One of the most difficult things one faces in the history of philosophy (and in many other fields of inquiry, I'd imagine) is generalizing from particular claims and arguments without destroying essential detail. On the one hand, to make sense of things we need to classify them, and classification involves generalization and abstraction from details. On the other hand, if you describe things vaguely enough, with no regard for possible distinctions, you can make anything seem consistent with anything, and anything seem inconsistent with anything, regardless of the actual state of the argument. For instance, you could take two positions that in fact are mutually exclusive and in fact are deliberately opposed to each other and treat them vaguely enough that you end up conflating contrary views. Likewise, you can take two very, very vague terms, like 'science' and 'religion', both of which cover any number of things, and talk about how 'science and religion are inconsistent'. Nobody in their right mind would argue that observing Passover makes it impossible to observe the stars; you certainly wouldn't mean that the Sikh doctrine of the Panth makes it impossible to study butterflies; you don't mean that the Sermon on the Mount contradicts the equations used in quantum physics; you aren't saying that the Catholic Church has commanded its members never to speculate about the Big Bang or the inflationary hypothesis. At least, if you were saying these things, you'd be a lunatic. So you must actually mean something very specific. And if you mean something specific, you have to ask yourself why you are talking about this very specific thing so vaguely, at a level where distinctions are lost and there is danger of unlike things being conflated and like things being misclassified. And if you meet someone who talks this vaguely about anything specific, you have to worry that it's a sign that they're playing a confidence game; only confidence men and snake-oil salesman talk that vaguely about specific things, because it's a way of talking about things that allows you to ignore relevant details and important distinctions. Another thing it allows you to do is to make inferences based purely on verbal similarity and not on anything of substance. Generalizations have to be made; but there is a form of generalization that involves equivocation. And that is to be avoided.
So the natural question is what steps should be taken to handle this potential problem. Are there domain-general guidelines that could be used? Or does it really vary from topic to topic? (Of course, both could be true.) How do you try to make sure that your abstractions and generalizations are abstract and general enough to work with but not so abstract and general that essential and relevant distinctions are lost?
