About Rationally Speaking

Rationally Speaking is a blog maintained by Prof. Massimo Pigliucci, a philosopher at the City University of New York. The blog reflects the Enlightenment figure Marquis de Condorcet's idea of what a public intellectual (yes, we know, that's such a bad word) ought to be: someone who devotes himself to "the tracking down of prejudices in the hiding places where priests, the schools, the government, and all long-established institutions had gathered and protected them." You're welcome. Please notice that the contents of this blog can be reprinted under the standard Creative Commons license.

Wednesday, April 14, 2010

The future of philosophy of science

I am in Tilburg, Netherlands, for a conference on the future of philosophy of science. Ah!, you might say, and what would that look like? I hope to write at least another entry or two in the course of the next few days to give you a flavor of what some of my colleagues here think, but let me start with my own views (not because they are better, but just because I have easy access to my own notes...).
Noted (and notorious) physicist Richard Feynman once quipped that “philosophy of science is about as useful to scientists as ornithology is to birds,” thereby encapsulating both scientistic arrogance toward a non-scientific discipline and a pernicious misunderstanding that many scientists have about philosophy.
Let me start by noting that the point of ornithology is most definitely not to be useful to birds, but that despite the fact that birds live their entire lives without even knowing of the existence of ornithology, we as human beings value that particular activity for its intrinsic rewards — both aesthetic and in terms of accrued knowledge and understanding about the world. It is also hard not to point out that the very survival of many bird species may, in fact, partially depend on the research conducted by ornithologists, particularly those interested in conservation biology and whose work informs the decisions of regulatory agencies concerned with species extinction. So much for an ironically flawed analogy, Mr. Feynman.
More seriously, I see the future of philosophy of science along three major lines of inquiry: as an independent discipline that studies scientific reasoning and practice; as a discipline contiguous to theoretical science; and as a crucial simultaneous watchdog and defender of science in the public arena. The first role is rather traditional for philosophy of science, the latter two are more recent developments, and are still very much evolving.
Philosophy of science started out as an independent field of inquiry into how science works, and has been practiced as such for most of the 20th century. This aspect of the discipline is of no particular concern to scientists, unless they wish to inform themselves about what’s “under the hood” of science itself. Contrary to popular perception, philosophers have made much progress in this area, though of course progress in philosophy does not arise from settling empirical questions (that’s science, you know), but rather from the increasing clarification of conceptual issues.
One such issue is the so-called problem of induction, first formulated by David Hume. Induction is the type of reasoning from specific examples to general statements about the world that characterizes much of scientific practice. Hume’s question was how, exactly, do we know that induction is a reliable rational tool. As it turns out, the common answer — often given by scientists themselves — will not do. That answer is that induction “works” (a statement usually accompanied by a visible smirk by self-professed “pragmatists” who can hardly be bothered with philosophical hair-splitting). But how do we know that induction works? Well, obviously, because it has worked in the past. Ah, yes, but that answer is itself a form of induction, which means that we are now justifying induction inductively, thus engaging in circular reasoning (a logical fallacy). Oops.
As it turns out, Hume’s question has spurred decades of thoughtful discussion, which have resulted in a number of ingenious attempts at solving the problem of induction. The most famous (and failed) attempt, of course, was Popper’s idea that scientific hypotheses cannot be proven true, but can be falsified. The issue of induction is not yet settled, but progress has been made in the sense that a number of proposals have been examined, some of which have been found wanting to the point of being essentially discarded, while others are still at least partial contenders and are being constantly refined (for a good introductory discussion of several of these ideas see James Ladyman’s Understanding Philosophy of Science).
In my talk I list a number of other issues concerning the foundations of science about which philosophers have made progress, including the distinction between the (often a-rational) context of discovery vs. the (largely rational) context of justification of scientific theories; the Duhem-Quine thesis that undermines falsificationism; the idea of theory-ladeness of observations (which therefore cannot simply be assumed to be neutral arbiters allowing discrimination among rival theories); the underdetermination of theories by the data (which has found a spectacular example in the ongoing floundering of string theory); and the ongoing debate between realists (who think that scientific theories in some sense really describe the world as it is, at least approximately) and anti-realists (who think that scientific theories are merely empirically adequate, but in no meaningful sense “true”), again a discussion that finds important applications in real science, particularly in quantum mechanics, where various schools of (realist) “interpretation” of the theory are battling it out amongst themselves and against the (antirealist) “shut up and calculate” approach.
The second area of development of philosophy of science is what philosopher Hasok Chang (in his book Inventing Temperature) labeled “the continuation of science by other means.” This is a joint effort between philosophers and conceptually minded theoretical scientists, which has flourished in both fundamental physics and in evolutionary biology (not to mention in math, though I don’t think of math as a science).
