Collecting Metaphors

José Mario Gutierrez Marquez




I have a small practice (meaning around 20 architects) already since more than 20 years together with my two Italians partners, our size is small but I would dare to describe our reputation as unblemished. I mention this because it retains relevant for you to understand that we are very, very busy architects and we don’t have much time left in our daily routine and even less the competence to reflect or interrogate our own cognitive behavior, which is what I am going to try now.


I am also teaching Architectural Space and Design at the Bauhaus Universität Weimar where, on the contrary, I have found the time for some reflection and to collect some information. These are notes taken for my lectures in the University, they have no pretension, no aspiration to be in any way regarded as the result of rigorous, methodical interrogation, if they have any value, I suppose it could be that they give an insight for better minds on the role of metaphors in the design mechanisms of our discipline.



Note 1: Finding metaphors / Lexicon


I suspect that Architects have always been collectors of metaphors without being aware of it, or in the linguistic patois…we have been collecting potential sources to future yet unknown targets. I could even argue that architects are in a symbiotic relationship with metaphors since unknown time, but again, we are hardly aware of it.


Some of us call them “images”, some of us believe they are just casual formal similarities, some of us, like Le Corbusier, with subliminal intuition, collected them under the name of “Objects of poetic reaction”, a very appropriated name.

Le Corbusier selected these objects for their potential to become sources in a yet not initiated, cross-domain transference of attributes from made things to things yet to be made, a collection of architectural riddles.

Fig. 1: Le Corbusier, “Objects of poetic reaction”.
Fig. 1: Le Corbusier, “Objects of poetic reaction”.

I found another good example a couple of years ago, when a well known architecture office from Berlin (AFF architects)(1) was invited to present their work in an architectural gallery. Instead of the usually expected models and drawings from projects and buildings, they decided to show, accurately disposed on a single big table a collection of objects and around it, hanging on the walls of the show-room, pictures, post-cards and images grouped in clusters or “families” that showed a common “relationship”.


(1) AFF Architekten, “In Love, To:”, exhibition in the DAZ (Deutsches Architektur Zentrum Berlin), January 2011,

Fig. 2: AFF Architekten, “In Love, To:”, exhibition in the DAZ (Deutsches Architektur Zentrum Berlin), January 2011.
Fig. 2: AFF Architekten, “In Love, To:”, exhibition in the DAZ (Deutsches Architektur Zentrum Berlin), January 2011.


All of us architects understood the concept without further explanation because, I believe, we all have our private collections of metaphors in form of pictures, objects and even gestures.


In our own practice becoming aware of the presence of metaphors and their role in our design routines has been a gradual process. When we discuss ideas for a project, concision is expected, stringency a necessity; complex compositional operations need to be explained in a very short, economic, compressed way. But, inevitably, the transcription from three-dimensional concepts into language is arduous, we have to do without the superfluous, we have to do without detail; we have to be unburdened from unnecessary specificity in order to keep the mechanism of our exchange light and agile. The conceptual core of an architectural approach consists of a cluster of interconnected blocks and is always a trial and error exercise before it finally crystallizes.  


We thought that the use of metaphors was a way to oil the mechanism of our exchanges, short-cuts to avoid long explanations. Verbs like “polarize”, “dock” or “extrude”, terms like “excavated solid”, “cast” or “inlay” or concepts like “Brownian motion” or “interference pattern”, obviously loaned from other disciplines’ technical vocabulary, were constantly used in our internal debate, referring to very complex compositional operations.


This vernacular vocabulary has been growing through the years, distilled by our experience, accumulating a number of terms that, in time, have become a code for compositional operations, a conceptual archive.  The first time that we tried an inventory of these terms, we named it “lexicon”, we became also aware that our “lexicon” was actually a collection of metaphors.

It was then clear that these metaphors were no longer mere “figures of speech”, no longer Aristotelian rhetoric devises, they were templates for concepts, sophisticated tools that have mutated from being a device to catalyze our exchanges to a storage system for complex concepts. In architecture, similar problems are usually solved in similar ways, and our metaphors, ready-made concepts for compositional operations that have solved similar problems in the past, kept reemerging spontaneously at each recurrent situation.


These are some examples of our lexicon when we were still unaware of the cognitive linguist approach to metaphors and how we retrieved the stored information and how we transcribed what we retrieved.



Note 2: Understanding Metaphors 


„Designing a product is designing a relationship“, Steve Rogers 


Maybe for you to better understand this quote, I should first explain some aspects of the design process in architecture. 


I believe we architects secretly envy disciplines like music, literature or dance because, unlike ours, it’s possible to test and if necessary correct the progress of the work before the final performance, in other words rehearsal is possible and an instrument of the design process. In these disciplines the feed–back resulting from rehearsals, of the same nature and complexity as the final product, is an essential part of the design process. The creative process from the performing arts is a constant and unmediated accretion of understanding through the insights won by rehearsal. In the performing arts experience is constantly feeding on experiment.


Alas, in architecture there is almost no room for experiments, no chance for rehearsals. The design process in architecture takes us from an initial abstract concept (usually a complex interrelated set of concepts) to a physical object, a constructed object...a building. It has no feed–backs from insights obtained by rehearsals, there are no “takes” in architecture, architecture could be understood as a performing art without rehearsals, there is only one irreversible performance. 


