Wednesday, April 22, 2009

Does the Argument from Reason have a basis in Peirce?

See this discussion on Bill Vallicella's blog.

24 comments:

Doctor Logic said...

Sure. Reasoning is normative. And you can't get a rational justification for reasoning without a circular argument.

But the dodgy part is the idea that there's no will unless there's "free" will, and that reasoning cannot be deterministic.

Ignore physicalism for a moment. Suppose there's some sort of deterministic dualism. Can't there be a deterministic reasoning system?

Suppose there's an angel that can reason. Once it identifies its premises and axioms, it always reaches the "necessary" logical conclusions. Wouldn't the angel be a deterministic reasoner? Adding non-determinism to the angel just makes the angel a faulty reasoner.

Ilíon said...

And it's not a particularly difficult insight -- for instance, even I, whose "formal training" in philosophy consists merely of an intro-to-philosophy class (*) in college back in the late 70s, seem to have figured it out on my own (**).


(*) Which class consisted mostly of a bewildering Beastiary of Philosophers-Through-the-Ages, rather than anything I'd care to sink my teeth into.

(**) At the same time, I've read several of Lewis' books and papers, and a few of Plantinga's papers, and, naturally, the flash of insight I had is probably derived from them. I discovered your blog because I'd thought to myself, "Surely, I am not the first person in the history of thinking to have realized this? (What a scarely thought that would be!)" So, in trying to search the web to find that my idea, and the argument entailed by it, is not sui generis, I came across you

N said...

I'm not sure about a basis (I haven't read enough Peirce to know), but I think his views about mind, science, and reality do at the very least have the 'same spirit' as the argument from reason.

Finney said...

A while ago, I suggested that the afr is analogous to the moral argument from free-will.

I think this argument makes that point clear:

(1) if someone is determined to his actions by external forces he doesn't have free-will. (2) then he cannot voluntarily guide his actions to conform to moral values. (3) then what we call morals are merely epiphenomenal projections of a behavior that develops entirely without regard for objective morals. (4) Morals have no objective reality; they're the creations of a a-moral behavior.

(2') if someone doesn't have free-will, he cannot voluntarily guide his beliefs to conform to rational norms or logical laws. (3') then what we call logical laws are merely epiphenomenal projections of a mental behavior that develops entirely without regard for logic or rational norms. (4') logic has no objective reality; it is the construction of non-rational mental behavior

Crude said...

I could agree to analogous, but I don't think the AfR relies on libertarian free will to work, does it? It seems that the argument could be made with no reference to free will whatsoever - it's a question of what causes the change less than whether the change was free, and the argument attempts to show that whatever is causing that change, mechanistic materialism cannot be it.

At least, that's my understanding of it.

legodesi said...

I think the afr says that what causes our beliefs are reasons, purposes. The argument is that teleology is a basic feature of people. If people are determined by physical stuff, teleology cannot be basic. So if teleology is a basic feature of people, then the aspect of people that is purposeful cannot be fully determined by natural causes. It has to be free of it. So free-will is a necessary pre-condition to be rational. Not the only pre-condition, but one of them. That's my take on it.

Ilíon said...

Crude: "... but I don't think the AfR relies on libertarian free will to work, does it? It seems that the argument could be made with no reference to free will whatsoever - it's a question of what causes the change less than whether the change was free, and the argument attempts to show that whatever is causing that change, mechanistic materialism cannot be it."

The AfR (and its penumbra) denies -- perhaps I should say, refutes -- all types of determinism (or "mechanism"); no, it both relies upon the reality of "libertarian free will" and refutes the denials of "libertarian free will." This is not a viscious circle, as I'll explain (though, one fully expects persons of a certain mind-set to refuse to see it).


1) Definitially, entities which do not "possess" free-will (as we clumsily phrase it -- better would be to say "entities which are not free-wills," which is to say, entities which are not beings of whom it is appropriate even to say "who") simply cannot reason.

1a) 'To reason' just is to voluntarily submit one's will, in accord with the "laws of logic," to the truth one sees/understands.

For example, if the truth of some proposition 'p' logically entails the truth of further proposition 'q' and if one says to oneself (or to another), "Yes, I can see that the truth of proposition 'p' logically entails the truth of proposition 'q'," then one still cannot honestly be said to have reasoned if one's presumptive act-of-will is determined by anything "outside" one's will/self: only free beings may act.

1a.1) Persons refusing to submit their wills to truth-in-accord-with-logic may certainly be engaging in rationalization, but they are not, properly speaking, engaging in ratiocination.

1b) Moreover, the first step in the process of reasoning -- that of understanding that the truth of proposition 'p' logically entails the truth of proposition 'q' -- *also* cannot be done by deterministic entities: only free beings may understand.


