1. Introduction
In contemporary philosophical reflections on multimodality, an important distinction is that between instance-multimodal and type-multimodal features (see Briscoe, Reference Briscoe2021; O’Callaghan, Reference O’Callaghan, Stokes, Matthen and Biggs2015a, Reference O’Callaghan2017). More specifically, an instance-multimodal feature is one such that some of its particular instances are represented through the joint functioning of several modalities, while others are represented unimodally. By contrast, a type-multimodal feature is one such that all of its particular instances are represented multimodally. In other words, there is a difference in content between experiences representing instance-multimodal features and those representing type-multimodal features. Type-multimodal features are represented only by multimodal contents (i.e., contents that arise from the functioning of several modalities), whereas instance-multimodal features are represented by contents that occur either multimodally or unimodally. Further, I use the term “type-multimodal content” to refer to the contents by which type-multimodal features are represented, and the term “instance-multimodal content” to refer to the contents by which instance-multimodal features are represented.
The most common example of a type-multimodal feature is flavor, as flavors are always represented as the joint functioning of gustatory, olfactory, tactile, and trigeminal mechanisms (see Auvray & Spence, Reference Auvray and Spence2008; Connolly, Reference Connolly and Brown2014; Matthen, Reference Matthen and Matthen2015; O’Callaghan, Reference O’Callaghan2015b; Smith, Reference Smith and Matthen2015). In this paper, I propose a reductive account of flavor perception. More specifically, I argue that even if flavors are type-multimodal features—and are represented by type-multimodal contents—their contents are not irreducibly type-multimodal. It means that such type-multimodal contents can be plausibly analyzed in terms of a combination of contents that are not type-multimodal.
As a starting point, it should be noted that philosophers of perception commonly assume that our ordinary perceptual experiences exhibit some form of multimodality (e.g., Briscoe, Reference Briscoe2016; Macpherson, Reference Macpherson2011; O’Callaghan, Reference O’Callaghan, Samuels and Stich2012, however see Spence & Bayne, Reference Spence, Bayne, Stokes, Matthen and Biggs2015). It seems that a single conscious episode possesses contents associated with functioning of mechanisms that constitute distinct modalities. The assumption regarding the multimodality of ordinary perceptual experiences is typically supplemented by a conviction that there are stronger and weaker types of multimodality (e.g., Fulkerson, Reference Fulkerson2011; Kubovy & Schutz, Reference Kubovy and Schutz2010; O’Callaghan, Reference O’Callaghan, Stokes, Matthen and Biggs2015a; Richardson, Reference Richardson2014; Stevenson, Reference Stevenson2000). The weakest form, which may be named “minimal multimodality” (see Bayne, Reference Bayne, Hill and Bennet2014; Connolly, Reference Connolly and Brown2014; O’Callaghan, Reference O’Callaghan, Stokes, Matthen and Biggs2015a, Reference O’Callaghan2015b, Reference O’Callaghan2017; Macpherson, Reference Macpherson2011 for variants of this idea), occurs when within a single experience there are several unimodal contents associated with distinct modalities and an experience represents merely a co-occurrence of elements represented by these unimodal contents. In such an experience every element of content is unimodal, in the sense that it occurs in virtue of mechanisms related to a single modality, and multimodality consists only in the co-presence of unimodal contents.
A stronger form of multimodality occurs when an experience has content that occurs due to the joint functioning of more than one modality. A common example concerns the representation of multimodal spatial relations that are experienced as connecting elements associated with distinct modalities (see Briscoe, Reference Briscoe2017; O’Callaghan, Reference O’Callaghan, Bennett and Hill2014, Reference O’Callaghan, Stokes, Matthen and Biggs2015a; Richardson, Reference Richardson2014). When a visual object is experienced as being to the left of a sound, the spatial relation is not experienced solely due to the functioning of vision or audition, but requires the functioning of both of these modalities. In consequence, an experience representing multimodal relations is not multimodal merely in the sense of having various unimodal contents, but in the sense of having a genuinely multimodal content.
A further distinction regarding the multimodality of experiential contents can be characterized using the notions of type- and instance-multimodal contents introduced earlier. An experiential content representing an audiovisual spatial relation, as mentioned above, is an instance of a content representing spatial relations. Some instances of this content are multimodal; for example, they occur only through the joint functioning of vision and audition. However, content concerning spatial relations also has instances that are unimodal. For example, a spatial relation between visual elements is represented unimodally, by vision alone. Thus, contents concerning spatial relations are instance-multimodal: some occur only multimodally, while others occur unimodally. It is also possible, however, that some experiential contents are type-multimodal: they have no instances that occur through the operation of a single modality.
The most popular candidate for a type-multimodal perceptual contents is contents representing flavor (see Briscoe, Reference Briscoe2021; Connolly, Reference Connolly and Brown2014; Matthen, Reference Matthen and Matthen2015; O’Callaghan, Reference O’Callaghan, Stokes, Matthen and Biggs2015a, Reference O’Callaghan2015b, Reference O’Callaghan2017; Smith, Reference Smith and Matthen2015). Flavors are features referred to by names and descriptions such as “vanilla flavor” or “cold sweet coffee flavor,” which are represented by contents of flavor experiences. These are experiences that appear to occur in the mouth, but are mainly exteroceptive, i.e., they attribute features to external objects located in the oral cavity. It is commonly claimed that flavors are perceptually represented by the combined functioning of different sensory modalities (however, it is still debated which of them are constitutive of flavor perception and which merely modify the flavor experience, see Skrzypulec, Reference Skrzypulec2021; Smith, Reference Smith2013; Prescott, Reference Prescott2015). In consequence, a flavor is not represented by any single perceptual system, so each instance of content representing flavors occurs multimodally. The thesis that some feature F, such as flavor, is type-multimodal, may be expressed by the following “type thesis”:
(Type thesis) Every perceptually represented instance of F is represented in virtue of the functioning of more than one perceptual modality.
Nevertheless, authors who propose that flavors are type-multimodal usually also postulate an additional thesis regarding the perceptual non-reducibility of flavor perception. It is claimed that contents of flavor experiences are not merely “collections,” “sums,” or “unstructured mixtures” of unimodal or instance-multimodal contents, but constitute an “organic unity” or a “structurally complex property” (see Connolly, Reference Connolly and Brown2014; O’Callaghan, Reference O’Callaghan, Stokes, Matthen and Biggs2015a, Reference O’Callaghan2015b, Reference O’Callaghan2017). This additional postulate is necessary if one wants to assert that flavor perception goes beyond minimal multimodality and instance-multimodality (see Briscoe, Reference Briscoe2021). Otherwise, flavor experiences are simply another example of experiences whose contents can be characterized in terms of combinations of unimodal or instance-multimodal contents. The non-reducibility requirement can be expressed by using the following “non-reducibility thesis”:
(Non-reducibility thesis) A content of a perceptual experience representing F is not merely a combination of unimodal or instance-multimodal contents.
