In his 1989 article `Reflections on the development of semiology of music' Jean-Jacques Nattiez offers an excellent review of the history of musical semiology. In it he gives an historic perspective on the fundamental issue of the nature of musical signification. Nattiez distinguishes between intrinsic and extrinsic significations within musical semantics, finding the theory of the former to be to a large extent founded on the work of Nicolas Ruwet and the notion of music as a language that signifies itself [Nattiez, 1989, pp. 30]. Jean Molino summarizes Susanne Langer's idea of music as the `unconsummated symbol' and captures the essence of the problem: ``On the one hand, the unchallengeable presence of evocation; on the other, the impossibility of exploiting it'' [Molino, 1990, pp. 126-7]. Molino aims at a theory in which music is understood as networked communication or exchanges between individuals. As we will discuss more thoroughly in the next section, the sender and receiver do not have to come to the same understanding of the message, or the `trace' as Molino would call it, hence there is no need for a understanding of the `code' which is significant to the semiosis favored by Umberto Eco. Eco points to the problems with connecting the investigation of a sign with the object to which it refers. It is impossible to attribute logical statements such as `true' or `false' to the semiological investigation of music and for Eco these are pre- or postsemiotic problems; ``The signs are of interest to semiotics as social powers'' and further ``Any attempt to establish the referent of a sign will force us to define this referent with the terminology of an abstract entity.'' This is what Eco calls the ``cultural convention''. [Eco, 1971, pp. 61-6]
Defining a cultural context as the referent resolves some issues in the analysis of performed music as a social fact. The listener or concert-goer can be defined as belonging to a cultural entity with predetermined understandings of the context of the performance, but also of the cultural markers within the music. This cultural entity may then be used as a code to decipher the message (the music as a symbolic system). However, in our study we are looking at a not yet existing work - a work in progress - and we are not primarily interested in the symbolic understanding of music as it is materialized in the physical world. Our focus is geared towards the understanding of the actions that lead to production of musical content. Following Eco's model we might try to approach this symbolic system in relation to a common context, or subculture created by the agents involved in it. Both composer and performer are working within the frame of their own cultural contexts which defines their respective understandings of the evolving work. The subculture is a result of interaction, and negotiation ('What is it we are developing?', 'How are we talking about it?', etc.), between the two agents and their inherent cultural contexts. Their mutual expectations and their understanding or imagination of the work in progress is of importance when they attempt at co-ordinating their actions, for instance towards a definition of the performance instructions. The musical work becomes the sign or the message, the agents the signifiers and the subculture the signified. Where, traditionally, we may tend to regard the composer/performer relation as a hierarchic structure in which the role, even the purpose, of the performer is to fulfill the composer's intentions (whether he is dead or alive), this mode of analysis allows us to look at the two agents as part of a larger system that may also contain many other agents.
But to fully understand the dynamics of the context, or subculture as we call it, we also need the tools to move to a lower level of analysis. The tripartite model suggested by Molino for analysis of music, though certain aspects of it remains problematic, appears to be a flexible method for our study at this stage.
The ambition of musical semiology has been to provide tools for an analytic understanding of the total symbolic fact of the musical work [Nattiez, 1990, pp. 34]. Molino argues for a three level symbolic analysis; ``the poietic, the esthesic and the `neutral' analysis of the object'' [Molino, 1990]. Three modes of analysis all representing the same work of art. The analysis at the different levels does not necessarily have to lead to the same conclusions or results but, according to Nattiez, it may help us to understand all aspects of the musical work:
...recognizing, elaborating, and articulating the three relatively autonomous levels (poietic, neutral and esthesic) facilitates knowledge of all processes unleashed by the musical work, from the moment of the work's conception, passing through its `writing down', to its performance. [Nattiez, 1990, pp. 92]
Leaving the problematic concept of the neutral level aside3, a rudimentary definition of the two terms `poietic' and `esthesic' from a musicological point of view indicates that an analysis of the (external) poietics of the work takes ``a poietic document - letters, plans, sketches'' as its point of departure whereas an analysis of the (inductive) esthesic ``grounds itself in perceptive introspection'' - that which is ``perceptively relevant'', that which one hears [Nattiez, 1990, pp. 140-3]. The three ``families of analysis'' correspond to a:
semiological `program' [...] that has three objects:
[Nattiez, 1990, p. 15]Though the `material reality' and the `physical traces' are not as self evidently defined as a result of only the poietics of the work, it is the processes themselves rather than the analysis of the processes that are of interest to us in this paper. (In the study that we performed following the methods developed here it will also be clear that neither the poietics nor the esthesics belong to only one aspect of the work.) The term `poietic' can be traced to the Thomistic philosopher Étienne Gilson whose definitions are less concerned with the analysis and more with the actual processes. According to Nattiez:
With `poietic' Gilson understood the determination of the conditions that make possible, and that underpin the creation of an artist's work - thanks to which something now exists which would not have existed, except for them. [Nattiez, 1990, pp. 12-3]Taking this short statement as a definition it may be argued that also acts of interpretation (and analysis) involves a poietic dimension.
Nattiez further discusses the issue of where the poietic process ends and the esthesic begins in score-based music (ibid, pp. 72). For Nattiez this is in essence an ontological discussion: What is the musical work, is it the graphic sign alone or is the musical work incomplete before it is realised as sound in performance? Contrary to our discussion in Section 2.1, Nattiez finds that the greatest difference, between the score and the acoustic trace left by a performance, is that while the score is ``an invariable physical reality'' there are just as many acoustic realisations as there are performances. The performance is the borderline between the esthesic and the poietic field. By focusing on the act of interpretation as it is performed between the score and its sonifications (``the interpretants that insinuate themselves between the score and its performance'' (ibid)), he draws the conclusion that analysis of the neutral level has to be applied to ``the graphic sign alone, because that sign precedes interpretation'' (ibid). Where Nattiez sees the production of a musical work as a linear process, we tend to regard it as an oscillating interaction between all of the different agents that are involved in the process, though, in this article, we limit the discussion to include only the performer and the composer.
As we suggested in section 2, the process of writing down a musical work is not a unidirectional poietic process but should rather be understood as an interaction between esthesic and poietic processes. This to an extent that makes it difficult to define the end of the poietic process as well as the beginning of the esthesic. The acts of musical composition that Nattiez gathers within the poietics can in themselves be analyzed by using the same method that he applies to the total fact of the musical work. According to us, Nattiez gives too little consideration to the generative processes (to repeat the quote: ``from the moment of the work's conception, passing through its `writing down', to its performance'' [Nattiez, 1990, pp. 92]), articulating the problem in ontological terms. It seems that Nattiez draws conclusions about ``processes unleashed by the musical work'' from a purely analytical understanding of music. This perspective is still dependent on the view of composers as `true creators' and works as `ideal objects': stable and fixed artworks that should make up the primary object of study for musicology.
What we are concerned with in these studies is almost the opposite: To understand the actions that lead to musical content and the significance of the interactions between the agents involved in these processes. A description of the generative phase of musical production preceding notation might provide a better understanding of the nature of the musical work evading the detour into abstract ontological reasoning. Hereby we also avoid the difficult and much debated issue of music as a signifying system.