Concepts and Percepts--Are They Different? Are They Certain?

Hi Laj,

I would think that the idea that there is information not contained in physical states of the brain would be a defence of some form of indeterminism.

It would support nonphysicalism, but not necessarily indeterminism, so far as I can tell. Supposedly, nonphysicalists can still be determinists.

I'm not sure how the phenomenon of backward causation would impact your test, but there is some debate about whether the laws of physics specify a particular direction for causation, and whether the past is as invariant as common sense accepts.

Yeah, I here ya.

First, I'm using a traditional (and probably flawed, given Relativity and QM) view of time for purposes of the test. I figured it was fair to talk about the past in this traditional way because it's also the way we talk about the future, at least in modal logic.

Second, I know that physicists are uncertain as to whether the past is invariant, but I don't understand their views well enough to adjust my test accordingly. It's difficult for me think outside the traditional view of time.

Third, I figured my test was more on point because it directly refutes Nathan's definition of determinism. If determinists accept only one physically possible future, then we should look for a second physically possible future. Interestingly, the "reverse" test doesn't work so well: If indeterminists accept multiple physically possible futures, we don't falsify indeterminism by locating one physically possible future. Here indeterminism is falsifiable only if we can justify why this one physically possible future is the sole physically possible future -- a falsification that I wouldn't know how to devise. At least as Nathan has defined them, I'd say that because determinism includes a more straightforward falsification, it is more scientifically supportable than is indeterminism.

Jordan

Post 261

Sunday, May 29, 2005 - 5:31pm

Laj:

Nathan,

I will not debate a person who cannot see the problem in fixing material an hour after posting it without explaining the nature of the changes, especially if they don't amount to more than corrections of misspellings. That such a disregard of posting ethics is being defended pompously and obnoxiously as moral behavior is galling to say the least, and in a situation where the spirit of trust is low, I have no desire to start rechecking my posts against revised formulations of arguments.

I also have no desire to debate such a person over e-mail either.

Consider this to be my last post to you under any circumstances on this forum.

LOL

I can see why you would not wish to debate me.

Of course, in your "one future at any moment" universe you have no genuine CHOICE but to withdraw (if you actually do).

And I have no choice but to "pompously and obnoxiously" state my case.

Why you're getting so damned excited over the inevitable is a mystery to me!

Oh, I forgot, you have no genuine choice in the matter. You're only ACTING like you do. LOL

I haven't seen anyone else whine about edited posts, while almost everyone I've had exchanges with has done it. I don't think you're going to evoke a lot of moral outrage on SOLO, as tolerance seems to be the prevailing culture on this.

Nathan Hawking

Post 262

Sunday, May 29, 2005 - 6:23pm

Jordan,

It would support nonphysicalism, but not necessarily indeterminism, so far as I can tell. Supposedly, nonphysicalists can still be determinists.

Actually, I know a few dualist determinists, including a past version of myself. However, I think that even a dualist determinist will have a hard time conjuring determinism out of physical dissimilarity, but I may be wrong.

Third, I figured my test was more on point because it directly refutes Nathan's definition of determinism. If determinists accept only one physically possible future, then we should look for a second physically possible future. Interestingly, the "reverse" test doesn't work so well: If indeterminists accept multiple physically possible futures, we don't falsify indeterminism by locating one physically possible future. Here indeterminism is falsifiable only if we can justify why this one physically possible future is the sole physically possible future -- a falsification that I wouldn't know how to devise. At least as Nathan has defined them, I'd say that because determinism includes a more straightforward falsification, it is more scientifically supportable than is indeterminism.

I took the definition for determinism from Peter Van Inwagen (cited in Dennett's Freedom Evolves).

I get the thrust of your argument, but I'm not sure if a simple solution isn't for the indeterminist to define his position as the negation of determinism, and apply your falsifiability test to it.

Post 263

Sunday, May 29, 2005 - 7:09pm

Laj:

The whole point of the Dennett article was that
1) there is an underestimated cultural element to creativity.

Memes happen. That hardly makes the case that all apparently-creative acts are inevitable.

One very significant problem with your view is reasoning from SOME cause-effect to a rigidly universal causally-programmed universe.

Similarly, because SOME creative-looking things come from the likes of weaverbirds and computers, it is excessive to hold that this invalidates ALL tests for consciousness or volition which entail creativity.

If, as I suspect, volition arises as an emergent property of vast amounts of computing power, it will probably be because of an algorithm which allows that sentient machine to steer a non-rigid present into whichever future it might choose. In other words, determinism at one level does not demand or imply determinism at all levels.

