General

Re: Induction == Empiricism?

Daniel James Sanchez — Sat, 31 Jul 2010 17:28:08 GMT

Induction is inferring generalities from particulars, as opposed to deduction which is inferring particulars from generalities.

Empiricism is the position that knowledge is derived from sense experience alone, as opposed to rationalism which is the position that knowledge is derived from reason alone.

Induction is not empiricism, because one can perform induction without any sensory data. For a very simple example, "every odd number times two equals an even number" can be known deductively, simply from considering the definitions of "even" and "two". Or someone, like a child, can be convinced of it inductively, by thinking, "well 3 times 2 is even, 5 times 2 is even (...) 375 times 2 is even. The latter reasoning is inductive, yet it is not empirical (it does not involve any sensory data).

Re: Induction == Empiricism?

ViennaSausage — Sat, 31 Jul 2010 16:45:24 GMT

"Actually empiricism is a whole school of thought insisting that knowledge comes from sense perceptions and observations. And induction would be a means to obtain knowledge by observing facts and drawing conclusions from them. For empiricism induction would be the prefered method to obtain knowledge."

Thanks Torsten!

This is what I was attempting to articulate, but erred in the confusion between a theory of knowledge (empiricism) versus a method of reasoning (induction/deduction).

Re: Induction == Empiricism?

Torsten — Sat, 31 Jul 2010 16:23:18 GMT

Of course not. Empiricism is a theory of knowledge; induction is a logical procedure. They're not even similar concepts.

Actually empiricism is a whole school of thought insisting that knowledge comes from sense perceptions and observations. And induction would be a means to obtain knowledge by observing facts and drawing conclusions from them. For empiricism induction would be the prefered method to obtain knowledge.

Re: Induction == Empiricism?

Jon Irenicus — Sat, 31 Jul 2010 15:34:15 GMT

With the dialogue thus far, I gather than one can use either deduction or induction as a form of argument from empirical data/observation. However, one cannot induct an argument from an axiom, since there is no data/observation available.

1. All men are mortal. Socrates is a man. Therefore Socrates is mortal.

That is deduction.

2. The Sun has risen reliably for all the existence of mankind. Therefore, it's reasonable to infer the Sun will rise tomorrow.

Induction.

Re: Induction == Empiricism?

AJ — Sat, 31 Jul 2010 11:35:37 GMT

"The proposition that "my sense data is accurate" is no less dogmatic and arbitrary than any other axiom."

First of all, see my edit above. It addresses "which comes first."

Here it is for readers who don't want to scroll up: Sense, then model (inductively hypothesize), then deduce...then sense, then model, then deduce some more. It's a continuous loop. So-called a priori reasoning just emphasizes some sometimes-nonobvious things that can be done in the modeling portion of the cycle. It essentially says, "Hey guys, your theory is broken, and I can prove it by pointing you to a model that is impossible for you to refute, but that simultaneously proves your theory wrong."

Second, what can you possibly mean by "accurate" if you don't trust your sensations as being sensations? Whether all this is a dream or there is a "real world," I am indeed experiencing whiteness as I (apparently) look at this screen. In other words, the words are getting in the way here, as they always do.

Or let's just take it to the practical level and say, what on earth is the purpose of theorizing like this if we are going to suspect the very pain and pleasure we experience as being possibly "inaccurate" or "unreal"? The point of any action is to experience more pleasure and less pain (now or in the future). Agreed, the reasons we might feel a pain could be completely fabricated, but the pain itself is very "real" in that it matters to me (and I assume you).

Re: Induction == Empiricism?

AJ — Sat, 31 Jul 2010 11:26:14 GMT

"Only propositions can be premises, because only propositions have truth-value."

I hope you mean, only thoughts can be premises. Propositions are just sets of signals (i.e., words) designed to cross the inferential gap between people, to get the listener to understand what the speaker intends in his own mind. Language is, after all, just a communication device (and perhaps a useful memory aid for our internal reasoning).

Light fishes last.

Is this saying that fishes that don't weigh very much have a long shelf life? Or is it saying that light is an entity that always waits for everyone else to finish their fishing before it starts? Or is it just three words thrown together randomly?

Certainly we can take one meaning (that is, one thought in our own mind that we suppose this string of words corresponds to) and evaluate whether it follows from our beliefs or not. But in the end we cannot help but evaluate a proposed thought or notion, not a proposition.

We speak in terms of propositions for obvious reasons of convenience (we can't see each other thoughts!), and because it's often possible to phrase things clearly enough to get most listeners to reliably reproduce a thought that corresponds with what we actually intended to convey:

There have been black dogs in the world.

There is some room for misinterpretation, but by and large we can safely assume that in a sufficiently non-rigorous context people will all think of the same idea when encountering this sentence. When the context gets very fundamental, like we are taking about basic questions of epistemology, all sorts of new ways to misinterpret this sentence arise. At some deep enough level of analysis, words become extremely patchy in their ability to cross the inferential gap separating speaker and listener. This is in constant evidence in all debates on these matters (which I would say are typically less like debates and more like "circuses of mutual misinterpretation").

Now here is the big thing I haven't mentioned that changes everything and accounts for why my ideas here may seem so alien: I believe all conscious thought is sensation, and I am prepared to prove that this is "obvious" even if the words I will use to do so will have to be very carefully chosen so as to obviate misunderstanding.

So that is (part of) why I call sensations "premises": if we call propositions "premises," we are really calling thoughts premises, but I say all conscious thoughts are themselves sensations. Therefore (at least some) sensations are premises. While this technically has the form of an empiricist result, the notion that "thoughts are all sensations" is a complete game-changer - AFAIK no one thinks this besides me. I would lay out my explanation of why this is so here, but I better do one thing at a time so I will stop here.

