Rebirth of Reason

Post to this threadMark all messages in this thread as readMark all messages in this thread as unreadBack one pagePage 0Page 1Page 2

Post 40

Monday, October 7, 2013 - 7:35amSanction this postReply
How is induction not "pattern-finding"? 
Michael Philp:Optical illusions tell us that our senses and our perceptual mechanism have a nature and identity. Apparent contradictions between percepts and what we think we are seeing.
Ed Thompson: I  agree with Michael P. that illusions actually illuminate a counter-intuitive, underlying objectivity to sense-perception.
These are learned.  They are not inherent in nature.  I have examples from numismatics that also support the claim that other peoples do not have our vocabulary of line, figure, and space.  They do not see coins the way we do.  In fact, some ancient Celts created three-dimensional images with coins that could only be seen when viewed obliquely. The Renaissance taught us to view everything front-on. 
Eliot Temple: "What most people mean by "induction" varies a great deal.  Most professional philosophers ...  Most people ...  Another option ...   I recently spoke with some Objectivists ...  One strategy is to elicit from people *their* ideas about induction, then address those. That poses several problems. For one thing, it means you have to write a personalized response to each ...  How do you argue with people who have only a vague notion of what "induction" is...  "
You do not have to argue with anyone about anything. You seem to have a point to make. Make it and move on to the next thing. 

Understanding what other people mean by the words they use has been a serious problem since the invention of words. Color-blindness was not described until John Dalton explained his own in 1798. It is a conjecture in linguistics that people do not invent words for "purple" and "brown" until after they have differentiated "green" from "blue."  Lower case positivism, objectivism, realism, instrumentalism, and rationalism, are easy to use as synonyms. 

A Christian conservative may decry the lack of morals caused by rationalism and humanism.  You do not get far arguing what those words "really" mean - that rationalism is not empiricism and so on. You can "win" the argument by pressing them to the wall with demands for definitions and then proving them not just wrong but ignorant  You probably will not convince them of much.

Post 41

Wednesday, January 15, 2014 - 5:38pmSanction this postReply

You're obviously correct that induction is wrong, a lot.
My point, following Popper, is that it was never right.

Both the inductions of Bacon and Mill state that the accumulation of data will gradually create a picture of causality. Hence. a proper scientific method will yield a valid statement of why things happen on its own

For Bacon, it was the rejection of Aristotle's form and telos; neither need to be postulated. Hypotheses, rather, focus on effective cause, alone.

For Mill, it was the formal representation of admitted causal properties with admitted effects--the famous 'four possibilities' of an ABCD -- efgh relationship. Classically speaking, his method more or less equates 'cause' with probability: small wonder that statistics was just being developed!

Here, Popper was the great nay-sayer, stating, "What's to keep a researcher from accumulating a mound of data that only favors his/her own hypotheses?"

In this case, the mounds would be irreductable, thereby throwing the argument back into who's more consistent with such and such a theory...pre-Baconian scholastico-Aristotelian redux, as it were.   

In other words, data-piles  talk to each other by means of the possibility of refutability. For example, hypotheses: the velocity of light is either constant, or varies with respect to direction. Hypotheses: Lorentz modifies Newtonian mechanics. These are testable, hence refutable.

Hypotheses: 'Class struggle is the motor of history': or, 'Our dream world is a symbolic domain in which all of our repressed desires is played out'. Non-testable, hence, nonscience.

Now perhaps ther's another usage of 'induction' that means...
"Reasoning from particular to general with respect to the theoretical platform of said inductivist'.
But because this daffy-nition, is agreeable to everyone, it's sterile.

What's always been at stake is why certain hypotheses are proposed to be tested, and why some data is considered more important that others.

For this, Popper either had no answer, or answered poorly. In short, refutability is only part of the story, as important as it is.

So whatever might be a satisfactory outcome (eg, do the laws of physics lie?), it won't include 'induction'. That talking horse was sent to the glue factory a long  time ago; he had nothing interesting to say.


Post 42

Saturday, January 18, 2014 - 7:13pmSanction this postReply
Bacon endorsed a more original Aristotelian conception of forms while attacking the scholastic conception of substantial forms that the scholastics mistakenly called Aristotelian. Bacon's project was essentially a return to Aristotelianism, but looked (to many, including himself sometimes) like an attack on Aristotelianism because what was going by the name was no longer Aristotle's own view. Some, like William Harvey and some involved with the new humanist Aristotelianism that got going across Europe in the 1540s, saw Bacon's project for what it was. He thought of it as Aristotelian. Also, Bacon did not use hypotheses. His first vintage is a primitive integration (or simply the smallest valid integration you can make given the facts). A proper vintage is not a hypothesis. It is a minimal interpretation that takes you from concepts to propositions. That is a huge step forward.

Mill was an "ultra-empiricist," and he rejected any contribution of the mind as a part of induction, any "subjective" element of knowledge. The only important element of the acquisition of knowledge was the "objective" element, the facts and observations of them. He ignored the mental aspects of induction, just like enumeration does: the inference is reached merely from observations, with no thought as to what the mind does to form the induction. Depending on your metaphysics, Mill's rules lead to either misintegrations or a dismissal of true induction as an impossible pipe dream. Herschel admitted that he's hypothetical but Mill did not, which makes his rules problematic. Personally I'd say that everyone who reads Mill should revisit Mill's rules and then immediately read Mercier's work.