Volition as looking into the future and selecting between alternate possibilities is what a chess playing computer does. If you admit this is possible in a computer, then what aspect of volition requires indeterminism?

The aspect that selects from among many possible futures and actualizes ONE of them.

You betray yourself again with the use of the word "selecting." If only one choice is possible, in what sense is an entity, human or machine, "selecting"?

NONE, in any meaningful way.

The generalizations about implying and not implying determinism at all levels are not helpful.

Of course it is. Your main argument rests solely on your assertion that there is at any moment one and only one possible future.

If any portion of the universe is NOT deterministic, then we are admitting more than one possible future.

How a volitional being selects from among those multiple futures, whether through a fully-deterministic algorithm, or by somehow actually INCORPORATING physically indeterminant materials (like quantum mechanisms) is an area of ongoing study.

Our ignorance of the mechanism does not establish determinism by default.

I've been looking hard for something in your posts that amounts to more than postulating that volition is indeterministic (in other words, a research program based on this claim) but it just isn't there.

Have I ever claimed that "volition is indeterministic"? Not to my knowledge. If I said otherwise, I misspoke. To the best of my recollection, all I've ever said or written is that the actual volitional mechanism MAY incorporate some indeterministic submechanism, but that it may also be completely deterministic and depend on an algorithm which brings about volition as an emergent property.

My denial of rigid determinism by way of randomness is directed at your claim that there is only one possible future. Clearly, if there are random events in the universe, there is more than one future possible.

If, for example, uranium atom decay is a random occurrance, animals living near pitchblende ore would be subject to random mutations, and the life forms available to natural selection would be random. In addition, if quantum randomness translates to some degree of randomness in weather, the environment which selects from among those mutations would also not be fixed.

In other words, given some randomness in the universe, there are many possible futures at any given moment. Volition, however it arises, would be that faculty which allows selection from among those futures.

I notice that you're sidestepping Dennett's execrable little piece of logic to focus upon mine. I don't blame you--it's a lost cause.

The problem with your first question is that it's reductionist. It assumes that volition and influence are two different things.

That's exactly where you came into this discussion, your assertion that determinists do not disavow volition as a self-influencing phenomenon. 'Why,' you declare, 'we believe people influence themselves.' But you betray yourself with your assumptions.

So, the answer to your question is: Volition IS an influence, one among many in a whole. It can be stronger and predominate, or weaker in some circumstances. It makes no sense to treat it reductionistically, as it is a wholistic phenomenon.
Yes, my position is reductionist. If reductionism is a problem for you, then so much for your position - it has nothing to offer scientists. A reductionist view of volition simply argues that volition can be analyzed in terms of influences (or determinants).

Your view or portrayal of science is simplistic and narrow. Science also deals with wholistic systems. But I have no desire to enlarge upon that.

If you argue that volition is not a proper subject of reductionist analysis (which I stressed from the first instant was the basis for my acceptance of compatibilism), why didn't you start there, instead of maintaining this facade of pseudoscientific thinking while really intending to claim that volition is something irreducible?

This, I have repeatedly stressed, is the problem with an indeterministic view of volition - sooner or later, the indeterminist has start setting up arbitrary defenses against reductionist scientific inquiry.

See above. Treating something wholistically, as a system, is hardly an "arbitrary defense," nor is science limited to reductionist thought.

Translated: Please explain human psychology.
Your Answer Translated: I don't have answer, so my position is scientifically useless, but by giving snide answers, I can sound smarter than the compatiblists who are trying to give good scientific answers!

You mean good scientific answers to how 'the future is inevitable but we still have freedom, volition, choice.'

Go ahead. I'll wait.

As opposed to the clear and coherent answer you are unable to give as to how 'the future is inevitable but we still have freedom, volition, choice.'

If you are unable answer that in a brief, clear and coherent manner, why don't you just say so?

I've allowed you to pepper me with questions, and have answered them all at some length, but you keep evading that simple one. Is that because you have no
clear and coherent answer?

First of all, snide attempts to disguise ignorance as intellect doesn't qualify.

I've answered your question repeatedly.

I don't drink, so I'm not sure how I could have missed that.

I do not accept that indeterminism is a necessary component of volition, freedom and choice, and anyone who claims that indeterminism is a necessary component of those ideas is free to do so if he cannot understand determinism as anything other than fatalism.

Yah. That would be about like your previous "answers."

