A minor follow up, just so that readers of this weblog know where I’m “coming from.” I accept the contentions that:
- Most cognition is implicit. That is, we do not have access to it to so that it is amenable to reflective analysis.
- Human cognition exhibits operational modularity. That is, different mental tasks rely upon different sets of mental competencies. Whether modularity is “massive” or not, whether it has a biophysical grounding in brain architecture or not, is not of much concern to me.
My ideas are strongly influenced by contemporary cognitive science, they’re not an a priori assumption that I hold because of some deep philosophical belief, or an inductive inference derived from a survey of human history. Rather, laboratory experiments on American college students suggest that as a species we fancy ourselves as much more reflective and rational than we truly are. These findings have made contemporary human behavior much more intelligible to me, and human history as well.
Now, it does turn out that people who are two standard deviations above the norm in intelligence are more rational than people who are around the human norm. These people form about ~2% of the American population, and, they are the ones who are likely to expend their marginal time on discussion about abstract topics. “Big Think” if you will. But, even these people are not very rational or reflective by absolute criteria, or at least not as rational or reflective as they think they are. Rationality has to be graded on a curve.
Hume,
The recent developments in cognitive science are certainly fascinating. There is, however, something about the logic of your position that I am not clear about. Are your statements about cognition and human rationality absolute statements as such, or are they themselves subject to the cognitive and rational limitations you suggest apply to human beings universally? In other words, is the rationality of your post something that we should grade on a curve like everything else?
What I mean is this: You write “we fancy ourselves as much more reflective and rational than we truly are”. I get the impression that, in writing this post, you did not think you were merely fancying yourself being rational and reflective, but that you truly were being rational and reflective. And I have no doubt that you were. But I wonder how you know when you are being truly reflective and rational, and not just fancying yourself being reflective and rational. I’m not trying to be flip… I’m really wondering about the logic of how this works.
This applies especially to your sentence in bold. You write “even these people are not very reflective or rational by absolute criteria,” which seems to imply that you, at least, have access to absolute criteria, for otherwise how could you judge that anyone’s rationality is falling short of it? I’m just wondering how one knows when he actually is thinking in terms of absolute criteria, and not just thinking he is when he really is not.
“…laboratory experiments on American college students suggest that as a species we fancy ourselves as much more reflective and rational than we truly are.”
Are there laboratory experiments on Americans over the age of 50 to counter the findings of the more likely self-absorbed college students?
I thought that was a really great point Mr Tye.
Of course a person who wanted to tighten up his reasoning, and render it transparent to verification, could always resort to formal symbolic methods. Logic or Mathematics.
I understand that not every problem is amenable to math, but the kind of limitations Mr Hume is talking about, are surely a good reason to use formal methods whenever we can.
Naturally, rationality requires more than logic. It also requires grounding in empirical validation. Otherwise we end up with pseudosciences like “Theology”.
Donna B, persons over 50 are totally rational, all the time. Well known fact.
Daniel,
Thank you for the compliment.
I would ask you a similar question to the one I asked Hume. The proposition that we can tighten up reasoning and render it transparent to verification through formal symbolic methods, is not itself a proposal made in terms of formal symbolic methods, so isn’t it subject to all the limitations discussed by Hume?
We agree that formal methods should be used whenever they can… the problem is that the question of when formal methods are appropriate is not itself a matter of formal methods. Nor is the question of how/when formal methods are better than other methods. So if reason is suspect whenever formal methods are not used, reason is suspect in its entirety – including when we use formal methods (because we can’t know if it if there use is appropriate in any given case.)
There’s a serious risk you’re going to get into an infinite regression here.
At some point you forced to conclude that theorem proving or proposition proving are actually computations carried out (carefully and systematically) in our brain-computers. The main purpose in resorting to formal methods is to render the logic transparent and verifiable.
Our brain logic is a product of natural selection and has been selected because it works well in constructing models of the world. From this I might conclude that all our knowledge including our logic is empirical and we have no other source of knowledge.
