Thursday, April 22, 2010

On The Necessity of Mental Causation

This is from my reply to Keith Parsons in essay "Some Supernatural Reasons Why My Critics are Wrong" (a title that was given to my essay by someone else), in Philosophia Christi (Volume 5, no. 1, 2003).

But think for a moment about wjhat it is to be persuaded by an argument. If we are thinking in common-sense terms, we would hve to say that what goes one when we are persuaded by Parsons's argument that Arizona State will not be in the BCS this year is that we conisder the epistemic strength of the premises, the grounding relation between the premises and the conclusion, and then accept the conclusion as a result of conisdering the evidence presented in the argument. To be convinced by an argument is for the reasons presented in the to play a causal role in the production of the belief. If the argumetn is causally irrelevant to the belief, then we cannot say that the argument was persuasive. This can often be cashed out counterfactually: If I really am persuaded by Parsons's arugment, then it cannot be the case that I am such a partisan of the Arizona Wildcats that I would think the worst of the Sun Devils' prospects even if the Sun Devils had a Heisman trophy candidate at quarterback, oustanding and experienced running backs and wide receivers, a rock-solid offensive line, and was returning everyone from what had been the stingiest defense in the Pac-10 the previous year.

On the one hand, the reasons have to persuade me in virtue of their being reasons. The logical force of the argument has to have a causal impact on belief. It has to make a difference as to whether I form the belief or fail to form the belief in question. And that, by the way, is bound to make a difference as to what I do with my body. I am going to behave differently if I think the Devils have a good chance to take the Pac-10 title than if I don't. And that is going to affect what the particles in the physical world do. But if the physical is causally closed, that means that only the physical can affect where the particles in the physical world go, and, the physical is defined as lacking, at the basic level of analysis, the central features of the mental. So the only way this kind of causal relation could possibly exist, would be if we could analyze the mental in physical terms as a kind of macro-state of the physical. Just as the word "planet" is absent from physical vocabulary, but a whole bunch of particle-states add up to there being a planet, perhaps "S's belief that P" can be added up  from a set of physical states. But that seems to me to be just impossible. Add up the physical all you like, and you aren't going to get "S's belief that P." The physical leaves the mental indeterminate. Yet, if science is to be possible, is has to be determinate whether, for example, Einstein is plussing or quussing when he is adding numbers in the course of developing his theory.

So, I argue that you need mental causation for the possibility of science, but you can't get that without affirming what seems to be an implausible reductionism, that conflicts with the indeterminacy of the physical.

36 comments:

Doctor Logic said...

Victor,

Can you expand on this a little:

Add up the physical all you like, and you aren't going to get "S's belief that P." The physical leaves the mental indeterminate.

Why, exactly, does the physical leave the mental indeterminate?

Victor Reppert said...

One reason is because to make an intentional attribution involves the normative, and you can't add up descriptive statements and get to something normative.

With the physical, logical relations are not relevant. If we are talking about where this particle is and where that particle is, logical relationships are not involved. We are talking about the spatiotemporal location of objects. But logical relations are not localized. The exist necessarily. So you are trying to go from the contingent to the necessary, and that's not going to work.

I take it you are going to defend reductive materialism, and not the sort of non-reductive materialism that Clayton is advocating.

Doctor Logic said...

Victor,

I take it you are going to defend reductive materialism, and not the sort of non-reductive materialism that Clayton is advocating.

Well, yes, that is what I would do...

I wasn't sure if Clayton was advocating non-reductive materialism or just defending its plausibility (I didn't read the entire thread).

However, I think you and I would agree that NRM isn't very compelling. From my perspective, NRM is non-explanatory, and the concept of over-determination can be taken to absurdity. I might as well suppose that you over-determined what I'm writing here. (Perhaps over-determination will be more interesting to theologians!)

I can tell you why your claim about the mental being indeterminate isn't persuasive to me.

First, as a naturalist, I can't find a place to put objective normativity. This means that any normativity I encounter is only going to be objective with respect to the goals of an agent, and those goals might in-turn be due to evolved or conditioned inclinations. That is, an agent with goals has normativity of a subjective variety. If intentionality requires normativity (and that sounds like a possibility), then intentionality would be related to an agent's goal (e.g., its goal of comprehending the world it lives in). This picture doesn't threaten naturalism, as far as I can see.

As for logical relationships, they are not necessary in a vacuum. They are necessary only to avoid contradictions that would render the world unintelligible to us. Your wording suggests that the contingent-necessary distinction is a metaphysical one, but it's just a common or garden distinction. Necessity is always with respect to something. For example, the mutual exclusivity of my existing versus not existing is necessary for the world to be intelligible to me.

Once we admit that an appreciation for logical relationships is necessary for a function or ability (inference and intelligibility), then we're not contradicting naturalism.

Steve said...

DL,

I've never known what to make of the sort of line you are putting forward here. You say that for something to be comprehensible to us, it the laws of logical must apply to them.

You seem to be commiting yourself to the idea that some things may be real (but unintelligible) which involve logical contradictions.

Are you allowing this, at least in theory? If so, then for many that will immediately amount to a reductio of your position.

Steve Lovell

Doctor Logic said...

Steve,

Whether or not the world contains incomprehensible zones I can't be sure. However, that wasn't my point.

My point is that "logical is as logical does." Essentially, logical thinking is necessary for an agent (a) to be labeled 'logical', and (b) to be able to make inferences to intelligible facts.

Victor is phrasing our awareness of logical relationships to be an awareness of an ultimate metaphysical necessity, i.e., as if we are communing with some part of reality that physics cannot see. Yet, every system of physics has consistency embedded within it (at least at some scale). If physics didn't incorporate logical consistency, then there would be no causality at all. That is, physical laws tell us that "A implies B," never "A implies B and simultaneously not B". That wouldn't be a law at all.

If a biological cell can evolve to become sensitive to general physical properties, e.g., temperature, concentration gradients, etc., why can't a more complex structure evolve to become sensitive to the general physical property of consistency in causality?

Material minds can be sensitive to abstract logic if they have two ingredients. First, they need to be sensitive to specific instances of inference, i.e., inferring from specific rules to consequences. For example, "if there is thunder, then there is rain nearby. There is thunder. Therefore, there is rain nearby." We can imagine that a brain could learn to make this inference without learning about logical consistency generally. I don't think this is controversial.

Second, if the brain can make generalizations from the specific to the abstract, then it can learn about abstract logic. Though many philosophers won't be familiar with the physical mechanism that does this (auto-associative neural networks), I don't think this is controversial either.

