HUNTING DESIGN MEMES

IN THE ARCHITECTURAL STUDIO

Student projects as a source of memetic analysis

 

Roel Daru, Eindhoven University of Technology, The Netherlands

r.daru@iae.nl

 

Abstract

The current practice in design programming is to generate forms based on preconceptions of what architectural design is supposed to be. But to offer adequate morphogenetic programs for architectural design processes, we should identify the diversity of types of cultural replicators (memes) applied by a variety of architectural designers. In order to explore the variety of replicators actually used, around hundred 4th year architectural students were asked to analyse two or three of their own past design assignments. The students were invited to look for the occurrence of evolutionary design processes. They were requested to try and find some traces of memetic ‘transmission’, ‘variation’ and ‘selection’ in their own design assignments. Some of them have described their design processes in very broad terms, others on the contrary have made their analyses in a much more detailed way. Although  the students were also encouraged to identify non evolutionary processes in their own design work, no one came up with more than a few exceptions in some particular phases where for instance no variant was considered and developed.

 

Introduction

How does an architectural design come into being? Does it evolve? Is it possible to apply Darwinistic evolutionary principles from biology to a cultural phenomena like architectural design? To shed more light on these follow-up  questions an exploratory study is carried out. About architectural designing, many theories and propositions were made in the past. There are impressive lists of literature about design theory, design methods, empirical design studies and so on. The literature might be devided in many different ways, but often one of the types considered will be evolutionary. Here I will explore this line of approach usually taken by people working in the fields of generic art and architecture.

 

According to Philip Steadman’s book about ‘The Evolution of Design’ (1979) the historical roots of evolutionary explanations extends back beyond the 19th century. In the 19th century a theory of cultural evolutionism was established, based on the analogy with biology before and after Darwin (1859/1985). Here I want to proceed on this theory of cultural evolutionism, but now based on the recent findings of molecular biology as an analogy. This theory is already elaborated by Derek Gatherer (1999) for the technical design world, where useful inventions have the upper hand. In this paper, I want to extent Gatherer’s explanations to the world of architectural design, where (besides usefulness) expressive innovations are at stake. To back up his theory Derek Gatherer has discussed the work of inventors like Edison. The work here presented is a step further on the way to empirical testing, because I asked the (upcoming) designers themselves, instead of consulting literature about their work (or work of practising architects). Most students were curious about their own design activities and eager to analyse their own design projects. Fellow teachers asked me about what they had overheared from conversations between students.

 

In this paper I will argue about the theory used, the method to approach the theory, the results of the applied method, conclusions as a reply to the questions postulated and some remarks about the utilisation of the results.

 

Theory

To answer the question about the origins and mechanisms of architectural designing, we will explore here the possibilities of application of Darwin’s theory of (natural) evolution. But than as it is restated first by Dawkins (1983) in a generalised way (to include every evolving system) and second as it is applied by Gatherer to explain the evolution of (mostly technical) design(ing). In all those cases three features should be involved to call a system evolving in the Darwinistic sense. A true evolving system should have:

       retention or heredity, in the sense that offspring inherit characteristics from their parents or more general heredity information could be replicated or transmitted somehow,

       variation or differences as errors creeping in in the replicator to avoid offspring with identical features, making it impossible to choose among them and preventing the system to evolve,

       selection or preferences by an environment in which some varieties are better in surviving than others thanks to their non-identical features, propagating their frequency in the population involved.

 

If some of these features or criteria are missing, no real evolving system will ever come out of it. But with all the three features in place, evolution is unavoidable. In order to testify that the design process is an evolving system in the Darwinian sense, we should identify all the three features present in all the processes of architectural design. In addition to this, to be a true Darwinian process of evolution, variations should be produced at random. As Derek Gatherer has remarked, in design the consequence of this theoretical position is that our ideas are not produced at will (as a conscious effort) but are generated in our brain in an arbitrary way.

 

Brainstorming is based on this assumption: in a free flowing session ideas should be produced from the subconsciousness in an associative way in reaction of what someone else has suggested before, but without exerting any criticism at all (variation at random).  These criticism should be applied afterwards and will than result in the choice of valuable ideas (first selection of the ‘environment’ within the brain). Brainstorming is a formalized way of what our brains are doing all the time: extracting ideas (memes) from our communal meme pool in a haphazard manner, associating them in an uncontroled way amongst one another, making quick mutations and recombinations in a arbitrary fashion and then exerting selective pressures for the first time (in the brain) and perhaps in a half conscious way only.