While it is common to accept both the type thesis and the non-reducibility thesis, I will develop a different position. I assume the type thesis in the context of flavor perception, but I argue that despite a common belief, there is no particularly strong evidence to support the non-reducibility thesis. I demonstrate this by showing that there is a reductive account of flavor experiences that is consistent with the philosophical and empirical evidence for the multimodality of flavor perception and provides a genuine alternative to the position that attributes non-reductive type-multimodality to contents of flavor experiences. According to this reductive account, contents of flavor experiences are combinations of unimodal and instance-multimodal contents. In other words, I argue that philosophers may be too quick to accept both the type thesis and the non-reducibility thesis in the context of flavor perception, since there is at least an equally strong reductive alternative.
This is not to say that the type thesis is uncontroversial. In particular, it has been proposed that flavor is a unimodal feature represented by a unified flavor perceptual system (see Macpherson, Reference Macpherson2011; Matthen, Reference Matthen and Matthen2015; Mizrahi, Reference Mizrahi2017; Richardson, Reference Richardson2013 for variants of this approach), or that there are more general difficulties that undermine attempts to determine whether features are unimodal or multimodal (see Wadle, Reference Wadle2021), However, despite these difficulties, the type thesis is often endorsed and constitutes an influential position in debates on flavor perception (Briscoe, Reference Briscoe2021; Connolly, Reference Connolly and Brown2014; O’Callaghan, Reference O’Callaghan, Stokes, Matthen and Biggs2015a, Reference O’Callaghan2015b, Reference O’Callaghan2017). In consequence, for the purpose of the paper I will assume that there is no separate flavor perceptual modality and that the flavor experiences are multimodal, as this is an assumption which is shared by both my reductive approach and the usual approach, which accepts both the type thesis and the non-reducibility thesis.
It should be noted that my approach does not entail that flavor experiences are only minimally multimodal. In fact, I propose that flavor experiences not only possess unimodal contents, but also instance-multimodal contents, and they do not represent merely a co-occurrence of features, but a relational combination of features. In the above perspective, flavors are type-multimodal as they are always represented by the joint functioning of many modalities.Footnote 1 However, they are not non-reductively type-multimodal, because they are represented by a combination of unimodal and instance-multimodal contents. In consequence, flavor experiences are not merely minimally multimodal as they do not only have unimodal contents (they also have instance-multimodal contents), they do not merely represent co-occurrence of elements (they represent relational combinations of elements), and their contents cannot be divided without reminder among modalities (they involve instance-multimodal contents which, in the context of flavor perception, require more than one modality to occur).
In this paper, I will follow the usual representationalist position, where content specifies what is represented by a perceptual state. In addition, and following the mainstream representationalist conviction, I assume that content, at least in usual cases, determines the phenomenal character of experiences. Consequently, showing that a certain content is either unimodal or multimodal makes it plausible that the corresponding phenomenal character is also unimodal or multimodal. In consequence, and under the above assumptions, my considerations about experiential contents are also relevant for these positions, which characterize the multimodality of flavor experiences in phenomenal terms.
Of course, every approach to multimodality of flavor experiences has to assume a certain mapping between features represented in such experiences, and modalities that are able to represent them. Furthermore, I will rely on the mainstream approach to such mapping (expressed in influential works by Malika Auvray, Barry Smith, and Charles Spence), which are typical points of reference in philosophical debates on flavor perception (see Auvray & Spence, Reference Auvray and Spence2008; Spence et al., Reference Spence, Smith, Auvray, Stokes, Matthen and Biggs2014). According to this adopted approach, in typical flavor experiences, features are represented by gustation, retronasal olfaction, touch, thermal perception, and the trigeminal system. Gustation (the system processing information from chemical receptors in the oral cavity) is able to represent features such as sweet, salty, sour, bitter, and umami. Features such as “lemony,” “cherry,” or “coffee,” which often allow for the recognition of a flavor type, are represented through the functioning of retronasal olfaction, with gustation also playing a certain role, as will be discussed later. Intuitively, the tactile system represents features related to texture and hardness, while thermal sense represents temperature (although thermal stimulation can also lead to the representation of features usually associated with gustation, see Green & George, Reference Green and George2004). The trigeminal system (the system of facial nerves) is responsible for the perception of spiciness, and also the coolness evoked by certain stimuli, such as mint.
In the subsequent section (Section 2), I distinguish three possible ways in which contents of flavor experiences can be non-reductively type-multimodal: (a) by being simple contents, (b) by being complex contents with a non-reductively type-multimodal element representing a component of a flavor, and (c) by being complex contents with a non-reductively type-multimodal element representing a mode of combination unifying a flavor. In Sections 3–5, I discuss these options in turn and argue that none of them has a strong justification. Consequently, I propose that it is plausible to characterize flavor experiences in terms of combinations of, at most, instance-multimodal contents.
2. Non-Reducibility Thesis
According to the non-reducibility thesis, perceptual states representing flavors are not merely states having unimodal or instance-multimodal contents. In consequence, the multimodality of content exhibited by the flavor experiences is of a different kind than the multimodality of content of the majority of multimodal experiences, which can be analyzed in terms of unimodal and instance-multimodal contents. In philosophical works, three types of multimodal experiences have attracted the greatest attention: minimally multimodal experiences, experiences involving multimodal calibration, and experiences representing multimodal relations. Below, I show that neither of these experiences satisfies the non-reducibility thesis. If experiences representing flavors can be also adequately characterized in such terms, then their multimodality is not distinct from the multimodality of usual multimodal experiences, whose content is not multimodal in a non-reductive way.
It is clear that the minimally multimodal experiences do not satisfy the non-reducibility thesis as their content is a combination of some unimodal contents (Bayne, Reference Bayne, Hill and Bennet2014; Connolly, Reference Connolly and Brown2014; Macpherson, Reference Macpherson2011; O’Callaghan, Reference O’Callaghan, Stokes, Matthen and Biggs2015a, Reference O’Callaghan2015b, Reference O’Callaghan2017). In consequence, their multimodality can be adequately characterized simply by referring to these unimodal contents and their conjunction.
