Music and Architecture. Revisited
COGNITIVE STRUCTURE OF MUSIC AND CONCEPTUAL ARCHITECTURE
Even to this day, the original Pythagorean ideas of harmonic proportion serve as the cornerstone of Western music traditions. From medieval plainchants to Baroque counterpoint and especially in the 4-part harmony of Bach’s chorales, the notion of musical perception is historically understood more through harmony than any other aspects of music such as rhythm, timbre and texture. Complete sets of rules and methodologies for composition and analysis of harmony, which is the root of western music theory, reached maturity as early 400 years ago. Compared to that, and indeed among all the “physical arts”, architecture is probably the least developed in studies of cognitive aspects.
Austrian theorist Heinrich Schenker was a pioneer in musical cognition studies. His argument of reducing the understanding of a complete work to a hierarchy of ideas relies entirely on musical harmony. Originating from Schenkerian notions of musical perception, in the 1980s linguist Ray Jackendoff and music theorist Fred Lerdahl developed a more systematic and graphical method of intuitive music analysis that they called the Generative Theory of Tonal Music. According to Lerdahl and Jackendoff, earlier attempts in explaining musical cognition (excluding Schenker’s) suffered from rigidity due to preoccupation with numerical studies in rhythm and harmony, or from the ambiguity of understanding intuitive affects of the musical experience. Comparatively their method has the advantage of avoiding surface qualities in music but instead focusing on deeper cognitive structures, and avoiding the need for an “experienced listener.”
Inspired by Chomsky’s generative linguistics, the theory that Lerhdahl and Jackendoff explain in Generative Theory of Tonal Music also inherited the notions of intuition and universality. In this light it is possible to compare their theory with the architectural thinking seen in Peter Eisenman’s investigations in the 70s.
Eisenman rarely mentioned music; Note on Conceptual Architecture only acknowledges visual arts at those points when Eisenman was seeking a conceptual art form. According to him, conceptual art is distinguished by its ability to change the primary experience of art from the visual and sensual to the mental and intellectual, or even by its suppression of the viewer’s or user’s sensual response. He also differentiates between art as an intellectual experience and as a mere vehicle for such experience. Such art may or may not have a physical reality, and while architecture has bigger trouble escaping built form, it can still be conceptual by involving pragmatics (functions) in the play of concepts.
Eisenman distinguished between two types of architecture. The duality seems not to be strictly between conceptual or non-conceptual but, as Mario Gandelsonas observes, between semantic and syntactic. Semantic architecture focus in application and modification on culture-based meaning, as in Venturi or Le Corbusier’s works, while syntactic architecture, like Terragani’s and Eisenman’s, possesses a more universal structure which may subsume the importance of surface meaning. The idea of the artwork in itself does not make for conceptual art: in conceptual art the concept has to exist in the object. Eisenman was also concerned with how the creators' intention can be carried over, through the means of a formal structure, so as to be apparent to the viewer. He believes this is achieved not by using signs to express meaning in the work, but through a deep structure of inherent relationships. It is here that Eisenman mentioned the universality of a syntactic structure without the need for cultural reference.
In architecture, Eisenman differentiates between a real concrete level of formal relationships and a conceptual level. The former is what the user can feel through her senses (perception, hearing, touching, etc.) and the latter is the abstract level as it exists in the object. Eisenman compared this duality to Chomsky's linguistics.
This is where the Generative Theory comes in. Not just because of the Chomsky connection, but because Eisenman’s idea of art is actually an old idea in music -- specifically, an idea seen in pre-20th century music.
