It's the Thought That Counts

i agree with Jason Yust that American academic music theory is structurally racist—both in overemphasizing a white, European, and notated canon (Figure 1) and in according undue prominence to functional-harmonic conventions.1 I also agree that the term post-tonal is irredeemable, plainly suggesting that tonality is a thing of the past. But where Yust rejects the word tonality altogether, I think the problem lies in the combination with the prefix post-.I will explore this point by considering four arguments found in Yust's essay: that the word tonality is useless, that the word is manifestly racist, that the word confuses two kinds of meaning, and that the word was invented by a racist. In each case I will say why I believe tonality can still be used in an ethical and intellectually responsible fashion. Readers who disagree are welcome to avoid it; in my opinion, tonality is neither necessary nor irreplaceable. And it may be that there are some theorists whose conception of tonality is so metaphysically freighted as to be counterproductive and beyond repair. But I personally would prefer to retain this relatively common word.One of Yust's arguments is that the term tonal is not useful. I think it has at least two useful meanings. The first is to distinguish most of the world's music from the tradition that begins with Schoenberg, and which systematically rejects such common musical properties as harmonic consistency, conjunct melodic motion, acoustic consonance, tonal center, and consonant macroharmony. Here tonal is more or less synonymous with non-atonal. Yust pooh-poohs this definition, but many of my favorite composers, from Stravinsky, Messiaen, and Shostakovich to Cecil Taylor, Matthew Shipp, and Sonic Youth, mix tonal and atonal techniques; I often find the tonal/atonal distinction useful when thinking about their music.2 In what ways are Cecil Taylor's solo improvisations tonal and in what ways are they atonal? How about Strauss's Elektra? Such questions are interesting, in part, because tonal and atonal are associated with identifiable traditions of twentieth-century music making, and because musicians sometimes move between those traditions in interesting ways. When Liszt wrote his "Bagatelle without Tonality," he was clearly signaling a comprehensible compositional intention.Another useful meaning identifies music that exhibits a set of tertian harmonic conventions according to which tonics go to subdominants that go to dominants that go to tonics ("TSDT harmony"). These conventions originate in the early sixteenth century, gather force for many decades, and come to dominate European notated music from around the time of Corelli to around the time of Mahler. Call this tonality 2, or functional tonality; it typically contrasts with the freer form of tertian diatonic music known as modality (Figure 2). Like non-atonal, I find the term useful when considering music that draws from both traditions. To what extent is Marenzio's music functionally tonal?I agree with Yust that these dual meanings are problematic, and for that reason, I use functionality, functional tonality, or functional harmony to refer to tonality 2, reserving tonality for non-atonality. Meanwhile, Yust agrees that my tonality (i.e., non-atonality) is clear and well defined, but dismisses it as useless—almost as if he were disappointed that a once-great word could be brought so low. To my mind this is the weakest point of Yust's essay. Uselessness is very much in the eye of the beholder, and Yust's argument here seems fundamentally tribal: everyone in the music-theory club should talk like this. I would much prefer that music theorists pursue diverse interests in an intellectually vigorous and musically compelling manner. If the cost is a little terminological heterogeneity, then so be it: tell me how you are using your words and I will do my best to follow along.This attitude is underwritten by my belief that isolation is a significant challenge for the field. Music theory does not have a lot of contact with, or influence on, other disciplines; in part this is because so much of our effort has been expended on idiosyncratic analytical methods, often applied to music that is marginal by nontheorist standards. But partly it is because of an unusually strong desire to regiment our own language, both by rejecting common terms (Schenker 1906 1954, modulation; Powers 1992, dorian mode; White 2020, referring to famous people by their last names) and by imposing unusually strict criteria on their use (Caplin 1998, cadence; Hepokoski and Darcy 2006, second theme).3 It might be appropriate for some fields to construct their own private languages; academic music theorists, however, participate in a dialogue with composers, improvisers, performers, and nonacademic theorists. In this context, regimentation can constrain the imagination while isolating academics from the outside world. So Yust's proposal, to respond to racism with yet more regimentation, strikes me as a step in the wrong direction.As a child, my Eastern European grandmother taught me a racist term for "scam" or "swindle"; I unthinkingly used it until someone kindly pointed out that it denigrates the Romani people. Its plain meaning is offensive, and we should avoid it for that reason.4 A different issue surrounds terms like grandfathered (or grandfather clause) which, I believe, can be uttered in nonracist ways. In this case, the problem is not that these uses enact racism as that they erase it. The cleaned-up, bureaucratic use of grandfather