Where scholarship meets technological innovation . . .
Commentary
Matteo Ravasio
Peking University, Beijing, CN
Ravasio, Matteo. 2025. “Expressiveness, Resemblance, and Imagination.” Journal of Philosophy of Emotion 7, no. 1: 20-24. https://doi.org/10.33497/2025.summer.4.
Abstract: In these comments to Marta Benenti’s book Expressiveness, I make two points. The first concerns the relation between the proposal defended by Benenti and what are generally referred to as resemblance theories of expressiveness. My aim is to understand how far we can take Benenti’s case against the role of resemblance in expressiveness. The second is related to a debate internal to the philosophy of music, namely, the one concerning formalism. While taking a position with respect to this issue falls outside of the aims and scope of Benenti’s book, I note that her proposal may have interesting consequences for the future of this central debate.
Keywords: expressiveness; resemblance; imagination; philosophy of music; formalism
Marta Benenti’s Expressiveness is a major achievement. The book is equally successful at offering a thought-provoking overview of existing research, as it is in developing a promising new account of expressive properties. In these comments, I shall try to address both these aspects of Benenti’s work, addressing its relation to existing proposals, and its potential role in the future of the debate.
My first point is a critical consideration regarding the book’s verdict about resemblance theories of musical expressiveness. My second goal is to ask what kind of impact Benenti’s account of expressiveness could have on broader considerations in the philosophy of music, particularly with regard to the problem of musical formalism.
EXPRESSIVENESS AND RESEMBLANCE
I shall introduce my first point by going back to one of the first contributions to the analytic debate on expressive experience, namely, Richard Wollheim’s 1968 article “Expression.” Wollheim starts from a
[page 20]
commonplace distinction between expressing a state and merely communicating it, the latter case being assimilated to asserting that one is in such-and-such a state. According to Wollheim, it is tempting, and yet misleading, to reduce this distinction to one between the behavioral versus the linguistic manifestation of psychological states. On the one hand, there are cases in which behavior is revealing of our feelings, and yet “what we do is not to be thought of as expressive” (Wollheim 1968, 229-30); on the other hand, language may also be on occasions regarded as highly expressive, even though it typically is not. Having shown that the difference between expression and communication cannot be preserved by simply appealing to the linguistic or behavioral mode of expression, Wollheim suggests that it may be captured by elucidating the nature of the connection between an act (whether behavioral or linguistic) and the psychological state it is supposed to express. If the connection is law-like, and comes in the form of a constant conjunction between the state and its manifestation, such a manifestation should count as an expression; if the connection is rule-governed, then the manifestation in question is a mere assertion that one is in a certain state. But even this proposal encounters difficulties, Wollheim observes, for the appeal to constant conjunction fails to account for the appropriateness of some manifestations of one’s psychological states. Appropriateness is characterized by Wollheim as “the way expression seems so finely matched or adjusted to the inner state of which it is the outer correlate, that we can see the one in the other” (Wollheim 1968, 233). Constant conjunction being a contingent relation, it cannot account for appropriateness, for we could imagine a constant conjunction to hold between a state and a manifestation, such that the manifestation in question would not be deemed appropriate. Wollheim’s solution, in line with his general theory of expressiveness, is to appeal to the notion of correspondence as the element that allows us to discriminate between manifestations that are appropriate and those that are not, regardless of their constant conjunction to emotional states.
Jenefer Robinson’s (2007) more recent distinction between expression and expressiveness, and the related claim that not all expressions are expressive, may be taken as a reformulation of Wollheim’s central intuition.[1] Robinson points out that we may recognize happiness in a smiling emoticon, and yet this would hardly count as expressive, in the sense of conveying some sense of what it is like to be in a happy state.
Of course, Robinson and Wollheim disagree as to the reason behind these ascriptions of expressive properties, but they agree on the general point that we should distinguish between two kinds of emotional manifestations, and that while one kind is merely correlated to a certain emotional state, the other conveys what it feels like to be in that state, or at least appears in some sense to match in a nonarbitrary way the state it expresses.
