The Use Of Analogy In Argument

[DavidOdden]

I'm really surprised to see you espouse this view. This seems similar to the case Francis Bacon made against deduction, in favor of his revised induction. I don't see how you can come to the conclusion that deductive inference cannot create new knowledge, besides through deduction from a syllogism. And then, have you said something new, or not? Doesn't this approach prevent one from abstracting beyond perceptual level concepts?

Keeping in mind that I was addressing classical deductive logic, I would turn the question around and ask what is an example of using deductive logic to create significant new knowledge? Abstraction -- concept formation -- cannot be done through deductive inference, so if I were bound to use only deductive inference, I don't see how I could form any concepts (I could never get the open-endedness of concepts). If I as fact "It's raining" and similarly "The dog is wet" then deductive inference allows me to say "It is raining and the dog is wet" -- not such a big deal. More dramatic-seeming pieces of deductive inference are of the type "If we assume that all men are mortals and also suppose that Socrates is a man, then we can derive the new knowledge that Socrates is a mortal" -- again, I don't see this as a significant new piece of knowledge. Now, not all deductions of this form are as trivial, but not all universal statements are so easily admitted into the set of valid premises. Once you're validated the universal statement (which isn't a deductive process), there's not much you can do with it that constitutes new knowledge. The important new knowledge comes in integrating a set of disparate statements into a single universal one.

I presume I'm not the only person saying this, but I can't point with certainty to anyone else who thinks that deduction isn't all it's cracked up to be. Maybe I'm a nutcase, but I stick by my position until someone can show the flaws in it.

Edited January 25, 2006 by DavidOdden

[pastoral engineer]

(reffering to line about Wittgenstein)

It's my understanding that this is an expression of the philosophical tradition known as nominalism. It can be traced back to the medieval philosopher Roscelin, who claimed that universals did not exist, but were instead mere flatus vocis: "puffs of air."

Actually, he would not have said mre puffs of air. Rather the information in the speech is not restricted to the words, but occurs with the action of speaking (writing, etc.) The effect the speech has on the audience is part of the whole and takes into account the context of the audience. Merely discussing such issues does not further a philosophical argument, but the community formed in the discussion and any resulting action or transformation begins to form a greater degree of communication.

Basically, you have your discussion and you have your meta-discussion. By taking into account what is going on in the meta-discussion, you have a more objective picture of what is going on the discussion. Recognizing that there is no easy solution to the problem of universals allows one to get closer to the actual solution. It is similar to carrying error calculations along with your work in science. By knowing all the ways you are likely wrong, you become more right.

This is the sense where the value of an argument is not whether it is logically solid, but whether it is convincing to the audience. In this more important sense, pure logic is not always successful and analogy may be more so.

Clinging to pure logic when even that which is supposedly axiomatic still needs to be checked (involving prior and prior premises) is a poor substitute for achieving the goal of persuading the audience.

[softwareNerd]

This is the sense where the value of an argument is not whether it is logically solid, but whether it is convincing to the audience. In this more important sense, pure logic is not always successful and analogy may be more so.

Polemic which is convincing in the sense of being beguiling is not good polemic. Good polemic is one that best allows the audience to see that what is being stated is true.

Truth is: correspondence with reality. Good polemic is constructed to demonstrate such correspondence in an honest fashion. [Leaving aside outlier situations like negotiating with criminals.] Good polemic thus also contains the seeds for its own refutation, and does so because of its clarity. Analogy is "pointing to reality". In demonstrating correspondence to reality, it is useful to point to other similar things in reality.

Having said that, I have little idea if anything new is being said in this thread. Pastoral engineer, you seem to be repeating the same theme in each post: i.e., that analogies are very useful and often more convincing than formal deduction-from-premises arguments.

[Hal]

Language is simply grammar and definitions this is precisely a formal system that is subject to the incompleteness theorm. You begin with your definitions

No, thats a horribly idealised view of how language works. Natural language definitions are not (and cannot be) precise in the sense that a formal system demands. There are no rules of inference, nor are there rules for constructing new sentences from existing ones.

You also dont seem to understand Goedel's theorem. Its not a general statement about 'axiomatic systems', its a statement about a particular class of formal systems, namely those powerful enough to allow the formulation of number theory. You cannot formulate number theory within the 'formal system' of ordinary language because a) there are no axioms, there are no rules of inference, c) there is no decision procedure for what constitutes a valid theorem of the system, d) the meaning of natural language sentences are inherantly tied to their use in reality rather than being something you can specify by purely formal semantics, e) language use is not rule-governed.

edit: I notice you cited Wittgenstein. Assuming youve read his later work, its baffling how you can hold this ultra-positivist view of language, because its precisely the sort of thing he was reacting against.

