Rosenhouse on scientism

December 18, 2011 • 8:55 am

Jason Rosenhouse is always worth reading (except, for me, when he writes about chess!), and his latest post at EvolutionBlog, “What is scientism?“, is a penetrating analysis of the slur that’s often levelled at atheists and scientists by accommodationist and believers. Jason was inspired by a peevish post from (surprise!) Michael Ruse, who, feeling that philosophy is being dissed in discussions about “ways of knowing,” argues that mathematics and morality are sources of “genuine knowledge.” Jason responds:

In the context of science/religion discussions, this definitional morass seems supremely unhelpful. It’s far too abstract. The real issue is very simple. If you are going to make assertions about how the world is, then it is on you to provide evidence for that assertion. Then people can decide for themselves if they think your evidence is any good. What science (defined in some reasonable, everyday sense) provides is a set of investigative methods that everyone regards as legitimate. In this it differs from religion, which points to sources of evidence, such as personal experience or the contents of holy texts, that are considered by many to be of highly dubious validity.

Jason, who of course is a mathematician, argues that much of mathematics is indeed part of science. “Pi,” for example, comes from empirical observations about the ratio of a circle’s circumference to its diameter.  That’s true, of course, but that seems an exception, and much of mathematics is simply logical deduction from axions.  Math is a useful tool for understanding the world, but with few exceptions the discipline itself, absent empirical observation, doesn’t tell us truths about the world.  But Jason’s willing to buy that, too:

But just for the sake of argument let’s suppose we are absolutely determined to define our terms in such a way that mathematical knowledge is not part of science. Very well. Since I am happy to grant that mathematics provides knowledge, I will consider scientism to be refuted. In its place I will suggest a new notion called “scienceandmathematism,” which is defined as the idea that science and mathematics are the only reliable routes to knowledge. Happy now?

Jason argues, correctly, that while science can inform moral judgments, in the end statements about right or wrong (or, in Ruse’s case, whether one should feel ashamed of an action) are opinions, based on subjective value judgments.  I think that’s true: even Harris’s justifiable claim that well-being should be the criterion for morality is not something that can be justified through science.

Finally, Jason takes on the really annoying claim (one made by theologians like John Haught) that because we can’t philosophically justify a priori the exclusivity of science as a way to find truth, then science devolves to a faith—like religion. (The object here, of course, is to drag science down by analogizing it to faith.)  And Jason’s response is the one I always give:

But why can’t we justify scientism on scientific grounds? I would think there is a plausible argument to be made that our confidence in scientism is an inductive inference from the persistent success of science coupled with persistent lack of success of all other routes to knowledge. Ruse earlier defined science as a generalization from experience. Is that not precisely the basis for a confident assertion of scientism?

Certainly the distinctively religious ways of knowing that people have suggested over the years have frequently proven themselves unreliable. Philosophical and ethical analysis are certainly valuable activities, but it seems strange to me to describe them as ways of knowing. What they provide is not knowledge, but clarity.

Indeed.  If you want to see a philosopher’s justification of scientism—in this case “philosophical naturalism,” read Barbara Forrest’s paper published in 2000 in Philo, “Methodological naturalism and philosophical naturalism: clarifying the connection.” The paper is free at the link; do read it.

Finally, in response to an almost incoherent statement by Ruse, Jason brings his piece to a close:

I don’t know what it means to say, “[I]f it isn’t science it isn’t genuine.” Genuine what? What I do know is that an assertion that science is the best, and perhaps the only, way of genuinely knowing the world is not a diss to the humanities. It certainly is not a rejection of mathematics, philosophy or ethical reasoning. And if you are going to argue that the assertion is false then it is your burden to point to a better way, and to indicate the knowledge provided by that alternate method.

There can be no knowledge about the universe that doesn’t derive from reason and empiricism, or that can’t be tested by empirical observation. Broadly construed, that’s science.

45 thoughts on “Rosenhouse on scientism

  1. A little bit of a tangent here and a pet peeve of mine. I don’t think Harris argued you could demonstrate with science that well-being should be the goal of morality. He argued that well-being should be the accepted goal by conscious organisms and that science can tell us the best ways to maximize it.

    1. I disagree; I think that Harris contended that morality IS the maximization of well being. I don’t disagree that science can tell us the best ways to maximize it, and I don’t think many of his critics disagreed about that, either. They contended that Harris’s claim that science tells us WHAT IS MORAL was not a claim about science informing morality, but that science tells us that well being IS morality.

