Wednesday, June 16, 2010

What is Naturalism? Part I: An Overview

What is naturalism? Or more precisely, what is metaphysical naturalism—naturalism viewed as a theory about reality? This is the question I intend to explore in a series of posts that will, hopefully, do a couple of things once we’re finished: first, clarify some issues that come up repeatedly on this blog; second, set the stage for more precisely characterizing (in future posts) the roughly Hegelian method for trying to understand reality which I tend to favor.

So, what is metaphysical naturalism? To say that it’s a theory which rejects the existence of the supernatural isn’t especially helpful, since what’s really at issue when we ask this question is precisely how to draw the distinction between “natural” and “supernatural.” For this reason, some choose to define metaphysical naturalism by reference to science. On this understanding, to be a metaphysical naturalist is to believe that what can be studied scientifically exhausts what is “real.”

This approach strikes me as in the ballpark, assuming we can specify the “scientific method” of inquiry in a sufficiently narrow way, and assuming that we can do so without appeal to the natural/supernatural distinction. A key move in contemporary efforts to characterize the scientific method (and one I’m sure I’ve resorted to myself) is to say that scientists are methodologically naturalistic. In other words, to do science is to look for naturalistic explanations rather than supernatural ones. Obviously, if we define or characterize the scientific method in these terms, we won’t be able to define naturalism by reference to science without finding ourselves right in the middle of a vicious circle.

What I propose to do here is pursue a definition of metaphysical naturalism (and by implication metaphysical supernaturalism) in relation to a method of inquiry that at least characterizes much of what scientists take themselves to be doing. Whether this method is coextensive with the scientific method is an issue I won’t take a stand on here (I suspect it's not, especially not at the level at which theoretical physicists operate).

With respect to both the method of inquiry I wish to describe and the definition of naturalism in relation to this method, my intention is to appeal to Kant’s distinction between phenomena (in brief, appearances) and noumena (the realities that underlie appearances).

In this post, however, I want to offer a kind of overview of the position I will be developing. By necessity, this overview will gloss over lots of details which will be fleshed out more carefully in subsequent posts—so concepts and distinctions that aren’t as clear as one may wish for here will hopefully be fleshed out in future posts. My objective in this post is to provide enough of an overview of the forest so that, once we’re in the midst of the trees, we’ll be less likely to get lost.

In briefest terms, then, what I will argue is this: there is a method of inquiry widely used by scientists which makes use of “phenomenal/empirical observations" to generate models of how the world works, models which are not themselves directly observed (phenomena) but which would have phenomenal effects that are. Such models are accepted on two conditions: first, we must be able to predict further phenomenal observations based on them; and second, those predictions must consistently come true.

This methodology ultimately generates a picture of the world which I’ll call (for lack of a better term) a “scientific picture.” And the scientific picture that has emerged today is very different from the way the world immediately appears to us (the phenomenal world). What I will propose is that metaphysical naturalism is the theory that the noumenal world in roughly Kant’s sense (what really is the case apart from our experience of it) is to be identified with this scientific picture.

More precisely, noumenal reality is to be identified with the scientific picture that would ultimately emerge once this methodology has been pursued to completion—that is, to the point at which no more refinement in the picture is possible because all actual and potential phenomenal observations have been perfectly modeled. Stated in these terms, it should be clear that, on this naturalist hypothesis, we will never actually arrive at a full understanding of the noumenal realm, but will steadily come closer and closer to it the more we rely exclusively on this “scientific” method of inquiry for developing our understanding of reality.

In any event, that’s a summary of what I take “metaphysical naturalism” to be. But a fully adequate articulation of this naturalist thesis—one which exposes both its strengths and weaknesses and helps us to better conceptualize the alternatives to it—will require a more in-depth look at several things, including Kant’s phenomenal/noumenal distinction.

My next post will therefore be about that.

7 comments:

  1. Eric,

    You write: “ there is a method of inquiry widely used by scientists which makes use of “phenomenal/empirical observations" to generate models of how the world works,[snip]

    I think it is more precise to say that science catalogues phenomenal facts and mathematically models them. The models are simply mathematical relations between phenomenal facts so that if one knows part of the data one can predict (either completely or probabilistically) the rest. Models of phenomena qua models of phenomena do not describe how the world works. That used to appear to be the case in classical physics, but in the 1920s physicists were able to model quantum phenomena *without* conceptualizing anything about how the world works, let alone modeling how it works. It was left to those who proposed “interpretations” of quantum mechanics to model how the (naturalistic) world works with all the known problematic results.

