In the introductory chapter of his book Practice: Architecture, Technique and Representation, Stan Allen opposes the categories âtheoryâ and âpracticeâ as shown in the following table.
However, Allen doesnât just argue that architecture is practice rather than a theory, but he also suggests a further differentiation between two competing categories of practice. He distinguished between âdiscursive practicesâ and âmaterial practicesâ as in the following table.
Allen argues that architecture qualifies as a material practice rather than as discursive practice for multiple reasons:
1) The Specificity of Building
Compared to other discursive practices as writing, film, new media, etc. architecture is ârelatively inert as discourseâ. Practicing architecture as discursive practice is âa troublesome pursuitâ. However, as material practice, architectureâs material and instrumental properties present great opportunities. Architectureâs real attraction is its source of creativity, operational power, and pleasure. [p.14]
2) Architectureâs Agency in the Public
Architecture is part of a complex social exchange. Because of the constant change of the public sphere, architecture cannot rely on existing norms and conventions of agency but needs to be agile and responsive to the fast-moving reality. Therefore, architecture must be conceived as material practice. [p.14/15]
3) Practical Consequences and Effects of Architecture
âThe ability of architecture to generate perceivable experiences and sensations in the world â practical consequences and effects â is more important than its conformance or non-conformance with some abstract set of theoretical criteriaâ. Architects work with lots of theoretical knowledge, but the impact of this knowledge is indirect. However, the variables of construction in architectural procedures create practical consequences and effects. [p.16/17]
4) Constant Change of Architectural Practice
Architecture influences culture and society but is also largely influenced by culture and society. Architecture has survived the constant changes of society by being constantly revised and changing itself. Architecture cannot rely on theory and past work. As material practice, âthe significant work of architecture is one that allows continual revision and rereading, teasing out new meanings as the context changesâ. [p.17/18]
5) The Writing of an Architect
Architects write texts as well, but the activity of writing of architects can be part of architecture as material practice. The writing of architects is constantly on the lookout for change and new techniques. It works from examples towards something new, and it does not work from principles to regenerate something that already exists. [p.20]
The reading is very much in accordance with our class which clearly treats architecture as material practice. Some of Allenâs arguments are very comprehensible in the context of the class:
4) Constant Change of Architectural Practice
ARC 311 gave us an understanding of the constant change of architecture and âthe new ways of thinking and seeing that have emerged with modernityâ as mentioned in the text (p.20). One of the biggest changes in modern architecture has been the shift to digital design and computation. The class taught us how to use computer programs and algorithms to facilitate the architectural design process. Furthermore, we learned how to use modern technologies and machines like the laser cutter, the Zund, or the CNC Milling machine. Another development in the architecture world is the rise of robotics, which we have also dealt with in the class.
3) Practical Consequences and Effects of Architecture
When producing volumes and shapes with some of the machines listed above, we generated perceivable experiences and sensations. Also, it was oftentimes apparent that whatever we produced was not conforming with theory. There was always a factor of doubt or risk, practical consequences and effects of materiality that werenât accounted for, etc.
5) The Writing of an Architect
Through our weekly readings we dealt with many writings of architects. However, we never tried to conform with the existing principles from the texts in our projects. We used the readings as base, but always pushed ourselves to go beyond previous work and create something entirely new and original.
In summary, our class really underlines Allenâs text because it presented architecture as material practice throughout the whole semester.
During the casting, we were interested in adapting the concrete not only by adding color but also by adding other materials for additional texture. Bubble wrap (small and large sizes) was used to test and push the boundaries of concreteâs physicality by producing texture and the permanent impressions of crumbling plastic air bubbles into this solid form.Â
This way of casting our foam speaks to Allenâs idea of the âtraceâ. We can see that the final product (the concrete cast) holds traces of the original foam pieces and the concepts of color and texture we hoped to imbue in them. Though, as Allen points out, the process of construction destroyed some trace (bits of the concrete broke and crumbled).
Exercise 7.2Â - Material Process: Casting Concrete
Our series of concrete cast objects investigates how surface and texture can be used to produce a variety of implied origins and uses for the concrete, breathing life and character into smooth grey concrete.
The casts were created by inverting foam-cut casts and using the negative space produced by a single profile curve to fill in concrete. Single tone and two-toned objects were produced, and the rough, crumbling edges of the concrete were intentionally maintained.