 

If it is possible however to introduce deliberately  variation or differences in these mental processes, we will have a Lamarkian process of instruction at work instead of a process of selection. With genetic engineering we have knowingly introduced this instruction approach in biology. In designing, it might be practiced from the beginnings of all design activities by human beings: this is the prevailing and counteracting hypothesis of the Darwinian position. Just as the sense of design in nature happens to be an illusion produced by selection, deliberate instruction in producing novel memes (ideas) seems to be  an illusion triggered by selection.

 

As mentioned above, the actual theory of cultural evolutionism is already elaborated by Derek Gatherer for the technical design world. Gatherer’s theory is based on ‘Universal Darwinism’ as concocted by Dawkins (1983). He asked himself if there could be an other kind of Darwinian replicator on earth, ‘even now staring us in the face?’ Something like a gene, but operating within a cultural instead of a natural environment. There the ‘meme’ (Dawkins, 1976) came in as an analogy of the gene. In genes heritability, variability and selections occurs and as such satisfy the criteria needed to identify them as replicators of a real Darwinistic evolving system. If memes could satisfy the same conditions, a second type of replicator will be discovered, functioning according to the same Darwinian rules of evolution.

 

But what is a meme? A lot of authors about memetics have tried to formulate all-embracing definitions, but as Susan Blackmore (1999) has commented after an extensive search none were sufficient enough to include all the cases where the concept of a meme could be applied reasonably well. To take her point of view: we should for the time being ‘keep things as simple as possible’. We should provisionally ‘use the term  ‘meme’ indiscriminately to refer to memetic information in any of its many forms; including ideas, the brain structures that instantiate those ideas, the behaviours these brain structures produce, and their versions in books, recipes, maps and written music. As long as that information can be copied by a process we may broadly call ‘imitation’, than it counts as a meme.’

 

Derek Gatherer (1999) writing in the context of the design of motor cars has compared a design meme with a ‘mental blueprint’: ...  ‘abstract notions of form’ ... ‘first created in the minds of the designers, and then used as templates from which the physical form of the motor car was constructed. The mental concept was thus expressed in physical terms’. Mental blueprints are what design researchers call declarative knowledge as opposed to procedural knowledge. Declarative knowledge in the design fields is about things (like buildings, spaces, lighting), their attributes (like form, function, attractiveness) and the relations between them (like branched, semilattice, network). Procedural knowledge on the other hand is about actions and plans of actions (like how to analyse, sketch or design). It could relegate among others to strategies and tactics of designing (knowledge about how to do or create things). Both types of knowledge are used by designers and memes could consequently be declarative as well as procedural, tracing the memetic conceptualisation  back to the broad outline given by Susan Blackmore. In the results I will present some distinctions made by me and the students to differentiate between certain types of design memes.

 

Method

If the theory is right, we should find in every (architectural) design process both: design memes and processes of retention, variation and selection exerted on those memes. The detection might be performed first in coarse and exploratory studies and (if the results are positive) afterwards in more finetuned and affirmative investigations.

 

In a first run-up study the detection might be performed indirectly like Gatherer has done by literature searching for appropriate traces of the mentioned heredity, variation and selection in the design processes described. A disadvantage of this approach is that you are dependent on what others have noticed occasionally for often quite other reasons and objectives.

 

A more direct and fast approach is to ask to reconstruct their own design processes from long term memory with the aid of their own remaining sketches, annotations, drawings, cardboard and computer models, etc.This still exploratory approach is choosen here, because the possibility existed to do it with a lot of students and a variety of design projects from several years of design studio work, and because the effort needed was modest and the results instructive in a preliminary phase to decide the next steps of investigation.

 

In other studies we have used a logbook based system of investigation. This is a more affirmative and time consuming manner in which the designers are asked to fill in a logbook with their remarks immediately after the end of their design sessions. This might be done in a free format like in in-depth interviews, or for an easy transcription with standard questions, to be filled in. The questions might be put forward directly about what he or she has done and thought of about design things and related other ideas and the role played by retention, variation and selection in those design activities. Posed indirectly the questionaire  might ask about what he or she has done, how it was done and why.