🙂 …but sadly untrue. I have personally falsified it more often than I care to admit; yet by my lights (though probably not David’s) I regard myself as rational almost to a fault. 🙂
I have long noted that libertarians and secular conservatives tend to “think” through their political positions, while Progressives tend to “feel” them. Presented with a novel idea, the former will ask themselves what they think of it, or better yet, what might be wrong with it. The latter only ask themselves how it makes them feel; and if the answer is uncomfortable or worse, they simply reject it on those grounds alone.
I do have an anecdote with too small a sample size to be significant, but interesting nonetheless. I once belonged to a large national club of predominantly retired folks, traveling around the country in their RVs. It had subgroups for those with common interests, and I innocently joined one with over one hundred members calling themselves “freethinkers,” before discovering that they were overwhelmingly Progressive ACLU type atheists, focused on anti-Christian political activism.
My brand of freethinking, which rejected their Progressive dogma every bit as much as it rejects religious dogma, soon became unwelcome on their forum. Not just for the heresy toward their Progressive orthodoxy, many were too easily offended by robust political debate that disregarded their fragile sensibilities.
I and a half dozen other libertarian oriented members broke away and started a new forum of “Reasonable Rationals.” As an early exercise in promoting comity, I suggested that we all take David Kersey’s “Temperament Sorter” test, so we could understand our individual inherent mind processes better. I was stunned that all but one of us tested as strong Jungian “rationals,” and he only missed it by one question.
This was an eye-opener for me, and I noticed over the next year that new members on our forum who also tested as “rationals” tended to fit right in, while those who did not were problematic and seldom stayed around very long. Anecdotal, to be sure; but I reckon there is something to the correlation worth considering. ◄Dave►
Yes ◄Dave►, Many “well known facts” are sadly untrue when verified.
I agree that rationality is partly a matter of temperament. Partly IQ, I suspect also partly training and practice and commitment.
Daniel,
The problem I have with all arguments of the type you are making is that they always assume more than the argument allows.
For instance, granted that brain logic (is there any other kind?) is a product of natural selection and constructs models of the world. How do we know whether these models are good or bad? The only way to verify a model is to check the model against reality. But how do we step outside our model-making minds to check their answers against reality? Aren’t we restricted to knowing our models of the world and not the world itself? Doesn’t it also follow that our ideas of natural selection are themselves only models of the world, and not necessarily how the world really works?
For myself, whether evolution is true or not (I’m happy to accept that it is), I don’t accept that the mind is essentially a model-maker, because of the logical problems I’ve just been stating. If that presents a problem for evolution, so be it.
I agree with you that an infinite regression is in the offing. But that’s a problem only for people who hold that formal methods are the only way to be truly rational and reflective, of which I am not one. I agree that formal methods are truly rational and reflective, but they are not the *only* ways to be so. When we justify formal methods, we are, I think, being rational and reflective yet not formal.
@◄Dave► I suppose the next question is whether your group of rational thinkers is better at reaching successful courses of action than the people who are emotional feelers.
Are you better at forming cohesive societies? Do those societies make better choices? Et cetera.
for instance, granted that brain logic (is there any other kind?) is a product of natural selection and constructs models of the world. How do we know whether these models are good or bad?
Natural selection eliminates the bad ones. What else can you mean by “bad” other than doesn’t work in securing surivival.
Doesn’t it also follow that our ideas of natural selection are themselves only models of the world, and not necessarily how the world really works?
I know nothing of the world except through my theories which are models of the world.
When we justify formal methods, we are, I think, being rational and reflective yet not formal.
We are being empirical by conducting computation experiments on our brain-computers.
Rather, laboratory experiments on American college students suggest that as a species we fancy ourselves as much more reflective and rational than we truly are.
Important point. We should aspire to and inculcate rationality, but it’s fairly well-established that even intelligent humans often make irrational, counterproductive and/or emotional decisions. Hume seemed to suggest as much: “let Reason be the slave of the passions,” etc.
That a person may perform very well on standardized tests of rationality (quantitative, analytical verbal, etc) obviously does not imply that the person is good or wise or “reasonable” in a sort of holistic sense. Indeed, rationality might be read “instrumentally”: the mad scientist or villain (say a Hannibal Lector-, or Unabomber-meme) often has a profound intellect, but that profound intellect makes it all the easier for him to enact his sinister plans.