My specific argument here does not provide an answer to the "hard problem" of consciousness (awareness itself), but it does answer the AfR with respect to logic. Zombies (if such beings existed) would make excellent inferences - they would simply have no conscious awareness of those inferences. We would not be able to say that zombies were illogical, or that they were insensitive to abstract logic.

Anonymous said...

We also wouldn't be able to say they actually make inferences. The only inferences that would take place would be our very own with regards to them.

As for whether intentionality is a problem for naturalism, it depends on how thoroughgoing the naturalism is or not. Have a look at naturalist Alan Rosenberg on this topic. From his response, with emphasis added:

It is of course obvious that introspection strongly suggests that the brain does store information propositionally, and that therefore it has beliefs and desire with “aboutness” or intentionality. A thoroughgoing naturalism must deny this, I allege. If beliefs are anything they are brain states—physical configurations of matter. But one configuration of matter cannot, in virtue just of its structure, composition, location, or causal relation, be “about” another configuration of matter in the way original intentionality requires (because it cant pass the referential opacity test). So, there are no beliefs.

No inferences either, since those would be an inference "about".

Anonymous said...

Correction: Alex Rosenberg.

Clayton said...

For the record (again), I wasn't advocating any view but pointing out that there didn't seem to be an argument against one of the materialist views on the table. I guess I'll just have to say that whenever a thread like this pops up.

Vic wrote, "One reason is because to make an intentional attribution involves the normative."

I've always thought that that sounded cool, but now I have two worries. First worry--most people reject reductive accounts of normativity and will deny that the 'ought' can be reduced to an 'is' regardless of whether the 'is' is about something natural or supernatural. The dominant view, I thought, was that you had supervenience of the normative without reduction.

Second, like I said, it sounds cool to say that an intentional attribution "involves the normative" but I'm not quite sure what that means. Don't animals have intentional states? I think dogs can have experiences and beliefs that have intentional content. I think the view that there are "oughts" that apply to the dogs _because_ they have contentful perceptual experiences or beliefs based on those experiences is just crazy. So, assuming that's right, what does the claim that "an intentional attribution involves the normative" mean?

If you say, well, the speaker who makes the ascription of the intentional state is under some normative requirement, that might be true. But, that's because speakers like us are under normative requirements whether we're talking about the mental lives of non-human animals or describing the weather. And, we wouldn't say that descriptions of the weather "involves the normative" in the sort of way that causes trouble for naturalism about clouds.

William said...

Clayton said:
--
I think dogs can have experiences and beliefs that have intentional content. I think the view that there are "oughts" that apply to the dogs _because_ they have contentful perceptual experiences or beliefs based on those experiences is just crazy.
--

I think that instinct may be a kind of "ought" with dogs?

Steve said...

DL,

I understand your response, and I also understand that the possibility of the world containing unintelligible zones was not your point ...

However it does seem to be a consequence of the view that your response embodies. That alone will, for many, be sufficient reason not to take your response seriously. Indeed, I deliberately phrased my query to you in such a way as to make such unintelligibility seem a live option. I assiduously avoided saying that your response seemed to commit you to the possibility of the logically impossible.

Such a "possibility" just is a consequence of the view your response embodies, but it seems clear to me that having such a consequence makes that view absurd, which is why I can't take your response to VR seriously on this point.

Is there something I'm missing here?

Steve

Rasmus Møller said...

William said:

I think that instinct may be a kind of "ought" with dogs?

Rasmus:

To paraphrase C.S. Lewis:

Ought we (dogs?) obey our instincts?

Doctor Logic said...

Anon,

Thanks for the Alex Rosenberg link.

Perhaps you can clarify how you're interpreting Rosenberg's claim about referential opacity. Presumably you interpret the claim to mean that there's a way in which:

"I believe that X"

can't be replaced by:

"The physical processes and configurations in my brain have a relationship with physical processes, configurations, and possibilities outside of my head."

My claim is that the substitution can be made, that the brain can make abstractions and generalizations, and that the brain can be aware of situations that don't meet its expectations. Part of this claim is that, once you can make abstractions of abstractions, you can abstract the world all the way to logical and mathematical systems.

What I'm used to seeing here on DI from AfR proponents are examples of how primitive machines (or worse, inanimate objects) are incapable of functioning like a brain in this respect. For example, Deep Blue does not make generalizations or abstractions, so pretending Deep Blue is a valid surrogate for human minds is not convincing to me in the slightest.

My question to you is, what example would you provide as an example of the kind of referential opacity Rosenberg is talking about?

Doctor Logic said...

Steve,

I understand your response, and I also understand that the possibility of the world containing unintelligible zones was not your point ...

However it does seem to be a consequence of the view that your response embodies...


I have an unfortunate tendency to tackle both horns of a dilemma (or both sides of the coin) at once, so brace yourself. ;)

First of all, I don't think I need to commit myself to the possibility of unintelligible zones. All I have to do is see the possibility of acting in a way that is not conducive to making an logical world intelligible (i.e., see the possibility of acting illogically).

However, I think it's worth exploring what grounds there would be for rejection of my claim if I did commit to unintelligible zones.

Consider the many worlds interpretation of quantum mechanics. In the MWI, every possible outcome of an event happens, but each outcome creates a new universe just for that outcome. Because (in the theory) our consciousness exists in only one universe, this is not a problem for us. From our perspective, only one outcome occurs for every event.

But suppose that there is a "bubble". Whenever we step into the bubble, we slip into one of the alternate (but nearby) MWI worlds in which JFK survived that infamous day in Dallas with nothing but a flesh wound. This bubble floats around the world, exposing us to an alternate and contradictory reality from time to time. This seems... manageable.

Of course, the more bubbles there are and the higher their density, the less comprehensible the world will be. If the foam of bubbles has infinite density, then the universe would become devoid of facts. Whatever. I just don't see why the possibility of a few bubbles (and consequent contradictions) is utterly absurd on its face.

Anyway, back to the other side of the coin. My real point is that, even in a logically consistent universe, it's not absolutely necessary for me to be a logical person. Indeed, I'm not a totally logical person. No one is. Being a logical person is necessary for having a perfect understanding of a logical world.

(We need to replace the world "necessary" in the English dictionary with "necessaryfor", so as to avoid these sorts of confusions.)

Logical necessity is an operational principle for making good inferences in a logical world. Nothing more.

The AfR complaint seems to be that physics is, at a fundamental level, insensitive to this principle. And I totally agree. Electrons don't have a conception of intelligibility, so they certainly can't be aware that certain kinds of relationships are required for intelligibility. The problem for the AfR is that this fact about electrons is irrelevant. What is relevant is whether there can be physical structures that can make inferences from the past to predictions about the future. After all, physicalism proposes a logically consistent universe.