 

One of the most detailed but also most time consuming approaches is by protocol studies where the designer is observed and videotaped in the act of designing and talking aloud about what he or she is simultaneously sketching, perceptually discovering in the sketch, thinking and deciding upon. The results can be analysed in minute details in order to detect, determine and affirm the presence or absence of the looked after memes and processes of retention, variation and selection at work. With the low cost digital video editing posibilities now available, it is much easier nowadays to perform this type of analysis.

 

The reconstruction method used was carried out with around hundred 4th years architectural design students. They got handouts and copies from publications explaining the theory of Darwinian (and Universal Darwinian) evolutions and the possible role of memes for the cultural world (which includes the technical and architectural design worlds as well). I asked the students to reconstruct (two or) three of their finished design projects, leaving the selection to themselfs, but asking to motivate their choices. The reconstruction should be done from their long term memory, assisted by the remnants of their design projects selected. The results should be reported in an essay about ‘memetics in architectural design’. Failures in detecting memes or processes of retention, variation and selection in their design work were even more interesting and would be equally well excepted and graded if argued in the same manner as those positively reporting on the same topics.

 

 To facilitate their work, I suggested to fill in a ‘project-process-matrix’ (figure 1) with (a) in the rows the three projects, (b) in the columns the three processes of retention, variation and selection and (c) in the cells the memes encountered eventually.

 

 

HEREDITY

VARIATION

SELECTION

Project 1

 

• Situation, program, concepual ideas and rules related.

• Good and bad examples from publications and excursions and mistakes from failed projects.

• By mutation and (re)combination of heredity elements.

• Incremental modifications of preselected concept.

• Rules application.

• By comparisons between options  or work of others.

• By judgements of teachers, design critics, fellow students and own esthetic and other preferences.

Project 2

 

 

 

Project 3

 

 

 

Figure 1: suggested Project-Proces Reconstruction Matrix

 

I encouraged them to hand in their provisional results in order to get answers about upcoming questions and to secure some confirmation or assurance for their approach. I made (and distributed) a summary about the most frequently asked questions and upcoming themes and topics.

 

Results

The students did not have encountered much trouble in completing the cells of the ‘project-process-matrix’ with appropriate memes. Both verbal and visual memes were quoted, indicated or reported. Most memes, whether visual or oral ones were named or described in titels and sentences. Some students refered to visual memes with names in the cells of the matrix and documented them with drawings and computer renderings in separate additions.  Others made depictions straight in the matrix cells in either thumbnail design sketches or equivalently scaled down drawings or 3D perspectives of computer generated models (figure 2).

 

Often, the terms used were clarified in subsection texts. To reflect the phases of a ‘project-design-process’, some students appended an extra left column with the number and phases involved (figure 3).

 

DESIGN PHASES

HEREDITY

VARIATION

SELECTION

01 Initiative

02 Feasibility

03 Projectdefinition

04 Brief/principles

05 Grids

06 Sketch Design

07 Analysis SD

08 Façade design

09 Preliminary Design

10 Analysis construct.

11 Options construct.

10 Stability calculation

11 Materials

12 Details

13 Final Design

• Visual, verbal and text memes about buildings, spaces and details.

• All sorts of inspiration patterns and diagrams and associations about objects like works of art, landscapes, anatomical illustrations.

• From libraries, lectures, literature.

• Comments from other students, tutors or lecturers.

• Spatial, architectural and/or constructive options.

• Variants about the building site within landscapes, city areas and/or river banks.

• 3D-models of human bodies and buildings.

• Ideas recombined from memory and/or  thinked up from own fantasy.

• With given or own criteria.

• About beauty, confort, applicability, costs, effectiveness, efficiëncy, feasibility.

• Adaptability to the site, function, organisation.

• Event and meaningfulness, identity and orientation and perceptual liveliness.

