The Role and Limits of Analogical Argument: A Reply to Aronovitch
| Date | 01 March 1997 |
| DOI | 10.1111/1467-9248.00072 |
| Author | Michael Lessnoffs |
| Published date | 01 March 1997 |
| Subject Matter | Article |
The Role and Limits of Analogical
Argument: a Reply to Aronovitch
MICHAEL LESSNOFF
University of Glasgow
Aronovitch's case for analogical argument in politics is extremely interesting,
thought-provoking and stylish, but I think he claims too much.1 Particularly
vulnerable, I suggest, is the claim that analogical reasoning as he recommends it
can provide `rational and non-relativist' foundations for normative claims in
politics. Admittedly, Aronovitch has written here not foundations but `founda-
tions, (in so-called scare quotes) which might suggest some hedging of bets. I
will return to this point; but there are several other problems besides.
These problems are divisible into the logical and the political, but I hope
to show that the two are connected. Aronovitch is well aware of one of the
logical problems, namely the relation of analogical reasoning to general-
izations. His stance throughout is to belittle and even deny any such relation:
good analogical reasoning `by no means implies that a generalization is really
wanted, [nor is it] implied or needed' (p. 81), although an accumulation of
analogies might result in one. This surely is false, as can be shown by examin-
ing some of Aronovitch's examples, starting with the ®rst. He suggests the
following reasoning: I dislike the writing of GarcõÂa-Marquez, which is
surrealist; Rushdie's writing too is surrealist; therefore (he infers) I will dislike
Rushdie also. This inference surely implies that I dislike surrealist novel(ist)s
generally ± a generalization that Aronovitch has in fact more or less smuggled
into the example (he says I have a sober, realist temperament, presumably
averse to literary surrealism). If the smuggled generalization were dropped, the
®rst proposition of this analogical argument would state a mere coincidence,
not a causal or explanatory relation, and nothing could be inferred. The fact
of the matter is that I do not much like GarcõÂa-Marquez, but greatly admire
Rushdie, even though both are describable as surrealist novelists. Maybe
Rushdie is just a better surrealist writer, or better than GarcõÂa-Marquez
appears, in translation, to me.
Aronovitch's analogy between GarcõÂa-Marquez and Rushdie is rather
untypical of his concerns, since it culminates in a prediction rather than a
normative judgment. This however makes no dierence to the point at issue.
Take the case (Aronovitch's) of candidates X and Y. Candidate X was admitted;
candidate Y has similar grades and recommendations to X; therefore Y should
also be admitted. Now, it is perfectly possible that when X was admitted...
To continue reading
Request your trial