Though it may not be the perfect choice for the word, I am reffering to all logical processes that attempt to begin with solid ground - a set of axioms - and attempt to build a structure of true facts. All such systems fail to create significant new knowledge as they all proceed directly from the set of axioms. They only result in an inherantly limited formal system.

This is incorrect: see the whole of mathematics. Axioms can have purely deductive yet highly non-trivial consequences. Edited January 25, 2006 by Hal

[dondigitalia]

No, thats a horribly idealised view of how language works. Natural language definitions are not (and cannot be) precise in the sense that a formal system demands. There are no rules of inference, nor are there rules for constructing new sentences from existing ones.

Hal is right. Definitions are always contextual, and are subject to change depending on usage and an individual's current knowledge.

Pastoralengineer, are you familiar with the Objectivist theory of concepts?

[DavidOdden]

Basically, you have your discussion and you have your meta-discussion. By taking into account what is going on in the meta-discussion, you have a more objective picture of what is going on the discussion.

I have a suggestion -- on that I very strongly urge you to follow. Just take this one statement of yours and explain it, and be prepared to defend it. I understand what a discussion is, and the term "meta-discussion" does not mean anything (has no referent) as far as I can tell. On the other hand, if you are talking about literal meaning and conversational implicature, then it is true that both things exist, but you should use the accepted terminology for talking about these things -- or at the very least declare "I don't like the expressions 'literal meaning' and 'conversational implicature' so I'm going to refer to these as 'discussion' and 'meta-discussion'." You should also make explicit what you mean by the vague expression "going on in the discussion". Are you referring to the purpose of the discussion? The conclusions drawn? The sum of evidence and logical structure?

There are significant signs of nihilism in your posts, which seems to arise from implicitly accepting as axiomatic that knowledge is impossible. I think that if you start to think about your claims more literally, you would probably realise that you don't really mean a lot of what you have said. I could be wrong, and I'm not saying that I think you are being insincere, just that you've let a false premise corrupt your epistemology. The remedy, I believe, is to understand the relationship between words, concepts and existents, and the Objectivist theory of that relationship is, from what I have observed, a correct recognition of that relationship. If you disagree, say why you disagree (disagree that it is, not that this is my finding).

[Bold Standard]

More dramatic-seeming pieces of deductive inference are of the type "If we assume that all men are mortals and also suppose that Socrates is a man, then we can derive the new knowledge that Socrates is a mortal" -- again, I don't see this as a significant new piece of knowledge. Now, not all deductions of this form are as trivial, but not all universal statements are so easily admitted into the set of valid premises. Once you're validated the universal statement (which isn't a deductive process), there's not much you can do with it that constitutes new knowledge. The important new knowledge comes in integrating a set of disparate statements into a single universal one.

I presume I'm not the only person saying this, but I can't point with certainty to anyone else who thinks that deduction isn't all it's cracked up to be. Maybe I'm a nutcase, but I stick by my position until someone can show the flaws in it.

I don't think you're a nutcase, but you seem to be making the mistake of assuming that man acquires knowledge in the same fashion as angels do, in the theory of Thomas Aquinas. Being creatures of pure form, angels, say Aquinas, come to grasp a universal directly from God; and in one act, they simultaneously gain complete awareness of every concrete and abstract implication from the universal. But human beings don't acquire knowledge in this fashion. Knowledge of the implications of a given universal requires a separate cognitive act for every deduction.

Knowing that all men are mortal, and that Socrates is a man, does imply that Socrates is mortal, but it does not guarantee that I, as a human being, will grasp that Socrates is mortal, unless I perform the distinct act of deducing it from my premises. As far as whether this new knowledge is "significant," well that would depend on context. If I (assuming I were living in ancient Greece) decide to miss a lecture by Socrates, because I think I can see him in twenty years, when I have a better grasp of philosophy-- not realizing that he's mortal and will probably be dead by then.. then it would have been significant.

Recognizing that there is no easy solution to the problem of universals allows one to get closer to the actual solution.

No easy solution, I'll grant. But that doesn't mean no solution.

This is the sense where the value of an argument is not whether it is logically solid, but whether it is convincing to the audience. In this more important sense, pure logic is not always successful and analogy may be more so.

Clinging to pure logic when even that which is supposedly axiomatic still needs to be checked (involving prior and prior premises) is a poor substitute for achieving the goal of persuading the audience.

