Principles of structure building in music, language and animal song

Abstract

Human language, music and a variety of animal vocalizations constitute ways of sonic communication that exhibit remarkable structural complexity. While the complexities of language and possible parallels in animal communication have been discussed intensively, reflections on the complexity of music and animal song, and their comparisons, are underrepresented. In some ways, music and animal songs are more comparable to each other than to language as propositional semantics cannot be used as indicator of communicative success or wellformedness, and notions of grammaticality are less easily defined. This review brings together accounts of the principles of structure building in music and animal song. It relates them to corresponding models in formal language theory, the extended Chomsky hierarchy (CH), and their probabilistic counterparts. We further discuss common misunderstandings and shortcomings concerning the CH and suggest ways to move beyond. We discuss language, music and animal song in the context of their function and motivation and further integrate problems and issues that are less commonly addressed in the context of language, including continuous event spaces, features of sound and timbre, representation of temporality and interactions of multiple parallel feature streams. We discuss these aspects in the light of recent theoretical, cognitive, neuroscientific and modelling research in the domains of music, language and animal song.

Keywords: Chomsky hierarchy; animal vocalization; comparative perspective; computational modelling; language; music.

Figure 1.

Hierarchical organization of nightingale song. Panel (a) depicts a spectrogram of ca 2 minutes of continuous nocturnal singing of a male nightingale. Shown are 30 sequentially delivered unique songs. The 31st song is the same song type as the second one (both framed). The average repertoire of a male contains about 150 unique song types, which can be delivered in variable but non-random order for hours continuously. Panel (b) illustrates the structural components of one song type. Individual sound elements are sung at different loudness (amplitude envelope in (i)) and are acoustically distinct in the frequency range, modulation, emphasis and temporal characteristics (spectrogram in (ii)). Panel (c) illustrates the structural similarities in three different song types (i,ii,iii). Song types begin usually with one or more very softly sung elements (blue, b), followed by a sequence of distinct individual elements of variable loudness (green, g). All song types contain one or more sequences of loud note repetitions (pink, p) and are usually ended by a single, acoustically distinct element (yellow, y). Panel (d) illustrates that the same song type (i,ii) can vary in the number of element repetitions in the repeated section (pink). Spectrograms courtesy of Henrike Hultsch. (Online version in colour.)

Figure 2.

Analysis of Bachs chorale ‘Ermuntre Dich, mein schwacher Geist’ according to the GSM proposed by Rohrmeier [36]. The analysis illustrates hierarchical organization of tonal harmony in terms of piece (piece), phrases (P), functional regions (TR, DR, SR), scale-degree (roman numerals) and surface representations (chord symbols). The analysis further exhibits an instance of recursive centre-embedding in the context of modulation in tonal harmony. The transitions involving denote a change of key such that a new tonic region (TR) is instantiated from an overarching tonal context of the tonal function x in the key ykey.

Figure 3.

A Venn diagram of the Chomsky hierarchy of formal languages with three extensions annotated with a comparison of the hypothesized classifications of human languages, (human) music and animal vocalization. The areas marked with ‘question’ signs indicate that further research is required to settle examples for the respective class of complexity in these domains. (Online version in colour.)

Figure 4.

A finite-state automaton generating a repertoire consisting of two sequences: ABⁿC and DBⁿE (with n > 0). Note that the finite-state automaton is redundant in the way that it contains multiple instances of the same structure Bⁿ.