Sharing uncertainty: Music in humanistic and scientific understandings

Abstract

The idea that science either can or does provide definitive ways of understanding music arose in the middle of the past century and almost immediately stimulated severe criticism, much of it from a perspective that came to elevate culture to the status to which science had been assumed to presume. It can be suggested that by the turn of the millennium the terms of this debate had begun to fragment. The idea of a positivistic science as the incremental accrual of a body of facts increasingly approximating to monadic truth was replaced by ideas of the sciences as multiple epistemologies that are mutually irreducible but somehow commensurable, as a set of related processes that result in highly instrumental yet still provisional bodies of knowledge. Simultaneously, the idea of culture as a stable and clearly identifiable set of features and practices was undermined by the realization of the political nature of the term and its applications—by a growing understanding of processes of cultural fluidity, hybridity, value, and identity formation—while the concept of the musical work was contested, as was the very idea of music as a discrete domain of thought and behavior. We have thus been left with a field of uncertainties that is too often used as an interdisciplinary battleground rather than as a meeting place to examine and negotiate the ideas and ontological commitments of our own and others’ disciplines. These uncertainties sometimes appear in plain sight but more often constitute implicit elements of disciplinary practice. We need to ensure that we negotiate encounters between disciplinary practices and commitments with our uncertainties in plain sight to develop mutual understanding across the broad worlds of humanistic and scientific research.

A problem arises when we try to make sense of music: are we aiming to understand a phenomenon that can be elucidated in terms of physical facts and biological capacities, or are we dealing with something that has to be addressed in terms that are particular, personal, and cultural? Music is at one and the same time both biological and cultural, but explorations of its sense and significance tend to be confined to either the scientific or the humanistic domain with, all too often, little consideration of—or even actual hostility toward—the alternative perspective. This article explores changing understandings of science and its applicability to music since the middle of the past century, noting how its positivistic manifestations were displaced from a central role in musicology while at the same time the bases of alternative, humanistic approaches were themselves called into question. Post-positivist conceptions of science appear plural, socially grounded, and united only in commitment to types of methods, while recent trends in the humanities have tended toward a monistic concern with issues of power. Neither domain appears in a position to claim certainty in respect of the accounts of music that they offer, and it is suggested that sharing an awareness and understanding of this uncertainty across the domains is necessary for our understandings of music to deepen and strengthen.

Some experiments on violins

In the mid-2000s I was part of an interdisciplinary team exploring how listeners distinguish between the sound of one violin and another. The problem was approached not by using real violins, but by breaking down the violin as a physical system (for a comprehensive account see Woodhouse, 2014) into its components that we could represent digitally, thus enabling manipulation of the elements that make up the sound of the instrument with something close to complete control; after all, were real violins with actual performers to have been used, it would not have been possible to distinguish between effects arising from the specific properties of particular instruments and those arising from the particular skills of the performers. A violin string can be represented as an acoustic system entirely independent of the resonances of the violin body; when bowed, it will produce a close approximation to a sawtooth waveform, and any one violin string is acoustically pretty similar to any other violin string (within limits). The resonances of the violin body, which filters and radiates the string signal, are what defines the sound of a given instrument and arise from the peculiarity of the construction of the instrument’s body (arched to brace against the tension of the strings, pierced to enable the body to act as a Helmholtz resonator, shaped to allow free access of the bow to each string separately, etc.) and the properties of the materials used—in general, Norwegian spruce, Picea abies (see Buksnowitz et al., 2007), together with protective layers of ground material (usually involving silicate) and varnish (see Barlow & Woodhouse, 1989; Fiocco et al., 2017; Invernizzi et al., 2020). Despite these complexities, the resonance characteristics of any violin body can be represented as a set of filters. Accordingly in a set of experiments real performances using a bridge-mounted force transducer—thus recording the sound of the string alone—were recorded and played to listeners through modifiable filter sets that were computer models of measured violin resonances. We found that listeners could reliably distinguish between a good and a poor instrument, and that the best listeners could distinguish between virtual violins when, in the filter set representing a violin, one of the peaks that corresponded to a low wood resonance mode was shifted by as little as a semitone—a level of performance acuity that was on a par with that reported by the best violin-makers (Fritz, Blackwell, et al., 2012; Fritz et al., 2007; Woodhouse et al., 2006).