Steve said...

DL,

Okay, first I didn't say that you were committed to the existence of unintelligible zones, but that your view committed you to the possibility of their existence. I don't think you've said anything to suggest otherwise. I don't see the relevance of the attempt to translate your view into one about illogical behaviour.

Second, on these "bubbles". I must first admit that I don't really know what you're talking about, but certainly on any natural reading of what you are saying these bubbles are just different zones where different natural laws may apply. If that is what you have in mind, then these are logically unproblematic. This would only be unintelligible in the relevant sense if you also affirmed the universal truth of laws of nature which fail to obtain in such a bubble. But certainly that's not a given. Your view commits you to the following:

(DL-ism) Possibly (P & -P).

The bubble scenario seems not to amount to P & -P, so while I agree with you in saying they don't seem totally absurd, I think you're response to VR commits you to possibilities much more radical.

If you think the bubbles are more radical, because in your example JFK both dies and survives, then it seems pretty clear that either this is nonsense, or it can be qualified in such a way as not to be an actual contradiction. If it's actual nonsense then I'm not happy with it's possibility as a consequence, and nor should you be. If it can be qualified to escape the contradiction then we are failing to consider sufficiently radical cases.

Steve Lovell

Anonymous said...

Doctor Logic,

My claim is that the substitution can be made, that the brain can make abstractions and generalizations, and that the brain can be aware of situations that don't meet its expectations. Part of this claim is that, once you can make abstractions of abstractions, you can abstract the world all the way to logical and mathematical systems.

You may misunderstand why I'm bringing up Rosenberg, so I'll clarify. Yes, I know you're claiming that physical processes can "have a relationship with physical processes, configurations, and possibilities" outside of your head. Say, this brain state is really "about" this other thing, or this given state of affairs, etc.

Here's the problem: Rosenberg is flat-out arguing your view is not a fully naturalistic one, precisely because you're saying - or so it seems - that brain states really and truly are "about" other things, and that presumably this isn't some useful but ultimately fictional way of speaking. (If it is, then you're not giving an account of intentionality of course, or a response to the AfR.)

My question to you is, what example would you provide as an example of the kind of referential opacity Rosenberg is talking about?

Well, you're claiming that you're not impressed with examples you normally see because they're simple machines or inanimate objects - and you even count Deep Blue as one of these. But those simple examples are (in my experience) given because the issue is not one of simplicity or complexity: Make the computer/software as "complex" as you wish. Ultimately all you have are (as Rosenberg said) "structure, composition, location, or causal relation". The only meaning/intentionality/aboutness that would be present is the (non-original, derived) meaning we assign to these things. (See Quine's inscrutability of reference.)

Are you saying that Quine's inscrutability of reference is incorrect?

Doctor Logic said...

Steve,

There's a difference between the following two statements:

(a) The cosmos is perfectly consistent and logical.

(b) Our thinking about the cosmos is consistent and logical.

My point is that (b) is "necessary for" the intelligibility of (a).

You are saying that if I reject the "absolute necessity" (necessity in vacuo?) of (b), then I admit the possibility that (a) is false.

However, your claim is known to be false. People who subscribe to (a) also believe that there are people think illogically. Therefore, (b) is not necessary in vacuo.

My claim that (b) is merely "necessary for" intelligibility is all I need to deal with the AfR.

That said, (and taking off on your tangent) I think your complaint against the possibility of unintelligible zones is closed-minded.

Consider this statement of yours:

If you think the bubbles are more radical, because in your example JFK both dies and survives, then it seems pretty clear that either this is nonsense, or it can be qualified in such a way as not to be an actual contradiction.

You say that the case in which JFK both dies and survives is nonsense, as if this should somehow discount the possibility of the situation arising. But, of course the situation in which JFK both dies and survives is nonsense. It's unintelligible to us, and that's the very point.

BTW, you're right that within each bubble, everything will seem consistent and hunky dory. The apparent contradictions occur when you cross bubble boundaries. If we pass through bubble boundaries faster than our ability to comprehend our environment (e.g., if the bubbles are a dense, frothy foam), we will lose all sense and intelligibility.

Here's another thought experiment. Suppose that God is upset with the world, and decides to destroy it by merging the logically contradictory quantum worlds together in a dense foam of MWI bubbles. And he plans to do this next Thursday. You seem to be arguing that God cannot possibly do this because, if he did, we wouldn't be able to understand the world after next Thursday (it would be nonsensical). But why does possibility hinge on intelligibility to us?

It seems that you're confusing the map with the territory. We would lack knowledge and sense of the world after next Thursday, but intelligibility to us isn't the glue that holds the universe together.

If I were to get hit by a bus on Tuesday, that would prevent me from having an intelligible picture of the world on Wednesday (cos I would be dead), but I don't think that makes the bus accident impossible.

Doctor Logic said...

Anon,

The only meaning/intentionality/aboutness that would be present is the (non-original, derived) meaning we assign to these things. (See Quine's inscrutability of reference.)

Are you saying that Quine's inscrutability of reference is incorrect?


Quine is saying that meaning is in use/usage (empirical/behavioral), not in the individual words. His is a view that strongly supports the naturalistic picture.

Whatever Quine says about machines applies equally well to us humans. So, for example, when you referred to "Quine" in your last comment, you might really have referred to "The Eiffel Tower", but you defined your terms in such a way as to mean the what I thought you meant as far as behavior and empiricism are concerned.

So, I'm wondering where this is taking you. Quine isn't laying down different rules for mechanistic minds and non-mechanistic ones. Whatever Quine does to destroy intentionality in machines, he simultaneously destroys in humans and angels. In effect, he's saying that intentionality is in usage.

Suppose you utter a simple sentence, like "My car is in the garage." I don't see how you can assign reference to all those terms without connecting those references to forms of usage.

Anonymous said...

Doctor Logic,

Quine is saying that meaning is in use/usage (empirical/behavioral), not in the individual words. His is a view that strongly supports the naturalistic picture.

Whatever Quine says about machines applies equally well to us humans. So, for example, when you referred to "Quine" in your last comment, you might really have referred to "The Eiffel Tower", but you defined your terms in such a way as to mean the what I thought you meant as far as behavior and empiricism are concerned.


It goes further than that. From the source:

"The doctrine due to Quine that no empirical evidence relevant to interpreting a speaker's utterances can decide among alternative and incompatible ways of assigning referents to the words used; hence there is no fact that the words have one reference or another."

[...]

"However ‘inscrutability’ is not quite the right term for Quine's doctrine, since the term implies something real but unknowable, whereas it is Quine's position that terms have no real unique reference."