Here's a list of the words your position has tried to define out of exitence:

volition
choice
freedom
inevitable
fatalism

Fatalism: "a doctrine that events are fixed in advance so that human beings are powerless to change them."
http://www.m-w.com/cgi-bin/dictionary?book=Dictionary&va=fatalism

You can't really explain, without burying us in bloated paragraphs, how we can "only have one possible future" without it being inevitable or fatalistic.

Otherwise, the burden falls on whoever defends indeterminism as a necessary aspect of volition to show what indeterminism confers that determinism doesn't.

For the last time: multiple futures.

To recap my clear and coherent answer.

1) Uncertainty, choice and volition do not require indeterminism.
2) My death is inevitable- so does that mean I lack freedom? Because freedom is construed as being compatible with causal determinism doesn't make it prison / bondage. Who is the jailer / slave-owner?

Answer this question, and you'll have the answer to your own: At the moment you began to write the text in that gray box above, how many possible textual outcomes were possible?

If you're honest and consistent, you'll answer "One."

There is your owner, slave. You are chained to a single future, according to your view.

3) I've also said that what is important is not inevitability, but an insight into causal necessity. Whether the future is determined or not will not enhance my powers to affect it. ...

I've stressed that an inevitable future, when fatalistically construed, is uninteresting and wrong. The future is not inevitable in any interesting sense of the word "inevitable".

You "answer" the question with a shrug. 'We have volition,' you say, 'but whether the future is inevitable is uninteresting.'

I should say you have not been interested. Had you been interested, you might have discerned that your entire argument rests on the question of inevitability. You might also have realized that random events, if they do exist, falsify your notion of inevitability.

Let's drop you from an airplane and see if the "inevitable" becomes a factor in whether you wish you were wearing a parachute.

You made the odd idiotic statement about bacteria knowing fully well that bacteria do not have a knowledge problem, the knowledge problem being part of what motivates the compatibilist position.

You've become incapable of writing a closing paragraph without being snotty.

As for who's being idiotic, I think equating ignorance with volition comes the closer. Your milage may vary.

I think we've about exhausted this discussion. We're starting to go in circles and repeat ourselves.

Nathan Hawking

Post 264

Friday, May 27, 2005 - 7:15pm

Nathan:
>>Sigh. Many Objectivists do LOVE their certitude. This is gonna be a long, tough sell.

I replied:
>Tell me about it...;-)

Actually, I retract that above statement. In practice Objectivists embrace uncertainty as much as the next girl.

Thing is, Ayn Rand has no better answer to the problem of uncertainty than she has to the related problem of precision we discussed on that other thread.

Like that problem, she thinks she can solve it by simply *changing the words around*, leaving the underlying situation untouched. So, just like using "absolute precision" or "perfectly exact" in place of "approximately" or "roughly", she uses *"contextual certainty"* to replace the usual "uncertainty".

If you are familiar with marketing jargon, this is known as "rebranding". The product itself remains the same, one just changes its name and packaging to try and create the perception of a competitive advantage. In this case, the allegedly product - "New Improved Absolute Certainty!"* - comes with a small print disclaimer -
"*Special conditions apply"

;-)

- Daniel

Post 265

Friday, May 27, 2005 - 11:41pm

Nathan:
>I recall Russell's timbomb reference, I think, but cannot recall details. Can you lay hands on the specific idea and present it here?

It's a pretty famous problem, so you might want to do a bit of reading on the subject for the background.

But I'll give you a quick overview.

Problem No.1 - reasoning from past events to future ones, or from the *known* to the unknown.

Say I've got 6 white eggs in a carton. Am I justified in saying "all eggs are white"?
Say I've got 100 eggs in a carton. Am I justified yet?
Say I've got 20,000 eggs etc. How about that?

Obviously the answer is no to all three. (If you think "yes" at say 20,000, or 20,000, 000, you'd have to explain why this is a magic justifying number, and 6 or 100 is not).

Now the situation does not change by adding "probably white" to the formula either, because there has never been a method of reliably estimating the probability of such a thing (tho many have tried) If you think the probability increases with each observation, it's kind of like the gambler's fallacy (in reverse). Actually, there is no "necessary connexion" between each observation, just like ther is no necessary connection between dice throws. Nothing in logic that takes you from the particular to the general. You simply don't *know.*

This problem itself is enough to destroy the idea we can find certainty, or probability, using inductive means. But there are others.

Problem 2.>Nate: The only thing I can think of might be the need to validate the principle of induction using induction.