Subsequently, Claudia Fritz, who had been the Research Associate on the project, conducted an experiment extending these perceptual findings to performance while testing the generally held proposition that Old Italian violins have some special musical properties (see, for example, Dünnwald, 1991). In her own words, “The experiment took advantage of the fine violinists, violin-makers, and violins gathered in September 2010 for the Eighth International Violin Competition of Indianapolis” (Fritz, Curtin, et al., 2012, p. 1); this was a unique set of circumstances that offered the opportunity for scientific access to the upper echelons of the world of performers, makers, and instruments, these latter ranging from some of the best new violins to instruments by some of the most celebrated makers of the past. She explored whether performers exhibited consistent preferences across instruments when playing blind, using a set of some excellent Old Italian violins and some good and outstanding modern instruments; the violins ranged in value from around US$1,800 to some US$10 million. Responses were variable but she found that professional performers did not reliably prefer the extremely expensive Old Italian instruments; indeed, the violin least liked by most performers was a Stradivarius. When published, this result caused outrage in the world of violin connoisseurship, perhaps because its results went strongly against common knowledge. Objectors claimed that because the experiment was conducted in a small space (a hotel room), the instruments had been judged in an unfavorable environment and the results were thus flawed and unacceptable. This led her to repeat the study under more controlled conditions in a Parisian recital hall; a near-replication of the original experiment (Fritz et al., 2014) with a larger sample of instruments produced more-or-less the same results as the previous one, and again the least-liked instrument was a Stradivarius. A further experiment (Fritz et al., 2017) explored the possibility that Old Italian violins might nevertheless project (be heard at a distance) better than other instruments; six violins (three Stradivari and three new) were used in live tests with audiences and it was found that new violins projected better than the Stradivarius and were generally preferred by the listeners. Hence there seems to be no objective bases for the contention that Old Italian instruments, and Stradivarius violins in particular, have unique and desirable musical properties.

Positivism and anti-positivism

Our experiments found that listeners could differentiate between good and not-so-good violins by the sound alone, and that even “good” listeners had a limited acuity in detecting changes in the acoustical make-up of violin sounds. We also found that excellent performers did not inevitably tend to prefer Old Italian or Stradivarius violins over extremely good modern violins; indeed, the general trend was in favor of the modern instruments. Science has been used to delineate the limits of our ability to discriminate between violin sounds and to shed light on performers’ preferences between violins; we could take these results to be definitive and indeed many scientists would simply do so. Around the middle of the past century, many musicologists might have concurred. The idea that science should provide definitive ways of understanding music appears to co-occur with a widespread acceptance of modernist (particularly, serialist) aesthetics, in the context of a positivistic view of science as the unitary foundational level of explanation to which all knowledge reduces.

At that time the prevalent unitary, reductionist, and materialist positivism conceived of science as a process of verifying theories framed in terms of formal logic (see, for example, Ayer, 1959; Russell, 1950). It provides a background to most, if not all, areas of writing and thinking about music, from composition to ethnomusicology, particularly in the United States and Europe. For example, most of the content of Die Reihe throughout the late 1950s and early 1960s, and the contemporaneous theorizing of US composers such as Milton Babbitt (e.g., Babbitt, 1960) is strongly committed to the idea that science provides the base levels of explanation for the materials of music. Positivistic claims on the understanding of music also appear to imbue fields seemingly more impervious to modernist impulses such as musicology; for instance, Charles Seeger (1958) asserts that “As a descriptive science, musicology is going to have to develop a descriptive music-writing that can be written and read with maximum objectivity” (p. 194). As late as 1989, a positivist agenda can be seen in M. Brown and Dempster’s reliance on Hempel’s nomological-deductive model (in which, given a set of general laws together with statements of particular facts, a conclusion can be derived deductively: see Hempel, 1962, p. 100) in proposing a version of Schenkerian music-analytic theory that could be construed as scientific.