Surely there are valid reasons to carry on arguments besides attempts to persuade your audience. What about attempts to learn, and to work out possible contradictions in your own conceptions? What about simply trying to understand the position of your audience, so that you may know how to persuade them in the future? I can think of a thousand other examples. Unless these fall under some "meta-persuasion" it seems "pure logic" is still important, since I want to be logical.

What's the use of persuading an audience with an argument that's not even logical, so that they will be persuaded half an hour later with another argument, or with the practical requirements of reality, like Hume when he left his books and tried to play backgammon?

[DavidOdden]

I don't think you're a nutcase, but you seem to be making the mistake of assuming that man acquires knowledge in the same fashion as angels do, in the theory of Thomas Aquinas. Being creatures of pure form, angels, say Aquinas, come to grasp a universal directly from God;

It is incomprehensible to me that you could think for a second that I believe in the existence of god, for not one thing I have said here or indeed anywhere for at least the past 45 years could in any way be interpreted as supporting that claim. Since I don't believe in the existence of god, and nothing I have said could be reasonably interpreted as implying that, then I cannot imagine how you deduce this piece of knowledge, that I hold that man acquires knowledge from said god. Now then, what did you really mean to say? I will tell you that I believe that man acquires knowledge via perception of reality -- by introducing new statements about existence and inductively forming new generalizations. This is the source of new knowledge.

But human beings don't acquire knowledge in this fashion. Knowledge of the implications of a given universal requires a separate cognitive act for every deduction.

I'm still waiting for an actual example of learning something significanc and new by deduction. Where is your example?

Knowing that all men are mortal, and that Socrates is a man, does imply that Socrates is mortal, but it does not guarantee that I, as a human being, will grasp that Socrates is mortal, unless I perform the distinct act of deducing it from my premises.

You're looking for a psychological guarantee that simply does not exist. Nothing at all will ever guarantee that a person will grasp any fact. At any rate, are you conceding that the best you can ever do with deduction is discover inanities such as "Yes, even Socrates is mortal"? It seems your entire argument is over elevating trivialities to the status of major knowledge. There are infinitely many such statements that we can form, for example given various premises based on gravity, I can deduce the new knowledge that this coffee cup will fall if I let it go, that coffee cup will fall if I let it go, so will yonder cup, this piece of chalk, this pencil, that pencil, his pencil and so on. You can invoke context all you want, I do not believe for a minute that the Ancient Greeks were the least surprised to learn that Socrates, Plato and Pyrrho all died.

At any rate, if the entirety of your point is that you can model the process of making explicit the formal "consequences" of word-relations via deductive logic, then I concede that point, though I do not consider that to be significant knowledge or new knowledge. It is a repackaging of old knowledge, involving nothing about reality that you were not already aware of, and constituted simply focusing on and emphasizing a narrow aspect of something that you already knew. Are you equating "focusing on, emphasizing" as constituting what it means to "know" something? For example, if I'm not right now focusing on the fact that I need to get gas in the next day, does that mean that I don't know that fact?

[Bold Standard]

It is incomprehensible to me that you could think for a second that I believe in the existence of god...Now then, what did you really mean to say? I will tell you that I believe that man acquires knowledge via perception of reality -- by introducing new statements about existence and inductively forming new generalizations. This is the source of new knowledge.

No, I didn't mean to imply that you believe in God or that Aquinas' angels actually exist. I just meant to illustrate the point that induction via perception of reality, while it certainly does provide knowledge, and is the ultimate source for all knowledge, does not by itself provide man with all the knowledge which can only be acquired by deducing from this knowledge. The example of Aquinas' angels is, in the spirit of the thread, just an analogy. But it's a helpful one for illustrating that which a human consciousness is not.

I'm still waiting for an actual example of learning something significance and new by deduction. Where is your example?You're looking for a psychological guarantee that simply does not exist. Nothing at all will ever guarantee that a person will grasp any fact. At any rate, are you conceding that the best you can ever do with deduction is discover inanities such as "Yes, even Socrates is mortal"?

No, because (and only because) I don't consider that point to be inane, but instead that it is new knowledge, not explicitly contained in the premises.

It seems your entire argument is over elevating trivialities to the status of major knowledge.

This is important-- I would never elevate deduction over induction, because induction is prior, and is the necessary foundation upon which deduction must be based if it is to be of any practical significance. But that doesn't mean that deduction is completely arbitrary or that induction alone can provide man with sufficient knowledge upon which to sustain his life.

There are infinitely many such statements that we can form, for example given various premises based on gravity, I can deduce the new knowledge that this coffee cup will fall if I let it go, that coffee cup will fall if I let it go, so will yonder cup, this piece of chalk, this pencil, that pencil, his pencil and so on.

