Go to Semiotic Review of Books Home Page
Go to SRB Highlights
Go to SRB Archives
This review appeared in abridged form in Volume 10(3) of The Semiotic Review of Books.
Having made such a claim, two tasks are set out for this review. Firstly, I hope to provide a useful literature survey of major works published in English in musical semiotics in the 1990s. To this end, there are many more works entered in the bibliography than are mentioned in the body of the text. Secondly, I hope to support my contention that musical semiotics has achieved an important synthesis, and to this end I have arranged the review thematically, exploring in some depth the various topics and issues that I feel are of most importance, indicating the ways they have been drawn together in recent work. I do not go into great detail about the subtle strengths and weaknesses of each (although the most important are made clear), and I do not deal at all with technical questions of musical analysis. My intention is to convey a broad overview of significant trends, and to indicate how specific works have contributed to these.
As Lidov suggests, musical semiotics is important not just for musicology, but for general semiotics as well. In fact, historically, musical semiotics has had a considerably higher reputation among workers in the latter area than the former:
For many music theorists, Henry Orlov's ten-year-old caution that 'simply because music may be described in semiotic terms does not necessarily mean that the terminology and theory of semiotics will help us to understand music better' still resounds. On the other side of the disciplinary perch, semiotics... has long recognized the value of treating music as a semiotic system (Dougherty 1994: 164).Music is important to a general semiotics because it is a particularly hybrid sign system. Musical sound engages a wide variety of resources for signification, from indexes of bodily states, through to the most abstract of cultural symbols. Aesthetic valuation seems to rest fundamentally both upon rule-governed stylistic norms, and upon radical individuation both in works and in their reception. Musical practice is embedded in multiple contexts and frames of reference (histories, performances, acts of composition, reception practices, and others). And finally, of all significant forms, music has perhaps the most subtle and complex relationship to verbal language and the sorts of referentiality with which it is (traditionally) associated. It is this richness that makes music an important source-case for a general semiotics. But it is also this richness which has often divided musical semiotics into competing camps, each one incomplete. I should take a moment, then, to indicate the general outlines of the new synthesis which, I believe, can finally begin to address this complexity.
There have been three key developments in musical semiotics in the 1990s. The first was a general swing in the field away from formalism for its own sake, towards a renewed interest in hermeneutics (albeit of a highly formalized sort). This hermeneutic turn has allowed musical semiotics to say much more about the specific links between formal detail and cultural value. The second development was the emergence of a coherent and powerful theory of musical gesture, a much better picture of the connections between music's corporeal and symbolic aspects, and the subtle interplay between literal and increasingly abstract signs of embodiment in musical practice. This development has allowed musical semiotics to be more specific about the mechanisms underlying musical signification at all levels of generality. The third development has been the emergence of a semiotic theory of musical personae and musical narrative. The bulk of my review will be taken up with examining the various components of these three developments, and showing how they represent, when taken together, a powerful synthetic potential.
Before proceeding, it is worth noting that during the 1990s there has also been important work done in the revision and revisiting of older approaches to musical signification. For example, interesting things continue to emerge in the areas of intervallic lexicon, similar to the direction pioneered by Cooke (for example Stefani 1995: 199), and in psychoanalytic theories of subjectivity, following Lacan, Kristeva, and Barthes (for example Schwarz 1997: 275). There have also been several extended efforts in the 1990s to develop a mapping between Peircian sign typologies and musical signification (to name only a few, see Baest and Driel 1995; Cumming 1999; Hatten 1994; Martinez 1996a; and Monelle 1991). Indeed, it seems that almost every major musical semiotic study of the decade contains a passage on mapping Peircian trichotomies into music, although few make that their chief order of business. In my opinion, most of this work is limited by a tendency to take Peirce's typologies as a given, and to simply try and map them into a new context.
An excellent overview of work in musical semiotics up to and including the early 1990s (although one which unfortunately excludes works in cultural studies and popular music studies) has been prepared by Monelle (1992).
Just how much technical detail is needed in order for the hermeneuticist to escape the charge of shunning the elements of structure? One answer would be that that depends entirely on the goals of the analysis. And it is here that the theory-based analyst's data may seem to harbor an excess, an unusable surplus of dubious value (Agawu 1998: 13).Agawu goes on to make some useful observations about the very nature of analysis:
The suspicion that analysis facilitates close involvement with the music as an end in itself is not so easily dismissed. Unlike, say, archival study, analysis does not always proceed cumulatively. The author of the 50th analytical essay on the 'Eroica' symphony is not obliged to demonstrate full acquaintance with the previous 49 essays... It may be, then, that certain analyses are better evaluated as we would evaluate a performance, not as contributions to a discursive, constructed field of knowledge but as events which make an immediate or delayed impact, or none at all... [For example], understanding musical structure for Schenkerians comes from the practical activity of doing, and it is precisely this hands-on approach that guarantees its greatest pleasure (Agawu 1998: 15).It is important to remember that there has been a complex, long-standing, and often contentious relationship between formal analysis and hermeneutic exposition in music scholarship. Since the 1960s, in North America, there has been a powerful institution of "music theory," in which formal, mathematically-inspired analysis of musical sound, devoid of any consideration of cultural context or significance, was advanced as a complete and self-contained field of study. Contrasted with this has been historical musicology, in which analysis is occasionally used, but which is more concerned with questions familiar to traditional historians of culture. The fields of criticism and aesthetics, both mainstays of musical thought in other times and places, were somewhat marginalized in the second half of the twentieth century by the two dominant schools of thought. In trying to re-situate analysis as an interested social practice, then, Agawu is poking at some of the most cherished foundations of contemporary music theory. He is also trying to find a performative understanding of formal analytical activity which can aid in the reconciliation of formalism with hermeneutics. As we will see, this sort of maneuver has been attempted by other music semioticians as well, most notably Hatten, but not with the kind of revisionist flair demonstrated here by Agawu. In music scholarship, as in other fields, the degree of formalism one seeks, and the degree to which one bases analysis in positivistic assumptions, is as much a political as a methodological choice. Throughout this review, I will tend to favour a moderate position. Formalism is an intrinsic and necessary part of semiotics. But the hermeneutic direction taken by much recent work requires a flexible definition of just what sort of activity analysis is, and just what kind of truth claims can and should be made for it.
One of the most nuanced positions on the value of formalism can be found in the work of David Lidov, who is one of the few to systematically advance the study of articulatory mechanisms and formal structures in music semiotics in the 1990s. From the earliest days of musical semiology Molino, Nattiez, Ruwet and others were very concerned with articulation and segmentation. But this early work was limited by the dual problems of an overly-simple system (albeit an explicit and elegant one), and the doubtful pertinence of their parsing algorithms to actual listening or compositional practice. Lidov has not done much work on the second problem, but he has done quite a bit with the first, both in the area of general semiotics and in questions specific to music.
With respect to general semiotics, Lidov has pointed out that the formal analyses of early semiotics were limited by the adoption of the vague and overly-simple notion of difference (binary opposition) as the chief and perhaps only articulatory schema. In his own work, Lidov has replaced the vague notion of difference with a rich set of articulatory schemas, which can allow formal semiotic analysis of articulation to proceed without remaining mired in simple binarisms (see Lidov 1999, and the entry on 'Articulation' in Bouissac 1999). In general, Lidov makes a convincing case that systems are still an important part of human experience, and that semiotic analysis should still aim to be systematic, in the humanist sense of recognizing systems as a part of human culture if not necessarily the only part (Lidov 1999: Chap 8).
With respect to music, Lidov develops a distinction between pattern and grammar (Lidov 1999: 74-75, 154-155). The former refers to organizational features of individual works not strongly determined by rules of a style. The latter refers to systems of organization which pertain to a given style, and determine the less individuated features of works within that style. This distinction allows for a play between individuation and stylistic governance, without de-formalizing or mystifying the formal details of individuation. Along similar lines, Lidov suggests that, as a rule, in all semiosis there is a competitive relationship between structure and reference (Lidov 1999: 127). In all of these ways, Lidov continues to develop a musical semiotics which is systematic and concerned with formal articulation, and is in that sense continuing important trends present in musical semiotics from the outset. But he does so without maintaining the most grievous limitations of earlier work.
However, while Lidov is very particular and thorough on questions of general theory, his position on the value of particular analyses and interpretations is not so different from that of Agawu:
Regarding [my analyses], I do mean what I have said, but semiotic theory is not wedded to particular interpretations. The prize I covet is not to persuade someone to adopt my understanding of these works. What I would like more would be that another person who has better evidence or deeper insights should want to explicate a different understanding through a study of foundational elements and relations of semiosis... Semiotic analysis should celebrate individuality. I hope the tools deployed are safe and inviting to play with. They don't amount to a rigid method (Lidov 1999: 248-249).This is an important clarification, given Lidov's overall commitment to furthering the interests of formal and systematic thinking. And it is in this spirit which I also evaluate the degree of systematicity in the work of various music semioticians. Semiotic work should, in my opinion, be explicitly formal (and elegantly, even rigorously so). But it should not be positivistic or prescriptive. I am critical of work which demonstrates significant self-contradiction or leaves important terms and formal problems undefined. But I am also critical of work which allows the requirements of a formal system to overrun the more important task of exploration, or which fails to show how a particular formalism can lead us to a greater insight about the specificity of the works to which it is applied.
