davidm

luvluv recently started a post wondering: Can everything that exists be proven to exist BY MAN? luvluv's analogy, about color vision, was flawed, and was readily exposed as such. But later in the thread Clutch offered the following, which no one responded to. I offer it again on Clutch's behalf, because it was the strongest post on the thread, and the analogy that luvluv should have used. Here it is:

<Clutch>luvluv has chosen a maximally hopeless way of framing the point. There may be better ones

Eg: Rabbits cannot grasp that there are infinitely many primes. It's not that we haven't figured out the right way to teach them; it's not that they feel they have better things to do than learn this. It takes a particular cognitive sumthin' to understand that there are infinitely many primes, and whatever it is, rabbits don't have it. Number theory, as the phrase goes, is Cognitively Closed to rabbits.

Choosing examples appropriately (this one might work for every non-human species, really) we have what appears to be substantial grounds for concluding that some truths can only be grasped using a particular cognitive sumthin' that humans don't have, too. It's a sort of cognitive Copernican principle, really: why suppose that for all other biological cognizers there are cognitive closed truths, but not for us? That puts us at the center of things, cognitively speaking, in a way that requires special justification

This line of thought is raised by a few people, including Noam Chomsky and Colin McGinn. I do not find it very plausible, but it's a far more forceful way of framing the question of principled knowability than luvluv's sensory version, trading as his/hers does on equivocation between the subjective qualities of property detection, and the existence of the properties themselves. </Clutch>

It's a strong line of argument, and the line of argument that I would use if I were a theist, or an attorney for theists in the court of public opinion. Clutch writes that he doesn't find the argument very plausible, but why? I find it quite plausible, and would be glad for any comments to move this debate off the tiresome "God exists!" "No, he doesn't" rigmarole.

Vorkosigan

It's an interesting argument. I'll have to think about it, but let's toss out some observations...

1) humans are not limited by their cognitive equipment, but actively create new modes and tools for analysis, and use those to create still other modes, and so on. Human cognition is augmented by tools both hard and soft. Is HUMANS + COGNITIVE TOOLS closed? Hard to imagine....

2) ...which brings me to whether computers will face these same limitations...if all they need to is tell us the answer, the issue of Cognitive Closure may not apply...

3) another problem is, since the universe is all there is, the issue is not, whether against all infinite possibilities of knowledge humans are Cognitively Closed, but whether against the universe and all possible knowledge therein humans are Cognitively Closed. Thus, the argument may not be as devastating as it appears at first glance, because the object to be studied -- the universe - is large, but definitely finite.

Vorkosigan

Clutch

My reasons for finding the argument less than compelling are essentially those sketched by Vorkosigan in his (1).

In brief, it would not be arbitrary to suppose ourselves different from all other biological creatures, since we have something that none of them -- so far as we know! -- has: a compositional language capable of expressing indefinitely complex thoughts. And, closely related to this, we have an ability to compress such thoughts via cultural refinement, and extrude key components of them into our environments in the form of cognitive tools.

We have many historical examples of the discovery of ideas so revolutionary, complex and abstruse that only a handful of people worldwide could understand them -- and then, within one generation, these ideas have been streamlined and clarified in ways that find them taught to high-school students. I do not say that this irrefutably establishes the principled learnability of all truths, but I do think it makes it *unclear* that there are principled limitations on human learning -- modulo the Fitch cases I also mentioned in the post davidm quotes.

rainbow walking

I would venture that the more pertinent question is: "Can anything that may exist outside the cognitive abilities of man still affect man and remain outside his cognition? If cause and effect are relative I see no way to reconcile such a possibility as logical.

davidm

Rainbow Walking writes: I would venture that the more pertinent question is: "Can anything that may exist outside the cognitive abilities of man still affect man and remain outside his cognition? If cause and effect are relative I see no way to reconcile such a possibility as logical.

To use the rabbit analogy again, we could say that the sun is at least partially, if not almost completely, outside the cognitive abilities of the rabbit. The rabbit can see light, but that's about it, presumably. Certainly we would never expect the rabbit, even in principle, to speculate about the sun, or formulate ideas about it. Yet the sun, obviously, crucially affects the rabbit. So I believe the problem mentioned above can be logically reconciled. It could be that truths outside our cognitive realm are affecting us all the time, but we have, by definition, no way of correctly recognizing or interpreting them.

