Current Controversies in Philosophy of Perception

Bence Nanay

• Publisher: Routledge
• Year: 2017
• ISBN:
  9781317556725
• Edition: 1
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

This book provides an up-to-date and accessible overview of the hottest and most influential contemporary debates in philosophy of perception, written especially for this volume by many of the most important philosophers of the field. The book addresses the following key questions: Can perception be unconscious? What is the relation between perception and attention? What properties can we perceive? Are perceptual states representations? How is vision different from the other sense modalities (like hearing or smell)? How do these sense modalities interact with one another? Contributors are Ned Block, Berit Brogaard, Alex Byrne, Robert Kentridge, John Kulvicki, Heather Logue, Mohan Matthen, Bence Nanay, Matt Nudds, Casey O'Callaghan, Adam Pautz, Ian Phillips, Susanna Siegel and Wayne Wu.

PART I
Are Perceptual States Representations?

CHAPTER 2
Experiences Are Representations

An Empirical Argument

ADAM PAUTZ

To a first approximation, representationalism is the hypothesis that experiences are representational states akin to beliefs. I will sketch an argument for representationalism on the basis of an inference to the best explanation. I will not attempt to show that it is superior to every alternative. Instead I will focus on its main rivals—namely, “naïve realism” and the “inner state view”.

My plan is as follows. I will start with contemporary naïve realism, defended by John Campbell, Bill Fish, and Michael Martin, among others. I will argue that it violates internal dependence: the empirically determined role of the internal processing of the brain in shaping phenomenal character. This will bring us to the inner state view, defended by Ned Block, Brian McLaughlin, and David Papineau, among others. I will argue that, while it accommodates “internal-dependence”, it fails to accommodate the essential “externally directedness” of experience. Finally, I will argue that only representationalists can adequately explain both of these features of experience. The upshot is a largely empirical case for representationalism.

At the end, I briefly address what I consider to be the most profound objection to representationalism. Representationalism faces an overlooked modal puzzle about what experiences are possible.

1. Naïve Realism Violates Internal-Dependence: An Empirical Refutation

I will introduce “naïve realism” by way of what I consider to be the best argument for it (Pautz 2010, pp. 283–295). To illustrate, suppose you see a blue ball (Figure 2.1). The most basic philosophical question about experience is this: what grounds the qualitative character of experience? Now there is something to this thought: it is obvious without argument that in such a case the character of your experience is grounded in nothing but your being acquainted with the bluish color and round shape of some concrete item. If so, then there are a couple of live options as to what this item is. It may be a mental image (a “sense datum”) created by the brain (the “sense datum theory”), or it may just be the ball itself. Since there are well-known problems with the first option, we have a case for accepting the second. The result is naïve realism: the character of your experience is grounded in your being acquainted with the color and shape of the physical object.

Figure 2.1 A blue-looking ball

In general, naïve realists hold that, even before sentient creatures evolved, external items possessed multiple objective (mind-independent) sensible properties: color properties, smell properties, loudness levels, and so on. Objects also possessed viewpoint relative but objective shapes, like being elliptical from here. The role of brain is not to construct experience. Rather the brain “opens the window shutter” to reveal objective properties of the items in the world. That is, when the brain responds to these objective sensible properties in the biologically normal way, this enables the mind to “reach out” and become acquainted with them. This long causal process is the superveniencebase of worldly acquaintance. In such veridical cases, the qualitative character of your experience is fully grounded in what external states you are acquainted with. The naïve realist John Campbell (Campbell and Cassam 2014, p. 27) also mentions your “point of view” as a factor, but then he says (p. 28) that this too is just a matter of which external states in the scene you are acquainted with (together perhaps with your own location in space).

Naïve realism is externalist. The brain configures qualitative character only to a very limited extent: only to the extent that it selects what objective external states we get to be acquainted with. For instance, pigeons have different color experiences than humans, only because their different sensory systems enable them to be acquainted with different objective colors (constituted by UV light). As Campbell (2010, p. 206) puts it:

[Naïve realism holds] that qualitative properties are in fact characteristics of the world we observe; our experiences have the qualitative characters that they do in virtue of the fact that they are relations to those aspects of the world. So looking for the qualitative character of experience in the nature of a brain state is looking for it in the wrong place; we have to be looking rather at the [properties] of the objects experienced.

So much for what naïve realism is. Is it right? I think that there are serious problems with the simple argument for it as just outlined, even if I consider it to be the best argument for naïve realism (Pautz 2010, pp. 295–297). I also think that there are much stronger arguments against naïve realism. Many argue that naïve realists cannot adequately explain illusion and hallucination. Here I will set this issue aside and develop a new line of argument: even in “normal” cases, naïve realism is inadequate. Psychophysics has shown that, even in normal cases, qualitative similarity is very poorly correlated with external physical similarity. At the same time, neuroscience has shown that neural similarity is the only accurate predictor of qualitative similarity. In short, the typical situation is that there is “good internal correlation” even while there is “bad external correlation”. Naïve realists like Campbell, Fish, and Martin neglect the scientific facts. They have it the wrong way around. Looking for the basis of qualitative character in the external world is looking for it in the wrong place; we have to be looking rather at the brain.

For example, Figure 2.2 shows reflectances typical of purple-looking grapes, a blue-looking ball, and a green-looking leaf.

