A few weeks ago, some of my philosopher friends were sharing this Chronicle of Higher Education article, "Is Philosophy Obsolete?", the tag-line of which reads, "No. It helps make us coherent to ourselves, and that's a never-ending project."
I don't think "seeking coherence" can be an adequate way of defending, or doing, philosophy.
First, there appear to be conclusive reasons to think that coherence per se is not truth-conducive (see sections 7&8 here). A set of beliefs can be as coherent as you like--but, for all that, they can be totally out of touch with the truth. This is a longstanding objection to coherentist theories of epistemic justification, and it seems to be a definitive one.
Second, while some might be willing to bite this bullet--suggesting that philosophy needn't be about getting at truth (I've heard a few people suggest this!)--I don't think that's a bullet we should want to bite. Making false views more coherent can be a very dangerous endeavor. Among other things, it can make false and harmful theories or belief-sets seem more plausible to those who already endorse them, as well as to others. To take just a few examples, consider Sir Robert Filmer's defense of the divine right of kings in his 1680 book, Patriarcha. Or consider young-earth Creationism or the views of the Flat-Earth Society. Or, of course, consider racism and white-supremacy. These doctrines can be rendered increasingly coherent, and indeed, their proponents have spun all kinds of coherent defenses of them--defenses which their proponents think constitute evidence for the view in question. But this, clearly, is a very dangerous thing to do. We should want to avoid approaches to inquiry that can lend an "air of legitimacy" to views that are false.
Finally, notice that coherence in one obvious sense lends itself to a kind of status quo bias. If our task as philosophers is to simply make our beliefs more coherent, we must begin with whatever beliefs we have--which is plausibly what led Filmer and those of his ilk to defend the divine right of kings. That doctrine fit with many status quo beliefs at the time, so of course it seemed defensible on coherentist grounds. Similarly, people resisted Copernicus' heliocentric theory of the solar system because it contradicted "commonsense" at the time. People resisted (and some still resist) the theory of evolution on such grounds. And yes, many physicists and philosophers initially greeted Einstein's theory of relativity not with open arms, but rather condemnation as obviously inconsistent with the "a priori fact" that space and time (obviously!) must be absolute.
For these reasons, I don't think philosophy or science should be in the business of rendering "commonsense" coherent. They should be in the business of discovering the truth--for, as we see in history, the truth often defies commonsense (or, at least, what people take to be commonsense at a given time). But how? What does it take for philosophy to be truth-conducive? And, are we currently using truth-conducive methods? Like a seemingly growing number of philosophers, I have a variety of concerns, some of which I would like to share here and explore in this post and several to come.
I would like to begin with a story. When I was an undergraduate philosophy and psychology double-major at Tufts University in the mid 1990's (yes, I'm getting old!), I had an experience in my psychology major that has always stuck with me. In brief, the experience goes like this. During my undergraduate years, modern empirical psychology and neuroscience were still relatively new disciplines. Just several decades before, in the early- to mid-20th Century--and indeed, up through the 1970's and 80's--psychology was in many respects a speculative endeavor, dominated by a variety of competing, wide-ranging theories (i.e. entire "schools of psychology") based on how things seemed to particular investigators. There was Freudian Psychoanalysis, Humanism, Instinct Theory, Behaviorism, Jungian Analytical Psychology, Piajet's theory of childhood psychological development, and many other theories to boot.
In many respects, these various theories were absolutely fascinating. Freud attempted to reduce human action to three interacting parts of the mind (the Id, Ego, and Superego); Humanists to drives to "self-actualize"; Instinct Theories to a variety of innate drives and inclinations; Jungianism to "individuation" (or integrating opposing psychological forces); and so on. At the same time, as a student (and reasonably skeptical person), something seemed to me quite amiss. The methods each school used to defend their theories seemed to me epistemologically dubious--and not just in one respect, but many.
First, each school appeared to defend their theories largely on the basis of how things "seemed" to them. Freud, for instance, developed his theory largely on clinically observing and talking with subjects, interpreting their behavior according to how things appeared to him. And other schools proceeded in a broadly similar way: Piajet developed his theory of childhood psychological development by observing children and interpreting their development according to how things seemed to him. And so on.