Right. This is another serious problem. The principle of induction is allegedly the only way we can have any knowledge of certain scientific truth. Yet the certain truth of this principle can be known only *inductively* itself, and thus requires reference to a higher principle to be certain, and so on into an infinite regress, which is a logical no-no. So it's game over here too. (I'm not sure if the Wikipedia is quite correct to define this problem as merely circular)

Now you may insert "X% probable" if you want to instead of certain, and you will find yourself in the exactly same predicament.

>If that's it, I'd hardly call it a fatal problem, if indeed it is a problem at all.

Fatal enough. If you cannot use it to justify certain truth with certainty, or probable truth with certainty, or even probable truth with any probability, what use is it?

Problem 3. I;ll throw a third related problem in there just for good measure - that is, in induction we make observations which, as they build in number, we turn into theories.

However, this idea, commonsensical as it sounds, is illogical for the following reason: *we must already have a theory before we make an observation*

Popper used to demonstrate this in his class on induction by walking in on opening day and demanding that his class "Observe!" He would then sit down at his desk in silence. Eventually, some timid student would raise her hand and say, "Excuse me sir...but *observe what*...?"

Anyway, it's a complex subject, this is just the raw overview, read up on Hume etc to get the full deal. But overall, it wasn't looking good for the basis of human rationality - until Popper came along.

- Daniel

Post 266

Saturday, May 28, 2005 - 3:54pm

Jordan writes
>Nathan, Why not just pretend that Dennett's position is�Laj's position? It seems like Laj has put his eggs in Dennett's basket, so if you refute Dennett, you'll refute Laj.

Good suggestion. Nathan?

>I know you point to 747's as evidence of volition, but I don't see that your volition indicates an indeterminist's world.�I can see how allowing for more than one future leads to uncertainty, chance, randomness, etc, but I don't see how it�leads to�intentionality, motive, will, etc.

Neither do I. Neither does physics, which recognises only determinism or chance, and can make precise predictions for both. "Will" is not in the picture to the slightest degree. This is the problem for "willful" physicalist! (needless to say I don't agree with Laj either, but I at least understand his point of view)

>Daniel, I should say that I like Popper, but like I said before, I don't think he escapes the problem of induction -- because I think we're still stuck with induction as a means of gaining knowledge --�although he�gave a valiant try. Of course, you're welcome to persuade me of otherwise.

Well, you are obviously familiar with the raven paradox, so I don't need to go into it too much. Simply, no amount of ravens observed will ever confirm the statement "All ravens are black". However, a *single* white raven will *falsify* this theory. There is a logical asymmetry between verifiability and falsifiability that we can take advantage of in growing human knowledge.

Of course this still means only that we can know what is false better than what is true. It means that epistemological certainty is still impossible, other than tautological propositions - and perhaps not even then, as this may only mean we have not been imaginative enough to think of ways of criticising them!

But this lack of certainty is no problem, because *certainty is not necessary* for human knowledge to progress. This quest for certainty is the central mistake, as Popper says. For human knowledge proceeds by trial and error, by proposing imaginative theories and then testing them by rational means; that is, by argument and experience. They will survive by Darwinian selection - death by a single falsification - not by inductive confirmation.

Of course Popper's theory itself is subject to criticism too, and so far has survived as far as I can see. And of course, such survival does not make it true (although it may well be). Nor does it make it "confirmed" by any kind of inductive process - this is a common line of attack on Popper, and may be what you are referring to? - merely *not falsified*.

Remember, inductive inferences are supposed to have an *amplificatory* effect; that is, they become more certain in future the more they are repeated in the past. Whereas Popper's theory is simply the best to have survived so far. This in itself is no indication of its ability to survive in the future, any more than the fact a species of animal has survived in the past makes it any more certain that it will survive in the future.

That's it in a nutshell: imagination and reason working hand in hand in a clear schema (as opposed to just the usual denial of the value of one or the other), and a ruthless survival of the fittest method of improving our theories (as opposed to dogmatic paralysis or limp relativism).

It works for me. If you are interested I would encourage further investigation.

- Daniel

Post 267

Sunday, May 29, 2005 - 1:59pm

Nathan:
>If, as I suspect, volition arises as an emergent property of vast amounts of computing power, it will probably be because of an algorithm which allows that sentient machine to steer a non-rigid present into whichever future it might choose. In other words, determinism at one level does not demand or imply determinism at all levels.

OK, now you've stated your theory more clearly. Human consciousness is the "emergent" outcome of vast amounts of computing power ie: human consciousness is an algorithm, just one of incredible complexity.

The way you use "emergent" in this sense seems to be in the sense that I might use it ie: a property arising unexpectedly from, but not commensurable with, the algorithms that produce it. And I would broadly agree.

