Sunday, September 23, 2007

Agents and reasoning.

Continued from here.

Central to my account of reasons-explanations of action are the values and rules of practical reasoning. Reasons (goals) are connected to actions (means) because the latter are appropriate to the former. This is cashed out in terms of those actions for those reasons either conforming to a rule of practical reasoning or realizing/exemplifying/etc. some value of practical reasoning. Furthermore, though, reasons must be objectively valuable in order to be genuinely reasons, and in order for an agent to begin trying to reason a way to achieve the goals. So, it follows that I need an account of practical reasoning and of agency such that actions follow from goals as reasons, and such that agents can engage in this sort of reasoning as well as be able to recognize and take action towards certain goals. Let me start with agency.

I should note that empirical results from cogsci and the like will probably require altering this account. A priori reasoing is a fine place to start, but it needs to be sensitive to the a posteriori. So, this is just an initial sketch. That said, I have thought for a while that the intuitionists of the early 20th century (from HA Prichard and GE Moore in the early part of the century, all the way up to WD Ross in the middle part) were on to something with the claim that moral value is perceived. We can, after all, just see that some things are right or wrong. But I don't think the intuitionists took this quite far enough. There really is no hard line between mind and world: the world is the way it seems to us, and the way things seem to us is constantly affected by the way they are. So, in part, our perceptions are shaped by things, but things are also shaped by our perceptions. I don't have a choice whether I feel the floor under my feet, but that what I feel is a floor comes from me. Similar points can be made for other sensory modalities. Given that, while it is true that some things -- shape or size -- seem imposed on us from outside, perceived because of the way things are, others -- colour and value -- must be perceived in things around us because of the way we are. Of course, these are not complete categorizations; and any line we draw between these categories is tentative, porous, and subject to revision.

But, from the above it follows that perception of value is not mysterious or unanalyzable. It is, instead, fundamental to the way we make the world (and the world makes us): perception of value is as fundamental as any other peception. The world to us is value-laden; and, since this is the only world we can ever know, the world itself is value-laden. How we do this is, of course, a fully empirical question.

It must follow that agents are constituted as value-perceivers. But, what is the value they perceive? If explicit reasoning has any relevance to what we do -- and surely it does -- it must be the case that the values we perceive are the values we use in reasoning. That is, what I am calling goals -- perceived values -- must be able to serve as premises in practical reasoning. Practical reasoning must have values in its premises because of the standard is-ought problem. Practical reasoning's conclusions are oughts -- what I should do, in order to achieve certain goals. If its premises are not normative in any way, then this sort of conclusion will never follow. But this means that practical reasoning never has relevance for our actions: it is not the source of the oughts that bind us. Since this is absurd, by reductio at least one of practical reasoning's premises must have normative content.

Now, reasoning itself is a normative activity. There are rules we must follow to proceed from premises to conclusions correctly. So, there must be rules to practical reasoning, and we agents must be such that we can follow these rules. As said, we must be constituted to perceive value; it would be parsimonious, then, if the values we perceive were one and the same as the rules we follow. For then rather than two processes, there is only one: perception is rule-recognition. (I'm eliding a little here, I know, but I think the basic point sound.)

There are two possible objections to this move. First, recognizing and following rules are different processes, and it is rule-following I should be concerned with. Second, values and rules are different in other normative areas, such as morality and aesthetics, and thus should be different in the basic practical realm.

I am not convinced by the latter. It strikes me as a fairly dogmatic assertion of a claim, and not an actual argument. For there to be an argument, there would have to first be some important difference bewteen following rules and realizing value. But is there? Take some rule: you should never lie to people. This can correlate, though, to a negative value: lying is bad. Or take another rule: you should be charitable. The value is obvious: charity is good. Generally speking, I think we can capture any rule, one should (or should not) φ, with a value, φ-ing is good (or bad). While I realize this is simplistic, it is at least suggestive that the distinction between rule and value is not a hard one.

The former objection is an instance of a more general problem for me. While I have said both that value is perceived and thus used in practical reasoning, and that pratical reasoning yields conclusions about what one shoudl do, there is yet a gap in both. One can fail to reason practically, and one can fail to follow through on the conclusion of a piece of practical reasoning. I suspect that causation will bridge these gaps. That is, I reason practically if I am caused to, somehow: probably something to do with the connections between the perceptual and higher-level cognitive systems. So a full -- not simply reasons -- explanation of action has a causal element. Similarly, I act on the conclusion of a piece of practical reasoning if (essentially) I cause myself to -- or, in older language, I will myself to. (Not language I use because I don't agree with the implied faculty of will, any more than the implied faculty of reson that does reasoning. Reasoning and willing, in my view, are activities of complex physical systems like us; there is no one independent and unique centre that is responsible for any particular activity.) Since this goes into action more generally, not just actions as they are done for reasons, it is beyond my scope here. But it is clearly an interesting and improtant question in its own right, and I do not mean to deny that.