Our design mechanism recourses to all kinds of notations to encode what we can’t rehearse, constrained to simulate and test our compositional operations in an abstract way (no rehearsals), we developed “conceptual scripts”.  The design process is actually a transcription process between our “conceptual scripts” through levels of increasingly specific notations that negotiate between them and the final constructed building. 


During this process, the assigned relationship between the two sets of conceptual/abstract parameters (“conceptual script”) and perceptual/physical ones (the building) endures an intense “shelling” from functional, economic, technical and even psychological demands. We gradually integrate these demands in a long journey across increasingly complex notations (models, sketches, 3D simulations, plans, details). We must not lose sight of the original conceptual set and its attributes in the multiple transcriptions leading us to the physical final product. The consistency of the relationship between concept and building must be preserved, the assigned attributes must not be ”lost in translation”.


So back to our quote: In the architectural design process we constantly watch the consistency of the assigned “relationship” between a conceptual domain and a physical domain, we design the relationship, we edit incoherence, dissonance, we keep track of the consistency of the cross domain map. I am certain that there is a clear correlation between the concept of “cross domain mapping” in cognitive linguistic and the transcription steps in the architectural design mechanism. 


Having reached the field of cognitive linguistic we became conscious of lacking the necessary training to find our way. Unsure on our feet, we tried a first interrogation of the term and soon left behind Aristotle, Brodsky, Borges and many others to finally reach the cognitive linguistic approach to the mechanism behind metaphors. We came in contact with Cognitive linguistic and the work of Lakoff, Johnson, Fauconnier, Turner and many others. We met conceptual approaches and terminologies that finally brought light in many recurrent situations in our design routines.



Note 3: A window is a picture - Cross domain mapping


A very clear example can be found in the disposition of the windows for our design of the city library in Köpenick. When trying different schemes of placing the windows on the outer wall, placing them became “hanging” them, we started to understand our windows as if they were pictures being hanged. 


Once this metaphorical “equation” WINDOW-PICTURE settled in our design routine, it started to model our understanding of the window itself.


If we try an explanation of what happened, taking assessments from an article by Gentner and Bowdle and apply them to our design process, we can observe the stringency of the correlation:


“Metaphor comprehension begins with a symmetric (non directional) alignment process.“ (Gentner and Bowdle, 112)


In our case the alignment was “a window is a picture, a picture is a window” and consisted, as far as I am aware, of the following correspondences:


1. Window, orthogonal frame = Picture, orthogonal frame (most of the time)

2. View  = “Picture”

3. Mounted on a wall = Hanging on a wall (most of the time)


Let’s go back to Gentner and Bowdle: 


“If an alignment is found, then further inferences are directionally projected from base to target.“ (Gentner and Bowdle, 112)


So once established, the alignment started to “dictate” further adjustments, it became directional from source (the picture) to target (the window) - like:


1. If Pictures have no movable parts and no hinges, it follows that we would have to do without them on our windows.

Consequently we modified our ventilation system, the ventilation was performed no more by the windows but by skylights and windows in secondary rooms connected to the reading hall, in this way we avoided the necessity to open the windows and did without hinges and openers. 

Fig. 3: BFM Architekten: Mittelpunktbibliothek Köpenick
Fig. 3: BFM Architekten: Mittelpunktbibliothek Köpenick "Alter Markt"

2. If a window is a picture, then it has to hang on the wall and not be inserted in it.

To avoid what we call “cold bridges” (another nice metaphor for you there), meaning missing places on the continuity of the insulation that allow energy to flow between inside and outside, we have to place thermal insulation around the window, we did it in such a form that a frame was needed to cover it, a frame “on the wall”.

Fig. 4: BFM Architekten: Mittelpunktbibliothek Köpenick
Fig. 4: BFM Architekten: Mittelpunktbibliothek Köpenick "Alter Markt"

So the metaphor “a window is a picture” had hijacked our conceptual approach and was dictating the attributes of the physical outcome: No hinges or movable parts, a frame placed on the wall and, of course, we still have a view asa  picture. 


And still Gentner and Bowdle keep guiding us: 


„Predicates connected to the common structure in the base, but not innitialy present in the target are projected as candidate inferences in the target“ (Gentner and Bowdle, 112, Emphasis in the original)


Our general scheme of the windows still didn’t have a final configuration, during the design process we were testing “random” dispositions of the windows of the reading hall, when at some moment someone mentioned the “Petersburger Hängung”. 


Fig. 5: „Picture gallery in the St. Petersburg Eremitage“ (Eduard Hau, 1860)
Fig. 5: „Picture gallery in the St. Petersburg Eremitage“ (Eduard Hau, 1860)

This was a way of hanging pictures on the Eremitage Museum in San Petersburg – also known as salon-style hang – and actually not related to our looking for a disposition of our windows in our reading hall, but our debate was already contaminated by the equation WINDOW=PICTURE, and “inferences” were emerging without being asked.

Due to our space concept with double height spaces docking at the façade, we were free to place windows unconstrained by floor height, so we “hanged“ our windows as it was done on the Eremitage as you can see in the images. 

A zen saying states:


„When the flower arranger arranges the flowers, he also arranges his mind and the mind of the person who looks at the flowers“

Metaphors work as conceptual forces, inevitably leaving traces of the pressure applied on the target conceptual domain that we try to grasp through them.






Gentner, Derdre &  Bowdle Brian (2008) 'Metaphor as Structure-mapping'. In: Gibbs, Raymond W. (ed.) Cambridge Handbook of Metaphor and Thought, Cambridge: Cambridge University press, 2008, 109-128.