspects of these activities show the ‘true’ meatpacking district? I will return to this later.

The last of the three sites for my investigation was the Westerdokseiland, an area just west of the central station in Amsterdam. This compound was founded in 1832 and served mostly as railway sidings. Due to its proximity to both the harbour and a part of the historical city, it became a valuable piece of property. Since it was owned and managed by the railway company, it only unlocked its potential at a very late point in the timeline of the gentrification of Amsterdam’s ‘rough’ neighbourhoods. Here I was faced with a different obstacle. The original area featured in the interventions in Graafland’s book had already been transformed and was at the completion stage of a later plan. It was therefore impossible to show the original condition and inappropriate to show the current one, since it obviously would not have pleased my client to have his film filled with someone else’s design. The only elements that I could use were the construction works in progress and the layout of the main and surrounding areas. Thus in order to make a film about this subject I needed to show anything but the subject; I had to ‘depict an absence’ (like using a footprint in sand to show a foot).

As I mentioned earlier, the construction of the three films served as a blueprint (or at least as a source of inspiration) for students to develop their own understanding of this type of cinematically aided cartography. In order to achieve a level of abstraction that could serve as a conceptual framework, I needed to go through the process myself and invent a new cinematic logic.

Cinematically aided cartography


The first film completed was the one on the Tsukiji fish market in Tokyo. As I said, the biggest issue was depicting the site’s size and processes. As an effect of its decentralised labour, economic and infrastructural philosophy, there was no central or quasi-objective perspective imaginable. At this point it should be said that the marketplace is an exponent of a much larger system of cultural and social values. In a nutshell, the quasi-religious expression of respect that typifies the Japanese ethos towards food preparation and food consumption necessitates this particular type of supply chain. It is not solely the daily availability of fresh fish and seafood that enables the consumption of these raw or delicately processed foods: it is the fundamental socio-cultural structure that requires this type of organisation in order to consume this type of food in the specified way.

Graafland’s architectural intervention here was a shelter specifically designed for the group of day labourers. A significant number of the 60,000 workers who inhabit the area are mostly single, uneducated men who submit to a system of daily selection in order to secure some income.

Since most the workers are migrants, predominantly from the poorer regions of Japan, the system proves very inefficient: much of their hard-earned wages are spent on food and shelter. The net revenue that they manage to send home is extremely small or non-existent. Indeed some workers even accrue debts in order to stay in the vicinity of the market. The shelter was designed to provide affordable, clean and safe lodgings right on the edge of the huge market complex, next to Namiyoke shrine, which was frequented by the target group. Taking all these factors into consideration, it seemed best not to focus on the men directly: any depiction of an individual as representative would have proved limiting and rather difficult to achieve. Besides, it was not the right way to start the narrative; day labour was an effect, not a cause.

This problem needs to be addressed from two sides: the individual in relation to the collective on one side and the systemic tendencies on the other. Within this division there needs to be room for careful consideration of what is a cause and what is an effect. For example, if an economic system drives people to search for labour outside their habitat, this need could be addressed on an individual level: (temporary) employment can be found elsewhere. However, if not one, but thousands of people make the same transition, than the migration itself becomes an economic system. All these movements have effects at an individual level, yet it is also this individual level that ultimately forms the economic system. The relation between individual and collective, and the relation between collective and systemic, are highly subjective and fluid concepts. If we focus on systemic problems at an individual level, then it becomes clear that the majority of problems are the same for all individuals. This in itself does not make the situation a collective problem, yet it could be addressed by a collective solution.

I decided to create an axis that included both sides (which is ironic given the contra-Cartesian sentiments in this book). On the horizontal axis I placed the size of the system, with its economic, social, cultural and political values and forces; and on the vertical axis, the day-labourer workforce, with its huge collective problems at an individual level. Both axes include interrelated values: individual/collective/systemic, yet they both show a different perspective. The one shows the relation ‘from big to small’ and the other ‘from small to big’.

This could be called a nesting of scales. Axis thinking can evoke the creation of two opposites that do not actually exist, or two elements that are presented as being unchangeable, forcing logic to fill all the blanks in between, even though these might not exist or be relevant. My approach to the axis took the relation between the two scales as the centre line in my investigation. This meant that neither end of the relation was defined in a concrete or precise way. The only way to describe a point in this diagram is to express it in terms of ‘bigger or smaller than’ another point in the diagram. Regardless of exact figures, we can always place any two elements on a scale that relates the two: one is always greater than the other, unless they are the same size. This could be called the ‘differential axis’. Subsequently, I began the mental exercise of collecting all the elements that were visually eloquent and organising them along the axis in preparation for the actual editing, which I began once I was back in Europe. Suddenly all the elements proved to be relevant: after all, the ends of the axes were not defined, only their mutual relation, meaning that the system is always presenting 100 per cent of its assemblage.