      1. Harris’s claim … that science tells us that well being IS morality.

        I don’t think that the claim that human well-being *is* morality can hold, since our moral senses are the product of evolution, and thus if we ask the cui bono? question then the answer would be gene propagation, not the well-being of humans (of course there will be overlap between those two, but the concepts are clearly distinct).

        Thus I agree with you that what “science tells us” is that morality judgments are the subjective feelings of humans, a view I defend here.

    2. Perhaps the better starting point is to say that well-being *is* (not “should be”) the goal of conscious organisms, since otherwise you have to justify the “should”.

    1. Really? Why?

      I think this is a very superficial view of mathematics. Yes, it is not strictly based on empirical evidence and statistical analysis of mountains of data (but it has provided the tools of probability and statistics).

      But what axioms there are are intuitively motivated by observation. The generalizations of number, sets, space and dimension, objects in space, movement of objects in space, changes with respect to time and with respect to one another, the modeling of fluids and flow, the algebraic manipulation of vector spaces, the analysis of multidimensional spaces, the algebra of groups and fields, the analysis of vector spaces and differential geometry, the analysis of complex numbers and the complex plane and function spaces and topology and differential equations etc. are all motivated by intuitions gained from observing natural reality.

      These abstract generalizations are so closely connected to nature, objects, and space that physicists routinely find ready made tools available providing them a language to describe and calculate predicted results of every new observational discovery they make.

      If mathematics were merely axioms and logical deduction without reference to nature, it might have as much application in the natural sciences as abstract expressionism or post-structural cultural criticism.

      Instead mathematics provides all scientists with new views upon their observations, and ways to think about and model the observed systems.

      1. +1.

        To add to that:

        Proofs are heuristic devices, and we know it. No one has been able to prove general proofs by proof theory.

        Hence proofs can be erroneous, either by application or inherently. Developing math is very much an empirical trial-and-error muddle.

      2. If mathematics were merely axioms and logical deduction without reference to nature, it might have as much application in the natural sciences as abstract expressionism or post-structural cultural criticism.

        Except that the language of mathematics is so general that the question is not whether its abstractions apply to something, but which of its abstractions apply.

        From this angle, mathematics is a branch of philosophy, and science a branch of mathematics.

        1. “science a branch of mathematics.”

          I suppose there are many ways to categorize things; I don’t know if any such categorization is fundamental or normative. Viewing things from different angles and for different motives can lead to alternative categorizations.

          There is applied mathematics and numerical analysis, which certainly seem to be the “hard scientific” branch of mathematics in a sense.

          But when physicists are doing their work, the motivation is to understand nature, not to advance mathematics, which kind of argues against the idea of science as a branch of math.

          What I find interesting in thinking about this is that while it is true that mathematics can be thought of as totally independent of empirical measurement and unconstrained by the laws of nature, there is a sense in which empirical scientific observation would be almost meaningless without mathematics.

          How could biologists analyse genomes and draw conclusions about evolutionary links based on genetics without math?

          How could physicists make sense of the data from particle accelerators, and extrapolate from that to draw conclusions about what is happening inside the sun or during the first few trillionths of a second after the big bang without mathematics?

          There is a sense in which empirical observation and measurement would be a meaningless miasma of data overload without mathematics to connect empirical measurement to meaningful conceptual conclusions and to predict phenomenon towards which new experiments can be directed.

          In this sense mathematics is a kind of adaptable scaffolding, structured according to biological, physical, or chemical models, that enables all empirical scientific observation as raw data to assume intelligible shape and meaning. Seen in this way I guess empirical data is like a boundary at which mathematics gains purchase on material reality.

  2. Excellent, Jerry. Thanks for alerting us to Jason’s post.

    Here are a few other good resources on the “scientism” issue…

    Subjective beliefs are not the same as objective knowledge

    Dan Dennett on “Scientism” 

    Defending Science – Within Reason: Between Scientism and Cynicism – by Susan Haack

    A defense of naturalism — and scientism—and-scientism/

    “Six Signs of Scientism” – by Susan Haack

  3. re: “Math is a useful tool for understanding the world, but with few exceptions the discipline itself, absent empirical observation, doesn’t tell us truths about the world.”

    But, isn’t there a sense in which math is entirely grounded upon empirical observation? We observe that the universe obeys the basic properties of addition. This seems painfully obvious to us, but this is because our intuitions about addition are based on a 100% consistent string of observations from our first day. The universe could have been otherwise – some bizarre world in which even addition worked differently. Just as we observe the meaning of Pi, we can also observe the outworking of basic mathematical principles in our universe.

