Where scholarship meets technological innovation . . .
Commentary
Marta Benenti
University of Murcia, ES
Benenti, Marta . 2025. “Replies to Copenhaver and Odenbaugh, Alcaraz León, Ravasio, and Song.”
Journal of Philosophy of Emotion 7, no. 1: 34-42. https://doi.org/10.33497/2025.summer.6.
Abstract: In this section I try to reply to my commentators’ questions, criticisms and suggestions. Where possible, I address them one by one, following the order in which they were raised, while sometimes I gather them together in order to elaborate more systematically on the topics they raise. I am extremely grateful for these comments and I am at the same time aware that a number of issues that deserve attention will go neglected. I take this as a prompt to develop these and related topics further in the future.
Keywords: representationalism, metaphors, cognitive permeation, conceptual content, response-dependence, aesthetics, resemblances, imagination
REPRESENTATIONALISM, METAPHORS, COGNITIVE PERMEATION,
AND NON-CONCEPTUAL CONTENT
As Copenhaver and Odenbaugh point out, the account I put forward admits cognitive permeation. First, I am deeply grateful to Copenhaver and Odenbaugh for suggesting this word instead of the more common yet slightly disturbing (yet evocative) “penetration” – had I come up with this before I would have used this massively in the book! Let me keep this as a good intent for the future. Second, I must admit that I am uncertain about the actual explanatory potential of cognitive permeation in theories of perception, especially the perception of so-called high level properties, so that I declare myself open towards alternative solutions to cognitive permeation – provided that they manage to preserve some room for non-conceptual contents . Third, the appeal to cognitive permeation follows the intuition that the application of concepts to perceptual experience can modify their phenomenology. Therefore, it is plausible to think that conceptual interventions, especially those mobilizing emotion concepts, are responsible for some expressive experiences. At least, emotions
[page 34]
concepts are responsible for making expressive experiences richer, more complex, available for interaction with further – possibly higher level – experiences and mental states.
As an explanatory tool made available by the contemporary debate in philosophy of perception, I exploit cognitive permeation by distinguishing between what conceptual interventions can do at the level of perceptual content and what they do at the level of phenomenal character. Only the latter kind of conceptual permeation is actually relevant for my account of expressive experience. I argue that, when we undergo recognitional experiences, concepts seem to shape the experience of expressive patterns, be they visual or auditory.
In particular, the intervention of emotions concepts is seen as plausible in experiences featuring detectable resemblances between the targeted object and typical manifestation of affects on behalf of people. In such cases, I submit, emotion concepts are applied metaphorically. This allows me to insist that such concepts apply in virtue of properties that are actually shared by objects and typical expressions of emotions, rather than in virtue of properties that objects are taken to possess “metaphorically” – as, for instance, Nelson Goodman suggests. According to Goodman’s nominalism, what property an object possesses depends on the predicate that applies to it. Therefore, objects can possess properties in two distinct fashions, either literally or metaphorically. A painting dominated by shades of grey colour, for instance, is said to “exemplify” the colour grey, thereby possessing the property of being grey literally. Instead, such a painting would possess the property of being sad only metaphorically – or metaphorically belong to the class of sad things (Goodman 1968, 51). I counter Goodman’s view by adopting a two-stages strategy. First, I suggest that emotions are better understood as clusters of heterogeneous features which include not only internal, felt states and the related intentional objects, but also expressive contours. As such, expressive contours, like the being sad of a behaviour, can be said to belong literally to the class of sad things. When we perceive a musical gesture as sad, thus, the concept of sadness permeates the phenomenal character of our experience so long as the music possesses features that typically characterize human behaviours. The conceptual intervention is metaphorical because the paradigmatic realm of application is that of human affects and related behavioural manifestations. Yet, the perceptual features that allow for the application are actually possessed by music as well as by humans.