In other words, to accept materialism (naturalism) means that there is no actual 'thinking about' or 'reasoning about'. There is no actual intentionality. No one really adds in a naturalist world, no one does science, and no one reasons.

So, I'm wondering where this is taking you. Quine isn't laying down different rules for mechanistic minds and non-mechanistic ones. Whatever Quine does to destroy intentionality in machines, he simultaneously destroys in humans and angels. In effect, he's saying that intentionality is in usage.

Actually, Quine is saying what must follow given a materialist (naturalist) view of the world. He's not saying "intentionality is in usage", he's saying that intentionality isn't really there at all - it can't be, given those materialist/naturalist commitments.

Suppose you utter a simple sentence, like "My car is in the garage." I don't see how you can assign reference to all those terms without connecting those references to forms of usage.

Again, Quine's view isn't that you can truly assign reference so long as you 'make connections'. You can't assign a determinate reference at all. Terms "have no real unique reference", because the physical is unable to have this, given what matter is. Thus, if we are entirely physical (and we'd have to be on naturalism).. well, you see where this goes.

Doctor Logic said...

Anonymous,

I don't think your interpretation of Quine is correct. For example, I think Quine thought that an individual's version of meaning was exempt from inscrutability, and yet he remained a naturalist. If your interpretation were correct, Quine should have defected from verificationism, yet he didn't.

I think the way to look at this is by analogy to formulations of a scientific theory. We could have many formulations of a scientific theory, all of which predict the same (correct) experimental results. No formulation is empirically better than any other (although efficiency of calculation may be different). None of the different formulations are wrong, so there's no unique formulation. What you are doing (by analogy) is saying that, if there's no unique formulation, there's no scientific theory at all!!

Of course, if we have differing formulations, we have different means of computing the same thing, and we can each independently know what we should expect to see happen next according to the theory.

However, setting that aside for just a moment, I would like to hear your alternative story.

Quine is not so much assuming that our minds are naturalistic, but that we have only naturalistic clues for translating or learning languages. Are you saying you're getting non-naturalistic information when you learn, say, French?

Like me, you grew up learning your native tongue from empirical data passed on to you by your parents. You were no better off in such circumstances than the anthropologist Quine speaks of. So, in your picture, what fixes your translation of English or fixes your references?

Anonymous said...

Doctor Logic,

For example, I think Quine thought that an individual's version of meaning was exempt from inscrutability, and yet he remained a naturalist. If your interpretation were correct, Quine should have defected from verificationism, yet he didn't.

There's another possibility: Quine's views on the mental were ultimately incoherent, but he stuck to them anyway. From the SEP entry on intentionality:

"From the same conclusion, Quine (1960, 221) presented an influential dilemma with both epistemological and ontological implications. The first horn of the dilemma is to accept the “indispensability of intentional idioms and the importance of an autonomous science of intention” and to reject a physicalist ontology. The second horn of the same dilemma is to accept physicalism and renounce the “baselessness” of the intentional idioms and the “emptiness” of a science of intention. This dilemma has been influential in contemporary philosophy of mind."

What you are doing (by analogy) is saying that, if there's no unique formulation, there's no scientific theory at all!!

I'm saying what follows given naturalism (materialism) and a commitment that there is no actual intentionality and/ir aboutness. How can you really have a scientific theory when 'aboutness'/intentionality is unreal? How can you do it when there's no fact of the matter about what we're thinking about (as there would have to be, assuming physicalism)?

Note that given the above, Quine isn't saying that you can't accept "indispensability of intentional idioms and the importance of an autonomous science of intention". He, at that specific juncture, is just arguing you can't really do so and be a naturalist.

Are you saying you're getting non-naturalistic information when you learn, say, French?

[...]

So, in your picture, what fixes your translation of English or fixes your references?

"Non-naturalistic information"? We have Rosenberg and Quine saying that there is no 'aboutness'. There are no beliefs. What kind of information are you saying I could have ever really gotten, given that? Information "about" something? When you say 'in your picture', are you asking for my belief? How do I learn, without learning about?

As Victor has said, there's multiple responses to this available if naturalism is given up. But showing what follows given a consistent naturalism goes a long way towards encouraging a look elsewhere.

Steve said...

DL,

First, I think you've probably been misled by my use of your own terms. I'm merely using the word "unintelligible" as a short-hand for logically contradictory. I don't have any problem with humans not being able to understand the world. My problem is with accepting the possibility of the world really being, in itself, contradictory.

I don't follow your comments on (a) and (b) and the latter being necessary for the former.

You seem to be saying that if people are stupid then the world is incoherent. Perhaps your thought is: irrational behaviour is behaviour which is rationally inexplicable, but the existence of rationally inexplicable behaviour entails that the world is "unintelligible".

If this is your thought, then while I think something like this is true, I don't see it's relevance.

As I see it, your view commits you not just to the possibility of (b)'s falsehood but the possibility of (a)'s falsehood ... and not because of (b)'s falsehood ... but directly.

On your view ... "that can't be true" is really short hand for "if that were true we wouldn't be able to understand it".

In other words, the "can't" is a feature of thought not of reality. In which case you implicity allow the falsehood of (a) directly and not as an implication of the falsehood of (b)

Also, I don't care about how things seem in each bubble. I care how things are in the actual world, of which such bubbles are imagined either to be or not to be a part. If they are not imagined to be part of the actual world, then obviously that is not a problem. If they are part of the actual world (in the thought experiment), then you say that it is both the case that

(JFKa) JFK survives
(JFKb) JFK dies

Now a person moving between two zones one containing JFK and another containing a JFK counterpart is logically possible, and those two people may suffer different fates. However, let's fix the reference of JFK and then follow that individual. He either dies or survives. Follow him around. Perhaps imagine you are JFK. Could it really be the case that the JFK who's place we are imagining our imagining ourselves in, both dies and survives. Or is it merely that we can't imagine it?

I agree that I can't imagine it. If you think this isn't a good reason to reject the possibility, then you are undermining all reductio arguments. Do you really wan't to do that? Is that a price you are willing to pay for avoiding the conclusion of the AfR? Be my guest.

All rather reminds me of a colleague advancing the following analysis of logical possibility:

(LP) Some proposition P is logically possible if and only if either (a) we can imagine a world in which P or (b) P is true.

Seems just as crazy to me now as it did then.

Steve Lovell

Doctor Logic said...