    Such models are accepted on two conditions: first, we must be able to predict further phenomenal observations based on them; and second, those predictions must consistently come true.

    To nitpick: If one can use a new proposed model to predict up to now unknown phenomenal facts and these facts are verified then this of course greatly increases the warrant for this model. On the other hand such is not necessary. A new scientific model has sufficient warrant if it is successful in mathematically relating known phenomenal facts, even if it doesn’t make any new predictions. For example if one would find a mathematical model that would mathematically relate quantum phenomena with gravitational phenomena then this would be a great scientific discovery even if no new predictions were made. On the other hand, empirically speaking, it has always been the case that better and more general models did produce at least some successful predictions about unknown phenomenal facts. Such is the case with Darwinism which models the phenomena related to our observation of biological organisms, but makes hardly any new predictions.

    More generally: If it is the case that science does not as a matter of fact model how the world (the noumenon in Kant’s terminology) works, or how the world is, then the idea that naturalism is the claim that the world is as science models it makes no sense. And, surely, the principle of charity does not require one to try and embrace the other party’s fallacious thinking. My point here is that many a naturalist’s notion of naturalism is based on a factual misunderstanding about what science does. (I know that this is a strong claim, because it implies that many a scientist too misunderstands what science does; on the other hand anybody who understands the philosophical distinction between phenomena and noumena will see that the data on which scientific knowledge rests are insufficient for deciding about the noumenon. – Incidentally wouldn’t it be simpler to speak about the distinction between phenomena and reality, i.e. about how reality seems and about how reality is?)

    When I put myself in a naturalist’s shoes then it seems to me that the defining characteristic of naturalism is the idea that all that exists is ultimately of a mechanical nature. One way to define “mechanical” is to say that all possible knowledge about causality can in principle be expressed mathematically (including using probability theory), or, in yet other words, that reality is such that it evolves according to fixed (and blind) mathematical laws (including probabilistic laws). Thus defined, naturalism is the negation of free will. Which strikes me as a more appropriate definition of naturalism than Alvin Plantinga’s definition that naturalism is “the idea that there is no such person as God or anything like God”. For a naturalist would deny that there are effects caused by free will I
    (which therefore do not fit with its mechanistic understanding of reality) even if the agent of that free will were some being quite unlike God.

    ReplyDelete
  2. Hi, Dianelos-

    Firstly, Darwinism is rich with predictions that have been fulfilled, so you might better stick to your knitting there. It is also largely non-mathematical, and is a matter of imaginative modeling, so I think your general line of reasoning is incorrect. We all agree that the problem of induction is serious, as put by Hume, et al. so all these models are tentatively describing past events and correlations, relying on the regularity of the world (which expresses its mechanical nature) to assure us that such modeling and predictions will continue to hold true in future time.

    Thus we predict the sun's demise by stellar models, though it may be hit by a stray black hole at an earlier time. There are places, such as quantum mechanics, where a mathematical system may be very successful at describing and predicting reality while our intuitive/imaginative modeling lags behind. Our intuitions may be left forever behind in that field, being insufficient to the task. But that is no reason to make our intuitions dispositive and flee to theology instead!

    After all, theology and supernaturalism attempt an imaginative modeling of reality, or sur-reality, just the same as sciences do. Our brains are modeling machines, after all. All that naturalists want is that such modeling be logically tied to empirical observations, (thus also falsifiable), and that the theories/models developed be what is minimally necessary to address the problems at hand rather than being ornately excessive, clearly imposed from our creative writing / mythical traditions.

    I don't think naturalism demands any particular view of the necessary contents of reality, and it is open to anything that fulfills the above methodological approach. So I think Eric is going in the right direction above. It just seems that naturalists only pay attention to mechanistic phenomena, since that is all that have been reliably found outside our rich social and imaginative lives, themselves apparently based on mechanisms in turn.

    --- Wow- cool new page theme!---

    ReplyDelete
  3. Hi Burk,

    You write: “ Darwinism is rich with predictions that have been fulfilled

    Right. My wording was not the best, but what I meant was that even if in some sense it is the case that Darwinism makes less predictions than other major scientific theories it does not mean that Darwinism is any less scientific (as some creationsists claim).

    [Darwinism] is also largely non-mathematical, and is a matter of imaginative modeling, so I think your general line of reasoning is incorrect.

