Tuesday, November 03, 2009

Supernaturalism and falsifiability

A redated post.

It is sometimes said that supernatural claims are unfalsifiable, while naturalistic ones are falsifiable. I don't see how this argument can be defended. First of all, we need to have a clear idea of what it is that makes a thesis supernatural. What is the supernatural?

But suppose we have such an idea. What does it take to falsify something? If what we mean is that the evidence logically entails the falsity of the thesis. That's a standard that would make every claim unfalsifiable. It's always possible to "fix" a disconfirmed theory to fit the evidence.

Falsification occurs, according to another definition, when all well-informed adherents of the thesis admit that the thesis is false. That is more plausible. But it seems to me that we could reach a situation where all well-informed defenders of a claim give up, whether the thesis is supernatural or natural.

It may be the case the well-informed people may find it harder to give up when religious beliefs are at stake. But how does supernaturalism render claims unfalsifiable?

38 comments:

Anonymous said...

Maybe it would be helpful if you defined precisely what "supernatural" means? Personally, I'm leaning to the view that terms like supernatural and natural are more harmful than helpful in trying to gain an accurate (truthful) understanding of reality.

Victor Reppert said...

Yes. I want those terms translated into English.

Anonymous said...

(NBAA)

Since I'm the one who (accidentally) started off this whole thread on falsifiability, I should clarify my position. I don't know if _every_ supernatural claim is unfalsifiable in principle, but it does seem to me that theism is unfalsifiable (although I am prepared to be convinced that I am wrong).

As I mentioned previously, a theory is falsified if a potentially disconfirming observation is made, and (most or all of) the proponents of the theory agree that it has been falsified.

In the case of theism, therefore, the standards of evidence are those set by theists, not by naturalists. However, the disconfirming evidence cannot involve divine intervention of any kind, since otherwise it would be self-refuting. (If God were to appear and declare "Theism is false!", this would confirm theism, not falsify it.)

I personally do not know of any evidence of this type (i.e. "naturalistic" evidence) that would convince the vast majority of theists that theism is false.

I am happy to accept that there is potentially disconfirming evidence for particular brands of theism (VR cited Pilate's court records as possible evidence against Christianity, for example), but even in cases like these I suspect that the adherents would not, as a group, renounce their beliefs. The Christians I know personally would almost certainly insist that Pilate's records were forged or perjured rather than renounce their Christianity.

So I ask again, what possible evidence could there be that would convince the majority of the world's theists (Christians, Jews, Muslims, Hindus, Buddhists and animists alike) that theism in general is false?

PS: I hope no one is going to cavil that Buddhists are not theists because "pure" Buddhism makes no reference to God or gods. Most Buddhists do believe in minor deities, and they pray as much as anybody else.

PPS: I have started putting my signature NBAA (nobrainsatall) at the top of my posts, so that I can be distinguished from the other ANONs out there.

Edwardtbabinski said...

VIC: Falsification occurs, according to another definition, when all well-informed adherents of the thesis admit that the thesis is false.

ED: That may be true of logical agrumentation dealing with a particularly narrow and limited subject, especially one about which all agree there is no major mystery.

The difficulty arises when you consider the inherent flexibility of philosophical and theological speculation in a broader sense.

Would you agree that there is more than one coherent philosophical view of the world, and more than one coherent Christian theological viewpoint on a host of matters pertaining to the Bible?

Would you also agree that the major schools of philosophy have been with us a very long time, perhaps even back as far as the pre-Socratics and that well-informed adherents continue to exist who support each major view?

The philosopher Simon Blackburn recently reviewed Frankfurt's little book, "On Truth," and had this to say that I find relevant:

Samuel Johnson who could write so prolifically thought that he was making a sound philosophical point by kicking a stone and declaiming: "Thus I refute Berkeley!" G.E. Moore got the same effect by holding up his hand and declaring that he was more certain of his hand than of any philosophical argument that purported to cast doubt on it -- which allegedly prompted Wittgenstein to remark that Berkeley had never intended to deny that his underpants continued to exist after he had put his trousers on over them.

Similarly, I do not think there was any recent philosophical movement that could have been stopped in its tracks by pointing out that it is easier to find your way about in daylight than in the dark, or that if someone tells you that a bottle contains gin and you act accordingly, you have a beef against him if it contains kerosene. Individuals, on occasion, might have carelessly let loose remarks that seem to imply the opposite, but I think these were rhetorical lapses rather than central doctrines to which they were wedded, and they probably misspoke themselves as they tried to say something more interesting. Easy triumphalism is always to be avoided in philosophy.

The doctrines of individual postmodernists may have varied, but the epicenter was surely the idea, as old as Protagoras, that man is the measure of all things. This means that there is no single way the world requires to be understood, and no process of science or history that is free from the engagement of the prepared mind. Even if the world were an open book -- which it is not -- books require reading; and reading requires intelligence; and intelligence requires language, learning, acculturation, tradition, and their innumerable influences on our resulting nature. Postmodernism recognizes that many of these influences can be expected to vary from person to person, place to place, time to time, class to class, gender to gender, tradition to tradition. Books require reading, and what they can then expect to get are readings, in the plural. What postmodernism then does with this thought is itself open to interpretation. Some may conclude that "anything goes," but that is unwarranted and untenable. Others may merely counsel increased awareness and toleration, a message with which it is hard to disagree, as is the message -- in the light of what William James called the trail of the human serpent everywhere -- that dogmatism, and the bigotry of certainty, might be soft-pedaled a little.