As Lerdahl and Jackendoff point out, Schenkerian music analysis is weak in figuring out the structure of early 20th century serialist music (Schoenberg, Stockhausen), and completely powerless in understanding postmodernist music (John Cage, Philip Glass), probably due to the increasing trend of modern music to invoke cultural reference or semantic surface elements. But music ideals between the Renaissance and the 20th century were not based on these same kinds of ideas; in fact during the epoch of absolute music (second half of the 18th century) it was believed that the ideal form of music should be absolute, i.e. devoid of substantial or concrete meanings. This is slightly different from Eisenman’s description of sensual art, for music has never shied away from invoking emotion in the listener. Thus the similarity of absolute music and conceptual architecture lies in their non-referential nature, their lack of specific surface meaning, which in turns implies a reliance on deep / syntactical structures.
Both Lerdahl and Jackendoff and Eisenman face the question: is it valid to find analogies to linguistics, specifically Chomskian linguistics, in non-linguistic disciplines? Gandelsonas argues that the effects of Eisenman's use of linguistic terminology may be limited (or "cannot be referred in positive terms") because, among other reasons, in architecture there is limited knowledge about linguistics and semiotic concepts, and therefore transference to architecture may be problematic. To Gandelsonas, the implicit transposition of linguistic ideas into architecture, without defining the distinction of concepts between the two or acknowledging their differences in context, limits Eisenman’s theory. Gandelsonas also points out that instead of directly comparing the syntactic aspect of architecture to that in language, it might be more appropriate to use literature as a model because, like architecture, literature was performed by a few within a community of language users.
Gandelsonas had pointed out the Western and bourgeois background of the notion of universality, which forms the backbone for the Generative version of Schenkerian theory in music and also serves as a byproduct of Eisenman’s syntactic structures in architecture, since he imports idea from linguistics. Without being too specific, Gandelsonas seems to imply that the idea of an autonomous architectural theory is undesirable, and while that’s not what Eisenman was doing, Gandelsonas implies he was heading in that direction. Not surprisingly, Lerdahl and Jackendoff’s Generative Theory is often discredited by musicologists for similar reasons: for being rigid, subjective, and worst of all, Western oriented.
My goal here in relating hierarchal music structures to architecture is to demonstrate that some very traditional notions of music can be interpreted architecturally. Indeed, Eisenman didn’t continue his investigation into this kind of universal syntactic structure after the 1970s, and I am not sure if his method is a good vehicle for comparison or translation between music and architecture. Yet I have to stress that as an analytical too to understanding music cognition, Schenkerian analysis is still a core part of Western music theory—and in architecture we did have similar studies in Eisenman’s work. Therefore if we are interested in the relationship between traditional music and its application in architectural design, the Schenker-Eisenman analogy may provide some inspirations to architects.
But the true potential of music in architecture lies in the study of another architect / composer:
IANNIS XENAKIS
Iannis Xenakis believed that music is a strong condenser, perhaps the strongest of all the arts, meaning it has the ability to condense the collective of all human knowledge into increasingly abstract form. (As proof, he uses a comparison across history between parallel realizations in music and mathematics.) This “music” that he devoted his life to is, however, rather unique. He spoke of the ideal composer—an “artist-conceptor” who would have the ability to freely navigate across a “general morphology” of science which consists of various domains such as mathematics, logics, physics, paleontology, history... a sort of universality, but one based upon and oriented towards form and architecture. This is what he committed himself to.
XENAKIS: INTUITION
Xenakis was a strong advocate for intuition over rationalism. Rules are tools or principles for learning, and Xenakis rejected the rules that have made up so much of the Western art-music tradition: he was against all conventional ideas of harmony and counterpoint, the basic and centuries-old principles of music as a craft. Xenakis once said “we have to find a way to mix the properties of sound on a more abstract level without having to think about melody, time, harmony, etc. As a first step we choose sets out of the elements of the sound and play with them.”
In rejecting rules like this, Xenakis was one of the first 20th-century composers who realized the limitations in the total control developed with serialism—the ultimate in compositional rationalism in that it was strongly rule-driven and thought to be the definitive extension of Bach, Beethoven, etc. Xenakis rejected systems, and in doing this he chose to emphasize what musicologist Leonard Meyer called secondary parameters. Meyer divided primary from secondary musical aspects. The primary ones are those that are syntactical, and usually arranged in class relationships, while the secondary things are statistical & non-hierarchical. Pitch is a good example of the primary, while texture & timbre are examples of the secondary. Most musicians would probably agree that primary things like pitch tend to be rule-governed while secondary aspects like texture are more of an intuitive part of the composer’s expression.