But then, oddly, you give an example that is the complete opposite:

>A few months ago I wrote a game which required searching a path which had, if I recall, about 2 trillion possible forks--brute force search was out of the question, as I could allow only one or two seconds on a 1 GHz machine. The algorithm took several days to devise. In the end, it did in two seconds what might have taken days or months to do linearly.

>So, was the program "creative," or the programmer?

The programmer of course. This question barely seems worth asking. If I use a calculator to do an equation it might take me a week to do by hand, and the calculator does it in a fraction of a second, do I now consider the calculator is "creative" seeing that it found the answer and I didn't? What about if I do it on an abacus? An algorithm is simply a set of instructions towards a *goal that you determined* (that goal might be:"search for something", "generate random numbers", "create beautifully unpredictable patterns" etc. Even the fact that your instructions might not do as you determine - they might produce something unexpected, for example - does not mean that your instruction set is "conscious")

So you will have to admit this example is either:
1) An act of emergent, conscious, "creative" volition by an algorithm - albeit at an extremely primitive level - in which case, your Nobel Prize is assured.
2) Irrelevant to your argument.

- Daniel

Post 268

Sunday, May 29, 2005 - 9:53pm

Tsk, tsk, Daniel,

Thing is, Ayn Rand has no better answer to the problem of uncertainty than she has to the related problem of precision we discussed on that other thread.

Like that problem, she thinks she can solve it by simply *changing the words around*, leaving the underlying situation untouched. So, just like using "absolute precision" or "perfectly exact" in place of "approximately" or "roughly", she uses *"contextual certainty"* to replace the usual "uncertainty".

Now I see why they bothered to list the argument ad nauseum as a fallacy! ;-)

Nate T.

Post 269

Sunday, May 29, 2005 - 11:57pm

Alright Nathan, I'll answer your proposed ape/man conundrum. Restated once more ...

--------------
PROBLEM: Now tell us, without referring to lineage, which differentia make one being a pan and another a homo sapiens?

* Rationality? Chimps are rational.
* Bipedalism? Chimps can use their feet alone, and some people are born without them.
* Anatomy? Some human are born with unusual anatomy.
* Speech? Chimps are as capable as many humans who use sign language.
If you cannot, that should be sending you a signal that your concept-formation by classification has some serious defects and/or limitations.
--------------

First of all, you haven't marshalled compelling argument as to why lineage is forbidden--you've merely stipulated this (and that, itself, is substandard argumentation). That said, I will ignore it and proceed--according to your terms.

It is rationality that differentiates apes from us. Results from ape trials are most parsimoniously explained (ie. best explained), by appeal to the 4 perceptual powers of awareness (ie. perception, memory, imagination, and "crude associations"--a mix of memory and perceived, but non-integrated, particulars). Here are some results of some ape trials to make this point clear and unmistakable:

--------------
"Both chimpanzees performed substantially and reliably above chance in collecting a quantity of dots equal to the target numeral, one chimpanzee for the numerals 1 to 7, and the second chimpanzee for the numerals 1 to 6."
From:
http://www.geocities.com/mjberan/counting_pub.html
--------------

The first point to take away from these results is that the apes did not decisively learn the numbers. What would you say, Nathan, about a 1st grader performing "reliably above chance" in counting to 7? Would you say she is exercising her mind's capacity--or would a red flag go up, pointing to a possible developmental disorder for this 6-year-old child?

The second point (again, pointing to a "perception-only" conclusion) is that the total amount of numbers used, lies within a margin of error for short-term memory (ie. for perceptual powers). Crows can perceive (and remember) up to 3 distinct things. Humans can perceive (and remember) about 7-10 distinct things--such as a 7-10 digit telephone number. It is thus highly plausible that apes can perceive (and remember) 6-7 distinct things.

Here is another ape trial (with evidence pointing in the same direction) ...

--------------
In Boysen�s tests, where choosing the smaller of two quantities of candy resulted in receiving a greater reward, chimpanzees chose the smaller quantity 27 percent of the time.

However, in otherwise identical trials that used numerical symbols rather than candies, they were able to choose the smaller quantity 66 percent of the time.
From:
http://nationalzoo.si.edu/Publications/ZooGoer/2002/4/orangcount.cfm
--------------

Not only are these success rates paltry (again, imagine 1st graders getting these same percentages--after repeated testing!)--but they show the unmistakable superiority of symbol memorization (where numerical symbols led to more than twice the success rate), over a rationalization of true counting prowess.

