The Analytic-Synthetic Dichotomy

[Brendan Hutching]

Brendan chose not to respond to my introducing perception by infants and eagles.

Does an infant make a judgment that mommy now is the same person seen hours earlier? Must an eagle make some sort of judgment that the hare it sees today is similar to the hare it preyed on earlier? What sort of judgment in each case?

Merlin: “Brendan chose not to respond to my introducing perception by infants and eagles.”

Because it wasn’t directly relevant to my argument, which was about the status of ‘similarity’.

But to answer your question, clearly some sort of judgement is occurring in both cases, so I’d concede that there are perceptual judgements that don’t necessarily involve what we would call conceptualisation.

But that doesn’t make ‘similarity’ a really existing category, which is my original point.

Brendan

[tjohnson]

The usual argument is whether similarity is in objects or in our perceptions/concepts, so this is a perhaps finer question. I will bridge-sit here, and suggest only that there seem to be different types of similarity: for instance things we perceive sensorily as being similar, and things we know intellectually to be similar (these last we may even perceive to be different, like the length of lines in the optical illusion). There also seems to be a complex interaction between the two out there on the frontier.

According to special relativity, there is NO absolute length of an object - the length that we measure will depend on our state of motion. In an optical illusion, we don't know what the length IS, but rather we think we know what it will be when we measure it, and the result will be quite different from what it appears to be.

Edited June 12, 2007 by general semanticist

[Michael Stuart Kelly]

Brendan,

What do you mean by category?

Did Rand's crows see a category?

And if similarity is not a category in the sense you mean, do people see categories?

If all you are saying is that the abstraction (selective comparison and compression into a unit) itself is not in reality except in the mind that did the abstracting, I agree. But then, perceiving a whole entity from partial sensory data is a form of abstraction, so whole entity perception itself is only in the mind, yet the whole entity exists.

If properties of entities did not exist in such a state that they could not be perceived as similar, they would not be so perceived. This seems obvious to me. And what more, this is a perfect case of demonstrating that the mind's function is to process information about reality as it exists, thus the mind is a part of the same reality.

Michael

[Dragonfly]

Ba’al: “If X is a mental event how come it causes a change in electrical potential in the brain? How can a non-physical cause have a physical effect?”

think the puzzle is due to the dissimilarity between the cause and the effect. If I decide to turn left rather than turn right, I can explain my decision in terms of ‘purposes’ and ‘desires’. However, the electrical changes in the brain and muscular contractions in my arms are explained in different terms.

It is an example of the flawed notion of "downward causation", which I discussed here.

[BaalChatzaf]

The usual argument is whether similarity is in objects or in our perceptions/concepts, so this is a perhaps finer question. I will bridge-sit here, and suggest only that there seem to be different types of similarity: for instance things we perceive sensorily as being similar, and things we know intellectually to be similar (these last we may even perceive to be different, like the length of lines in the optical illusion). There also seems to be a complex interaction between the two out there on the frontier.

According to special relativity, there is NO absolute length of an object - the length that we measure will depend on our state of motion. In an optical illusion, we don't know what the length IS, but rather we think we know what it will be when we measure it, and the result will be quite different from what it appears to be.

The invariant quantity under a Lorentz Transformation is not length, it is the space-time interval

dx^2 - dt^2. The length changes under a Lorentz Transformation are analogous to the apparent change in the length of an item due to its angle to the line of sight.

Ba'al Chatzaf

[Brendan Hutching]

What do you mean by category?

By categories I don’t mean objects, I mean ways of thinking. An example is the phenomenon of confirmation bias, where we place greater credence on information that confirms what we already know, and discount information that disconfirms current knowledge.

No doubt there is an evolutionary reason for this phenomenon, and it seems to be a very widespread human trait, manifesting itself in a wide variety of knowledge areas, from politics to sports to science to religion etc.

This bias doesn’t appear to be learned, but inborn, part of a more general class of inborn expectations, for example of regularities, and that the future will resemble the past. The expectation of regularities leads to such endeavours as seeking underlying principles to explain diverse phenomena, whether those principles are religious, philosophical, scientific, aesthetic etc.

The perception of similarity between like objects seems to be a type of expectation of regularities, an outcome of a preference for similarity over difference. That doesn’t mean that certain types of objects are not similar in various ways, only that this doesn’t establish that the similarity is in the objects rather than the perception.

