I had taken physicist Lawrence Krauss to task for his intemperate (and badly informed) remarks about philosophy just a few months ago. Krauss — apparently prompted by Dan Dennett — issued a non-apology of sorts in Scientific American shortly thereafter, thereby showing he had learned precious little from what should have been an embarrassing incident. Now he has done it again, this time in a debate with philosopher Julian Baggini published by The Guardian. I think Baggini went a bit too easy on Krauss, but the reason I’d like to comment on the specifics of the debate (other than the self-indulgent one of poking a bit more fun at Krauss, who is a big guy and can take it) is that I think we can learn something of general import from these ongoing skirmishes between philosophers and scientists.
The title of the piece featuring the debate is “Philosophy vs science: which can answer the big questions of life?” which seems prima facie a bit ill conceived, given that one could reasonably argue that science and philosophy address different kinds of “big questions in life.” Be that as it may, Baggini begins with a proper acknowledgment of the successes of science and states that he doesn’t feel proprietary about the domain of philosophy, but, he says “there are some issues of human existence that just aren’t scientific. I cannot see how mere facts could ever settle the issue of what is morally right or wrong, for example.” (As readers of RS know, I quite agree, pace Sam Harris.)
Krauss, predictably, will have none of it, categorically stating that as far as he is concerned, it’s all “merely distinguishing between questions that are answerable and those that aren’t. To first approximation, all the answerable ones end up moving into the domain of empirical knowledge, aka science.” Krauss is clearly wrong here, and he ought to know it. Despite some scientists’ tendency to consider logic and math as “sciences” (they aren’t, by any sensible definition of science — indeed, logic is clearly a branch of philosophy), the answer to neither logical nor mathematical questions hinges on empirical knowledge, so we have at least two clear (and ample) examples of Baggini’s basic contention that not all questions admit of scientific answers.
But Baggini brought ethics, not math, into the discussion, so Krauss aims his response straight at moral decision making: “Getting to your question of morality, for example, science provides the basis for moral decisions, which are sensible only if they are based on reason, which is itself based on empirical evidence.” Seriously? I would have thought the other way around: “empirical evidence” isn’t something that stands on its own. Facts are facts only within a given theoretical framework (as Darwin famously observed), and reason and evidence better play with each other in a mutually reinforcing fashion. It just isn’t as simple as “just the facts, madam.”
Krauss continues: “Without some knowledge of the consequences of actions, which must be based on empirical evidence, then I think ‘reason’ alone is impotent.” Well, yes, but I don’t know any philosopher who would claim otherwise. Let’s say you are a consequentialist when it comes to moral decision making. You do rely on empirical evidence (broadly construed, more on this later) to make specific decisions, because after all you care about the likely consequences of your actions, so you need some way to estimate what such consequences might be. But the prior decision to adopt a consequentialist stance in ethics is not the result of “facts,” it derives from your philosophical reflections on what ethical judgments are and how best to carry them out. You need to read Mill, not Einstein.
But of course Krauss has drunk the Harris cool-aid in full: “Ultimately, I think our understanding of neurobiology and evolutionary biology and psychology will reduce our understanding of morality to some well-defined biological constructs.” He, like Harris before, confuses an understanding of the mechanisms of X with the epistemic status (true, not true, undecidable, etc.) of X. To see the point simply, imagine if a neuroscientist told you that he wanted to scan your brain while you were trying to demonstrate the Pythagorean theorem. Said neuroscientist would be able to tell you all sorts of fascinating biological facts about which parts of your brain were involved in the task, how much of each neuropeptide you were using while carrying it out, etc. What he would have absolutely nothing to say about is whether you got the proof right. You’d need a mathematician for that.
Baggini makes a similar point very nicely and reasonably: “No factual discovery could ever settle a question of right or wrong. But that does not mean that moral questions are empty questions or pseudo-questions. We can think better about them and can even have more informed debates by learning new facts. What we conclude about animal ethics, for example, has changed as we have learned more about non-human cognition.” Yup, exactly: science can (and should) inform moral thinking, but it doesn’t determine it.
