Appraising Analogical Arguments

[darkchild]

"Appraising Analogical Arguments"

I am self-studying an oldish text on introductory logic by Copi. The text lists some criteria by which arguments by analogy may be judged. Here are three of them:

"...the number of respects in which the things involved are said to be analogous."

"...the number of disanalogies or points of difference between the instances mentioned in the premisses and the instance with which the conclusion is concerned."

"...the more dissimilar the instances mentioned in its premisses, the stronger is the argument."

I am confused because the third criterion seems inapplicable to arguments by analogy. This criterion is illustrated by the following example: a particular student is likely to do well in college if ten students of similar background (such as high school grades) also do well, and this argument is strengthened if the analogy involves ten students of different backgrounds. But if the ten other students are of different backgrounds, in what sense is an analogy being made; what is/are the point(s) of comparison? In fact, such an argument seems to be the exact opposite of an argument by analogy, because to argue for the collegiate success of one person based on the collegiate success of a bunch of people whose only similarity to the first is their success in college is to effectively claim that neither background nor anything else effects one's probability of doing well in college. Furthermore, the third criterion does not jibe with the first two.

 

[Stephen Tashi]

Not all studies of "logic" are studies of "mathematical logic". As I recall, Copi's books do contain treatments of mathematical logic - also known as "symbolic logic". However, the passages you are quoting don't deal with mathematical logic. They deal with the "logic" used in arguments by columnists, lawyers and debaters. (For example, the subject of ad hominem arguments is a topic for that type of logic, but not for mathematical logic.)

You aren't posing a question that has any mathematical interpretation. Of course, given members of the forum have such wide expertise, perhaps some expert in the logic of debaters will comment on it.

 

[darkchild]

Not all studies of "logic" are studies of "mathematical logic".

I know all of this. I didn't post the question because I thought that it is related to mathematical logic. "Logic" is in the name of the forum, so I assumed that discussion pertaining to all types of logic were permissible. If inductive logic is out of place in this forum, surely statistics is as well.

 

[Stephen Tashi]

I'll leave it to some moderator to comment on whether the logic of debaters is a topic for this section of the forum. (I don't have an axe to grind one way or the other.)

Mathematical questions about statistics are not out of place.

 

[blue_raver22]

I understand the importance of evaluating arguments by analogy in a logical and systematic way. In this case, I believe the third criterion listed is not applicable to arguments by analogy.

Analogical arguments rely on the similarities between two or more things to make a conclusion about the thing being discussed. Therefore, the more similarities between the instances mentioned in the premises and the instance with which the conclusion is concerned, the stronger the argument. This aligns with the first two criteria listed.

However, the third criterion, which suggests that the more dissimilar the instances mentioned in the premises, the stronger the argument, does not fit with the concept of analogy. Analogies are based on similarities, not differences. In fact, the more dissimilar the instances are, the weaker the analogy becomes.

In the example given, the argument is strengthened by including students with different backgrounds. However, this does not make the analogy stronger. In fact, it weakens the analogy because the success of these students may be attributed to other factors besides their background.

In conclusion, as a scientist, I would suggest that the third criterion be disregarded when evaluating arguments by analogy. Instead, we should focus on the number of similarities and the number of disanalogies between the instances mentioned in the premises and the instance with which the conclusion is concerned. This will provide a more accurate evaluation of the strength of an analogical argument.