Cumming has written about the difficulty of expressing, or even describing to yourself, certain musical meanings. These meanings are both distinct and individuated, but difficult to put into words. Cumming describes them as ineffable:
To speak or write about [such an] experience is to ask 'Do you hear it this way too?' It is to name the object of perception, seeking its confirmation in someone else's listening experience. The description is no substitute for an experience of the work. It is more an act of ostension, a verbal pointing to qualities of harmony and texture that open up in time, than it is an attempt to capture precisely what is pointed to. Verbal adequacy is not sought here because the object of attention is something that can be shared and recaptured without further analysis... One of the pleasures of musical conversation is to establish intersubjectivity in relation to a work... For this purpose there needs to be a shared attention to the experience of the work. When description is ostensive, you have to hear the work, not merely grasp its structural procedures (Cumming 1996: 117-118).Cumming points out that the ineffable is an important part of music's power and appeal (Cumming 1996: 119). Her contribution is to highlight the fact that the ineffable is a semiotic phenomenon: it involves hearing the music as something (albeit an unnameable something), it is a conscious phenomenon, and furthermore it involves and inspires a large part of the discourse surrounding music. Since the phenomenon is semiotic, it is tempting to ask how it relates to the kind of formal structures and articulations with which semiotics is concerned:
A sense of ineffability may be the result of a synthesizing response whose object is undefinable because it fails to conform to any singly named feature of the music, its communication requiring a language of persuasion that embraces the provisionality of metaphor, rather than one of assertion that looks for certainty in music theory's categorical terms. Because such a response exceeds the domain of current theoretical categories, it will fail to be captured by a formalized model that depends on them. It is, then, simply false to imply that an account of cognitive processing supersedes critical reflections on artistic experience at a higher level. The cognitive problem of the inarticulate is not the 'same' problem as the aesthetic one... Mapping ineffability onto a low-level processing incapacity is to threaten the experiential ecology in which a verbal 'aphasia' is felt. This is not, however, to suggest that aesthetic and cognitive modes of discourse are irreconcilable with one another" (Cumming 1996: 119).It is also not to say that the ineffable is unrelated to formal perception, and to the models with which musical semioticians analyze it. Specifically, Cumming suggests that ineffability is perhaps a phenomenon of structural hierarchy, or if you prefer, of prolongational duration. Certain impressions are formed over longer time spans. These impressions are more general and vague than those afforded by briefer events. Long-term impressions are seemingly more unstable than shorter ones, and have a less binding connection to the passing moment-to-moment details (Cumming 1996: 123-125). Under this reading, ineffable meanings occur when a listener registers certain structural potentials and relations without naming them (or without being able to name them). This represents a kind of rapprochement between the aesthetic and analytic, since it reinstates the affective response as a legitimate component of structural perception (Cumming 1996: 126-128).
What I find most significant in Cumming's discussion of ineffability is that it shows how musical semiotics has moved considerably beyond basic questions like "can music mean anything," and "how can music refer to anything." The answers offered to these questions will be explored below. For now I simply want to note that the field has moved on to considerably more subtle sorts of meaningfulness, ones more specific to music and less driven by a linguistic model, and that these questions are being considered through a careful balancing of formal analysis with other factors. Similar work includes Pierce's analysis of the character of a musical style or work (Pierce 1995). It is also worth noting that this work can be seen as a continuation of the sorts of concerns expressed in cognitive and semiotic research of the 1980s into expressive nuance.
Before proceeding, I should note that Lidov has helped to highlight a general-semiotics underpinning for ineffable meaning by developing the idea of the processive sign. Some signs include extended temporal engagement as a sign factor:
A processive sign is a sign in which the representamen, the object, or the interpretant is a process" (Lidov 1999: 182). Works of art are one example. "With works of art we do not merely know that x represents y as z: x induces us to sustain a perspective in which we see, hear, or understand y as z. When engaged with art, we are cast into its perspective. The perspective sustained is an interpretive process" (Lidov 1999: 184)It should be clear how the processive sign relates to ineffability. It has a certain specificity (its perspective), but can only be experienced in-process. Trying to step outside of the processive sign and express its modality with non-processive (for example linguistic) resources can meet with only limited success.
Hatten has two principal concerns. On the one hand, as in other hermeneutical work, he wants to enumerate the expressive meanings associated with the works he analyzes, and their cultural significance. On the other hand, he wants to specify the mechanisms by which they achieve this expressive significance, looking at not only what the music means but how it acquires this meaning.
I am committed to a semiotic approach, which I construe as involving both structuralist and hermeneutic approaches to the relationship between sound and meaning. A structuralist approach, in this construal, is concerned with mapping associations (correlations) of structures and meanings in a manner that reveals their oppositional organization. Markedness is a semiotic valuation of oppositional features that... accounts for relative specification of meaning, the coherence of meanings in a style, and the emergence of meaning within an expanding style competency. A hermeneutic approach is concerned with interpretation beyond the more general oppositional meanings secured by correlations. Although guided by stylistic correlations, hermeneutic inquiry expands the theoretically stable bases of a structuralist modeling to encompass the subtlety, ambiguity, and allusive richness implied by any truly artistic competency... Hermeneutics is not essentially systematic or deductive, although its results may well be amenable to a later structural explanation, and indeed must be if those interpretations are to expand the systematic base of style understanding (Hatten 1994: 2).It is this specific mix of the hermeneutic and the structural which is taken to productive extremes in Hatten's work. It is also worth noting that Hatten's careful choice of analytical object has helped him to succeed. He is aware that different issues attach to the analysis of a style as opposed to individual works, although the two levels are in constant interplay. In manner similar to Lidov's distinction between pattern and grammar, Hatten states:
There are two competencies involved in musical understanding, the stylistic and the strategic. These correspond respectively to the general principles and constraints of a style, and the individual choices and exceptions occasioned by a work (Hatten 1994: 29).While Hatten's specific analyses often focus on the strategic level, he is very careful to situate stylistic competence as the main target of his analysis. By doing so, he guarantees that the relationships he uncovers will be of a wide-enough applicability to yield the kind of hermeneutic insight he desires. And by limiting his attention to the style of a single composer, he guarantees that the system he discovers will have enough coherence to be productively analyzed in a structural manner.
But it is in the specifics of his method that Hatten's work truly stands out. On the most general level, his claims and interests are not unusual:
Correlations and interpretations are conceived as mappings of expressive oppositions onto oppositions in musical structures. Correlations typically involve general cultural units or expressive states defined by basic semantic oppositions in a culture (happy vs. sad; tragic vs. nontragic). These cultural units are mapped onto general stylistic types, as oppositionally defined by traditional or other theories (major vs. minor). Interpretations, on the other hand, further specify or contextualize expressive states as they relate to entities ... actually manifested in musical works (i.e., tokens of their stylistic types) (Hatten 1994: 30).Hatten is exceptionally careful and effective in the way he chooses to analyze these relationships. First, he draws attention to the many problems encountered by encoding/decoding theories of correlation (Hatten 1994: 33). He then suggests that the theory of markedness might be the most useful structural linguistic idea for importation into music theory, especially since Shapiro has already extended the concept for use outside of natural verbal language (Hatten 1994: 34):
Marked oppositions organize the correlational mappings between structural and expressive oppositional pairs... The marked terms 'line up' with each other and encompass a smaller area... This 'congruence' between the markedness of signifier and signified is considered basic to markedness in language (Hatten 1994: 37).In order to clarify this model of how expressive meanings are tied to musical structures, Hatten also relies upon Peirce's notion of the diagrammatic, a structural iconism in which similarity rests in the structure of oppositions rather than properties. Such relations can motivate seemingly arbitrary choices (Hatten 1994: 167). Hatten goes on to develop many examples of markedness relations in musical systems, showing how they underlie basic semantic correlations (for example, the markedness of the minor key relative to the major allows it to have special associations, such as sadness or tragedy).
Hatten goes beyond correlations of simple musical elements, to make extensive use of Ratner's theory of topics (Ratner 1980), which has also been revisited and substantially expanded by Agawu (1991). A topic is a musical figure which has acquired specific connotations, sometimes even a denotation, within a musical culture. For example, certain intervallic patterns signify "hunting" in certain repertoires. Between the basic correlations of simple resources (major/minor, various dissonances and consonances, etc), and the more elaborated topics within a style, Hatten has established a rich network of basic musical meanings. The next step is to show how they can be elaborated into more complex works.
In this connection, Hatten develops a theory of expressive genre. An expressive genre can be understood as an elaborated topic: it is a higher-level correlation between a selection of characteristic topics and resources, and a more complex cultural unit. For example, the pastoral as an expressive genre involves the use of a certain subset of the topical universe of classic music, deployed so as to elaborate the cultural idea of the idyllic countryside. Expressive genres not only imply paradigmatic choices of material, but also a quasi-narrative in that they often require particular sorts of deployments through time and particular transformations (for example, another expressive genre posited by Hatten is tragic-to-transcendent).
Expressive genres serve to place interpretive activity in the proper realm. They are cued by basic oppositional features such as mode, high/middle/low style, texture, tempo, and thematic exploitation of familiar topics... Once genre is recognized or provisionally invoked, it guides the listener in the interpretation of particular features (such as the abnegation or resignation progression) that can help flesh out a dramatic or expressive scenario (Hatten 1994: 89).The one thing remaining for Hatten's theory is to explain style change, and to leave the door open for more subtle sorts of meaning, less thoroughly determined by stylistic norms (to edge towards what Cumming has called the ineffable).
The development of correlations provides a stable framework for interpretive growth as further articulation of the hierarchy of oppositions... But another kind of interpretive growth may alter the constitution of semantic fields in more radical ways, suggesting the need for a more flexible model to complement the hierarchical one implied thus far (Hatten 1994: 161).In order to get at these radical changes in semantic fields, Hatten explores the theory of metaphor. He points out that fixed correlations within a style may be regarded as frozen or conventionalized metaphors. A musical metaphor proper is more dynamic than this, and "requires a more creative or integrative act on the part of the listener, one that leads to an emergent meaning--and probably a more complex meaning" (Hatten 1994: 165).