Xenakis went a bit further when he said that art is by nature intuitive: “art… has an aspect which needs no proof. Suddenly we understand something: that is revelation, intuition. That’s what renders art more free than the natural sciences… the revelation for beauty occurs immediately, directly, to someone ignorant of art as well as the connoisseur. This is the strength of art and so it seems, its superiority over the sciences”
Xenakis was interested neither in “rules” nor in aesthetics, the two things that had actually formed the foundation for so much of the Austro-German classical musical tradition. In 1954, he found probability theory to be the compositional tool he needed – a total avoidance of rules. Probability is a study of masses: and in Xenakis’s Pithoprakta and Metastasis listeners perceive overall global shapes (as opposed to apparent small-scale, local organization or disorganization), successions of events, the growth of one stage or tendency into another, the surprise of seeing a momentary pattern coming out of an otherwise nebulous state.
XEANKIS: UNIVERSALITY
In a conversation with Michael Serres, Xenakis mentioned that the concept of a general morphology of human knowledge can be traced back to antiquity; the idea of proportion as applied to architecture on man-made forms was such an example. This is different from endorsing the Pythagorean studies—he mentioned earlier that there probably is no analogy between harmonic proportion and mathematical invention, for the relationship between the two is more one of cause-and-effect than a proof of universality.
Yet, even more than Eisenman, Xenakis aimed for true universality, a truly non-contextual art. In his Texas Stretto House, Steven Holl overlapped Baroque fugal technique and Bartók’s modern application of the ancient Greek golden section, but Xenakis saw most techniques from the Western music tradition as hindrances instead of tools. To Xenakis, counterpoint rules are not universal at all, even though they rely strongly on the kinds of syntactic structure that qualify for universality to Eisenman. For Xenakis such rules are products of a particular cultural context. For instance, polyphony of the early Middle Ages consisted almost entirely of octaves and fifths, the same intervals that were later forbidden. Xenakis thought that emotional response to music was also subject to change — part of a piece’s “subjective coloring” and not of “the music itself.” To him the same music will evoke quite different reactions in two or three hundred years, because the terms of reference of the listener will be completely different. The intuitive, by contrast, is sudden and non-rational and timeless: “Suddenly we understand something; that is revelation, intuition … Beware of aesthetics… replace it with something that is manageable, treatable.”
XENAKIS: ARCHITECTURE
Eisenman mentioned that one drawback of conceptual architecture compared to other conceptual arts is that it has to have a built form. If art, science, and even the whole of human knowledge are indeed a continuous general morphology as Xenakis said, then looking at his music compositions per se would be sufficient to understand the motives behind the works he built during his time in Le Corbusier’s office.
Pair 1: Monastery of La Tourette vs. Le Sacrifice
Mikesch Muecke suggested that Xenakis’ musical and architectural investigations may begin with his involvement in the Modular. In Le Sacrifice he applied the Fibonacci series in construction of the entire musical edifice (the first 8 numbers), turning the idea of musical rhythm from an externally determined framework into a “density” issue. The same principle is used in La Tourette: in order to develop the façade as a “windowed outer skin” as La Corbusier requested, Xenakis experimented with permutations of a set of window panes of different widths to obtain a rhythmical motif. By using his intuition with regard to density, Xenakis solved the problem of too many elements making it impossible to aesthetically control the resulting configurations: rather than considering the individual distances between the upright casings, he focused on considering zones in the façade where a higher or lower number of casing per length unit would be required and then decided how the transition between these two states would occur: fluently or abruptly. He derived the increasing width of the windows in golden sections from the Modular.
