Epiphenomenal Mind

An Integrated Outlook on Sensations, Beliefs, and Pleasure

William S. Robinson

• Publisher: Routledge
• Year: 2018
• eBook ISBN:
  9780429787706
• Topic: Philosophy
• Subtopic: Mind & Body in Philosophy

According to epiphenomenalism, our behavior is caused by events in our brains that also cause our mentality. This resulting mentality reflects our brains' organization, but does not in turn cause anything. This book defends an epiphenomenalist account of philosophy of mind. It builds on the author's previous work by moving beyond a discussion of sensations to apply an epiphenomenalist outlook to other aspects of mental causation such as beliefs, desires, pleasure, and displeasure. The first four chapters of the book argue for a dualistic theory of sensations and develop an epiphenomenalist version of dualism. The remaining chapters discuss propositional attitudes and valence. The author also responds to potential objections to epiphenomenalism by considering how sensations, intelligence, or understanding might be built into a robot. This book will be of interest to scholars and students in philosophy of mind who are interested in consciousness, mental causation, and how our mentality is situated in the world.

1 Sensations

1. As one of my article titles has it, Red is the Hardest Problem. If we can achieve a full understanding of what red is, how it is in the world, how it comes to be in the world, we will have understood the problem of what consciousness is, and we will have solved the Hard Problem of consciousness.

There is, of course, nothing special about red. Any color could occur in its place: so could taste qualities (sweet, oreganoish, vanilla-flavory) or odor qualities (sulfurous, floral, vinegary), sound qualities (C#, resonant, shrill) or tactile qualities (slick, rough, warm). Similarly, we could just as well focus on the qualities of our bodily sensations, such as itchiness or nausea.

These qualities, and other qualities in the same sensory modalities, are known to philosophers as qualia. The term often extends to cover emotional qualities, such as the quality of the feeling one has when angry, or the feeling one has when embarrassed, elated, jealous, and so on. I take this extended use to be justified. Emotions, however, are complex, and I do not want to make my results here dependent on agreement about all aspects of the theory of emotions. So, I shall mainly be concerned with sensory qualia – qualities like those in the preceding paragraph.

Qualia are qualities, which is to say that they can occur in more than one instance. We will eventually need to have a term that denotes an instance of a quale. We shall see, however, that how we are to conceive of an instance of a quale is a controversial matter.

This chapter will address this controversy. Its broader aim is to explain a view of sensations that will, among other things, justify the claim of centrality for phenomenal qualities in the study of consciousness. It begins by explaining a neo-Lockean view that will then be supported with an argument that avoids reliance on conceivability. Later in the chapter, I compare this view favorably with several physicalist views. The result will be a substantial case for the need to develop dualism for sensations, a task that will be pursued in the following three chapters.

A Theory of Sensations

2. We may begin with a view that has its roots in Galileo and Locke (although the physical details are updated).

When we see a red object, light rays are reflected from its surface into our eyes. These light rays are electromagnetic disturbances and they are completely described by their energy and frequency (or, wavelength) composition. The character of the reflected light depends (a) on the character of the light that falls on the object, and (b) on the molecular structure at the surface of the object, which determines which percentage of each wavelength is reflected. When the reflected light encounters our retinas, it causes changes in opsins in our cone cells. These cone cells are specialized neurons, and they synapse onto other neurons in our retinas. The latter synapse onto further neurons, eventually causing neural activations in primary visual areas in the back of our brains. These activations have effects in neurons at many other places in our brains, and these effects may eventually lead to actions (through activating motor neurons). Some subset of these neural activations also causes an instance of red – an instance that may be called ‘our experience of red’, or ‘a red experience’. This instance of red is an episode of consciousness.

Although Galileo wrote earlier, Locke’s use of “ideas” is more suggestive of the foregoing view than is Galileo’s location of sensory qualities in the “living body”. So, I shall call it a ‘neo-Lockean’ view.

The neo-Lockean view is not a popular view today. But let us make explicit some claims that would seem natural if one held it. The most instructive corollary of the view is that it is committed to two senses of ‘red’. There is the red object: this is an object that can reflect light. There is the instance of red that is an episode in our consciousness: this is not an object that can reflect light. Moreover, the redness in our consciousness has a claim to be the central meaning of the term ‘red’; for it does seem that the meaning of the word ‘red’ is a color that we find in our conscious experience. The red that is in the object can be defined in terms of the red in consciousness. Namely, a red object is one that, in normal lighting conditions (e.g., daylight), will produce red in the experience of people with normal sight.¹ In the sense of ‘red’ in which redness occurs in an episode of our consciousness, red objects are, on this view, not red. They are objects with surfaces whose molecules resonate at certain wavelengths. Resonating requires energy, so wavelengths at which molecules resonate are wavelengths at which energy is absorbed, i.e., not highly reflected, by the object. A full accounting of everything outside our bodies that is relevant to seeing things can be given in terms of reflectance percentages across the spectrum of frequencies to which our eyes are sensitive. Redness does not come into play until we ask what an object looks like to a normal person in normal lighting. If the normal result is experiencing red, then we call the object ‘red’. This latter sense of ‘red’ is redness by relation. To explain what that means, consider that healthy foods are not foods that are in good health; they are foods that produce good health in us. Analogously, red objects are not red in the basic sense; they are objects that produce red, in the basic sense, in our consciousness².

