Dena Shottenkirk, 2022.

My research focuses on the ways epistemology dovetails with aesthetics, with an emphasis on perception. Beginning with my 2005 PhD dissertation “The Effects of Nelson Goodman’s Nominalism on his Epistemology and Aesthetics”, which was subsequently published by Springer (2010) under the title Nelson Goodman: Nominalism and Its Aftermath, I have developed a body of research at the intersection of perception, epistemology, and aesthetics.

Coming from a background in the visual arts, both as a practicing artist and as an art critic, I was aware of the inadequacy of Goodman’s aesthetics. My original concern with Goodman was to draw out the causal relationships between his logic/nominalism, his epistemology, and his aesthetics. Until then, the disparate areas of his writing had been dealt with separately, and instead I demonstrated that his nominalism causally constrained what was available to him in his epistemology, which likewise constrained what was available in his subsequently developed aesthetics. For example, it was in his nominalism and his “calculus of individuals” wherein he defines qualia (presentation of color, time, and space) as the basic phenomenal unit – for him, a concrete particular; this sets the stage for him to forbid platonist properties and abstract objects, meaning accounts, classes, and fictive references. All these constraints are directly translated into his epistemology wherein his commitments involved denials of universal/objective truth, natural kinds, the autonomous object, and univocal human responses; an epistemology where empiricism is denied and coherentism embraced, and the relativistic worldmaking takes center stage.

The cumulative sum of his metaphysics and epistemology is seen in the aesthetics, which is a semantic account of reference. Let me lay out what I saw as the problem as that will explain the trajectory of my subsequent research. My basic criticism of Goodman’s aesthetics was founded in his central claim that we understand what the artwork is about once we have accurately ascertained the referential relations between the terms used to describe the artwork and what the terms denote or what the terms are denoted by.

But if I were told that that the gray painting of the sea is one that “does not denote the color gray but is denoted by the predicate ‘gray’” I would not necessarily have understood anything: knowing that others legitimately attribute “gray” or even “sadness” to the painting does not insure that the bearer of this information has the adequate intentional attitude, conscious awareness, or emotional frameworks that makes normal cognitive functions possible. Stating my point bluntly: predicate matching fails to explain people’s passion for art.

An aesthetics that forbids intensions, properties, fictive entities, non-semantic meaning, natural symbols, or a central role for emotion, is not an aesthetics that can satisfactorily explain the phenomenal and representational complexities of what it means to experience art. And that is what we are asking for: what is it to stand in front of an artwork and experience it? That experience is not reducible to the referential description after the experience. What we want to know is how to logically characterize the moment itself of experiencing the artwork.

This then led me to investigate the construction of meaning in aesthetics (in opposition to Goodman’s stricture on reference) and to research the social and communicative functions of art, bringing forward Goodman’s positive insight that aesthetics is part of epistemology. In 2009, I published a monograph on censorship in art (Cover Up the Dirty Parts!, Cambridge Scholars Press) wherein I examined how art gains meaning both for the individual viewer and for the collective group. I included not only the empirical history of censorship with a focus on the reactionary tactics of the Christian right, but the philosophical theories supporting the cognitive functions of free, uncensored speech within an arts context, including points of view such as J.S. Mill’s, and addressing concerns with the vagueness of terms such as “obscenity” by drawing from Wittgenstein’s writings on vagueness and rule following.

While a less technical book than the Goodman one, it did though incorporate Goodman’s notion of the projection of predicates (though in reverse!) and how that functions to determine meaning in a larger social context. For example, what now counts as emblematic of central European art of the 1930s is expressionism and surrealism: a time of focus on the subconscious and the Freudian and Jungian debate, a time of introspection, as well as a time of anger over the growing authoritarianism and racism in politics. But that projection of predicates (from the past into the present) was dictated in the decades after, not by the people of the 1930s; it was the art historians, the artists, and the larger public from the 1940s and 1950s who looked back on that earlier era and chose expressionist and surrealist artwork and their associated predicates as significant and thus emblematic of the era. To state it more directly: the (reverse) projection of predicates is those from the past pulled into the present from the innumerable ones that could have been pulled into the present, but it is now these particular chosen ones that are The Past. Pop art and Warhol (not conceptual art or minimalism) are the 1960s, because the present-day world needs those past predicates (slick, commercial, on top of the world, etc.) in order to give precursors to the present.