Brendan

[merjet]

The error is to claim that ‘similarity’ can literally be observed, ie physically perceived. But ‘similarity’ is not a perceptual quality. Rather, we perceive specific objects, not ‘similarity’.

So when the child observes the three objects he cannot “see that there is something similar”. What he sees are three objects, not three objects plus “something similar”.

Since ‘similarity’ cannot be present in the perception of objects, it must be inferred or thought. In that case, in the above example, the child is not “seeing” something similar, but rather thinking “something similar”. Therefore, ‘similarity’ must already be present in the mind and brought to the experience of observing the three objects.

The perception of similarity between like objects seems to be a type of expectation of regularities, an outcome of a preference for similarity over difference. That doesn’t mean that certain types of objects are not similar in various ways, only that this doesn’t establish that the similarity is in the objects rather than the perception.

I note a significant shift in position here. First similarity could not be perceived, then it's "in the perception". Maybe in a few more days, he will say that objects can be similar, even while not being perceived.

[tjohnson]

The error is to claim that ‘similarity’ can literally be observed, ie physically perceived. But ‘similarity’ is not a perceptual quality. Rather, we perceive specific objects, not ‘similarity’.

This position seems totally backwards to me. It is only after perceiving a number of similar 'things' that we can form a concept. After observing a number of 'things with hard, rough, slender, long parts coming out of the ground with branches and green things on them' we call this image 'a tree'. A child would ostensibly perceive these similarities as well but wouldn't know what the concept name was or it's definition until it was told.

[BaalChatzaf]

I note a significant shift in position here. First similarity could not be perceived, then it's "in the perception". Maybe in a few more days, he will say that objects can be similar, even while not being perceived.

Similarity is an equivalence relation produced by the operation of our nervous system. Similarity is In Here. Similarity is a judgment call, not a fact. If there were no sentients in the cosmos there would be no similarity.

Ba'al Chatzaf

[merjet]

Similarity is an equivalence relation produced by the operation of our nervous system. Similarity is In Here.

Really? So two objects could not be similar while not being perceived? What about a cause-and-effect relationship? Is it also just "in here" and not "out there"?

[tjohnson]

Similarity is an equivalence relation produced by the operation of our nervous system. Similarity is In Here. Similarity is a judgment call, not a fact. If there were no sentients in the cosmos there would be no similarity.

Ba'al Chatzaf

Do you count dogs as 'sentients'? If not, how does he find his master with his nose if not by 'similar' chemicals given of by the master? The same could be said of almost any organism, in fact, even, molecules recognize similarities with chemical specific receptor sites.

[Laure]

I note a significant shift in position here. First similarity could not be perceived, then it's "in the perception". Maybe in a few more days, he will say that objects can be similar, even while not being perceived.

Similarity is an equivalence relation produced by the operation of our nervous system. Similarity is In Here. Similarity is a judgment call, not a fact. If there were no sentients in the cosmos there would be no similarity.

Ba'al Chatzaf

Hm, Ba'al, I predict that you also think that "if a tree falls in the forest, and nobody's there to hear it, it doesn't make a sound". In my view, similarity is "out there". Some similarities are more relevant to us as humans, and those are the ones we notice, but our noticing something doesn't cause its existence.

[BaalChatzaf]

Similarity is an equivalence relation produced by the operation of our nervous system. Similarity is In Here. Similarity is a judgment call, not a fact. If there were no sentients in the cosmos there would be no similarity.

Ba'al Chatzaf

Do you count dogs as 'sentients'? If not, how does he find his master with his nose if not by 'similar' chemicals given of by the master? The same could be said of almost any organism, in fact, even, molecules recognize similarities with chemical specific receptor sites.

It is the dog's nervous system that size the molecules cause "similar" smells. If there were no dogs, no such thing would be the case.

I am not saying similarity does not have a physical cause. I am saying the physical cause is In Here. If our nervous systems worked differently there would be no such "similarities".

Ba'al Chatzaf

[bmacwilliam]

Hm, Ba'al, I predict that you also think that "if a tree falls in the forest, and nobody's there to hear it, it doesn't make a sound".

Philosophically, the question is meaningless until you define sound.

Bob

[BaalChatzaf]

Hm, Ba'al, I predict that you also think that "if a tree falls in the forest, and nobody's there to hear it, it doesn't make a sound".

