Musical Grammar and the Creative Process

New things are made with old things. In this sense, all creation is basically a combinatorial process – though not just that. It also implies novelty; but novelty comes into play in a dynamic space between random indeterminacy and routine regularity. This raises two basic questions about human cognition: how rules of combinatoriality arise, and how the freedom beyond the rules may be exploited.

It is plausible that these questions can be generalized over various domains. A particularly tempting hypothesis is that certain forms of musical, hierarchical-temporal ordering share cognitive resources with linguistic syntax. This has particular relevance for the metrical-harmonic tonal system which has come into being before 1600, and has been in use and constant transformation ever since. As a collaborative creation which has emerged from numerous interrelated individual practices, tonal organization has at least a superficial analogy to linguistic grammar. Empirical evidence favoring such a cognitive overlap (Patel 2013, Koelsch et al. 2013) has been contested (Bigand et al. 2014). Much depends however upon the definition of ‘syntax’, which varies between the musical and linguistic frameworks.

On the basis of these considerations I propose the following hypotheses.
(1) Tonal music has a harmonic-metric syntax which extensively exploits certain standard patterns (schemata).
(2) In this particular aspect (and possibly other), Western tonal music shares cognitive resources with other domains, in particular linguistic grammar.
(3) The development of a cognitively plausible theory of musical syntax may therefore benefit from the study of linguistic grammar.
(4) At a basic level, the creative imagination takes off from the surfaces and interfaces of these musical-grammatical schemata.
(5) The study of these schemata may allow us to interpret the musical score as partial evidence of the composition’s genesis.

The most influential attempt to model a theory of music upon linguistic grammar is Lerdahl and Jackendoff’s Generative Theory of Tonal Music (1983). Of many possible points of criticism, the most pertinent for present purposes are the following. (1) The theory follows Generative Grammar in its minimization of a possible cognitive overlap between domains. (2) Following Generative Grammar, which analyses and explains sentence structure, it pictures the composition as a huge sentence, an idealized and timeless totality constructed in the listener’s mind. (3) Despite its stated aim of being a cognitive theory of music, it heavily relies on non-explicated concepts from traditional didactic music theory. (4) As a parser, rather than a fully developed grammar, it attempts to account for passive understanding only.

Generative Grammar constitutes one paradigm within linguistic theory. Its aim is to be strictly algorithmic and self-contained. Its potential attraction for music theorists depends on the formalist aesthetics of ‘absolute music’, paradigmatically embodied in the work of Heinrich Schenker.

Various alternatives in linguistics may be broadly characterized as ‘functionalist’. These include Cognitive Grammar, Construction Grammar (Goldberg 2006), and Emergent Grammar (Hopper 2011). With different methodologies these study language in terms of its functions and of general principles of cognition and behavior, rather than as the product of an innate, inalterable and highly specialized faculty. Viewed as a process of giving and recognizing meaning, language appears to be something makable and in the process of being made. The clue to linguistic creativity is the ability to adopt and apply patterns or schemata at various levels of complexity and abstraction: from the word (a phonological-syntactic- semantic schema) up to the full sentence cliché.

Within music theory and historical musicology, appreciation of the constructive- combinatorial element involved in musical creativity has steadily grown since the 1970’s. The first comprehensive (re-)construction of a thesaurus of musical compositional commonplaces or schemata (Gjerdingen 2007) is based on a didactic tradition of the 17th -19th centuries, and has been reintroduced effectively as a tool in improvisation. Its theoretical underpinnings and analytical application are however under debate. Most problematic are (1) questions of the structural levels at which these schemata are situated, and of their interrelations; and (2) the precise definition of ‘schema’, as against ‘prototype’ or ‘model’. These are problems which call for a broader scientific, and particularly comparative musicolinguistic framework.

Many musical patterns imply certain restricted combinatorial possibilities. To the extent that the structure of the composition can be explained as a realization of these possibilities, the compositional process becomes transparent (Dreyfus 1996). Music most valued in the classical-romantic repertoire involves many complexities not reducible to these combinatorial factors. But with sharper distinctions between the individual-particular and the general-schematic, and a consistent rooting of surface schematism in the foundation of tonality - in other words, with a musical grammar - score analysis may provide fragmentary insights into the creative process.

The composition, as symbolized in the score, reflects its genesis only indirectly and incompletely. Still, there are secondary factors which lend the composition some transparency towards its creation. Trivially (but importantly), the factors time and labor have often precluded extensive revision and rethinking, in particular in the Baroque and Classical eras. Under such stress, composition approaches improvisation on paper. Less trivial is the fact that Classical-Romantic aesthetics favors a view of the composition itself as a kind of discourse or thought process. This, to be sure, is an imaginary discourse or thought process; essentially a piece of fiction, not a direct reflection of the composer’s actual musical thinking. Even as such it implies an idealized image of ‘musical thought’ which may have guided the composer in his actual thinking. Following this discursive ‘thought’ attentively, the listener is (ideally) recreating the composition. Since all thought is closed off from the world, we may feel privileged in sharing this act of creative communication.

The argument outlined above will be illustrated with the analysis of an excerpt from the works of Beethoven or Schubert.