Something akin to creative metaphor in language may be achieved in a musical work when two different correlations are brought together to produce a third meaning. I will refer to such figuration of musical meaning more generally as troping ... Although one may observe a strictly hierarchical and logical derivation process... one may also discover a more network-like, less predictable juxtaposition of disparate types... When there are contradictions between juxtaposed correlations or their divergent realms of meaning, then conditions are ripe for a truly metaphorical interpretation... In order for tropological interpretation to be warranted, there must be a musical event that contradicts stylistic expectation... If successfully reconciled... a new meaning has been troped from the contradiction of two older ones. That meaning might involve either a simple synthesis of two separate meanings ('my end is my beginning' or 'in my beginning is my end') or a third meaning that emerges from their interaction ('nothing is ever finished', or the play of incongruity that yields humor) (Hatten 1994: 166-169).This concludes my synopsis of Hatten's early work on musical hermeneutics. As we will see later, he has recently expanded it to take into account the role of embodiment and gesture. Before considering the broader implications of musical hermeneutics, though, I will give a brief synopsis of the work of Tarasti (1994), in order to convey a feeling for the close cousin of musical hermeneutics, musical narratology.
Like Hatten, Tarasti wishes to find a unified method for discussing correlations between musical sounds and cultural meanings, from the very specific and local through to the most general and sweeping. He attempts to do this by mapping, systematically and term-by-term, Greimas' trajectory of narrative generation into a musical context. Tarasti hopes that his theory will illuminate the dynamic, processual, and emotional (modal) nature of music.
Tarasti divides existing musical-semiotic studies into two categories; structuralist (characterized by a search for deep structures and minimal significant units) and iconic (characterized by a search for irreducible significant structures directly in the musical surface). Tarasti argues that neither approach can fully illuminate musical narrativity, and suggests a critical synthesis. This is similar to Hatten's rhetorical move of dividing approaches up into the hermeneutic and the structural, and then arguing for a reconciliation. From the structuralist camp Tarasti borrows the notion of isotopy as a foundation for his analysis of structural levels and formal features, and from the iconic camp he takes the energeticism of Ernst Kurth to be a key to understanding how actorality can be generated though these formal isotopies. For Tarasti, musical discourse is the musical surface. Tarasti sees this surface as the final step in a generative process analogous to Greimas' generative trajectory, and in order to prepare for a detailed examination of this trajectory, he presents two preliminary goals: (i) to enumerate the models and forces that guide the formation of the musical discourse; (ii) to introduce a formalism that can faithfully describe the dynamic nature of this process.
Tarasti begins to map out the specifics of his narratology by noting how musical processes of tension and release can be given interpretations using the language of modal logic. He shows how formal hierarchies (like those posited by generative approaches) can be interpreted as enacting 'will' (the dominance of one element over another) and 'must' (the perspective of the subordinated unit). He further specifies musical equivalents for the modalities of 'being' and 'doing' (initially in a way that makes them nearly identical to the classical notions of paradigm and syntagm).
All musical features presented or posited in Tarasti's analysis, from the concrete (e.g., a particular accompaniment figure) to the more abstract (e.g., a dissonance prolonged in the middleground), are presented as instances of isotopies. Isotopy, then, becomes Tarasti's general term for any kind of musical structure, and along with logical modalities yields a powerful vocabulary for unifying the discussion of various levels of musical meaning. Tarasti divides all musical isotopies into three categories--spatial, temporal, and actorial--organized in a generative trajectory, from achronic spatial background to kinetic actorial foreground. Tarasti's basic interest is in showing how the isotopies, modalities, and overall narrative trajectory are realized in absolute, nonprogrammatic music, through an entity he calls the 'thematic actant'. The modalities of the thematic actant can be correlated to musical features and events as follows: will = kinetic energy, the inclination of the music to move forwards; know = the sounding elements themselves, in so far as they introduce new information; be = the simple, consonant, non-actorial presence of a musical element; must = the state of an element which is modalized by another; can = an element's ability to modalize another; do = an actual musical event, a tension or desire to move.
This brief synopsis should give a feeling for Tarasti's approach. Despite the similarities with Hatten, there are important differences. Firstly, Tarasti's work is strongly marked by his decision to stick so closely to a highly elaborated the specialized theory (the narratology of Greimas). Hatten, by contrast, prefers to use simpler, more general, more easily detachable methods and develop them freely. Second, Tarasti is more specifically committed to the idea of musical actorality, making his work more explicitly narrative and also introducing important unresolved questions of exactly how musical actors are constituted. (As we will see, answering this question has become one of the most recent undertakings of musical semiotics as a whole). Finally, it should be noted that the theory of isotopy bears important resemblances to topic theory, but demonstrates important differences as well. Both posit consistent correlations between musical features and cultural meanings. The theory of isotopy is more powerful in that it allows features that are widely distributed through a work, or exist on a high structural level, to be drawn together into the same correlation. However, the theory suffers from vagueness, especially since Greimas himself used the term in many different ways. Topics are easier to grasp, and a certain amount of the flexibility shown by isotopy can also be achieved by topics through theories like that of expressive genre, which specify relationships between particular sets of topics.
It is also important to notice another feature shared by both musical narratology and musical hermeneutics: they both re-instate the musical surface as a crucial site for analysis. Since the ascendency of Schenkerian analysis in the 1960s the musical surface has seemed to become, for many theorists, something that required immediate and radical reduction in order to uncover deeper structures. While Hatten and Tarasti both recognize the importance of relating surface details to underlying processes, and relating the specificity of individual works to the broader features of a style, they also both insist that much of the experience of music, its vibrancy and life, occur right on the surface. Similarly, musical hermeneutics and musical narratology re-instate the use of metaphorical verbal language as a valid analytical tool:
A crucial although easily overlooked distinction between hermeneutic and theory-based analysis is that whereas the former relies on language, the latter often uses graphs, symbols, and a metalanguage to convey its findings. The enterprise of musical hermeneutics... is inconceivable outside a realm in which the polysemic nature of verbal language plays a central role (Agawu 1998: 11). Theory-based analysis becomes hermeneutic only at the point at which it dissolves its conceptual props into a more open and flexible narrative space (Agawu 1998: 19).Hatten and Tarasti are of course not the only two researchers to move in such a direction. Others can be found in the bibliography, and I would especially like to mention the work of Abbate (1991) and Cumming (1997b), which is important in the way it moves the discussion of narrative towards a consideration of musical personae. It should also be noted that Grabócz (1996) has also done extensive work in the application of Greimassian theory to music.
Despite the importance of their contribution, there are criticisms that can be made of Tarasti and Hatten. Regarding Tarasti, most of the specific problems stem from his fairly uncritical adoption of the work of Greimas, which is today seen by many as a deeply problematic theoretical heritage, full of arbitrary choices and unexplained mechanisms. In this connection, I can cite Lidov's general comments:
I mention Greismassian theory because of its influence and suggestiveness, but I have not been able to discover any grounds for believing it. It is tempting to see the seed and essence of all thought in the first acts of binary distinction... yet we must be wary. Perhaps binary oppositions are merely the easiest distinctions to handle... Peirce argued that the whole history of science was a progress from oppositions to continuities. Even if binary oppositions are foundational, every caution must be exercised in attributing such structure to higher-level constructs (Lidov 1999: 136-137).To be fair to Tarasti, he does attempt to get beyond the simple binarism of Greimas' narrative trajectory by placing emphasis on modal relations and the affective dimensions of actorality in the musical surface. However, and ironically, several readers (myself included) have been unable to escape the impression that, for all its systematicity, the analyses produced by this method seem in some sense deeply arbitrary. Pederson has criticized Tarasti for presenting the appearance of formal method but offering mostly subjective impression as the substance of his analysis (Pederson 1996). Now firstly, it should be noted that this is not necessarily a problem. Hermeneutic work is about insight, and if the formalism serves as a method for Tarasti to convey his insights, where is the harm? If the insights are compelling, and the theory an effective language for their exposition, then there is value in that. But it does not sit well with the scientific pretensions of Greimas' theory as a whole. Secondly, it should be noted that the same comment could be made about Hatten, but generally is not. In fact, most critics state the opposite, as does Pederson, who suggests that Hatten is more securely grounded in a cultural and stylistic context, and that his method for analyzing expressive genre is more securely systematic than that of Tarasti (Pederson 1996: 186-187). To an extent this is true: Hatten does use a less arbitrary, simpler theoretical model, and he does ground his specific decisions more firmly in historical detail. However, Hatten still does not really consider the dynamics of actual listening, or come to grips with multiple reading practices. Underlying his work is the assumption that there is a correct stylistic competency, which can be encapsulated in the responses of an idealized and fictitious listener. All of which serves as a convenient way for his own perspective to be presented as stylistic law. Perhaps I am overstating this point, but to me it seems crucially important if the broader social meaning of the analytic work itself is to be taken into account. I will return to this.
Work in this direction was substantially influenced by an article by David Lidov (1987), as well as by Coker (1972), Cone (1974) and Kivy (1980). Sheinberg (1995) has also pointed out that the work of Gombrich is applicable here. In all of this work, the central concern is to determine how music can present materials which we tend to hear anthropomorphically (modally and actorially). To explain this, it is now common to assert that there are structural iconicities between energetic states of musical sound and energetic states of the body (as well as indexical relationships, such as the audible effect of performance movement, or the physical effect of some sound parameters on the listener). Just as Hatten developed the theory of expressive genre from the simpler theory of topics and markedness, the theory of musical personae presumes that these energetic indices and icons, perceived anthropomorphically, are elaborated and assembled into full-fledged musical actors, and that as a result we often hear music as if we were conversing with another sentient being; the music seems to have agency. (An influential non-semiotic development of the same idea can be found in Cusick 1994). On the most basic level, this effect rests upon the fact that music is a manipulation of energetic states (sound configurations). These have properties which can be seen as bearing strong resemblance to the energetic state of other sorts of things (bodies, landscapes, and so-on). As the energetic state of the music changes, it appears to "move." Our perceptions in this regard are doubtless partly the result of inbuilt perceptual grouping mechanisms, but also doubtless partly determined through cultural learning.