Philosophically, the question is meaningless until you define sound.

Bob

The vibration of the air is Out There. The effects of the vibrating air vibrating our ear drums and stimulating the auditory nerve is In Here. I define sound as the effect that auditory nerve stimulation has on our consciousness. If everyone were stone deaf there would be no sound, but there would still be lots of air vibrations.

Ba'al Chatzaf

[bmacwilliam]

Hm, Ba'al, I predict that you also think that "if a tree falls in the forest, and nobody's there to hear it, it doesn't make a sound".

Philosophically, the question is meaningless until you define sound.

Bob

The vibration of the air is Out There. The effects of the vibrating air vibrating our ear drums and stimulating the auditory nerve is In Here. I define sound as the effect that auditory nerve stimulation has on our consciousness. If everyone were stone deaf there would be no sound, but there would still be lots of air vibrations.

Ba'al Chatzaf

Exactly. And that's why the question is more or less meaningless.

Bob

[Dragonfly]

The point is that it depends on the definition of sound. One definition is that what our consciousness perceives, in that case there is of course no sound in the forest if nobody is there. But sound is also defined as vibrations in the air (check any dictionary), and in that case there is of course sound in the forest. This supposedly philosophical question is therefore a trivial matter of definition.

[Michael Stuart Kelly]

The vibration of the air is Out There. The effects of the vibrating air vibrating our ear drums and stimulating the auditory nerve is In Here. I define sound as the effect that auditory nerve stimulation has on our consciousness. If everyone were stone deaf there would be no sound, but there would still be lots of air vibrations.

Bob,

By this reckoning, we arrive at one of my own speculative questions, which is this: Do we perceive all there is to reality, or just enough to survive? In other words, do our sense organs cover it all?

I am talking about kind, not degree.

If this were true, it would give a whole other dimension to the term "human knowledge." Also, I bet you could fit the mind/body problem in there right nicely. You could probably fit the quantum mechanics observation thing right along in there with it.

Michael

[Michael Stuart Kelly]

You guys are talking about the fundamental axioms. One doesn't exist without the others.

There is no knowledge without a consciousness. There are facts, though. The knowledge of sound needs consciousness (and an appropriate sense organ). The fact of sound (however you want to call it) does not.

My "A is A" itch is starting to act up.

Michael

[Daniel Barnes]

Wouldn't this be better as a new thread about the 'problem of universals' rather than the ASD?

[tjohnson]

The vibration of the air is Out There. The effects of the vibrating air vibrating our ear drums and stimulating the auditory nerve is In Here. I define sound as the effect that auditory nerve stimulation has on our consciousness. If everyone were stone deaf there would be no sound, but there would still be lots of air vibrations.

Ba'al Chatzaf

Yes and SOME of those vibrations would be similar, with or without an observer, like when the wind whistles through the trees. I'm not sure there is any way to prove this either way. :cry: What about electrons, wouldn't it be safe to assume one electron is similar to the next? We can't sense electrons so you're argument about similarity only existing in perception won't apply there.

Edited June 13, 2007 by general semanticist

[BaalChatzaf]

The vibration of the air is Out There. The effects of the vibrating air vibrating our ear drums and stimulating the auditory nerve is In Here. I define sound as the effect that auditory nerve stimulation has on our consciousness. If everyone were stone deaf there would be no sound, but there would still be lots of air vibrations.

Ba'al Chatzaf

Yes and SOME of those vibrations would be similar, with or without an observer, like when the wind whistles through the trees. I'm not sure there is any way to prove this either way. :cry: What about electrons, wouldn't it be safe to assume one electron is similar to the next? We can't sense electrons so you're argument about similarity only existing in perception won't apply there.

Similarity is a -judgment- made by an observer. No observers, no similarities.

Ask a normal visioned person to sort items by similar colors and then ask a color blind person. You would get a different sort from each. Similarity is In Here, but it is based on what is Out There. Think of an observer as a kind of potato sorter. The sorting of a pile of potatoes is dependent on the hole size.

Similarity judgment has a subjective component.

Do you see a similarity between a donut and a teacup? If you are not mathematically trained you would say there is little or no similarity. If you are a topologist you would say they have the same structure. Similarity is only partly Out There and mostly in your head.