Anon,

Here's a quote from the SEP article on Reference:

Nevertheless, Quine does not go so far as to say that our words fail to refer in any sense. His view is rather that it makes sense to speak of what our words refer to only relative to some purpose we might have in assigning referents to those words.
...
Quine's views on underdetermination can be applied to one's own language. The result is that the available evidence no more forces the speaker to the conclusion that by ‘rabbit’ he means rabbits, than it forces him to conclude that by ‘rabbit’ he means undetached rabbit parts or time-slices of rabbits. If a speaker observes himself using the word ‘rabbit,’ the evidence he amasses will give equal support to all three theories, as well as to many others. So, according to Quine, for any given body of empirical evidence, there will be numerous competing theories as to what the words one uses refer to. And there will be no principled way of adjudicating between these theories.

Dualism doesn't solve this problem. Dualism is an unprincipled, non-explanatory way of non-adjudication. Dualism just says that reference is absolute, and we magically know what the references are.

You say:

I'm saying what follows given naturalism (materialism) and a commitment that there is no actual intentionality and/or aboutness.

I really don't see your case for this. For example, take inscrutability of reference (IoR). IoR doesn't imply that we lose intentionality in any way that matters. Truth values of sentences are preserved, and empirical predictions are preserved. So, how can it be the case that science crumbles if we admit IoR?

As I see it, IoR says that, while there is only a subset of manuals that preserve meaning and consistency with empirical fact, that set has a large number of elements, and reference is fixed only tautologically with respect to a particular manual. If that manual is the physical brain, then the deal is done. It doesn't matter if another brain with a different structure can be constructed such that all truth values and behaviors are identical (i.e., there's another brain with the same empirical validity). The difference in reference just isn't the critical component for intentional thinking.

Suppose we could somehow flip a switch and change the "rabbit" reference to "undetached rabbit parts" in my mind. What changes? Not the word or the number of syllables. Not the inputs to which I associate the word. Not any aspect of my experience that I could detect. Indeed, it seems that a metaphysician oblivious to Quine's argument could find himself contemplating whether rabbits actually are time-slices of rabbits.

Doctor Logic said...

Steve,

On your view ... "that can't be true" is really short hand for "if that were true we wouldn't be able to understand it".

In other words, the "can't" is a feature of thought not of reality. In which case you implicity allow the falsehood of (a) directly and not as an implication of the falsehood of (b)


You're operating on a premise that is built on all sorts of realist assumptions that go far beyond the question at hand. You are assuming that we can create a distinction, i.e., there is a real world that is independent of our thoughts, and we can know this by various aspects of experience. For example, I can't make a cup of tea with my mind, which is why I think teacups have a reality different from my mind, etc.

Now, I agree that we can make such distinctions, subject to certain assumptions. But what are the assumptions?

Non-contradiction is one of those foundations. If non-contradiction goes bye-bye, then so does the distinction (and all distinctions). But, on what grounds can we make the assumption of non-contradiction? We can only make the assumption on pseudo-pragmatic grounds. We will never have a rational argument for the assumption because any such argument would assume a distinction of some kind.

You say:

I agree that I can't imagine it. If you think this isn't a good reason to reject the possibility, then you are undermining all reductio arguments. Do you really wan't to do that? Is that a price you are willing to pay for avoiding the conclusion of the AfR? Be my guest.

What I am saying doesn't undermine reductio arguments. Reductio arguments end at contradictions with implicit foundational axioms.

I'm saying that every argument implicitly assumes a non-contradictory playing field and some logical axioms. A reductio takes us to a contradiction. However, these games don't prove our foundational assumptions to be true because that would be circular.

Getting back to the original debate (if I can remember what it was)... AfR proponents are trying to say that we know that the playing field is non-contradictory, and we know the foundational axioms of logic, and we can't possibly know these things by physics. However, this view is incorrect because the notions about a rational universe are axioms, i.e., they're just assumptions. They cannot be "known" because they cannot possibly be proven. People are acting on impulse, not by their knowledge of abstract logic. Clever folk may have a formal picture of what we're doing, but they don't have proof. And if non-contradiction is mental by compulsion, it can be mental by physical compulsion.

Steve said...

DL,

I never said I was arguing for the truth of the law of non-contradiction. That would be absurd. Nor do I think it can be argued for from a basis of pure logic. However I do think the law obtains and applies to reality.

You, however, think it is a matter of pragmatics that we accept the rule at all. If this is what you think, then I don't the AfR will change your mind about naturalism. That's not a problem with the AfR it's just an indication of how crazy views have to be to avoid the AfR.

You do undermine reductio arguments. Reductios attempt to show that something cannot be the case because assuming it is the case leads to contradictions. However you think that something being contradictory only shows that we can't understand it, not that it's false. Of course if they wan't to hold on to the law of non-contradiction then they'll have to reject their initial position, but you've jettisoned the law of non-contradiction already.

If this is what it takes to avoid the AfR then you're more than welcome to this position.

Steve Lovell

Doctor Logic said...

Steve,

However you think that something being contradictory only shows that we can't understand it, not that it's false.

No, that's not quite what I said. You're missing a distinction.

I can devise a system of propositions to describe reality (a theory). When I create a theory about reality, I implicitly assert that there is a good mapping from the theory to reality, i.e., that the theory predicts or describes reality. It's this assertion about the system/theory that can be false. If the theory doesn't correspond to reality, then the theory is said to be "false". However, systems of propositions don't have truth values. Only propositions do. Systems can be consistent or inconsistent, but not true or false.

It's possible one could devise an inconsistent theory, and end up with a system of propositions for the theory which has an internal contradiction. In this case, the system or propositions would not be false in itself. However, an assertion that there is a good mapping from the inconsistent theory to a consistent world would be false.

You're going a lot further than this. First, it seems like you're trying to create a reductio for the law of non-contradiction itself, and that won't work because it's circular. Second, you're saying that the world must be consistent, always and everywhere, because if it isn't, we won't be able to find a consistent theory to describe it. But our desire to create a consistent map of the world doesn't force the world to be consistent, always and everywhere.

Finally, I'm not saying that reductio arguments do not work, as you allege. I'm saying a reductio provides justification to reject arguments that lead to contradictions on the grounds that such arguments are not rational, cannot be true, and lead to systems that are unintelligible.

All of this is a distraction from the point about the AfR. You're claiming that absolute logical necessity is something our minds are aware of, and that mechanistic minds could not be aware of. Well, even if absolute logical necessity was sensible (which I have been contesting), logical necessity *for* intelligibility certainly does exist, and it's well within the reach of mechanisms. Even if I accepted absolute logical necessity, a mechanism could evolve an instinct for logical necessity on the basis of necessity for intelligibility.

Anonymous said...