The governing issue here as in all my research is the argument that art is something that can only be understood within the framework of epistemology. The world at large – including the art that is found in it – comes at us with a plethora of data: massive bits of information, some of which is attentional and noticed consciously, some unconsciously, and much not noticed at all. We edit, we select. My paper “Global Grammar” (Polish Journal of Aesthetics, 2018) argued not only is art a part of epistemology, but that in our era after both modernism and post-modernism where art is practiced without the stylistic codifications found in those movements, it instead encapsulates the fully formed aesthetic vocabulary owned by the individual artist; in other words, meaning in art is now understood through an analysis of the symbol system and expressions of the individual and not the individual as representative of the larger symbol system adopted by the stylistic group.

This attention to the meaning inherent in an artwork, and its ability to be constitutive of social facts, is a part of my research that I continually revisit. I am concerned with how it is that art becomes an epistemic experience, and what precisely constitutes that experience. One way of stating it is to say that the individually experienced phenomena (e.g., the artist’s experiences) become part of socially constructed facts through, firstly, the translation of the artist’s experience into art, and secondly, the absorption of that art through the process of audience viewing and acceptance. But that is far from a complete story. Slightly more completely: the art (again, a translation of the artist’s experiences) is the catalyst in the viewer’s 1) naming of (the art as) objects, 2) expression of cognitive affordances, and 3) building of intentional beliefs; eventually, if that art is thought of as paradigmatic and thus incorporated more fully into the culture, those three catalyzed results feed into the construction of the social facts of the group. Thus, the non-linguistic experience of the artist is converted into the linguistic practice of the group. In this, listening to others is the most crucial epistemological act a person can do. And art is a form of listening. We pay attention—in that moment of experiencing an artwork—to the subjectivities and perceptions of another. We listen to their viewpoint, their truth, their experience. And we take what is useful, what seems uniquely true and previously unnoticed. We learn. We learn to see an object we’d not seen before and learn to care about that formerly unseen object. Art creates new objects, that are experienced by the viewer as now ontologically distinct entities with associated linguistic expressions.

Another way of thinking about it is to carve out the more general and related point (though potentially separating it from the social): How do we go from seeing to thinking? Or to phrase it more pointedly: How do we go from perception to what is taken to be the quintessential structure of thought e.g., concepts? The point here is the relationship between the moment of perception and what’s thought of as the process of cognition.

In pursuance of this question, I co-edited a volume titled Perception, Cognition, and Aesthetics (Routledge, 2019), wherein I included my article “Gist Perception”. In that article, I examined gist perception in contrast to object perception, rejecting a strict divide between perception and cognition. Philosophy has historically been focused on object recognition, particularly interpreted as an analysis of how objects become predicable for us. But if we examine the process of gist perception, a very different perspective opens up. Gist is holistic perception, an efficient system that has evolved for obvious evolutionary reasons. Gist experience – sometimes called “preattentive perception” shows great computational efficiency, and is often defined as the first 300 milliseconds (ms) of perception. But much happens even earlier. At 50ms subjects were able to reliably discern unity or order (e.g., horizontals, verticals, grids, stable patterns, etc.) after only a single glance (Schwabe et al., 2018:18). In addition, subjects are unlikely to change their views regarding that particular kind of judgment upon longer or repeated gist experience. It is a stable judgment, and research has repeatedly confirmed that our bodies are quite good at assessing this particular facet of the scene. One has to wonder why? And the answer is obvious: we value order, as disorder is often a source of threat. Order means safety. That is one reason that gist is important: because it registers our subjective reactions to the data. It gives one’s own bodily reaction to that fact: Like or dislike / yes or no / desire or aversion; these are all dichotomous choices that reveal our assessment of our relative safety to our surroundings. It is our bodies’ reading of the environment and is an essential ingredient to the emotional component of experience. Value is not “added on” to empirical experience; it’s an integral part of it.

But there are other reasons gist is important, and these revolve around issues of art as an epistemic experience. In short, there isn’t a good account of gist experience within aesthetics. Let me try to present the problem and thereby motivate the search for a solution.

