Sunday, June 27, 2010

Naturalism Defined

I've gotten involved in an interesting discussion of naturalism at Metamagician and the Hellfire Club. The discussion was instigated by the work of Richard Carrier, who is neither a philosopher nor a scientist, but a historian who writes about both. Unfortunately, I think he should probably stick to history. His attempt to define "naturalism" and "supernaturalism" is marked by fallacy and inconsistency. (Update: I point out several problems with his arguments in a comment attached to this post.)

Carrier's definitions are as follows: "naturalism is true if everything that exists is causally reducible to the nonmental. . . . If naturalism is true, everything mental is caused by the nonmental, whereas if supernaturalism is true, at least one thing is not." Yet, naturalism and supernaturalism are not exclusively theses about minds or mental entities. For example, Carrier excludes the possibility of David Chalmers' variety of naturalistic dualism. (More on naturalistic dualism below.) Furthermore, some naturalists may adopt the Rylean view that the way we talk about mental entities and events is categorically unlike the language we use to talk about physical events. Mental entities are not causes or events in a causal chain. Rather, they are dispositions. So for a Rylean naturalist, it is not the case that "everything mental is caused by the nonmental."

Carrier misses the defining features of supernaturalism and naturalism. In my posts at Metamagician and the Hellfire Club, I offered a replacement definition: Naturalism is the view that science can describe all of the causes. To put it another way, naturalism is the view that no causes are theoretically outside the bounds of scientific discovery.

Here's the main substance of my posts (edited and revised for cohesion), in which I elucidate and defend my view:

Supernaturalism is not the belief that there are some mental entities which have not been caused by physical entities. Rather, supernaturalism is usually defined as the belief that nature is subservient to another realm. Supernaturalists believe that what happens in nature is ultimately caused by events outside of nature, that what is outside of nature holds dominion over nature. Also, "natural" seems to mean "describable with scientific methodology." So supernaturalism is a view about the limits of science to account for nature. According to supernaturalists, science can describe nature, but it cannot describe the ultimate causes of nature--it cannot account for what holds dominion over nature. Naturalism is the rejection of this view. Naturalism is the belief that science can describe all of the causes.

A key issue is how we understand the word "science." I don't think science is defined by a single set of laws or methods. New scientific methods are developed as science progresses. So, when I refer to "science," I'm not referring to anything that is limited by our current capabilities or knowledge. I'm referring to the process of expanding shared capabilities and knowledge.

Naturalism is a philosophical position, not a scientific one. (I'm disagreeing with some well-known and outspoken scientists here, like Sean Carroll. I've expressed my disagreement with Carroll before. So be it.) The reason it is impossible to scientifically falsify naturalism is that, to do so, you'd have to have scientific evidence for something which was not natural. Since "natural" is defined as what is describable by science, this is a priori impossible. This doesn't make naturalism irrefutable. It only makes it a matter of philosophical, and not scientific, concern.

Earlier I said science is the process of expanding shared capabilities and knowledge. I meant knowledge of historical facts and causal relationships, and not knowledge simpliciter. In addition to scientific knowledge, we have knowledge of the rules of discourse. This sort of knowledge logically (if not historically) precedes scientific discovery.

As a philosophical position, naturalism is a matter of analytic truths, and not synthetic ones. It's a way of defining boundaries in our language. As I see it, naturalism is true by definition. However, if the definition were shown to be incoherent or inconsistent with other analytic truths, then naturalism would be open to doubt.

The definition of naturalism I am proposing is not arbitrary, and it is not defined to privilege a particular side in current debates. Rather, it is based on the history of 20th century philosophy. There's a clear tradition in 20th century philosophy to regard naturalism as a view about science and its limits. Naturalistic dualists, for example, regard mental entities either as fundamentally unlike the entities currently postulated by the sciences (a la Sellars), or as fundamentally unlike any other entities any science could ever postulate (a la Chalmers)--yet, at the same time, they stipulate that mental entities are theoretically describable via some as-yet-unknown scientific methodology. They define the mind as within the boundaries of scientific discovery, and that is why they call themselves naturalists. Other naturalists (e.g., Dewey, Quine, Dennett, etc.) reject dualism but still identify naturalism with a devotion to science. So, while there may be a traditional set of questions that naturalists have tended to be concerned with, I don't think that is reason to question the defining attribute of naturalism as a view about science--specifically, a rejection of the limits on science which supernaturalists attempt to impose.

There are many attempts to distinguish varieties of naturalism, such as epistemological vs. ontological naturalism, and metaphysical vs. methodological naturalism: I think the best way to approach such discussions is to focus on what naturalism is primarily about, and that is the philosophy of science.

I don't see an important distinction between metaphysical and methodological naturalism. It seems to me that those who accept methodological naturalism but not metaphysical naturalism are in a very peculiar position: They want to protect religious belief from scientific scrutiny, but are at the same time denying their ability to make that religious belief relevant. Either that, or they want to limit methodological naturalism (and thus science itself) to only a subset of what can be discovered. In that case, they are saying that methodological naturalism is not a philosophically grounded position at all, and so shouldn't be given any weight in arguments about the supernatural. So, frankly, I don't trust any attempt to distinguish metaphysical from methodological naturalism.

Also, I don't see much point in trying to distinguish between ontological and epistemological naturalism. It seems to be more a semantic difference, though maybe I'm missing some of the philosophical subtlety there.