While the above sketched proposal can be invoked in cases in which resemblances between objects that share the same features are available, there exists a set of cases in which it cannot be claimed that conceptual interventions permeating the phenomenal character of the experiences are triggered by and anchored to shared features. In the book, this set of cases is exemplified by the experience we can have of a sad minor chord: it is hard to point at those features that such a perceptual stimulus as a minor chord shares with typically sad behaviours. Here is where the Wittgensteinian notion of secondary meaning comes into play. Likely owing to its elusive nature, secondary meaning is not much exploited in aesthetic literature. It is neither absent, though.
Stephen Davies (1994; 2011), for instance, contends that emotion words are not used metaphorically to describe expressive passages of music. On the contrary emotion words refer literally to emotional states and to their related behavioural expressions. Therefore, when we make expressive descriptions we are not referring metaphorically to any absent psychological affective state, but we are rather referring literally to something (an auditory pattern in the case of music) that resembles a behavioural expression of an affective state. Davies account is indebted to Frank Sibley’s notion of quasi-metaphors, according to which certain concepts that are at work in the aesthetic discourse have gradually lost their metaphorical status to become part of the standard vocabulary of art criticism. Yet, neither Sibley’s nor Davies’ accounts appear to exploit the explanatory potential of secondary.
Secondary meaning as it is introduced by Wittgenstein is not paradigmatic – i.e., one’s understanding of secondary meaning depends on the understanding one has of the literal meaning of the label – but it is applied
[page 35]
literally – in that one cannot substitute the employed word with any other. It can be considered either as a primitive expression, as opposed to a description, or as a quasi-metaphorical description. In the Brown Book, Wittgenstein examines the word “strain” which identifies both physical and mental phenomena. If asked to justify this double use, he observes, it is difficult to point at what physical and mental strain have in common, that is, in what way they resemble each other. The use of the word “strain” does not seem to be constrained by the possibility to recognise any manifest resemblance:
it would have been wrong to say, “I called them both ‘strain’ because they had a certain similarity,” but that you would have had to say, “I used the word ‘strain’ in both cases because there is a strain present in both.”
Drawing an analogy with the case of the blue color he adds:
Should you say we use the word “blue” both for light blue and dark blue because there is a similarity between them? [...] To say that we use the word “blue” to mean ‘what all these shades of colour have in common’ by itself says nothing more than that we use the word “blue” in all these cases (Wittgenstein 1958, 133-134).
This latter example is particularly fitting when one tries to account for the sad chord case. There seems to be no similarity between a minor chord and human expressions of sadness. Yet, the word “sad” appropriately applies to both items. Such a label applied to the experience of a minor chord – I want to suggest – permeates the phenomenal character of the expressive experience one has of it.
Are we justified in applying such emotion words to simple stimuli that do not show any relevant resemblance to affective manifestations? According to Malcolm Budd, if we seek a justification for the application of emotion terms to musical stimuli in the phenomenology of our experiences, all we find is the suitability of those terms to describe what we hear – in the very same way a shade of blue justifies the application of the word ‘blue’ (Budd 2008, 149). Is this conclusive? I don’t think so. Because, powerful and evocative as it seems, secondary meaning is a disputable notion whose semantics and psychological underpinnings are not entirely clear. I only believe that such a tool, laying half-way between metaphorical and literal applications of aesthetic and affective predicates can be further exploited in a promising way precisely in virtue of its liminal nature.
This being said, I am aware of the risk, mentioned by Copenhaver and Odenbaugh, that overintellectualizing expressive experience prevents from accounting for children’s and non-human animals’ experiences as expressive of more or less complex affective states. While I have to concede that concepts can permeate the phenomenal character of our expressive experiences, I want to insist that those experiences’ content remains non-conceptual. Besides allowing to ascribe expressive experiences to subjects that are poorly conceptually equipped, endorsing a non-conceptual content view for expressive experiences permits one to account for learning processes. Indeed, it seems that we might come to know something about what is pride by listening to a triumphant piece of music, or despair by contemplating an artistic installation. And we may do that prior to possessing the right concepts to describe and classify those states as such. More generally, it seems that we can enrich our knowledge about emotions thanks to expressive experiences that we might nonetheless be ill-equipped to conceptualize.