Why is it not true that anything goes? The answer is that among the cacophony of different readings there may be some that are better tuned to the way the world is or better adapted for finding that way. To take a fairly clear example, geologists derive their knowledge of the age of things from a principle of uniformity -- that where a feature is explicable by processes that happen and take time now, then it is right to suppose that it happened in the same way and took a similar amount of time then. A creationist may contest this, but the creationist will rely on the same principle -- to judge the age of the white-haired gentleman in front of him, or to date his holy texts, or to say how long ago was the landing on Plymouth Rock. In rejecting the geology, he is making exceptions from rules that he generally accepts. And this cannot in turn be a habit adapted to finding the way of the world, since such habits will lead to inconsistent results -- the holy text that for you trumps scientific method will likely disagree with the sacred oracle that for another person does so. Reliance on uniformities, without arbitrary exceptions, emerges as the only rational option.

Scientific method is one thing, and in spite of this fairly simple example it is a very complex one. But it gives us a field where we can hope for rational convergence. The pressure of experience is itself enough to flatten out any differences in the variously prepared minds that come to the problems of science. The political importance of all this may be more evident when we consider matters with more human content: not so much Berkeley's underwear, or the gin or kerosene in the bottle, but religious, political, moral, and historical interpretations of who we are, where we have been, and where to go. Here convergence is less certain.

Anonymous said...

Ed, you are completely irrational.

Victor Reppert said...

Many theists have maintained, for example, that apparently gratuitous evil counts against theism, just not enough to overthrow belief based on the total evidence. Of course Loftus is right, theists normally expect the final truth to be revealed when they die, so they may not in fact set their beliefs aside in the face of premortem counterevidence. I can certainly imagine postmortem disconfirmation of the idea of a good God: I get popped into hell and find Peter, James, John, C. S. Lewis, my late father and mother, and Jesus himself roasting there with me, while Adolf Hitler, Jeffrey Dahmer, and Bertrand Russell are up in heaven laughing at me.

Christians do believe, however, that they stand in a personal relation to God; it's a little like a marital relationship. You don't want to start accusing your partner of having an affair unless you really have good reason. And there is the personal experience of Christ's presence which has to be included as well (although pace the Mormon Church and William Lane Craig, I oppose using it as an all-purpose trump card). Lewis's On Obstinacy of Belief is helpful here.

Does any evidence, in my book, count against theism? Sure it does. I wish it were clearer than it is that God exists, and I wish there were less suffering in the world (unless I'm in a really bad mood). I also admit that it's easier for an atheist to imagine a convincing scenario than it is for me to imagine an unconvincing scenario. I think a certain amount of "obstinacy of belief" is rational for anyone who is committed to a world-view. World-view change is something like a paradigm shift in science. Usually it takes a lot of stuff to move someone who has carefully thought about it and has adopted one world-view or another. I don't think that because atheists have an easier time constructing disconfirming scenarios that that is any reason for saying that atheism is falsifiable in a way that theism is not.

Anonymous said...

Wait, I thought Dahmer got saved in prison before they killed him.

Victor Reppert said...

That's right about Dahmer. But I wouldn't expect his reward to be the privilege of laughing at me.

Anonymous said...

(NBAA)

VR wrote: "I don't think that because atheists have an easier time constructing disconfirming scenarios that that is any reason for saying that atheism is falsifiable in a way that theism is not."

Just one scenario will do.

Naturalism may or may not be falsifiable: it depends on the attitude naturalists would take in the face of a potentially disconfirming observation. (My belief is that naturalists as group _would_ reject naturalism in the face of hard evidence for the supernatural. You of course are free to believe otherwise.)

But unless you or someone else can come up with a potentially disconfirming observation for theism, theism is quite simply unfalsifiable.

Victor Reppert said...

No, NBAA, I think that is a mistake. I don't have to be able to envision the scenario in order for my belief to be falsifiable. Is an object unbreakable because I don't know how to break it? Besides, I just gave a scenario of me in hell being laughed at by various miscreants and atheist philosophers and said that would falsify it. Why is that not good enough?

Most real-life world-view changes involve collapse of confidence in several evidential areas. Ask Loftus if you doubt me. The fact that an atheist can specify some dramatic scenario that would prove his belief wrong isn't what makes his belief falsifiable.

Anonymous said...

(NBAA)

VR wrote: "Is an object unbreakable because I don't know how to break it?"

If neither you nor anybody else knows how to break it, then yes, it is unbreakable. This is what "unbreakable" means. Do you have another definition perhaps? (Maybe you regard an object as unbreakable only if it is _logically_ impossible to break it. In which case there are no unbreakable objects in the real world, and the word should appear only in fiction and philosophy.)

VR wrote: "Besides, I just gave a scenario of me in hell being laughed at by various miscreants and atheist philosophers and said that would falsify it. Why is that not good enough?"

Once again, your scenario would be evidence against Christian theism (or, as you say, a "good God"), not theism in general. All it would prove is that God prefers miscreants and atheist philosophers over Christians like you.

I have already stated my willingness to accept that Christian theism is falsifiable (God appearing and affirming that He is a Buddhist is probably the simplest scenario). The question is, and always has been, is theism in general falsifiable?