Neo-Lockeanism is sometimes vilified as an ‘error theory’. It implies that ordinary people are mistaken when they think that colors are literally out there upon the surfaces of objects. In a holdover from the heyday of ordinary language philosophy, this aspersion on the wisdom of ordinary people is often greeted with horror and disdain. Sometimes it is rather testily insisted that one use a different term – ‘phenomenal red’, or ‘red-ish’ – for the property (or, some may say, the alleged property) instantiated in experiences. This is so, despite the fact that the same objectors are quite happy to reject many other views that are commonly found highly intuitive by most nonphilosophers.

The neo-Lockean view accepts that the correct, detailed theory of perception diverges from what most people who have not encountered the details of visual processes ordinarily think. But it is also true that the so-called ‘error theory’ does not impute any ordinary error to most people. That is, it is still true on the neo-Lockean view that bullfighters’ capes, certain pool balls or roses, old-fashioned fire engines and so on are red objects. That is, they really are objects that normally produce red in the consciousness of the normally sighted. It is not a consequence of the neo-Lockean view that most people are generally mistaken when they use color language for its ordinary, nonphilosophical purposes. But it would be only an implication of error of that kind that would be an objection to a view.

Just to be clear, the neo-Lockean view does allow for ordinary cases of error. Cars of several colors look black under mercury vapor lamps; people can seem to be talking in low voices if one’s ears have not yet ‘popped’ after landing at an airport. These are cases where conditions are not normal, and one may be deceived if one does not take the abnormality into account (or, is unaware of it). They are cases where it is not true that the object in question would continue to cause the same sensations if conditions were normal. If one judges otherwise, one is making an error, and an error that has a special relation to how one is perceiving. But there is no suggestion from a neo-Lockean view that such errors are widespread. The widespread error is only the theoretical error of supposing that we are in direct contact with colors (and other phenomenal qualities) in external things, when in fact colors (in the basic sense) are only in our experiences that are caused by surfaces, light waves and neural events in locations from the retina to higher visual areas³.

3. On the neo-Lockean view, there is a natural suggestion for how to think of an instance of a quale. Namely, an instance of a quale is an episode of consciousness of a certain kind. Consider looking at a red rose, and then looking at a white one. A salient difference between your successive states of consciousness is the difference between the colors. The suggestion is that you have two episodes of consciousness, and at least one of the differences between them is the red of the one and the white of the other. This is not a difference that your consciousness is merely related to; it is a difference in the kind of consciousness you are having at each moment.

The neo-Lockean way of looking at qualia suggests a dualistic view, and that is the way the view is to be understood here. Red objects such as roses and pool balls reflect light; instances of qualia do not. If, per impossibile, all the neural activations that cause an instance of redness in consciousness were to occur while brain surgeons had removed a portion of the skull and were viewing all the neurons involved, they would not see anything red (except, perhaps, blood, which they would see just the same if the person under the knife were experiencing any other color). Qualia, on the neo-Lockean view, are instantiated in consciousness, and consciousness, on this view, is not conceived of as an object alongside roses and stop signs.

4. The neo-Lockean way of looking at instances of qualia could be usefully captured by a terminology according to which an episode of consciousness that instantiates a given quale is called a sensation of the given kind. Or, to emphasize the fact that consciousness changes so rapidly, and that episodes of consciousness are events, we might call instances of qualia ‘qualitative events’. Or, we might speak of ‘red experiences’, ‘blue experiences’ and so on.

We need to make a choice of terminology that will help us to be clear not only about the neo-Lockean view but also about rival views. To this end, I am going to restrict ‘qualitative event’ to mean ‘an episode of consciousness that is literally an instance of a quale’. On the other hand, I shall use both ‘F sensation’ and ‘sensation of F’ as generic terms, to mean items such that qualitative event theory is just one account of what they are. Those who do not like the neo-Lockean view will say that there are no qualitative events. They will still be entitled to say that there are sensations of various qualitative kinds; they will just have to give an alternative account of what sensations are. We shall soon look at some of these alternatives.

This addition to neo-Lockeanism stretches the connection with the historical Locke to a degree that warrants introducing a new label. I shall use the one I introduced in Robinson (2004), namely, ‘qualitative event realism’ (often abbreviated to QER).