The second type are experiences involving multimodal calibration in which, due to multimodal processing, some elements are not merely represented as co-occurring, but their qualitative character undergoes modifications (Briscoe, Reference Briscoe2016). Probably the most discussed example is the McGurk effect, in which the simultaneous visual representation of a facial expression pronouncing the syllable /ba/ and auditory representation of the syllable /ga/ causes a perception of the syllable /da/. In this situation, the perceptual system receives incongruent information through distinct sensory channels and resolves this conflict by representing the intermediate /da/ syllable. Sometimes, the McGurk effect is interpreted simply in terms of an experience representing a conjunction of unimodal, visual and auditory elements (see Connolly, Reference Connolly and Brown2014). However, there are also accounts that treat the McGurk effect as involving an instance-multimodal content (see Macpherson, Reference Macpherson2011). In particular, one may observe that while the syllable /da/ may be, in principle, represented unimodally in virtue of auditory mechanisms, in the particular circumstances leading to the McGurk effect, it is represented due to joint activities of both visual and auditory mechanisms.
The third category is constituted by experiences representing multimodal relations. The most widely discussed of such relations are spatiotemporal relations and instantiation relations (Briscoe, Reference Briscoe2017; Kubovy & Schutz, Reference Kubovy and Schutz2010; Macpherson, Reference Macpherson2011; O’Callaghan, Reference O’Callaghan, Bennett and Hill2014, Reference O’Callaghan, Stokes, Matthen and Biggs2015a, Reference O’Callaghan2015b, Reference O’Callaghan2017). In virtue of representing multimodal spatiotemporal relations, an auditory element may be represented as, for instance, occurring earlier than a visual element or as being positioned to the left of a visual element. Representing multimodal instantiation leads to experiences of multimodal binding in which a single element, such as an object or an event, is experienced as possessing features associated with distinct modalities.Footnote 2 For example, a dog may be experienced as being brown and as barking. In such cases, the relational element is represented by an instance-multimodal content, because spatiotemporal and instantiation relations may be represented due to the cooperation of mechanisms belonging to distinct modalities, for instance, when they combine visual and auditory elements, but they may also be represented in virtue of functioning of a single modality, for instance, when such relations combine merely visual elements.
In order to investigate whether, in opposition to the above typical experiences, the content of flavor experiences is non-reducibly type-multimodal, it is worth asking under what conditions the non-reducibility thesis would be true. First, the non-reducibility thesis is true if contents representing flavors are simple contents that are not a combination of some other contents. In this case, a flavor experience cannot be analyzed in terms of an occurrence of a combination of unimodal or instance-multimodal contents. If one accepts that flavors are type-multimodal, i.e., they cannot be represented by any single modality, then such simple contents representing flavors would be non-reducibly type-multimodal contents.
Second, it may be the case that flavors are represented by complex contents whose elements represent various components of flavors, but one of elements of such complex content is non-reducibly type-multimodal content. In this case, the non-reducibility thesis is true because flavor experiences must be analyzed in terms of having at least one element of content that is non-reducibly type-multimodal.
The third option concerns not the contents representing components of flavors, but rather contents representing the ways in which components of flavors are combined. It may be the case that even if all components of flavors are represented by unimodal or instance-multimodal contents, there is also a non-reducible type-multimodal content that represents a relation that combines components of flavors into unified wholes. This also makes the non-reducibility thesis true, as in this case the flavor experiences have a non-reducibly type-multimodal content representing a certain mode of combination. In the subsequent sections, I discuss the above variants and argue that none of them has a stronger justification than the reductive alternative.
3. Simple Flavor Contents
The idea that contents representing flavors are simple is not particularly popular among philosophers of perception, as flavors seem to be experienced as constituted by a variety of qualitative components (see Briscoe, Reference Briscoe2021; Macpherson, Reference Macpherson2011; O’Callaghan, Reference O’Callaghan, Stokes, Matthen and Biggs2015a). I believe that such doubts are justified, and the complex character of contents representing flavors can be demonstrated by comparing contents representing flavors with contents representing colors—which are usually treated as paradigmatic examples of simple contents—and contents representing shapes, which are commonly interpreted as complex contents.
The simplicity of color-contents does not mean that colors are not experienced as having any components, which can vary independently from each other. In particular, colors are usually understood as combinations of hue, saturation, and lightness (see Palmer, Reference Palmer1999, pp. 97–99). In perceptual experiences, one can distinguish that two colors have the same hue or that a saturation of a color changes in time while other aspects remain constant. However, the experiential components of colors are not “detachable” in the sense that they are not experienced without being experienced as components of some color. For instance, one does not experience a red hue without any saturation and lightness, or pure saturation without any hue.
The lack of detachability differentiates experiences of colors from experiences of features such as visual shapes (or at least from some of the visual shapes). Shapes are visually experienced as having parts such as lines, vertices, or simpler shapes. In contrast to elements of colors, many parts of shapes are detachable. A line, for instance, demarcating where a shape ends, can be experienced as a component of some shape, but can also be experienced as a separate visual element without being represented as constituting some more complex whole. Similarly, if a shape is composed of a square part and a circular part, each of these parts can also be experienced without standing in a parthood relation to a larger visual element.
While I do not attempt to formulate a definition of complex perceptual contents, I believe that a comparison between colors and shapes provides a plausible sufficient condition of complexity: if a perceptual feature has some detachable components, it is represented by a complex content. Given this criterion, it is clear that flavors are represented by complex contents, because various elements of flavors are such that they may be also experienced separately, without being represented as constituents of any flavor. In particular, this seems plausible about tactile, thermal, trigeminal, and gustatory elements that can be evoked by stimulating the corresponding sensory mechanism even outside the context of the multimodal flavor perception. To deny this, one would have to demonstrate that it is impossible to, for instance, experience the same level of tactile hardness when eating some food and when chewing some tasteless item, or that tasty soup and tasteless liquid cannot be represented as being equally warm.
The above points are true even if one accepts that chemical senses use a representational format by which chemical features are experienced in a synthetic way, without experiencing them as composed of simpler components (see Young, Reference Young2015, Reference Young2024, pp. 73–79). Indeed, classical studies on the orthonasal perception of odor mixtures show that the ability to recognize the components that make up complex olfactory stimuli is rather limited (e.g., Livermore & Laing, Reference Livermore and Laing1998). However, even if such a synthetic mode is dominant in the chemical senses, this does not mean that contents representing flavors are simple. Flavors are also experienced as having non-chemical components, such as tactile and thermal, and it is well recognized that people are able to distinguish tactile and gustatory components of flavor experiences (see Stevenson, Reference Stevenson2009, pp. 115–138).