But well before 1989 a view of science as ontologically foundational was being strongly rejected within musicology. The idea that understanding is inescapably cultural explicitly disallows any ontological primacy for an objective and culturally neutral science. For instance, Tomlinson followed Geertz (1973) in rejecting science and its apparent manifestations within the world of musicology (such as formal analysis) as appropriate frameworks for understanding music, claiming that only Geertzian thick description, “a means of understanding individual actions in particular cultures [that] stresses the local features of the description, and is not conducive to theoretical generalization” (Tomlinson, 1984, p. 100) is capable of addressing the ways in which we understand music in all its—and our—cultural specificity. The idea that cultural difference and cultural particularity are at the center of how music should be understood, both in terms of what is to be understood and how those understandings are to be achieved, is exemplified in the claim that “both music and social movements . . . are best figured not as organic totalities but as constellations of heterogeneous mediations” (Born & Barry, 2018, p. 458). In other words, from the perspective of Born and Barry, both music and, indeed, social movements that might elsewhere be conceptualized as cultural, are better conceived of and analyzed in terms that are embedded in specific contexts and local frames of reference than in terms that claim general or even universal applicability. Along these lines, Kramer implicitly excludes any generalizable form of enquiry from any significant role in understanding music when he claims that “subjectivity as . . . a socially constructed position made available by the music and occupied to a greater or lesser degree by the listener . . . is not an obstacle to credible understanding but its vehicle” (Kramer, 2003, p. 126). Hence unitary, reductionist, and positivistic science can have no claim on the “identities, differences and constitutive outsides of disciplinary and aesthetic formations” (Born, 2010, p. 230) that frame both music and how it can and should be understood. Indeed, some musicologists have claimed that science itself disintegrates when subjected to critical-theoretical interrogation (concerned with framing and reframing ideas of societal dominance and power: see, for example, Agger, 1991) that situates it in cultural context. Todd Titon declares that

Latour’s . . . work on scientific laboratory culture underlined Foucault’s claim that knowledge did not progress toward greater understanding but instead was captive to a social consensus among an interpretive community of qualified scientists . . . subject to the general prejudices prevalent in society. (Titon, 2013, p. 13)

Science itself appears to be conceived of as a cultural pursuit, proper to the dynamics of particular historical societies and susceptible of being appropriated to political ends—scarcely the unitary and foundational level of explanation proclaimed by the positivists.

However, such an interpretation of Latour’s ideas is wide off the mark. Latour (1999) makes clear that he believes that although scientists are not immune to the “general prejudices prevalent in society,” neither is the conduct of science wholly shaped by those prejudices; science can indeed result in an accumulation of knowledge. How it might do this is open to question. One possible positivist route could be to accept that a theory about states of affairs in the world constitutes scientific knowledge according to whether or not it manifests a logically transparent structure (Reisch, 2007) and is susceptible to confirmation; on this basis the role of experiment would be to confirm the truth of the theory and express that fact within a superordinate framework of logic (see, for example, Hempel, 1962). The simplest objection to this proposal is to note that on these grounds, establishing the universal applicability of a theory in respect of the class of phenomena it purports to explain would involve exhaustive testing of all possible instances—an evident impossibility—hence Popper’s (1962) suggestion that theories should be framed so that they are experimentally falsifiable rather than confirmatory or verifiable. The fundamental operations of science would thus consist of attempts to formulate well-formed hypotheses in such ways that they could be disconfirmed by experiment (see also Sankey, 2013).

Science as the sciences

Nevertheless, philosophies of scientific method grounded in either falsificationism or verificationism can be viewed as equally problematic. From the perspective of the logical empiricists such as Hempel, Carnap (1950), and others (including, effectively, Popper), the conduct of science results in empirical facts that can be framed within an observation language that can then be related, via a set of correspondence rules, to a set of axioms containing theoretical terms that can be expressed in terms of formal logic (see, for example, H. I. Brown, 2005). But there is no a priori obligation for theories that are, or that purport to be, explanatory about empirical, scientific facts to be capable of being consistently re-expressed—even via a complex chain of translations—within the terms of the formal, Principia logic preferred within the logical empiricist tradition (see H. I. Brown, 1977, p. 31) or, in fact, within the terms of any particular logic; indeed, in respect of science, Feyerabend (1962) makes the claim that “a formal and ‘objective’ account of explanation cannot be given” (p. 95). Hence the idea that there exists a generalizable foundation in formal logic for asserting the validity of scientific explanation seems to rest more on faith than on logical necessity, irrespective of whether or not one adheres to verificationist or falsificationist approaches—even those such as are proposed in Lakatos’s (1970) theory of scientific research programs under the banner of “sophisticated methodological falsificationism” (p. 116). Lakatos’s approach distinguishes between core and auxiliary theories, and proposes that auxiliary theories are more likely to be susceptible to falsification than core theories; however, if a large enough proportion of auxiliary theories are falsified, core theory must change to fit. What constitutes a large enough proportion of falsifications will depend on the consensus of the scientific community; an element of social judgment concerning what constitutes scientific knowledge is thus allowed to infiltrate the domain of logical empiricism.

