The notion “adaptation”, being central to the target article, while used comparably to analogous work on perceptual effects of systematic alteration of sensory input, does not obviously correspond to the unambiguous physiological notion of adaptation, i.e., a decrease over time in the responsiveness of sensory receptors to changed, constantly applied environmental conditions (e.g., Held 1965; Noguchi et al. 2004; Smithson 2005). Distinguishing semantic adaptation (a remapping of color terms and building immediate semantic connections to their proper object referents) from phenomenal adaption, Grush et al. focus on phenomenal aspects of regaining both stimulus constancy and original color arrangement in spite of changes in input. Given that adaptation to numerous alterations in visual input has already been reported in various studies (Kohler 1962, 1963; Anstis 1992; Heuer & Hegele 2008), Grush and colleagues also hypothesize the possibility of some form of adaptation to a version of the color-inverted spectrum. They designed a series of experiments to assess phenomenal adaptation of visual experience under color rotation by 120°, which leads to “tomatoes […] causing red qualia again, even if the subject is wearing the rotation gear” (Grush et al. this collection). The definition of “phenomenal adaptation” in the target article is “a return to normalcy” and “a gaining of color constancy under rotation”. Phenomenal adaptation then, can be understood as the regaining of phenomenal qualities of the pre-rotated color experience while using the rotation equipment for some time, such that experienced colors are stable, constant, and non-rotated, i.e., just like in normal color vision under standard conditions.
But there are reasons to think that adaptation is not necessarily a phenomenally-conscious phenomenon. Such an assumption is supported by research with blindsight patients exhibiting, in their unconscious perception, spectral wavelength sensitivity and several other features of color vision adaptation (Stoerig & Cowey 1989, 1991). In addition, adaptation should not be confused with habituation, which is an attentional phenomenon over which subjects reveal some conscious control (Webster 2012). This could help to explain some of the difficulties Grush et al. encountered when trying to prove the occurrence of phenomenal adaptation under color rotation—which are described below.
For most of the target paper, Grush et al. treat “phenomenal adaptation” as if it has a single, obvious, and straightforward meaning. Only at the end do they briefly hint at different readings of such adaptation depending on the understanding of the notions “phenomenal” or “qualia”. That is, their investigation is driven by certain implicit assumption of phenomenal qualities. However, these terms are quite controversial in philosophy and one may wonder what actually is examined in the study.
Philosophers denying the existence of phenomenal qualities or qualia (e.g., Churchland 1985, 1989; Tye 1995, 2000; Dennett 1988) may understand them in a specific and narrow sense, either as consciously-accessible properties of non-physical mind-dependent phenomenal objects, mental images called sense data (Lewis 1929; Robinson 1994), or intrinsic non-representational properties (Block 1990; Peacocke 1983), or non-physical, ineffable properties of experiences given to their subjects incorrigibly (Dennett 1988, 1991). Nonetheless, qualia may be endorsed in a broader sense, namely as phenomenal character. This use of the term generally refers to introspectively accessible qualitative aspects of one’s mental life, and it is hard to deny that these exist. Phenomenal character of an experience is “what it is like” for a subject to undergo the experience (Shoemaker 1994, 2001; Chalmers 1996; Nagel 1974). While engaging in introspection and focusing attention on the phenomenal character of experience, one is aware of and gets access to certain phenomenal qualities that make up the overall phenomenal character of the experience.
Since there is no single definition of the term “phenomenal qualities” or “qualia”, there might be also more than one reading of the notion of “phenomenal adaptation”. Depending on the particular understanding of phenomenality and the sort of mental states that can have phenomenal qualities or enter phenomenal consciousness, there seem to be different ways of interpreting the phenomenon of adaptation and thus the possibility of different kinds of adaptation. A related matter discussed in the philosophy of perception—between those supporting phenomenal liberalism and those who propose phenomenal conservatism (expansive and restrictive conceptions of the domain of phenomenal consciousness)—is whether there are high-level properties in the content of perception and whether cognitive states have a distinctive and proprietary phenomenology (Bayne 2009; Prinz 2012).