4. Type-Multimodal Components of Flavors
Even if contents representing flavors are not simple, the non-reducibility thesis can still be true if every flavor is represented by a content containing an element that is non-reducibly type-multimodal. It is easy to see that this is not true for many of the elements of flavor-related contents. As noted in the previous section, tactile, trigeminal, gustatory, and possible auditory and visual elements of contents representing flavors can also occur outside the context of flavor perception. Consequently, they are at most instance-multimodal contents. It may be asked, however, whether the same is true about contents representing features such as “lemony” or “cherry” that seem important in determining the character of flavor experiences, but cannot be straightforwardly ascribed to any single perceptual modality. Due to the fact that such features usually refer to a category of food, like lemon in case of “lemony,” later I name them category-related features.
In psychological literature, it is generally accepted that activities of retronasal olfaction are crucial for representing category-related features such as “lemony” (see Auvray & Spence, Reference Auvray and Spence2008; Spence, Reference Spence2014). In consequence, it may seem that contents representing category-related features are simply unimodal contents that occur due to the activity of retronasal olfactory mechanisms. These are mechanisms that process chemical stimuli arriving at olfactory receptors from the oral cavity and contribute to flavor perception. They are, therefore, different from orthonasal mechanisms, in virtue of which smells in the external environment are felt.
Nevertheless, the situation regarding contents representing category-related features is more complicated. Empirical studies suggest that the mere retronasal activation of olfactory receptors does not lead to a perception of flavor-related features attributed to items in the mouth. What is actually required is a spatiotemporally and qualitatively congruent activation of olfactory and gustatory receptors (see Lim & Green, Reference Lim and Green2008; Lim & Johnson, Reference Lim and Johnson2011, Reference Lim and Johnson2012; Spence, Reference Spence2022). In consequence, one may propose that contents representing category-related features are type-multimodal as they always occur in virtue of the joint functioning of more than one sensory modality.
However, proponents of a reductive approach to flavors may resist this conclusion by pointing out that contents that represent category-related features occur not only in flavor experiences, but also in virtue of orthonasal olfaction. For instance, we are able not only to attribute the “lemony” feature to the eaten food, but also to experience that there is a lemony smell around us. If category-related features that are represented in flavor experiences are also represented by orthonasal olfaction, then contents representing category-related features are not type-multimodal but instance-multimodal. This is because such category-related features are represented both multimodally (in flavor perception) and unimodally (in orthonasal olfaction). It should be noted that this argument is independent of whether orthonasal and retronasal olfaction are two distinct perceptual modalities or two parts of a single olfactory modality (see Smith, Reference Smith and Matthen2015; Wilson, Reference Wilson2021). In the first case, the contents representing category-related features are instance-multimodal: they occur both multimodally (in flavor perception, by using retronasal olfaction) and unimodally (in orthonasal olfaction). If the second option is true, contents representing category-related features are also instance-multimodal: the single olfactory modality can generate them together with other modalities (in flavor perception) and unimodally (by using mechanisms of orthonasal olfaction).
I believe that (a) there are prima facie reasons for the claim that category-related features represented in flavor experiences are also represented orthonasally, and (b) arguments against this claim are not particularly strong. In particular, I believe that there are three reasons why it is likely that category-related features represented in flavor perception are also represented by orthonasal olfaction. First, both flavor perception and orthonasal olfaction share the fundamental mechanism for representing category-related properties: such properties are represented by patterns of cell activation in the olfactory bulb, which receives input from chemical receptors in the nasal cavity. A common conviction in the literature on olfaction is that different activation patterns serve as “templates,” which allow for the recognition of category-related features such as “lemony” or “fishy” (e.g., Batty, Reference Batty, Biggs, Stokes and Matthen2014; Richardson, Reference Richardson, Young and Keller2023; Wilson & Stevenson, Reference Wilson and Stevenson2007; Young, Reference Young2024, pp. 74–76). Because the major mechanism for representing category-related features is shared between flavor perception and orthonasal olfaction, it is likely that orthonasal olfaction represents category-related properties that are represented in flavor perception.
Second, behavioral results show that presenting the same stimulus both orthonasally and retronasally is able to enhance the ability to identify category-related features. However, presenting incongruent stimuli through each route can impair identification, especially in orthonasal experiences (see Sun & Halpern, Reference Sun and Halpern2005). This suggests that both orthonasal mechanisms and retronasal mechanisms, which are crucial for flavor perception, aim to represent the same category-related features of stimuli. If each sensory channel carries the same information about the relevant feature, then identification becomes easier, while incongruent information makes it more difficult to identify which feature is actually present. If flavor perception and retronasal olfaction represented distinct category-related features, it would be more difficult to explain why information about such features from orthonasal and retronasal channels interferes with the other, as each of them could simply inform about a distinct feature.
Third, from the functional perspective, the mechanisms responsible for representing category-related properties in flavor perception and orthonasal olfaction have the same goal: to inform about the chemical composition of the stimulus and its relevance for the organism. Given the fact that flavor perception and orthonasal perception are closely related—sniffing often precedes eating and allows one to decide whether eating something is the right decision—the category-related features relevant for flavor perception are also relevant from the perspective of orthonasal olfaction.
While the above considerations show that there are serious prima facie reasons for believing that category-related properties represented in flavor perception are also represented by orthonasal olfaction, these are defeasible reasons that may potentially be undermined by further argumentation. In consequence, it is necessary to investigate why one may want to deny that category-related features represented in flavor experiences are also represented orthonasally.
First, one may argue that despite the common mechanisms in the olfactory bulb, further differences in the processing of chemical stimuli in flavor perception and orthonasal perception justify the claim that multimodal flavor perception and unimodal orthonasal perception do not share contents regarding category-related features (see Goldberg et al., Reference Goldberg, Wang, Goldberg and Aliani2018; Hummel et al., Reference Hummel, Heilmann, Landis, Reden, Frasnelli, Small and Gerber2006; Young, Reference Young, Mroczko–Wasowicz and Grush2023, for reviews). In particular, there are differences in cortical processing between orthonasal and retronasal stimulation. For instance, it has been shown that when an olfactory stimulus is presented together with a gustatory stimulus, there are differences in neural activity in structures such as the anterior cingulate cortex, depending on whether the olfactory stimulus is presented ortho- or retronasally (see Blankenship et al., Reference Blankenship, Grigorova, Katz and Maier2019; Small et al., Reference Small, Gerber, Mak and Hummel2005; Welge–Lüssena et al., Reference Welge–Lüssena, Husnera, Wolfensberger and Hummel2009).