The idea that scientific knowledge and practice is socially mediated is generally associated with Thomas Kuhn, whose theories have been interpreted as circumventing some of the problems of logical empiricism by stressing that science is (at least in part) determined by what scientists do and debate, and that the meaning of scientific concepts rests on their relationships to other concepts. Kuhn (1962) proposes a retreat from the absolutist realism of positivism, suggesting that science normally proceeds within generally accepted paradigms that are susceptible to change not only on the basis of lack of fit with experimental results but also on grounds of the acceptability of new paradigms within the community of scientists. In addition to theories and experiments, the actions of scientists and their interactions with their peers thus acquire agency in making and, indeed, remaking scientific knowledge. Latour (1999) goes somewhat beyond Kuhn in situating science in the social world, ascribing agency not only to scientists and their psychologies but also to the objects of study with which they interact in shaping scientific knowledge, rather than grounding this solely in observation and logic (or in experiment and theory). This is most clearly delineated in his account of Actor Network Theory (ANT), in which he claims that characterizing modern societies requires recourse to “a network-like ontology and social theory” that is neutral with respect to the entities and actions it describes, specifying “only what the recording device should be that would enable entities to be described in all their details” (Latour, 1996, p. 370). Within ANT,

An “actor” . . . is a semiotic definition—an actant—that is, something that acts or to which activity is granted by others. It implies no special motivation of human individual actors, nor of humans in general. An actant can be literally anything provided it is granted to be the source of an action. (Latour, 1996, p. 373)

In this, Latour seems to demonstrate a “commitment to ontological equality” (Campbell, 2021, p. 418), extending and, arguably, stretching beyond breaking point the idea of agency in according the capacity to act equally to the human and the nonhuman. The indiscriminate nature of this universal extension of agency can be interpreted as giving rise to more problems than it resolves, especially in its disregard for the significance of the concept of volition for ideas of agency in the context of human behavior (McCann, 1974; Zhu, 2004; see particularly Frith, 2013). Latour’s account of the locus of agency in the conduct of science seems to suffer from overreach in comparison with Hacking’s (1999) approach, which distinguishes between interactive and indifferent kinds. Hacking identifies interactive kinds as entities and concepts that may be influenced by the ways in which a scientist may employ them or engage with them (such as gender, or even sex), while indifferent kinds are those of which the properties and identities appear indifferent to any such encounter (such as plutonium).

However, Hacking’s distinction points to a further complication in framing science; it appears to map onto a distinction between human and physical sciences, the former likely to be concerned with interactive kinds and the latter with indifferent kinds. But even that distinction may be too crude (Khalidi, 2010). Fodor (1974) proposes a form of scientific pluralism in the existence of special sciences that address phenomena that cannot be wholly accounted for in terms of the laws of physics. He suggests that rather than conceiving of other frames of scientific understanding as reducing to that of physics, they instead supervene on physics: that is, any property expressible in a frame of scientific understanding other than physics cannot be changed without some corresponding change having occurred in the scientific frame of understanding that constitutes physics (see, for example, Steinberg, 2014). As he puts it, “not all natural kinds (not all the classes of things and events about which there are important, counterfactual supporting generalizations to make) are, or correspond to, physical natural kinds” (Fodor, 1974, p. 113). The idea of a positivistic science as the incremental accrual of a body of facts increasingly approximating to monadic truth that is reducible to—and hence expressible in terms of—the so-called hard science of physics is replaced by ideas of the sciences as multiple epistemologies that are mutually irreducible but in some senses mutually comprehensible—as sets of related processes that result in highly instrumental yet still provisional bodies of knowledge (see also Thagard, 2013).