Back to the main point. We know something about what agents must be like -- as perceivers and reasoners -- but how do they reason? What rules do they follow and what values do they realize? This is a difficult and general problem. Some rules are obvious: the rules of logic. Some values are also obvious: the aforementioned propriety, as well as basic pratical values (e.g., what is achievable). Beyond that, the answers are probably quite complex and difficult and, again, beyond the scope of my project. However, as long as there are rules and values of practical reasoning, then my project is basically well-founded. I may be wrong on particulars -- any given φ may not be done for the ψ I cite, bcause I don't have a full theory of practical reasoning -- but the general shape -- the dependence on practical reasoning -- seems supported.

Next week, I'm going to finish up this series and talk about the distinction between the practical and the moral; motivation; and some ontology and epistemology, just for fun.


Ron said...

things are also shaped by our perceptions. I don't have a choice whether I feel the floor under my feet, but that what I feel is a floor comes from me

How does that shape/change the floor? Are you actually saying that objects are changed by our thoughts? Something like solipsism on a miniscule scale? Will it's atoms adjust in some way, even at the subatomic level? I mean: you can *call* a floor any sound you feel like attaching to it, but I don't see how what you call a floor (or how accurately you perceive it) is gonna change the floor one tiny bit.

ADHR said...

The idea is this. There's some object, a whatever-it-is. We attach various concepts to it -- so, we may consider it to be an instance of our concept of "floor". Now, that has some implications for what we are supposed to do with the object. You can sleep on a floor. In our culture, you're not supposed to eat on the floor. You can walk on a floor. Floors exist in buildings; if there's no walls or other features typical of buildings, then it's not a floor at all -- it's the ground. And so on.

That's the basic point I'm making. Observation is conceptualized in a way that we can't escape. And in conceptualizing observables, we create norms governing the use and observation of the observables.

From this, as I tried to argue (quickly), it follows that we never really know the thing-in-itself -- only the thing-as-it-appears.

Ron said...

So the thing is not shaped by our perception, just that our perceptions affect (are) what we know of it.


ADHR said...

Not quite. That's where I start, but I then take a sharp turn. Not only are our perceptions/conceptualizations all we know of objects, but they are all we are entitled to say the objects are. That is, we have no warrant for claiming there is any more to objects than what we perceive and conceive.

Ron said...

We think, therefore they are?

I don't see how that holds. We can know by past experience that the earth, say, was always a rough sphere even when some of us thought it was flat. It didn't morph according to our perception, just as it doesn't swing back and forth in shape even if some folks think it's flat.

Would we not be justified in surmising, then, that "the universe" is pretty much the same way. That is, regardless of how we perceive it or how accurately we perceive it, it is what it is and it has the qualities it has.

ADHR said...

Sure, we know now that the Earth is, was, and always has been roughly spherical. Before we knew that, though, conceptualizing it in other ways (e.g., flat, vortex, perfect sphere) was perfectly cogent. Until we had something we couldn't explain using our prior set of concepts and perceptions -- that is, a contradiction in the set -- we didn't have to change anything. Once the contradiction really set in, we had to re-conceive things. Once we'd re-conceived things, this re-conceptualization became the right one, and the old conceptualization became a mistake. If in future there's some further conflict in our conceptualizations -- say it turns out that a roughly spherical Earth contradicts some observations we make of distant stellar motion -- then we have to resolve the conflict by revising something again. And however we revise it, that revision will become correct, and what it was revised from was always and forever a mistake.

The idea isn't that there's no real universe out there, somewhere. The idea is that we never get direct access to that universe, ever. All we get access to are conceptual schemes, structures of theory that help us make sense (indeed, that make it possible to make sense) of the unfiltered stuff that bombards our unconscious cognitive systems. The "stuff" is there, but we never get to it. And since we never get to it, we're not entitled to say anything about what it's like -- even whether it differs from or is the same as the way it seems to be.

Given that, the distinction between the way things "really" are and they way the "just seem" to be is moot. There's no really real -- there's just the real for us. All that exists is what we can know. After all, if we can't know it, how can we know it's even there?