He continues to give a basic lesson in philosophical reasoning to Krauss: “What is disparagingly called scientism insists that, if a question isn’t amenable to scientific solution, it is not a serious question at all.” And that is one of the best definitions of scientism available. That there are questions that are not answerable scientifically and yet are perfectly meaningful should be clear to anyone who considers ethics, logic or mathematics, among other things.
Krauss comes close to acknowledging the point, and then manages to screw up again, all in one sentence: “I do think philosophical discussions can inform decision-making in many important ways, by allowing reflections on facts, but that ultimately the only source of facts is via empirical exploration.” The part after the “but” is either not true or tautological. Depending on what you mean by “facts” it may not be true: do mathematical theorems count as facts? If so, they do not require empirical exploration to be determined true or false. If we stick to straightforward empirical facts, of course their only source is, well, empirical! But we are talking about knowledge and understanding, for which facts are a necessary but not sufficient condition.
Krauss then proceeds to gleefully engage in more logical fallacies, chiefly the naturalistic one (deriving an ought from an is), a favorite of Harris too: “Take homosexuality, for example. Iron age scriptures might argue that homosexuality is ‘wrong,’ but scientific discoveries about the frequency of homosexual behaviour in a variety of species tell us that it is completely natural in a rather fixed fraction of populations and that it has no apparent negative evolutionary impacts. This surely tells us that it is biologically based, not harmful and not innately ‘wrong.’” First off, scriptures don’t “argue,” they simply lay down the (allegedly divine) law, without argument (the chief difference between religion and philosophy). Second, and most importantly, so what if it had turned out that homosexuality wasn’t “natural”? Would it follow that it is wrong? Vaccinating your kids is not natural, but surely it is the right thing to do. Conversely, the tendency for human beings to settle disputes by violence is natural, but most of us would agree that it is the wrong way to go about it. There simply is no logical connection between something being natural/unnatural and the same thing being right/wrong, just as Hume had warned (did Krauss read Hume? I doubt it). Baggini, incidentally, points out the same to Krauss, though in a more restrained fashion. Kudos for his patience.
Krauss then changes the subject (his words) and wants to convince Baggini that there is no difference between “how” and “why” questions. An interesting approach, because — in his mind — if we can show that there are no “why” questions then science becomes the ultimate arbiter of any meaningful issue. (Again, this is prima facie contradicted by math and logic, which are not concerned with “why” and yet do not require empirical evidence to settle their questions, but let’s go along with it for a moment.)
His example is Kepler’s quest for “why” there are six planets in the solar system (according to the best astronomical knowledge of the time). Kepler thought he got the answer when he realized that there are only five Platonic solids (the faces of which are composed by regular polygons). Each Platonic solid can be circumscribed by a sphere, and if such spheres separate the orbits of the planets, then there can be only six planets. Kepler, as we know, was wrong on the number of planets and on his proposed explanation, but astronomers made progress when they turned his “why” into a “how” and figured out that the answer was to be found into Newton’s laws of mechanics.
The history of science aspect of Krauss' story is likely a bit simplistic, but of course nobody would object to his account of “why” vs “how” in science. However, Baggini immediately points out that why questions come to the forefront whenever we consider agency: “if we want to know why someone made a sacrifice for a person close to them, a purely neurological answer would not be a complete one. The full truth would require saying that there was a ‘why’ at work.” Indeed. Now, I don’t want to get into yet another debate about free will, because people simply tend to talk at cross-purposes in that context. At the very least we can agree that — epistemically speaking — we need to deploy high level cognitive concepts like motivations, reasons, etc. to make sense of human behavior. Which means that “why” questions are very much alive and well, thank you very much.
At any rate, even if one were to concede to Krauss that in some broad sense everything comes down to “how” things work, all the above considerations about the insufficiency (and sometimes irrelevance) of empirical facts to settle certain questions still hold, and philosophy gets to live another day.