Doctor Logic,

You quote,

His view is rather that it makes sense to speak of what our words refer to only relative to some purpose we might have in assigning referents to those words.

Right. In other words, there's no actual 'meaning' in words or - and this is very important - brain processes. These things only have meaning relative to a pre-existing mind (original intentionality must show up somewhere). They don't and (given naturalism) can't have meaning on their own. More on this later.

Dualism doesn't solve this problem. Dualism is an unprincipled, non-explanatory way of non-adjudication. Dualism just says that reference is absolute, and we magically know what the references are.

Who's talking about dualism? Certainly not me. All I'm talking about is naturalism, and what follows given naturalism. In other words: If 'aboutness' really does exist, if beliefs really do exist, if intentionality really does exist, then whatever may be true about our world... a thoroughgoing naturalism is not it. And if science requires these things, well...

Either way, there's a variety of other options left on the table after naturalism is removed. But that's not important here, and the AfR is not a direct argument for any of those views anyway.

I really don't see your case for this. For example, take inscrutability of reference (IoR). IoR doesn't imply that we lose intentionality in any way that matters. Truth values of sentences are preserved, and empirical predictions are preserved. So, how can it be the case that science crumbles if we admit IoR?

[...]

As I see it, IoR says that, while there is only a subset of manuals that preserve meaning and consistency with empirical fact, that set has a large number of elements, and reference is fixed only tautologically with respect to a particular manual. If that manual is the physical brain, then the deal is done.

A thoroughgoing naturalism not only implies that we lose intentionality in a way that matters - we lose it entirely (as Rosenberg admits). Again, see the two horns of Quine's dilemma. The problem may be that you think the IoR is only a statement about some particular frailty of language or third-person perspective. But IoR applies to the empirical world across the board, including brains. Brain states are no more "determinately about" anything than are words.

You replied earlier that you don't think Quine's IoR applies to the speaker himself, just everyone else around him. But that suggests you think sure, no amount of empirical data is able to fix determinate meaning... but there still IS determinate meaning all the same.

There's a word for such a view. It is not "naturalism".

Steve said...

DL,

I think refusing to talk of falsehood of systems is only a semantic issue.

I don't follow when you write:
I'm saying a reductio provides justification to reject arguments that lead to contradictions on the grounds that such arguments are not rational, cannot be true, and lead to systems that are unintelligible.

Well, if we are playing semantics (!!) no arguments are true. Ever. Arguments are valid or invalid, sound or unsound. But ignore that, I still don't follow this sentence.

What are you saying reductio arguments still do on your view? I can't see how you can say that they show that something "cannot be true" (your words above) since you allow that contradictions may be true. But if contradictions may be true, then why think that reductios provide reason for rejecting the position they reduce "to absurdity" when the absurd might still be true?

Perhaps I've completely misunderstood that paragraph of yours, but I did say that I didn't follow ...

Steve Lovell

Doctor Logic said...

Anonymous,

If your interpretation is correct, then we should be able to concoct a simple example of what you are talking about.

Quine is talking about variations in reference subject to empirical constraints. Quine primarily considered the position of a linguist trying to decode an alien language, but there are even more empirical constraints that derive from brain function.

For example, there are cortical columns that recognize rabbits. A set of neurons that switch on only when rabbits are seen. And these columns are connected to the auditory cortex that can play back the sound of "rabbit" in your head.

Universals, like "animal", are neural generalizations from all specifically recognizable animals, and "animal", too, is bound to the auditory cortex and the spoken word "animal", and to the part of the brain responsible for recognizing the written word "animal".

Then there are causal/historical facts about development that are empirically observable. For example, one might be able to see empirically that the subject's conception of "detachable" is not linked to rabbit, i.e., the person has never made any link, either inclusive or exclusionary, between rabbit and undetached rabbit parts.

There are a lot of constraints there.

Now, one way to add inscrutability to reference is to decorate the empirical facts about rabbits with unobservable, unverifiable facts. That is, instead of "rabbit", the person refers to "rabbit + unobservable property A" or "rabbit + unobservable property B" etc. such that every sentence about observable properties of rabbits is unaffected. I'm not sure that this actually does escape empirical testing, but let's suppose that it does. What difference does it make to the intentionality?

My sentence "the rabbit is sitting on my lawn" might translate to "the rabbit (with A) is sitting (in fashion B) on (in fashion C) my lawn (with D)" where A, B, C and D refer to unobservable properties, or refer to properties that have no effect on the empirical consistency of the original sentence. BTW, this empirical consistency means that scientific statements are immune from the problem to the extent that they are making empirical predictions.

I think another possible transformation is to exchange "rabbit" with its complement. Something like "rabbit = rabbit-shaped hole in reality". Assuming this can get past an empirical study of the brain (which is questionable), it doesn't make any difference. Rabbit-shaped holes in reality that have all the same observable properties as rabbits are, well, rabbits.

Doctor Logic said...

Steve,

Oops! I should not have said "cannot be true" in that vague way. I should have said that the conclusion (i.e., concluding proposition) of the argument cannot be true. And it cannot be true because the true-false distinction breaks down when contradictions are admitted. The conclusion would neither be true nor false.

But if contradictions may be true, then why think that reductios provide reason for rejecting the position they reduce "to absurdity" when the absurd might still be true?

A reductio argument works like this:

1) Implicit premise: The system being reasoned about is intelligible, i.e., consistent and a conclusion can be reached by avoiding contradictions, i.e., following rules of logic.

2) Present the explicit premises.

3) Show that valid logical operations on the premises lead to a contradiction.

4) The presence of the contradiction means either the system is unintelligible (absurd), and there are no true or false propositions about the system OR one of the explicit premises is false.

You are interpreting me to say that if it is possible that the system is unintelligible, then I can pick my own truth values for the explicit premises or for the conclusion. You're suggesting that I'm saying that I have the option of sticking to my guns and claiming a proposition about the system (either the conclusion or one of the explicit premises) is true because the system could be absurd.

Of course, this is not at all what I am saying. Rather, I'm saying that a reductio has no power to prove that a system is intelligible. A reductio has power only subject to the assumption that the system is intelligible.

So, let's get back to the AfR. What we know about logical necessity is that it holds subject to the assumption that the world is intelligible. It doesn't hold in the absence of all conditions. It's not necessary for the world to be intelligible. At best we might say that if the world is not intelligible, then there's no necessity of any kind.

Anonymous said...

Doctor Logic,

If your interpretation is correct, then we should be able to concoct a simple example of what you are talking about.

Well, if my interpretation is correct and true (by some miracle), we can't really 'concoct a simple example'. But that's okay, because I'm not truly 'talking about' anything.