Lidov refers to the movements represented by music as its motional object. For Lidov, a musical persona is an interpretant of the motional object (Lidov 1999: 219). In other words, for Lidov, the process begins and ends with the musical work defined as a bounded text, and the play of personae (which opens up a much wider field of reference and influence) is limited to being a tool through which the work is elaborated. Most music semioticians seem to follow Lidov in this, since they are mostly embedded in the tradition of Western art music, and are consequently invested in maintaining a central position for discrete, bounded musical works. A very different perspective would emerge if we were to designate the virtual musical persona as the object, and I will return to this possibility later.
Besides motional objects, music also has emotional objects, and the two are related in a specific way. Lidov and Cumming share a similar model of this relationship: "Music represents specific characteristics of movement that are indices of emotion" (Lidov 1999" 221). For Lidov, two sorts of indexicality are involved here: the general idea of gesture, and the specific essentic forms of Clynes (Lidov 1999: 222-223). Cumming develops this idea further. She suggests that simple indices of physical states become representamena for iconic signs of more elaborated physical states and emotions (Cumming 1999).
Cumming has also related her discussion of ineffability to the theory of embodiment. She points out that traditional theories of musical structure have had a difficult time accounting for feelings of ineffability, and suggests that many of these problems arise because structural models don't take account of embodied schemata, but engage in an "exclusion of sensory or proprioceptive response as a component of structural understanding" (Cumming 1996: 129). Cumming cites the embodied cognitive theory of Lakoff and Johnson as a corrective here (Johnson and Lakoff 1980; Lakoff and Johnson 1999). According to Lakoff and Johnson, all of cognition proceeds by metaphorical elaboration of basic schemata representing common bodily experiences. While Lakoff and Johnson do not present their theory in a semiotic light, Cumming shows how it is compatible with a particular brand of Peircian thought. "In Peircian terms, the structural relationship requires an embodied interpretant... When 'gesture' is found it is not, then, merely structure, but a structure interpreted with physiological reference and affective connotation" (Cumming 1996: 129). Important here is not only the claim that the interpretant needs to be corporeal, but also that affective response is a key part of structural relationship, and that the conventional bifurcation between these cannot be sustained (Cumming 1996: 132). But their relationship needs to be kept clear, to allow for the experience of the ineffable to remain unreduced. "Musical feelings do not translate automatically into structural descriptions, but contain an added factor of gestural interpretation. An understanding of musical gesture may be an embodied understanding, especially for the performer who does not attempt to articulate verbally what it is that he or she knows when realizing a structure through physical movement" (Cumming 1996: 133).
I should also note at this point that an independent but substantially-similar theory has been put forth by Shepherd and Wicke (1997). Following Zuckerkandl (1956), they view musical sound as a continually-shifting energetic texture, amenable to being heard as iconic with bodily and affective states. Their principal contribution, in my opinion, is the careful manner in which they analyze the actual perception of these iconic relations, and the importance of this perceptual modality for musical signification. Shepherd and Wicke situate the perception of musical gesture in what they call the sonic saddle, which is "the continually unfolding sound-image... experienced as the material ground and pathway for the investment of meaning" (Shepherd and Wicke 1997: 170). Although the connection is not explored by Shepherd and Wicke, it can be noted that the sonic saddle bears a strong resemblance to William James' specious present. For Shepherd and Wicke, musical gestures are perceived in the first instance as affective states. In other words, Shepherd and Wicke situate affective iconicity as a basic perceptual mechanism. This is not to say that their theory is essentialist, and throughout the book they are at pains to explain that the immediacy of gestural interpretation does not amount to a strong determination of musical meaning, for at least two reasons: the relevant habits of perception are in part learned rather than being completely innate, and the complexity of musical signification always leaves room for individualistic synthesis and negotiation of meaning. Given their phenomenological orientation, Shepherd and Wicke explore many issues of subjectivity relevant to the discussion of musical personae, and given their background in cultural studies, they also do much to situate the theory of gestural and affective iconicity in a broader culture-theoretic framework, which I will not explore here, but to which I would refer the reader enthusiastically.
Like ineffable musical meanings, musical affects and personae are not completely arbitrary, but also not strongly determined. As a result, questions of verifiability and truth tend to naturally arise whenever they are discussed. Cumming deals at length with problems of intersubjective agreement in the affective characterization of music:
If I say 'gentle pathos', someone else says 'abject grief' and a third gets really poetic with 'ineluctable desolation' where is the agreement about content? Can there be a difference in the language used to describe expression, and yet an accord about what is expressed? Acknowledging these variations, the level of verbal agreement does seem to rest with a generic term, the category of pathos or 'sadness'. At a finer level, it is simpler to point out moments of change, intensification, or reversal [of expressive nuance] than it is to find a shared emotive terminology for changing degrees of intensity in these moments. A possible variation in the choice of descriptive terms by different people does not, however, render these nuances an ineffable mystery... It merely establishes a variation in individual response to the intensity of a shifting musical affect, whose formal characteristics can be agreed about in an uncomplicated manner (Cumming 1997a: 335).Cumming is not the only worker to be concerned with questions of this sort. In general, musical semioticians remain primarily interested in describing trans-personal, intersubjectively-shared sorts of musical meaning, maintaining the (usually tacit) assumption that the stylistic competencies they posit need to be verifiable and consistently applied by social actors in order to be valid. I will not contest this point, since it is part of a coherent and tenable position. But I will suggest in passing that it is not, for me, nearly as large a concern. Beginning as I do from a dialogic theory of subjectivity, and concerned as I am with the multiplicity of reading practices to be found within and between interpretive communities (which are never homogeneous), I am not interested in consistency for its own sake, so much as in finding a consistent framework within which to understand the mechanics of variation. The distinction I am making here may ultimately prove more rhetorical than theoretical, insofar as disagreement is only possible within a common frame of reference (as pointed out by Frith 1996, Fish 1980, and many others), and insofar as a single competency can include room for polysemy and negotiation of meaning. But it is still worth noting that semioticians of affect are still on occasion troubled by lingering echoes of positivism, and that their concern is not entirely necessary.
Cumming also broaches another subtlety of affect theory, one which to me is more interesting. She revives the question, posed earlier by Langer among others, of why we would listen to music which portrayed undesirable affective states if it actually made us experience those feelings (Cumming 1997a: 336). The related question is, do we actually feel the expressive content of music, or do we detect it in a less emotionally-engaged way? Cumming suggests something intermediate, that musical involvement "has its own way of contextualizing affect" (Cumming 1997a: 337). For example, in the case of works that evoke grief:
I do not deny that in playing or listening to these works I 'feel' their pathos in some sense. I would not choose to play them at a time when I am already experiencing grief, if there were no connection between their expressive content and the state I am in. If, in this state, I 'mirror' what the music provides, I do not, however, become more saddened, but find a greater objectification of the grief. The musical appearance of pathetic 'sighs' within a context of relative stability, and without a referential object, allows for a recognition of intense passion, and yet the gaining of distance and perspective. Music allows a framing of emotion as distant from immediate life implications, and a concomitant gain in 'perspective' on what is expressed, even while the 'mirroring' effect of identification with the emotion is going on" (Cumming 1997a: 337-338).This suggestion bears strong relationships to Lacanian psychoanalysis and to theories in the semiotics of acting, both of which Cumming explores (Cumming 1997a: 339-340). Actors maintain a double-awareness of the emotion they are portraying, on one level actually experiencing it, but on another staying at a distance, having a semiotic consciousness of the fact that they are mirroring it in order to portray it. "A similar negotiation could be postulated to exist in response to musical gestures, where both natural and conventional elements are involved. If it is possible to both mirror a gesture, experiencing its affective content, and to be aware of its conventional standing within a style, its 'sadness' (or even tragic pathos) may be framed and made tolerable" (Cumming 1997a: 341).
The most recent work on the semiotics of musical gesture and affect of which I am aware is by Hatten (1999). Feeling that certain details of expressive nuance were escaping the grasp of his earlier analysis of Beethoven, Hatten develops a theory of gesture substantially similar to that which we have seen in the work of Lidov and Cumming, and incorporates it into his existing hermeneutic model. For Hatten, gesture is a concept that can help explain the "temporal and textural gestalt of shaping and shading" in musical expression. His definition of gesture is basically the same as that of Lidov: a gesture is a molar bodily time-form which is marked for significance (Hatten 1999). Hatten notes that, physiologically speaking, there are many degrees of freedom in gestures, and in addition, different gestures can often have similar effects. For these reasons, we can't expect an exhaustive and specific description of which gestures have which effects musically (Hatten 1999). Interestingly, in looking for source material on the study of gesture, Hatten takes a tack opposite to that of Lidov. Lidov wants to consider movement as a general category, which he carefully distinguishes from formalized contexts such as dance, where the formal expectations tend to limit vocabulary, emphasize style, and cater to contexts of visual reception (Lidov 1999: 216). By contrast, Hatten feels that we should "draw on what we can learn of culturally conditioned human gestural languages, especially those that are preserved due to their artistic or rhetorical development" (Hatten 1999). Also, Hatten is deeply occupied with the question of how to use gesture to integrate scores (to create an appropriate gestural layer in the performance of a scored work), and with the historical reconstruction of gestural languages. As a result, much of his work is specific to a particular pianistic performance tradition, whereas the work of Lidov and Cumming has a slightly broader applicability.