Benenti’s appeal to James Russell’s (2003) theory of emotions can be seen as providing an alternative account for the distinction that Wollheim and Robinson are after. The view in question involves two main notions: core affect, and affective qualities. Core affect is a neurophysiological state specified in terms of two dimensions (pleasure/displeasure and activation/deactivation). Affective qualities are those properties of the world that “can be experienced as inherently connected to our affective dimension” (p. 159), that is, properties that have a disposition to change our core affect. Affective qualities are thus “inherently connected” (p. 159) to core affect, in that they have a disposition to modify it along its two dimensions, in a way that is determined by the perceptual features of the objects that possess such qualities. This model has the advantage of being able to explain cases of expressive experience that are hard to square with canonical resemblance theories of musical expressiveness, such as the expressive character of colors, chords, timbre, etc.
The suggestion I want to advance is that Benenti may not have taken far enough the case against the role played in expressive experience by resemblance. It is not merely that a resemblance between a perceptual configuration need not be consciously experienced, or that not all expressive perception may be explained by appealing to such resemblances. It is also the case that some resemblances between, say, music, and expressive behavior, however manifest and salient they may be, are not going to result in any sort of expressive experience. This is not
[page 21]
tantamount to denying that such perceptual stimuli are going to be consistently described as ‘sad’, ‘happy’, or in terms of other emotion words. Rather, it is simply to point out that such cases should not be mistaken for cases of expressive experience, and are best described as something else. Tentatively, one could speak of physiognomic expressiveness in the case of appearances that are typically connected to emotional states, and talk of phenomenal expressiveness whenever a perceptual configuration is related to the felt quality of an emotion. Not all cases of phenomenal expressiveness are physiognomically expressive, and not all cases of physiognomic expressiveness are phenomenally expressive. In these latter cases, we simply “read off” emotional states from a certain perceptual configuration, but this configuration does not tell us anything with regard to the state in question.[2] On this view, the mistake of resemblance theories of musical expressiveness would be that of offering a characterization of the ascription of emotions to music that conflates two distinct cases. Emotion characteristics in appearance, to employ Stephen Davies’s terminology,[3] are not all created equal, and this has repercussions for the correct characterization of the experience of expressive music.
Thus, a first option for Benenti is that of more explicitly rejecting resemblance theories as offering a picture of expressive experience that is not only partial, but also misleading. Alternatively, Benenti could accept that no unified account of expressiveness can be provided. While some applications of emotion words to music are grounded in the music’s intrinsic connection to our emotional life, other such applications are merely guided by a resemblance between the music’s ebb and flow and the way we outwardly manifest our emotions.[4] This latter option would be in line with a preference for pluralistic accounts of musical expressiveness, perhaps most representatively championed by Malcolm Budd.[5]
EXPRESSIVENESS AND FORMALISM
The second consideration I want to offer is an attempt to point to some interesting consequences that an account of expressiveness such as Benenti’s could have on a debate central to the philosophy of music: the issue of formalism and of its relation to the claim that music is expressive of the emotions.
Peter Kivy, borrowing Philip Alperson’s formulation, famously characterized his own view of music as a kind of “enhanced formalism” (Kivy 2002, 90). The enhancement comes from the idea that, in addition to the standard dynamic and gestalt properties ascribed to pure instrumental music by traditional formalist doctrines (most famously, Eduard Hanslick’s), Kivy also includes expressive properties as constitutive parts of the musical fabric. Expressive properties “come with the territory” (Kivy 1990, 183), in that they are part of the perceptual fabric of music, and are going to be grasped by anyone who is familiar enough with the music so as to be able to follow it with some understanding. At the opposite end of the spectrum, we find anti-formalist conceptions of music. What exactly makes them anti-formalist views will depend on the definition of musical formalism we embrace, but for the purpose of these remarks a list of typical features will suffice: an appeal to the composer’s intention to portray or represent emotions; the claim that the music is about emotions; the idea that a composer’s transient emotional state or underlying character is relevant to the understanding of the expressive import of their works; the view that engaging with works of music mandates the conscious involvement of extra-musical imagery–all of these are characteristic symptoms of a conception of music and its appreciation that veers away from the formalist orthodoxy.