Yes, you can. Many professional philosophers could not-- including David Hume and his followers. He says, just because the sun rose this morning, we have not one iota of a reason to suppose that it will rise tomorrow. Of course, he's wrong. And his flaw was in defaulting on induction, not deduction. But giving up on deduction, if done consistently, would eventually put a person in a similar epistemological predicament.

You can invoke context all you want, I do not believe for a minute that the Ancient Greeks were the least surprised to learn that Socrates, Plato and Pyrrho all died.

That's because the Greeks believed in deduction, and that it resulted in new knowledge. If you can imagine a creature who was capable of induction, but whose cognitive abilities stopped there, it would be very much surprised by this phenomenon. It would only be able to grasp general principles, but could not apply them to the concretes with which it came into contact. What good would that be? It would literally be an absent minded professor, who's life could only be sustained by trading information with some other creature who was capable of deducing concrete instances from its broader abstractions. My only point is that such a creature would not be a human being, we need both induction and deduction to grasp reality.

At any rate, if the entirety of your point is that you can model the process of making explicit the formal "consequences" of word-relations via deductive logic, then I concede that point, though I do not consider that to be significant knowledge or new knowledge. It is a repackaging of old knowledge, involving nothing about reality that you were not already aware of, and constituted simply focusing on and emphasizing a narrow aspect of something that you already knew. Are you equating "focusing on, emphasizing" as constituting what it means to "know" something? For example, if I'm not right now focusing on the fact that I need to get gas in the next day, does that mean that I don't know that fact? [bold added]

It's possible to arrive at such premesis as "all men are mortal" and "Socrates is a man" without ever being specifically aware that "Socrates is mortal." It would be easier to elude a deductive conclusion when dealing with more sophisticated premises. You needn't focus on something in perpetuity in order to say that you know it. But you can hardly claim to know something if you've never even been aware of it-- even if it does follow from your premises. Can one deduce Ayn Rand's epistemology, the necessity of Laisse Faire Capitalism for eudaimonia in the widest sense, and, say, Tara Smith's meticulous approach to ethics from general principles all contained in the ancient works of Aristotle? I'd say, yes, with surprisingly few and minor clarifications and at most one or two new inductive principles. But that doesn't mean that Aristotle would have had the faintest notion of any of those things. If he had seen them, it's likely he would have said, "Of course! Why didn't I think of that?" The reason is that he didn't engage in extensive enough analysis and deduction to come to all of those conclusions. But you can hardly blame Aristotle for some of the wrong conclusions he came to. You can't say, "Stupid Aristotle! How could he have allowed those statist premises from Plato to remain in his politics? Didn't he know A is A?!" That's unfair, true as it might be, because Aristotle was not Aquinas' angel.

Who knows what new and vital ideas will be deduced from Objectivism in the future? We're not Aquinas' angels. All we know for sure is that it will not contradict what we already know..

[DavidOdden]

This is important-- I would never elevate deduction over induction, because induction is prior, and is the necessary foundation upon which deduction must be based if it is to be of any practical significance. But that doesn't mean that deduction is completely arbitrary or that induction alone can provide man with sufficient knowledge upon which to sustain his life.

The problem is that you're arguing against a position that I don't hold. My position is that knowledge cannot arise by deduction alone. If you allow induction in logic -- as you must, for logic to be at all meaningful -- then you are not creating knowledge by deduction, you are creating knowledge by induction and deduction. You seem to be assuming that I hold that deduction is entirely superfluous. I am saying that pure deduction is inutile.

Can one deduce Ayn Rand's epistemology, the necessity of Laisse Faire Capitalism for eudaimonia in the widest sense, and, say, Tara Smith's meticulous approach to ethics from general principles all contained in the ancient works of Aristotle? I'd say, yes, with surprisingly few and minor clarifications and at most one or two new inductive principles.

Well if you're allowing induction in the argument, then you aren't deducing the conclusion, which is fine with me. I would be interested to see that deduction (say, applied to Smith's metaethics). I would be impressed to see this.

[Bold Standard]

The problem is that you're arguing against a position that I don't hold. My position is that knowledge cannot arise by deduction alone. If you allow induction in logic -- as you must, for logic to be at all meaningful -- then you are not creating knowledge by deduction, you are creating knowledge by induction and deduction. You seem to be assuming that I hold that deduction is entirely superfluous. I am saying that pure deduction is inutile.Well if you're allowing induction in the argument, then you aren't deducing the conclusion, which is fine with me. I would be interested to see that deduction (say, applied to Smith's metaethics). I would be impressed to see this.