Perception is always defined as the act of being aware of something through the senses, and generally it begins at the point where it’s seen as an openness to the world. In this, it is often thought that it commits one to describing the objects and their properties, e.g., the red apple, and explaining how we come to know that. In this article, I argued that to do an epistemological analysis of how-we-know-the-world with the construct that one starts from the antecedently existing object and then traces how that object becomes an instance of experience to the perceiver, is surely to misconceive the relationship. What we need is not an analysis of how objects come to be predicable information for us but how it is that we take our experiences – primarily our gist experiences – and convert that into named entities. And that is largely a social process. For the existence of an object – a named entity – assumes a substratum of past sensorial experiences by many individuals who, through their cumulative and shared articulations of that seen, sensorial experience. In other words, tracking the experience as it becomes a public object.

The having of a gist experience is the first stop our bodies make in the interaction with the external world. It is the first move in the editing process. I include in the process both the existence of nonconscious perception as well as nonconceptual perception. As an essential move in the editing process, gist gives us the basis that we use when we move to the creation/naming of objects. We negotiate entities, we decide to call one set of data “an object” – e.g., a particular name, and decide that politically and contentiously. Not all is seen or recognized. Some bits of perceptual data are consciously recognized by us, while others are experienced only nonconsciously, and other bits of data are completely not experienced at all. This triage is the consequence of habitual readings of external reality, social fact recognition and the biases that trail as a consequence of what we allow ourselves to see and recognize and what we filter out. Gist is what we use to make objects.

My recent work continues at this intersection of epistemology, aesthetics, and perception. In a paper titled “Aesthetics from the Visual Artists’ Viewpoint” (Contemporary Aesthetics, 2021), I also argued for rejecting a strict divide between perception and cognition, but in the instance of this paper I switched from the viewpoint of the viewer to the viewpoint of the artist, using the philosophy of Gareth Evans to address criticisms made of a well-known position in perception/aesthetic theory (e.g., Dustin Stokes’ criticism of Bence Nanay). In giving an analysis of the difference between an aesthetic perception and a nonaesthetic experience, I analyze how an artist perceives using low-level features and contrast that with the nonaesthetic perception involved in attributing a property to an object. Low-level features are those basic, and often nonconceptual, parts of initial gist perception such as color, orderliness, contours, shapes, textures, contrast, illumination, etc. The problem of how to characterize aesthetics is thus solved by looking at three things: 1) recent research into gist perception, 2) Gareth Evans’ notion of nonconceptual information, and most importantly, 3) the way artists rely on low-level features in visual experience in looking at the world.

This reliance on low-level features is an important illustration of how an analysis of perception can proceed without resorting to a description based on objects. While not conflating the distinction between conceptual and nonconceptual with the distinction between low-level and high-level (e.g., semantic) features, it is though uncontentious to point out that high-level semantic perceptual features are defined in terms of the application of governing concepts (it is an “apple”), and it is the latter’s function vis-à-vis general terms that gives high-level features their ability to identify objects as F of x. I draw on Evans’ The Varieties of Reference to address concerns of representation and semantics as he significantly emphasized the role of sensing, most pertinently in the oft-quoted footnote: “…the senses yield non-conceptual information, whereas language embodies conceptual information…” Evans’ view is that objects are not only perceived in terms of being an instance of an named category – “apple” – but in addition, he instead proposed that part of the information that we receive from perceptual experience comes from sensing and that that is non-semantic. That seems right to me.

The continued concern with the distinction between high-level and low-level features and the role of gist is seen in another paper I published titled “A Tale of Two Reds” (Erkenntnis, 2021). The article began with a story-telling device contrasting two tales: one of looking for red clothes in a closet, and the other seeing the color red in a gallery. Relying heavily on research in vision science, I again addressed Dustin Stokes’ criticisms of Bence Nanay as a way of expanding the concerns into the larger realm of how viewers (not artists, this time) perceive properties/predicates in the two different contexts of aesthetic versus non-aesthetic experience. The discussion centers around notions of object, property attribution, distributed and non-distributed attention, and finally on our reliance on low-level features in the viewing of artworks.

To fill in a few blanks here in regard to those distinctions: Distributed attention points to a way of gaining perceptual information that is not rooted in object-recognition or in the process of foveating on a particular entity. To distribute one’s attention is to look broadly and at global features, and though Nanay does not discuss gist perception, he at least seems aware of perception that focuses on global features. Thus, according to Nanay, aesthetic perception is focused attention on an object and distributed attention to properties. He distinguishes this from expert visual tasks, which are distributed with regards to objects and focused with regard to properties. An example of the latter would be identifying the blue socks from an array of different socks.