For example, there are cortical columns that recognize rabbits. A set of neurons that switch on only when rabbits are seen. And these columns are connected to the auditory cortex that can play back the sound of "rabbit" in your head.

The claim being put forward here is that brain states are not truly 'about' anything. If your example of this starts off with "So, there are cortical columns that recognize rabbits...", you're not really speaking to the claim at all, because the fact that cortical columns and brain states can't really be "about rabbits" is at issue here. So the whole example goes down before it even begins.

I suspect you don't understand the stakes and claims in play here. Again, quoting Rosenberg (himself a naturalist) with emphasis added:

Like public speech and writing, our introspective stream of consciousness doesn’t record or report what the brain is actually doing, because the brain can’t store or manipulate information in words and sentences of any language, including mentalese. Conscious introspection is not just wrong about sensory experience, it’s no guide to cognition either. Whatever the brain does, it doesn’t operate on beliefs and wants, thoughts and hopes, fears and expectations, insofar as these are supposed to be states that “contain” sentences, and are “about” things, facts, events that are outside of the mind. That the brain no more has original intentionality than anything else does is the hardest illusion to give up, and we probably won’t be able completely to do so till neuroscience really understands the brain.

[...]

If linguistic meaning is anything at all it has got to be something like what the philosopher of language Paul Grice discovered about it: at bottom it’s a nested set of beliefs and desires that speakers have about their listeners and themselves. Grice’s own set of conditions necessary and sufficient for linguistic meaning might need to be fine- tuned. But he showed at least to a first approximation what linguistic meaning consists in. Scientism must treat this conclusion as devastating to any attempt to take semantic meaning seriously as a fact about reality. If there literally are no beliefs and desires, because the brain can’t encode information in the form of sentences, then there literally is no such thing as linguistic meaning either. It’s just a useful heuristic device, one with only a highly imperfect grip on what is going on in thought. Consequently, there is no point asking for the real, the true, the actual meaning of a work of art, or the meaning of an agent’s act, still less the meaning of a historical event or epoch.

I'll add a point. If one digs in their heels and insists Rosenberg is wrong here - that there really is determinate meaning in the world, that we do have beliefs, that we do have expectations, that we do reason and reason about, etc.. then whatever we are committing to, it isn't naturalism.

So, what about you? Do you have beliefs? Expectations? Do you have thoughts and brain states 'about' things? Do your thoughts have actual meaning with which you can reason about things?

Steve said...

DL,

I think we both feel this is going round in circles ... but we may be making some small amount of progress.

You write that when a reductio succeeds it shows:
the system is unintelligible (absurd), and there are no true or false propositions about the system OR one of the explicit premises is false.

Well it certainly doesn't show that there are no true or false propositions about the system. One proposition about the system is "there are no truths or falsehoods about it". Clearly it can't show that. Presumably that isn't quite what you meant. I think you mean that a reductio can show:

The system being reasoned about is either false or unintelligble and therefore itself neither true nor false.

However I given your example "bubbles" example I don't think you can say that the unintelligible is neither true nor false. The problem is that both P and -P are true, not that neither is true.

Suppose I construct a reductio about the "Bubble world". You say that despite the success of the reductio the Bubble world theory may be correct. But then if it may be correct then why should the success of the reductio mean that we should reject the Bubble world theory. Do you think we should reject the Bubble world theory because it leads to contradictions? I do. But you seem to think it may still be correct. But what is the difference between accepting the contradictions of the Bubble world and accepting the contradictions shown in a reductio? I don't see any.

I understand why you want to go back to the AfR here. But I don't think we've actually left it ... though we are more in a sideline. The issue is that, if I'm right, then your concept of reason has a disconnect from truth, which on it's own is sufficient to undermine our confidence in our ability to know. So I'm really saying that in taking this line of response you are implicitly allowing another variant of the AfR to go through.

Steve Lovell

Doctor Logic said...

Anonymous,

I recognize your argument from authority, but somehow it just isn't the slightest bit compelling to me. (Who the hell is Rosenberg, anyway?) I have no idea why you think that line of argument would work on me.

So, please make your case with an argument and a clear example. If you understand the argument, it should be easy for you to come up with an example.

Look, we both think we have some form of intentionality. Your job is to show me why our intuitive (folk?) concept of intentionality contradicts naturalism. In particular, please explain why you think a cortical column cannot recognize anything.

A neural network is self-organizing, and, without help from us humans, finds patterns in its inputs. These patterns are abstract. Mere exposure of the network to a single instance of an elephant trains the network to recognize all instances elephants. The networks even recognize elephants with missing limbs as elephants. Human intervention is not required for this learning and recognition to take place. And the network is content-addressable - if it sees a trunk protruding from behind a wall, it infers that there is probably an elephant behind the wall. In what way is this not recognition?

Intentionality and recognition are not quite the same thing. Intentionality is reference to a faculty for recognizing a thing. So when I refer to an elephant, I refer to my faculty for recognizing elephants. When I say "an elephant is in the room", I mean that my faculty for recognizing elephants and my faculty for recognizing "in" and "room" are all firing in coincidence. And because I refer to my faculties, I can refer to fictional situations or counterfactuals. I can say that "if an elephant were in the room, the room would smell bad."

Now, on what grounds is this picture flawed? If I refer to my faculty for recognizing elephants, am I failing to refer to elephants in any way of importance?

Anonymous said...

Doctor Logic,

I recognize your argument from authority, but somehow it just isn't the slightest bit compelling to me. (Who the hell is Rosenberg, anyway?) I have no idea why you think that line of argument would work on me.

Argument from authority? I'm just trying to get you to actually engage the claims being made here, because it seems you're not getting them at all.

I certainly don't mind if you shrug off Rosenberg and reject his views without argument (Me, I think he's crazy and his position is ultimately incoherent). I'll simply note that if you're serious about there really being 'aboutness' (among other things) in the world, then by Rosenberg and Quine you're no longer a naturalist, and move on.

Look, we both think we have some form of intentionality. Your job is to show me why our intuitive (folk?) concept of intentionality contradicts naturalism. In particular, please explain why you think a cortical column cannot recognize anything.

I'm not a naturalist, so my thoughts on whether there can be true 'aboutness' in a cortical column is uninteresting here. But I'll note that you didn't directly answer my questions. So one more time...

Do you have beliefs? Expectations? Do you have thoughts and brain states 'about' things? Do your thoughts have actual (and I will add, determinate) meaning with which you can reason about things?

In what way is this not recognition?

[...]

Now, on what grounds is this picture flawed?