Admittedly, the structure of a so-conceived non-conceptual content deserves to be inquired further. The account I provide in the book is not yet satisfactory, limited as it is to the much elusive notion of rythmos and to the appeal to gestalt properties. Yet, my attempt has been that of complementing these phenomenological notions with the core affect model for emotions, which hopefully offers further support for a non-intellectualistic view of expressive experiences.
[page 36]
RESPONSE-DEPENDENCE, AESTHETICS, AND BI-STABLE FIGURES
The book is admittedly traversed by a tension between the need to ground expressive experience in perception and the attempt to do justice to our justificatory – and therefore normative – practices. On the one hand, expressive properties appear as anchored in perceptual properties, thereby looking “objective”, while on the other hand their attribution seems highly dependent on “subjective” variables. Although I deal with theories of expressiveness that treat expressive properties as fully response-dependent, the account with which I end up is “perceptualist” in that it considers expressive properties as perceptual properties whose experience can be permeated by concepts, but that remain non-conceptual at their core.
As Alcaraz León remarks, such a conundrum has been masterfully addressed by John McDowell. At the risk of trivializing McDowell's view, I would summarize his perspective as claiming that our perceptual experiences do not provide us with givens, nor brute facts. Rather, we perceive the world as already imbued with those concepts that allow perceptual contents to justify our beliefs. In this sense our responses to perceived values – or evaluative properties – are constitutive of the judgments that are based on those values. Accordingly, perceiving a musical gesture, or even a minor chord as sad or subdued would amount to evaluating it as sad or subdued, that is to correctly (or incorrectly) apply a predicate to it. As Alcaraz León rightly says, by endorsing this view there would be no need to look for a source of expressive features that be independent from the responses of the experiencer yet non reducible to merely perceptual properties.
Yet, this approach leaves me unsatisfied in two respects. First, if our expressive experiences are already conceptually structured, that is, their instantiation depends on our evaluative response, then it seems hard to explain how we learn to apply certain emotion concepts to certain experiences, i.e, to evaluate them. Second, as Copenhaver and Odenbaugh pointed out, this response-dependent strategy that relies on the conceptual structure of what is given to us in perception, ends up excluding from expressive experiences those who are not equipped with the right concepts. Clearly, these two worries are interrelated: how can a young child learn to apply the concept of sadness to something that sounds a certain way if she cannot access the relevant perceptual saliences that are shown to her, within her linguistic community, before she acquires the capacity to apply that concept?
In order to avoid these problems, I would rather preserve some room for non-conceptual contents of expressive experiences. On this perspective, expressive properties are, in their basic instantiations, closer to perceptual than to evaluative properties. Therefore, expressive ascriptions are not true or false (nor correct or incorrect). Rather, the sort of criteria my accounts gestures towards amounts to accuracy. On the one hand, in order to establish whether an ascription of expressiveness to an object is accurate we do check its perceptual and affective grounding. Accordingly, even if they do not suffice for establishing the expressive character of, for instance, an artwork, a landscape, or a complex interior design arrangement, affective qualities that are perceived as constitutive of such characters are also perceived as affectively charged rather than neutral. On the other hand, disagreement about expressive ascriptions can be legitimately dealt with by appealing to higher level cognitive (i.e. conceptual) interventions such as beliefs, memories, imaginings, background and contextual knowledge, and so on. In other words, the perceptual and affective grounding of expressive properties does not undermine the layer of the experience that is response-dependent. On the contrary, it is meant to do justice to the continuity between “merely perceptual” properties and affectively charged perceptual properties while still acknowledging their phenomenal specificities.
I suspect that difficulties in accounting for the normative side of expressive experiences arise if, in line with Alcaraz León’s stance, one patterns expressive properties after aesthetic properties. Although in the aesthetic literature expressive properties are often listed among aesthetic properties, expressive properties can be
[page 37]
instantiated in experiences that are not aesthetically relevant. So, there seems to be good reasons to consider them as exceeding the domain of the aesthetic. Take for example the experience of a minor chord sounding sad: I would not be inclined to consider it as an aesthetic experience – unless we want to blur the boundaries between what is merely perceptual and what is aesthetically relevant. My intuition is rather that, in this case, we are dealing with a perceptual experience.
