Thoughts on my view of induction and deduction?

[LauricAcid]

A quick point: I didn't say that the Objectivist view is that attributes are not observed prior to concept formation. I recognize the assertion that there is a perceptual stage prior to a conceptual stage. But it woud be odd indeed to say that one can understand the relation between the nature of an entity and its mode of action without having formed concepts of the attributes that are the nature of the entity.

[dougclayton]

A quick point: I didn't say that the Objectivist view is that attributes are not observed prior to concept formation. I recognize the assertion that there is a perceptual stage prior to a conceptual stage. But it woud be odd indeed to say that one can understand the relation between the nature of an entity and its mode of action without having formed concepts of the attributes that are the nature of the entity.

Actually, I believe that is implied in what you did say. You said previously:

Thus, I would think, to identify the nature of something, one must identify its attributes. But attributes are identified by concept formation and concept definition.

You are correct that to identify the nature of something, one must identify its attributes, for "nature" means "identity" and one's identity is nothing more than the sum of its attributes (which, for concrete entities, can be identified perceptually). But you present this identification as necessarily happening after one has formed a concept of the object in question--and this I don't understand.

In the post I am responding to, you add "the relation between the nature of an entity and its mode of action." Given that its nature is nothing more than its attributes, why would the additional task of recognizing their relation to its mode of action require concept formation?

I will point out, again, that there is a big difference between the knowledge that "balls can roll" and "this particular object can roll because it has this-or-that feature." (I don't mean to insult your intelligence if this is obvious, but I suspect there may be some equivocation between the two notions.) The first requires concept-formation and the attendant determination of essential and unessential attributes, because those relate to concepts. But where there is a single unclassified entity, there are no concepts yet.

Did that address your response, or did I miss your point?

[hunterrose]

What's this if business? This is such a fundamental fact about reality on Earth that no normal adult human anywhere doubts that. Size and color are the two features that are the most robust basis for all biological taxonomies everywhere on Earth, meaning that they are the most obviously variable across species. So this is knowledge that you do not need to doubt in the slightest.

Well, I didn't mean that I think one way or another on the swans; I meant that it's possible that someone, somewhere isn't aware of the limits of the color range of a specific organism.

That withstanding, while we might argue that color is a wide-ranging trait, there are still limits. I doubt there are any purple cardinals, orange polar bears, or green men. The inductors may realize observe that the swans are milk white, dusky white, pure white, etc. various shades of white, but have no reason to believe that the swans' colors extend beyond these limits.

Now, I have no idea what you mean by "automatically". If you mean, would a being with a conceptual faculty and no knowledge whatsoever instantly say "The proposition 'all swans are white' is invalid", then no.

Not exactly. I mean that [it seems] it was said that "swans are white" is invalid no matter how narrow the context of knowledge involved.

If it was meant to apply to all contexts of knowledge, it seems a bit strange:

x is invalid in all contexts. We can know that it is invalid, regardless of the limits of our particular knowledge?

If it wasn't meant to apply to all contexts of knowledge, which ones did it apply to? The ones in which it is known or suspected that all swans aren't white? If that were the case, it wouldn't seem to tell how to determine if an induction is valid - other than show it is incorrect.

[dougclayton]

The inductors may realize observe that the swans are milk white, dusky white, pure white, etc. various shades of white, but have no reason to believe that the swans' colors extend beyond these limits.

But they have no reason to believe the color is bound by those limits, either--and that's the point. They know only that swans tend to be white (in some given area), or are only known to be white. This is an entirely different claim.

Not exactly. I mean that [it seems] it was said that "swans are white" is invalid no matter how narrow the context of knowledge involved.

It is not the conclusion that is invalid no matter what the context, but the approach. It is wrong to observe N swans and conclude, "Now I know that all swans are white, because I have looked at all the swans in my town and every single one of them was white." It does not matter what one's context of knowledge is, because until there is some cause--some reason--that all swans would be white, one should not act like one knows it.

Remember that when we talk about the "all swans are white" claim here, we are not talking about the conclusion, which biology might someday be able to prove (assuming we hadn't already found black ones, that is). If there is something that causes swans to be white the way that their bones are caused to be hollow, that's fine. What we are critiquing is the method of induction via counting.

[DavidOdden]

Well, I didn't mean that I think one way or another on the swans; I meant that it's possible that someone, somewhere isn't aware of the limits of the color range of a specific organism.

There's something wonky about the question: it seems to me to involve contradictory assumptions. When you're talking of "swans" you're talking of two animal genera represented over a number of species. It's not rocket science, but it is at least bird science, and in order to be making valid claims about the nature of the genera coscoroba and cygnus, you gotta have a certain background knowledge, which would include more than your standard high school biology. This would therefore entail you knowing about color variation between (and within) species. If the hypothesis were "no swans are white" (a dumb hypothesis) or "no crows are white" (not as dumb), restrictions on color might be relevant -- but here, you have to deal with the improbable situation of uniformity in color. In multi-species genera, it is not common to find all species being a single color.

The contradiction is this. On the one hand, for your hypothetical case you want to assume a person who is more ignorant of the natural world than pretty much anyone alive, someone unaware of the fact that animals differ wildly in color. Show me this person! And yet, this person also has the presumed scientific sophistication to be making a claim about multiple genera.