Even here a further distinction must be drawn. Those sets of related processes typically encompass forms of methodological reductionism, wherein a complex phenomenon is conceived of as comprising interacting variables that can be independently manipulated in experiment. In effect, in designing an experiment the complex phenomenon to be investigated is operationally reduced, and it is the operational reduction, rather than the full-blown, messy, contextually embedded, real-world phenomenon, upon which the experiment will be conducted. In the case of the violins experiments referred to earlier, the violin is operationally reduced to two variables, the string and the body, each independent of the other and interacting in restricted ways (energy output from the string acts upon the body and not vice versa); only values of the body (represented as a filter set) are manipulated. Moreover, the duration of the stimuli that were evaluated by listeners was restricted to 300 ms, one-third of a second, extensive pilot testing having shown, through a process of epistemic iteration (Eronen & Romeijn, 2020), that longer musical phrases led to responses that were far too variable. Consistent results were obtained using highly reductionist, indeed, psychoacoustical, methods, and the explanatory value of those results is accordingly limited in its generalizability.

In the case of the violins experiments, the findings, though based on listening to highly reduced stimuli, converge with informal but consistent reports from expert makers about the limits of perceptual discriminability between instruments. Moreover, there is a mass of evidence (see, for example, Woodhouse, 2014) that the acoustical characterization of the instrument used in the experiments, in which it is reduced to two principal variables, retains the most significant and generalizable features of the instrument and constitutes a very close approximation to those aspects of the real-world violin that determine the acoustical differences between instruments (although see Bissinger, 2006; Gough, 2018). However, it is not possible to claim that the limits of discriminability that were identified necessarily extend beyond circumstances in which the bases for evaluation are severely circumscribed, as in the perceptual studies. In particular, it is not possible to make any predictive claims about how players might be able to discriminate between instruments.

Indeed, when we attempted to extend the experimental paradigm beyond listening to performance by using a performer playing an electric violin (effectively, a bodiless set of strings) as input to the digital virtual violins and asking the player to evaluate the difference between two violins, we received a response that had not been anticipated. When a switch was made from one digital virtual violin (filter set) to another, the performer spontaneously announced, “Oh, this one’s much easier to play.” Yet the physical instrument that she was playing, the electric violin, had not changed in any way; what had changed was the relationship between the performer’s actions and the fine details of the sound produced, and the player responded to the change in sound as though it were a change in the feel of the instrument. In effect, for the player, a change in the relationship between action and sound was experienced as a change in the quality of the action—a fact that should have been less surprising to us than it was, as for any expert performer the actions they undertake in playing a particular instrument and the sounds that are consequently produced are almost perfectly correlated (see Pfordresher, 2019; but also see Wollman et al., 2012). For a performer, the instrument is thus an element in an auditory–haptic complex rather than something that simply emits sound (as in our perceptual experiments), and has to be judged accordingly.

The extent to which inferences may be made on the basis of experimental results will depend on the extent to which the operational reduction that guides the conduct of the experiment constitutes a more, or less, comprehensive representation of the phenomenon in question. Thus while the methodologies of science are typically reductionist, that reductionism does not necessarily carry over to scientific explanation (see Schultz, 1996, on operational versus constitutive—explanatory or epistemic—reductionism in practical science). As Hüttemann and Love put it,

Explanatory and methodological reduction can be decoupled because they do not entail one another: methodological reductionism does not guarantee explanatory success and a successful explanatory reduction does not imply that methodological reduction is the most favorable strategy of inquiry. (Hüttemann & Love, 2011, p. 524)

Science, then, is better conceptualized as the sciences; there appears to be no principled way of reducing scientific theory and knowledge to successively more fundamental forms—rather, other sciences supervene on those that can be considered more fundamental, such as physics; the sciences can guarantee to provide neither absolute certainty nor comprehensive understandings of the world; and the sciences share commitments to methods rather than to specific kinds of knowledge. Given this last property, one might reframe the suggestion that “All the facts that there are . . . ‘supervene on’ the facts of basic physics” (Fodor, 1998, p. 6) as “All the facts that there are that can be determined by the application of acceptable scientific methods . . . ‘supervene on’ the facts of basic physics.” However, in the end, what the sciences do offer are interconnectable, ramified, and testable theories that can enhance our capacity to predict events, circumstances and situations. And when we try to understand or explain music from a scientific perspective, it seems to require to be the focus of a plurality of sciences (see Currie & Killin, 2016).

