In my previous entry in this series, "Metaphilosophical concerns", I traced out some similarities between philosophy as it is practiced today and pre-scientific psychology. Several decades ago, before psychology increasingly adopted scientific methodology,
- Theories in psychology were based primarily on how things "seemed" to different investigators.
- Without any clear protections against forms of cognitive or emotional bias, motivated reasoning, etc.
This resulted in opposing theoretical camps--Freudianism, Jungianism, Behaviorism, Humanism, Instinct Theory--whose practitioners tended to think that their favored framework was better supported by arguments than their rivals. Unfortunately, this tended to result in very problematic forms of argument, such as post-hoc reasoning where each framework would marshall their own theoretical resources to "explain" the same phenomena. If, for instance, a patient resisted treatment, Freudians would say, "Of course - that's just what our theory of reaction formation predicts!"; Humanists would say, "Of course - that's precisely what one would expect from a humanistic drive to self-actualize!; and so on. The end result was a bunch of theoretical stalemates, in which proponents of each perspective maintained their framework was the most coherent with the evidence. The premises that one group of people found "plausible" did not seem all that plausible to people on the other side of the debate.
Of course, this all seems a bit quaint now. Turns out, contra Freud, that sometimes a cigar really is just a cigar! Psychology moved beyond its more speculative origins, and has increasingly adopted a far more scientific approach--one requiring theories and arguments not to based on how things "seem", but instead on hard data: observational and experimental, data. This isn't to say that psychology is perfect today. As the current replication crisis in a number of sciences (psychology included) indicates, psychologists are still struggling with methodological issues. Nevertheless, we shouldn't overstate those issues. There are many, many results in psychology that have been replicated successfully many times over, and psychology has in many respects developed a large body of established knowledge--much like biology, physics, and other science.
As I explained in my previous post, the situation in philosophy seems disturbingly similar to the history of psychology. Philosophy throughout history--but especially in 20th Century analytic philosophy (as I will explain momentarily)--has in large part been based on how things "seem" to different inquirers. Unfortunately, things often seem very different to different inquirers--resulting in a wide variety of competing camps with different theoretical frameworks, different views about which premises are "plausible", and so on. Compatiblists, for instance, may find compatibilist premises "plausible"--but free will skeptics and libertarians not so much. Similarly, Derek Parfit and T.M. Scanlon may think that reasons are primitive, but constitutivists certainly don't think so. And so on. We could go on. The general point is this: much like pre-scientific psychology, all we have to go on in philosophy (as it is usually practiced) are different people's judgments about what "seems true" as a premise--this despite the fact that, all too often, a premise that "seems true" to one person "seems false" to another--which in turn raises the spectre of whether "progress" in philosophy is actually a matter of getting closer to the truth or more of a sociological phenomenon of what some privileged group of people think the truth might be (in much the same sense that the dominance of Freudian psychology was sociological rather than a matter of truth).
Notice that I am not claiming to answer these questions one way or the other. For all we know, philosophy as it is currently practiced may be progressing toward truth. But, the problem is, for all we know it isn't. The Freudians certainly thought they were getting at truth via their arguments--but, as we now know, much of Freudianism is off-base. Without clear, reliable methods to protect against systemic bias, motivated reasoning, and so on, the truth-aptness of philosophy--much like the truth-aptness of speculative psychology--can be reasonably in doubt.
This, finally, brings us to the big question: what might philosophy do to best address these methodological concerns? Currently, philosophers appear to me to be largely addressing methodological concerns like these in a piecemeal fashion--all too often utilizing, unfortunately, the very methods in question to think about these methodological issues! For instance, there is currently an ongoing debate over whether intuitions play an important role in philosophy, and if so, whether intuitions are reliable. Personally, I think the debate over whether philosophy uses intuitions is a red-herring. Whatever we want to call them (intuitions, evidence, whatever), philosophers appeal to premises--and the problem is that philosophy does not have any systemic protections against bias or motivated reasoning corrupting the premises we appeal to. Yes, p may "seem true" to you as a premise, but (A) what seems true and what is true can (as we all know from the history of science) be two different things, and (B) there are very rarely premises in philosophy that seem true to many people (hence, persistent stalemates across a wide variety of philosophical issues).
So, one thing we could continue doing is approaching these issues in a piecemeal way--for instance, by debating "phenomenal conservativism" (i.e. whether seemings are evidence or not). But alas, this only seems to me (wink, wink) to push the metaphilosophical question back a step, as those arguments come down to disagreements over how things seem. Rinse and repeat. Is there a better way?
In Chapter 1 of my book, I argue that there is. I argue that the general lesson of history--from the emergence of modern physics from pre-Socratic and Aristotelian speculation, to the emergence of modern chemistry from alchemy, to modern psychology--is that a genuinely truth-apt discipline cannot reply on how things "seem." We must instead insist upon much more strict epistemic standards. Which ones? Following the lead of these sciences, I defend at length the following principles of argumentation and theory selection (ch. 1):
- Firm Foundations: theories based on common human observation—or observations that are taken to be obvious, incontrovertible fact by all or almost all observers—should be preferred over theories based on controversial observations that may seem true to some investigators but not to others.