Figure 3: summarized example of some students output

(without specific contents of the design projects)

 

 

An illustrated version of this type of matrix is shown in figure 4. The project illustrated was a design by the student Martin Koster as part of a training job (architectural office of Marlies Rohmer, Amsterdam, 1999). The given meme was about ‘wild housing’ (‘unregulated architecture’) for an exibition to be held in 2001 in Almere. The first inspiration meme were about a shed, which was transformed (a) by a growth principle of mutations and combinations and (b) by ideas about (1) chaos, (2) a common zone, (3) variants between separation walls, (4) a field of patio housing and (5) terrace housing in blocks of four. The second (and last) inspiration meme was about a boiler house.

 

Inspiration source: shed

 

Given meme: Het Wilde Wonen

(unregulated architecture)

Design meme: expansion patterns

 
MEME 
 
SELECTION
 
variation
 
HEREDITY
 
MEME 
 
SELECTION
 
VARIATION
 
HEREDITY
 
MEME 
 
SELECTION
 
VARIATION
 
HEREDITY
 

Transformation-meme: groeien

 
Figure 4 : Martin Koster for Architectenbureau Marlies Rohmer

Inspiration meme : boilerhouse

 

MEME 
 
VARIATION
 
HEREDITY
 
SELECTION
 

MEME 
 
VARIATION
 
HEREDITY
 
MEME 
 
SELECTION
 
VARIATION
 
HEREDITY
 
SELECTION
 

SELECTION
 
VARIATION
 
HEREDITY
 
MEME 
 


As an other alternative some students converted their phases in a flowchart of heredity, variation and selection (figure 5).

 

INVERSED CONE DIVISION

SPIRAL

SPIRAL-ROOF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3 Divisions

 

 

 

 

2 Divisions

 

 

 

 

Undivided

 

 

 

 

Propeller Rings Helix

 

 

Heli-coidal roof

Termi-nated under win-dow

 

 

 

Flat roof

FI-

NAL

 

RE-

SULT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Retention

X

X

 

X

 

 

 

 

 

X

X

X

X

X

 

Varia-tion

X

X

X

X

X

X

X

X

X

X

X

X

X

X

 

Selec-tion

 

 

 

 

 

 

X

 

 

X

X

 

X

X

 

 

 

 

 

 

 

 

 

 

 

(c)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(a)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The arrows in the matrix indicates where some retention, variation or selection originates (in most cases the former sketch design). Sketches with an (a), (b) or (c)  denotes an selection made between them (either (a) continues or (b) or (c); this is indicated by the arrows beneath the lowest X-crosses).

Figure 5: flowchart of heredity, variation and selection phases, from student Wouter de Wit

In the retention or heredity cells ancestor memes were identified by every student without any difficulty (figures 1-4).  The refered memes happen to be divided in situation, program, concept and rules regarding memes:

Situation memes were referred to in the sense of past starting conditions to be taken into account as the heritage of the project involved or questions about the building site of the given project.

Program related memes or issues were about prefered performance requirements or the application of already testified performance effects. 

Conceptual memes or ideas are about partial or whole architectural solutions. Wholistic conceptual solutions refering to typologies like functional ones (schools, shops, offices) or structural examplars (towerblocks, rows, courtbuildings) or precedences as formative ideas, spatial organisational concepts and partis of architectural buildings.

Rules related memes are discernable in sayings like ‘function follows form’ or ‘less is a bore’ or other heuristic search and decision rules mentioned by the students as inheritable principles to steer their design activities.

 If famous architects are mentioned they might refer to one or a mixture of situation, program, concept and/or  rules related memes as used or associated to those architecs and worthy to be copied. Visions of teachers, design critics or the students themselves might also be composed from a combination of situation, program, concept and/or  rules to be applied for deciding what to use and what to avoid in copying for their own project design processes. Good as well as bad examples from past and present times were important. Some students especially noted mistakes from a failed  project in order to dodge imitation and replication deliberately.

 

Some  information came firsthand from reality as it was observed on excursions, trips or outings and eventually documented with photographs, sketches and pictures or maps and leaflets and so on. But most useful suggestions came from teachers or fellow students with tips to look in some books or textbooks, to some specific manuals or volumes, periodicals or magazines, journals or newspapers.

 

In the variation or differences cells students had different experiences and opinions. They reported either a lot of variety, no variety at all or something in between. The subject matter of variety were shapes of buildings, part of buildings or details. Some differences were reported about rooms, the configuration of spaces, the composition of façades or building masses. Variety was also mentioned about building materials, construction systems, the building site or surroundings and so on. In most cases the students did not remembered how and when exactly they got the idea to vary something.