Ba'al Chatzaf

[Michael Stuart Kelly]

Bob,

This is semantics.

Metaphysics: Similarity means that different entities (and attributes, actions, relationships, etc.) exist in such a manner that a consciousness can organize knowledge (similarities and differences) about them. Such existence is "out there" just waiting for an "in here" it come along.

Epistemology: Similarity is a mental form of identifying reality. This is an "in here" and just waiting to bump up against an "out there."

No amount of repetition will ever make the metaphysical part stop existing. All your argument is showing is that you do not accept that a person is talking about the metaphysical part. You want to delete one of these two meanings of "similar."

That will never happen. Switching meanings on a poster in the middle of a discussion certainly doesn't help. Both meanings exist in today's language. I foresee a long and healthy usage of both, despite anyone's protestations.

Michael

[Gregory Browne]

That's why he selected 'Ice is solid' as an example of an alleged synthetic truth: because it is not part of the definition.

Indeed, therefore it is a synthetic truth.

You still haven't proved that non-definitional truths are synthetic, because you need to prove that it is only definitional truths whose denial yields a contradiction.

Peikoff selected it because it is the kind of analytic truth that people mistake for a synthetic truth.

It was falsifiable because a necessary connection between the solidity and the floating on water was not demonstrated.

Hindsight is always 20/20. When he wrote that article, Peikoff was no doubt absolutely convinced that such a necessary connection was sufficiently demonstrated, and that the statement therefore in his terms was an analytic statement. He was proven wrong.

First, neither Peikoff nor I equate analytic truths and necessary truths: I say that all truths are analytic but not all are necessary; he says that (in one sense) all truths are analytic, but not all express necessary facts.

Second, Peikoff was wrong in believing that "All ice floats on water" is an analytic truth and a truth expressing a necessary fact, and that there was a necessary connection that had been sufficiently demonstrated ---but not because it was a synthetic truth or a truth expressing a contingent fact, but because it was not a truth at all.

More importantly, that sentence was only an illustration. His general claims are:

(1) If a statement is true, then it is analytically true

and

(2) if a statement, other than one expressing a human free will choice’ is true, then it expresses a truth about a necessary fact.

These are conditionals, and a conditional is never made false by the falsehood of its antecedent (the clause just after the “if”).

As far as I know, ice is solid because of its chemical structure, and the kinds of ice that float on water do so because of their chemical structure, and that it is because their structure is less dense that that of liquid water that they can float on it. I think that all of these are necessary facts.

You need to give some definition of 'definitional truth', or, better, a definition simply of 'definition', that allows us to sort definitional from non-definitional truth.

You need to elaborate the distinction you make. Otherwise it will just sound like you are saying that physics and math are obviously different, and the truths of former are synthethic and truths of the latter are analytic, which needs a lot of arguing for.

No, that ploy will no longer work. I have always to argue for this and to argue for that, while you think it's sufficient to state "which Peikoff and I deny". Merely denying is no argument. I've given examples to illustrate my point, which you ignore, so perhaps it's now time that you give a refutation of my examples.

Yes, you do have to argue for what you claim, unless it is self-evident or your opponent concedes it, because you are claiming the existence of these dichotomies: it is not up to your opponents to prove that the dichotomies do not exist, just as it would not be up to the opponents of somebody who believed in ESP to prove that ESP does not exist.

Moreover, Peikoff made arguments against the existence of the dichotomies anyway, and you began the debate with a post criticizing Peikoff's claims in the ASD. So all the more is up to you to make the case that the dichotomies are valid.

I have not ignored your examples: I have addressed them.

Now I assume that you want to persuade me, and everyone else here, that your claims are correct. And I assume that you want to do so rationally: that is, you do not want to persuaded me by arguments that are defective in any way. Well in a debate its is a defect to start from premises your opponent does not concede. It would be irrational for me to let myself be persuaded by arguments based on premises that I believe to be false.

You have given arguments for your original claims, but the premises are themselves disputed.

Now it seems very likely to me that those premises are premises that you have taken for granted, and for a long time, because they are widely taken for granted among those you deal with. You probably find it very difficult to even conceive how anyone can doubt premises that seem so obvious to you. But they should not be enough for you or for any scientist. You should be able to defend your premises right down to their foundations.

Nonetheless, I will be happy to argue against the claim.