Motivation, instinct, intentionality, conditioning, etc. enter the “agecy equation” at some point, and then the philosopher, even a Hume is of little avail.
What do you test your models of the world against? I hope you don’t say the world, because you know nothing of the world – you only know your models of the world. I don’t see how you can know if your models are true or false.
Yes, our model of natural selection is that it eliminates bad models, but that only pushes the question back to the natural selection model, doesn’t it? How do we know if the natural selection model is good or bad?
By the way, the idea that the mind is a model-maker that knows reality only through its models isn’t new, and doesn’t really owe much to science. It was first proposed by Immanuel Kant back in the 18th century in his Critique of Pure Reason. Kant was a philosophical genius and thought through the implications of this position about as thoroughly and deeply as anyone has done. Consider Hume’s point #1 in light of this passage from the Introduction to the CPR:
“All that seems necessary for an introduction or preliminary is that there are two stems of human cognition, which may perhaps arise from a common but to us unknown root, namely sensibility and understanding, through the first of which objects are given to us, but through the second of which they are thought.”
evolution by natural selection is itself an empirical process.
It is the result of survival/replication experiments conducted by competing mutant forms of DNA.
I suppose that depends on one’s subjective definition of “better.” My observations have been that “thinkers” tend to support the idea of a voluntarily cooperative society, organized around the principle of equal opportunity for free individuals to thrive in any non-coercive manner of their choosing, and to excel and succeed beyond the norm, through their own diligence and industry. Conversely, “emoters” tend to support the notion of altruism, collectivism, and the suppression of individual achievement, for the “benefit” of equality of outcome, which they hold is the very purpose of society.
Speaking for myself, as a proponent of individual Liberty, who is more than willing to cooperate and trade with my neighbors; but most unwilling to become a slave in a collective – my answer would have to be, “yes.” ◄Dave►
What do you test your models of the world against? I hope you don’t say the world, because you know nothing of the world – you only know your models of the world. I don’t see how you can know if your models are true or false.
This is starting to get to the point where I’m tempted to bring out Ockhams Trusty Razor.
Seriously what do I care about metaphysical speculations? I am a Darwinian agent. I only need to survive. Metaphysical speculations are a luxury that cut into serious breeding time.
Yes you might be right. The world might be an illusion.
Meanwhile there’s this dangerously cute potential mate over here. Children to bring up. Can you see how maladaptive all this speculative stuff is?
Should we be indifferent to the laws of non-contradiction, excluded middle, and identity? Should we be indifferent to a claim that a thing is not what it is?
Any theory about evolution necessarily supposes that we live in a universe and that we are actually aware of the the true nature of it. In other words, Darwinian theory takes for granted epistemology and metaphysical knowledge.
This indifference and undercutting of metaphysics, as well as epistemology, has left a train of wreckage in the sciences and humanities.
Actually google E.T. Jaynes-he has this all worked out. The first 3 chapters of his magnum opus are online. Read it now.
Bealu, ET Jaynes is a very big influence on me too.
As he points out, there is a logical structure, actually Bayesian Probability Theory, that seems to underlie scientific reason.
But that’s just another way of saying that those, based on their natural gifts, who would rise to the top in that sort of society, are in favor of that sort of society.
Reminds me of when Sen. Bill Bradley once said that he would prefer that the contest for the Presidency be settled by a jump shot from the top of the key.
What do you test your models of the world against? I hope you don’t say the world, because you know nothing of the world – you only know your models of the world. I don’t see how you can know if your models are true or false.
We don’t and we will never know if “our models are true or false”.
As it goes, “All models are wrong but some are useful”.
The belief that there is an “ultimate” rationality, i.e. that there is an all encompassing “true model” that we strive to discover strikes me as yet another religious stance of the Platonist kind.
The Principle of Sufficient Reason has been proven wrong by Gregory Chaitin.
Rationality only pertains to a given model, and it is the process of model building which brings in questionable assumptions during the elaboration of concepts which form the basis of a model.