2) Atheism -- the denial that there exists a Creator-God -- if it is even to attempt coherence, implies naturalism/materialism, much as naturalism/materialism asserts atheism.

2a) If atheism is the truth about the nature of reality, then likewise is naturalism/materialism the truth about the nature of reality.

2b) If atheism is the truth about the nature of reality, then there exists no such thing as "freedom" -- there are no acts, but only events; and these events are either determined or, possibly, random (and there exists no one able to tell which is the case). But, in no case is any event ever an act, it's simply logically impossible were atheism true.


3) If atheism is the truth about the nature of reality, then there can exist no beings capable of reasoning.

3a) There exist beings -- namely, ourselves -- who can and do reason.

3b) Therefore, because we see the truth that there do exist beings who can (and do) reason, we see/understand the further truth that atheism -- the denial that there exists a Creator-God -- is necessarily false.

3c) Therefore, since we *see* that atheism is necessarily false, we have the logical (and moral) obligation to complete the act of reasoning by admitting this truth -- and further acting accordingly.

Crude said...

That helps a lot, Ilion. Thanks!

The AfR continues to impress me. "One" argument that seems to do a whole lot.

Doctor Logic said...

Ilion,

You say that, in a physical world, there is only determinism and randomness. I agree. What I would like to know is how you can get around the dichotomy in any consistent worldview.

Forget materialism for a moment, and consider the following in any picture. By definition, the outcome of an event is deterministic when the outcome is determined by things in the past or things that are outside of time (constants). For example, if I were deterministic, the temporal events that led up to my writing this are within time, while other influences like the gravitational constant are outside of time (i.e., not labeled by a time index). Anything outside of time appears to us to be a constant.

The set of all things that determine an outcome, i.e., the set of all things in the past and all things outside of time, is a rather big set! The only stuff not in this set are things in the future. If we assume the future does not exist or does not impact the past, then there is absolutely nothing whatsoever left to determine the outcome of an event. (And if the future did determine the past, that wouldn't help avoid a deterministic picture.) Therefore, to the extent that an outcome is non-deterministic, an event is fundamentally random. Fundamentally random as distinct from epistemically random. Fundamentally random because the outcome literally depends on nothing whatsoever.

Now, I have not assumed materialism in this analysis. And I see no way of getting around this argument. You can't use the supernatural to get around the argument any more than you can use the supernatural to create square circles. So even if you accept dualism or the existence of God, you can't get around the fact that determinism and randomness are a complementary pair. There's no "free" alternative. To invoke the concept of "free" as a third option after determined and random would be like me saying that an argument can be consistent, inconsistent or "free".

Crude said...

Since DL is taking the tact that libertarian free will can't work, rather than defending LFW, I'm going to take a different tact.

The (at least, as wikipedia formulates it) AfR doesn't say anything explicitly about will - it seems aimed entirely at the mechanical material-monist metaphysical perspective, and in that vein it's devastating.

But taking an aristotilean/thomistic view of persons and agents seems like it goes a long way towards providing a (naturally, non-materialist, and perhaps necessarily theistic) solution to the AfR that doesn't involve reference to LFW (it doesn't deny or affirm it, it just makes it irrelevant to the question, if I understand it right.) Once you deny a purely materialist-mechanical conception of the world and recognize that certain things naturally tend towards certain goals, the justification and basis of reason falls out practically on its own. Back to Aristotle and Aquinas we go.

And just to work this into the original post, I think Peirce (not sure if he was an Aristitotilean, but he certainly seemed to have been a realist about universals, etc) helps indirectly back this up with his neglected argument for God. Namely that the advancement of science more and more suggests that the universe itself is a product of a mind (since it has a fundamental intelligibility, etc. And I think this has been radically strengthened by computer simulations), which may in turn bolster the justification for seeing formal/final causes not just in human concept but actually inhering as part of nature.

Doctor Logic said...

Crude,

Actually, Vallicella mentioned determinism in the referenced post.

As for the AfR, we can go back to the Pierce quote:

Now, that which is uncontrollable is not subject to any normative laws at all; that is, it is neither good nor bad; it neither subserves an end nor fails to do so.What does "uncontrollable" mean? AfR proponents can't help themselves see teleology everywhere, so they jump to the conclusion that uncontrollable means non-physical, and non-deterministic. But does that make any sense? Why should control always be non-deterministic?

Here's the root cause of the faulty intuition. Suppose I say that that, yesterday, I could have decided to eat the beef instead of the fish. When I say this, I don't mean that if I time traveled back to the point of the decision (so that the initial state was identical), I could have wanted the fish as I did the first time, but yet chosen the beef anyway. If I meant that, I would actually be denying that I had the free will to eat what I want.

Of course, what I really mean is that, if I time traveled back to the decision point, AND the situation differed in that I wanted to eat the beef instead of the fish, then I could have eaten the beef instead of the fish.