Nathan, unless you can marshall compelling evidence and reasoning (as I have done here), I consider this problem solved. For some strange reason, however, I predict that you will merely argue that I've only STATED that the problem is solved, as if I've really said nothing here, in this post that deals with the evidence and its rational interpretation--and this causes me to be anxious about (but not to eagerly anticipate) your reply.

Ed

Post 270

Sunday, May 29, 2005 - 10:27pm

Nate T.
>Now I see why they bothered to list the argument ad nauseum as a fallacy! ;-)

I was trying to make it inductively true!...;-)

- D

Post 271

Monday, May 30, 2005 - 2:34am

Daniel:

I started a new thread which presents my proposal for the problem of induction.

Nathan

Post 272

Monday, May 30, 2005 - 3:26am

Face it, Ed - this is all such a bloody load of crap it's not worth even using hipwaders.....

Post 273

Monday, May 30, 2005 - 3:36am

Daniel:

Jordan writes
>Nathan, Why not just pretend that Dennett's position is Laj's position? It seems like Laj has put his eggs in Dennett's basket, so if you refute Dennett, you'll refute Laj.

Good suggestion. Nathan?

No, I've explained why this is unacceptable "argument form" in several other posts.

>I know you point to 747's as evidence of volition, but I don't see that your volition indicates an indeterminist's world. I can see how allowing for more than one future leads to uncertainty, chance, randomness, etc, but I don't see how it leads to intentionality, motive, will, etc.

Neither do I. Neither does physics, which recognises only determinism or chance, and can make precise predictions for both. "Will" is not in the picture to the slightest degree. This is the problem for "willful" physicalist! (needless to say I don't agree with Laj either, but I at least understand his point of view)

"Volition indicates an indeterminist's world" because if only one future is available, it's not volition. If volition, then more than one future, then NOT determinism.

>Daniel, I should say that I like Popper, but like I said before, I don't think he escapes the problem of induction -- because I think we're still stuck with induction as a means of gaining knowledge -- although he gave a valiant try. Of course, you're welcome to persuade me of otherwise.

Well, you are obviously familiar with the raven paradox, so I don't need to go into it too much. Simply, no amount of ravens observed will ever confirm the statement "All ravens are black". However, a *single* white raven will *falsify* this theory. There is a logical asymmetry between verifiability and falsifiability that we can take advantage of in growing human knowledge.

Please see my solution to the problem of induction in a new thread.

Thinking of induction as "confirming" something in the absolute sense is to use the wrong terminology.

Of course this still means only that we can know what is false better than what is true. It means that epistemological certainty is still impossible...

I agree with this, by and large. "Certainty" is only possible for a rather minor class of problems. But don't sell induction short. There is a reason why it's effective.

But this lack of certainty is no problem, because *certainty is not necessary* for human knowledge to progress.

I agree.

This quest for certainty is the central mistake, as Popper says. For human knowledge proceeds by trial and error, by proposing imaginative theories and then testing them by rational means; that is, by argument and experience. They will survive by Darwinian selection - death by a single falsification - not by inductive confirmation.

Be careful here, though. Without induction you'd never get knowledge off the ground. By what process do you suppose babies begin to acquire much of their knowledge?

Of course Popper's theory itself is subject to criticism too, and so far has survived as far as I can see. And of course, such survival does not make it true (although it may well be). Nor does it make it "confirmed" by any kind of inductive process - this is a common line of attack on Popper, and may be what you are referring to? - merely *not falsified*.

How does Popper handle induction?

Nathan Hawking

Post 274

Monday, May 30, 2005 - 3:52am

Daniel:

Nathan:
>If, as I suspect, volition arises as an emergent property of vast amounts of computing power, it will probably be because of an algorithm which allows that sentient machine to steer a non-rigid present into whichever future it might choose. In other words, determinism at one level does not demand or imply determinism at all levels.

OK, now you've stated your theory more clearly. Human consciousness is the "emergent" outcome of vast amounts of computing power ie: human consciousness is an algorithm, just one of incredible complexity.

The way you use "emergent" in this sense seems to be in the sense that I might use it ie: a property arising unexpectedly from, but not commensurable with, the algorithms that produce it. And I would broadly agree.

That's progress.

But then, oddly, you give an example that is the complete opposite:

>A few months ago I wrote a game which required searching a path which had, if I recall, about 2 trillion possible forks--brute force search was out of the question, as I could allow only one or two seconds on a 1 GHz machine. The algorithm took several days to devise. In the end, it did in two seconds what might have taken days or months to do linearly.