However this doesn’t mean that when it comes to “theological” discussions we should accept from the fundamentalists the rejection of evidences that themselves are forced to acknowledge for all practical purposes.
I’m not sure I’m making my point clearly, and I apologize for that.
My point is not whether the model-making view of the mind is right or wrong. I’ve said nothing about theology. My point is that those who argue for the model-making mind do not hold consistently to its logic. They make statements that implicitly go beyond the model-making constraints they themselves have placed on thought.
For example, let us consider the following statements in light of the logic of the model-making theory:
1. We don’t and never will know if “our models are true or false.”
2. All models are wrong but some are useful.
3. Rationality only pertains to a given model.
Since our minds are essentially model-makers, it follows that 1, 2 and 3 are all themselves models. If we believe 1, then the following is a logical consequence:
1a. We don’t know and never will know that “we don’t know and never will know if our models are true or false.” So, in fact, we may be able to know if our models are true or false.
2a. If all models are wrong but some are useful, then #2 is itself wrong and some models may be true but useless.
3a. If rationality only pertains to a given model, then the rationality of model #3 only pertains to itself, and says nothing about the rationality of other models or other modes of thought.
The point is that when talking about the model-theory, its advocates implicitly make claims that go beyond the model-making constraints, and so falsify the theory in its very proclamation. “Rationality only pertains to a given model”, for instance, is a meta-model statement intended to apply not only to itself, but to all models universally. It’s just the sort of Platonic statement the theory denies can be made.
Fortunately, experience shows that when they are permitted to do so, society as a whole benefits tremendously from their contributions. Squelch their incentive to achieve, and society unquestionably loses. I have often wondered if all those striving for a collectivist society somehow believe that they will be among the select few privileged overseers of the agrarian masses toiling in the fields. They might be able to coerce their serfs to contribute manual labor to the collective; but there is no way to compel them to think. Even in their glorious Soviet Union, those who did were rewarded with better living conditions. ◄Dave►
. . . he said, channeling Ayn Rand.
Well, there are never perfect solutions, are there. It’s more like “which set of problems would you rather deal with?”
I’m interested in what happens psychologically and emotionally, when the gap between the “have too much” and the “have not enough”‘s gets really large, to a tipping point. In human History, this sort of thing is often not pretty.
How are we to argue with a person who, judging by the arguments they offer, is uncertain about whether we live in a universe?
It doesn’t embarrass me that Ayn Rand would agree with me. 🙂
If you haven’t read Atlas Shrugged recently, try doing so again as an adult. Then look around you at what has happened to this country. It is remarkable how prescient she was.
Frankly, I am past caring about the problems extant in our ghetto infested metropolitan districts. I simply avoid them like the plague. It is high time these folks learned a skill set beyond sticking their hands out and expecting to continue being overfed in the manner to which they have become accustomed.
Academically, this may be interesting; but on a practical level, not so much, once one reaches my state of apathy. I possess nothing I did not earn and the notion that someone 3,000 miles away, whom I have never met, has a claim on my property because he has less, strikes me as ridiculous. That I need fear his retribution, because I did not share my earnings with him, is even more so.
Perhaps the elites in the nearby suburbs, who haven’t the good sense to keep and learn how to bear arms, might come to regret disempowering the ghetto dwellers with their altruistic Progressive policies; but there is little risk that the mayhem would spill out here in flyover country. Were it to, my competent redneck neighbors would make short work of them, long before they got near me.
All in all, I could care less if the coastal cities fell into the oceans. The lifeforms living in them have devolved into something so alien to me and traditional American culture, that I no longer consider them my countrymen. ◄Dave►
@David Tye
Just having had a glance at your blog I much appreciate that you took care to reply 🙂
But you are the one mixing up model and meta-model levels, model and meta-model levels are not bound to be expressed in the same language and a meta-model doesn’t necessarily apply to itself.
At the very least whenever you pretend to communicate you have to use natural language and stick to some bare minimum of logic that is to be agreed upon by all parties.
This bare minimum is roughly at the level of propositional logic, the plain modus ponens syllogism without the aristotelian confusions about essence (Socrate “is” a man or whatever) and existence (though all A are B there isn’t necessarily any A).