Much variety was obtained by modifying or by mutation and combination from the retention or heredity memes already acquired about the situation, program, concept and/or rules.In the end this is  leading  to some unity in all that variety (unity in variety).

But perhaps most students focused very fast to one basic design and reported their incremental modifications afterwards made to adapt their design to all incoming or already obtained considerations. As a result variety was obtained out of unity (variety in unity).

Some students however observed the absence of variety in one of their particular projects. For instance because the decisions were transfered to some piece of music and then the melody was translated in geometric compositions and measures by deductive rules of transcription. But in this case indepth questioning reveiled necessary decisions remained to be taken outside the automatically applied rules of musical/geometrical compositions and measures: decisions producing variety for comparison and choice.

Last but not least some noticed variation and differences obtained in specially organised teamwork design courses or by way of their participation to an architectural competition. Remarkable is, that no student until now reported about  errors, luck, chance, coincidences or incidental occurrences as sources of variation.

 

In the Selection or preferences cells the students had some difficulty to identify their selection phases, because decisions about preferences were made continuously and it even started before memes were allowed  by them to enter as ancestor memes. Some decided to mark the formal period of the in between design studio presentation as their most important selection phase. Most selections were made by checking once again the retention memes about the situation, program, concept and/or rules, but now checked and compared to the actual ruling conditions as it is infered from all the incoming information up to then. Selections were made by comparisons between alternatives amongst one another, compared to the program and site or compared to the work of other students or existing buildings. Important were also the judgements from teachers, visiting design critics or fellow students as wel as their own esthetic preferences and considerations. The beauty of the solution was mentioned as wel as performance criteria like about light and shadow effects, transparancy, resistance against vandalism, continuity of the flowing of spaces, building physics, construction, transport, assembly, styling, governmental codes and regulations, the budget, judgments of authorities like teachers and so on.

 

In general the exercision of retention, variation and selection is conditional for the success of memes. But then, is for instance a performance criterium not succesful in designing because it does not change as a consequence of the design process? The performance criterium is often imitated  as well from one design process to another with some retention of the gist of what is required, but perhaps undergoing some adjustments and thus with variation creeping in and then decided upon what to select. In other words, we should differentiate the memes involved in the design process if we want to see sharper what is going on in memetic terms.

 

The following differentiation (figure 6) appeared to be useful and is based on the simpel sequence of input, transformation and output.

Figure 6: types of design memes

 

Input memes are those predestined  to be transformed in output memes. Transformation memes are those undergoing some manipulation like the recombination with other memes. Output memes are the modified or adapted memes of the design solution. To transform an input in an output however contextual requirements, governmental regulations and so on should be known and taken into account without trying to chance those type of environmental information: another type and group of intervening design memes. To transform input to output designers are needed and consultants, computers and other resources, showing another group of design related (vehicles of) memes.

 

Dependent of the circumstances other words might be needed, so input, throughput and output memes should be called different if necessary within the objectives or context one is attending (figure 7):

       the input memes might occasionally  be called ‘given’ memes, ‘problem’ meme, ‘inspiration’ meme, ‘old’ or ‘existing’ meme, ‘ancestor’, ‘elder’, ‘progenitor’, ‘forebear’ or ‘forefather’ meme.

       the transformation memes might be differentiate in for instance in ‘design’ meme, ‘variety’ meme, ‘throughput’ or ‘change’ meme.

       the output memes might be distinguished in new meme, solution memes, transformed memes, descendant or offspring memes.

       the contextual memes might be tell apart in selection meme, influencing or environmental meme, steering meme and control meme.

       the resource or help memes might be separated in editing  and instrumental memes and production memes as call memes to be processed somewhere else (like the execution of a rendering program).

Input memes that get transformed and exported as output might be named lucky or  success memes, those that failed to be exported are failure or flop memes.

figure 7: vocabulary of design memes

 

The approach taken in this exploratory study is very gross. The projects taken for consideration were from months to a few years ago and should be reconstructed from long term memory with the aid of what is left of filed away documentation. 