But this linguistic failure gets twisted in people's heads so that they think that determinism contradicts free will. It simply doesn't contradict free will. Determinism is necessary for free will.

What determinism does contradict is the sort of libertarian free will that makes sense of Christian justice. As I showed above, libertarian free will is plainly incoherent.

And I think the confused intuitions about free will tangle the AfR debate. If will and control are not fundamental, but reduce to mechanics, that doesn't mean there's a lack of control. There's still an "I" that is controlling things.

Crude said...

I know Bill V's post references determinism, but other formulations of the AfR (Including the wiki version I referenced, and most versions I've come across) do not. It's possible reason requires more than simply a non- or more-than- material basis, but it still seems to me that the AfR as Victor Reppert gives it does not deal with LFW. Either way, there are multiple arguments about Christian justice within LFW and compatiblist schemes, so that's not my concern here.

The AfR highlights that if you're going to reduce everything to mechanistic materialism, you've removed reason from the picture straightaway - precisely because of what materialism leaves out to be what it is. If you're placing intentionality, aboutness, etc at the ground level of reality or making these things a real and fundamental feature (which you need to if you're going to have reason), you're no longer dealing with a purely mechanistic materialism, and arguably (unless the definition changes yet again) not naturalism either.

Finney said...

Doctor Logic,
I'm not sure if I entirely follow your reasoning.

The trouble with physical determinism is that our actions are determined by non-rational agents, and so they are not rational. If you define determinism to make it such that our minds can freely make choices to decisions (making our decisions determined by our mind's free choices) then there's no problem, since those actions are determined by a rational source.

Doctor Logic said...

Finney,

The trouble with physical determinism is that our actions are determined by non-rational agents, and so they are not rational.No, this is a non sequitur. We do not expect components of the whole to have all the properties of the whole. Molecules do not have "oceanic" properties, and yet oceans are made of molecules.

However, this is beside the points I was trying to make.

Let's take physicalism out of the picture. Let's assume a deterministic dualism.

Let's imagine a being who cannot fail to reason correctly. Such a being would be deterministic. Pose a logical problem to this being, and it will reason to the consistent solution every time. (Or, at least, it will do so for finite proofs.)

Now, either you have to say that such a picture is somehow impossible, or you have to admit that such a being is reasoning. In which case, reasoning makes sense in a deterministic system.

Once you admit that a deterministic system can reason, then we're no longer particularly concerned with the mechanism (if any) of the reasoning process. As long as the being is causally connected to the abstractions it is reasoning about, there's no faulting the being's reasoning.

Finally, having reached this point, showing the possibility of physical minds is a matter of showing that a physical mind can create abstractions. I expect you think that physical minds cannot create abstractions, and so your dualism claim sticks. However, I'm trying to get to a precise statement of what the AfR is really saying about physical minds.

Crude said...

It wouldn't just be abstractions. Intentionality and 'aboutness' would play a key role as well, and quite possibly realism about universals as well.

Doctor Logic said...

Crude,

It wouldn't just be abstractions. Intentionality and 'aboutness' would play a key role as well, and quite possibly realism about universals as well.Agreed, although I think universals would fall under abstractions.

In studying a rabbit, I automatically learn to recognize other instances of rabbit, not only the one instance I've been studying. This means that I implicitly create an abstraction of rabbits in my mind. The rabbit abstraction is also a universal rabbit recognizer. It's universal because it is capable of recognizing rabbits that do not actually exist. I can even reason about rabbit universals by referring to "that which my rabbit recognizer would recognize".

Finney said...

"We do not expect components of the whole to have all the properties of the whole"

This is completely different. I'm not saying that rational thought cannot be composed of non-rational constituents (though my argument doesn't rule this out); I'm saying that rational thought cannot caused by non-rational causes. It's nothing to do with non-rational properties of a rational "whole".

"In which case, reasoning makes sense in a deterministic system."

I'd say that this is only possible of the being that cannot fail to reason correctly is being determined by an external rational agent who has determined the beings actions. I would re-state that determinism itself is not the issue, but determinism of ultimately non-rational causes.

Ilíon said...

Finney: "I'm not saying that rational thought cannot be composed of non-rational constituents (though my argument doesn't rule this out); ..."

Ultimately, rational thought (at any rate, the rational thought in which we human beings engage) is rooted in, and depends upon, the intuitive, which is non-rational (definitionally).

Doctor Logic said...

Finney,

I'm saying that rational thought cannot caused by non-rational causes.Again, I disagree, and would like to see a more formal argument.

For starters, I'll throw in another counterargument.

Consider the Pythagorean Theorem which follows from the axioms of Euclidean geometry. A rational belief in the theorem is caused by the mathematics. But the mathematics isn't rational. At best one might consider the mathematics to have been caused by something rational, but let's set that aside for a moment.

