What's the current status of the "paradox of analysis"? And are there any strong and widely accepted resolutions?

Question

It would seem that figuring out a solution to the paradox of analysis would be of prime importance to philosophers, especially considering the fact that conceptual analysis seems central to philosophical practice.

The wiki page, however, seems to list under proposed resolutions simply biting the bullet and claiming analysis gives us no new knowledge (which seems slightly crazy, as writing bachelor = bachelor seems to give you something epistemically different from bachelor = unmarried man), or going the Quine route and rejecting conceptual analysis outright (which most philosophers don't seem to accept judging by a survey of the field).

I'm almost positive the wiki page neglects to mention other possibilities when it comes to resolving the paradox, and I'm hoping the good people here can point me in the right direction.

Answer 1

A good paper to read on this subject is an old classic: Gilbert Ryle's Systematically Misleading Expressions. (Proceedings of the Aristotelian Society, 32: 139-170 (1932). Also in his Collected Papers, vol 2.)

Ryle's view is that ordinary non-philosophical use of language frequently contains "improper" usages, by which he means usages that, while having a clear meaning to the non-philosophers who are using them, are systematically misleading when we try to subject them to a more rigorous understanding. This is not to say that such usages are defective, only that philosophers, when they subject them to examination, find that they need to restate them in a different way to avoid the apparent misleading implications. Ryle gives as examples:

1. Quasi-ontological statements. "Carnivorous cows do not exist" is not, despite appearances, a statement about carnivorous cows, because there are none. Rendering it instead as "nothing is both a cow and carnivorous" avoids this difficulty. "Jones hates the thought of going into hospital" does not entail the existence of objects called thoughts, at least one of which Jones hates, but can be analysed as something like "Jones becomes alarmed whenever he thinks about what might happen to him in hospital."
2. Quasi-platonic statements. "Unpunctuality is reprehensible" does not mean that there exists some universal - unpunctuality - that should be ashamed of itself, but rather that any person who is unpunctual deserves to be reproved by other people for being unpunctual.
3. Quasi-descriptions. "Poincaré is not the King of France" does not entail that Poincaré and the King of France are different people, whereas "Tommy is not the King of England" plausibly does.

These are straightforward examples, where sentences are being paraphrased to avoid being misleading. By extension, they are part of the whole process of abstracting, generalising and looking for a satisfactory logical form, which is what makes up philosophical analysis. The purpose being to find a form that is simple, powerful and expressive, while avoiding contradictions, confusions, paradoxes and unwarranted ontological commitments.

This approach resembles what Quine (Word and Object, section 33) calls paraphrasing, though interestingly Ryle claims that the analysans has the same meaning as the analysandum, while Quine says not.

Answer 2

The most recent thing I know about is Frank Jackson's work on the topic. Here's a blurb from the Stanford Encyclopedia article on analysis:

"One recent defence of conceptual analysis, with a qualified rejection of Quine’s critique of analyticity, has been offered by Frank Jackson in his book, From Metaphysics to Ethics (1998). On Jackson’s view, the role of conceptual analysis is to make explicit our ‘folk theory’ about a given matter, elucidating our concepts by considering how individuals classify possibilities (1998, 31-3). To the extent that it involves ‘making best sense’ of our responses (ibid., 36), it is closer to what Quine called ‘paraphrasing’ (1960, §§ 33, 53) than the simple recording of our ordinary intuitions (Jackson 1998, 45). Jackson argues for a ‘modest’ role for conceptual analysis, but in so far as he admits that a certain “massaging of folk intuitions” may be required (ibid., 47), it is not clear that his conception is as neutral as he suggests. "

So the idea here is that what we're giving is an analysis of the folk concept--which might be illuminating insofar as that folk concept might be confused, imprecise, or otherwise defective. Then the philosophically interesting bit is creating a more precise, more perspicuous concept which takes over the important explanatory roles from the folk one, hopefully preserving enough of common sense for the solution to make sense!

Answer 3

Analysis does not need to be informative if there is always another kind of knowledge for it to act upon.

For 'a posteriori' knowledge, we can obviously form referents by actually indicating real objects, so no analysis needs to be informative. Analysis can then be applied to information, and does not have to provide information itself.

But it is clear that some knowledge is 'a prior', that we have leverage on information from some basis that is integral to our minds. Even to grasp what is pointed to when someone points at a table and calls it a table, one needs a set of underlying organizing conventions that limit the range of the potential implicit pronoun created by pointing. And even below that, one needs to know that pointing is. This set of conventions allows analysis to get a foothold by the process of elimination. But if these underlying conventions do not limit the options to a finite set, no information really can be conveyed.

Many people accept Kant's notion that synthetic a priori knowledge is indeed possible. There seems to be formational content present for all of us that provides enough structure for us to learn meanings from tacit references and implicit indications.

Plato's notion of anamnesis suggests that, at base, all knowledge is simply recombinations of this kind of knowlege -- that we already know everything we can learn, but need to apply analysis. So in extreme models of the mind, one does not ever have to be able to generate new synthetic knowledge. One only has to trust that it is there.

Most thinkers will not go that far, but they will go part of the way. From a naturalist point of view, we are evolved beings and we are formed from information, in the form of genetic code. So it is unlikely that none of that information ends up in our minds at birth as a basis for bootstrapping learning.

So the paradox is not really paradoxical. Analysis may be thoroughly incapable of generating information on its own. But it does not need to. Synthetic knowledge can be the ultimate source of all information, with analysis providing a different aspect of understanding than 'information'.