    1. However, non-Euclidean geometry was developed during the 19th century without being based on empirical observations. That it could serve to describe space-time was proposed by Einstein much later, and the observation of the deflection of the light of a star by the sun by Eddington was in fact the first empirical observation of the curvature of space.

      1. That spatial geometry might be non-Euclidean was suggested by Riemann, along with the first fully general version of non-Euclidean geometry. It was a pretty obvious idea, actually. But theories of forces as manifestations of spatial curvature were found either to add nothing new to our understanding, or to be empirically incorrect. Einstein combined his earlier realizations about special relativity, which made space and time inseparable, with the older idea of curvature forces to produce general relativity, in which gravity is really a curvature of spaceTIME.

      2. Non-eucliean geometry is still a branch of geometry. It still deals with concepts that we are all familiar with: how long is a curve, how big is a closed surface, how curvy is a surface, how large is an angle… etc. These concepts are generalized but they are still rooted in our everyday experience. Modern algebraic geometry has become very abstract but not so abstract that its contents are completely unrecognizable as geometry, as we typically understand the word: the shape of things and their spatial relationships.

      3. Spherical geometry is not euclidean (the fifth postulate is not true on a sphere), but spherical geometry should be very familiar to most people. Theorems about spherical geometry typically can be easily verified empirically. More importantly, the development of the theorems were motivated by empirical/pragmatic reasons.

        Non-euclidean geometry actually amplifies the case that geometry is empirical. Euclidean geometry was developed as a model of reality. It is very limited as it fails to make prediction about empirical things (such as length of a curve on the sphere), which motivated the development of non-euclidean geometry.

  4. … and much of mathematics is simply logical deduction from axions.

    This seems doubtful. I think axioms are the more likely culprit.

    1. Not quite sure what they are supposedly guilty of, but I’ll certainly agree that they can be somewhat problematic. And that, in part anyway, because they seem to be either inductive hypotheses at best – as the previous comment on the more accurate correspondence of non-Euclidean geometry to actual reality would suggest – or articles of faith at worst.

  5. The charge of “scientism” is hurled at atheists, skeptics, and scientists who want to treat a sacred belief like any other hypothesis. So they confuse the issue by placing fact claims into categories normally reserved for meaning statements, moral assertions, aesthetic expressions, and so forth.

    You believe in God for the same reasons you love your mother. It’s a choice.

    I hate that sly and sloppy tactic.

    1. Aren’t we biologically programmed to bond with (~love) our mothers (and they with us), such that it takes an unusually abusive mother or ungrateful child to break that bond?

      We are clearly not biologically programmed to believe in god/dess/es, but conditioned to do so from an early age – and indeed, that conditioning is intimately entangled with that bond. (I would think that an abusive earthly father and a devout mother would be a particular potent combination for engendering belief.)

      It would be interesting to study the nature of different groups’ relationship with their god: those raised as atheists who had converted to theism and those raised as theists. How badly do people raised without the need for a “loving Heavenly Father” need one, as distinct from finding some kind of supernatural being an essential part of their worldview?

      1. We are clearly not biologically programmed to believe in god/dess/es

        But haven’t social anthropologist, psychologists, et al. identified a natural tendency to infer agency? Thus, deities… 


  6. What they provide is not knowledge, but clarity.

    Sometimes. Mostly though I find philosophers and priests going at it like wet cats in a sack when ever it comes to defining morality and reality.

    Which makes much of what they do ‘clear as mud.’

  7. I would like to challenge all of those ‘other ways of knowing’ to produce just one piece of useful technology.

    Because our generation has grown up surrounded by technology, we simply take it for granted. We do this even as technology advances at such an incredible rate that five years is enough time to render your state of the art toy completely obsolete.

    Take away science and we would be sent back to a world of subsistence farming with no effective medicine no transport faster than walking speed and no communication other than talking.

    Take away all of those other ways of knowing, no-one would notice the slightest difference.

  8. “Pi,” for example, comes from empirical observations about the ratio of a circle’s diameter to its circumference.  That’s true, of course, but that seems an exception, and much of mathematics is simply logical deduction from axions.  Math is a useful tool for understanding the world, but with few exceptions the discipline itself, absent empirical observation, doesn’t tell us truths about the world.

    No and no. Math tells us a great deal about the world without making a single observation because it rules out many possibilities as logically impossible. And π is defined and computed with geometric-algebraic abstractions, not empirical measurements.