Opposite? How so?

>So, was the program "creative," or the programmer?

The programmer of course. This question barely seems worth asking. If I use a calculator to do an equation it might take me a week to do by hand, and the calculator does it in a fraction of a second, do I now consider the calculator is "creative" seeing that it found the answer and I didn't? What about if I do it on an abacus? An algorithm is simply a set of instructions towards a *goal that you determined* (that goal might be:"search for something", "generate random numbers", "create beautifully unpredictable patterns" etc. Even the fact that your instructions might not do as you determine - they might produce something unexpected, for example - does not mean that your instruction set is "conscious")

I think you took exactly the opposite of my point. My point WAS that in software this simple (including the chess-playing machine), it's difficult to point to anything the machine is doing which is "creative."

(In some rather radical and complex software we may have had the emergence of genuine volition or creation unbeknownst or as yet undemonstrated. But we're not talking about that.)

So you will have to admit this example is either:
1) An act of emergent, conscious, "creative" volition by an algorithm - albeit at an extremely primitive level - in which case, your Nobel Prize is assured.
2) Irrelevant to your argument.

It is neither. You have mistaken the point I was making. I was saying that Laj had not made his case for genuine creativity in weaver birds or software. That something RESEMBLES created things does not make them intentionally created any more than order in nature suggests a designer God.

In my view, something is not "created" in the fullest sense unless the generative entity had a choice to do it differently or not at all. In that respect, cooling silicon dioxide cannot meaningfully be said to be "creating" quartz crystals in the sense of "creativity."

This is just another of Dennett's determinism-is-really-freedom blind alleys.

Nathan Hawking

Post 275

Monday, May 30, 2005 - 6:40am

Hi Nathan,

As a law student who has studied internet law and copyright in some detail, I just wanted to keep Laj and SOLO safe.

In the weak sense, to which I subscribe by and large, I believe that randomness has been strongly corroborated (to use your word).

Are you familiar with Bohm's alternative to the Copenhagen interpretation? His has the same empirical weight as Bohr's view but avoids randomness as an attribute of the physical.

Worth a try. Maybe you'd be interested in reviewing my determinism thought-experiments if they ever mature (they're incubating for now).

Sure. Sounds like fun.

Anyway, I don't want to get bogged down in the modal logic of the "2nd past test." If my note to Laj didn't clarify the test for you, here's another simple rewording of it: If there's only one physically possible future, then there should be only one physically possible past. Let's look for a 2nd past. Finding that 2nd past falsifies the view that there's only one physicially possible future.

How can we detect a 2nd past? Well, we have lots of ways of detecting the past (just think archeology and atronomy). I think the best evidence would come from us focusing on one event and finding distinctly contradictory evidence as to its past. Of course, if we accept that contradictions can't simultaneously exist, then we can rest assured that we'll never find a 2nd past and that determinism, though falsifiable, is pretty safe from falsification.

How do you think "existence" and "identity" and "consciousness" get us to total certitude of what we're seeing.

I wasn't alleging anything like that, and what�s this got to do with our discussion? I've lost its relevance. As you suggested, let's defer this and the talk of axioms-to-certainty-of-particulars.

[The definition of volition] escapes determinism by positing more than one possible future, which is the exact opposite. (The purpose of a definition is not to refute an argument, but to identify something as distinct. This definition serves that purpose, I think.)

That's not exactly what your definition did. Your definition posited perception of more than one future. It said nothing of whether volition requires that those extra futures exist. We can "perceive" of lots of stuff that might happen but won't; I wouldn't call that misperception.

Hiya Laj,

I took the definition for determinism from Peter Van Inwagen (cited in Dennett's Freedom Evolves).

I know. ;)

I get the thrust of your argument, but I'm not sure if a simple solution isn't for the indeterminist to define his position as the negation of determinism, and apply your falsifiability test to it.

I think that'd be cheating because we can't successfully define one thing only as the null class of another. That is, lots of stuff is non-determinism: cats, WWII, jumping...or more in the ballpark, the view that no futures exist. For indeterminism to be meaningful in this discussion, it requires some positive definition: the view that more than one future exists. Not sure how relevant that response was...