Oh, I see, you're arguing against deduction divorced from induction. In other words, against the "floating abstraction" rationalist approach. I didn't think of the fact that if induction is present at the base of your knowledge, it's in a sense married to every act of deduction you perform from it. I was taking for granted that there must be induction for deduction to take place; but there could be, instead, a junk heap of unwarranted beliefs, prejudices, non-sequiters, and, as I mentioned, floating abstractions.

I'll have to set that Smith project as a long term goal. But I'll let you know if I'm able to do it!

One last point on Aquinas' angels. I don't think I made this point clear enough, but the essential characteristic of their means of acquiring knowledge is that they gain knowledge in one act of abstracting from a universal, not that they get it from God. The reason that they have to get it from God, according to Aquinas, is that being creatures of pure form, they have no sense organs-- no eyes, ears, mouth, etc., which exist as physical means of perceiving reality. So in order to interact with people as they sometimes do in the JudeoXian mythologies-- for instance, in the legend of Balaam, the angel would simply contemplate "valley," "donkey," and "man," and he'd know where to stand and when to speak, because he would have grasped every particular which could possibly be subsumed under those universal concepts. (At least, that's how I understand the issue as it was explained to me.)

I think it's fascinating as a contrast to human perception, and a little crazy that Aquinas put so much thought into defining to the last detail these completely imaginary beings..

[Thoyd Loki]

Analogy is not an argument. Bad analogies can muddy an argument. (I remember Plato's Republic being full of them, to the extent that I found it tedious to read.)

However, good analogies can be useful tools of thought. When considering X, one might say (for instance):

• X is like Y
  
• Y implies Z
  
• so X implies something like Z
  

Doesn't work as a formal argument. However, if X and Y are truly similar, "refuting" the "argument" forces one to ask what indeed is different about them. It forces one to look for differenting facts. It forces one to "chew" further on X and Y and deepen one's understanding of the two concepts.

I started thinking along these lines when I read the following:

http://www.nizkor.org/features/fallacies/straw-man.htmlAre there other fallacies that break a formal argument, but might be useful as tools of thought? For instance, how about "slippery slope". Suppose I claim X, and someone says that X is a slippery slope to Y. As long as the two are related (i.e. the objection is not completely arbitrary) considering why there is no such slippery slope can strengthen one's understanding of the subject.

Is the question mark at the end of my sentence in the quote in post #1 of this discussion invisible to everyone but me? I did not make an argument. I asked a question. I did not know that the analogies would be unrecognized as being lifted verbatim from Atlas Shrugged. But, then again, if you cannot recognize the difference between a statement and a question, that book was probably too hard anyway!

[Groovenstein]

If you think that someone has misinterpreted or misstated your comments, or has falsely imputed opinions to you, it is best to state that and show why. Your suggestion that Atlas Shrugged was too hard for anyone who has thus far participated in this thread is neither constructive nor appropriate.

[Thoyd Loki]

If you think that someone has misinterpreted or misstated your comments, or has falsely imputed opinions to you, it is best to state that and show why. Your suggestion that Atlas Shrugged was too hard for anyone who has thus far participated in this thread is neither constructive nor appropriate.

Sure, I will go through it a third, and final time. Here is the original post of mine that was quoted by someone else and posted on the opener of this thread.

Here is the direct question that I asked.

Do you mean to say that melody is a primitive vulgarity in music, as plot is a primitive vulgarity in writing, as logic is a primitive vulgarity in philosophy?

Instead of an answer to the question, I get a stupid link to a fallacy website. Trouble is, there was no argument posted.

If my question suggests that you have committed a fallacy, you do not get out of it by trying to tell me I have constructed a fallacious argument when I have made none. You say: "No, that is not my view." And then, if you please, you clarify what you mean. Or maybe even: "No, that's not my view. How did you get that out of what I said?"

A question is a request for information. It is not the positing of information, but the soliciting of it. I did, in that question posit my view of what he meant, but the meaning, that I was checking this against him for verification, was clear. Thus, the whole basis of this thread is based on erroneously quoting me in the act of fallacy, when, in fact, none was made. I cannot even say I was quoted out of context, the damned question mark is still there.

[softwareNerd]

Is the question mark at the end of my sentence in the quote in post #1 of this discussion invisible to everyone but me?

I am not sure if you quoted my post because it contained your quote (and someone else's response) within it, or if your comments were directed to what I wrote. Assuming the second, my post was actually in support of the appropriateness of asking the question you did.