VR wrote: "The fact that an atheist can specify some dramatic scenario that would prove his belief wrong isn't what makes his belief falsifiable."

I have spelt out the definition of falsifiability (as used in the Duhem-Quine thesis) three or four times now, and "specify[ing] some dramatic scenario that would prove his belief wrong" is precisely what makes his belief falsifiable (provided that he is sincere when he says that the "dramatic scenario" would be falsifying).

What part of the definition do you think this situation doesn't satisfy?

Victor Reppert said...

I was thinking of the classical definition of theism in which God is supposed to be omnipotent, omniscient, and perfectly good. That would be refuted by any of several scenarios. And I am not talking about Christian theism either. On my view an Cartesian Evil Demon would not be God.

What I think you are asking is what it would take to confirm the claim that there are no intelligent causes underlying the universe. And I've got the problem that if that were really true I wouldn't be able to trust the reasoning that got me there. But if we set that difficulty aside, I suppose having a gapless physical cause trace on everything in the universe would do the trick. If science progressed to the point where we could all be LaPlacean demons who could predict every occurrence before it happened based on physics, then I would probably have to say that I would have to reject all nonphysical causes as superfluous.

Anonymous said...

"I was thinking of the classical definition of theism in which God is supposed to be omnipotent, omniscient, and perfectly good."

Then I don't see why your example would disprove the existence of such a God. Why should the fact that you end up in Hell falsify the existence of such a God? I'm sure some theistic apologist could show that it would not be logically impossible for this perfectly good God to send you to Hell and pick others whom you hate to go to Heaven.
Sam

Victor Reppert said...

All I was trying to do was establish falsifiability. So if you try to argue that it shouldn't take that much to convince me, you are missing my point. I'm not proposing a god of the gaps, I was proposing a no god of no gaps. But I admit that some of the gaps seem rather conceptual, such as the problem of the contingency of the universe. There are some conceptual issues for an adequate naturalism that would have to be resolved to be greater satisfaction.

Remember the Parsons star-spelling scenario is logically compatible with naturalism. It could be an advanced technology, it could be an accident, it could be all sorts of things. So if I have to come up with a scenario that is logically incompatible with theism, I'm out of luck, and so are you.

Victor Reppert said...

If that's the case, John, then why don't you agree with me (and, by the way, Keith Parsons in God and the Burden of Proof) that we should get out of the business of trying to pin irrationality charges on our opponents? I think there are some religious philosophies that are irrational, but for the most part everyone is rational but not perfectly rational. I try to offer what to me are the good reason for being a believer, you for being an unbeliever. World-view commitments are like marriages. You go from one to another with an understandable degree of reluctance.

We'd all like to have that argument that takes the opponent by the scruff of the neck and forces them to change or else be irrrational. It's just not out there. It's not the argument from reason, it's not the presuppositionalist's TAG, it isn't any version of the argument from evil. You can't infer from the fact I have trouble creating a scenario where I would give up my belief that I would hang onto my beliefs no matter what. I seriously doubt that you were thinking, while you were a believer, gosh, if such and such were to happen, and I were to re-evaluate this argument that way, I'd have to become a skeptic.

I don't think philosophy is about proclaiming that you have such a strong argument that the idiots on the other side have to give up. I think I can give good positive reasons for my beliefs and have arguments for why some critical objections are not what they're cracked up to be.

One of my reasons for rejecting the inflated claims made on behalf of the argument from evil is that I have studied many philosophical arguments in the course of a lifetime of study and I have found none or next to none that achieve that degree of success. (Butler's argument against psychological egoism is, I think, the closest.)

Yes yes I know. the Bible says (in Romans) that the reality of God is clearly seen in the things that are made and that the unbeliever is without excuse. I just don't know of any arguments that are up to that daunting chore.

Victor Reppert said...

It's best not to mislead people with rhetoric. I know, when you are passionate, it gets tempting. I think Lewis falls prey to this mistake at times (calling atheism a "boy's philosophy" is a good example).

I'd have to look through your posts to see passages which sound like irrationality charges. But I'm pretty sure I won't come up empty.

In fact some people on "my side" have been disappointed that I don't make stronger claims for my arguments than I do.

Anonymous said...

(NBAA)

VR wrote: "If science progressed to the point where we could all be LaPlacean demons who could predict every occurrence before it happened based on physics, then I would probably have to say that I would have to reject all nonphysical causes as superfluous."

As JWL pointed out, we are closer to this goal than we were 500 years ago. But it is unlikely that we will ever reach it. In fact, many people believe that quantum indeterminism makes such a goal impossible.

So you are suggesting a potentially disconfirming scenario that modern science regards as very unlikely or impossible. (And it certainly won't happen in our lifetimes.) Furthermore, I seriously doubt that the vast majority of the world's theists would reject theism just because scientists could (in principle) predict everything. Would theists reject theism if scientists were able to create artificial life in the laboratory, or create a whole mini-universe in a particle accelerator? I don't think so.

Now, you might respond by saying that the kind of scenarios that naturalists postulate when arguing that naturalism is falsifiable (such as the appearance of God or star-spelling) are just as impossible according to the current beliefs of theists. This may well be true. Theists seem to believe all sorts of odd things. But I suspect that any claim by theists that God doesn't intervene in the world in a testable way would steer perilously close to deism.

Anonymous said...