      1. the only verification of “logic” is that it works empirically

        Logic works whether it is seen to have empirical manifestations or not. That there is no evidence that the universe is not completely logically ordered is what makes “scientism”—including math and its implications—the only viable source of truth about the world.

        1. “Logic works whether it is seen to have empirical manifestations or not.”

          How do you know that it “works”, except through empirical manifestations? Indeed, what do you even mean by “works” divorced from all empirical manifestations?

          1. How do you know that it “works”, except through empirical manifestations?

            We don’t know anything about anything except through observation (thank you scientism!). But that that doesn’t alter the fact that mathematical truths are necessarily independent of physics and observation. This is nothing more than the “working Platonism” practiced my most/all mathematicians/scientists.

        2. If only it was so simple. Mathematics rules out many mathematical possibilities. But it has nothing to say on physics. No part of physics have been derived from pure math.

          Also, math is not based on logics even if it incorporates it. Proofs are heuristic devices, and math from arithmetic and onwards is based on consistency. (See Gödel on that.)

          Similarly, physics is not based on logics even if it incorporates it. Theories are empirical devices, and physics from theories and onwards is based on testing. (See Popper on that.)

          1. Ouch, “and physics from theories and onwards” should be “and physics from mechanics onwards”.

            But maybe that was obvious, that is how mechanics were derived in the first place.

          2. physics is not based on logics even if it incorporates it

            I really like Feynman’s speculation about the possible role of simple logical operations as the underpinning of physics:

            It is possible—and I’ve often made the hypothesis—that physics will not ultimately require a mathematical statement, that the machinery will ultimately be revealed—it’s just a prejudice, like one of these other prejudices.
            It always bothers me that, in spite of all this ‘local’ business, what goes on in no-matter-how-tiny-a region of space and no-matter-how-tiny a region of time, according to the laws as we understand them today—takes a computing machine an infinite number of logical operations to figure out.
            Now how could all that being going on in that tiny space?! Why should it take an infinite amount of logic to figure out what one tiny, stinky bit of space-time is going to do? So I made the hypothesis often that the laws are going to turn out to be, in the end, simple like the checkerboard, and all the complexity is from size.

  9. “much of mathematics is simply logical deduction from axions.”

    Those would be some kind of fundamental particle that exist by definition… Axioms?

  10. Here’s a tidbit from David Deutsch’s The Fabric of Reality, which I cannot recommend enough:

    “Abstract entities that are complex and autonomous exist objectively and are part of the fabric of reality. There exist logically necessary truths about these entities, and these comprise the subject-matter of mathematics.”

    Since the book is an eye-opening defence of science, broadly construed, and its implications, it couldn’t be more relevant in the light of (predictable, and mostly predictably ignorant) allegations of ‘scientism’.

    1. That is one of my bibles too. =D

      But much of what Deutsch writes on math and simulation is bunkum. (For example, Boltzmann Brains and how to define probabilities over multiverses is a living problem in cosmology.)

      Such texts reach is larger than its grasp of course. As you say an eye-opener, but as always one shouldn’t open ones eyes so wide that they fall out.

  11. It is enough to observe that science, in the distinct definition as the area and its process, is the best way to get to facts. That makes these people start raving, since they want equally valuable “alternatives”.

  12. Do you suppose that if two people have knowledge of all the empirical facts of a matter, that by definition there is nothing for them to possibly disagree about?

    For example, if you and your spouse can predict, down to the atomic level, the empirical consequences of taking your vacation at the beach vs. taking your vacation in the mountains, that it is somehow a conceptual impossibility that two rational people might hold conflicting views about which would be better?

          1. Then it seems as though Prof. Coyne (and the preponderance of posters here) are now presented with a counterexample to the idea that the scientific method is a panacea for the solving of human problems.

            That is to say, there are genuine areas of our experience in which the mere accumulation of facts is of no help, and yet other, nonscientific methods are.

            1. There is an important distinction to make regarding your excellent point: the genuine areas of our experience you are talking about are entirely restricted to subjective experience and subjective evaluation.

              For everything that is outside of the human skull (and a lot that’s inside the skull too) the “mere” accumulation of facts is of tremendous help in providing a basis for improving the agreement with reality of our subjective opinions and ideas.

              And supposition about the objective world that is entirely derived from subjective evaluation with no input from empirical facts, such as for example the metaphysical claims of religion, does not count as knowledge about the objective world.

              It may help us decide where we would enjoy spending our vacation though.

Leave a Reply