It's possible we can extend our cognitive reach via computers, or possibly through genetic alterations of our own brain. But if so, that would establish the point that there is a cognitive realm worth aspiring to that is outside our current reach

Clutch mentions our "compositional language capable of expressing indefinitely complex thoughts" as a refutation of cognitive closure for humans, but I'm not persuaded. This idea seems to imply that evolution has produced, in our brain, either the maximal possible sort of intelligence, or else the minimal sort of intelligence that is capable, either directly or thorough computer/genetic augmentation, of apprehending all possible truths. Either could be true, but I see no evidence that they are.

In any event, if there is a straw for theists to grasp at, this one is much sturdier than the usual array. Several papers on this site have demonstrated the incoherence of the supernatural. If I were a theist, I would argue that the supernatural only seems incoherent because it is cognitively closed to humans, but still affects us directly, as the sun affects the rabbit. I would further argue that this theory could be testable, provided we are able someday to enhance our intellects through various means so that we are able to apprehend what we currently cannot. (None of this, of course, implies the existence of god.)

Clutch

I attempted to be absolutely explicit about *not* believing this. To no avail, it seems.

luvluv

ummm.... yeah thanks david that is better. I think from now on I will use that one instead...

Just a thought:

If we had hard limits to our knowledge, would we even know it?

Does the rabbit "know" that it cannot formulate thoughts about the nature of the sun?

rainbow walking

rw: I think your analogy fails to do the question justice. Rabbits possess five senses similar to a human and likely some of them are more sensitive to their environmentthan ours. The only cognitive skills they display are those directly related to their survival and procreation. Like most other animals they have adapted to a particular niche in the food chain. Additionally, your claim that light is outside a rabbits cognition is false because it is a proven fact that a rabbit will freeze up when caught in the glare of an approaching vehicle. Freezing is one of a rabbits survival techniques so it shows that they have cognitively detected a possible danger and responded.

rw: What's more likely is that if such a dimension exists we would require some evidence of it to inspire our search. In the absence of a method of detection there exists no direction to conduct an investigation. Theists have only listed "faith" as their methodology. Can this be scientifically attuned? We know that belief and faith are human attributes developed to explain experiences. All believers eventually admit to some experience or another that they rationalize as evidence from which they derive their faith. But the experiences themselves, when closely examined, all can be attributed to natural explanations or remain totally incomprehensible.

rw: Since you are postulating truths that are un-knowable because of their alleged indetectability, it will be difficult to verify their truthful existence. From this perspective I fail to see how this strengthens the theist's arguments. You are still left with another unverifiable undefined something and are basically just multiplying the drama.

rw: Then your argument would fail for the same reason all the other theistic claims have failed: lack of evidence. It's just another argument from ignorance.

rw: And your argument fails to account for the possibility that our intellects are not the means of apprehending this dimension as most theists will immediately argue that it is a matter of faith. In as much as faith is not without substance you already have ample supplies of substance to investigate, and it has been and still being investigated. All to no avail I might add.

davidm

Dave:: Clutch mentions our "compositional language capable of expressing indefinitely complex thoughts" as a refutation of cognitive closure for humans

Clutch: I attempted to be absolutely explicit about *not* believing this. To no avail, it seems.

Question for Clutch: How have I misinterpreted what you said? I had no intention of doing so. I would be glad for any clarification. I understand your following point:

" I do not say that this irrefutably establishes the principled learnability of all truths, but I do think it makes it *unclear* that there are principled limitations on human learning -- modulo the Fitch cases I also mentioned in the post davidm quotes."

I doubt I said you irrefutably established anything. I saw your argument as weighing against the idea of cognitive closure. Am I wrong? Maybe "refutation" was too strong a word. Again, I'd be glad for any clarification.

To RW: I specifically didn't say that light is outside a rabbit's cognition. I said that a correct explanation for the source of the light is outside a rabbit's cognition. Could it not be that a correct explanation for various aspects of our reality is outside our cognition? There are other parts of your post I want to respond to, but I'm a bit tired right now. Hope the foregoing isn't incoherent due to fatigue.

Clutch

davidm,

Yes, refutation is too strong a word for what I offered, and I tried to be clear about that. Perhaps "defuse" would be a better term; it is simply unclear what limits on truth-expressibility, if any, apply to human languages and cognition.

The essence of the CC argument is that it would arbitrary or unmotivated to propose that humans are different from other species in the relevant respect. But our having a compositional representational system capable of expressing thoughts of indefinitely great complexity is an *enormous* relevant difference between us and other species. So it is simply unclear that the CC argument's reliance on the Copernican metaphor is appropriate.