clause is troubling because the original grandfather clauses were racist instruments of an oppressive society. It is reasonable to feel that we should be remembering the grandfather clauses as they were, not repurposing their name for other ends.5I will shortly explain why I think the plain meaning of post-tonal is intrinsically problematic, comparable to a slur though less offensive. I do not feel the same way about tonal itself: when isolated from its prefix, the term neither enacts racism like a slur nor erases it like grandfather clause. Indeed Yust agrees that tonal can be given a clear, nonracist, and purely structural definition roughly synonymous with non-atonal. I think that makes it hard to mount a compelling case against the term. If judgments about a word's utility require a delicate balancing of pros and cons, then reasonable people will disagree, and this will make it hard to forge the consensus required to eliminate the term in question.About three-quarters of the way through his essay, Yust writes that the term tonality has two different kinds of meaning: an "internal" meaning that describes musical structure and an "external" meaning that make reference to the context of its composition ("by whom, where, when, and for what purpose"). He says that the "principal claim" of his essay is that "the term tonality is problematic—racist, colonialist, and a hindrance to scholarly progress—because of the way it conflates these internal and external meanings."My initial reaction is that a similar complaint could be made about many music-theoretical expressions. Consider galant, which was popularized by Matheson in 1713. Matheson used the term to describe certain stylistic features such as contrapuntal and harmonic simplicity, identifying composers like Vivaldi and Telemann as exponents of the galant style. But galant is also used to describe a certain period in European music history, something like 1720–60, in which a lot of music had the structural features in question. (This historical use often excludes composers like Vivaldi and Telemann.) I agree that we should distinguish the galant style, a structural term that could describe music of any period, from the galant era, a historical period during which the galant style flourished. But I think it would be going too far to reject galant just because people are not always precise about this distinction. It is after all reasonable to say that the music written in a particular time and place is often characterized by certain structural properties. Galant, like classic rock, does double duty because it labels properties that did in fact become popular during the relevant period.Yust rightly notes that post-tonal makes simultaneous reference to a historical period (ours?) and a certain type of musical organization (typically set-theoretical or twelve-tone). To me the problem is not the blurring, but the specific form that blurring takes. Post-tonal originates in and plainly rearticulates the Hegelian belief that cultural products can be situated in a teleological order, with later creations superseding their predecessors. This picture led numerous musicians to think that functional harmony was characteristic of, and appropriate to, the eighteenth and nineteenth centuries, while atonality was characteristic of, and appropriate to, the twentieth—even if certain individuals happened to write other music during those times. (Though this view might seem old-fashioned and implausible, I suspect it is held by a fair number of living composers, particularly outside the United States.) This is misguided on many different levels, not least because there was a lot of non-atonal music written in the twentieth century, some of it great, and almost all of it undervalued by Hegelian modernism.Yust rightly says that the ambiguity of tonality contributes to the problem. Some composers, wanting to explore alternatives to functional harmony, mistakenly believed that they had to abandon tonality in the more general sense. Thus we can find Hegelian modernists sliding from relatively unobjectionable claims like "one can make compelling music without the conventions of functional harmony" to the much more questionable "any musician who has not experienced the necessity for the dodecaphonic language is useless."6 This produced some dramatic misjudgments about the significance of atonality itself. In my view, the twentieth century was no more an atonal century than the 1970s were the Beefheart decade. (And I say this as someone who enjoys both atonality and Captain Beefheart.) The term post-tonal all but states "tonal music used to be acceptable, but now 'we' write differently"—where "we" encodes a racist, elitist, and generally problematic view of the musical world. Here Yust and I are 100 percent in agreement.But while the plain meaning of post-tonal is clearly problematic, I do not think the same is true of tonal on its own. For one can use tonal without adopting Hegelianism, and avoid it while retaining very Hegelian opinions. Indeed, I hear echoes of Hegel in Yust's own essay: "When we see the full story, from the creation of the concept of tonality to the invention of atonality and its institutionalization, I believe it becomes clear that delegitimizing non-white music has been the agenda all along, even if those of us carrying out this agenda have in most cases not been fully aware of it" (emphasis added). Just whose agenda are we talking about? The idea that every user of the term tonality is consciously pursuing an agenda of delegitimizing nonwhite music seems