To mention just a few examples from the field with which I am most familiar, evolutionary biology, our understanding of important concepts such as species, natural selection, genetic drift, levels and targets of selection, and the distinction between “selection of” and “selection for” are all instances where science itself has benefited from the input of philosophers. To zero in on just one specific case, Samir Okasha’s book on the levels of selection is the most lucid discussion of the mathematics and theory behind group selection that I have seen in a long time, and his argument that species selection is possible while clade selection is incoherent ought to be considered by any serious biologists interested in macroevolution.
The third area where I see an interesting future for philosophy of science is in what I broadly term “science criticism.” The term has an unfortunate connection with certain postmodern approaches and with the so-called “strong programme” in the sociology of science that has been famously (and — largely — justly) been mocked by Alan Sokal with his famous hoax perpetrated at the expense of the editors of Social Text.
But it seems to me that serious philosophy of science ought to reclaim science criticism as a legitimate area of inquiry that also provides an important service to society at large (which, ironically, was also the aim of the mostly misguided postmodernist critique of science). Science is important not so much because of its intrinsic value in satisfying human curiosity, but because it provides answers to practical questions — ranging from how to cure cancer to how to annihilate entire cities (the latter obviously illustrating the dangerous dark side of the scientific enterprise). That is why so much taxpayer money goes into science, not to satisfy a small group of biologists’ obsessive curiosity about, say, the sexual habits of a particular species of moths.
But scientists themselves should not be the only guardians of the huge societal resources that go into science, nor the only ones to make decisions about how to use the outcome of their work. Yes, there are politicians who hold the purse and can push that fatal button to launch the atomic strike, but politicians are not particularly knowledgeable about either the practice of science or the ethics of scientific discovery.
Enter serious philosophy of science, a discipline grounded in the humanities, and yet practiced by people who also have to develop an in-depth understanding of science — both the process and its outcomes. Philosophers, working together with (not in opposition to) scientists, have a huge role to play in furthering societal dialogue about science, including both criticism and defense of science. Let me briefly mention one example of each type.
My colleague Jonathan Kaplan has been an intelligent critic of some practices and assumptions common in medical genetic research, where much that concerns the general public is done using either questionable methods or debatable assumptions about the complex issue of the interaction between nature and nurture. Jonathan discusses, for instance, what we mean when we talk about a “genetic” disease — such as phenylketonuria — which happens to have a relatively simple environmental cure (stay away from phenyl-alanine, which is clearly stated on every can of coke you drink). His type of nuanced discussion ought to be part of both the decision making process about funding of medical research, as well as of how the results of such research are explained to the general public and applied in medical practice. It’s not that the philosopher becomes the ultimate arbiter of worth, but it can hardly be argued that thoughtful contributions by people external to medical research, and yet familiar with its methods and assumptions, have nothing of value to bring to the table.
As an example of philosophy coming to the defense of science, of course, I only need to point to the many crucial contributions of philosophers in the ongoing debate about creationism and intelligent design. This is a societal, not a scientific controversy. But precisely because of that, it is all the more important as it has practical consequences for the public education of the next generation of citizens (not to mention for the continuing funding of evolutionary biological research). In this context, I only have to mention that Judge John E. Jones III, who presided over the famous Kitzmiller v. Dover case in 2005, relied heavily in his decision against the teaching of intelligent design on the arguments advanced by two philosophers, Barbara Forrest and Robert Pennock. The Judge concluded that ID has no standing in public education because of three factors:
“(1) ID violates the centuries-old ground rules of science by invoking and permitting supernatural causation;
(2) the argument of irreducible complexity, central to ID, employs the same flawed and illogical contrived dualism that doomed creation science in the 1980’s;
(3) ID’s negative attacks on evolution have been refuted by the scientific community.”
Of these, the third argument relies on the results of scientific research, but the first two are inherently philosophical (the first one is about the proper epistemic domain of science, which does not extend to the supernatural; the second one relies on a logical fallacy, contrived dualism).
Last year was the 150th anniversary of Charles Darwin’s Origin of Species, a milestone for the scientific and naturalistic understanding of the world. But it was also the 50th anniversary of C.P. Snow’s essay on the two cultures, focusing on the counterproductive divide between science and the humanities. It seems to me that modern philosophy of science is the discipline best suited to productively bridge that divide, rooted as it is in a humanistic understanding of the sciences.


  1. It was also a witty thing to say, should we cut Feynman some slack for being funny? A very odd thing to say, though, for someone who said:


    ( or http://tinyurl.com/ykfvyz5)

    I can't fit a cigarette paper between that and Popper.