The fact that this kind of experience can be characterized in terms of valence and arousal does not make it normative in the sense in which aesthetic ascriptions are normative. However, characterizing this sort of perception in terms of valence and arousal (and maybe further dimensions) offers a plausible answer to the question how and why we may point at these experiences to ground basic expressive ascriptions.
The comparison with the case of bi-stable figures is relevant here. Alcaraz León suggests that expressive experiences are to some extent akin to the experience of patterns that represent distinct objects, such as well known Jastrow’s duck-rabbit. They argues that what one sees as represented by the figure does not depend on the perceptual pattern being intrinsically “representational”. Similarly, expressive experiences would not require that the perceptual patterns by which they are enabled are expressive per se. While the case of the weeping willow and that of the musical gesture are consistent with this remark in that their explanation relies on a combination of resemblance detection and conceptual intervention, it is once again the case of the minor chord that is at stake here. In order to defend the proposed account, another example of bi-stable figure that is discussed in the literature is worth mentioning, namely the case of the black and white cross. Unlike with the duck-rabbit figure, one is neither required to possess the concept of cross, nor to detect the resemblance between the pattern one sees and a cross. Similarly, one does neither need to know how sadness is typically expressed behaviourally, nor one has to perceptually detect the resemblances it bears to sadness manifestations, in order to hear a minor chord as sad. Rather, one can take the chord as an instantiation of sadness: that is how sadness can sound - analogously to: that is how a black cross looks like. In this case, the perceptual patterns are heard - I submit - as constitutively affective, that is, as valenced, which makes them apt for anchoring (although not fully justifying) basic expressive ascriptions.
RESEMBLANCES, PHYSIOGNOMIC EXPRESSIVENESS, AND IMAGINATION
Besides thanking Ravasio for bringing to my attention Wollheim (1968), a text that I have neglected in the book, I am grateful for them pushing me to elaborate on an important aspect of my own account, namely its pluralistic nature. The distinction between different classes of cases of expressive experiences is meant to introduce a plurality of explanations none of which excludes the others. While resemblance based accounts provide powerful and plausible explanations for expressive experiences that actually rely on the existence of resemblances - and on our capacity or natural tendency to detect them - they are ill equipped to account for more basic instances of expressive experiences. Therefore, I suggest to amend and complement such views with a theory of basic expressive experience that exploits what empirically informed theories of emotion can tell us. The outcome is a layered view that acknowledges multiple components of all expressive experiences, while at the same time admitting for experiences that instantiate only some of such components.
I fully agree with Ravasio why we should not conflate what they suggest to call physiognomic expressiveness and phenomenal expressiveness. Yet, both classes of cases can legitimately be treated as including properties represented in expressive experiences. Moreover, I insist, the reason why these phenomena that, upon theoretical analysis, deserve distinct explanations, do belong to the class of expressive experiences is not merely conventional. That is, it is not only because we talk about experiences that represent physiognomic and phenomenal expressiveness in the same or a similar way that it is correct to consider them evenly legitimate contents of expressive experiences. It is rather because of their intimate connection to emotions, considered not
[page 38]
only as felt affective states, but as complex objects made of multiple components. This is why the componential approach to affective states is particularly appealing: it promises to include in the definition of emotion a wide range of features, from feelings to intentional objects, from core affect to behavioral manifestations. As such, emotions can be seen as interacting with our perceptual experiences in multiple ways, among which are expressive experiences. The latter, thus, consist in the experience of objects as manifesting emotions either in virtue of resemblances between their perceivable profiles and typical emotions manifestations, or in virtue of the combination of valence and dynamism they display, or else in virtue of a compound of these elements.
