Monday, October 10, 2011

A history of music cognition?

One of the pioneers in the field that would come to be called music cognition was H. Christopher Longuet-Higgins (1923-2004). Not only was Longuet-Higgins one of the founders of the cognitive sciences (he coined the term in 1973), but as early as 1971 he formulated, together with Mark Steedman, the first computer model of musical perception. That early work was followed in 1976 with a full-fledged alternative in the journal Nature, seven years earlier than the more widely known, but, according to Longuet-Higgins, less precisely formulated, Generative Theory of Tonal Music of Lerdahl and Jackendoff. In a review in Nature in 1983 he wrote somewhat sourly:
‘Lerdahl and Jackendoff are, it seems, in favor of constructing a formally precise theory of music, in principle but not in practice.’
Although Lerdahl and Jackendorff’s book was far more precise than any musicological discussion found in the leading journals, the importance of formalization cannot be underestimated. Notwithstanding all our musicological knowledge, many fundamental concepts are in fact treated as axioms; musicologists are, after all, anxious to tackle far more interesting matters than basic notions like tempo, meter or syncopation, to name a few. But these axioms are not in actual fact understood, in the sense that we are not able (as yet) to formalize them sufficiently to explain them to a computer. This is still the challenge of ‘computer modelling’ (and of recent initiatives such as computational humanities) – a challenge that Longuet-Higgins was one of the first to take up [Excerpt from Honing, 2011]. Longuet-Higgins, H. C. (1983). All in theory — the analysis of music Nature, 304 (5921), 93-93 DOI: 10.1038/304093a0 Longuet-Higgins, H. C.  (1976). Perception of melodies Nature, 263 (5579), 646-653 DOI: 10.1038/263646a0 Honing, H. (2011). The illiterate listener. On music cognition, musicality and methodology. Amsterdam: Amsterdam University Press.

1 comment:

  1. Great little post on music cognition. I never knew what axioms meant until you referred to them in your post. Also because I had never seen the word axioms used in writing or in general. Now I can finally describe my writing correctly! Although, I didn't fully understand the last part. What did you mean by formalizing these axioms to explain them to a computer?