Well, it's similar to the problem with your 'example' with rabbits. You're just out and out declaring the network does these things (It finds patterns! The patterns are abstract! It recognizes! It makes inferences!) But the only way any computer 'does' these things is not due to any intentionality original or innate to it - it's the meaning we ourselves assign. What's more, you yourself should accept this given your defense of Quine's views in this thread (but then again, you also suggested that such statements only applied to third parties, and not to the speaker himself - which implies that the physical facts did not exhaust said speaker's description.)

Think of it this way: What does it mean for a physical network to 'recognize elephants'? Do you mean that the physical network somehow has an innate, original, determinate 'aboutness' directed at elephants? If so, bad news! You've rejected naturalism as Quine, Rosenberg and others have defined it, arguing intentionality, aboutness, and the like are real and ground-level constituents of reality.

Now, do you mean that there is no 'aboutness' in the physical network, and that "recognizes" is simply shorthand for 'A chain of mindless and totally efficient causation that itself is devoid of aboutness or recognition, but from my (observer-relative) position I find useful to refer to as 'recognizing elephants''? If so, good news! You're still a naturalist on this topic. But this means the purely physical network only 'recognizes elephants' relative a mind by your own view. That's a problem.

So, which is it? Or is it some unmentioned third?

Doctor Logic said...

Steve,

Sorry for the delay. Working around the clock at the office this week.

I think you mean that a reductio can show:

The system being reasoned about is either false or unintelligible and therefore itself neither true nor false.


Well, I don't think systems are true or false. Only propositions can be true or false. However, systems can be valid or not valid. And systems that are not valid are unintelligible.

It seems to me that we can only see the world in terms of true and false statements. Consider this example of a system I'll call Q:

x = 5

y = 3

x + y = 9

Knowing that this system Q is contradictory means (by definition) that we know the system is invalid. However, we would not say that the system is false.

I think we're agreed on this so far.

The difficulty arises when we try to turn Q into a theory about the world. If Q is a theory about the world, then we're saying that there's some sort of correspondence between Q and observations about the world. Q fails as a theory if the correspondence doesn't work.

It seems to me that a contradictory system like Q can never have a correspondence to any world, consistent or inconsistent.

Ignoring the principle of explosion for the moment, Q seems to say, at least, that x = 5 and x = 6. What does that mean for observation? Does it mean sometimes we observe x=5, and sometimes we observe x=6? I don't think correspondence makes any sense for multivalued ranges. If it did, then Q could serve as a model for a consistent world, too.

So, we've talked about consistent and inconsistent systems, and, with correspondence, we can talk about good and bad (or even true or false) theories. However, I don't see where we have proven that inconsistencies cannot exist in the world. What we have shown is

1) that there can be no good theories about inconsistent worlds, and

2) an inconsistent theory cannot correspond to observations (of consistent or inconsistent worlds).

These do not imply that

3) the world cannot be inconsistent. (NOT)

A reductio tells us that our system is inconsistent, and that our theory cannot correspond to the world. However, correspondence is not exactly the same thing as existence. It's not obvious to me that the (in-principle) inability to create a theory of X means X cannot exist.

You are saying that, since I won't rule out inconsistencies in the world, I can't use the inconsistency of Q to rule out Q as a theory of the world. But I can do precisely that. Q cannot correspond to any world, consistent or otherwise. Even if the world is inconsistent, an inconsistent system cannot correspond to Q.

So, I say there's necessity for consistency in our theories if there is to be correspondence, i.e., a good theory. But I don't see how to prove necessity for non-contradiction in the world itself, nor can I imagine on what basis such a proof could be made.

Doctor Logic said...

Anonymous,

Since you don't have a counterexample to illustrate Quine's IoR argument in a relevant way, I'll consider that line of your attack to be shelved.

Do you have beliefs? Expectations? Do you have thoughts and brain states 'about' things? Do your thoughts have actual (and I will add, determinate) meaning with which you can reason about things?

Sure I do. But you're begging the question if you assume from the start that those things contradict naturalism. The contradiction is for you to prove.

Do you mean that the physical network somehow has an innate, original, determinate 'aboutness' directed at elephants? If so, bad news! You've rejected naturalism as Quine, Rosenberg and others have defined it, arguing intentionality, aboutness, and the like are real and ground-level constituents of reality.

I totally disagree. I'm proposing that the network is composed of things that do not have intentionality (e.g., molecules), and yet the function and configuration of the network give it intentionality. That is the statement that intentionality is our name for an effect that reduces to a causal system of non-mental components. I'm certainly not claiming that matter is fundamentally mental stuff.

What is intentionality for us humans? Or better yet, what would it mean for us to be mistaken about intentionality? How do I know that my own thoughts are about things? What does intentionality mean?

If I think of an elephant, then I am reflecting on my faculty for recognizing elephants. I didn't directly create that faculty myself. The neural networks in my brain learned to recognize elephants without anyone's help. That faculty was impressed on my mind by elephantine causes.

This is different from, say, an elephant's footprint. Elephant footprints don't recognize elephants, despite having been caused by an elephant. Also, footprints don't reflect on what they recognize. That is, they don't recognize that they recognize. Neural networks are, in principle, capable of both recognition and recognition of recognition.

It seems to me that the difficulty you have is in seeing a physical reaction to a class of things as being a valid form of recognition. So, let me ask, does our subconscious mind recognize things?

Suppose I find myself distracted, and realize that I can't recall the last two miles of driving. During that time, I successfully maneuvered around obstacles, other cars, etc. Didn't I recognize those obstacles without being conscious of them? And if so, weren't my driving evasions 'about' those obstacles? It seems to me that I did recognize the obstacles, but that I wasn't self-aware of my recognition of them.

What do you think?

Steve said...

DL,

I now think this is quite funny. Suppose the world is as described by Bubble Theory, and that reality is therefore contradictory.

You are now saying that Bubble Theory doesn't correspond to the world and therefore cannot be true since Bubble Theory implies a contradiction.

If you mean to say this then your example cannot do the work you want it to do since on your own admission the theory would be literally nonsense.

Alternatively if it can do the work you want it to do, then it contradicts your view that a theory which implies a contradiction is nonsense.

Also, although it doesn't really matter, I don't understand why you think theories are never true or false. A theory is a set of propositions which are asserted together. Therefore they can be combined into a single conjunctive proposition. This, presumably can be true or false. If the conjunction is true then the theory is true. If the conjunction is false then the theory is false. Obviously there is be some debate about propositions on the fringes (auxilliary assumptions) of the theory, but the conjunction of the core propositions is sufficient. Anyway, I think this is just a matter of terminology but your terminology makes sensible conversation more difficult than mine.

Steve