Unfortunately, I am not in the position to say much about applications of the proposed view to specific debates, but I am glad that it provides a handhold for further discussions, especially in the aesthetics domain. While preserving the non-conceptual core of the experience’s content, the theory is open to the possibility that top-down interventions shape expressive experiences, enriching them and making them more complex. Imagination is particularly powerful in creating connections and mobilizing top-down influences. So long as it admits modifications of the experiences on behalf of imaginative interventions, the proposed account is not a purely formalist account. As Ravasio remarks, however, it does not contain any normative commitment as to whether certain forms of imaginative engagement are required (or at least are to be preferred over others) in order to appreciate expressive objects. The aesthetic dimension of expressive experience, I suspect, should be discussed on this level, taking into account cultural, historical, and critical perspectives on artworks while at the same time trying to do justice to the intrinsic expressive character of their perceptual structure.
AROUSALISM, NON-ARTISTIC OBJECTS, AND EXPRESSIVENESS BEYOND EMOTIONS
Similarly to when one asks what an aesthetic experience is, or what a perceptual experience is, people disagree about what an expressive experience is. The lack of a widespread, shared definition of expressiveness and expressive experiences in the literature makes it difficult to put forward an account that be comprehensive enough. Some have the strong intuition that expressive experience paradigmatically concerns artifacts and artworks, so that expressive experience about other sorts of objects shall be considered as derivative cases. Others believe that expressive experiences paradigmatically involve imaginative engagement, so that very basic expressive experiences that do not seem to involve imagination are excluded from the class. Yet others believe that feelings are crucial components of expressive experiences. This disagreement conceals what I take to be the common ground of theories of expressiveness, namely the intuition that they are phenomenally perceptual experiences: experiencing something as expressive amounts to perceiving it as characterized by affective qualities. This assumption makes the family of expressive experiences admittedly large, including extremely complex experiences such as the appreciation of expressive musical pieces, as well as extremely simple experiences such as the recognition of minimally expressive chords.
As it has clearly emerged from this discussion, one must be careful when accounting for all these experiences at the same time. In particular, allowing for a specific explanation for the case of simple expressive experiences, one risks to lose sight of the problems raised by arousalistic explanations of complex expressive experiences. Song hits the nail on the head when they suggest that “expressiveness cannot be identified with a disposition of perceptual features to alter our core affect, or else a kind of arousalism creeps back in”. That is, a theory that wants to avoid explanations that put all the epistemic burden on the feelings felt by subjects, it should avoid dispositionalism which is intrinsic to the theory of core affect. The question indeed arises as to whether the kind of properties core affect accounts for (be it a theoretical model or an empirical fact) are actually minimal emotional reactions rather than expressive properties.
There are two things that I would like to insist on here. On the one hand, the theoretical framework of core affect is introduced in the proposed account in order to explain properties that are actually perceived as
[page 39]
expressive. Perceptual features like chords or shades of color enter the perceptual radar already charged with degrees of valence and dynamism, regardless of their eliciting affective reactions. Therefore, expressive experience of these features is not conceived as subject to actually felt feelings.
On the other hand, it is true that “affective qualities” in the core affect framework are treated as dispositional properties. Yet, their nature is not explored further by the account - presumably because it is not the focus of the theory, but only a theoretical counterpart of core affect. This leaves room for elaborating further on how perceptual properties can be detected as affectively charged without necessarily eliciting emotional reactions. As already said, the notion of rythmos recruited in the book is meant to help out in this elaboration, at least backing the idea that dynamic properties of our environment are seen and heard as intimately connected with the affective dimension, without this implying arousal reactions.
This is admittedly not enough to settle the concerns of Song, which are well grounded. However, I hope it helps to see the direction that can be taken to avoid falling back into fully arousalistic explanations of expressive experiences, and preserving the distinction between properties such as the being sad of a chord, and its being saddening.