But let me get somewhat back on track and see if I can make my point another way. Consider: (1) "all events are determined" is to (2) "all ravens are black" as (3) "some events are random" is to (4) "some ravens are nonblack." If I've seen lots of ravens, and they're all black, then I've lent weight to (2), which means I've lent weight to (1). That is, by seeing a bunch of ravens, it gets more probable (even if just minisculely so) that (2) is true. But I've lent no weight to (4) because I have no support for it, which means I've lent no weight to (3). In other words, the truth of the proposition, "some ravens are nonblack," is no more probable now after I've seen 10,000 ravens than it was when I had seen just 6. So (1) and (2) are superior to (3) and (4) in that with the former two, we can have a better chance at approximating truth. Not sure if that made sense.

Hi Daniel,

I'm familiar with Popper's arguments; I just don't think they overcome the problem of induction. While Popper hands us a tool (falsification) for shoving aside false theories, he doesn't give us much for ascertaining which theories are true. As I recall, when Popper is confronted with the problem of which of two as-of-yet-unfalsified theories we should choose or use, he responds that we should choose or use the one that has survived the severest of tests. And what is a severe test if not the degree to which it is doubted given past evidence? So far as I can tell, everytime we rely (or choose or use) the past, we're stuck with induction.

-Jordan

Post 276

Monday, May 30, 2005 - 7:04am

Daniel,

Neither do I. Neither does physics, which recognises only determinism or chance, and can make precise predictions for both. "Will" is not in the picture to the slightest degree. This is the problem for "willful" physicalist! (needless to say I don't agree with Laj either, but I at least understand his point of view).

What do you think of evolutionary theory? I am a willful physicalist, though as usual, I'm sure that I'm guilty of perverting the meaning of "will" to a libertarian dualist. But I'm quite sure that what I consider a "will" does everything real wills do.

Laj

Post 277

Monday, May 30, 2005 - 7:33am

Dan,

I want to clarify aspects of your answer to Nathan and some of the things I've been hearing from you about strong AI.

Let's see if you really believe that computers can never be creative (which to me is a variety of the essentialism you mock all the time) or whether you think that the line between who is responsible for being creative gets blurred as things get more complex.
---------------------------
Determinist: Computers can play chess.

Indeterminist: Computers cannot play chess because playing chess requires that you choose between moves. A computer could only make the move that it was going to make given the position on the board, so it wasn't really playing chess. It is only appearing to play chess.

Determinist: But my computer beats you in every game. Are you saying that because my computer doesn't make crappy moves, that it isn't playing chess?

Indeterminist: Yes, I'm saying that in part. The computer needs to be able to choose crappy moves to play good chess. At the very least, it needs to be able to consider options.

Determinist: But my computer does consider options. It evaluates promising moves using the algorithms programmed into it and selects the best move based on its evaluation function at the end of the search.

Indeterminist: Yes, but your computer cannot decide not to make the best move. It needs to be able to decide not to make the best move. It needs to have genuine choice to play chess.

Determinist: If I stuck in a randomizer that made my computer likely to choose different opening positions and play crappy moves once in a while, would it now genuinely play chess?

Indeterminist: No, because it would not be responsible for making the moves that it did. They would all be a result of its programming, which ultimately tie it back to the programmer.

Determinist: So you are saying that the programmer is responsible for the computer's moves. But many programmers are not great chess players but their software beats some of the best human players in the world.

Indeterminist: Well, the algorithms are deterministic all the way down. There is no genuine choice between moves there.

Determinist: So what would be required for my computer to play chess and not just appear to play chess?

Indeterminist: I don't know, but neither do you!
-------------

Dan, please say it ain't so?

Post 278

Monday, May 30, 2005 - 7:54am

Jordan,

But let me get somewhat back on track and see if I can make my point another way. Consider: (1) "all events are determined" is to (2) "all ravens are black" as (3) "some events are random" is to (4) "some ravens are nonblack." If I've seen lots of ravens, and they're all black, then I've lent weight to (2), which means I've lent weight to (1). That is, by seeing a bunch of ravens, it gets more probable (even if just minisculely so) that (2) is true. But I've lent no weight to (4) because I have no support for it, which means I've lent no weight to (3). In other words, the truth of the proposition, "some ravens are nonblack," is no more probable now after I've seen 10,000 ravens than it was when I had seen just 6. So (1) and (2) are superior to (3) and (4) in that with the former two, we can have a better chance at approximating truth. Not sure if that made sense.

I get your point without the explication. I guess my intuition (which might be a function of a limited imagination or limited exposure to relevant facts) tells me that determinism and indeterminism are mutually exclusive in an empirical manner in this case.

So I wouldn't consider it cheating in this case. I have nothing against indeterminism other than the resistance to some interesting psychological science (and philosophy) that it fosters.

Post 279

Monday, May 30, 2005 - 2:18pm