    It seems to me that any scientific theory describes a mechanism, and a mechanism is by definition amenable to mathematical modeling. Even if Darwin himself did not use much math, he did describe a mechanism (and what is actually an algorithm and thus an abstract mathematical object). Today, I understand, a lot of math is used by Darwinists, and with obviously good reason: Science is a quantitative field of knowledge, and no matter the prima facie plausibility of Darwin’s theory as a case of “imaginative modeling”, if one doesn’t actually check mathematically how well that model works, one will not find out if it is a successful scientific model. (As for the issue of Hume’s discussion of induction, this is completely outside my meaning. For me truth is justified by its practical success, and as long as I find induction to be a successful epistemic principle I will continue using it.)

    All that naturalists want is that such modeling be logically tied to empirical observations, (thus also falsifiable), and that the theories/models developed be what is minimally necessary to address the problems at hand rather than being ornately excessive, clearly imposed from our creative writing / mythical traditions.

    As I argued in my previous post the models that naturalists have been proposing lately are non-falsifiable, and, in my judgment at least, much more “ornately excessive” than anything that the theistic imagination has come up with so far. For example, according to the many-worlds interpretation of quantum mechanics each time you switch on a light you give birth to an enormous (perhaps infinite) number of parallel universes, each with a copy of you in it. According to the same interpretation there are universes out there where we shall all be bodily raised three days after we die.

    ReplyDelete
  4. I thought it would be helpful to paste Rich Griese's comment here so that it's easier to engage with it on this site. Here it is:

    Rich Griese says:

    "This guy is a supernaturalist, and like many supernaturalists they go out of their way to try to equate religion and science.

    "They have a few standard techniques. One of them is to say... of evolution is just a belief, so if you are going to teach the belief of evolution in schools, you should also allow the teaching of the belief of Christian creationism in school.

    "It is very similar to some of their attempts to constantly equate supernaturalistic belief with science. Another of the tricks is that when you start to talk about if we can know that gods exist, to kind of move to the gambit of talking about if we can really 'know' anything.

    "I have noticed that there is kind of two paths here. If you accept their sales pitch that supernaturalistic gods exist, then it's all well and good. but... when you start to move towards trying to point out that this stuff is all speculation... they will try to sell the idea that EVERYTHING is speculation, so what does it really matter.

    "Well, I think this is BS. We know that my car exists, we know that my high school exists, we know that gravity exists... to try to use sophist like arguments to try to make it seem like we don't know anything simply because it is pointed out to people that gods have never been demonstrated, is pretty pathetic if you ask me.

    "I am sure in this series of his we are going to be introduced to a LOAD of BS, so strap on your BS detector."

    ReplyDelete
  5. In response to this, I want to stress that I do not equate religion and science--as anyone who takes the time to undersand my philosophical position will readily see.

    Neither does Kant. For Kant, science studies the phenomenal world--the only realm about which KNOWLEDGE is possible. And science gives us knowledge about the phenomenal world.

    What it doesn't do is give us knowledge of the noumenal, which Kant says is unattainable by any means, scientific or otherwise. For Kant, religion gives us NO knowledge of the phenomenal OR noumenal world. But he does nevertheless seek to delineate a species of religion that reason can endorse--but reason endorses it not as something known, but as POSTULATES about the noumenal realm that are necessary in order for PRACTICAL RATIONALITY to be possible--in other words, in order for it to be possible for reason to motivate behavior.

    Not all religion does this work, and so not all religion meets the demands of practical reason. Thus, Kant thinks there are criteria for distinguishing between rational and irrational religion. But religion that meets these criteria is not and never will be knowledge, since religion makes postulates about the noumenal realm and all knowledge is of the phenomenal realm.

    In short, science gives us knowledge of the phenomenal realm, while religion makes postulates (which can be more or less in line with the requirements of reason) about the noumenal realm. There is, as such, no parity between science and religion for Kant.

    My own view is in many ways Kantian, although there are important differences. These differences aside, I share with Kant the conviction that there is no parity between science and religion. Science helps us to understand the workings of the empirical world, while religion makes postulates about that which transcends the empirical world. This is one reason why I find "creation science" and its inheritors so confused, and why the attempt in Rich's comment to associate me with creation scientists is not merely pesumptuous but wrong.

    I could go on, but I'll stop there.

    ReplyDelete
  6. "... POSTULATES about the noumenal realm that are necessary in order for PRACTICAL RATIONALITY to be possible"

    If neither science nor religion traffic with the so-called noumenal, then it is hard to understand why anything about it is necessary for anything else, let alone for practical rationality, which would seem rather confined to phenomena. Am eager to hear more!

    ReplyDelete