Phenomenal conservatism (e.g., Tye 1995, 2000; Carruthers 2005; Braddon-Mitchell & Jackson 2007; Nelkin 1989), proposing austere perceptual phenomenology, i.e., that the contents of perception are exclusively of the sensory sort, promotes sensory adaptation. Phenomenal adaptation, understood as sensory adaptation, has been described in the literature as adaptation to various distortions and systematic alterations of sensory input employing single or multiple modalities. For example, subjects wearing prismatic goggles or lenses inverting a visual scene in terms of color and spatial arrangement can adapt in the course of time to these new settings and become able to act normally, because they develop new visuo-tactile contingencies that allow them to get around and efficiently see and reach for objects (Held 1965). Importantly, such an adaptation directly affects perceptual experience and cannot be explained by correcting judgments. But this may not be the whole story about phenomenal adaptation, since phenomenal character does not have to be limited to sensory experiences, although traditionally it is said to be. In line with phenomenal liberalism, contents of perception may contain high-level properties such as kind properties (e.g., the property of being a tiger; recognizing that something belongs to a certain kind—seeing a tree as a pine tree; Siegel 2006; Bayne 2009), but also causal (the property of one thing’s causing another; Strawson 1985; Siegel 2006; Butterfill 2009), and generic properties (the property of being nonspecific; Block 2008; Grush 2007). These properties are abstract, generalized, and cognitive in their nature, yet they can enter into phenomenal contents. Consequently, a liberal conception, allowing cognitive states to possess phenomenal qualities, and phenomenal character to be ascribed to conceptual contents, endorses cognitive phenomenology and thus would opt for phenomenal adaptation in the cognitive aspects of experience.
The debate surrounding cognitive phenomenology involves many different versions and strengths of the claim that the domain of phenomenology extends beyond the sensory (Strawson 1994; Siewert 1998; Pitt 2004; Bayne & Montague 2011; Horgan & Tienson 2002; Kriegel 2002, 2007). Irrespective of its particular varieties, such a view raises alternative interpretations of Grush et al.’s results. It suggests that phenomenal adaptation may be present not only in sensory but also in cognitive aspects of experience. Both perceptual and cognitive states determine how we experience the world and adapt to changes in our surroundings, because they both exhibit their own phenomenal characters—something it is like to be in such a state for the subject (Chalmers 1996, p. 10; Strawson 1994; Montague & Bayne 2011).
Moreover, conceptual contents seem able to modify the phenomenal character of perceptual states; they can cognitively penetrate our perception (Raftopoulos 2005; Macpherson 2012; Siegel 2012). An interdisciplinary approach to the cognitive penetrability of perception assumes that there are various ways in which conscious perception can be affected by cognition—i.e., by thoughts, beliefs, desires, judgments, intentions, moods, emotions, expectations, knowledge, previous experiences, and memories (Frith & Dolan 1997; Bar 2003; McCauley & Henrich 2006; McCauley & Henrich 2006; Raftopoulos 2009; Vuilleumier & Driver 2007; Stokes 2012; Deroy 2013; Wu 2013; Vetter & Newen 2014; Briscoe 2014; Nanay 2014; Lupyan 2015). In other words, higher cognitive states not only have causal influence on the contents of perception, they are also explanatorily relevant in accounting for the processing of perceptual systems. It has been shown that semantic contents and categories play a critical role in perception, even in early sensory processing (cf. Mroczko et al. 2009; Mroczko-Wąsowicz & Nikolić 2013). This may be exemplified in the connection between language and color vision. For example, languages with a larger number of generic color terms such as Russian have an impact on color perception (Winawer et al. 2007).
Such an integrative cogno-sensory approach, combining high-level cognitive and low-level sensory aspects, is also manifested in recent theories of concepts relating the possession of concepts to perceptual adaptation in various ways (Machery 2009; Prinz 2010; Noë this collection). For instance, being sensitive and showing a discriminative response to certain kinds of objects or combinations of features corresponds to having concepts for the related kinds of objects (Machery 2009; Deroy 2013, 2014). According to the ability-based account of conceptuality, one can reveal skillful understanding of concepts in a perceptual, practical, or emotional way, meaning that the possession of concepts is a condition that informs and is informed by our able engagement with things (Noë 2012, this collection; cf. Wittgenstein 1953). This indicates a close interdependence between conceptuality and sensorimotor processing.