Problems with culture

In exploring and explaining aspects of human life, science is usually counterposed with culture as a key—perhaps the key—term within humanistic disciplines. Attempts to oppose what is construed as science’s determinism or its epistemic imperialism have focused on the idea that humans are cultural animals, and it can be argued that by focusing on the human animal, science has tended to underplay or even ignore the role of culture in shaping the human world, in terms of our perceptions, our actions, and even our biologies. ¹ The human world seems socially or culturally constructed; as Littlewood puts it, “To deny cultural construction is to suggest that we can really experience the world out there as it is. Some hope” (Littlewood, in Ingold, 1996, p. 101). The idea that the world is socially constructed comes into focus in the work of Berger and Luckmann (1966), who postulate that “[o]ur paramount reality (everyday reality) exists as an intersubjective reality” (Vera, 2016, p. 6). The human world is consequent on the ways in which humans interact and reach agreement (implicit or explicit) concerning how and what things and events are and should be; it is thus, for many sociologists and anthropologists, a world that is beyond the reach of scientific enquiry and that can only be addressed by exploring cultural facts and processes. As James puts it,

The very notion of a “world” implies some sort of coherence, if not as a simple integrated whole, then at least as an ensemble within which the possibility of cross-linkages of interpretation can be imagined; if the cultural element were withdrawn, separated or denied, the ensemble of fragmentary action and experience would fall apart. (James, in Ingold, 1996, p. 88)

However, more-or-less simultaneously with the shift in the concept of science from a domain of absolute certainty to one of fragmented instrumentality, the idea of culture as a stable and clearly identifiable set of features and practices has been undermined by the realization of the political nature of the term and its applications, by a growing understanding of processes of cultural fluidity, hybridity, hegemony, and value and identity formation. Over the past century, definitions of the term culture within sociological and anthropological thinking have been variously founded in frameworks that are economic, racial, class-based, hegemonic, identitarian, intersectional, and essentialist, among others. These apparently diverse frameworks have increasingly tended to be reduced to theorization in terms of power, although those power-based theorizations themselves present considerable epistemic diversity. For example, Foucault (1982) sees power as foundationally disciplinary; Bourdieu’s (1977) theory of habitus sees power as immanent in social structures; for Latour (1984), power is wholly performative; for Forst (2015), social power is noumenal (grounded in narratives of justification) rather than phenomenal. All these views can be seen as prefigured in, if not preempted by, the writings of Gramsci (Hoare & Nowell-Smith, 1971), whose concept of hegemony (“political leadership based on the consent of the led . . . secured by the diffusion and popularization of the world view of the ruling class”; Bates, 1975, p. 352) has come to be central to humanistic thinking about societal structures and dynamics, although the various routes by which it has come to serve as a conceptual tool have themselves given rise to extensive debate concerning its scope and dynamics (see, for example, Crehan, 2022; Kurtz, 1996).

Rodseth suggests that approaches privileging power as the primary determinant of human action and interaction have tended unsustainably to narrow the field of how culture and its dynamics can be understood; as he notes, “even if power is always a cultural phenomenon, culture is not always a political one” (Rodseth, 2018, p. 400), and to conceive of culture solely in terms of power is to exclude the necessary dimension of cooperation that can be shown to underlie productive human interaction in the first place (Boyd & Richerson, 2009; Curry et al., 2019). A similar narrowing can also be seen to inflect theories that focus on issues of identity; as Bernstein puts it,

observers argue that identity politics does not challenge the social construction of status categories, ignores the intersection of identities, forces those with multiple identities to privilege some aspects of identity over others, fails to recognize diversity within groups, imposes a uniform identity on groups that are diverse, and essentializes a group’s identity. (Bernstein, 2005, p. 57)

Theories of intersectionality, often founded on identitarian concerns but seeking “to focus attention on the vexed dynamics of difference and the solidarities of sameness in the context of anti-discrimination and social movement politics” (Cho et al., 2013, p. 787), again center on issues of power. The tension within intersectionality theory in attending simultaneously to identity and inequality can lead to an impracticably broad focus, while it can also result in a significant contraction in explanatory scope, as noted by Bilge (2013). Explicit attempts to rescue the culture concept by developing more inclusive accounts can also be sabotaged by an unfeasibly broad focus. Bennett, for example, proposes that culture can be construed as “a conjunctural and pliable articulation” of relations between “a people, place and way of life . . . that derives its distinctive qualities from the creative, form-giving capacity of the people concerned” (Bennett, 2015, p. 246). While this proposal fulfills many of its immediate purposes in cleansing the culture concept of the essentialism that contaminated its usage within racial theories in the sciences of the late 19th and early 20th centuries (see, for example, Jackson & Weidman, 2005; Marks, 1996), its very inclusiveness appears to render it almost too open-ended to be practically useful.

