Heavy metal seems particularly appropriate terrain for such methods. As it has gained in popularity, metal has grown in stylistic innovation and pluralism. The term “heavy metal” is now used to designate a great variety of musical practices and ideological stances. Moreover, metal has contributed to the development of many discursive “fusions”: metal-influenced pop, rock, rap, funk, and so on. But as Jameson argues, the proliferation of styles within a genre and the concomitant lessened capacity of the norms to explain divergent practices do not mean that genre is no longer a fruitful analytical category. On the contrary, the recent expansion and diversification of heavy metal musical practices and their audiences make it all the more imperative to map the norms that make such fusions and transformations intelligible.6
The analytical notion of discourse enables us to pursue an integrated investigation of musical and social aspects of popular music.7 By approaching musical genres as discourses, it is possible to specify not only certain formal characteristics of genres but also a range of understandings shared among musicians and fans concerning the interpretation of those characteristics. The concept of discourse enables us to theorize beyond the artificial division of “material reality” and consciousness. Discourses are constituted by conventions of practice and interpretation, and, as John Fiske puts it, “Conventions are the structural elements of genre that are shared between producers and audiences. They embody the crucial ideological concerns of the time in which they are popular and are central to the pleasures a genre offers its audience.” Genre, then, is “a means of constructing both the audience and the reading subject: its work in the economic domain is paralleled by its work in the domain of culture; that is, its work in influencing which meanings … are preferred by, or proffered to, which audiences.”8
Traditionally, only language has been thought to be discursive. But recent usage has opened up the concept of discourse to refer to any socially produced way of thinking or communicating. The literary critic Tzvetan Todorov has analyzed the relationship of genre and discourse in a way that helps clarify the relevance of these terms to music.9 Building on Bakhtin’s theory of speech genres, Todorov argues that discourses are made up not of sentences but of utterances. That is, they are constituted not of abstract rules or patterns but of the concrete deployment of such abstractions in real social contexts. Sentences are transformed into utterances by being articulated among themselves in a given sociocultural context. For music, this implies that any formal or syntactical patterns an analyst may recognize must be interpreted as abstractions from utterances or speech acts that can only be said to have meaning in particular, socially grounded ways.
Genres, according to Todorov, arise from metadiscursive discourse. The discussion in chapter 1 of the competing definitions and understandings of heavy metal promoted by fans, business interests, critics, and others was meant to demonstrate precisely this point. It is from the discourse about discourses that concepts of genre are formed, transformed, and defended. Genres then come to function as horizons of expectation for readers (or listeners) and as models of composition for authors (or musicians). Most important, Todorov argues that genres exist because societies collectively choose and codify the acts that correspond most closely to their ideologies. A society’s discourses depend upon its linguistic (or musical) raw materials and upon its historically circumscribed ideologies. Discourses are formed, maintained, and transformed through dialogue; speakers learn from and respond to others, and the meanings of their utterances are never permanently fixed, cannot be found in a dictionary. Thus, the details of a genre and its very presence or absence among various social groups can reveal much about the constitutive features of a society.
Like genres and discourses, musical meanings are contingent but never arbitrary. There is never any essential correspondence between particular musical signs or processes and specific social meanings, yet such signs and processes would never circulate if they did not produce such meanings. Musical meanings are always grounded socially and historically, and they operate on an ideological field of conflicting interests, institutions, and memories. If this makes them extremely difficult to analyze, it does so by forcing analysis to confront the complexity and antagonism of culture. This is a poststructural view of music in that it sees all signification as provisional, and it seeks for no essential truths inherent in structures, regarding all meanings as produced through the interaction of texts and readers. It goes further in suggesting that subjectivity is constituted not only through language, as Lacan and others have argued, but through musical discourses as well. Musical details and structures are intelligible only as traces, provocations, and enactments of power relationships. They articulate meanings in their dialogue with other discourses past and present and in their engagement with the hopes, fears, values, and memories of social groups and individuals. Musical analysis is itself the representation of one discourse in terms of another, the point being to illuminate the social contexts in which both circulate.
Many critics and historians of rock music have been dismissive of any sort of musical analysis. Peter Wicke, for example, claims: “Rock songs are not art songs, whose hidden meaning should be sought in their form and structure.”10 Wicke is right to be dubious of the sort of reductive musicological analysis that simply abstracts and labels technical features, but he is wrong to assume that the specific details of rock music are insignificant. He accepts uncritically a highly problematic dichotomy when he argues, “Rock is not received through the critical apparatus of contemplation, of consideration by visual and aural means; its reception is an active process, connected in a practical way with everyday life.”11 To argue that critical scrutiny of the details of rock music is inappropriate because people don’t hear that way is like arguing that we can’t analyze the syntax of language because people don’t know that they’re using gerunds and participles. But more important, reception of all music is “connected in a practical way with everyday life,” however hard some people may work to hide the social meanings of their music. The danger of musical analysis is always that social meanings and power struggles become the forest that is lost for the trees of notes and chords. The necessity of musical analysis is that those notes and chords represent the differences that make some songs seem highly meaningful and powerful and others boring, inept, or irrelevant.12
The split between academic contemplation and popular understanding is not a function of repertoires but rather of interpretive ideologies. As recent musicological colonizations of jazz and even rock have shown, any cultural text can be made over into a monument of neutralized order. But too many analysts of popular music are unaware of the extent to which this process has already remade what is now “classical” music. They assume that traditional musicological methods are simply appropriate for the traditional musicological repertoire and that popular musics do not warrant such analysis. Yet much recent work in musicology has been directed toward undoing the formalist depoliticization of classical music. The problem is not with musical analysis per se but with the implicit or explicit ideological context within which such analysis is conducted. Rock songs, like all discourse, do have meanings that can be discovered through analysis of their form and structure, but such analysis is useful only if it is grounded culturally and historically and if it acknowledges its interests forthrightly.
This is where I differ from most semioticians of music. Jean-Jacques Nattiez, for example, might agree with much of what I have asserted thus far, but there are two important divergences.13 First, his concern is primarily metadiscursive; he seems more interested in debating definitions and concepts than in analyzing actual music and musical activities. Second, while Nattiez recognizes the conventional basis of semiological meanings, he seems to want to retain some sort of absolute notion of truth, against which interpretations can be measured. That is, he stops short of recognizing the conventional basis of semiology itself; he