Consciousness Versus Cognition

Multiply seventeen by twenty-three in your head. Then drink something bitter.

Both episodes are mental and the difference between them is not one of difficulty. The multiplication is work your mind does with information, and a pocket calculator does the same work without anyone supposing there is a mind in it. The bitterness is not work at all. It is a way things are for you while it lasts, and the calculator has no analogue of it.

The previous article treated the mental as one thing. It is at least two, and separating them is what turned a general puzzle into a set of tractable questions plus one that has resisted everything.

One word doing too many jobs

Ned Block called the concept a mongrel, and the label is fair. The word is applied to a creature (the patient regained consciousness), to a relation between a creature and something else (she was conscious of a smell), and to a state (that belief was conscious at the time). These come apart. A creature can be awake and unaware of a stimulus that is nonetheless influencing its behaviour, and there are respectable theories on which a state can carry information the system uses without anyone being aware of it.

Most disputes in the area involve two people using the word for different things and each finding the other's claims obviously false. Fixing the vocabulary is not pedantry here. It is most of the progress.

The processing side

Cognition is the part of the mind that looks like something a physical system could do, because we build physical systems that do parts of it. A chess program represents a position, evaluates options, and selects a move. A face-recognition system categorises inputs. Nothing in these achievements requires a special kind of substance, and the explanations run entirely in terms of states that carry information and processes that operate on them.

That is not a claim that cognition is finished business. Nobody has a complete account of how a brain stores a memory or parses a sentence, and the details are the work of several sciences. The point is about the kind of difficulty. Explaining how a system discriminates red things from green things is a hard engineering and empirical problem with no mystery about what a solution would look like. You would specify the mechanism and check that it does the job.

The felt side

Nagel introduced the phrase in a paper of 1974 and chose bats as his example because they are mammals, undeniably experiencing creatures, and equipped with a sense we lack. A bat navigates by emitting cries and reading the returning echoes. There is presumably something it is like to perceive the world that way, and it is not like anything we can get at.

The argument is often misreported as the claim that we cannot imagine being a bat. Nagel's point is sharper. We can imagine ourselves behaving as a bat behaves, hanging upside down and catching insects in the dark, and that is imagining what it would be like for us to do those things. What is wanted is what it is like for the bat, and our imaginative resources are drawn from our own experience, so they cannot reach it. The trouble is not a shortage of information about bat physiology. Someone could know every fact about the bat's nervous system and still not be in a position to say what its experience is like.

That generalises into a difficulty about method. The physical sciences make progress by moving away from any particular point of view, describing the world in terms available to any observer whatever their sensory apparatus. Experience is essentially tied to a point of view. A form of understanding that succeeds by discarding points of view is being asked to capture something that is nothing but a point of view, and Nagel's question is whether it can.

The term has a history worth knowing because it affects how arguments are read. Charles Sanders Peirce introduced the singular form in 1866, and C. I. Lewis used the plural in something close to the current sense in Mind and the World-Order in 1929, where they were features of the given in experience rather than properties of external objects.

Since then the word has picked up freight. Some writers use it neutrally, as a name for whatever it is that distinguishes the experience of red from the experience of green, and use it without commitment about the nature of that. Others use it for features that are by definition intrinsic, private, directly known, and not analysable in functional terms, which builds a substantial and disputed thesis into the vocabulary. When a philosopher denies that qualia exist, they are usually rejecting the second package rather than claiming that nothing distinguishes red experiences from green ones. Reading a denial of qualia as a denial that experience has any character at all is the standard way of misunderstanding those debates.

The other half of the split

Block drew the distinction between this and phenomenal consciousness in a paper of 1995, and the argument for drawing it is that the two are defined in different terms and can in principle come apart. Access consciousness is a functional notion, specified by what a state is available for. Phenomenal consciousness is specified by how it is, and no amount of availability obviously adds up to a felt character.

Whether they actually come apart is a separate and contested question. The case that is always cited runs in one direction.

Why the split matters

Return to the problem from the previous article. The four claims that could not all be true were stated about the mental in general, and once the mental is divided the pressure turns out to be concentrated in one half.

For cognition, the shape of a physicalist account is clear even where the details are missing. A cognitive capacity is specified by what it does, and to explain it you find a mechanism that does that. When neuroscience identifies the circuitry underlying working memory, nothing is left over that anyone thinks needs separate explanation.

For phenomenal consciousness the same strategy has no obvious purchase. Suppose you had the complete mechanism for pain: the receptors, the pathways, the processing, the behavioural output. A question appears to remain about why that mechanism is accompanied by hurting rather than running in the dark, and it is not clear what could answer it. The advanced tier gives this difficulty its standard name and works through the arguments built on it, including the argument that a person who knew every physical fact about colour vision would still learn something on seeing red for the first time, and the argument that a physical duplicate of you with no experience at all is coherently describable.

The division also explains a pattern in the literature. Theories that account elegantly for what consciousness does tend to be accused of changing the subject, and accounts that take the felt character as basic tend to have nothing to say about why it should be connected to anything. That is the shape of the field, and it follows from the two halves being specified in incompatible vocabularies.

Next

Two categories have been introduced without saying what the items in them are. Beliefs, desires, pains, and intentions have all been called mental states, and what makes something a state of that kind, and how the kinds differ from one another, has not been touched. That is the next article.