An Objection Raised and Answered

5. Before proceeding to discuss alternatives to QER, I want to state and respond to an objection to it. This objection was recently put to me by a philosopher who is not unsympathetic to the claim that dualism in some form may be true. “What”, he asked, “are the bearers of qualia?” Behind this question was a tacit assumption that qualia would need bearers. Such bearers might be brain events. In that case, qualia bearers would be perfectly respectable physical items, and dualists could accurately describe their view as ‘property dualism’, i.e., a dualism of properties but not of individuals that bear those properties. But if qualia bearers are not brain events, it looks as though one would need a special ‘substance’ to bear qualia, and this would be a more radical complication of one’s ontology.⁴

6. I do not think that brain events are bearers of phenomenal qualities. A brain event is the occurrence of a specific kind of activity in some group of brain parts. For example, if a certain three neurons were to fire at rates that stood in a ratio of 5:3:2 over some interval that would be a brain event of a certain kind. Now, suppose one said that such an event instantiated redness. What can that mean? It surely cannot mean that if one were able to observe those neurons when they were firing in that way, they would look red. Of course, it is perfectly intelligible that a brain event and a sensation can be related, such that one has a red sensation when and only when a certain kind of event occurs in one’s brain. But that relation might be causation, or correlation. Declaring that the relation is instead one of instantiation seems to me to introduce a mysterious new kind of ‘instantiation’ that is not enlightening and merely serves to hide a problem.

The generic type of our sample event is specified by its being a set of ratios of firing rates in a group of neurons. The particular duration, the particular neurons involved and the particular set of rates do not follow from the generic description; however, it does follow from that description that in each instance of the event type, there will be some duration, some neurons involved and some set of rates. This connection makes it natural to attribute the particular rate values in a particular instance to the event whose individuating description implies that the event will have some property of that kind. Similarly, a particular event of the type in our example will cost its possessor some number ei of ergs. This number does not follow from the specification of the event, but it is nomologically required that an event of this kind will cost its possessor some quantity of energy. Again, this connection makes it natural to attribute the particular value in a particular instance to the particular event of the given type.

It is not a consequence of the description in our brain event example that there be any color, or indeed, any phenomenal quality at all that an instance of the specified event type should possess. An attribution of a phenomenal quality to such an event would thus be an attribution that is not underwritten by a generic connection; it would be an arbitrary assignment to the event as bearer by fiat. We should, however, avoid doing metaphysics by fiat.

It may be objected that dualists are committed to a nomological connection between brain events and qualitative events. So, even on their view, there is a connection that could give a nonarbitrary underpinning to the attribution of phenomenal qualities to brain events.

The dualists’ nomological commitment, however, is to the relation of causation. Causation already presupposes that the cause and the effect are distinct events. Thus, dualists face no inconsistency in allowing that brain events cause qualitative events while denying the kind of connection that would justify counting brain events as bearers of phenomenal qualities.

Of course, there are many kinds of brain events besides the example I have used in this section. It may be that firing ratios offer too crude a description – perhaps absolute time intervals, or ratios of time intervals are needed to describe more defensible candidates for brain events that might be bearers of (or causes of) qualia. Perhaps dendro-dendritic connections or events in microtubules will prove important (Hameroff & Penrose, 2014). Electromagnetic field theories like those of Pockett (2002) or McFadden (2013) may gain support. Perhaps some glial cells may turn out to have more to do with our mental life than mere provision of a kind of skeleton, or performance of repair and trash management roles (Fields, 2010; Anastassiou & Shai, 2016).

The point that was made with my example would, however, apply to any such candidates. More generally, arguments that use neural firings or their properties and relations as illustrations can usually be easily adapted so as to apply to the other kinds of brain events. This generalizability should be assumed throughout this book, unless otherwise noted.

7. It might be thought that there is a problem in thinking of possible qualia bearers as events. But it is surely not our brain parts that are red, or tasty. Aside from pathologies such as tinnitus, our sensations arise, last for a brief interval, and are succeeded by other sensations. Supposing that our neurons are, e.g., red for a short time and then change to some other color would not relieve the arbitrariness of phenomenal quality attributions that was explained in §6. The only individuals that seem available for bearers are thus arrangements of transitory conditions in some set of brain parts. The parts are present before and after the sensation occurs; it seems it can only be brain-parts-in-a-state-of-activity of various sorts that could be plausibly nominated as candidates for bearers of phenomenal qualities. But that is just a long-winded way of saying that the available candidates for the bearers are events.

8. ...

Table of contents

• Cover
• Half Title
• Series Page
• Title Page
• Copyright Page
• Table of Contents
• Preface
• 1 Sensations
• 2 Developing Dualism
• 3 Epiphenomenalism
• 4 Experience as Such
• 5 Mental Causation
• 6 Believing and Desiring
• 7 Robots
• 8 Unconscious Processing
• 9 Valence
• Epilogue
• References
• Index