The resulting complexity of façade for La Tourette can be read as a physical polyphony of light and shade: while each level of windows has its own distinctive rhythm, together they form a continuous texture of a synchronized whole. This principle eventually became the cornerstone of the complex rhythmical polyphony in his musical compositions. (I should mention, however, that Xenakis has never been interested in any analytical reading of artworks in terms of hierarchical structure, so comparison to linguistics should be kept to a minimum.)
Le Sacrifice isn’t the only composition where we can find analogies with La Tourette. On a macro level, Xenakis’ compositional thinking in Metastasis had already appeared in the overall design of La Tourette (which Xenakis actually had heavy involvement in well before he designed the façade of the building). In Metastasis he started with a top-down mentality where, instead of beginning with a lone idea and expanding it onward, Xenakis conceived the whole in mind and worked his way down with elements and proportions. In short, to judge from the example of La Tourette he found architectural problems the same as those in music.
Pair 2: Philips Pavilion vs. Metastasis
Xenakis’s next investigation was the idea of continuous change, as both an aural and a built phenomenon: “… I was interested in the question of whether it is possible to get from one point to another without breaking the continuity. In Metastasis this problem led to glissandos, while in the Philips Pavilion it resulted in the hyperbolic parabola shapes… What is a straight line in two-dimensional space? The continuous change of one dimension compared to the other. The same happens in the pitch versus time domain: the straight line is the continuous change of pitch versus time.” The glissandos in Metastasis, the most famous elements of that piece, are therefore carried over almost as-is into physical form in the Philips Pavilion. I would say that music and architecture, from the composer’s / architect’s intuitive intention to the development of the work to its understanding by listeners / users, have never been so close in history. Tackling both problems with the same solution, Xenakis created two entities that can only be perceived as a whole and not as a product of small elements.
And the significance of this parallelism is more than a compositional motive; it also works as an Engineering solution, expanding that “general morphology”. In an earlier project like the Unité d’Habitation de Marseille, Xenakis had already solved Le Corbusier’s structural problem of creating a free-standing platform with the strength of reinforced concrete. In the Philips Pavilion he took a radical approach: beginning by experimenting with sections of conoids, he arrived at hyperbolic paraboloids (or S-curves), also built with reinforced concrete. The strength of the shape was so strong that after the expo they actually had problems demolishing it.
XENAKIS: CONCLUSION
An Architectural perspective would allow us to focus more on the aspect of Xenakis translating musical work to architectural applications, but in the works above it is not difficult to see that his music and his architecture affect each other deeply. One proof of that affinity is his idea on time in music…
“I suddenly realized that it is not true that music is only time, as Stravinsky claims in one of his books. Messiaen is also of a similar opinion: music is nothing without time. In fact, music is basically outside time and time serves only for it to manifest itself.”
Before the invention of modern notation in the 14th century, music was conceptually similar to time, as it was perceived as a collection of successive and discrete “nows.” i.e. immeasurable and indivisible. Xenakis was not the first one who broke this barrier; Philip Glass, Earle Brown and John Cage’s works also investigate the non-hierarchical temporal possibilities of music. But it was Xenakis who was the most confident with this concept, as he probably had the most profound understanding through the interactions between his music and his architecture.
Xenakis died 8 years ago, but his architectural career ended long, long before that. Thus the above two architectural examples of his are probably primitive with respect to his overall life’s studies. Metastasis and Pithoprakta were merely the foundation on which he developed his ideas of stochastic music, and sadly he didn’t have the chance to make the same kinds of developments in architecture.
Where does this leave us? I feel comfortable in saying that so far no composer has continued Xenakis’ method with the same conceptual vigorousness and thorough understanding. As for comparing him to contemporary architects—I am really not sure. I could compare works of composers like Glass, Cage or Earle Brown to Liebeskind or Hadid’s architecture; I see some vague similarities in the conception of their works. But for Xenakis I am proposing Greg Lynn as the closest comparison.