A second type of axis focused on the system of parallel handling. As I explained, all the processing is done in a small-scale, decentralised manner, which ultimately leads to an overall throughput of 700,000 metric tons of fish and seafood every year. The only way to develop an understanding of such a magnitude is to repeat the same types of images to the point that the repetition itself becomes visible. By this I mean disrupting the cinematic process of homogenisation and the harmonious assembling of a montage in favour of the creation of a more fundamental insight into the virtual of the subject matter. This was a difficult procedure, not so much conceptually or technically, but because I had to undermine my own standards as a filmmaker: I had to challenge my own aesthetic instinct.

New York

The second lesson that can be learned is based on the New York experience. The historical layering of the connotative values of the area of the meatpacking district meant that regardless of whichever layers I picked, I would always leave some out. The intervention was based on a set of premises that, in part, no longer existed. Some had shifted and some were ‘universal’ in the sense that homelessness is not an effect that can be solved simply by modifying a set of parameters. The causes and effects are much too complex to be addressed on a single level. In addition, none of the layers had actually vanished, the area was still claimed by meatpacking activities, S&M practices, clubbing hipsters, garbage men, fashionados, business people and sightseers, all at the same time. Immediately after my return I made a short edit of some of the meatpacking activities and while showing this to my professor and his visiting colleague from Columbia University, the guest reacted immediately, saying, ‘That must have been filmed a long time ago,’ whereas, in fact, the footage was not even a week old.

This is indicative of the constant interchange between denotation and connotation that is provoked by that area: it is what people think-feel it is. The process of gentrification is an apparently unavoidable element in the development of every city, and in that sense this situation was comparable to that of Amsterdam. Yet, given that the proposed intervention was also part of the system of gentrification, it was most likely that the film sequence could serve its original purpose very well, even if the conditions themselves had changed. This made me decide to focus on the conditions it had reflected in the first place, without any attempt to historicise events.

I told one story without pretending that others did not exist. Obviously, this was a meta-story not a specific story, and the choices of what to depict were amply provided by the subject matter itself. Apart from waiting for the photogenic meat delivery, I did not force anything to be in or out of the shots. Just as in Tsukiji, I spent several nights and days in the area, firstly to familiarise myself with it and to become ‘known’ there, and later to film. In Tokyo especially, it was imperative to demonstrate respect for the system and the flows within it. The seemingly unregulated swarming of labourers followed very exact paths and structures: standing one metre closer or further away from a certain point literally made a crucial difference; the moment I unwittingly blocked an invisible pathway, I immediately got into trouble. The method for New York, however, was that any particular story is true as long as all stories are included.


The third and final lesson to be drawn from the making of this trilogy arose from the experience with the Westerdokeiland project. In contrast to the meatpacking district, all layers of denotation and connotation had already been stripped away, not only physically, because of the demolishing of all old structures, but also mentally, because the area had never been inhabited or used by ordinary citizens, so there was simply no form of social history there. I could have employed this freedom to focus on the architecture itself and utilise the architect’s arguments for the intervention as a thread in the film. Unfortunately, his plan was never actualised, and so any arguments concerning that specific site were unsuitable for inclusion in the production. The only option that remained was to raise the level of abstraction: to make a film about a plan in any city. The more generic the arguments (generic in the sense of arguments based on a plan, not on the site or an actual intervention), the more detailed the visualisation needed to be. Luckily, that proved to be easy: construction was still in progress, so shots of the work were effortlessly obtained, and shots of the site’s surroundings were equally easy to produce. It became a film about interventions in general; the narrative was provided by the design, not by the actualisation of the plan.

Photo-credit: John Margolies, photographer, Library of Congress Prints and Photographs Division Washington, D.C. 20540 USA. No known restrictions on publication.

Title: Supersonic Car Wash, Ogden, Utah, USA, 1991.

[i] Marc Augé, Non-Places: Introduction to an Anthropology of Supermodernity (New York and London: Verso, 1995).

[ii] Arie Graafland, The Socius of Architecture (Rotterdam: 010 Publishers, 2000).

[iii] In this context, parallel-serial indicates the simultaneous acting of similar processes; e.g., multiple lines of the same type of processing without a central core. Rather than one line processing all tasks (serial), or several lines doing different tasks (parallel), parallel-serial does both.