Whenever I speak about models I am indeed “within” another model of some sort but NOT within the most general models which can be desirable for various purposes.
I am only within the most minimal model which anyone can have, the “intuitive” meanings of natural language.
If you deny this (that there is a simplest “core” meaning to common discourse) then your own discourse is devoid of any meaning too.
If you deny that there can be a multiplicity of modeling rules which can be agreed upon thru conventions established by the arguing parties you are the one pretending to “flatten” the discourse to your own very low level of discourse, game over of course…
Let me review your objections.
1a. We don’t know and never will know that “we don’t know and never will know if our models are true or false.” So, in fact, we may be able to know if our models are true or false.
No, you cannot use this statement to prove its own inconsistency, because (sorry for my miswording) it is a not a statement about “it’s own truth”, it is a statement about the total lack of evidence for the absolute validity of any model that could be “validated” by reference to another one.
This because the “validating” model (since truth or falsity only has meaning within a model which encompasses some logic) would have to be itself validated, which opens up an infinite regress with NO starting point.
I must apologize for having used the words “true” and “false” where I should have used “perfect” and “sometimes failing” to emphasize the distinction between the meta-discourse, the statement, and what is the subject of the statement, all general models which truth can be gauged against another “validating” model.
This is only another perspective to the “foundational problem” of the choice of axioms in mathematics.
Please note that all the above is written in natural language with no references whatsoever to any “cryptic” or controversial axioms beside the common sense assumptions of natural language understanding.
2a. If all models are wrong but some are useful, then #2 is itself wrong and some models may be true but useless.
When I say wrong I mean “not perfect”, so yes indeed, #2 is wrong but useful (exact in most cases) and your refutation is true but useless.
Thanks for your help 🙂
3a. If rationality only pertains to a given model, then the rationality of model #3 only pertains to itself, and says nothing about the rationality of other models or other modes of thought.
Again, “rationality only pertains to a given model” is a sentence from the “lowest common model”: natural language.
It means that rationality is a property which is definable about the “internal workings” of any given model but has no meaning as a comparison between models. A model may be rational or not (goes haywire when you try to use it) but no rational model is “more rational” than any other.
Unless you are willing to build a meta-model to establish a rationality metric between competing models (in some domain…) and try to sell this model as the “true rationality”, it is not a given.
As for “the rationality of other modes of thought”, it falls under the same predicament, beside the crude criterion (do you go bonkers using this or that “mode of thought”?) having a rationality scale would require a meta-tought.
All these arguments about rationality and truth are actually misnomers for arguments about evidence.
Rationality is about sound methods of proof, truth is about proving (once you’ve got your premisses), evidence is about probing (is this, here, now).
The religionists LOVE to maintain the confusion…
We may be in agreement… I’m not sure. I agree that the “lowest common model” is natural language. Since it is the lowest common model, natural language cannot be undermined or gainsaid by other models. I was trying to show this point by showing that the denial of it is logically contradictory.
The assertions #1, #2, and #3 are all made in the natural language model. Are the assertions true or false, reasonable or unreasonable? The answer to this, if there is to be any, will have to occur in the natural language model and take the rationality of the natural language model for granted.
What often happens is that people say proposition #1:
We don’t and never will know if “our models are true or false.”
and think they are saying something that has been demonstrated by empirical science when it is a proposition that, if it is not to be self-contradictory, can only be made at the level of natural language. Typically, however, if I make natural language objections to #1, my objections are summarily dismissed because they do not fit the model of scientific rationality. Such a dismissal demonstrates, as you note, a confusion between the different levels of discourse. That’s all I was trying to show, poorly it seems.
We are in agreement that rationality is about sound methods of proof. Just what constitutes “sound methods”, of course, is a question that will have to be resolved at the level of natural language… as is what qualifies as “proof” and what constitutes “evidence.”
If you posted on my blog and were expecting a response, I apologize for not responding… sometimes people comment and I think they are not expecting a reply, they are just making a comment.
Pingback: BlogBites. Like sound bites. But without the sound. » Blog Archive » Rationality has to be graded on a curve.