However, the mere fact that the same chemical stimulus is often processed differently when administered orthonasally or through the oral cavity is not sufficient to conclude that category-related features represented in flavor perception are not also represented orthonasally. In particular, the differences in processing may not reflect differences in the features represented, but other facts, for instance, that orthonasal olfaction represents a category-related feature as instantiated in the space surrounding the subject, while flavor perception represents it as instantiated together with other flavor-related features possessed by an object positioned in the mouth. Furthermore, the suggestion that category-related features represented by flavor perception are also represented by orthonasal olfaction does not require that in both cases the same features are represented when the same stimulus is provided. For the proposal in question, it is sufficient that if a category-related feature is represented in flavor perception when the stimulus S is provided, this feature can also be represented by orthonasal olfaction when a stimulus is presented that may be different from S. For example, it may be the case that when a certain substance M is given as a stimulus, only multimodal flavor perception represents the “vanilla” category-related feature, but orthonasal olfaction can also represent the “vanilla” feature if the stimulus contains a higher concentration of M, or contains M combined with another substance.
Another attempt to reject the idea that multimodal flavor perception and orthonasal olfaction represent the same category-related features may rely on the observation that there seem to be significant phenomenal differences between representing category-related features in flavor perception and orthonasal olfaction. In particular, it is observed that the same stimulus—for instance, coffee or cheese—may evoke phenomenally distinct experiences when attributed orthonasally rather through the oral cavity (see Auvray & Spence, Reference Auvray and Spence2008; Smith, Reference Smith and Matthen2015; Young, Reference Young2024, p. 48). It may be proposed that such phenomenal differences undermine the claim that the same contents representing category-related features are generated both multimodally, in flavor perception, and unimodally, by orthonasal olfaction. Differences in phenomenal character are commonly accompanied by differences in experiential content, and their presence may suggest that flavor perception and orthonasal olfaction represent distinct chemical features when processing the same stimulus.
However, a proponent of the reductive approach has alternative and plausible ways of explaining phenomenal differences between flavor perception and orthonasal olfaction that do not require the assumption that distinct contents representing category-related features occur in each case. First, based on the empirical state of the art, it is difficult to conclusively determine how significant the phenomenal differences between flavor and orthonasal experiences are. The classic studies by Rozin (Reference Rozin1982) showed that people have difficulty recognizing retronasally presented stimuli if they were previously presented only orthonasally. While later studies confirm that retronasal identification is worse than orthonasal identification (e.g., Halpern, Reference Halpern2004; Puttanniah & Halpern, Reference Puttanniah and Halpern2001), people have the ability to recognize the same stimulus both orthonasally and when attributed through the oral cavity. For example, in the study by Sun and Halpern (Reference Sun and Halpern2005), participants had near perfect accuracy in recognizing orthonasally presented stimuli such as cinnamon or strawberry, while retronasal accuracy in recognizing the same stimuli ranged from 70 to 100%. Such quantitative differences do not support a strong claim that category-related features represented in flavor perception are not represented orthonasally. Other reported that phenomenal differences between retronasal and orthonasal experience of the same stimulus concern the experienced location of a sensation, the stimulus threshold required to generate an experience, and the perceived intensity (see Diaz, Reference Diaz2004; Goldberg et al., Reference Goldberg, Wang, Goldberg and Aliani2018; Heilmann & Hummel, Reference Heilmann and Hummel2004; Hummel et al., Reference Hummel, Heilmann, Landis, Reden, Frasnelli, Small and Gerber2006; Small et al., Reference Small, Gerber, Mak and Hummel2005). However, such results only show that in many contexts it may be more difficult to represent a category-related feature when it is processed by using retronasal olfaction than orthonasal olfaction, but they do not show that some category-related features represented in flavor perception are not represented by orthonasal olfaction. Consequently, it is not obvious that phenomenal differences between flavor perception and orthonasal olfaction are significant enough to provide support for the thesis that in each of these cases different contents representing category-related features occur.
Second, the same feature may be experienced in a phenomenally different way depending on what other features are represented together with it. For example, an orthonasal experience of cheese may represent only a category-related feature together with some additional olfactory features, whereas in a flavor experience a category-related feature is represented together with gustatory, tactile, and trigeminal features (see Wilson, Reference Wilson2021 on this point). Furthermore, the same stimulus may be experienced differently when represented orthonasally and in flavor perception not due to difference in contents representing category-related features, but because its relevance to the subject is different in each situation. In the case of the chemical senses, relevance is likely to relate to the possible positive or negative effects on an organism (see Castro & Seeley, Reference Castro and Seeley2014; Keller, Reference Keller2016; Martinez, Reference Martinez2015), and it may change from when something is merely present in the surrounding space to when it is about to be ingested. Finally, phenomenal differences between orthonasal and flavor perception may arise from distinct top-down influences from mental imagery and memory (see Spence, Reference Spence2023), and not from the fact that in each case different features are represented. Consequently, despite phenomenal differences, there is no strong justification that the same category-related features are not represented multimodally in flavor perception and unimodally in orthonasal olfaction.
In consequence, it is not easy to find conclusive evidence that contents representing flavors are non-reducibly type-multimodal in virtue of their constituting components, because these components, including category-related features, may plausibly be interpreted not as type-multimodal but, at most, as instance-multimodal. However, a proponent of the non-reductive approach may claim that the genuine type-multimodal features involved in flavor experiences are rather relational features concerning how various characteristics of eaten entities are combined. For instance, perhaps a non-reductively type-multimodal feature represented in flavor experiences concerns the fact that various retronasal, gustatory, tactile, and other characteristics are jointly attributed to an object in the mouth. In the subsequent section, I argue that this strategy is also not successful.