For some, the explanatory powers of the culture concept have become so depleted that the only solution is to bypass culture by decentering the subject at its core, the human. Recent theorizing—somewhat paradoxically situated within humanistic traditions, albeit those imbued with the spirit of critical theory—has erected the idea of the posthuman as a means of making sense of the apparently increasing permeability of the concept of the human in the light of interpretations of developments in technology and in the sciences (particularly, in genetics). This work appears to offer ways of reconciling aspects of scientific and humanistic intellectual traditions, at least in part by adopting “STS [Science & Technology Studies] as a controlling interdisciplinary approach” (Bolter, 2016, p. 7), and hence opening up the possibility of bridging the humanities and social sciences. However, the frameworks that it propounds tend, on closer inspection, to reproduce deficiencies common to recent elaborations of the culture concept: in particular, the reductionist reliance on power as an axiomatic term of analysis. This is clearly evident in the claim that

a posthuman approach requires . . . careful cartographies of the different degrees and the extent to which any one of us can be said to be “human” . . . The aim of a cartographic method is to provide a politically grounded and theoretically infused account of the webs of power relations we are all entangled in. (Braidotti, 2016, p. 15)

Tales of the posthuman thus seem themselves to be susceptible to something like the criticism raised by Rodseth (2018) a propos power-based theories of culture; they narrow rather than extend the field of how the human can be understood.

Quite apart from the increasing insubstantiality of the culture concept, many ideas that have been considered central to humanistic engagements with music have also been subject to destabilization, in part through the types of debates that have surrounded concepts of power or identity and in part through a heightened need to make sense of music beyond the boundaries of WEIRD (Western, Educated, Industrialized, Rich and Democratic) cultures (Henrich et al., 2010). The idea that music has some immanent existence and can be engaged with in the form of identifiable and discrete entities termed works that are created by specific individuals has been undermined by situating the idea in its historical context (see, for example, Goehr, 1994); the socio-economic nature of the work-concept, and its cultural particularity, is highlighted in scholarship that shows its close ties to the emergence of the idea of ideas as intellectual property, particularly in England in the 18th century (Barron, 2006; Hunter, 1986). The idea of a Kantian notion of a disinterested engagement with the beautiful or sublime is challenged by the development of concepts of beauty related to everyday utility that may contribute to meaning or significance, as in Davies’s (2006) notion of functional beauty. Similarly, the universality of an autonomous aesthetics is contested by the ways in which at least some non-western cultures bind something like beauty to other domains, particularly the social. As Perman notes in respect of performance in Ndau communities of southeastern Zimbabwe,

The ethics (or ethical values) of ceremonial and performative efficacy in performance are synonymous with what is often thought of as aesthetic judgment. Ndau ideas of beauty are thus tied to social and ethical values [hence] Beauty is not an autonomous ideal [in Ndau culture]. (Perman, 2011, p. 227)

Indeed, the idea of music as a discrete and universal domain of human thought and behavior has itself been called into question. In part, this uncertainty has arisen because it has become evident that at least some cultures simply do not distinguish phenomena that appear cognate with western notions of music as constituting a unitary category; an instance of this is the way in which wosi (songs) in the Trobriand communicative taxonomy constitute a subdomain of the predominant Trobriand communicative register biga sopa (joking or lying speech) (Senft, 2018). And, in part, it receives empirical support from studies that indicate that similar processes underpin interaction in both music and at least some registers of speech (see Robledo et al., 2021), leading to the suggestion that as interactive media, music and phatic speech (speech that is oriented toward establishing and maintaining social relations rather than toward exchanging information extrinsic to the immediate social interaction: see Malinowski, 1923) both constitute subdomains of a superordinate domain of affiliative communicative interaction (Cross, 2022).

Sharing uncertainty

In summary, scientific knowledge is neither unitary nor certain—but then neither is humanistic knowledge. What remains of positivism in the sciences is a set of interconnected frames of understanding that share methodological commitments and that are not mutually reducible but mutually “commensurable” (in terms of each being reciprocally comprehensible from the others’ perspectives, after Lakoff, 1987, p. 322). These interconnections do not appear to be shared with humanistic frames of understanding, which seem to possess their own, perhaps more malleable, transitory, contextually bound and often historical, linkages. From both perspectives, humanistic and scientific, knowledge is not a terminus but a process of understanding the world that is always provisional and, particularly in respect of a hard topic such as music, necessarily multiple and heterogeneous (Cross & Tolbert, 2021).