(NBAA)

VR wrote: "Remember the Parsons star-spelling scenario is logically compatible with naturalism. It could be an advanced technology, it could be an accident, it could be all sorts of things. So if I have to come up with a scenario that is logically incompatible with theism, I'm out of luck, and so are you."

Once again you (like Parsons before you) seem to misunderstand the notion of falsifiability. Falsifiability (in a scientific context) does not mean _logical_ falsifiability. Sure, naturalists faced with the star-spelling scenario would look very hard at the evidence before rejecting naturalism. They would check to see if any of the stars had travelled faster than the speed of light to get into position, and to see if the conservation of energy or momentum had been violated. They would scan that region of the sky for any unusual radio emissions, or strong electric or magnetic fields. They would calculate the probability that the star-spelling formation was just a conicidence.

And if the results of these observations and calculations went the wrong way (for naturalism), naturalists as a group would reluctantly conclude that naturalism was false. So naturalism would be falsified. End of story. I stress again that the standards of evidence used to evaluate naturalism are those set by naturalists (just as those for theism are set by theists), not by idealist philosophers.

I don't know how many times I need to repeat that falsification of a theory means a potentially disconfirming observation, plus agreement from the theory's supporters that it has been falsified. It does _not_ mean logical falsification. But just in case my position is unclear, I will expand again:

You are probably aware that Karl Popper introduced the notion of falsifiability into general currency in an attempt to solve the problem of scientific induction. In his first stab at the problem, Popper claimed that a contingent universal statement (that is, a statement of the form "All X are Y", where Y is not part of the definition of X, nor logically connected to X in any way) is scientifically meaningful only if it is falsifiable.

The Duhem-Quine thesis was then raised as an objection to this claim. Because it is logically possible, according to the DQ thesis, for the proponent of a theory to refuse to recognize that the theory has been falsified in the face of any number of potentially disconfirming observations, Popper's falsifiability condition has no value as an _a priori_ criterion of meaning.

Despite the consequent failure of Popper's program of "naive falsification", it is still clear that Popper (and the Duhem-Quine thesis) did not equate falsifiability with _logical_ falsifiability. A contingent statement, by definition, cannot be logically falsified. If you seriously think that Popper was foolish enough to claim that a contingent statement is meaningful only if it can be logically falsified, and that the Duhem-Quine thesis does nothing more than make the obvious and trivial observation that contingent statements can never be logically falsified, then you are flogging a very limp straw man.

Anonymous said...

(NBAA)

Sam wrote: "I'm sure some theistic apologist could show that it would not be logically impossible for this perfectly good God to send you to Hell and pick others whom you hate to go to Heaven."

I love your work.

Jason Pratt said...

{{We'd all like to have that argument that takes the opponent by the scruff of the neck and forces them to change or else be irrrational.}}

Personally, I would consider such an argument to be a dark enchantment mentally equivalent to a forced seduction or even a rape attempt. So even if I was presented with such a thing, I would reject it on moral grounds. And I'm someone who _does_ use deductive argumentation (and thinks more highly of it than Victor does on this topic, including in the AfR.)


{{In fact some people on "my side" have been disappointed that I don't make stronger claims for my arguments than I do.}}

For instance. {g} Although to be more precise, I do agree that Victor is proper and correct not to present his case in his book as a deductive argument. I even said as much in my amazon review.


While I understand NBAA's notion of falsification, I recognize naturalism to be a _philosophical_ position first (not a process of inquiry and inference to conclusion); and thus (in theory) open to falsification by propositional logic. I don't recognize naturalism to be legitimately established by the process of investigation of events within nature, but even if I did it would still be a manifestly _logical_ establishment (specifically of an expectation by induction.)

Furthermore, I don't consider falsification of a theory to require agreement from the theory's supporters that it has been falsified. I _DO_ agree, however, that it would be illogical (among other things, all in favor of the supporters) for an opponent to insist that a supporter consider the position to be falsified when (by reverse tautology) the supporter himself does not accept the falsification.


That being said, philosophical naturalism will always be in a potentially vulnerable position (unless deductively established or dogmatically asserted a priori) for disconfirmation, by the nature (pun half-intended {g}) of the case. Insofar as supernaturalism involves the introduction of behavior into one system from beyond that system, disconfirmational evidence might arrive any time; might have already arrived; might arrive and not be noticed as such; and the results would be analyzable (at least in principle if not in practice) from within the system. Supernaturalism is not in a symmetrically parallel position: an analyst of a system studying behaviors of that system, is not in a position to _thereby_ rule out the existence of undetected influence from outside the system in system behaviors. (Except perhaps in unrealistically simplified examples.)

Supernaturalism _is_ (at least potentially) symmetrically parallel to naturalism in regard to falsification by propositional analysis, though.


As usual, I will note that a discussion of naturalism/supernaturalism is _NOT_ (in principle, and should not be in practice) the same as a discussion of atheism/theism. The categories are distinct and should not be confuted. (Which is why there can be particle- and astrophysicists who infer the existence of a more fundamental reality not identifiable with our natural universe and upon which the natural universe depends for its existence, yet not be theists in any way. This is aside from the question of whether they are logically correct to draw such an inference about the existence of what in metaphysical language would be logically identifiable as a 'supernature'.)

Jason Pratt

Anonymous said...

NBAA wrote:
"I love your work."