I gave several versions I found of the definition of analytic statements, for example: "analytic truth is a priori, necessarily true, independent of reality", "a proposition that is true by definition".

Which one do you want to argue for? You will claim that they are all the same, but that claim needs to be argued for.

I don't have to argue for anything. Both will do for me. If you think there are essential differences then you'll have to argue for that.

You should assume that they are different until they are proven the same because they have different definitions (as given by most philosophers, including their most prominent advocates, the Logical Positivists):

Analytic truth--1. A truth whose predicate is contained in its subject.

2. A truths whose denial is a self-contradiction (I will concede what this definition is equivalent to definition 1)

3. A truth that it is true in virtue of the meanings of the terms in it.

Necessary truth—a truth whose negation is impossible

A priori truth—a truth which cannot be known by empirically (i.e., from experience)

Definitional truth—a proposition that is true by definition

So why do you assume that tautologies about Deep Kinds (such biological and chemical kinds) will always be trivial, while tautologies in math will not always be?

While the first are only a question of definitions and semantics, the logical conclusion then follows in most cases in one step from the definitions. And if you think that the tautologies in mathematics are all trivial,

I certainly don't think that tautologies in math are all trivial.

Rather, I doubt that all tautologies about Deep Kinds are trivial.

Indeed, I deny that they are all trivial, since I agree with Peikoff that all truths are tautologies, and so I oppose the common view that tautologies are inherently trivial--and I am glad to see that even the Logical Positivist A. J. Ayer, and you, agreed with me on that.

However, even if I agreed with you and the Logical Positivists that only truths expressing nominal definitions are tautologies, I would still deny that tautologies about Deep Kinds are trivial.

It can be self-contradictory even if you cannot be sure of its truth.

But in that case you don't know that it is self-contradictory, so you can't say that the statement is self-contradictory, in contrast to self-contradictory analytical statements, where you can categorically say that they are self-contradictory, they don't need any empirical evidence for that. That is the essential difference that Peikoff and you can't argue away.

I don’t deny that there is a difference: I just deny that it is the difference you think it is.

The difference is between truths about Shallow Kinds, on the one hand, and truths about Deep Kinds, Narrow Classes and individuals, on the other. Regarding the former, you will have certainty if you analyze the concepts right to come up with your basic truths, and then make deductions without making any mistakes. For example, we can analyze the concept of triangles to come up with the basic truths of trigonometry, we can analyze the concept of bachelors to come up with statements such as ‘All bachelors are unmarried’, and we can analyze the concept of ice to come up with statements such as ‘All ice is solid’. Regarding the latter, you may not have certainty, or any degree of knowledge, because you have to discover the basic truths: you have discover that water is H20, and that gold has atomic number 7, and you’d have to discover that all ice floats on water—if it did, which in fact it does not, which also had to be discovered.

However the above facts do not require us to conclude that truths about Shallow Kinds are not known by empirical means: we can still say that all concepts, including those of triangles, bachelors and ice, are empirical, and if that is true we can say that truths discovered by analysis of those concepts, such as the truths of trigonometry discovered by analysis of the concept of triangles, are known empirically. Nor do the above facts require us to conclude that truths about Shallow Kinds say nothing about the world, just because they can apply even to other possible worlds, too.

Nor do the above facts require us to conclude that truths about Deep Kinds, Narrow Classes and individuals are not necessary, just because we have to discover the basic truths about each of these, nor to conclude that they can never be certain, just because we have to discover the basic truths about them. Nor do the above facts require us to conclude that denying truths about Deep Kinds, Narrow Classes and individuals will not yield a contradiction.

But this matters little to me, because, as I have said, scientists are not authorities in this field. It is you who have been invoking their authority. Let's use our own arguments for our claims.

Of course scientists are authorities in this field, perhaps not in all the linguistic niceties, but certainly in the foundations of physics and mathematics, of which most philosophers don't know anything (Rand and Peikoff included). Only philosophers with a solid scientific background like for example Bernard d'Espagnat can be taken seriously in this regard.

You are making an Argument from Authority: you are saying that we ought to believe in these dichotomies because all or most scientists believe in them, and they are authorities in the matters, and so we should take their word for it.

However, an Argument from Authority is valid only when the authority is genuine authority—that is, an expert—on the given subject, and to be an expert that person must much more about the subject than non-experts.