By any natural measure, it is not the case that the reflectance of the ball objectively resembles the reflectance of the grapes more than the reflectance of the patch of grass. (In fact, if anything, the opposite is true.) Nevertheless, the blue appearance of the ball resembles the purple appearance of the grapes much more than the green appearance of the leaf. So there is “bad external correlation”.

Since the explanation of similarity in color appearance is not to be found in the external world, there is reason to think it must reside in the brain. And this is exactly what recent neuroscience suggests. Neuroscience demonstrates “good internal correlation”. As Brouwer and Heeger write,

The visual system encodes color by means of a distributed [neural] representation [in area V4] … similar colors evoke similar patterns of [neural] activity, and neural representations of color [in V4] can be characterized by low-dimensional “neural color spaces” in which the positions of [experienced] colors capture similarities between corresponding patterns of activity.

(2013, p. 15454)

Figure 2.2 Reflectances of grapes, a blue-looking ball, and a leaf. From MacAdam (1985)

So, your internal V4 neural representation of the blue-looking ball resembles your V4 neural representation of the purple-looking grapes more than your V4 neural representation of the green-looking leaf. This is the only available explanation of the resemblance-order among your color experiences.

Now let’s take a parallel example involving smell. Consider the chemical properties in Figure 2.3 below.

It is not the case that the middle chemical-type, R-limonene, resembles citral more than R -carvone. Rather, it resembles R-carvone more closely. Nevertheless, the perceived smell quality of R-limonene resembles the perceived smell quality of citral much more than the perceived smell quality of R-carvone. R-limonene and citral smell different, but their apparent smell qualities can both be described as “citrus-like”. That is why I call them citrus smell “1” and “2”. By contrast, R -carvone smells “minty”. This is another case of “bad external correlation”.

At the same time, neuroscience demonstrates “good internal correlation”. Howard and co-workers (2009) found that “spatially distributed ensemble activity in human posterior piriform cortex (PPC) coincides with perceptual ratings of odor quality, such that odorants with more (or less) similar fMRI patterns were perceived as more (or less) alike”. For instance, they found that your PPC neural representation of R-limonene resembles your PPC neural representation of citral more than your PPC neural representation of R-carvone, in perfect agreement with the character of your smell experiences. As Margot (2009) says, “Because the chemical structure of the odors in [the citrus] odor category are very different, this is strong support for the idea that the PPC codes odor quality rather than structural and chemical similarity”.

Figure 2.3 The first two chemicals smell citrus-like (“citrus 2” and “citrus 1”); the third smells minty. From Margot (2009)

Figure 2.4 From Goldstein (2009)

Finally, consider auditory experience. As Figure 2.4 illustrates, if you continuously increase the “voice onset time” of a speech signal (the time between opening the lips and the onset of vocal fold vibration), then suddenly at 30 ms there will be a big, categorical change in the audible quality, from /da/ to /ta/.

This categorical change in the perceived sensible property corresponds to no categorical change in the objective stimulus. It corresponds only to a categorical change in your neural representation in the brain (Chang et al. 2010).

In addition, even under normal conditions, perceived loudness is related in an enormously complex, non-linear fashion to a number of objective physical properties, including intensity, frequency, and “critical bands”. By contrast, it is related in a simple fashion to the total neural activity produced by a sound, according to standard models (Moore 2003).

Given bad external correlation and good internal correlation, naïve realists’ externalist approach fails. To show this, I offer two arguments.

First, the argument from irregular grounding. Naïve realists hold that sensible properties (color properties, smell and taste properties, loudness levels) are brain-independent properties of physical items. What, in their view, is the relationship between these objective sensible properties of these items, and the underlying ordinary physical properties: reflectances, chemical properties, and acoustic properties?

Bad external correlation (illustrated in Figures 2.3 and 2.4) means that sensible properties and the underlying physical properties fall into different resemblance-orders. This rules out the view that sensible properties are identical with the underlying physical properties (Pautz 2016).

However, naïve realists might hold onto their view by claiming that sensible properties are irreducible objective properties of things that are grounded in, but distinct from, the corresponding physical properties. In that case, their resemblance-orders needn’t match (Allen 2015).

But this requires irregular grounding: totally unsystematic and arbitrary grounding connections. Here are some examples. (I) As Figure 2.2 shows, the ball’s reflectance resembles the leaf’s reflectance more than the grape’s reflectance. Still, naïve realists must hold that these reflectances ground objective colors—namely blue, green, and purple, which stand in a totally different, autonomous resemblance-order (blue evidently doesn’t resemble green more than purple). They must hold that this is just a quirk of reality with no explanation. (II) Likewise they must hold that the chemical structures in Figure 2.3 ground objective smell qualities that stand in a totally different resemblance-order than they do. (III) They must hold that if varying the voice onset time for a speech signal continuously, then at precisely 30 ms there is a big, discontinuous “jump” in its irreducible, objective audible quality, from /da/ to /ta/. Since they take this audible quality to be independent of our neural response, they have no...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• List of Figures
• List of Contributors
• Part I Are Perceptual States Representations?
• Part II Is Perception Thin or Rich?
• Part III Non-Visual Sense Modalities
• Part IV The Multimodality of Perception
• Part V Is Attention Necessary for Perception?
• Part VI Can Perception Be Unconscious?
• Appendix: Other Controversies in Philosophy of Perception
• Index