In âThe Jig, the Nudge, and Local Ecologyâ, Crawford writes about the âjigâ as a method of human experts which helps them improve their memory and work capacities. Crawford states that the maximum capacity of our working memory as humans is very limited, however, humans are able to âvastly extend [their] intellectual capacitiesâ (p.35). The reading in respect to our class makes me want to compare the memory and work capacities of humans and robots.
Humans use spatial âjigsâ to cue the environment in order to make themselves focus on a specific activity without constantly having to think about it. They make activities easier for themselves by âinformationally structuring the environment as they go alongâ (p.33). This prevents them from forgetting steps or getting distracted form their work or activity. For humans, this tactic is very helpful because âthere is only so much room in [their] head[s]â (p.32).
Unlike a human, a robot is unable to improve its memory capacity by arranging its working environment in a certain way. A robot only does whatever it is programmed to do (at least this is true before artificial intelligence). However, this also means that a robot really doesnât need to put cues in its environment: As it is programmed by humans and doesnât have its own will or feelings, it will not get distracted or forget steps. If anything, the programmer of the robot, i.e. the human, has to cue the robotâs environment. As Professor Parascho showed us in the workshop last week, she had to measure out space, define coordinates as points of orientation and set up planes, etc. for the robot to work. This is similar to the way that Crawford describes unskilled human labor: âThe jig is [âŚ] rigidly deployed by someone other than the worker him- or herselfâ (p.34).
Robotsâ memory capacity is a lot higher than humansâ, however, there are also limits to the amount of information robots can receive and store. As Professor Parascho demonstrated, it can happen that the robot keeps moving and stopping and moving and stopping (rather than moving in one continuous motion) because the rate at which information is sent to the robot is not as fast as the robot processing this information. This opposes the following quote from the text which refers to a machine: âThe busier it gets, the more âonâ it isâ (p.34). The example by Professor Parascho shows that indeed, machines have limits as well, and if it gets too busy, they might even stop.
Overall, this reading and relating it to the capacities of robots raises a debate over the efficiency of robots vs. humans (or at least skilled experts). What is more efficient: A robot with a very high memory capacity but no potential for improving this capacity by its own cognition, or a skilled human expert with a significantly lower memory capacity but the ability to cue its environment with jigs and use other cognitive methods to extend its intellectual capacities?
Recreating Canovaâs Pauline Bonaparte Borghese as Venus Victorius
This series investigates questions of authorship, materiality, and limitations of computational tools.
While the wire cutting ABB could cut and create one face of the sculpture in about a minute, the materiality and the tool (temperature/diameter of wire) prevented the tool from capturing nuances from the sculptureâs profile. Interestingly, the tool can perform undercuts, which this particular shape exemplifies.
The shape also leads to questions about traces â through the process of creating a profile curve, iterating it into a mesh, cutting it into a foam block, and then casting it in concrete, the process exemplifies that would could be product could also become trace, and vice versa. Another layer to the trace/product dilemma is that the profile curve itself is created from an image of the sculpture, which adds additional layers to Allenâs conception of remainders and indexes.
Modern computational tools and technology, while looking forward, can also be productive ways of re-negotiating the classical â iterating a face of the sculpture in both foam and concrete produce different contexts and frames to read the very same curve, which is an interesting consequence to unpack.
Time lapse video of the âcuttingâ process with robot and hot wire.
Note that the hot wire, that moves through the foam very slowly, "melts awayâ the material rather than actually cutting it.
While cutting the foam, we experimented with the speed of the robot arm to see how that affected the quality of the cut surface â a slower cut resulted in a more granulated, rough surface, while a slow and steady cut produced a smooth finish to the topography. When cast in concrete, it will be interesting to observe how these differences manifest in the topography.
Photographs of all three series side by side.
Interestingly, in the perspectives chosen in these photographs of the final shapes, the initial curve created in Rhino becomes evident through the shadows of the shapes more than through the actual foam volumes themselves.
These three cuts investigate curves through duplication and manipulation.
The yellow and green series were created by mirroring and flipping the same curve without disrupting the geometry of the curve at all. The resulting concrete cast that we hope to produce from these foam iterations will be formed from the negative space that is a result of the flip+mirroring functions, filling in negative space that is produced when to identical forms are inward-facing and forced together.