Second, as we see here, things apparently seemed very different to different people in different psychological schools. Indeed, competing schools would often interpret the same behavioral phenomena according to their favored framework. For instance, a patient who resists treatment might be interpreted by a Freudian as engaging in reaction formation; by a Humanist as attempting to self-actualize; by a Behaviorist as responding to treatment as an aversive stimulus; etc.
Third, there appeared to be few, if any, systemic controls against bias or motivated reasoning (i.e. seeing "the facts" how one one wants to see them). For instance, because Freudians could fit just about any observation into their theoretical framework (explaining whatever behavior was in question in terms of the Id, Ego, and Superego), anytime one attempted to use observational evidence to argue against a Freudian, one could hardly expect to get anywhere. Freudians would typically just fit the observations into their theoretical framework, engaging in post-hoc reasoning, "verifying" Freudian theory. But of course this was methodologically problematic. Given that the same data could be explained by rival theories, claiming it "supported" the Freudian model was little more than motivated reasoning: seeing the data as "supporting" Freudianism when, in reality, the data did not unambiguously support the theory at all (indeed, a major problem here was that dominant schools of thought didn't make any determinate predictions).
Finally, though, at precisely the moment I was having these worries as an undergraduate (taking undergrad courses on the theories discussed above), psychology was going through vast changes that began to systematically address these problems. Like physics and biology before it, psychology was becoming a science. Whereas psychology had previously consisted of different "schools of thought" based on theorists' intuitions, over the course of the twentieth century the discipline progressively rejected its previous methods in favor of the scientific method, increasingly insisting upon controlled experiments, predictions, statistical analysis. And, much like other emerging sciences before it, psychology's use of the scientific method ended up producing results that challenged and disconfirmed many of the very things that seemed true to previous investigators (see e.g. here and here).
Now turn to philosophy. While I would be the first to say that a lot of philosophy is absolutely fascinating and worth thinking about--and I don't want readers to think I am attempting to "slag" philosophy like Dennett recently did (I'm really not, more on this below!)--I think there are some disturbing similarities that might lead us to wonder whether philosophy could use some serious methodological improvements.
First, notice that many of the metaphilosophical concerns philosophers are pressing today broadly cohere with the kinds of concerns raised about pre-scientific psychology above. A good number of philosophers--myself included--appear to be increasingly concerned about metaphilosophy, particularly the reliability of basic, common philosophical methods including:
- The "method of cases."
- Conceptual analysis
- The use and reliability of intuitions
- Philosophical "stalemates" that seem to result from each "side" having roughly equal "moves" available.
- Potential sociological confounds
- Problems calibration
- Inadequate protections against systemic bias and error.
Second, consider how philosophy, like psychology before it, has tended to splinter into a variety entrenched "schools of thought" (or opposing "camps"). In moral philosophy, for instance, there are realists, anti-realists, quasi-realists, Kantians, consequentialists, contractualists, intuitionists, constitutivists, reasons-fundamentalists, etc. In the free will debate, there are free will skeptics, compatibilists, semi-compatibilists, libertarians, etc. In the philosophy of mind, there are physicalists, dualists, panpsychists, mysterians, internalists, externalists, etc. And so on.
Third, consider how each school or camp in philosophy often appears to defend their views. If we look at a variety of philosophical debates--compatibilism versus incompatibilism, physicalism versus antiphysicalism, moral realism versus anti-realism--the various sides typically seem to begin from premises that seem attractive to people on their side of the debate, but not the other side. I know this from experience. When I read the compatibilist literature, or literature defending reasons fundamentalism, and some other debates, I just don't find many of the premises being appealed to attractive. I recognize full well that some people find them plausible. I just don't--and neither do others in my "camp." Which, I think, is worrisome. In practice, one rarely sees, say, compatibilist arguments convincing incompatibilists, or physicalist arguments convincing anti-physicalists, etc. Instead, each side seems to think their premises or arguments are more plausible, leading them to defend their view against the other's side's counterarguments and provide what they take to be a plausible, coherent response to the "same data." Of course, the other side tends to disagree--but that's the problem. There doesn't appear to be any independent measure of which side's premises are actually "more plausible" than the other side's. All we have to go on is the judgments of those party to the relevant debate. Which, offhand, is plausibly why so many philosophical debates seem to turn into "stalemates."