Regression sucks. It's not about counting cases. I've been saying this and Doug has been saying this -- you have to move away from mere correlation, and look for causation.

[hunterrose]

It is not the conclusion that is invalid no matter what the context, but the approach... It does not matter what one's context of knowledge is, because until there is some cause--some reason--that all swans would be white, one should not act like one knows it.

What we are critiquing is the method of induction via counting.

I'll agree that "counting inductions" don't derive cause-effect relations, but "all swans are white" isn't necessarily an induction via counting.

When you're talking of "swans" you're talking of two animal genera represented over a number of species. It's not rocket science, but it is at least bird science, and in order to be making valid claims about the nature of the genera coscoroba and cygnus, you gotta have a certain background knowledge, which would include more than your standard high school biology.

I'm not checking, but I take it from your post that swans are of two species. Learn something new every day

Nonetheless, I don't think knowledge of genera, DNA, or even genetics, is necessary to validate inductions about living beings. Is it really necessary to find genetic cause for "all humans are volitional" before such a claim is valid?

A person may draw the cause-effect relation that 'white' is an inherent quality of swans. If genetics had shown that known swans didn't have a gene for other colors, "all swans are white" would be correct, but IMO no more or less valid than the prior, less "scientific" cause-effect relation.

If "all swans are white" isn't a counting induction, is the white swan induction invalid because it doesn't have a deep, "scientific" cause-effect relation, or because "inherent quality" cause-effect relations are taken to be invalid?

The contradiction is this. On the one hand, for your hypothetical case you want to assume a person who is more ignorant of the natural world than pretty much anyone alive, someone unaware of the fact that animals differ wildly in color. Show me this person! And yet, this person also has the presumed scientific sophistication to be making a claim about multiple genera.

Perhaps we disagree about this "animals differ wildly in color." I think virtually everyone would agree that there is some variance in color between animals of the same type. But, without knowing of contrary swans, swans may simply vary in the shade of white, or be limited to colors that could reasonably be associated as white (light grey, perhaps.) That might be the extent of "wild," as far as swans and particular people's knowledge goes. There is a rather large variance in what is considered white, and I don't think it unreasonable or naive that a person might believe swans are limited to that variance.

[jrs]

To those who really want to understand the Special Theory of Relativity (rather than merely paying lip service to it), I recommend reading at least the first part of Einstein's original paper.

http://www.fourmilab.ch/etexts/einstein/specrel/www

[DavidOdden]

Nonetheless, I don't think knowledge of genera, DNA, or even genetics, is necessary to validate inductions about living beings. Is it really necessary to find genetic cause for "all humans are volitional" before such a claim is valid?

Some such knowledge is required, if you are going to make claims about genera. If you are only making a claim about a particular kind of bird, like the swans of Spennymoor, then it isn't necessary to know anything about the same species of swan elsewhere, or about other species of swan. In the context of your knowledge -- specifically, given what one thinks "swan" refers to -- a person could make a valid inductive generalization "all swans are white" if in their knowledge context "swan" refers only to those northern-hemispheric birds which are indeed white.

If "all swans are white" isn't a counting induction, is the white swan induction invalid because it doesn't have a deep, "scientific" cause-effect relation, or because "inherent quality" cause-effect relations are taken to be invalid?

In this specific case, because there was reason to doubt the claim. The conclusion needs to be certain, and there were good reasons to consider alternative hypotheses. The conclusion was not tested in the southern hemisphere -- in. half of the world. It was known even then that animal color is highly variable according to variety of animal (even if they did not have a proper grip on "species"), and it was known that animal varieties varied geographically. These facts mean that the white swan theory is not certain, and cannot be considered certain, until it is checked that there are no non-white swans in the rest of the world. And indeed it turned out that there were plenty of non-white swans.

Black swans were discovered in 1697 and Linnaeus presented his taxonomy about 40 years later (and Mendel's book was over 100 years after that), so the question of causation -- as we understand it now -- would have been meaningless in that era. I have a policy of not extrapolating too much from the standards of modern science back to ancient science, especially when the old-days answer turns out to be "Swans are white because god made them be so". As far as I am concerned, the white swan induction was invalid because of the (temporary) failure to check it in about half of the world. I decline to pass judgment on their overall approach to science (indeed, I have no idea whether this was a scientific conclusion at all in that era).

I very strongly suggest that you abandon all questions about color in biology, if you want to understand induction.

[Free Thinker]

Betsey Speicher summed up what I've to realize perfectly:

"You determine what an object's characteristics are -- all of them -- using your senses. When you integrate entities into concepts, the defining characteristics are those characteristics that separate units of the concept from everything else you know (your context) that isn't a unit of the concept."

[AisA]

Fair enough, but, as I mentioned above, if by just observing balls roll we can say it is self-evident that they roll on account of their nature, then, by this method, what precludes one from saying that swans are white due to their nature?

Remember the context in which one first grasps that pushing a ball on a flat surface will cause it to roll. By the time one is able to grasp that principle, one has encountered many other types of objects which will not roll when pushed: chairs, tables, etc. Thus by observing similarities among balls and differences with other objects -- a process he must go through to form the concept of "ball" -- the child grasps a causal connection between the shape of the ball and its behavior when pushed on a flat surface.

No such causal connection has been made to show that the nature of swans causes them to all be white. This is the distinction between observing correlation and identifying causation.