The blue series was created out of two curves mirrored onto each other. One of the curves, while maintaining the same geometry, exaggerates each of the curves, producing a dynamic surface that is sharper on one end, and shallower on the other. The topography of this surface is of particular interest in ways that it can be cut and cast in concrete.
In âPlotting Traces on Processâ, Stan Allen leads an interesting discussion about the concept of the index, which he defines as âa sign which refers to the object that it denotes by virtue of being really affected by that objectâ (p.48). I find that one of the most interesting questions discussed in the text is the following: âIs an index still an index if the trace is processed through a coded representational system, even if all the information is preserved intact?â (p.52). Allen argues that if information is converted into âdigital code that permits its manipulationâ, the âevidential authorityâ and the character of the index are bypassed (p.52). Therefore, while a photograph can be considered an index, a digital scan of that photograph cannot be considered an index. However, later in the text Allen denotes that in architecture a drawing can function as an index. This seems contradictory to me, because drawings â and especially the digital drawings of the modern day â can be manipulated just as well as digital scans of photographs. This raises an interesting question about the role that manipulation plays in the definition of an index.
The gist of the reading is the idea that indexical signs and traces can point toward the past. With indexical signs âcausality is invertedâ and we can point back in time (p.48). This is evident in many of our ARC 311 projects. In Exercise 6 (Robotic Toolpaths), the whole project was about the robotic arm tracing curves with light and the camera tracing this movement of the light. When looking at our final photographs, we can reconstruct the movement of the robotic arm. Similarly, in Exercise 7 (Hot Wire Cut), the hot wire was tracing along two curves forming a surface, and this movement of the wire can be understood in hindsight by looking at the final volume. Therefore, our final products of Exercises 6 and 7 would be considered indexes by Allen. On the other hand, other projects we have done in this class oppose this concept of the index. For instance, in Exercise 5 (Milling a Mesh Surface), we cannot reconstruct the movement of the mill by looking at the final volume. It is unclear along which paths the mill moved during the milling process, and it is unclear which parts were milled during the âHorizontal Roughingâ part and which parts were milled during the âParallel Finishingâ part of the milling process. Therefore, our final mesh surface from Exercise 5 does not constitute an index.
In this last series, we wanted to try out the glow sticks one more time. We noticed with the sphere that the glow sticks created images that were very blurry, so we were interested in trying a rather simple curve as opposed to the complex sphere. We tried many different drawing speeds of the robot and different camera setting, but most images drawn with the glow sticks still turned out very blurry.
In this series of spheres, we experimented with different light sources. We created two spheres with two different curves, one more complex than the other. We drew both of them with a light blinking in three different colors (bottom right image) and were interested whether the initial curves would still be visible when the image was created with a blinking light source. The blinking of the light source created a interesting pattern of blurry dots in both pictures, but both are still distinguishable and relatable to their initial curves. In the bottom two pictures (yellow spheres) we drew the spheres with glow sticks. The light source was a lot bigger and at the same time less bright, which is why the images turned out to be just a circular blurr.
Our group was also interested in creating Euclidean objects for the robot to draw. This series of cubes investigates how the perception of the shape changes when rotated in space and viewed from different camera angles.
On of the main features of drawing with a robot is the ability to draw in a three-dimensional space. Therefore, our group wanted to create mainly three dimensional shapes and objects. I was reminded of of an artwork I have seen multiple times in the Kunsthalle in Mannheim, Germany, which consists of bed frames floating in space (see top image). Therefore, I was interested in drawing other everyday objects (chairs, tables) that would then be floating in space when drawn by the robot.
In the first part of Sarah Ahmedâs text âWilling Partsâ, she differentiates between two different main arguments: the subordination of parts to the whole body vs. the mutual dependence of parts.
Largely by referencing Blaise Pascal, Ahmed demonstrates the idea that the will of every part âshould tend toward the generalâ (p.99). âHaving one shared will is deemed necessary for the happiness of each memberâ (p.100). A part, that does not submit its will to what the whole of the body wills, is called âthe willful partâ (p.99), who âcompromises the happiness of the whole bodyâ (p.100).