5. Unique Combination
As characterized in Section 2, the last way in which the non-reducibility thesis may turn out to be true is that experiences of flavors involve non-reducible type-multimodal contents representing how elements of flavors are combined. For instance, it is claimed that flavors are experienced as an “organic whole” and that they cannot be reduced to a conjunction or a co-instantiation of components (Briscoe, Reference Briscoe2021; Connolly, Reference Connolly and Brown2014; O’Callaghan, Reference O’Callaghan, Stokes, Matthen and Biggs2015a, Reference O’Callaghan2015b, Reference O’Callaghan2017). Such statements may be interpreted as suggesting that the source of non-reducible type-multimodality of flavor-related contents lies in contents representing the way in which components of flavors are combined. Further, I name the ways in which components of flavors can be combined “perceptual modes of combination.”
In order to properly investigate whether the experiences of flavors involve type-multimodal mode of combination, it is useful to consider the common unimodal types of perceptual modes of combinations. If the way in which the structure of flavors is represented can be analyzed in terms of such modes of combination, then contents representing flavors are, at most, instance-multimodal. Below, I discuss four major perceptual modes of combination: parthood, grouping, binding, and infusion. To avoid a possible terminological confusion, it should be noted that by modes of perceptual combination I mean the ways in which perceived components are represented as combined (e.g., that one component is a part of another) and not processes that lead to experiences of combined components (e.g., processes in virtue of which objects are experiences as divided into parts). Sometimes, the same term is used in both cases. For instance, “binding” may describe that an object is experienced as combined with several features (e.g., O’Callaghan, Reference O’Callaghan2017) and also as the name of a process in virtue of which features are attributed to a single object (e.g., Treisman, Reference Treisman1998).
One such common mode of perceptual combination consists of representing some components as parts of a larger whole. For instance, in vision several surfaces may be perceived as parts of an object and in audition some sounds may be perceived as a melody. The parthood is usually represented when the combined components stand in some relation of spatial or temporal connectedness but are separated by the presence of some qualitative discontinuities, demarcating where one part ends and another starts (see Hoffman & Richards, Reference Hoffman and Richards1984; Palmer & Rock, Reference Palmer and Rock1994; Xu & Singh, Reference Xu and Singh2002). On the other hand, perceiving components as combined by parthood does not significantly rely on the presence of similarity between elements. For instance, a visual object may be composed of surfaces with various shapes and colors.
The second common perceptual mode of combination is grouping. A perceptual group is usually composed of disjoint but spatially or temporally proximal components. In opposition to combination by parthood, experiencing objects as composing a group does not require the presence of connectedness and edges between combined elements. On the other hand, grouping strongly relies on similarity, as components are usually grouped in virtue of sharing some quality like color, shape, size, or a common pattern of motion (see Elder and Goldberg Reference Elder and Goldberg2002; Kubovy et al. Reference Kubovy, Holcombe and Wagemans1998).
Both combination by parthood and grouping allow for a certain independence of components from the whole they constitute. For instance, parts of a visual object can also be represented separately, without being represented as being parts of some whole. The same is true about grouped components, which can be perceived also as ungrouped. Such independence is not present in the third common mode of perceptual combination: binding, which combines features with an individual, like an object or event, that instantiates those features (see Briscoe, Reference Briscoe2017; Kubovy & Schutz, Reference Kubovy and Schutz2010; Macpherson, Reference Macpherson2011; O’Callaghan, Reference O’Callaghan, Bennett and Hill2014, Reference O’Callaghan, Stokes, Matthen and Biggs2015a, Reference O’Callaghan2015b, Reference O’Callaghan2017). For instance, seeing an object as being red or hearing that a sound is loud are examples of unimodal binding. In binding, a feature depends on an individual in the sense that a feature, such as redness or certain pitch, is only represented as bound with some individual. Perceiving multiple features as bound with a single individual, for instance, when seeing that an object is both red and round, does not require the presence of similarity or edges between bound components. However, due to the fact that bound features are represented as characterizing a single individual, they are also represented as being spatiotemporally co-located.
Binding concerns the relationship between an individual and its features. However, perceptual binding is also usually associated with a mode of combination that concerns relations between features attributed to the same individual. Such a mode of combination, sometimes called “perceptual infusion” (see O’Dea, Reference O’Dea and Wright2008), makes it possible to perceive that bound features not only characterize the same individual, but also characterize each other. For example, when looking at a red circular object, it does not appear that shape and color are two separate, unrelated features. Rather, the redness is represented as having a circular shape and the shape is represented as red.
5.1. Flavor-related binding
The grouping mode of combination does not seem to adequately account for the organization of flavor experiences. In particular, flavors do not seem to be represented as composed of disjoint components that are unified by standing in some similarity relations. On the other hand, the components of flavors are likely to be experienced as organized by the binding relations. Gustatory, olfactory, tactile, trigeminal, and other components are represented as features characterizing a common object which, relying on the attributed features, is assessed in respect of whether it should be eaten (see Auvray & Spence, Reference Auvray and Spence2008). However, it should be asked whether binding, which seems to occur in the case of flavor experiences, is in fact an instance of the same mode of combination that occurs unimodally, for instance, in visual or auditory experiences. It may be the case that flavor-related binding is only superficially similar to unimodal binding, and that in fact, flavor-related binding is a distinct mode of combination. To investigate this issue, the question must be asked of what the possible reasons are for treating flavor-related binding as a different mode of combination than binding that occurs unimodally. There are two major reasons that would allow flavor-related binding to be distinguished as a distinct mode of combination. First, it may be the case that there are some important differences in properties of binding in flavor perception than in vision or audition. For instance, if flavor-related binding were constituted by symmetric relations and bindings occurring unimodally were constituted by asymmetric relations, that would strongly suggest that flavor binding is a distinct mode of combination. Second, even if flavor-related bindings have the same properties as unimodal bindings, it may still be the case that its function differs. While it is possible that distinct functions can be served by the same mode of combination in different contexts, one may still propose that a significant difference in function provides a reason for believing that the modes of combination are distinct.
Nevertheless, neither of these worries has a strong justification in the case of flavor-related binding. Unimodal, for instance, visual or auditory, bindings are asymmetric modes of combinations that connect features with an individual and are associated with certain dependency as, in normal cases, features are not represented without being represented as bound with some individual. These characteristics are also present in flavor-related binding. For instance, in visual binding, redness is attributed to an object, but not vice versa, and in flavor-related binding, a mintiness is asymmetrically attributed to the eaten object. In fact, it is quite implausible that binding could be a symmetric relation. Making it symmetric would, rather, transform it into an infusion relation (see Section 5.3). However, even if a symmetric account of binding can be developed, the binding that occurs in flavor experiences is not a symmetric relation.