    Speaking of Popper (and drifting off topic, because I am a Popper obsessive) I don't agree that the Duhem-Quine thesis undermines him. I think it makes things a lot more complex than a nice dichotomy between existential and universal statements, but the base logic still holds. And just to secure my credentials as a crank: induction doesn't exist!

    The more on-topic point is that I think one of Popper's greatest errors was with the criterion of demarcation. Not the criterion itself (which is fabulous) but with the goal of distinguishing science from non-science. Far too limited! I'd like to see the philosophy of science expand its horizons and treat science not as "something done by people in lab coats" but any attempt to understand the factual nature of the world.

    That partly links into your "third area". Popper's criterion not only provides a great way of distinguishing rational factual assertions from irrational factual assertions (and so could help cut through the bullshit in public discourse) but also helps distinguish factual assertions from other assertions.

    "But scientists themselves should not be the only guardians of the huge societal resources that go into science, nor the only ones to make decisions about how to use the outcome of their work."

    I agree and think much of the problem is with an idea that there is "science" which scientists know about and that this is "their territory". But they are only competent qua scientists on factual assertions, and science funding mixes up a load of non-factual assertions.

    In trying to demarcate science from other stuff we miss out on demarcating what we should: that there are factual assertions and non-factual assertions (as you pointed out Sam Harris could do with a refresher course in this).

  2. Tired, late hour, and of course for matters like this my lack of education is philosophy is fatal, so I will only raise two very small points apart from first saying that this is a very nice, eloquent post that makes a good case for philosophy of science.

    Nevertheless, (1) I personally, in my daily research, feel like the bird does about ornithology. It is nice to know that there are philosophers working on the problem of induction, sure, but I simply don't care about this particular issue. Partly because I would never get anything done if I did, partly because I see science as a natural, only very formalized and sophisticated extension of how we also try to understand the material world privately on a daily basis, starting as toddlers, and we also never worry about the induction when we learn that stinging nettle hurts and that smiling at mommy makes her smile back.

    (2) The Dover issue leads us back to what I wrote a few threads back. If science cannot examine the supernatural, what would a scientist do if we were to discover that sacrificing frogs to the Mayan water god Tlaloc reliably but miraculously caused rain clouds to form, P < 0.0001 and all? Walk away and say, "not my business, epistemic limits, can't examine this in more depth, gotta go"? Or would that by definition not be supernatural any more because it is reproducible? (Well, admittedly not to me either, as I would also consider a god part of nature if it existed.) But what, then, is supernatural? Nothing, it seems, making that point one a bit weird - it must either forbid scientist to examine certain evidence, or it says the same as point three, that there is no evidence for ID. I'm not an American, so this does not touch me as much, but that argumentation from Dover seems a tad unwise to me. I would have simply have said point three is the killer, and that's it.

  3. Tony Lloyd wrote:

    I'd like to see the philosophy of science expand its horizons and treat science not as "something done by people in lab coats" but any attempt to understand the factual nature of the world.

    I think many members of the general public think of science as "something done by people in lab coats". Yet when asked to name a famous scientist, people commonly cite Albert Einstein, a theoretical scientist who didn't spend much (any?) time in a lab coat.

    It is well known to philosophy of science that doing science doesn't necessarily mean taking measurements yourself (much less conducting experiments). However I disagree with including "any attempt to understand the factual nature of the world" under the umbrella of science. Such a broad characterization would admit all sorts of activities—certain religious practices, for example—that have little to do with what is generally understood as science.

  4. "The issue of induction is not yet settled.."

    No kidding :-)

    Falsification works, within its limits. Popper was right, but tried to take falsification too far. I don't see an issue here. Scientific falsification and philosophy of science are not mutually exclusive realms. Each has a place - why try to undermine one with another? Seems counterproductive.

  5. Mintman,

    my point was not that the bird ought to care about ornithology (except when the ornithologist comes to the rescue to avoid extinction!), but simply that ornithology is worth pursuing in its own right.


    yes, I liked Feynman a lot too, and I appreciate the humor. But his quip was a perfect introduction to my point.


    I think the Duhem-Quine thesis does fatally undermine any useful falsificationism. There is still quite a bit worthy of consideration in Popper's general ideas, but the thesis does show that one cannot, in principle, zero in and falsify a theory independently of its corollary assumptions. And the history of science backs this conclusion up spectacularly (e.g., Copernican theory, Newtonian mechanics).

  6. Completely agreed, and understood.

    By the way, when the religious fundies or even only cost-cutting short-sighted politicians gain so much momentum that they manage to threaten this bird's habitat, I doubt they will be cowed by a bold philosopher of science stepping in front of them... both groups are inherently immune to these kinds of arguments. Intellectual curiosity about the underpinnings of science sounds like the more convincing reason to appreciate its philosophy.

