The Failings of Analytic Philosophy
In my last post, I discussed what is good about analytic philosophy. After that, I was going to note some of the shortcomings of analytic phil, but the post was getting too long. So now, here's what's wrong with analytic phil, as currently practiced.
The main problem: too analytic.
Background: analytic statements are, roughly, statements that you can see to be true just from understanding the meanings of words. Like "all rhombuses have four sides" and "the present is before the future" [if the present and future both exist]. There are issues about how exactly to define "analytic" sentences, but let's not worry about that.
Analytic philosophers used to think that philosophy was or ought to be a body of analytic knowledge, and that analytic knowledge was essentially about the meanings of words, or the relationships between concepts, or something like that, and did not concern substantive, mind-independent facts. So they spent a lot of time talking about word meanings, how to analyze concepts, and boring stuff like that. They never did succeed in analyzing anything, though.
I don't know how many people still think the job of philosophy is to analyze language/concepts. I don't think it's very many. But the field retains leftover influences of that early doctrine. And the central problem with this is that most questions that are amenable to typical analytic-philosophy methods are just not very interesting.
More specifically, I see three things that we're doing too much of.
1. Fruitless Analysis
We spend too much time trying to analyze concepts. For instance, there are dozens of theories that start with "S knows that P if and only if . . .", followed by some set of conditions -- which get increasingly convoluted and hard to follow as time goes on. This has occupied a large portion of the literature in epistemology over the last 50 years.
There are two problems with this kind of philosophizing. One is that the analyses are always false.
That's controversial; some philosophers think that they themselves have correctly analyzed one or more philosophically interesting concepts. But most will agree that there are almost no successes, and that no philosophical analysis has attained general acceptance. For every attempted analysis, there are many philosophers who would say that that analysis is wrong and has been refuted.
For discussion of why no one ever successfully analyzed anything, see my “The Failure of Analysis and the Nature of Concepts” in The Palgrave Handbook of Philosophical Methods (2015).
This probably is not just about to change. We've had a lot of highly intelligent, highly-educated, dedicated people working on the analysis of various philosophically interesting terms, for decades now. If we don't have a single clear success by now, I don't think we need to keep doing this for another 50 years. There ought to be a time when you move on.
Another reason this is fruitless is that the analyses we devise would not be particularly useful, even if one of them were widely accepted. The analyses that epistemologists now debate are so complicated and confusing that you would never try to actually explain the concept of knowledge to anyone by using them. So what is the point?
Perhaps the value of these analyses is purely for the theoretical understanding of philosophers. But understanding of what -- how a specific word is used in a specific language? The exact contours of a conventionally defined category? Is that what we need to expend decades' worth of reflection by a host of highly sophisticated minds to figure out?
2. Semantic Debates
A fair amount of debate in analytic philosophy, even when it is not directly about the analysis of some word or concept, looks to me like essentially semantic debates. And as I've just suggested, I don't find semantic debates especially interesting.*
*Counterpoint: maybe our current conceptual scheme reflects the accumulated wisdom of our society, about what are the important and useful-to-discuss phenomena in the world. And maybe that's why figuring out the correct account of that conceptual scheme is helpful?
Example 1: Justification
The debates between internalism and externalism in epistemology look semantic. Roughly, the debate concerns whether justification for a belief is entirely determined by the subject's internal mental states (or states the subject has access to, or something like that).
Ex.: Reliabilists (the most common kind of externalists) sometimes say that a belief is "justified" as long as the subject formed it in a reliable way, whether or not the subject knows or has reason to believe that the belief-forming method is reliable. Internalists say this is not enough.
That looks to me semantic. As an internalist, I don't deny that reliability exists or is good. I just don't think that's what "justification" refers to.
Example 2: Composition
There are debates in metaphysics about the "existence" of various things. These include debates about when a composite object exists.
One view holds that there are no composite objects. That is, if you take some elementary particles, there is nothing you can do to them that will make them collectively comprise a larger object. So tables don't exist, people don't exist, etc. (Don't worry. There are still particles arranged table-wise; it's just that they don't add up to a single object.)
Other philosophers say that any objects compose a further object. If you have an object A, and an object B, then there is always a third object that has both A and B as parts. E.g., there's an object composed of my left eye and Alpha Centauri.
Still other philosophers say that some but not all ways of arranging simple things make them compose a further thing.
All that strikes me as a very semantic sort of debate to have. (There are arguments that it isn't semantic in the literature. But it really feels semantic.)