As to the crucial relation between expressive properties and non-artistic objects, I regret not being given Song’s suggestion earlier. While dealing with my proposal, I have been looking for salient cases of expressiveness outside the domain of arts in general and of music in particular. And actually, everyday experiences of expressiveness have led my thinking throughout. Nevertheless, the literature I drew on is mainly concerned with arts, and most of the debate about expressiveness has taken and is still taking place in the aesthetics domain. This partially justifies the lack of non-artistic examples. One side issue which is related to this, moreover, is the one concerning Nature. The role of expressiveness in our relation to Nature and natural objects likely requires a specific discussion whose boundaries are only touched in the book. Not only landscapes as peculiar objects, but also the way in which expressive properties can influence our experience of Nature, along with the role of expressive properties in the aesthetic experience of Nature, are among the topics that merit further discussion.
To be fair, however, one should acknowledge that expressive properties acquire further philosophical interest when they are instantiated by artworks. While I am persuaded that one should refrain from reducing aesthetic ascriptions to expressive ones, I also think that artworks offer the opportunity to explore the most complex and therefore intriguing combinations of expressive features. Not only are expressive properties that we encounter in artworks often the result of intentional creative processes, but they also trigger processes of imagination, empathy, simulation, conceptual interventions and interpretive strategies. Such processes encompass, for instance, the conscious or unconscious attribution of intentions to artists, pragmatic implicatures, ascription and self-ascription of feelings, contextual, i.e. cultural and historical, considerations. Phenomenological and psychological investigations can in principle dissect such experiences of artworks one by one and arguably reconstruct their layered structure. Even more important, they can offer salient examples of possible interactions between expressive and aesthetic properties to be explored.
As already said, the view I presented is open enough to accommodate most of the inputs arising from aesthetic and artistic considerations. Among them there is the question raised by Song of whether there are other things, besides emotions, that can be expressed by objects. Clearly, this depends on how liberal one decides to be when assessing the meaning of “to express.” Adopting an extremely liberal interpretation, according to which the meaning of the verb “to express” amounts to all possible uses in everyday language, objects can express basically everything, from our own moods to political orientation, from religious beliefs to desires, regrets, and expectations. This is not surprising and is clearly attested by discourses in artistic criticism.
[page 40]
However, I believe that philosophical analyses require a narrower use of the verb, restricted to features that can at least be perceived. That is, on the one hand, I am inclined to exclude from the domain of expressive features, properties or qualities that cannot be detected perceptually - but, for instance, only imagined or thought. In this sense, it is unlikely that political orientation is something that can be actually expressed by an object, because it cannot be perceived. On the other hand, the aim of my research was to account for the role of emotions in expressive experiences, yet I do not want to claim that only affective states can be expressed by and perceived in objects.
More in general, although the semantic and pragmatic contexts of our descriptions do legitimize the application of the verb “to express” to several cases where non-psychological (but rather social, political, historical) states that are said to be “expressed” by objects, I favor a narrower approach that treat these latter cases as instances of representation, symbolization, reference, communication, inspiration, suggestion, metaphor, demonstration, and so on. At the cost of technicalizing the language - always for analytical purposes - the notion of expressiveness seems to me more effective and promising if limited to the phenomenon I have attempted to describe and explain, without prejudice to the possibility of broadening its scope through the exploration of borderline cases.
[page 41]
References
Budd, Malcolm. 2008. Aesthetic Essays. Oxford, UK: Oxford University Press.
Davies, Stephen. 1994. Musical Meaning and Expression. Ithaca, NY: Cornell University Press.
______. 2011. “Cross-Cultural Musical Expressiveness: Theory and the Empirical Programme.” In The Aesthetic Mind: Philosophy and Psychology, edited by Elisabeth Schellekens and Peter Goldie, 376-388. Oxford, UK: Oxford University Press
Goodman, Nelson. 1968. Languages of Art. Indianapolis, IN: Bobbs-Merrill.
Wittgenstein, Ludwig. 1958/2007. The Blue and Brown Books. Oxford, UK: Basil Blackwell.
Wollheim, Richard. 1968. “Expression.” Royal Institute of Philosophy Lectures, 1: 227-244.
[page 33]
Marta Benenti © 2025
Author email: martabenenti[at]gmail[dot]com