Thanks. The feeling is mutual. I really appreciate the clarity of thought you bring to the discussion here.
Sam

Anonymous said...

(NBAA)

JP wrote: "Furthermore, I don't consider falsification of a theory to require agreement from the theory's supporters that it has been falsified. I _DO_ agree, however, that it would be illogical (among other things, all in favor of the supporters) for an opponent to insist that a supporter consider the position to be falsified when (by reverse tautology) the supporter himself does not accept the falsification."

I am not entirely sure what you are saying here. Do you mean that it is possible for a theory to be falsified for its opponents, but not falsified for its supporters? If so, I agree that that this is a plausible and entirely natural way of reading falsification. There are certainly many theories that fit this description: Young Earth Creationism is a prime example.

However, within the context of the debates over Popper's program of naive falsification, "falsifiability" and "falsification" have peculiar technical definitions which are not consistent with this reading. And it is almost invariably these definitions that people assume when they ask "Is your theory falsifiable?" or invoke the Duhem-Quine thesis to argue against falsifiability.

JP wrote: "That being said, philosophical naturalism will always be in a potentially vulnerable position (unless deductively established or dogmatically asserted a priori) for disconfirmation..."

Here we come to a point that this discussion has probably been avoiding too long. What exactly is naturalism? My understanding is that there are two standard forms of naturalism:

(i) weak (or methodological) naturalism: the world around us is governed by certain testable regularities (which we sometimes call scientific laws) which are never experimentally violated

(ii) strong (or philosophical) naturalism: there exist no entities or forces which are (even in principle) capable of violating the scientific laws postulated by weak naturalism

My personal position is one of weak naturalism, and this is what I normally mean when I write "naturalism" alone. I am agnostic about strong naturalism, mainly because I can't think of any way to distinguish experimentally between strong naturalism (no law-violating entities exist) and bare deism (such entities do exist, but they choose not to violate the laws).

However, any evidence that falsifies weak naturalism will also falsify strong naturalism. And if strong naturalism is true then weak naturalism is true.

Strong naturalism has not been and will never be _deductively_ established. It is simply a position that is temptingly suggested by 500 years of the successful application of weak naturalism.

JP wrote: "Supernaturalism _is_ (at least potentially) symmetrically parallel to naturalism in regard to falsification by propositional analysis, though."

I disagree. Naturalism and supernaturalism are contingent statements about the way the world works. (Here I mean weak naturalism and its negation.) They can neither be confirmed nor falsified by a propositional analysis. Sure, a supporter of either theory could be trapped (rhetorically) into making logically inconsistent statements, but this would only prove that the supporter was confused, not that the theory was false.

JP wrote: "As usual, I will note that a discussion of naturalism/supernaturalism is _NOT_ (in principle, and should not be in practice) the same as a discussion of atheism/theism."

I agree with this statement entirely. The present thread started off with an exchange of the form "Is theism falsifiable?", "Well, what about naturalism, is it falsifiable?" without either party assuming that one was the negation of the other.

JP wrote: "Which is why there can be particle- and astrophysicists who infer the existence of a more fundamental reality not identifiable with our natural universe and upon which the natural universe depends for its existence"

I'm a bit confused here. Could you expand a bit. What "more fundamental reality" are you referring to? The Platonic realm of mathematics perhaps? Or maybe string theory?

Victor Reppert said...

I think there is some misunderstanding here. I don't think of falsification in the logical sense. Here's the context of my discussion of logical falsifiability:

VR: "I was thinking of the classical definition of theism in which God is supposed to be omnipotent, omniscient, and perfectly good."

Sam: Then I don't see why your example would disprove the existence of such a God. Why should the fact that you end up in Hell falsify the existence of such a God? I'm sure some theistic apologist could show that it would not be logically impossible for this perfectly good God to send you to Hell and pick others whom you hate to go to Heaven.

VR: Of course it would be logically possible. That's not the point. I think that there are evidential scenarios that would lead most theists to give up. I don't think of falsification as requiring logical falsification. But what you can't do when I offer a scenario is say that there is some way theists could logically get out of giving up. I think that certain developments in thought would result in most theists giving up.

NBAA: So you are suggesting a potentially disconfirming scenario that modern science regards as very unlikely or impossible. (And it certainly won't happen in our lifetimes.) Furthermore, I seriously doubt that the vast majority of the world's theists would reject theism just because scientists could (in principle) predict everything. Would theists reject theism if scientists were able to create artificial life in the laboratory, or create a whole mini-universe in a particle accelerator? I don't think so.

Now, you might respond by saying that the kind of scenarios that naturalists postulate when arguing that naturalism is falsifiable (such as the appearance of God or star-spelling) are just as impossible according to the current beliefs of theists. This may well be true. Theists seem to believe all sorts of odd things. But I suspect that any claim by theists that God doesn't intervene in the world in a testable way would steer perilously close to deism.

VR: It seems to be unlikely to me for theological reasons. Not that I don't think there couldn't be miracles today: I not only believe that there could be miracles today, I think there are miracles today. But show-off miracles are out of character for the Christian God.

And I'm not sure that there wouldn't be skeptics still in place if the star-spelling scenario were to occur. Not every atheist philosopher agrees with Parsons on this. Kai Nielsen, if I am not mistaken, has indicated in response to this kind of scenario that he would remain an atheist. The atheist debater Doug Jesseph in his debate with Bill Craig says that he would give up atheism if the lectern he uses is turned to purple smoke. Of course, if that were to happen I, a theist, would suspect David Copperfield, not God, as Loftus has indicated.