Hatten articulates an important critique of Lidov's earlier work on musical embodiment (Lidov 1987). In this earlier work, Lidov suggested that while musical signs often have a corporeal origin, the more elaborated musical signs (and by implication the more interesting and aesthetically valuable ones) are increasingly abstracted from their bodily origins. This is not so different from what we have heard from Cumming (that indices of simply bodily states become representamena for further icons of more elaborated states), but Lidov adds a value judgement which many workers have found problematic: really high musical achievement amounts to a kind of flight from the bodily origins of musical signs. It is to this implication that Hatten responds when he writes:
Instead of sublimation, perhaps we should be seeking a kind of emergence whereby the gesture maintains its characteristic potency while gaining a factor of generalization or type-formation (rather than abstraction)... If gestural expressiveness is an essential motivator for compositional form and structure... then we must find a way to incorporate gesture in all of its particularity, in all of its continuity, in all of its analog character, and in all of its temporal shaping and shading, as part of the very foundation of structural analysis (Hatten 1999).I would also note here (as have Hatten and Cumming in other places) that the theory of embodied cognition presented by Lakoff and Johnson forms a very good fit with this view of the body as present at all levels of musical signification. (In a recent article, I have explored the use of Mark Johnson's image schemata in musical analysis, see Echard 1999). Hatten goes on to present a list of eight presuppositions for a semiotic theory of gesture. These mostly summarize points made elsewhere in the literature, but his list is admirably complete and concisely stated. The highlights include:
"Gesture is movement interpretable as a sign, whether intentional or not, and as such it communicates information about the gesturer." Hatten suggests that the information thus conveyed can be categorized using Peircian trichotomies. "Another way of defining gesture is as movement that is marked as meaningful."Hatten goes on to consider the possibility of gestural troping, in order to allow gesture to be active at all levels of his hermeneutic system.
"Intermodality is one of the most fascinating aspects of gesture, in that gestural dynamics and shaping can be expressed through many different senses, all of which share characteristics of continuity through time."
Gesture also has important evolutionary aspects (the sensitivity to the gestures of other animals is an important adaptive trait). There are many biological, neurological, and other scientific connections to be explored here.
"Regardless of modality, gestures may share certain characteristics, being: analog as opposed to digital or discrete, continuous in a productive sense of continuity, having articulate shape, possessing hierarchical potential, characterized by a significant envelop (pre- and post-movement can substantially affect the quality of the gesture), being contextually constrained and enriched, typically foregrounded, being beyond precise notation or exact reproducibility, but amenable to type-token relationships, and thus potentially systematic to the extend of being organized oppositionally by type, as in gestural languages or ritual movements
Posture maybe be considered as gesture 'under a fermata'. A 'frozen motion' or pose may reveal the energy and affect with which it is invested, including that required to move into the pose. The posture thus 'reverberates' with the resonance of the implied gesture of an agent" (Hatten 1999).
This concludes my overview of what I take to be the most impressive achievement of musical semiotics in the 1990s: the development of hermeneutic and narrative theories of music which can begin to reintegrate formal analysis with critical reflection, coupled with an embodied theory of exactly how musical personae and affects are constructed and experienced. But despite the admirable scope of this achievement, it is just a start. As I noted early in my review, the branch of musical semiotics which has developed this work is strongly connected to traditions of Western art music aesthetics which have fallen considerably behind in terms of cultural theory and the next step is, in my opinion, the deployment of these models of musical signification towards a consideration of musical practice framed as an interested, intersubjective human activity deeply connected to questions of power, gender, class, and other issues engaged on a regular basis by cultural theory. In order to edge back in this direction, in an admittedly roundabout way, I would like to consider an article by Monelle (1996), which I think graphically shows how, far from confusing the issue, a more practice-oriented theory could at this point clarify some of the difficulties still left unresolved in musical semiotics.
Returning to a favourite topic of early poststructuralism, Monelle asks how we might want to define a musical text. He points out the many difficulties that have emerged in literary studies in connection with this question, and suggests (not entirely seriously, as it turns out), the following:
In a medium like ours, that is supposed to be self-referential or to have no meaning at all, the discernment of a text will seem less dependent on intention or interpretation. In music, the meaning is simply whatever the music means; the question of 'true' meaning is no more than empty talk. It ought to be easier to define the text in music... The score is, perhaps, the text. This would seem to match the traditional musicological view, in which scholars try to establish and authoritative text for music of the past. By this, they mean the musical aspects that can be written, and indeed were written by the composer. Of course, the realization of this score requires much cultural knowledge; but this study, called 'performance practice', is not considered a textual study (Monelle 1996: 245-246).Monelle is not happy with this option, nor with several other simple ideas about what a text might be. He is most interested by an idea presented by Rodolphe Gasché, who he paraphrases as follows:
A text is not the pattern of signifiers or signifieds, or even the patterned relation between them; on the contrary, it is the annihilation of opposition, the stage of resolution or fruition of the opposition of sign and meaning which constitutes the action of signification... The text is not form plus content, but the overcoming of form and content... [The musical text] is the score, not as performed, but as understood, its dialectics resolved into intelligibility (Monelle 1996: 247-248).Many questions are raised by this definition, only one of which is deeply engaged by Monelle:
If the text is that boundary where dialectics is resolved into a monism, where self and otherness confront each other, what is to be found on that boundary? Does it exist? Can it be known or interpreted? Like so many deconstructive ideas, this merging point of the text seems to be just nothing at all (Monelle 1996: 248).In other words, if the text is understanding but understanding can be infinitely deconstructed, then the text vanishes. However, at least in the case of music, this is where the strength of the art form lies. "The absence of musical signification, its dissolving under the studious eye, is not a limitation of music, but on the contrary its very life and beauty" (Monelle 1996: 249). In an apparent weakness, then, music reveals that it is actually more in tune with the basic nature of textuality than are many other, seemingly more transparently textual media:
The text, whether literary or musical, is profoundly abstract. It is not the score, not a performance, not an intention. It is also--and this is vitally important--not the work. The musical work is something somebody has made; it is a poiesis... But the text does not merely occupy a space defined by the composer's work. Its space is chiefly defined by certain other factors: in particular, by the universe of texts, which is to say by intertextuality" (Monelle 1996: 249).According to this view, texts are nodes in intersecting networks of intertextual reference:
The musical text, then, is a boundary between inside and outside, rendered problematic by the flow across the boundary and the interdependence of inside and outside. It is also an epistemic nexus, the meeting point of all its significations, indexical, iconic, and symbolic. It is not a transcendent essence, an abstract pattern, an object, an 'experience' (Monelle 1996: 255-256).This formulation is entirely brilliant, but I cannot help feeling that something has gone very wrong. Granted, Monelle is working in this instance within a poststructuralist tradition, and so the disappearance and différance of the text is to be expected. But I still feel that the result is too complicated and not illuminating. And I would suggest that this happens because Monelle, like most other semioticians who work within the Western art music tradition, chooses not to adopt a more culture-theoretic view of the subject. The one possibility that Monelle does not consider is that the text is a practice (in Bourdieu's sense of the term). Textuality is a construct maintained through socially-determined reading practices. Texts are the bounded entities upon which that cultural work is performed, and a particular text is usually not difficult to define if you start with a particular instance of use. True, the text outside of the context of use is nothing at all, but that does not mean that it is in the basic nature of a text to be nothing. It means that the text is an constituted as an object in the course of actual, situated, interested acts of reading, and needs to be analyzed with respect to these. There is nothing forced or unnatural about Monelle's formulation--it is where you are naturally led if you try to think about textuality without framing it in terms of social practice. Which is why I would suggest that certain semiotic questions can only be clearly addressed when they are so framed.
Music analysis can scarcely proceed without postulating a listener, yet the difficulty of specifying the relevant features of a listening subject has led writers to invoke a variety of constructs, some of them hypothetical, many of them designed to evade the challenge of providing an ethnographically secure characterization. Thus we have the naive listener, the competent listener, even the ideal listener (Agawu 1998: 10).It must be conceded that the avoidance of real social practice is sometimes grounded in theoretical rather than practical concerns. Many semioticians feel that semiotic systems have their own particular mode of existence and efficacity, and that it is not necessary to provide them with ethnographic grounding. For a defense of this position, see Lidov's entry on 'semiosis' in Bouissac (1999). It should also be noted that the simplification of the listener has allowed much to be accomplished. By not dwelling on the difficulties of real ethnographic grounding too soon, musical semiotics in the 1990s was able to focus upon other questions, and to make substantial advances. But I would suggest that its further questions need to be addressed through a theory of the subject, and of social practice, which engages in depth with the messy details of real lives. To this end, it is now time to move on to a brief examination of a very different musical semiotic tradition, that of cultural studies and popular music studies.
To oversimplify grossly for a moment it might be suggested that, while classical music semiotics has accepted a naive view of subjectivity in order to make significant advances regarding the understanding of musical structure, cultural studies and popular music studies have accepted a fairly shallow view of musical structure, and sometimes an ad-hoc approach to basic theory, in order to make advances in the understanding of how music, as a social activity, fits into broader patterns of cultural practice. Like most work stemming from British cultural studies, popular music semiotics has tended to use a Saussurian rather than Peircian model, and to deploy it in a manner similar to that of Barthes' Mythologies, identifying particular connotations and denotations of musical sounds, and then employing them as part of an analysis of broader ideological systems. An early work in this vein was Hebdige's analysis of punk rock (Hebdige 1979). The approach has been nuanced somewhat in recent works, for example that of Walser on heavy metal (Walser 1993), but there is still a marked tendency to think in terms of denotative codes, linked up in fairly simple ways with musical configurations.