Similarly, the characteristic dependency is present both in flavor-related binding and unimodal bindings. For example, a pitch is not normally represented without being represented as a feature of some sound, and elements attributed to eaten objects are not perceptually represented without being represented as instantiated by some individuals. This does not necessarily mean that these individuals have to be positioned in the oral cavity. For instance, outside the context of flavor perception, olfactory components can also be attributed to odors located around the subject. In addition, individuals instantiating flavor-related components do not have to be external objects. For example, it may be the case that some tactile and trigeminal components represented in flavor experiences may be attributed both to objects in the oral cavity and to fragments of one’s body. Nevertheless, these components are always, except for some possible cases of perceptual dysfunctions, represented as features bound to some individuals.
Furthermore, both flavor-related binding and unimodal bindings are associated with an experience of spatiotemporal overlap. For instance, in vision it is experienced that an object at a given moment is both red and round. The way in which components of flavor experiences are organized also exhibits the spatiotemporal co-location associated with binding, because components represented in flavor experiences are represented as simultaneously characterizing an object positioned in a specific place.
In addition, from the functional perspective, flavor-related binding and unimodal bindings seem to serve the same general purpose. In vision, audition, and flavor perception, binding serves a unificatory role; it consists of representing that various detected features characterize the same individual. For instance, visual binding represents that a square object is also red; auditory binding represents that a high-pitched sound is to the left; and when flavor binding occurs, it is experienced that the eaten thing which is warm is also bitter. Furthermore, in both flavor perception and other modalities, binding allows for object recognition. With binding operating in vision and audition, an object can be recognized as a dog or as a certain melody. In flavor perception, binding is also relevant for recognition. For instance, in order to recognize that something is fresh cherry fruit, or a cherry-flavored ice cream, or a cherry-flavored hard sweet, flavor perception has to differentiate between various combinations of gustatory, olfactory, tactile, and thermal features attributed to the eaten object.
The presence of both functional similarity and similarity of main characteristics of flavor-related binding and unimodal bindings makes it plausible that flavor-related binding is an instance of the same general mode of combination that also occurs unimodally. This is not to deny that the phenomenal character associated with flavor-related binding may be distinct from that associated with visual or auditory binding. However, as observed in the previous section, accounting for such differences does not require postulating that flavor-related binding and unimodal bindings are distinct relations. For instance, phenomenal differences may arise from the fact that in different contexts the same binding relation combines components associated with distinct modalities (e.g., it combines an object with two visual components or an object with an olfactory and a gustatory component).
5.2. Flavor-related parthood
In addition to binding, it is possible that there is some organization of entities represented in flavor experiences that can be accounted for in terms of “parthood.” For instance, a flavor experience may represent that there is a part of the food on the right side of the oral cavity that is spicier than the part on the left side. Similarly, some flavor experiences, such as those associated with drinking wine, may have temporal parts in which represented features of the liquid are changing.
However, there are no strong reasons to believe that parthood occurring in the context of flavor perception is a distinct mode of combination from parthood occurring in, for instance, visual or auditory perception. First, in the flavor experiences, parthood is an asymmetric relation. For example, an object positioned on the right side of the oral cavity may be represented as a part of the whole eaten thing, but not vice versa. Second, representing parthood in a flavor experience also requires the presence of some qualitative discontinuities. For instance, the right fragment of an item may be represented as a part of the eaten whole due to its higher level of sweetness than the left fragment. Third, the parthood occurring in flavor experiences is also characterized by a significant degree of independence of parts from the whole. This independence is larger than in the case of binding, and is similar to the independence characterizing parthood occurring in other modalities. For instance, there may be a flavor experience that represents an object having a sweet part in the right of the oral cavity and a non-sweet part in the left side. However, there is also a possible flavor experience that represents only a sweet object in the right side of the oral cavity, without representing it as composing a whole together with a non-sweet part to the left.
Furthermore, and also from the functional perspective, the parthood organization represented in flavor experiences does not seem to significantly differ from the parthood organization represented by other modalities. In all these cases, relations of parthood allow us to establish how various features are distributed within the structure of a perceived object and, in particular, where a region instantiating one feature ends, and a region instantiating a distinct feature starts.
5.3. Flavor-related infusion
Nevertheless, as pointed out by O’Callaghan (Reference O’Callaghan, Stokes, Matthen and Biggs2015a, Reference O’Callaghan2015b, Reference O’Callaghan2017), a flavor experience does not merely represent a co-instantiation of various features by an eaten object. The attributed features also seem to be combined with each other, such that they constitute some form of a qualitative whole. However, a proponent of the reductive approach may argue that this can be explained by postulating the presence of perceptual infusion relations between components of flavor experiences (see O’Dea, Reference O’Dea and Wright2008 for the introduction of this term). To plausibly develop such an option, two important worries have to be addressed. First, similar to the case of binding, one may propose that infusion relations present in flavor experiences are only superficially similar to those present in unimodal experiences, so flavor-related infusion is not instance-multimodal. Second, it may be the case that even if infusion itself is an instance-multimodal mode of combination, the features represented in flavor experiences are represented as combined in such a way that type-multimodal components are created.
My approach to the first issue is analogous to that applied in the case of binding, i.e., we should investigate whether flavor-related infusion and unimodal infusions have significantly similar properties, and function such that we may treat them as instances of the same mode of perceptual combination. O’Dea (Reference O’Dea and Wright2008) provides only a general description of infusion, according to which infused features characterize each other. This is satisfied both by the flavor-related infusion and unimodal infusions. In vision, for example, a color is experienced as having a particular shape, and a shape is experienced as being filled with a particular color. Similarly, in flavor perception, the olfactory fruitiness and gustatory sweetness are not experienced as disjoint features of the same entity. Rather, we feel that fruitiness and sweetness are phenomenally combined. Sweetness is not experienced as some pure gustatory sweetness co-occurring with some pure olfactory fruitiness, but rather sweetness is experienced as having some fruity characteristics (which are not present when sweetness is related to “coffee” olfactory elements), and fruitiness is experienced as having some sweet characteristics (which are not present when fruitiness is related to sour gustatory elements).