When it comes to the debate between formalists and anti-formalists, to accept Benenti’s framework does not entail a commitment to either side. Her proposal is valuable precisely because of how it provides some of the scaffolding necessary to defend either position. On the one hand, Benenti’s claim that expressive experience is perceptual at its core, and does not require conceptual interventions, is compatible with an enhanced formalist position à la Kivy. The fundamental abilities required in order to understand the music’s expressive character are those that allow us to grasp its audible moving forms. On the other hand, as Benenti herself notes towards
[page 22]
the end of the book, her account is “. . . liberal enough to include a wide range of complex conceptually and culturally related ascriptions…” (p.170). Presumably, these ascriptions will involve some form of top-down contribution, and may involve not only concepts, but also, as Benenti suggests (p. 171), the imagination. Obviously, of such top-down contributions one could not claim, as Kivy does of expressive properties, that they come with the territory. Additionally, Benenti also makes another important claim: top-down contributions do not leave expressive experience unaltered. Through such contributions, expressive experience is “enriched and integrated” (p. 171).
To understand this thesis, we could briefly contrast it with what appears to me as a statement of the opposite view. In a discussion of the persona theory, Stephen Davies expresses skepticism at the idea that imagining a persona expressing herself in the music may be a necessary element of the experience of expressive music. Such an imaginative experience, he suggests, may well be part of a series of modes of listening that we adopt as “heuristic devices” when “coming to grips with music’s expressiveness” (Davies 2005, 149). Without wanting to read too much into this suggestion, which Davies makes only in passing, one could describe things as follows. To think of the musical persona as a heuristic is to consider it as an imaginative intervention that allows us to retrieve an expressive quality of the music that was already there, and that would have been open to retrieval through alternative means–on Davies’s view, also without any imaginative engagement at all. On Benenti’s account, however, imaginative engagements with the musical material are not expressively inert. Among the questions the book leaves open, or at least only partly answered, this is perhaps the most intriguing one: how exactly do interventions from the conceptual and imaginative domain shape expressive experience? I would like to conclude these comments not with an attempt to answer this difficult question, but rather with a suggestion regarding how Benenti’s claim as to the role of imagination may become part of an anti-formalist strategy. The anti-formalist could not limit themself to the claim that expressive experience is, as a matter of fact, modified by top-down contributions in the way Benenti claims it is, for to the formalist could accept this claim while holding that it falls short of establishing that we should engage imaginatively with the musical material in any way whatsoever. What the anti-formalist would need is a commitment to the idea that some top-down contributions are required, or at the very least appropriate, when listening to a given piece, whereas others are not. An obvious place to look for this normative element, or at least for a criterion of appropriateness, is the musical culture of the time at which the piece was composed. If top-down contributions effectively modify our experience of a piece’s expressiveness, then it seems possible to hold that adopting an imaginative engagement similar to the one adopted by the piece’s composer may help us come closer to experiencing the psychological state that the composer intended their work to express. Robinson’s claim concerning the role of the musical persona in Romantic music is easily interpreted as going in this direction:
My own view is that some Romantic pieces of music should indeed be experienced as containing a persona whose unfolding emotional life is portrayed in the music. By saying they should be so interpreted, I mean that important aspects of the work are likely to be inaccessible to listeners who do not listen in this fashion. To this extent we must listen in this way in order to detect the work’s expressive structure. (Robinson 2005, 333)
[page 23]
References
Budd, Malcolm. 1995. Values of Art. London, UK: Penguin.
Davies, Stephen. 2005. “Artistic Expression and the Hard Case of Pure Music.” In Contemporary Debates in Aesthetics and the Philosophy of Art, edited by Matthew Kieran, 179-191. Oxford, UK: Blackwell.
______. 2006. The Philosophy of Art. Oxford, UK: Blackwell.
Kivy, Peter. 2002. Introduction to a Philosophy of Music. Oxford, UK: Oxford University Press.
Robinson, Jenefer. 2005. Deeper than Reason: Emotion and its Role in Literature, Music, and Art. Oxford, UK: Oxford University Press.
______. 2007. “Expression and Expressiveness in Art.” Postgraduate Journal of Aesthetics 4/2: 19-41.
Russell, James A. 2003. “Core Affect and the Psychological Construction of Emotion.” Psychological Review 110: 145–172
Wollheim, Richard. 1968. “Expression.” Royal Institute of Philosophy Lectures, 1: 227-244.
[page 19]
Matteo Ravasio © 2025
Author email: matteoravasio88[at]gmail[dot]com