This is not to say that early work in popular music semiotics did not occasionally reach a high level of sophistication. Just as in early classical music semiology, which was also Saussurian in orientation, semioticians of popular music explored a wide range of theoretical questions in considerable depth. A useful summary of this work can be found in Middleton (1990). And in recent years, the field has expanded to include theories of gesture quite similar to that put forward in classical music semiology. Shepherd and Wicke have already been mentioned in this connection, and another important contribution has come from Middleton (1993). In general, though, popular music semiotics has not come anywhere near the degree of subtlety and sophistication in addressing details of musical sound as have other branches of musical semiotics. But, as already noted, this is in some ways a deliberate trade off, and I will now consider some of the unique achievements and contributions allowed by this reduced attention to musical detail.
Consider the sign typology proposed by Tagg (1991). His focus is pragmatic, by which he means social use and pertinence must always be addressed in the typology. Tagg is not mapping an existing system (for example that of Peirce or Greimas) into music, but asking directly what social effects and uses music has been associated with, and basing his typology on these. Empirical methods, involving the response of non-specialist listeners to particular musical textures, are part of the process of constructing the typology. Tagg arrives at the following typology for musical signs:
|Anaphonessonic||sonic: structural homology with paramusical sound|
kinetic: structural homology with paramusical movement
tactile: structural homology with paramusical sense of touch
|Genre synecdoche||pars pro toto reference from inside one musical style to element(s) of another, hence to the genre of which the second style is part, hence to the complete cultural context of that style|
|Episodic marker||Short, one-way process highlighting the order or relative importance of musical events, e.g., anabasis, katabasis, cresc., accel., rall.|
|Style indicator||Unvaried aspects of musical structuration for the style in question.|
Admittedly, compared to a tightly-theorized investigation like Hatten's, this seems rather ad-hoc. It doesn't cohere as a system, and while Tagg claims that it encapsulates the results of extensive empirical work with actual listeners, it seems that the subtleties of real-life listening experience are not well represented here. That said, though, it must be noted that Tagg isn't seeking to make a self-consistent system, but to categorize a few of the particularly common ways he has found music functioning in cultural contexts. Similar typologies have been offered by Middleton (1990) and Stefani (summarized in Middleton). Such a typology is not meant to explain exactly how music constructs these effects, but to help in naming and tracing the various chains of signification that wind through particular cultural practices.
Tagg's work is still, essentially, a continuation of approaches associated mostly with the 1970s and 1980s. Other work in popular music semiotics has proceeded further into new directions. I would like to consider two examples here: Middleton's application of dialogic theory to the idea of authorship, and Cook's observations about the musical logic of television commercials. Middleton begins by noting that...
pop songs are often interpreted, by fans, critics and even academic analysts, in relation to traditional notions of 'authorship'. But in recent pop, such as the Eurythmics' hits, these notions are at the very least in tension with a more fragmented construction of subjectivity" (Middleton 1995: 465).
He draws upon the dialogic theory of Bakhtin in order to...
present the various parts of songs (i.e., both textural lines and structural sections) as interactive 'voices', each with its characteristic style-features. Such features are always culturally marked, through their multiple associations and their different positionings within various discursive domains. It is possible, therefore, to locate the styles, their features and their interrelations on a range of discursive axes (gender, ethnicity, etc.), making up a 'map' of the musico-discursive terrain, then to place the 'dialogue' constructed in a specific song in relation to these axes, this map" (Middleton 1995: 465).In many ways, this is not substantially different from Hatten's hermeneutic, and is less developed. But I would suggest that Middleton is making an important contribution by linking the theory of multiple musical voices to a dialogic theory of culture, and by pulling away from the naive view of a unified subject. It is also worth noting that the axes of meaning which he chooses to employ have considerably greater political import than those used by Hatten, Tarasti, and other workers in classical music semiotics.
Middleton goes on to consider at some length myths of authorship and identity in classical and popular traditions (and in this respect his article makes an interesting pair with that of Monelle, discussed earlier). He examines journalistic responses to the Eurythmics, especially the tendency to read autobiographical detail into the sexual relationships discussed in many of the lyrics, and also considers ways in which the band members themselves perpetuate a sense of individual agency and personal ownership with respect to their songs. Middleton does not deny the reality or value of these engrained assumptions, and is careful to develop a view of agency which is dialogic, but nonetheless allows room for the impression of subjective solidity and control. The core of this model is the claim that "since signs are social products--that is, discourse is always inter-textual, always and unavoidably referring to previous discourse--subjectivity is created in dialogue with other subjects, other discourses" (Middleton 1995: 469). The self is not a stable presence which acts as an ultimate arbiter of the real, but is an ongoing project, an energetic capacity that only realizes itself though dialogic exchange. "This approach to creativity enables us to retain a concept of agency, in a particular, closely qualified sense, while refusing the mystifications of traditional notions of authorship" (Middleton 1995: 469). This is important because it allows for subjective identities and effects to be noted, for example the feminist issues activated by Annie Lennox's equal creative partnership in the Eurythmics.
The Eurythmics are self-consciously eclectic in their stylistic references, which makes them a good subject for dialogic analysis. Part of their compositional skill is in interpreting and interweaving these styles, maintaining their associations but recontextualizing them and presenting complex subjectivites in the process (Middleton 1995: 469-470). Much of this could be understood in terms of musical narratology, and Middleton does note the connection, although he limits his comments to topic theory:
Mainstream music semiology might classify the various style-features [in the music of the Eurythmics] as topics. But this somehow fails to catch the way that, in popular music anyway, styles are intersected by multiple discourses, so that they stand at the conjunction of a variety of axes of social meaning. Nor does it deal adequately with the fact that... manifestations of the styles overlap and coincide, so that they have the character of interacting 'points of view'" (Middleton 1995: 470).To a degree, the first of these concerns is addressed by Hatten's theory of troping, and the second by the theory of musical personae, although in both cases the theories would tend to throw the phenomenon into a depoliticized light, which is one of Middleton's objections. One of the chief values in Middleton's deployment of dialogic theory is that it forcefully moves us towards the situated, interested, politicized dimensions of these musical signification processes. Middleton goes on to a detailed analysis of some Eurythmics songs, to show how the subjects they construct are polyphonic and heterogeneous, and how they deploy multiple sign systems (fashion, musical sound, and others) towards this end. He also provides an interesting map of how particular popular music genres and styles (ballad, pop, rock, blues, soul) enact particular gender constructions (Middleton 1995: 474).
To pursue another example of recent popular music semiotics, Cook is concerned with basic question of what music can mean, and how it does so. Making a distinction familiar from Kivy, Cook suggests that "it is helpful to contrast the concept of [communicated] meaning with that of effect" (Cook 1994: 27). Music clearly has effects. And it seems that these effects can be the basis for meaning and communication. But this second step has generated a tangle of schools and orientations, and little agreement (Cook 1994: 27). This is all familiar territory. What is new in Cook's approach is, first, the degree of importance he puts upon context in sorting out how effects can become meaningful. Even more importantly, Cook explores the idea that some contexts allow musics which are stylistically chaotic, or even incoherent, to function perfectly well, although according to a logic specific to the context and not a strictly musical logic. The example offered by Cook is music for television commercials:
Music for commercials need not [make sense in musical terms]. Music that is custom-written for a commercial frequently makes little or no sense when heard by itself, away from the context of words and pictures... Its logic is not the logic of concert music" (Cook 1994: 31).Cook also touches on something mentioned by Middleton in his article on the Eurythmics:
Traditionally, musicians compose with notes, rhythms, and perhaps timbres. Only with postmodernism has the idea of 'composing with styles' or 'composing with genres' emerged, at least as a consciously adopted procedure" (Cook 1994: 35).Now this is probably an overstatement. There are many examples, going back as far as we may care to look, of composers deliberately juxtaposing and modifying style components to achieve a particular effect. But the notion that stylistic juxtaposition can constitute its own compositional logic, displacing previously more fundamental logics, is new. Cook's ultimate point, though, is not that this new music is of a radically different kind, but rather that it highlights something true of all music: that meaning is only defined in relation to contexts, which will often involve a negotiation with logics besides that of internal musical rules. (It should also be noted that these arguments are considerably expanded in a new book by Cook 1998).
My contention is that the models of classical music semiotics and those of popular music semiotics have reached a stage of development where they can be productively synthesized. As an example of one possible exploration in this area, consider again Lidov's comment that musical personae are interpretants of the motional object of music. What this suggests is that the object being developed by the musical semiosis is the musical work itself. What happens if we decide to argue that, in some cases, musical personae are the objects, raw sound forms representamena, and the impressions of motion are interpretants? In this case, we would have a powerful model for considering the use of music as a tool of identity formation. It could help us to clarify many issues in the semiotics of popular music, including: the strong personal attachment people show for particular musical styles, the powerful feeling that one has a personal relationship with musicians one has never met, the ease with which certain musical features become emblems of attitudes and philosophies, and so-on. In short, this slight re-articulation of a theory from classical music semiotics could help substantially in illuminating one of the most interesting relationships in popular music, the one between real and virtual personalities in the media and in daily life, with special reference to the specifics of musical sound. In exchange, this exploration might also teach us something new about classic music, and the different (although not entirely alien) balance between sound, gesture, and persona in various aspects of that tradition.