For example, calling a medical examiner or other doctor as an expert witness on medical matters in a court trial (such as the time of death in a murder case) is valid because doctors know a lot more about medical matters than the rest of us do, and calling a ballistics expert as an expert witness on ballistics in a court trial is also valid, because ballistics experts know a lot more about ballistics than the rest of us. The same is true about relying on scientists in highly specialized, technical fields, because there is a lot of detailed information to learn in such fields, and they have learned it. For example, it is OK to appeal to authority of most atomic physicists about matters of atomic physics.

However, an appeal to the authority which asks us to believe what most scientists believe about general, basic scientific principles, including principles of general scientific methodology, and to believe it merely on the ground that most or all scientists believe it, is invalid. The basic principles of science, including those of scientific methodology, are principles that any intelligent person with a modicum of education should be able to grasp and analyze intelligently. So here we should believe what most or all scientists say about them, just because most or all of them say it.

Moreover, most of the topics in scientific methodology, including these alleged dichotomies of truth, are also topics in logic. Now would you ask us to accept scientists as authorities here, rather than logicians? And even logicians would not, or should not, ask you to accept what they say just because they are logicians.

Even Einstein didn't ask us to accept his views because he was an authority.

We should not take their word for anything

Uh-huh, you came up with "most scientists" who supported your view, and when I told you that my experience was different, you tell me that we shouldn't take their word for anything.

I don't believe that most scientists today support my views, though I doubt that they are unanimous in supporting yours, and I pointed that scientists have not always felt that way, nor have philosophers, nor do most philosophers today.

Yes, that in itself proves nothing about my own claims. But it was you who introduced the view of "all scientists" as an authority to provide reason for not arguing for your claims. In short: there is no authority in these matters.

Again, 'does not depend on reality' is a little vague, but I will assume it means 'true in all possible worlds, i.e., necessary'. This then is the classic argument of some Logical Positivists such as Wittgenstein, who said that necessary truths are not factual (i.e., not about the world) because they are true no matter how the world is. However, being true no matter how the world is, i.e., being true in all possible worlds, does not mean that it does not say anything about the actual world---on the contrary, it means that it says something about each and every possible world, including the actual world.

No. Its referents may refer to things in the actual world, but not necessarily. With the usual definition of unicorn the statement "a unicorn has a horn on its head" is an analytical statement, or in your words, it's true in each and every possible world. But does it say anything about the actual world? We can calculate the volume of a 254306-dimensional sphere embedded in 254307-dimensional Euclidean space, which also forms an analytic statement. Does it refer to anything in the actual world?

Yes: those statement do say something about the actual world (as well as every other possible world). In the case of the unicorn statement it says that nothing without a horn is a unicorn, and so if anyone attempting to prove the existence of unicorns brought in some hornless animal their claims should be rejected immediately. An analogous but more complicated statement could be made about the second, assuming that a sphere with more than 3 dimensions is even logically possible.

Also, note that ‘true in all possible world is definition’ (metaphorically) for ‘necessarily true’, not ‘analytically true’, and though I claim that all truths are analytic, I don’t claim that all truths are necessary.

I got the idea from you: you said that the physics of this world can make an geometry impossible--not merely false, but impossible---which is usually taken to mean 'necessarily false'.

No, I meant with impossible that it would not describe physical space correctly. You continue to confuse the applicability of a certain mathematical theory to a certain physical phenomenon with its validity. A physical theory that used that particular mathematical model would not be valid, but that doesn't invalidate the mathematical theory, it merely says that this isn't the appropriate theory for that particular phenomenon. Really, it isn't that difficult!

I wasn’t talking about applicability in so many words, but if you want me to address it, I will: how could a theory be applicable to the world if it is not “factual” (i.e., “about the world”)?

[Brendan Hutching]

After observing a number of 'things with hard, rough, slender, long parts coming out of the ground with branches and green things on them' we call this image 'a tree'.

Objects are individual and specific. No two objects are identical; every object is distinct and separate from every other object. A number of observations of different objects translates into a number of distinct and separate images. So why do we regard these distinct images as similar, as belonging to the same class?

You seem to be saying that it’s on the basis of perceived similarities. In effect, that is to say that we regard some objects as similar because we regard them as similar, which is to beg the question.

Brendan