But Krauss isn’t going to be happy with this, he really wants to push his scientism to its logical endpoint, thus insisting: “Certainly, we already understand many aspects of sacrifice in terms of evolutionary biology. Sacrifice is, in many cases, good for survival of a group or kin. It makes evolutionary sense for some people, in this case to act altruistically, if propagation of genes is driving action in a basic sense.” This pop evolutionary account confuses the origin of a behavior (say, the evolution of an altruistic instinct to protect one’s progeny) with the kind of moral reasoning that is of interest to modern human beings in complex societies. To return to the example of math from above, it would be like saying that evolution explains our ability to solve, say, Fermat’s last theorem because it was adaptive to be able to count a few items in the Pleistocene’s savannah. The latter is likely true, but the former doesn’t follow at all.
Now, what are we to make of all of this? I am not particularly interested in simply showing how philosophically naive Krauss is, as much fun as that may be. Rather, I want to ask the broader question of what underlies Krauss’ (and others’) take on science and philosophy. It’s a broader conversation that we have been having for a while on this blog and will no doubt continue, but I’d like to attract the reader’s attention to a fundamental mistake underlying pretty much everything Krauss has said throughout his debate with Baggini: he has (implicitly) built the following argument:
Premise 1: Empirical evidence is the province of science (and only science).
Premise 2: All meaningful / answerable questions are by nature empirical.
Premise 3: Philosophy does not deal with empirical questions.
Conclusion: Therefore, science is the only activity that provides meaningful / answerable questions.
Corollary: Philosophy is useless.
Now, P2 is awfully close to the philosophical (!!) position known as logical positivism (in the US, logical empiricism), which has been demolished by the likes of Quine, Putnam, Kuhn and others. I have argued above that there are plenty of questions whose nature is not empirical, or not wholly empirical, and yet are meaningful. P3 can be interpreted in more than one way: yes, philosophy isn’t in the business of answering empirical questions (just like math and logic), but it is a caricature of the field to claim that empirical facts are irrelevant to philosophical considerations, and no sane philosopher would defend such claim.
This would already be enough to show that both the Conclusion and the Corollary are false, and we could go home with the satisfaction of a job well done. But I want to add something about P1. This appears to be a widespread assumption among scientists, recall for instance Jerry Coyne’s argument that plumbing is a science because it deals with empirical evidence, which plumbers use to evaluate alternative “hypotheses” concerning the causal mechanism of your toilette’s clog. But there is a fallacy of equivocation at work here, as the word “science” should be used in one of two possible meanings, but not both: either Krauss, Coyne et al. mean that (a) science is any human activity that uses facts to reach conclusions; or they mean that (b) science is a particular type of social activity, historically developed, and characterized by things like peer review, granting agencies, complex instrumentation, sophisticated analytical tools etc.
(b) is what most people — including most scientists — mean when they use the word “science,” and by that standard plumbing is not a science. More importantly, philosophy then can reasonably help itself to facts and still maintain a degree of independence (in subject matter and methods) from science.
If we go with (a), however, some nasty consequences ensue. Besides the fact that we would have to grant the title of scientist to plumbers, it would follow that I am doing “science” every time I pick the subway route that brings me somewhere in Manhattan. After all, I am evaluating hypotheses (the #6 train will let me get to 86th Street at the corner with Lexington) based on empirical evidence (the subway map, the directly observable position of the stations with respect to the Manhattan street grid, and so on). You can see, I hope, that this exercise quickly becomes silly and the word “science” loses meaning.
An additional consequence of (a) is that two can play that game, turning the table on people like Krauss: if everything involving empirical evidence counts as science, then everything involving critical reflection counts as philosophy, which means that really we are all philosophers and that philosophy departments ought to get the lion’s share of federal funding and campus resources! Yes, this is a reductio ad absurdum, but it is only slightly more absurd than the scientism defended by Krauss and company. QED.