I have already mentioned Martinez in connection with attempts to map Peircian categories into a musical context. It should be noted that he has also explored their use in ethnography. For example, in that case of a Papua New Guinea culture where a particular piece of music is part of a tobacco-smoking ritual, Martinez offers the following analysis:
If the semiosis concerns one person of that culture who is familiar with the conventional meaning of the sign as an 'invitation to smoke tobacco', the sign will refer to its dynamic objet (the invitation, i.e., another sign) not by means of its pure acoustic materiality nor by the fact the sign is an index of the Non-Papua. Rather, in that culture there is a relation, dependent on habit, among the three relata (sign-object-interpretant). The sign refers to its object by a cultural disposition that enables the interpretant... The sign is in itself a legisign; to its object it is a symbol... the interpretant could be of an emotional kind (a euphoric feeling); energetic (the attitude of going in the direction off the village that uttered the symbol); or logical, as a partial syllogism such as 'if there is an invitation, then there will be a ritual' (Martinez 1996a: 64).This seems to demonstrate that, in some cases, Peircian categories can be useful in sorting out ethnographic data. Martinez has also applied them to understand the rasa system of Hindustani art music. His basic contention is that:
There are three terms that could be considered the main links in the sign chain that bears the aesthetic process in Hindustani art music: raga, bhava, and rasa. As a whole, these terms are engaged as the three relata of the semiosis: raga is the sign, the entity that represents an object, the bhava, and determines and interpretant, the rasa. In actual performance, a raga is sung or played. If performed correctly, it refers to the ragabhava and is interpreted, by the attuned audience, as a rasa (Martinez 1996b: 111).Martinez describes the relation here as essentially symbolic (in Peirce's sense of symbol):
Here, the raga will be interpreted as a sign that represents its object as a representamen per se, despite any relation of likeness or contiguity that it may have to its object. Therefore, the Peircian idea of symbol can be applied to understand the semiosis in the aesthetic enjoyment of classical Indian music" (Martinez 1996b: 111-112).Martinez does not deny, however, that affective engagement is a key part of that aesthetic experience, and he considers it at length, although we have seen that he takes the essential connection between raga and bhava to be symbolic rather than affective. While the experience of listening involves a visceral savoring of rasa...
there is a tendency towards a certain direction of meaning, that which is determined by the interpretative rule grounding the symbol. An actual raga-bhava, rasa entity only exists in order to be interpreted as such, which presupposes listeners able to comprehend it. Paradoxically, the relishing of an emotion in Indian music is a rational process" (Martinez 1996b: 112).There is much to be admired in this article, but a few reservations must be stated. Firstly, although the author refers frequently and authoritatively to Hindustani music theory, there is little or no attention to ethnographic details of actual listening and performing practice. Sweeping generalizations are made about the priorities of those involved in this musical culture, but we are given no way of knowing how the author knows that these are the actual, active priorities of anyone in particular. Second, and more immediately pressing, it is unclear just what the use of semiotic theory adds to our understanding in this case. While Martinez is often effective in his mapping of a Peircian perspective onto Hindustani music theory, he doesn't in the first instance give a compelling reason for doing so. From an ethnomusicological perspective, interested in understanding more about Hindustani music, there are no insights here that require a Peircian analysis, or that couldn't be more clearly expressed without it. Neither does the article really tell us anything new about Peirce or semiotic method. None of which is to say that it is a bad article, because it many ways it is a good one, and the exercise it describes doubtless has exploratory and pedagogical value. But it also stands as an example of a tendency demonstrated not just by Martinez but by many writers interested in semiotics, and one which should be resisted: the use of a semiotic model simply because such use is possible, and without careful consideration of whether it is really necessary. Before moving on, I should note that Martinez has also presented a more essentially Peircian argument in a book-length treatment of Hindustani music, where he argues that Indian musical aesthetics involves a movement in perception from thirdness (various conventional listening habits and related symbolic constructions) towards firstness (immersion in a quality of feeling)(Martinez 1997: 368). Given the scope and subtlety of Martinez's work, I do not want the reservations I have expressed about a single article to be taken as a blanket critique.
Another recent application of Peirce to ethnomusicology is presented by Turino (1999). This article is written at a more general theoretical level, and considers the usefulness of Peircian theory in explaining musical affect in an ethnographic context. The article is strong, but it does not add anything to the previously-existing theories of musical narratology and semiotic hermeneutics. Also, strangely, Turino cites none of the work discussed in this review, nor any preceding work in Peircian musical semiotics.
In summary, while ethnomusicology in a general sense remains a rich source for semiotic reflection, especially the research done in the 1980s with respect to iconicity, the field has not been centrally concerned with semiotic issues in the 1990s.
Without underestimating the valuable and necessary empirical work that has been (and still will be) undertaken, one of the most original and significant contributions sociology could make [to the study of music] is of a theoretical nature. In my opinion, it should consist in bringing forth a definition of music as a fully-fledged social phenomenon, one that would not only grasp its constitutive social dimensions, its various historical forms and foundations, but also recognize and account for its very specificity. Some steps have recently been taken in this direction. However, to my knowledge, there is as yet no fully developed conceptual framework which lives up to these two requirements... [This] means that to a large extent music is still granted only a limited or partial social character, and allowed only a mere aesthetic or artistic specificity" (Grenier 1990: 27).I would like to conclude by asking two questions. First, to what degree should Grenier's criteria be adopted for work in musical semiotics? And secondly, how far has the state of the art advanced in the 1990s towards meeting those criteria?
Accounting for the specificity of music as a signifying system is a goal which almost every music semiotician has embraced. However, elaborating its specificity as a social practice, its constitutive social dimensions, has not been widely adopted within semiotics. Even in cultural studies and popular music studies, where such a concern might be expected, semioticians have often stuck to simpler models of musical reference, in which music is seen (at least by implication) as a poor cousin to verbal language. This is why, working within the field of cultural studies in the late 1990s, Shepherd and Wicke still felt it necessary to argue that explaining the specificity and relative autonomy of music as a signifying system is a central uncompleted task for musical semiotics (Shepherd and Wicke 1997). In cultural studies and popular music studies, the fully social nature of music has been thematized, but its specificity has not. In classical music semiotics, the reverse has been the case: a theory of the specificity of music has been offered, but it is not a social theory.
The lack of immediate progress does not mean that semioticians are hostile to the idea of studying music as a social phenomenon, nor that they are unaware that eventually this will require their models to expand towards a more general practice theory (and towards each other as well). In cultural studies this has always been a stated goal, although in effect the theoretical models chosen have often militated against its achievement. In classical music semiotics, although the primary focus remains textual and individualistic, the attention paid to cultural meanings and the interest in expressive genres and topics which cut across various art forms indicates a willingness and even desire to say more about exactly how music is constituted, in practice, as a relatively-autonomous but fully social phenomenon. And I would suggest that this is the direction in which music semiotics should be moving, although I would concede that one cannot move too quickly. Bracketing of some of these questions has been necessary, and may continue to be necessary for some time, in order to make sure that our foundations are secure.
But we are nearly ready. The theories of musical narratology and musical hermeneutics, along with the theory of gesture and embodied affect, provide a powerful lense for viewing music in both its specificity and in its connections to other cultural forms. Within classical music semiotics, then, many of the music-specific tools are in place. And within cultural studies and popular music studies, much has been learned about how to trace the actual uses of music in ongoing cultural practices, to analyze the pragmatic cultural effects of its signifying modalities in the daily lives of specific people and communities. There is, as yet, no monumental work which has succeeded in putting these traditions together. And before such a thing can happen, we will need to work harder at bridging the considerable aesthetic and disciplinary gulf that separates the various traditions of semiotic work. Precedents exist, especially the work of the Prague school, which has been inspirational to both classical music semioticians and workers in cultural theory and popular music studies. Also, the latest (unpublished) work of Naomi Cumming goes a considerable distance towards uniting Peircian musical semiotics with feminist theories of subjectivity.
The state of the art in the 1990s has been the development of a powerful embodied theory of musical style and narrative within a kind of test-tube: the individualistic, idealized, only nominally-social world of the philosophical aesthetics of Western art music. This has been a great achievement, not only on its own terms, but also for what it might lead to next: a fully social theory of the particularity of music as a cultural practice.
Agawu, Kofi. 1998. Musical Analysis Versus Musical Hermeneutics. American Journal of Semiotics 13 (1-4): 9-24.
---. 1991. Playing with Signs: A Semiotic Interpretation of Classic Music. Princeton: Princeton University Press.
Baest, Arjan van and Hans van Driel. 1995. The Semiotics of C. S. Peirce Applied to Music: A matter of belief. Tilburg: Tilburg University Press.
Becker, Judith and Alton Becker. 1981. A Musical Icon: Power and meaning in Javanese gamelan music. In The Sign in Music and Literature, ed. Wendy Steiner:203-215. Austin: University of Texas Press.
Coker, Wilson. 1972. Music and Meaning: A theoretical introduction to musical aesthetics. New York: Free Press.
Cone, Edward T. 1974. The Composer's Voice. Berkeley: University of California Press.
Cook, Nicholas. 1994. Music and Meaning in the Commercials. Popular Music 13 (1): 27-40.
---. 1998. Analysing Musical Multimedia. Oxford: Clarendon Press.
Cumming, Naomi. 1996. Musical Ineffability and the Fear of Smiles. Semiotica 111 (1/2): 117-141.
---. 1997a. Keeping Up (Musical) Appearances. Semiotica 116 (2/4): 319-349.
---. 1997b. The Subjectivities of 'Erbarme Dich'. Music Analysis April: 5-44.
---. 1999. The Sonic Self: Unpublished manuscript.
Cusick, Suzanne G. 1994. On a Lesbian Relationship With Music: A serious effort not to think straight. In Queering The Pitch: The new gay and lesbian musicology, ed. Philip Brett, Elizabeth Wood, and Gary C. Thomas:67-84. New York: Routledge.