To be honest with you, I don't see what you could be getting at with the argument that theism isn't falsifiable. It seems to be a matter of envisioning some scenario and then asking whether there would be a consensus of adherents who give up. This seems to be a contingent fact depending on who is in the group of informed theists or atheists. I don't, for example, believe that you can draw a science-pseudo-science distinction based on falsifiability thus defined. I think that the fact that it is perhaps easier for naturalists to envision a "falsifying" scenario has to do with some features of the relative positions, which have nothing to do with the rationality of those beliefs. Therefore nothing interesting follows from the fact that theism isn't falsifiable in the sense you are describing.

Complaining that what I described is not likely to happen in our lifetimes doesn't mean much of anything. I think the only way you could be convinced that theism was falsifiable would be if you could actually falsify it.

Anonymous said...

VR wrote: "I think there is some misunderstanding here. I don't think of falsification in the logical sense."

My comment was very clearly marked as a response to your statement: "Remember the Parsons star-spelling scenario is logically compatible with naturalism. It could be an advanced technology, it could be an accident, it could be all sorts of things. So if I have to come up with a scenario that is logically incompatible with theism, I'm out of luck, and so are you."

If you were not thinking of falsification in the logical sense, then why did you insist here that falsifying scenarios need to be "logically incompatible" with their theories?

VR wrote: "It seems to be a matter of envisioning some scenario and then asking whether there would be a consensus of adherents who give up. This seems to be a contingent fact depending on who is in the group of informed theists or atheists."

Obviously it is a contingent fact. And we know who the adherents are in both cases: the 5 billion or more theists in the world on one hand, and the 60% or more of scientists, plus assorted atheist or agnostist philosophers, who believe in naturalism on the other. It is simply a matter of one or the other of us positing a scenario, and then thrashing out whether (i) the scenario is feasible and (ii) the adherents would really give up. This is the way contingent claims are always evaluated. Is there any problem with this?

VR wrote: "Complaining that what I described is not likely to happen in our lifetimes doesn't mean much of anything."

I beg to differ. Naturalism could be falsified tomorrow (or more realistically in a year or two, once all the tests are done). If your LaPlacean demons scenario is the standard of falsification for theism, then theism will not be falsified in our lifetimes, or in a hundred lifetimes. So you can go to bed tonight safe in the knowledge that you will not live to see theism falsified. I cannot say the same thing about naturalism. The evidential bases of the two theories are not equivalent.

VR wrote: "I think the only way you could be convinced that theism was falsifiable would be if you could actually falsify it."

I repeat: if you want to claim that theism is falsifiable, I am under no obligation to take you seriously unless you tell me how it could be falsified (and preferably in our lifetimes).

Jason Pratt said...

NBAA (and afterward): {{I am not entirely sure what you are saying here. Do you mean that it is possible for a theory to be falsified for its opponents, but not falsified for its supporters?}}

Yes. Which, as I pointed out, is practically a tautology--for unless the opponents (or proponents?) are merely being contentious (for whatever reason) then obviously the opponents will be opponents because they think the theory has been falsified somehow. By reverse tautology, the proponents are going to think the theory hasn't been falsified.

There is also the question of establishment, of course. An opponent might agree a theory hasn't been falsified yet, but still be an opponent because he disagrees that the hypothesis has been sufficiently established by induction. (Or alternately that the proposed point hasn't been sufficiently concluded by deduction.)

I'm trying to put all this as neutrally as I can, btw. And I do recognize that within organizations it would be irresponsible for _the organization_ per se to announce establishment or falsification without some kind of quorum among the organization. But I also consider organizations to be an abstract (though handy) way of describing a particular kind of personal interaction among a particular set of persons; and ultimately the individual is the one who has responsibility to believe or disbelieve. A person could responsibly accept the judgment of his peers when he himself happens to be unable to see the conclusion or weigh the liklihoods himself; but... hm... I'm not sure how to put this without sounding like I'm making an accusation... {pondering}

I would think it wrong for me to set up a notion of falsification that shifted the burden of responsibility off myself and onto a group. Granted, there is such a thing as personal loyalty, and this shouldn't be sneezed away; also granted, not everyone can be good at everything (which is one reason why we have organizations {s!}). But there is also such a thing as hiding behind the group when faced with apparent falsification threats. I think it would be wrong for me, when faced with a criticism I can't meet, to say, 'well... uh... but _technically_ it can't be falsified until all these other people agree it is, too!'

That's one sort of thing I'm trying to guard against in considering principles of falsification. Another sort for me to guard against (and which I would be specially tempted toward, as it happens), would be the abuse of personal resposibility in order to claim falsification on my own authority and insisting the opponent accept my authority _thus_ to accept the falsification. If a person honestly thinks something is wrong, and seems to clearly see why; then even if _I_ can see where the mistake is being made (I think), I wouldn't consider it right to insist the other person simply accept my rebuttal instead.


{{However, within the context of the debates over Popper's program of naive falsification, "falsifiability" and "falsification" have peculiar technical definitions which are not consistent with this reading.}}

Entirely granted, and not disputed.

Furthermore, I can see why you would think Victor was discussing falsification in the technical fashion peculiar to modern scientific induction theory (or the philosophy thereof.) Or, maybe _not_ thinking about it in those terms and thinking about it in the logical sense instead. Or...