implausible. The idea that every present use of tonality carries out Fétis's personal agenda isn't much better. But the idea that human beings are subject to free-floating Agendas, detached from individual agents and belonging to History or the World Spirit, is characteristic of Marxist and Hegelian historiography. It resonates with the teleology in statements like the following: Harmonic function is a tool that evolved in eighteenth-century Europe to make long-form dramatic instrumental music possible. For musical traditions unconcerned with these kinds of forms, functional norms would serve no purpose and simply be an unnecessary constraint. (Yust, this issue, p. 80; emphasis added)I would argue that functional harmony did not evolve to do anything, it just developed; we can find it in all sorts of places, whether it "serves a purpose" or not; and we can certainly write "long-form dramatic instrumental music" without it. Yust, who I consider Schenkerian, seems to be alluding to a dubious claim that local or within-key TSDT organization is inextricably bound up with global conventions (e.g., patterns of modulation among keys) that in turn create the possibility of long-term musical coherence, thus allowing composers to write longer and more dramatic pieces that feel more unified than they otherwise would. This leads him to locate the development of functional harmony in the eighteenth century, which began nineteen years after Corelli's op. 1. And it leads to the puzzling suggestion that we should not expect functional harmony in genres featuring short vocal works. But musical properties sometimes occur even when they are "unnecessary," and I hear TSDT harmony in the frottola, Germanic folk music, the Carter Family, and rock and roll.From my point of view, the important thing is that we stop endowing tonality and functional harmony with mystical attributes—that we abandon the idea that functional harmony makes long-form dramatic instrumental music possible, or that it relies on nonobvious linear connections between widely separated notes, or that it was superseded by atonality in the Great Chain of Musical Styles. Compared to Yust, I am less concerned with words and more concerned with thoughts. Both functional harmony and non-atonality are enormously complex phenomena produced by many different factors.7 I wish theorists would think more about what these factors are and how they interact. If the Schenkerian picture is wrong, then what creates the sense of melodic logic? If twelve-tone rows are imperceptible, then what features of that music do listeners actually enjoy? How should we think about atonal harmony if not with traditional set theory? How do two-, three-, four-, and seven-note logics interact in functional harmony? We are still at the early stages of understanding tonality in either of its two principal senses, in large part because our thinking is so influenced by nineteenth-century ideology. Avoiding the word tonality won't fix the problem. On the contrary, offering simple linguistic solutions to complex conceptual problems is symptom rather than cure.Finally, Yust notes that the inventor of the term tonality was himself racist. It is reasonable to ask why this matters. Pascual Jordan, one of the founders of quantum mechanics, was an enthusiastic Nazi, yet contemporary scientists use his ideas without worry. This is because physicists have been very successful at building an intellectual culture that reaches agreement about core principles. That agreement is underwritten by a relentless focus on what works—testing hypotheses, making predictions, and building things. Music is an art and will always be the subject of substantial disagreement. But music theory is an intellectual activity that sometimes borders on science. Would it be possible to create an intellectual culture that placed a greater emphasis on reaching agreement through rational discussion? On testing hypotheses? On trying to come up with ideas that are genuinely useful? Could we at least try to isolate the key questions around which reasonable theorists disagree?Why does the racism of Fétis and Schenker cause such consternation? I suspect it is because music theory is too close to its past: rather than treating earlier theorists as sources of ideas to be refined, tested, revised, and discarded, music theorists sometimes treat them as gurus or heroes or friends. We write books about what Fétis or Rameau or Hauptmann thought, applying the ideas of theorist X to the music of composer Y. As Yust points out, we teach "Schenkerian analysis" rather than "reductive analysis"—adopting an entire package of ideas, rather than carefully sifting the true from the false. People join theoretical schools like sports fans adopting a team.Thus we have produced thousands of Schenkerian analyses without bothering to test the method empirically (Temperley 2011). We have produced a comparable number of set-theoretical and twelve-tone analyses without establishing that these structures are perceptible. And none of this really matters because theorists don't focus on whether their ideas work—for example, whether they help musicians make better or more interesting music. (To the extent that we do consider these questions, we tend to ask, Does theory help me hear better?, which is a much harder question than we like to think.) Theorists talk mostly to one another, speaking their own idiosyncratic language and focusing on repertoires that are increasingly marginal. Our problems originate not in our words, but in the way we think and feel and act.