3. Defining Down the Issue
Okay, here is my biggest complaint. Philosophers will actually decide what questions to ask based on the consideration of to what questions they can apply purely a priori methods, especially conceptual analysis and deductive arguments. This often involves shifting attention away from questions that matter, to questions that are in the vicinity but that in fact do not matter at all.
Example 3: Authority
Say we're doing political philosophy. And suppose I have raised the issue (as I have been known to do) of why any government should be thought to have any moral authority over anyone. Why should those clowns in Washington get to tell us what to do, and why should any of us obey them?
As a contemporary academic political philosopher, let's suppose, you would like to say something about this. But it could be difficult, as I've formulated it. So you're first going to want to change the question to something more "analytic philosophical". How about this: how should an ideal "liberal" political order make decisions so that we would have reason to respect them, assuming that they were not independently unjust?
You then go on to discuss a theory of the ideal conditions for public deliberation among a group of rational agents who all regard each other as free and equal citizens. These conditions might include, for example, that everyone has a full opportunity to be heard, that all ideas receive fair consideration, and that the outcome of deliberation is determined only by the merits of the arguments.
I then point out (as is my wont*) that none of those conditions obtains in any society, so we still have no basis for political authority. You respond that that is an empirical matter outside your purview. Your interest was simply in describing an ideal.
*See The Problem of Political Authority, sec. 4.2.
That's lame. That's essentially replacing the question that matters with a question that doesn't matter, because the latter doesn't require getting off the armchair.
We can't answer whether any state actually has authority by reflecting on concepts. But that question is nevertheless what matters.
Example 4: God
When I worked as a TA in grad school, some of the classes covered the Problem of Evil. Here's a simple way of understanding the problem:
God, if he exists, is supposed to be all-knowing, all-powerful, and maximally good.
If God is unaware of all the evils in the world, then he is not all-knowing.
If He is aware of evil but unable to do anything about it, then he is not all-powerful.
If He is aware of evil and able to eliminate it, but unwilling to do so, then he is not maximally good.
But if God is aware of evil and both willing and able to eliminate it, then how can evil exist?
Here is a possible response: maybe God isn't all-powerful after all. (Or he could fail to be all-knowing, or maximally good, but the 'all-powerful' attribute is the one theists are most likely to give up.)
I saw this discussed in one of these introductory philosophy textbooks that the students were reading. The author (who was defending atheism based on the Problem of Evil) said something like "we are merely bored by such replies" -- I guess because it's not interesting to defend a thesis by redefining it. (Well of course you can defend the existence of "God" in some sense of that word!)
I found this kind of amazing. So if it turns out that there is an extremely powerful, intelligent, and good being who created the physical universe, but the being isn't capable of all logically possible actions, then that would be completely uninteresting to a philosopher, because ... it doesn't satisfy the definition of a certain word that we stipulated at the start? That sounds to me like caring more about word games than about reality.*
*In fairness, there are cases where defending a thesis by redefining it renders it uninteresting. E.g., if you defended "theism" by defining "God" to refer to nature, that would be uninteresting. That's partly because the new thesis would be uncontroversial, and also because it is too far from what we initially were interested in.
Why do we think the traditional philosophers' conception of God (the O3 world-creator) is the interesting thing to discuss? I suspect the answer is, at least in part, that this conception makes it easy to construct a priori, deductive arguments about "God", and spawns lots of fun conceptual/logical debates ("Is omnipotence logically coherent?" "Is it compatible with perfect goodness?" "Is maximal goodness compatible with free will?" Etc.)
In other words: the traditional definition creates jobs for armchair philosophers.
But that's not a rational, reality-oriented basis for selecting a definition. A rational basis for selecting a definition would be something like: "This is the definition that best fits with what we (seem to) have evidence for." Or: "This is the definition that enables us to formulate the questions that are important." (Granted, the existence of the O3 world-creator would be important. But a sub-omnipotent creator would also be important.)
Academic philosophers are so used to defining issues in this way (to create jobs for conceptual analysts, so to speak) that, if you try to discuss a normal point, philosophers will often misunderstand you, because they will try to get you to be making some 'conceptual point' that can be verified or refuted by analysis, deduction, and hypothetical examples. They default to hyper-strengthening or hyper-weakening issues. E.g., if someone wants to discuss whether A's are generally B, philosophers will try to talk about (i) whether it's conceptually possible for an A to be B, and/or (ii) whether it's logically necessary that all A's are B. This shows a greater interest in ideas, and how they relate to each other, than in the actual world.
Tl;dr: Analytic philosophers focus too much on playing with concepts, and not enough on thinking about the parts of reality that matter.