Heck, _I'm_ honestly more than a little confused about what kind of falsification he was talking about! {wry g!} (And moreso even after his comment subsequent to my previous post.)

As a practical matter, though, as well as a technical one, I don't consider scientific induction to be the basis of all logical analysis; consequently, my notion of falsification isn't dependent (so far as I can tell--and you seem to agree {g!}) on the type of falsification principle technically peculiar to scientific induction.

Rather than defend Victor, then (since I was, and am, confused about what kind of falsification he's appealing to and thus asking about), I figured it might be better (as well as more practical) to stake a position myself and go from there.

While I'm at it, I will also state that I, at least, would not consider the "fixing" of a theory to avoid disconfirmation due to evidence, to be the same as rendering all theories unfalsifiable. On the contrary, 'fixing' a disconfirmed theory to fit the evidence clearly involves the creation of a new theory altered from the previously falsified one. The first one is technically toast. {g}



{{Here we come to a point that this discussion has probably been avoiding too long.}}

I agree. And I would have preferred to see a discussion on this before going to the question of falsification. But since the discussion was already going... {shrug}. Besides, most (or all) of what I wrote would fit into (ii), which only has to do with the denial of anything 'outside' the system introducing effects 'inside' the system. But which I _don't_ consider to be robust philosophical naturalism in category distinction to robust philosophical supernaturalism. i.e. the (ii) kind of naturalism isn't what I find myself fundamentally up against in philosophical disputation.

{{What exactly is naturalism?}}

Methodological naturalism (what you listed as (i) ) I understand (and I recall Victor thinking the same way) to be more-or-less ideologically neutral in regard to the kind of debates we're usually having on this forum. Nature goes through certain kinds of reactive and counterreactive behaviors, which can (in principle, and frequently in practice) be discovered by means of observation, inference from observational data, and testing. Where limits to the observations and inferences exist, we can at least discover the limitations and infer to an understanding of why there cannot be further observations along that line. (Quantum indeterminacy being a paradigmatic example. The limits of observation in an expanding universe I suppose to be another example.)

The 'philosophical' naturalism you mention, however, even though philosophical in its proposition, is not the philosophical naturalism I routinely deal with as a metaphyscian--although it _is_ a proper corollary to it. Not wrong, therefore, but not fundamental enough (not _strong_ enough even) for the purpose of discussing and debating supernaturalism. (It also uses the loaded term 'violate', which I disrecommend.)

To take a popular example of the limitations of this definition vis-a-vis supernaturalism: it has long been a staple of sci-fi/fantasy for parallel worlds to exist and (to take perhaps the original example in modern fiction) for the Flash of Earth 1 to travel to Earth 2 and meet his DC Golden Age counterpart, and vice versa. Rigorously speaking, though, for purposes of debating the existence of the kind of supernaturalism typically being discussed by Christians (and other supernaturalistic theists, including on this site), Earth 2 and Earth 1 should not be considered supernatural to each other. Earth 2 does not depend for its existence upon the distinct (and distinctly different) existence of Earth 1, nor vice versa.

Your definition (ii), therefore, while not bad, doesn't go far enough in being useful for the kind of discussions that deal with questions of _super_-natural (instead of extra-natural) existence and operation. The issue isn't _only_ whether there are entities or forces which are (even in principle) capable of introducing effects into the evident system of Nature (much less whether there are entities or forces which are capable in principle of _violating_ the behaviors of the evident system of Nature). The issue is whether the evident system of Nature depends for its existence upon a more fundamental reality distinct from Nature. (This is another reason why the violation language is not technically appropriate to the debate at hand; the operations of the natural system would be dependent upon the supernatural system.)

Philosophical supernaturalism of the sort being proposed by, for instance, Christian theists, states (or proposes or concludes or hypothesizes, etc.) that there is another reality more fundamental than the evident Natural system, and distinct from it in its own existence, upon which the Natural system (and indeed everything in reality, if anything else than Nature and this Supernature does exist) depends for existence.

Philosophical naturalism (of this 'robust' sort) states (or proposes, etc.) that the evident Natural system is itself the most fundamental reality upon which everything in reality depends. The (ii) you mentioned would be a logical corollary to this position.

(Both naturalism and supernaturalism would stand over against cosmological dualisms and infinite regressions.)


{{My personal position is one of weak naturalism, and this is what I normally mean when I write "naturalism" alone.}}

That's fine. And I don't see it under (actual) debate anywhere nearby. Supernaturalism, and the question of its falsification, is not topically concerned with (i) (since (i) can easily be true whether or not supernaturalism is true.)

Supernaturalism _is_ topically concerned with (ii), since they would be mutually exclusive of each other; but (ii) doesn't go far enough to address the question of falsification of supernaturalism per se. (Admittedly, _no kind of_ induction about the natural system based on inferences about the natural system drawn in reference to observations of the natural system, is going to 'falsify' supernaturalism. But those aren't the only kind of inferences available.)