Dougherty, William P. 1994. The Quest for Interpretants: Towards a Peircian paradigm for musical semiotics. Semiotica 99 (1/2): 163-184.
Echard, William. 1999. An Analysis of Neil Young's 'Powderfinger' based on Mark Johnson's Image Schemata. Popular Music 18 (1): 133-144.
Feld, Steven. 1984a. Communication, Music, and Speech About Music. Yearbook for Traditional Music 16: 1-18.
---. 1984b. Sound Structure as Social Structure. Ethnomusicology .
---. 1985. Sound and Sentiment. Philadelphia: University of Pennsylvania Press.
---. 1988. Aesthetics as Iconicity of Style. Yearbook for Traditional Music 20: 74-113.
Fish, Stanley. 1980. Is There A Text In This Class: The authority of interpretive communities. Cambridge: Harvard University Press.
Frith, Simon. 1996. Performing Rites: On the value of popular music. Cambridge: Harvard University Press.
Grabócz, Márta. 1996. The Role of Semiotical Terminology in Musical Analysis. In Musical Semiotics In Growth, ed. Eero Tarasti:195-218. Bloomington: Indiana University Press.
Grenier, Line. 1990. The Construction of Music as a Social Phenomenon: Implications for deconstruction. Canadian University Music Review 10 (2): 27-47.
Hatten, Robert S. 1994. Musical Meaning in Beethoven: Markedness, Correlation, and Interpretation. Bloomington: Indiana University Press.
---. 1999. Lectures on Musical Gesture, 1999: Semiotic Review of Books Cyber Semiotic Institute. www.univie.ac.at/Wissenschaftstheorie/srb/cyber/cyber.html.
Hebdige, Dick. 1979. Subculture: The Meaning of Style. London: Routledge.
Johnson, Mark and George Lakoff. 1980. Metaphors We Live By. Chicago: University of Chicago Press.
Kivy, Peter. 1980. The Corded Shell: Reflections on musical expression. Princeton, N. J.: Princeton University Press.
Lakoff, George and Mark Johnson. 1999. Philosophy In The Flesh: The embodied mind and its challenge to Western thought. New York: Basic Books.
Lidov, David. 1987. Mind and Body in Music. Semiotica 66 (1/3): 69-97.
---. 1994. Foreword. In Musical Meaning in Beethoven, ed. Robert Hatten:ix-xi. Bloomington: Indiana University Press.
---. 1999. Elements of Semiotics. New York: St. Martin's Press.
Martinez, José Luiz. 1996a. Icons in Music: A Peircian rationale. Semiotica 110 (1/2): 57-86.
---. 1996b. Musical Semiosis and the Rasa Theory. In Musical Semiotics In Growth, ed. Eero Tarasti:99-125. Bloomington: Indiana University Press.
---. 1997. Semiosis in Hindustani Music. Helsinki: Acta Semiotica Fennica V.
Meintjes, Louise. 1990. Paul Simon's Graceland, South Africa, and the Mediation of Musical Meaning. Ethnomusicology 34 (1): 37-73.
Middleton, Richard. 1990. Studying Popular Music. Philadelphia: Open University Press.
---. 1993. Popular Music Analysis and Musicology: Bridging the Gap. Popular Music 12 (2): 177-190.
---. 1995. Authorship, Gender and the Construction of Meaning in the Eurythmics' Hit Recordings. Cultural Studies 9 (3): 465-485.
Monelle, Raymond. 1991. Music and the Peircian Trichotomies. International Review of the Aesthetics and Sociology of Music 22: 99-108.
---. 1992. Linguistics and Semiotics in Music. Philadelphia: Harwood Academic Publishers.
---. 1996. What is a Musical Text? In Musical Semiotics In Growth, ed. Eero Tarasti:245-260. Bloomington: Indiana University Press.
Pederson, Sanna. 1996. The Methods of Musical Narratology. Semiotica 110 (1/2): 179-196.
Pierce, Alexandra. 1995. Character and Characterization in Musical Performance: Effects of sensory experience upon meaning. In Musical Signification: Essays in the semiotic theory and analysis of music, ed. Eero Tarasti:285-299. Berlin: Mouton de Gruyter.
Ratner, Leonard G. 1980. Classic Music: Expression, Form, and Style. New York: Schirmer.
Schwarz, David. 1997. Listening Subjects: Semiotics, psychoanalysis, and the music of John Adams and Steve Reich. In Keeping Score: Music, disciplinarity, culture, ed. David Schwarz, Anahid Kassabian, and Lawrence Siegel:275-298. Charlottesville: University Press of Virginia.
Sheinberg, Ester. 1995. An Application of Ernst Gombrich's Projection Theory in Art to Musical Semantics. In Song and Signification, ed. Raymond Monelle and Catherine T. Gray:38-57. Edinburgh: University of Edinburgh Faculty of Music.
Shepherd, John and Peter Wicke. 1997. Music and Cultural Theory. Cambridge: Polity Press.
Stefani, Gino. 1995. The Voice of the Heart: A progressive semiosis on the interval of the sixth. In Musical Signification: Essays in the semiotic theory and analysis of music, ed. Eero Tarasti:199-207. Berlin: Mouton de Gruyter.
Tagg, Philip. 1991. Towards a Sign Typology of Music. www.liv.ac.uk/IPM/tagg/tagghmpg.htm
Tarasti, Eero. 1994. A Theory of Musical Semiotics. Bloomington: Indiana University Press.
Turino, Thomas. 1999. Signs of Imagination, Identity, and Experience: A Peircian semiotic theory for music. Ethnomusicology 43 (2): 221-255.
Walser, Robert. 1993. Running with the Devil: Power, Gender, and Madness in Heavy Metal Music. Hanover: Wesleyan University Press.
Zuckerkandl, Victor. 1956. Sound and Symbol: Music and the external world. Translated by W.R. Task. Princeton: Princeton University Press.
Christianen, Michael. 1995. Cycles in Symbol Production? A New Model to Explain Concentration, Diversity, and Innovation in the Music Industry. Popular Music 14 (1): 55-93.
Cumming, Naomi. 1994. Measures of Meaning (essay review of Hatten). Semiotic Review of Books 5 (3).
---. 1998. Encountering Mangrove: An essay in signification. American Journal of Semiotics 13 (1-4): 61-102.
Eddie, William Alexander. 1995. Ghost Town: A case for intonational analysis. In Song and Signification, ed. Raymond Monelle and Catherine T. Gray: 28-37. Edinburgh: University of Edinburgh Faculty of Music.
Hatten, Robert. 1990. The Splintered Paradigm: A semiotic critique of recent approaches to music cognition. Semiotica 81 (1/2): 145-178.
---. 1992. Is Music Too Definite For Words? (essay review of Agawu and Mosley). Semiotic Review of Books 3 (2).
---. 1998. Grounding Interpretation: A semiotic framework for musical hermeneutics.American Journal of Semiotics 13 (1-4): 25-42.
Keil, Charles. 1995. The Theory of Participatory Discrepancies: A progress report.Ethnomusicology 39 (1): 1-20.
Keil, Charles and Stephen Feld. 1994. Music Grooves: Essays and dialogues. Chicago: University of Chicago Press.
Karbusicky, Vladimir. 1995. On the Genesis of the Musical Sign. In Musical Signification: Essays in the semiotic theory and analysis of music, ed. Eero Tarasti:229-266. Berlin: Mouton de Gruyter.
Kurkela, Kari. 1995. Tempo Deviations and Musical Signification: A case study. In Musical Signification: Essays in the semiotic theory and analysis of music, ed. Eero Tarasti:301-317. Berlin: Mouton de Gruyter.
Lidov, David. 1992. The Lamento di Tristano. In Music Before 1600, ed. Mark Everest:66-90. Oxford: Blackwell.
Lidov, David. 1994. A Guidebook for Musical Semiotics (essay review of Monelle). Semiotic Review of Books 5 (1).
Lidov, David. 1995. Toward a Reinterpretation of Compositional Theory. In Musical Signification: Essays in the semiotic theory and analysis of music, ed. Eero Tarasti: 11-25. Berlin: Mouton de Gruyter.
Lidov, David. 1996. Our Time With The Druids: What (and how) we can receuperate from our obsession with segmental hierarchies and other tree structures. Contemporary Music Review 16 (5): 79-106.
McClary, Susan. 1992. Feminine Endings: Music, Gender, and Sexuality. Minnesota: University of Minnesota Press.
Monelle, Raymond. 1991. Structural Semantics and Instrumental Music. Music Analysis 10 (1-2): 73-88.
Monelle, Raymond and Catherine T. Gray, eds. 1995. Song and Signification: Studies in music semiotics. Edinburgh: University of Edinburgh Music Department.
Nattiez, Jean Jacques. 1990. Music and Discourse. Translated by Carolyn Abbate. Princeton: Princeton University Press.
Samuels, Robert. 1995. Mahler's Sixth Symphony. Cambridge: Cambridge University Press.
Shepherd, John. 1991. Music as Social Text. Cambridge: Polity Press.
Tagg, Philip. 1994. From Refrain the Rave: The Decline of Figure and the Rise of Ground. Popular Music 13 (2): 209-222.
Tarasti, Eero. 1994. The great synthesis of Leonard B. Meyer. Semiotica 101 (1/2): 125-137.
Tarasti, Eero, ed. 1995. Musical Signification: Essays in the semiotic theory and analysis of music. Berlin: Mouton de Gruyter.
Tarasti, Eero, ed. 1996. Musical Semiotics In Growth. Bloomington: Indiana University Press.
Tarasti, Eero. 1998. From Aesthetics to Ethics: Semiotic observations on the moral aspects of arts, especially music. In In The World Of Signs: Essays in honour of Professor Jerzy York: Walter de Gruyter.