Transmission, variation and selection in designing however takes place at least on four levels (figure 8):

       micro level, where the design activity is concealed in the individual brain. Such an activity consist of memetic inheritance, at random generation of variety and subsequent selection. It takes place mentally in milliseconds. It is on this mental level that protocol research is aimed at.

       meso level, where we can see someone doing things like spending time in the library consulting journals, books and so on (transmission), like doing some sketching (variation) or talking with someone involved in the decision making process like the teacher, critic or the client involved (selection). All those activities takes up multiples of quarter hours to hours.

       macro level, where the activities are planned on weekly and montly intervals. There we become aware of another rhythm of transmission (problem statement and documentation at the beginning of the project), variation (designing and coaching on a weekly basis) and selection (again on a weekly basis, halfway the project after some weeks time and at the end after two or more months).

       super macro level, where the personal development from successive design projects where identified and elaborated. Where for instance in earlier work form as such was a goal, while in later design projects form appeared only as a means to obtain other goals.

This study was intended to deal with the macro level, but the students often tried very succesfully to analyse their design project work on the other levels as well, because they feeled that those levels were most significant for designing.

1. oldest inner cerebral cortex, centre of emotions and subconsciousness, artistic and reflex behaviour, pragmatics; 2. middle cortext, centre of the senses and preconsciousness, technical and sensomotoric behaviour, syntagmatics; 3. newest outer cortex, centre of logical reasoning and consciousness, scientific and inferential behaviour, paradigmatics.

Figure 8: four levels of design studies

 

Conclusions

As put forward in the introduction, the central question we wanted to answer was about the possibility of application of Darwinistic evolutionary theory from biology to architectural design. The answers of the students makes it highly probable to answer the question positively. The students were unanimous about the occurrence of memes in their design activities. They could name and depict memes appropriate in their processes of heredity, variation and selection. They could even imagine, detect and acknowledge the functioning of brainstorming effects in the arbitrary production of novel ideas or memes, but were still confused with the consequences of their own findings. The appearing autonomy of the (architectural) designer is then at stake with all the accompanying ingenious creativity and originality that brought along.

 

Philosophical remarks were also made about the applicability of the theory. Like history, linguistics or bicycle riding, no one expected to design (or decide, write or ride) better by understanding how those processes take place. But despite those limitations, they appreciated the exercise to understand better there own design behavior and development over the years. Nevertheless knowledge about the mechanisms behind those processes are essential if we want to simulate them for fun or training like in computer games and simulators or to explore the (self-defined) available design spaces in morphogenetic design programs.

 

What remain to be done is at least to look with closer attention to the evolution of usual architectural design projects. Obviously, the theory needs some adjustments because designing turned out to be a not outspoken population based activity. Students do not design by breeding whole populations of designs and subsequently do not select among them in order to allow the survivals to breed again and again until some optimised or compromised solution will emergence in the end. This corresponds to the so called breadth first strategy, which is hardly applied in architectural design. There, the so called depth first strategy is nearly always used, exploring and developing only one (or a few) alternatives at a time until eventually something important forces the designer to abandon the looked for solution and start anew. But if we take a look from the meme’s eye view like most of the students did, we have already experienced something quite different. From this a lot of questions are left behind to be investigated.

 

A last but not least issue is about artistic design. Although memetics claims to explain all culture, the memetics of engineering is according to Gatherer ‘certainly more approachable than the memetics of art.’ Architecture is something in between and might serves as a bridge between the two. The results up to now are encouraging, but detailed analysis of the artistic elements in the design work of the architectural students (and of artists and architects in practice) remain to be done.

 


References

Blackmore, Susan: The Meme Machine, Oxford University Press, Oxford, 1999

Darwin, Charles: The Origin of Species, Penguin, London 1985 (reprint of 1st edition, 1859).

Dawkins, Richard: The Selfish Gene, Oxford University Press, Oxford, 1976.

Dawkins, Richard: Universal Darwinism, in Bendall, D.S. (ed.), Evolution from Molecules to Men, Cambridge University Press, Cambridge, 1983.

Gatherer Derek: The Memetics of Design, in Bentley, P.J. (ed.), Evolutionary Design by Computers, 1999.

Steadman, Philip: The Evolution of Design, Cambridge University Press, Cambridge, 1979.