Consequently, phenomenal adaptation, in light of the enactive theory, would mean enactive adaptation and learning a new set of skills in the form of new sensorimotor contingencies and related dependencies, such as behavioral dispositions, predictive possibilities, and cognitive, aesthetic, and emotional reactions. Enactive adaptation would entail an application of sensorimotor skills to conceptual understanding, and as such it could be seen as adaptation in the cognitive aspects of experience with altered expectations or beliefs about or sensitivity to kinds of objects encountered in perceptual experience. This phenomenally liberal reading would provide an appropriately more capacious notion than the adaptation of the pure sensory sort offered by Grush et al.
To sum up, departing from a distinction between phenomenal conservatism that accepts perceptual phenomenology with solely sensory contents and phenomenal liberalism that acknowledges higher-level contents of perception and cognitive phenomenology (Bayne 2009; Montague & Bayne 2011), I differentiate between adaptation of the purely sensory sort and adaptation in the cognitive aspects of experience. The distinction is used to show the contrast in understandings of the notion of “phenomenal adaptation” between the target article and this commentary. Grush et al. seem to suggest that phenomenal and (non-phenomenal) semantic adaptation are different forms of a more general phenomenon of adaptation. However, they do not give any explicit example of the genus of adaptation of which these later are a species. I contend, in turn, that there is no need to produce such subclasses of the notion; semantic adaptation involving higher-level non-sensory states may also be understood as phenomenal. Thus, the reading of adaptation I put forward pertains jointly to the phenomenal and semantic aspects of regaining of stimulus constancy; it assumes a recovery of prototypical color-object associations both in phenomenal experience and in semantic reference in spite of changes in input. This follows from phenomenal liberalism.
The proposed view is that being processed in the headway of phenomenal adaptation is phenomenal character, understood in an expansive liberal way that includes high-level contents. Therefore phenomenal adaptation is considered to be the adjustment of cognitive aspects of experience.
Dissociating semantic from phenomenal adaptation is problematic. This is because they are interconnected. It is hard to think about the occurrence of semantic adaptation without phenomenal adaptation taking place and vice versa – semantic adaptation is methodologically necessary for detecting phenomenal adaptation. This presumed correlation might be the reason why, when faced with difficulties finding phenomenal adaptation to a color-rotated scene, the investigators could not confirm any reliable Stroop results for semantic adaptation. In addition, it should be noted that Stroop-type tasks contain two components of competition—semantic and perceptual (Stroop 1935; Nikolić et al. 2007; Mroczko et al. 2009)—and as such they exhibit limitations in differentiating between semantic and perceptual aspects of the phenomena tested.
To assess the occurrence of phenomenal color adaptation under rotation, that is, the process of the normal phenomenal appearance of objects returning, Grush et al. used the memory color effect (Hansen et al. 2006), aesthetic judgments of food and people, and subjective reports from their test persons.
It has often been assumed that subjective introspective reports are a generally reliable mode of first-person access to one’s current conscious states or processes (Descartes 1984; Locke 1689/1979; Hume 1978; Brentano 1973; Husserl 1982; Chalmers 2003; Gertler 2001; Horgan et al. 2006; Horgan & Kriegel 2007; Varela 1996; Rees & Frith 2007; Hurlburt & Schwitzgebel 2007; Hohwy 2011). However, this assumption is also problematic. Arguments for introspective scepticism or even criticism of introspective methodology pose genuine threats to the trustworthiness of this approach (see Bayne this collection, for a discussion of such views). Because of this ambivalence one needs to be careful when using subjective reports as a source of or support for the results presented.
Certain doubts about whether subjective reports are trustworthy enough come from the fact that introspection delivers solely first-person, unverifiable, private data, and thus it is unscientific and often fallible (Dennett 1991; cf. Zmigrod & Hommel 2011). In addition, subjects tested are often uncertain or disagree about what the introspective access actually provides (Bayne & Spener 2010; see also Bayne this collection) and have difficulty describing their own conscious experiences (Schwitzgebel 2008). Nonetheless, this is not to deny that they have some first-person knowledge of phenomenal consciousness.
The specific reasons one may have for doubting the findings in the context of Grush et al.’s study are related to the fact that investigators were also the test subjects. We should avoid involving persons who know the hypothesis when conducting the experiments, who in this case tested and evaluated themselves at the same time. Knowing the research question and the expected or desired results may bias any study.