  7. Mintman,

    yes, I do not overestimate the power of philosophy myself, still the example of Judge Jones' decision is a clear instance of practical help given by philosophers to a science education cause.

  8. Great post Massimo. I can't get enough of the subject you address.

    What I would like to understand is how you would alternatively reconcile the problem of induction (and its lack of resolution) with the fruits of the empirical method of science. On the one hand you seem quite confident that there are scientific truths, i.e. that there is some definite meaning (at least to us as humans) to truth in this perspective. On the other hand you seem resolute to dismiss any rigorous undergirdings for this. How then can we not conclude that scientific truths should receive the same skepsis as any supernatural claim? I thought I knew the answer to disentangle that one (something like the explanatory and predictive power of science, which is in essence is a pragmatic take on things) but you made me question it again by suggesting that any pragmatism is bunk.Aren't you professing some sort of pragmatism yourself when you hail the fruits of science and philosophy by presenting benefits in real world cases?

  9. Fujaro,

    you may be reading too much in my criticism. I think the best answer is indeed some sort of pragmatism (which Hume himself endorsed), together with the realization that science is a web of knowledge, not an edifice built on a foundation.

  10. Dear Massimo, meeting you really made my day. I couldn't have wished for a better present for my 31th birthday.

    I will email you the details of the conference in NYC.

  11. "I think the Duhem-Quine thesis does fatally undermine any useful falsificationism."

    Help me out here. Falsification is a cornerstone of scientific method. Duhem-Quine states that no theory can be conclusively falsified via empirical means if the background assumptions are not proven (note operative words: "proven" and "conclusive").

    The hard sciences approach reality through probability analysis, which is always provisional to new data. Plain old experimental theory. Falsification is fundamental to achieving increasingly higher probabilities towards any given theory. You disagree?

    Ironically, Duhem-Quine couldn't exist without homage to the correctives of falsification. What D-Q ultimately seems to say is that "nothing is certain." Their argument seems more applicable to philosophy and epistemology than conventional hard science. Hard sciences deal with successive probability, not philosophical certainty.

  12. John,

    yes, I disagree. It is actually hard to find any case of falsificationism in the history of science. That's just not the way science operates.

  13. No amount of experimentation can ever prove me right; a single experiment can prove me wrong.
    -A. Einstein

  14. John,

    I'm quite sure Massimo did not mean that it's hard to find any case of a scientist invoking falsification in the history of science.

  15. Scott/Massimo, we must be talking about two different definitions of the word. Falsification is as embedded into the scientific method as hypothesis, measurement, and deduction. In many (most?) cases, deduction IS falsification. The very nature of informed observation -implies- certain embedded logical attributes.

    I perceive a continuum of hard science that spans from realist-positivist to subjective-relativist (after Polanyi). Granted, in the latter, falsification becomes dicey for a number of reasons. If that's your argument, I get it. If not, sorry - you've lost me and I would appreciate some help.

  16. John,

    Scott is correct, I meant that it is hard to find cases in the history of science where falsification has actually been used. For instance, the Copernican theory should have been falsified by several observations, including the apparent lack of stellar parallax, and yet scientists ignore that and forged ahead with the theory.

    You are right in a sense that falsification is embedded in deduction, that was Popper's idea. But the fact is that scientific hypothesis testing is hardly deductive.

  17. Fair enough. I work in technology, which might be called an applied science in that we're constantly coming up with ideas (hypotheses) and testing them in hardware and software. The ideas that don't meet our objectives (most of them) are "falsified." We use all the -ductives in the process: inductive, deductive, and (perhaps most importantly to our work) abductive.

    I would disagree that the scientific method is "hardly deductive." I can think of very few hard science processes that do not embrace all the -ductives in some manner.

  18. John,

    I think there's an important distinction between falsification of hypotheses (easy) and falsification of theories (hard). Falsification of hypotheses often amounts to little more than modus tollens:

    x -> y
    therefore ~x

    But falsification in the Popperian sense has to do much more than modus tollens, because it's very easy to create a statement that's falsifiable by modus tollens but that doesn't come close to being a useful scientific hypothesis. For example:

    The world was created by an infinite unperceivable chicken and G=9.81m/s^2

    So once you start using falsification to think about larger aggregates of statements (i.e. theories), it becomes less useful.

  19. Scott, yes exactly. This is my perception as well. I was simply responding to claims such as "so and so fatally undermines any useful falsification." When I hear the term "useful" I think of applied science, rather than theoretical. In the former, falsification is not only useful, it predominates.