The literature on Nanay has generated some concerns about whether or not the expert visual task paradigm is properly distinguishable from the aesthetic. Space doesn’t allow a reiteration of the specifics of the criticism leveled by Stokes (given though in detail in my article) but the outlines are as follows: Stokes argues that according to the parameters laid out by Nanay himself, expert visual tasks are just like aesthetic experiences – both focused attention on an object and distributed over properties; thus, a car mechanic is then doing the same thing as someone looking at a painting. Herein, of course, seems to lie a real problem.

For we are instinctively drawn to ask: do we attend to properties—the multiple tokens of the type red—in, for example, a gallery setting in the same way that we attend to the properties of red tokens while looking for clothes in the closet? To be more specific about the problem: What does it mean to attend to properties?

In answer to this problem I argue that there are two different kinds of distributed attention: (1) property attribution, and (2) gist perception. More specifically, my claim is that distributed attention is experienced differently in the different contexts of expert visual task and aesthetic experience. In the visual task, the red was targeted, named, and linked to that object. I demonstrate that property attribution exists in the expert visual task context in the application of the type/token distinction regarding an instance of that property, which will thus make it a distinctive kind of distributed attention. On the other hand, this is distinct from an instance where one is merely aware of the predicate red; this happens in gist perception. By examining gist perception, I show firstly the role of low-level features in all perceptual processes, and secondly the special role that low-level features play in the distinctive kind of distributed attention that we see in the aesthetic experience. An example of the latter would be looking at a beautiful tree that has red leaves. If queried later regarding the color of the tree, one would be able to say “It was red”, but that is not to say that in the exact moment of standing in front of the tree, that one was—in a manner identical with looking for red clothes—conscious of the red as being a property of the tree. One would not being saying, “oh, that leaf is red, and there’s another red leaf.” Instead, one would be scanning the low-level features of color, texture, form, etc. Thus, it does not seem plausible that we perform, in aesthetics, the same task of finding and checking the presence of property attributes in the object. I did not, in the gallery, go around the room and say, “There’s a red, there’s another red, and here is yet another one! These two match!” We are not looking for and semantically attributing properties.

The science behind all of this is important. Our eyes are a mechanical, computational mechanism that absorb the external world through a series of sampled episodes driven by interposed and extremely fast saccadic scans. The samplings occur at 3–4 per second; this makes for the obvious fact that a large proportion of our visual surroundings are not processed by us (e.g. Ballard et al. 1992; Land and Hayhoe 2001; Land et al. 1999; as cited Tatler et al. 2005, 643). What distinguishes the saccade target selections from those that are not targeted? Beginning in the 1980s, many studies began to appear that characterized visual perception as two-stage process: the first stage of “preattentive” mode, in which simple low-level features were rapidly processed in parallel over the entire field, and a second “attentive” mode, where the focus was to particular locations in the visual field (Koch and Ullman 1987, 115).

There are also a significant number of researchers who have emphasized the bottom-up role of “salience maps”, and the activity of low-level features in those salience determinations (Itti and Koch, 2000; Itti et al. 1998; Olshausen et al. 1993; Parkhurst et al. 2002). This gives us a model of vision that is initiated by a bottom-up oculomotor system governed by relatively uniform standard of saliency that is then, in both a feedforward and feedback loop, interactive with top-down object-recognition processes. This has obvious evolutionary advantages, as we need an interpretative map of the world that can quickly prune incoming sensory input, and the bottom-up processes do exactly that. It also provides a role for decision and top-down cognitive phenomenology.

Research has also shown the process to be iterative, particularly in the aesthetic context. Evidence for this kind of iterative process is supported by research that has shown that “When gist information in a painting is deemed to have sufficient interest to an observer, the second stage of aesthetic processing ensues” (Locher 2015, 78). I thus argue in this paper that iterative gist experience is the distinctive characteristic of the aesthetic experience. We glance quickly, and glance again with repeated saccadic trajectories, assessing and re-assessing the experience. Thus, aesthetic perception is a staged-process, or experiences of several different temporal events of separate gist scannings, significantly constituted by low-level features.

Finally, a note about the future: these last several articles will reappear in reconstituted and newly renovated form in my book Art as Cognition: Framing the Aesthetic Experience as Conversation (Springer, forthcoming). I finalize my initial complaint regarding Goodman: No. This is how we logically characterize the moment itself of experiencing an artwork.