Fourth, notice that aside from experimental philosophy, common philosophical methods used today--much like the methods of pre-scientific psychology before it--do not appear to involve any independent method for ensuring reliability. Although we commonly recognize that what seems true and what we want to be true may be very different from is true, current philosophical methods mostly leave it up to philosophers, as individuals and groups, to settle these issues using their own judgment. In philosophy, all we have--by and large--are the arguments themselves and each individual's judgment of whether an argument's premises are "plausible." But, wait a minute. As we know all too well from science (and everyday life), people's beliefs on a given issue can be systematically corrupted by all kinds of biases (confirmation bias, availability bias,), motivated reasoning, and sociological confounds. It seems like, if we are concerned with truth, we should want some method for isolating and protecting against these kinds of confounds--but, recent advances in experimental aside, this seems to be precisely we don't currently have in philosophy.
Finally, consider persistent concerns about philosophical progress and skepticism about philosophy. Whereas physicists, biologists, psychologists, etc., disagree on many things, there are also many widely accepted results. However, in philosophy, persistent disagreement on basic issues is widespread. Begin, for instance, with the philpapers survey. While there is significant agreement on some issues (71% of respondents accept or learn towards the existence a priori knowledge, 72.8% atheism, 81.6% non-skeptical realism, 59% are compatibilists), there is persistent wide disagreement on most issues:
Normative ethics: deontology, consequentialism, or virtue ethics?
Other 301 / 931 (32.3%) Accept or lean toward: deontology 241 / 931 (25.9%) Accept or lean toward: consequentialism 220 / 931 (23.6%) Accept or lean toward: virtue ethics 169 / 931 (18.2%)
Epistemic justification: internalism or externalism?
Accept or lean toward: externalism 398 / 931 (42.7%) Other 287 / 931 (30.8%) Accept or lean toward: internalism 246 / 931 (26.4%)
Knowledge claims: contextualism, relativism, or invariantism?
Accept or lean toward: contextualism 373 / 931 (40.1%) Accept or lean toward: invariantism 290 / 931 (31.1%) Other 241 / 931 (25.9%)
Knowledge: empiricism or rationalism?
Other 346 / 931 (37.2%) Accept or lean toward: empiricism 326 / 931 (35.0%) Accept or lean toward: rationalism 259 / 931 (27.8%)
I don't know about you, but I for one think it would sure be nice if there were some new method available for more definitively pushing these debates in one direction or another--especially a method known to be truth-conducive! On that note, consider some of David Bourquet and David Chalmers' findings in, "What do philosophers believe?". Disturbingly, B&C's study found that what professional philosophers believe--the theories, positions, arguments they find compelling--often vary significantly by gender (pp. 17-18), age (p. 19), geographical location (p. 20), and so on. Philosophers in Australasia are more likely to be consequentlists, Americans more likely to be deontologists, British more likely to be disjunctivists about perceptual experience, Europeans more likely to be Fregeans about proper names, and so on.
Now, of course, maybe this is the way philosophy has to be! Philosophy and science, we might say, are very different things. Science deals with hard data, and philosophy with the deepest questions and issues that can't be quantified--things that must, of necessity, remain more speculative. And indeed, unlike Dennett, I don't think philosophical speculation of this sort, untethered from other disciplines or everyday life, is pointless or to be derided. On the contrary, most of my work has utilized the very kinds of common methods I have raised concerns about here--and I don't think those speculations are pointless. Part of the beauty of philosophy is its depth and openendedness. My reason for raising these concerns, however, is that I think we need to be honest about just how tenuous our methods appear to be. I think this is important for three reasons. First, because it might help us develop greater humility about our methods, theories, and arguments. Second, because it might help us better understand and appreciate concerns that outsiders (physicists, the general public, etc.) have about philosophy. But third, and perhaps most of all, because I think honesty about problems with our methods might provide impetus for preferring new methods. In Chapter 1 of my new book, I present and defend what I believe to be a helpful new method in this regard--and in my next post, I hope to discuss that method a bit, both to see what people think of it, and to hopefully generate discussion of other possible methods.
In any case, what do you all think? Do you think all is well with philosophy's methods? Why? Or, do you think philosophy needs methodological changes? If so, what should they be?