Whenever I read about Lynn’s resistance toward pseudo-science or “happy accidents”, or how he avoids using scientific analogies when there isn’t a solid connection in architecture, I am reminded of Xenakis’ condemnation of Stockhausen’s and Boulez’s aleatoricism. In my view, the analytical method that Lynn used in designs such as his entry for the Port Authority Competition recalls Xenakis’ works. Not the Philips Pavilion or La Tourette, so much as his later compositions. Lynn himself may not agree, for he is especially modest about the project, in person and in his book: “completed in less than a week,” he says, “pseudo-quantitative indexing of statistical data”. But it is exactly this kind of insistence on conceptual purity that reinforces his resemblance with Xenakis. Lynn’s more recent studies in parametric design may be less comparable to Xenakis’ works, but in the end it is Lynn’s precision and vigorousness that matter the most.
MY OWN WORKS
As much as I have thought about the issue, in my six years of studio experience I have tried only twice to incorporate music. The first time I was a senior, when I had the idea of creating a truly integrative experience in music; not only did I imitate Eisenman's syntactic structure, but thought I could include Edmund Burke's idea of the sublime, some harmonic analysis, Schenkerian theory, and many other elements into one studio design. And I did come up with a product. Whether I achieved my claim or not, I have no idea. As embarrassing as it is too see it today, back then I did truly think that I could translate one sublime experience into another through the medium of architecture.
My second attempt happened just this year; this time I am more aware of other aspects in architectural design, thus music was relegated to a generative tool, or as Greg Lynn calls it, a way of “validating the design”: in other words, a manner of creating interesting shapes, façades, and so on. I feel quite dishonest about the process: I was claiming it to be the built form of a musical work when it wasn't. But in the end it is true that the design was conceived with a certain spark, and the spark was music.
I really can’t think of a good way to close this talk. I entered architecture school 10 years ago, and I have never stopped rethinking what design is, because I was never sure about that. Today I have presented to you something that I honestly believe has potential in architecture, especially since there seems to be a renewal of interest in the past few years. I don’t know if this will provide the answer to my problem, because I can’t even begin to compare myself to any architect I have mentioned, and especially not to Xenakis—I consider him a truly unique figure, whose standard is simply way too high to reach. But this is the direction that I want to continue working on.
Bibliographies
Alberti, Leon Bettista. Ten Books on Architecture. Translated by Cosimo (Italian) Bartoli and James (English) Leoni. Edited by Joseph Rykwert. Vol. 5, Tiranti Library. London: A. Tiranti, 1955.
Cook, Nicholas. Music: A Very Short Introduction. New York: Oxford University Press Inc., 2000.
Eisenman, Peter D. "Notes on Conceptual Architecture: Towards a Definition." Design Quarterly, no. 78/79 (1970): 1-5.
Gandelsonas, Mario. "Linguistics in Architecture." In Architecture Theory since 1968, edited by K. Michael Hays, 112-23. Cambridge, MA: MIT Press, 2000.
Lerdahl, Fred, and Ray Jackendoff. A Generative Theory of Tonal Music. Cambridge, MA: MIT Press, 1983.
Rappolt, Mark. Greg Lynn Form. New York: Rizzoli, 2008.
Taruskin, Richard, and Piero Weiss, eds. Music in the Western World: A History in Documents. Belmont CA: Schirmer, 1984.
Vitruvius, Pollio. Vitruvius : Ten Books on Architecture. Translated by Engrid D. Rowland. New York: Cambridge University Press, 1999.
Wittkower, Rudolf. Architectural Principles in the Age of Humanism. New York,: Random House, 1965.
Wright, Frank Lloyd. Frank Lloyd Wright: Collected Writings. Volume 1 1894-1930. Edited by Bruce Brooks Pfeoffer. 4 vols. Vol. 1, Frank Lloyd Wrights: Collected Writings. Rizzoli: New York.