What does appear increasingly to pervade both the sciences and the humanities are uncertainties about concepts and beliefs that had seemed axiomatic; if there is to be any value in synergy between the two domains—if each domain is to make sense of the other—then those uncertainties need to be shared. Across disciplinary boundaries, there needs to be a mutual recognition of those uncertainties and of the epistemic limits of any residual certainties that each domain might claim. This mutual recognition will rely on an awareness and an articulation of each domain’s proper uncertainties, and some forms of cross-domain translation processes are likely to be required for the emergence of a degree of mutual comprehension. It is likely that awareness and translation are probably the most critical attributes for cross-domain understanding and also the most difficult to achieve.

Mutual recognition of uncertainty is not a given; interaction between the sciences and the humanities can take several forms, depending on research outcomes and on inter-domain attitudes. Outcomes can be siloed: they can be parallel and proper to particular epistemic domains, and hence difficult or impossible to translate; the methods whereby they have been derived and the ontologies in which they are embedded may simply be impermeable to those conceivable within the other domain (see, for example, MacLeod, 2018). Where outcomes are evident across domains, requiring a degree of mutual inter-domain awareness and accessibility, they may reflect convergence in referring to results that are mutually confirmatory and enriching, in that they identify or refine questions that could be addressed within one or both domains. Alternatively, they can manifest as dissonance, reflecting conflicting modes of understanding indicative of terminal disjuncture or requiring further debate.

Inter-domain attitudes can be indifferent, arising from lack of awareness or of flexibility, or reflecting blinkered commitment to particular epistemic traditions; for instance, few scientists are particularly motivated to engage with critical theory or even with science and technology studies, while the majority of critical theorists are unlikely to undertake experimental studies of the types that would be recognized as such by most scientists. Attitudes can be collaborative, motivated by a drive for understanding—an awareness of the strengths and limitations of one’s own discipline and a confidence in confronting the exigencies of strange and novel ways of thinking. Or, in the worst case, they can be antagonistic, driven by a lack of confidence in vulnerable sets of unexamined beliefs and a sense of epistemic and ontological threat.

In all, even when there is an awareness that the same types of issues are being tackled—after all, we are generally addressing something we want to conceptualize as music—we need to be able to articulate to each other our agreements and disagreements about whether, and if so, why, what it is that we are looking at is, or is not, the same thing. We have to be able to share and debate our uncertainties to understand each other and the world a little better. Of course, underpinning all is the willingness to admit to uncertainty in the first place.

Coda

Claudia Fritz’s results showed there was no evident bias among performers toward Old Italian instruments. Unknown to her, at around the time of her experiments, an excellent violin maker and repairer, Tom Wilder, was completing a doctorate on the emergence of the idea of Stradivarius violins as not only superb musical instruments but also transcendental art objects, partly driven by his own practical awareness of the extent to which most Old Italian violins had been subject to extensive restoration and repair and partly by having gained access to the archives of the London violin dealers W. E. Hill & Son. ² In his dissertation (Wilder, 2017), he revealed how historical, social, craft, and commercial practices worked together across the 19th century both to develop the idea of the violin as more than just a means of making music and to sanctify the early Italian, Cremonese, instruments (displacing late-18th century preferences for the somewhat earlier instruments produced by the Austrian maker Jacob Stainer, favored by Handel). Using a range of sources including the Hill archives, he was able to show that across the 19th century Hill & Son made significant interventions in the market in violins that boosted the monetary and aesthetic value of their favored Stradivari instruments into orbit. The results of his historical and humanistic research clearly converge with those of Fritz’s scientific research; there are no good bases for regarding Stradivarius violins as unique or particularly special in relation to other good violins, whether old or new. Nevertheless, these findings do not seem to have had any real effect on the market for violins; the prices paid for Stradivari instruments at auction appear to have remained largely immune to such evidence.

But a codetta to this coda: in July of 2022, Christie’s auction house in London offered for sale the Hellier Stradivarius, of impeccable provenance and described as one of the finest Stradivarius instruments in existence. It was anticipated that the violin would sell for between £6 and £9 million. It attracted no bids at all. ³