However, because such a description is very general, one may be concerned that there are some more subtle differences between flavor-related infusions and unimodal infusions, which provides reasons for treating flavor-related infusions as a distinct mode of combination. To investigate this, let us consider the more specific characteristics of the flavor-related infusion. First, flavor-related infusions sometimes not only combine features but also introduce a form of dependency. For instance, as argued in Section 4, in order to obtain a flavor experience, there has to be a certain congruency between gustatory, olfactory, and maybe also tactile components. In consequence, an infusion between, for instance, an olfactory component and a gustatory component is not contingent, but in a flavor experience, an olfactory component is always infused with a gustatory component. On the other hand, in the case of the flavor-related infusion, such dependencies do not occur between all types of components. For example, an olfactory component can be represented without being infused with a thermal component. Second, flavor-related infusion is sometimes not merely related to experiencing that features characterize each other, but as an effect of infusion some of the combined components are modified. Probably the best-known example is the sweetness enhancement effect that occurs when gustatory sweetness is combined with an olfactory vanilla component (e.g., Stevenson et al., Reference Stevenson, Prescott and Boakes1999).
However, these features also occur in the case of unimodal infusion. For example, the visual infusion is also associated with dependency between some types of visual element, but such dependency is not present in the case of other types of components. To illustrate this claim, let us consider a visual component such as color. It is not represented without being infused with some spatial component related to shape, but can be represented without being infused with visual motion. Similarly, also in the case of visual infusion, modifications of combined features are observed, for instance, regarding the interactions between color, shape, depth, and recognition-related features (e.g., Bloj et al., Reference Bloj, Kersten and Hurlbert1999; Egus, Reference Egus1983; Wesp et al., Reference Wesp, Peckyno, McCall and Peters2000). Furthermore, flavor-related infusion and unimodal infusion do not seem to differ significantly in their purpose. In both cases, the function of the considered modes of combination is to provide information about relations between features of a single individual.
Nevertheless, even if flavor-related infusion is merely instance-multimodal rather than type-multimodal, one may still argue that it gives rise to the type-multimodal content of flavor experiences. For instance, a representation of hardness infused with sweetness appears to arise from type-multimodal content, since no single sensory system is capable of representing such a combination of features. However, as stated in Section 2, to justify the non-reducibility thesis in the case of flavor perception, it is not sufficient to show that certain contents are type-multimodal: they must be non-reducibly type-multimodal. In fact, a proponent of the reductive approach is well positioned to reject the claim that contents representing the infusion of hardness and sweetness are non-reducible. More specifically, such a proponent may argue that representing an infusion of features A and B consists in representing flavor features A and B as standing in the infusion mode of combination. Given that contents representing flavor features A, B, and the infusion mode are all, at most, instance-multimodal (as argued in Sections 4 and 5), representing an infusion of features A and B is not non-reducibly type-multimodal. This is because the content representing such an infusion is a combination of contents that are, at most, instance-multimodal. However, representing an infusion between features A and B is not merely a case of minimal multimodality. What is represented is not simply the co-occurrence of feature A, feature B, and the infusion mode of combination, but rather features A and B as combined by the infusion mode.
One may still attempt to counter the reductive picture of infusion given above by postulating that there is simply no common relation of infusion represented in both unimodal and multimodal experiences, but rather that what is represented are simply unimodally or multimodally infused features.Footnote 3 In consequence, there may be an experience representing sweetness infused with softness, and an experience representing redness infused with squareness, but there is no common element, i.e., an instance-multimodal infusion relation, represented in both these experiences. Such an interpretation may be motivated by the belief that it is difficult to identify the common content of the infusion relation shared between experiences. For instance, it seems that there is some content associated with representing some element A as being to the left of element B, but which is independent of the properties of A and B. However, perhaps in the case of representing sweetness infused with softness, the whole content is determined by the characteristics of sweetness and softness without any additional content associated with the infusion itself.
Despite the above points, I believe that the reductive interpretation is more plausible. In particular, when an infusion is represented—say, between sweetness and softness—the mode of combination introduced by infusion is “analytic,” in the sense that, despite infusion, sweetness and softness are still distinguishable as separate features (unlike, for instance, from the case of obtaining the color orange by mixing red and yellow). These features may undergo some modification due to infusion, for instance, sweetness may become less intense, but they can still be identified in the experience.
In consequence, the question arises as to what differentiates the infusion of sweetness and softness from the mere co-occurrence of sweetness and softness, as in both cases, sweetness and softness are represented. My answer is that what differentiates these cases is the presence of an infusion relation, which can be generally described as a relation of mutual characterization between the two features. In other words, the infusion relation is an additional element that is not present in the case of the mere co-occurrence of sweetness and softness.
Is such a relation a common element that can be represented both in the case of multimodally infusing sweetness and softness, and also in the case of unimodally infusing redness and squareness? I believe it is because in each example, infusion is a relation in virtue of which properties characterize each other. This is a common mode of combination, occurring in various types of experiences, which differentiates infusion from the mere co-occurrence of features, as well as from different modes of combination like parthood or instantiation. According to such a view, it is not the case that when sweetness infused with softness is represented, the whole content is determined by the characteristics of sweetness and softness. This is because there is a part of the content specifically connected to the infusion relation that determines that the mode of combination between the features is the mode of mutual characterization.
6. Conclusions
The way in which components represented in flavor experiences are experienced as organized can be accounted for in terms of perceptual binding and infusion. While the exact pattern of relations may be more complex than, for instance, those in audiovisual experiences, there are no strong reasons to postulate that the occurrence of some additional mode of perceptual combination is required in order to account for the way in which flavors are represented. Such a result does not support the non-reducibility thesis, since the considered binding and infusion are instance-multimodal (and not type-multimodal) modes of perceptual combination. Furthermore, as argued in the earlier sections, the other ways of justifying the non-reducibility thesis are also without strong support: contents representing flavors are not simple, and there is not strong reason to characterize them as including non-reducible type-multimodal contents. Overall, flavor experiences can be plausibly analyzed in terms of, at most, instance-multimodal contents. Thus, while flavors may be type-multimodal features (i.e., if the type thesis is true), there is a plausible reductive approach to contents of flavor experiences that denies the non-reducibility thesis.
Acknowledgments
The author would like to thank the anonymous reviewers for their comments, which significantly improved the paper. The early version of the paper was presented during ECAP11 conference in Vienna.
Funding statement
The work was supported by the National Science Centre (grant 2018/31/D/HS1/00363).
Błażej Skrzypulec is an associate professor at Jagiellonian University. His work focuses on the philosophy of mind and perception. In particular, he is interested in the structural aspects of sensory experiences, especially non-visual and multimodal ones. He is currently working on a project concerning the unity of the structural experiential aspects in multimodal perception.
Open access