By contrasting Pascal to Saint Paul, Ahmed articulates a different role the part can take on: âThe whole needs each [member of the body] equally in order to functionâ (p.104). Individual parts are dependent on each other, and therefore each individual part is valued for what it does.
Reading this text with respect to our class and specifically our exercise with the robot made me wonder which of these two arguments is more in accordance with the individual parts of the robot. A robot is comprised of so many different little parts, and the robot needs all of them to function. If one little part is damaged, the whole robot cannot function. This makes me think that the second argument is more fitting to the role of parts in robots. A robot is dependent on all its parts and the parts are mutually dependent on each other. A similar argument can be made with the parts of a computer.
Another main distinction in Ahmedâs text is the one between being âwillingâ and being âwillfulâ. Being willing refers to a part that is aware of its âstatus as partâ (p.100) and submitted to the will of the whole. A willing part promotes happiness of the whole body. Being willful, on the other hand, refers to a part that is breaking free from the will of the whole body and has its own individual will.
This distinction reminded me of the differentiation that both DeLanda and Kolarevic made between the predictability vs. unpredictability of computer design. Interestingly, in Ahmedâs text, breaking free from what one is expected and supposed to do, i.e. being willful as a part, is presented as rather negative trait: The willful part âthreatens the reproduction of an orderâ (p.99). The willful part âis the one who leaves the path of becoming part, breaking, or threatening to break, the tie that holds a community, a family, a nation togetherâ (p.115).
In contrast, DeLanda and Kolarevic both appreciate the capacity of computers and machines to go beyond what they are expected to do and do the unexpected. DeLanda states that âas an aid to design, [genetic algorithms] would be rather useless if the designer could easily predict which forms would be bredâ (p.9). Similarly, Kolarevics emphasized that what highlights craftsmanship as opposed to manufacturing is taking on risk, which implies taking on uncertainties and not being able to predict the outcome of design. In Sara Ahmedâs terms, the unpredictable things computers and machines can do would probably be seen as the willfulness of parts with a rather negative connotation.
The reading âThe (Risky) Craft of Digital Makingâ relates in many ways to last weekâs readings and to this weekâs exercise. Similar to last weekâs readings, Kolarevic recognizes the potential and the importance of connecting fields: âThe process was defined by parameters related to designed geometry, parameters pertaining to production [âŚ], and parameters related to the material itself [âŚ]. These parameters were interrelated, thus numerous design opportunities were explored through several iterations informed by continuous feedback loops between design and production.â (p.124) This quote also emphasizes the numerous unpredictable design opportunities that digital design offers us, as emphasized by DeLanda and Iwamoto as well. The feedback loops between design and production are one of the most important features of digital fabrication that Kolarevic repeatedly emphasizes. Technology has an incredible ability to streamline design and production and eliminate intermediary steps between design and fabrication.
The plywood panel material study by Jill Desimini and Sarah Weidner as part of their project âField Explorationsâ seems to be fairly similar to our Exercise 5.1 â Milling a Mesh Surface. As Kolarevic explains their project, he mentions several constraints and issues that we explored and faced in our milling process as well. Desimini and Weidnerâs âinitial production attempt was unsuccessfulâ due to multiple reasons (p.125). The grain of the wood was not taken into account and the milling feed-rate was too high resulting in âcomplete destruction of the materialâ (p.125). For our milling process, we were given a specific feed-rate to put into the computer by Grey, and we still experienced destruction of our material as visible in the pictures of the project below. Furthermore, Kolarevic states that âthe parameters related to the production [âŚ] were crucial to the overall success of the projectâ (p.125). We were able to see this in our project as well. It was crucial to the result of our project which of the components in the customâ âmeshâ âsettings of the grasshopper file we manipulated and by how much we manipulated them. This shows that although technology brings with it a lot of advantages and opportunities, for instance the ability to streamline design and production mentioned above, the human influence and human choice of design is still crucial to control the product.
The three pictures above show a destruction of our volume caused by the milling machine. The initial block of plywood consisted of four sheets of the material glued together. During the milling process, one of these sheets completely came off in one place. Unfortunately, this happened during âHorizontal Roughingâ part of the milling process. Therefore, we could not just glue the piece back on in the end, because it missed the entire âParallel Finishingâ part of the milling process. Unfortunately, this took away from the detail that we would have wanted for our more complex and more detailed half of the volume.