Re: Induction == Empiricism?

Jesse — Sat, 31 Jul 2010 11:14:07 GMT

AJ:

Sense, then induce, then deduce. If the words needed to describe this process were not so difficult, a child could explain it to us.

I disagree with this. Sensation can't be at the beginning (isn't this empiricism?) because sensations alone have no implications — not even ones that you can induce from. To proceed in this way, you would, at the outset and at the very least, have to include the proposition that "my sense data is accurate." As trivial as this may seem, it is a necessary step to move from sensation to induction. Once you admit this, your framework changes to "assume, deduce, test." The proposition that "my sense data is accurate" is no less dogmatic and arbitrary than any other axiom. (In fact, it is false.)

Re: Induction == Empiricism?

Jesse — Sat, 31 Jul 2010 10:55:21 GMT

AJ:

Oops, terrible word choice on my part. Data is a silly word, I agree - what does it really mean? Trusted perceptions? I should have said "premises" instead. IOW, premises that include all believed information and all other beliefs or imagined premises as well.

This shouldn't be controversial once we get past the language issues:

If you wake up one morning and experience a burning sensation in your hand, that sensation forms part of your starting premises. From there any reasoning about your situation would entail forming a model (a hypothesis) such as, "Perhaps my brass bedpost was heated by the sunlight and my hand is now resting against it." You could then try things like jerking your hand away, or looking to see if your hand is in contact with the bedpost. Those are new pieces of information that become part of your beliefs or provisional beliefs. You reason from those premises.

Okay, this sounds a lot better. Haha

Just to nitpick though, I don't think that it's a good idea to call sensations "premises." This can be misleading. Only propositions can be premises, because only propositions have truth-value. Certainly, sensations can lead you to make propositional statements, which can then become premises in an argument. It's important to keep this in mind though, particularly when the issue of universals come up, as there is no sensation which implies a universal proposition.

Re: Induction == Empiricism?

AJ — Sat, 31 Jul 2010 10:54:22 GMT

All reasoning rests on axioms which themselves cannot be proven; only tested. Moreover, these first principles (or axioms) must be propositional. Nothing can be deduced from a sensation.

Note we're saying the same thing here (as of course we must be; we're not fools here!). You experience sensations, then you create models (i.e., systems of first principles, axioms, hypotheses, etc. - they are categorically the same things, just with different degrees of certainty attached to them), then you figure out what logically follows from the models. Sense, then model (inductively hypothesize), then deduce...then sense, then model, then deduce some more.

It's a continuous loop. So-called a priori reasoning just emphasizes some sometimes-nonobvious things that can be done in the modeling portion of the cycle. It essentially says, "Hey guys, your theory is broken, and I can prove it by pointing you to a model that is impossible for you to refute, but that simultaneously proves your theory wrong."

Re: Induction == Empiricism?

AJ — Sat, 31 Jul 2010 10:43:12 GMT

Induction just means thinking up a model that seems to fit your beliefs (including the "data" that you trust and whatever other premises that comprise your beliefs about something - although the scientific method, for better or worse, only includes the repeatably measurable data among those "beliefs").
Deduction just means figuring out what logically follows from some set of premises, either what logically follows from your raw beliefs themselves OR from the provisional model you've come up with. In science, deduction usually means the latter.

Re: Induction == Empiricism?

ViennaSausage — Sat, 31 Jul 2010 10:36:42 GMT

With the dialogue thus far, I gather than one can use either deduction or induction as a form of argument from empirical data/observation. However, one cannot induct an argument from an axiom, since there is no data/observation available.

Re: Induction == Empiricism?

AJ — Sat, 31 Jul 2010 10:32:06 GMT

Oops, terrible word choice on my part. Data is a silly word, I agree - what does it really mean? Trusted perceptions? I meant it to include even your own beliefs, so I should have said "premises" instead. IOW, premises that include all believed information and all other beliefs or imagined premises as well.

This shouldn't be controversial once we get past the language issues:

If you wake up one morning and experience a burning sensation in your hand, that sensation forms part of your starting premises. From there any reasoning about your situation would entail forming a model (a hypothesis) such as, "Perhaps my brass bedpost was heated by the sunlight and my hand is now resting against it." You could then try things like jerking your hand away, or looking to see if your hand is in contact with the bedpost. Those are new pieces of information that become part of your beliefs or provisional beliefs. You reason from those premises.

Re: Induction == Empiricism?

Jesse — Sat, 31 Jul 2010 09:54:11 GMT

What I'm concerned with is AJ's assertion that "All reasoning is modeling (i.e., provisional induction based on data)."

This is a false dogmaticism of empiricist epistemology.

All reasoning cannot be based on prior "data" (by this I assume that you mean "empirical data") because all reasoning must start somewhere. To say that everything rests on prior data, empirical or otherwise, leads to an infinite regress. All reasoning rests on axioms which themselves cannot be proven; only tested. Moreover, these first principles (or axioms) must be propositional. Nothing can be deduced from a sensation.

Re: Induction == Empiricism?

Laotzu del Zinn — Sat, 31 Jul 2010 09:47:53 GMT

"This sounds an awful lot like empiricism..."

It is... because empiricism is not a method of argument. It is knowledge based on expirment, observation, and/or data. Deduction/induction are forms of arguing for/against said empirical (or otherwise gained) knowledge.

Re: Induction == Empiricism?

Jesse — Sat, 31 Jul 2010 09:41:28 GMT

AJ:

This isn't complicated:

All reasoning is modeling (i.e., provisional induction based on data)

This sounds an awful lot like empiricism...