What is a mental state?

Question

A great deal of the theory of mind is based upon supposedly obvious observations about our mental states. E.g. that they are immediately available, or that they are introspectable, at least in theory.

It seems obvious that the functions of the nervous system that ultimately control truly autonomous functions like peristalsis are not aspects of one's mental state. But someone recently suggested that the unconscious adaptations that keep one from stepping on rocks at the beach should be such.

These seem to stake out a vague borderland in the middle of which falls something like breathing, as something we can control at will, but usually don't, never learned or decided much about, and almost never notice. When we wait "with 'bated breath", we have made an unconscious decision not to breathe right then. But is that an element of our 'mental state'?

I can't say I could consider any of these processes even close to 'mental'. Mental-ness seems to involve will and decisions, but only ones that eventually become conscious. Still, I cannot say on what basis I would place those constraints, or where my favored position actually comes from.

What, if anything, have various schools of thought put forward as a 'litmus test' for whether something is or is not a mental state, and what are their respective justifications for their chosen definitions?

Is there a useful continuum of 'mental-ness' that clarifies or points up the sticking points between different basic positions or theories?

Answer 1

Let's start with some definitions, that I don't think are too idiosyncratic. A mental state is defined by its typical environmental causes and behavioral effects and relations to other mental states. For example, "pain" is the mental state characteristically caused by pinpricks and burns, which characteristically causes one to flinch, say "Ow!" and which occasions other mental states like surprise or anger.

A mental state may be conscious like the state I am in right now as I write these words, or unconscious, as in the case of your feet moving to avoid the rocks in your path while you walk. Your mind is often working processing information even when you are not consciously attending to this. (Have you ever gone to bed thinking about a problem, and woke up knowing the solution?) Likewise, even things that we think of as arch-examples of conscious mental states like the state of feeling pain can kept from coming to our conscious awareness by medical interventions. If you think, for instance, that the feeling of pain supervenes upon the firing of C-fiber neurons, then you should acknowledge that a person on a morphine drip is experiencing pain, because the C-fibers can be firing. The person doesn't feel the pain because the drug interrupts the ability of the lower-level systems to report to conscious attention. (Forgive my layman's understanding of the neurology here. I think something like this story is correct, but am open to correction by a trained medical professional.)

Further, there are lots of important results in the psychology of perception to demonstrate that there are unconscious mental states. Take blindsight patients. They have brain lesions which render part of their visual field blank to consciousness. If you put an object in the blank field and ask them "what do you see?" they reply "absolutely nothing." However, if you place, for instance, a door knob in the blank field and ask the patient to interact with it, then the patients always reach forward and grab the door know just as if they can see it. This suggests that what the lesion has destroyed is not the parts of the visual system responsible for recognizing the objects, but rather the parts responsible for reporting to higher-level conscious processes the content of the lower-level object recognition systems.

All of this is very strong evidence that consciousness is not an intrinsic feature of mental states. Many important mental states are things we can become conscious of, of course. But many important processes in our minds having to do with perception and action operate more or less well without the handholding of conscious attention.

Answer 2

Because your question aims at testing different proposals for the concepts of “mental state”, I would like to attack the problem from the side of an empirical science like neusoscience and from informatics.

In informatics the concept of state is fundamental for every kind of process models, like process algebras, Petri net etc. Because the brain is a distributed system, the concept of a global state will result from combining the local states of its subsystems.

Hence the first step to design a testable model of the brain is a list of the relevant subsystems and their connections. One needs a wiring diagram and its translation into a net. A further step is to determine the possible states of each subsystem and their possible state transitions.

As an example, input from neuroscicence for such a model would determine certain subsystems from pre-frontal cortex, from the motor areals of frontal lobe, and from the limbic system, which are relevant for triggering voluntary motor actions.

Cognitive science considers the representational content as a crucial characterisation of a mental state. That’s the question for the semantics of the states. The latter representing the learned experience of the system.

It seems easy to equate the semantics of the states of the motor cortex with the activation of certain muscles. But in case of the upstream systems the question seems much more difficult.

One kind of test would be whether the model reproduces some well-known stimulus-reponse behaviour under a given, but fixed subset of local states.

So far sharing some thoughts of a layman in neurocience. I wellcome any feedback from the experts.

Answer 3

The computational theory of the mind, in conjunction with Higher Order Theories of Consciousness can address your question, assuming you wold agree that HOTC provide a possible approach to qualia and Chalmers' hard problem. Mental states are software to the brain's hardware, and HOTC can account for the difference between unconscious and conscious metal processing.

The continuum of "Mental-ness" you describe would then be the different layers of software, from a bottom neurological level, almost physical in nature, then processes that control breathing and walking etc, all the way up to language and finally higher order social and scientific concepts (analogous to the layers of abstraction in programming languages and Operating systems, or maybe the OSI model of networks). Again, some variation of HOTC would allow for processes to run in the background, or in the foreground, depending on the subject's desires and decisions. And it is likely that, as with all things biological, the layers of abstraction form a continuum, instead of the discrete layers of machine software, although their might be a qualitative jump when language processing is involved.

This doesn't address your litmus test question, especially how other schools that don't agree with CTM would address this question. But I think that your question is only relevant within a functionalist worldview anyway. An eliminativist or an indentity theorist would simply reject your premises on the grounds that they're neural all processes anyway. Dualists are stuck with a purely knowledge based definition of mental states as their starting point. The moment they start trading in subconscious and semi-mental states, they pull their conceptual foundation out from under their feet and can no longer maintain the ontological separateness of the mental that defines their position.