So, here I am, rationally concluding that the PT is correct. You can think of my mind as exploring the logical relationships of Euclidean geometry, and, consequently, my mind's conclusion is determined by the mathematics. The extent to which I fail to conclude the PT is correct is the extent to which my mind fails to be determined by the mathematics.

There's one final step in the counterargument. If logical structures and mathematical systems reflect laws of computation, then a person has a causal link to the theorems by being in possession of a computing mechanism. This takes care of what I set aside a couple of steps earlier... logical structures are not Platonic things caused by rational thinking, but exist like laws of computing, waiting to be discovered by computers (i.e., us). The idea that contingent mathematical relationships (they are all contingent upon their respective axioms) are discovered is perfectly plausible.

Again, an important point is that, once one concedes that minds are compelled to conclusions by logical structures, it doesn't really matter what the mechanism of that compulsion is. "Freedom" is irrelevant to reasoning, and, hence, mechanistic systems can reason at least as well as non-mechanistic ones.

Doctor Logic said...

Ilion,

Ultimately, rational thought (at any rate, the rational thought in which we human beings engage) is rooted in, and depends upon, the intuitive, which is non-rational (definitionally).Well said. And intuition is not necessarily non-physical.

Ilíon said...

Curiously Mis-named Person: "Well said. And intuition is not necessarily non-physical."

*eye-roll*

Crude said...

Even in the AfR, it seems to be an unspoken assertion that there is a 'mechanism' for reason. That's trivial - it even draws out what the AfR's argument highlights as that 'mechanism'. (How it works - it relies on a source that did not begin to exist, etc.)

But 'a' mechanism is not 'any' mechanism. Not any old thing you can call a 'mechanism' will work. And a thoroughgoing mechanistic materialism, won't. Something that involves the physical can escape, but not the physical alone. (Possibly a radical redefinition of physical/material a la Strawson can make some headway, but if panpsychism is the answer, the question has done its job and then some.)

Merlijn de Smit said...

I believe that some kind of Argument from Reason is implicit in the following passage by Peirce:

(Peirce argues, as a philosophical realist, that sensory feelings are inherently communicable)
"I hear you say: »This smacks too much of an anthropomorphic conception.« I reply that every scientific explanation of a natural phenomenon is a hypothesis that there is something in nature to which the human reason is analogous; and that it really is so all the successes of science in its applications to human convenience are witnesses. They proclaim that truth over the length and breadth of the modern world. In the light of the successes of science to my mind there is a degree of baseness in denying our birthright as children of God and in shamefacedly slinking away from anthropomorphic conceptions of the universe."
(Collected papers I:316, italics by me).

Doctor Logic said...

Crude,

Even in the AfR, it seems to be an unspoken assertion that there is a 'mechanism' for reason. That's trivial - it even draws out what the AfR's argument highlights as that 'mechanism'. (How it works - it relies on a source that did not begin to exist, etc.)I neither think it's trivial nor do I think there's an unspoken assertion that there's a mechanism. Mechanisms are predictable, deterministic and explicable. In contrast, the supernatural is inexplicable, not predictable, and not a mechanism. I don't think you're using the term mechanism correctly. Showing that reasoning is a mechanism is a non-trivial and very important step on the way to disproving the AfR.

But 'a' mechanism is not 'any' mechanism. Not any old thing you can call a 'mechanism' will work. And a thoroughgoing mechanistic materialism, won't.You're just stating the claim, presumably because you think the AfR is making the argument for you. The AfR is founded on imprecise and wrong-headed intuitions, most notably about mechanisms. I'll mention the faulty intuition again...

When I say I chose Y, but that I was free to choose X over Y, I mean that I could somehow have chosen X when I really meant to choose Y. (FAULTY)This intuition is used to reject determinism and mechanism, and arrive at the faulty conclusion of the AfR. It's ironic because it looks to me as if that faulty intuition denies our own capacity for intention.

That intuition should really be written as:

When I say I chose Y, but that I was free to choose X over Y, I mean that, had I meant to choose X, I would have chosen X instead of Y. (CORRECT)Yet we have no intuition that tells us we were free to prefer X over Y when we in fact preferred Y over X.

For example, when I say that I chose not to stick my hand in boiling water, but I was free to choose to stick my hand in boiling water, I really mean that, if I had meant to stick my hand in boiling water (because I desired the consequences), then I could have stuck my hand in boiling water.

But I have no intuition that I could, at the time I made that past decision, want to stick my hand in the boiling water. And so, in my universe, given my state at the time, I had to choose not to put my hand in the boiling water.

Nowhere to be found in this analysis is the conclusion that I cannot be a deterministic mechanism. And if I am a deterministic mechanism, I can just as well be a physical mechanism as long as that mechanism can predict future outcomes, prefer one outcome over another and execute the corresponding choice.