- Internal Coherence: all things being equal, and subject to Firm Foundations, theories with fewer or no internal contradictions should be preferred over theories with more.
- External Coherence: all things being equal, and subject to Firm Foundations, theories cohering with more known facts and observations should be preferred over theories cohering with fewer.
- Explanatory Power: all things being equal, and subject to Firm Foundations, theories explaining more facts and observations should be preferred over theories explaining fewer.
- Unity: all things being equal, and subject to Firm Foundations, theories unifying disparate phenomena, showing how they have a common explanation, should be preferred over theories providing more fragmentary explanations.
- Parsimony: all things being equal, and subject to Firm Foundations, theories that successfully explain phenomena with fewer facts or entities should be preferred over theories explaining the same phenomena with more.
- Fruitfulness: all things being equal, and subject to Firm Foundations, theories solving more existing theoretical or practical problems should be preferred over theories solving fewer.
The most critical standard here, I maintain, is the first one: the principle of Firm (Observational) Foundations. Although I have only roughly characterized it here, broadly speaking it is in my view something like the founding, essential principle of modern science--the principle sound arguments and theories need to not be based on how things "seem" merely to some, but instead on virtually universal observations that can be replicated and recognized as true by human observers in general (in experiments, clinical observation, etc.). This, I contend, is the principle that ensures that we are dealing with truth, as opposed to merely things that "seem true", some of us think true, or want to be true. Further--and importantly--I argue that all other principles of argument and theory-selection should be subservient to to the principle of Firm Foundations. The lesson we should learn from the history of inquiry--from Freudianism, Humanism, etc.--is that "coherence" isn't epistemically valuable unless the very observational foundations of the theory are truth-apt.
What does this mean for philosophy? Can the principle of Firm Foundations be applied to philosophical questions? Some might think not--that because philosophers inevitably disagree over premises, there are no virtually universal observations that philosophy can be based upon. I think this may be true of some inherently speculative parts of philosophy (e.g. analytic metaphysics). But, in general, I don't think it is true. I think there are universal observations to base philosophy on--but that it requires a very different approach to philosophy!
A while back, I distinguished between analytic philosophy, continental philosophy, and natural philosophy. Analytic and contintental philosophy (as it were) have generally been based upon a priori "armchair methodologies": using philosophers' judgments about thought-experiments (such as linguistic reference, mental content, etc.). My proposal (viz. the seven above principles of theory selection) is that in order to adequately protect against systemic bias, motivated reasoning, etc., philosophy should strive as much as possible to be based on naturalistic observation and experiment. What does this mean? It means, I believe, that philosophy moving forward should prioritize what might call both "narrow" and "broad" experimental philosophy. Instead of basing theories of linguistic reference on philosophers' judgments about cases, or claiming "philosophical expertise", the principles of theory-selection above suggest that experimental data should be used to discover how people use words to refer, and whether philosophers have any expertise in these matters. By a similar token, instead of basing practical or moral philosophy on controversial claims about "reasons" or "moral facts", the above principles of theory-selection suggest what G.E.M. Anscombe famously wrote in her article, "Modern Moral Philosophy", nearly 60 years ago: namely, that "it is not profitable for us at present to do moral philosophy...until we have an adequate philosophy of psychology, in which we are conspicuously lacking." On the seven principles of argument and theory-selection above, we should construct theories of moral and practical reasoning not on the basis of premises that "seem true" to philosophers, but instead on well-informed empirical understanding of human practical and moral cognition--that is, on how human beings in general actually cognize "reasons", "oughts", and so on (things for which, I argue in my book, there is actually quite a lot of empirical knowledge!).
I realize that not everyone may agree with my argument or like these implications. However, I hope readers find them interesting and worth grappling with. As I mentioned in my previous post, despite making these arguments, I am not attempting to "slag" philosophy. After all, I too have consistently used the very kinds of armchair methodologies I am suggesting are problematic--and I still think it was work worth doing! Like I said in my previous entry, I think psychology was fantastically interesting and well worth doing before it became a modern science. By a similar token, I think a great deal of philosophy today is fascinating and well-worth doing. Rather than "slag philosophy", all I am trying to do is make a case in favor of a broad shift of methodological emphases and practices away from arguments based on how things "seem" toward arguments based on how things are (viz. experimental philosophy and empirical science), basing philosophy on naturalistic science to whatever extent it is possible to do so (and it might not always be possible!).
Finally, far from making philosophy "too naturalistic", I believe that such a methodological shift promises to make philosophy more exciting, not less, using naturalistic facts to help push philosophical debates forward and upend unjustified prejudices in the way that naturalistic results have a history of doing. For indeed, as I mentioned previously, one problem with basing arguments on how things "seem"--or "reflective equilibrium" (without insisting upon Firm Foundations)--is that these methods seem inherently conservative (in a troubling way). Einstein's theory and space and time contradicted what people thought they knew. So did Darwin's theory of evolution. The great thing about naturalistic observation is that--when theories are based upon them rigorously--they often end contradicting and overthrowing beliefs that once "seemed true" to just about everybody the armchair. That, to me, sounds exciting indeed--and, among other things, I believe that the science of practical reason and moral cognition promises to do just this in moral philosophy. But that will have to wait for another day...