Brownian Motion Diagram. By NivedRajeev. CC BY-SA 4.0. Source: https://commons.wikimedia.org/wiki/File:Csm_Brownian-Motion_f99de6516a.png

It is hard to pin down exactly what “information theory” is. I think this works both in favour of Iverson and against her in this chapter. Where it might once (in the years immediately following the publication of Shannon’s landmark paper in 1948) have referred to a very narrow set of mathematical techniques applied to the analysis of communications systems, the conceptual fecundity of Shannon’s idea saw it taken up in dozens of fields of investigation. Shannon publicly bemoaned the uptake of his ideas outside of the strict context of communications systems in a short position paper in 1956 called “The Bandwagon”, but by that stage, the genie was out of the bottle: “information” found many applications, of which music (or, really, phonetics) was just one and “information theory” resolves into something more like a heuristic that can applied in innumerable contexts, rather than a stable field or discipline in its own right. From the historian’s point of view, we are immediately presented with a fairly acute case of the classic presentist/historicist problem: does Iverson tell a story of how ideas that we now understand to be part of the common ground of information theory were used in the WDR studios? Or does she want to tell the story of how a configuration of techniques that was legible as “information theory” to the actors was deployed at the time, in tandem with its emergence within a constellation of ideas?

Iverson takes a lot of care to rely on contemporary documentary evidence in order to build out the picture of information theory as one of a precession of “invisible collaborators” in the electroacoustic studio, taking one of the ANT axioms to bear on a mathematical theory as an agent (or “mediator”?) in the analysis of how a number of WDR compositions came about. By drawing on Meyer-Eppler’s textbook (1959, rev. 1969) and seminar titles, Ligeti’s claim for the functional equivalence of determinacy and indeterminacy (1958, pub. in German 1960, trans. 1964—by none other than Cornelius Cardew), and Stockhausen’s “…..HOW TIME PASSES…..” (trans. 1959) [“…wie die Zeit vergeht… (1957)”], contemporary sources dominate in the picture—though the composers themselves are not always as explicit about “information” as I’d want them to be—which, so painted, demonstrates an obvious concern with the use of statistics and research into the perceptual constraints on the human ear in the composition of the works analyzed. This (I would argue) narrower notion of information theory is already tied up in (but not reducible to), as Iverson rightly points out, contemporary intellectual movements, most notably cybernetics: the topic of the Macy Conferences, which Iverson characterizes in brief as part of the wider “domesticating, socializing project” (Iverson 2019, 20) of information-theory’s rehabilitation from wartime technology to universal cybernetic solvent.

So far, so historicist. But the case of Xenakis’s treatment complicates this. Iverson argues that one reason that we have not recognized an affiliation between Xenakis and the WDR group is that his use of “information theory” has not been recognized as such. Xenakis was interested in the use of statistical models (mostly of gases) and, as Iverson points out, there is a close correspondence between the structure of the strictly communications-theoretical formulation of information (on the one hand: in the WDR work) and the thermodynamic one, ex statistical physics (on the other hand: in Xenakis). Because of this link, we are invited to conclude that this similarity between the mathematical techniques was either not really recognized by the WDR actors or (more suggestively) was suppressed or ignored because of the lukewarm reception of Xenakis by the Stockhausen clique. Iverson summarizes the point so far below:

Xenakis latched on to the probability aspect of information theory most strongly, whereas Stockhausen, Koenig, Ligeti, and others in the Cologne circle mined the musical implications of perceptual Gestalten and the sampling theorem. These are all legitimate components of information theory discourses. Beyond missing information theory as the common foundation, perhaps scholars tenaciously grasp onto the conclusion that Xenakis and the Cologne–​Darmstadt composers are “quite different” because they seem to have repelled each other. (135)

I don’t think the WDR clique were all that invested in an understanding of the scope of information theory that allows Iverson to tie Xenakis’s investigations into the fold, which emphasize the connections with thermodynamics (A conception of information that was indeed contemporary – but seemingly unknown or of little interest — to the WDR composers. This notion of “information” sits in a scientific lineage including people like Léon Brillouin, Leo Szilárd, and even Erwin Schrödinger, whose 1944 book What is Life? was very widely read and introduced readers to the closely-related if underdeveloped notion of “negentropy” which gives information-theoretic discourse an organicist tinge unexplored in Iverson’s account). Writing in the late 1980s, the scientist Jeffrey Wicken claimed that “the concepts of entropy and information have developed in two noninteractive traditions—thermodynamics and information theory” (1987, 176); Katherine N. Hayles discusses some of the reasons for this distinction in terms of metaphor-use by scientists in Chaos Bound (1990, 31–60). Much has moved on in the historiography of information since then (for example, the work of William Aspray and Bernard Dionysius Geoghegan) and as Brian Miller’s work shows, there’s quite a lot to be said about the intersection between information and musical style that has been left unstated in a lot of empirical research into music and can be revealed by careful tracing of the filiation of information theory’s big ideas not only in composition but also in music theory/music psychology.

The point I’m trying to make is fairly simple: we’ve seemingly slipped away from a story that seems to seek out a historicist (or, if you like, spatio-temporally “emic”) conception of information theory to one that relies on noting the similarities between various scientific or mathematical formalisms from the vantage point of the historian. In the story of Xenakis, these differences, unbeknownst to (or, at least, unremarked upon by) the historical actors themselves become apparent to the historian who decodes or reveals the hidden, invisible, or otherwise covert links between diverse practices. Neither approach is unimpeachable: I’m not saying either option is superior. But it seems important to note that aspects of both approaches to the concept are discernable in this chapter. It might well be the case that Iverson’s cases studies suggest that Wicken is incorrect: that, among others, music composition was a discipline in which these two traditions were conflated—giving the lie to my insinuation that these traditions remained separate in the musical imaginary, at least. To suggest that, however, would suggest the the musicians who composed under an information-theoretic régime stood apart from mainstream scientific thinking in making precisely such a conflation. Their relationship to scientific orthodoxy is perhaps one that is yet more ambivalent than that which Iverson’s chapter—correctly—identifies.

References