I would expect a Popperian to find an interesting association, hypothesize a cause, and try to shoot down the hypothesis before handing it over to other scientists to shoot down. If it survives bombardment, it sticks around as a convenient explanation. I don't see anything wrong with trying to shoot down an idea. But there are far better methods than forming a hypothesis. Hypo-thesis i.e. below a thesis; as in not worth of stating as a thesis, as in you're not ready yet, as in you're jumping ahead of facts. Also not everything can be tested. The concept of natural selection is not testable. It's a concept. Not a theory. It is not difficult to see what measurements are omitted (particular organisms, particular traits, particular adaptation, particular selection factors...Those would be omitted as if they were measurements. You leave them unspecified on the principle that such things matter metaphysically but aren't part of the selection criteria for conceptualization..). Falsifiability is structurally similar to the way Kant uses the idea of "universality" in morality (the idea that a moral principle has to apply in the same way to everyone to be valid). It's useful in a certain sense (it would tell you to look for important differences), but it's not really a fundamental issue - it arises because of mistakes, and focusing on it distracts away from a much larger pivotal issue.

Some additional thoughts on all this:

If you need a hypothesis simply to get some practical work done, try to falsify it. That will help you think of experiments. If you have to guess, use your guess to gain more information about reality. Name your guess for what it is. Then think of experiments. Then dump the guess and wait for a prerogative instance. (What I have yet to figure out is how to discover more prerogative instances.) If you have so little that you can't make a first vintage according to Bacon's rules,then you can't induce. All you can do is try to test your pre-wine; the mushed grapes that may or may not become wine. (I had wanted to work out Bacon's definition of a prerogative instance by abstracting characteristics from his examples of prerogative instances and then applying Rand's rule of fundamentality. I have yet to do so.)

(Edited by Michael Philip on 1/18, 9:28pm)

Post 43

Saturday, January 18, 2014 - 9:15pmSanction this postReply

How is induction not "pattern-finding"?

How is pattern indicative of causality? I know that a sunrise depends on the earth rotating. No pattern was necessary.

(Edited by Michael Philip on 1/18, 9:55pm)

Post 44

Wednesday, January 22, 2014 - 5:33amSanction this postReply
Michael Phillip,

I believe that you’ve made an excellent point in stating that Bacon attacked the Aristotelians of his day; there might be a ‘true’ Aristotle that’s left standing after the polemic is over.


This is because even though he was a fierce polemicist, Bacon was far more concerned in going after ideas than people.


For example, there was a ‘Paduan school’ that he simply left alone, because they wrote long essays on the ‘real’ inductiveness of Aristotle’s science. Those whom he attacked were of the more orthodox sort: deduction works because things present themselves to the senses as they actually are.


Well, not, said Bacon.


Central to the issue is Bacon’s observation that things do not necessarily present themselves as true. The Aristotelian model (accepted by Rand) says, basically, that if you observe properly and think properly, you’ll derive truth.


Transcribed into a philosophy of language, this indicates that a language- based concept adheres to its object as real. This new position was taken up and re-developed by Kripke


(Now Rand discusses this position in what I’ve seen as rather indirect, yet assertive. I’m still looking for specifics…)


Bacon’s response to either Aristotle or his scholastic followers was, in a nutshell, “The New Method”, or not the Aristotelian business as usual. On a more personal note, Bacon penned the famous expression, ‘Crooked Mirror’ to contradict Aristotle’s own ‘Tabula Rasa’.


In any case, ‘method’ means to devise a way of doing something, or performing any task, that doesn’t come naturally. So herein lies the rub. Science needs a method because normative thought is inadequate.


The key word to this process, as given,  is induction. No telos or formal cause can offer us a deductive indicator.


That science isn’t like daily thinking through somewhat resembles the recent work of Kahneman and Tversky. Normally, we use heuristics, science demands more. ‘Method’, in any case, would be a way of ensuring that one doesn’t lapse back into said heuristics.


Most all of philosci discussion begins with this critique of thought that began with Bacon. Science is a rig-up, a procedure that tries to isolate us against common bias, of which Bacon said there are three.


In this light, Popper‘s just another amendment, important as he might be. Induction presupposes far too much. If everyone inducts, yet comes up with different results, then of what use is the term…etc…? Well, let’s see if the researcher offered us data that was falsifiable….


On and on…assuming interest. Cartwright, for example, is big on saying that laws of science lie because of the extraordinary nature of the rig-ups we have to perform. What we do is develop ‘simulacra’. For her, cause always relates back to use, because that’s the only realistic standard we have.


Lastly, because Rand accepted both the Aristotelian model of science and language, her notion of ‘induction’ must be somewhat different than that defined by Bacon. Again, for him, we induct because concepts don’t just naturally adhere to things. We have to find a method to make it so.


Post to this threadBack one pagePage 0Page 1Page 2

User ID Password or create a free account.