{{I am agnostic about strong naturalism, mainly because I can't think of any way to distinguish experimentally between strong naturalism (no law-violating entities exist) and bare deism (such entities do exist, but they choose not to violate the laws). }}

That's very sensible, so far as it goes; although I reiterate that not all conclusions require experimentation. My belief that 2 oranges and 2 apples = 4 fruits does not depend on my never having caught them behaving differently. Granted, as a matter of history the principle may have been discovered while doing something inductive and akin to experimentation; it can still be taught that way, too. But it doesn't _have to be_ taught or learned that way. I'm pretty sure I didn't learn differential calculus that way. (Though maybe that explains why I'm not so great at it... {g!})

That being said, I agree (in passing) that the proposal of minimal deism is indistinguishable for all _practical_ purposes from philosophical naturalism (either the robust kind I've discussed, or (ii) as a corollary thereof.)


{{However, any evidence that falsifies weak naturalism will also falsify strong naturalism.}}

Not at all. Discovering that the system behaves in erratic ways (falsifying (i)) neither falsifies the claim that no tampering from outside the system can be happening, nor establishes such tampering is happening. It certainly doesn't falsify the claim of robust phil-nat, that the system is the one and only (and thus by default most fundamental) reality. At worst it only would be grounds for some kind of suspicion, and perhaps provide some data for drawing inferences about what might be tested or re-evaluated.

Please note that I am giving this rebuttal _in your favor_ (potentially).

{{And if strong naturalism is true then weak naturalism is true.}}

Not so--whether Nature is what I elsewhere call the Independent Fact or not, it might not operate self-consistently in all regards. If it doesn't it just doesn't, and we'd have to just put up with it (or pretend that it did anyway).

We would of course _like_ for reality to be self-consistent--because we can discern that logic is ideally self-consistent. From this notion has come all theoretical and practical science, whether Eastern or Western, since before the Common Era (which is only common historically by being Christian. {g})

But it's quite a theistic notion--even if venially so, so to speak. (Theistic, but not necessarily supernaturalistic. Not necessarily naturalistic either, though.)


{{I disagree [that supernaturalism is at least potentially symmetrically parallel to naturalism in regard to falsification by propositional analysis]. Naturalism and supernaturalism are contingent statements about the way the world works. (Here I mean weak naturalism and its negation.)}}

By supernaturalism, I do not mean the negation of the proposition that the natural system will always of its own accord behave coherently. (Where would the 'super'-naturalism be in that??)

Even weak naturalism could be falsified, though, by discovering the proposition necessarily entails logical inconsistency. (Whether or not someone can practically do this, is not something I am especially concerned about; but I can imagine someone ending up _having to_ make logically inconsistent statements for support of weak naturalism in the face of observations of apparent incoherence in natural behavior. The falsification might of course be overcome by an alteration to the notion of weak naturalism, which alteration might be minor and even easy to make; but the original notion would still be falsified.)

Fwiw, I have an expectation that weak naturalism (i) (or some closely related variant of it) will never be legitimately falsified; but that is because I expect it to obtain as true due to my belief in supernaturalistic (indeed orthodox Christian) theism: Nature will behave in an ultimately coherent fashion. If I ever found myself obligated to agree that weak naturalism (of the general sort you described) was falsified, I would need to go back and re-evaluate why I thought it obtained as a corollary from my Christian theism.


{{The present thread started off with an exchange of the form "Is theism falsifiable?", "Well, what about naturalism, is it falsifiable?" without either party assuming that one was the negation of the other.}}

Actually, the _present_ thread kicked off with John Loftus immediately switching categories to talk about belief and non-belief in God. {g} But in this case I wasn't critting a switch by anyone in particular--I just know from long experience (including on this site) that the categories tend to be elided between _very_ frequently, so I usually make a point of cautioning about it. Especially since I'm seeing the topics being jumped between pretty freely in this thread (and in related prior threads, too.)

{JP wrote: "Which is why there can be particle- and astrophysicists who infer the existence of a more fundamental reality not identifiable with our natural universe and upon which the natural universe depends for its existence"

NBAA: I'm a bit confused here. Could you expand a bit. What "more fundamental reality" are you referring to? The Platonic realm of mathematics perhaps? Or maybe string theory?


JP: I'm honestly not sure string theory counts; but I have certainly heard, in recent years, of some astrophysicists accepting a conclusion of Nature's contingent existence and so inferring a supernatural cause, i.e. an eternally existent super-nature where the collision of this-or-that has produced our own nature as a pocket reality. (In fact, I know I heard of this again in 2006. I would go dig up the reference, but the example is purely for sake of principle. It doesn't strictly matter whether anyone _has_ done it--they _could_ do it without necessary absurdity. And I have other things to do today. {g} If you really want to find the reference I'm specifically thinking of, though, I'm sure it has been discussed within the past year over at Christiancadre.blogspot.com.)

Similarly, I do hear (on the fringes as it were) about physicists inferring that quantum particles are popping 'into' our reality not out of non-existence but percolating in from a more fundamental kind of reality distinctly different than Nature as part of the ongoing 'creation' (and thus existence) of Nature, as it were. I don't recall these things being murmured in context of theism; and I don't see why they would have to be proposed in that context, either.


Platonic idealism, on the other hand, is not what I was thinking about; and I'm somewhat doubtful anyone is currently holding as a position that the realm of abstract mathematics is somehow generating and sustaining Nature. (Personally I would argue rather strenuously that this is impossible.) But that kind of position has been linked to supernaturalism before, in the history of Western thought. (And even more feasibly to theism of some kind.)


Jason Pratt

Anonymous said...

(NBAA)

JP, your last comment was very detailed and wide-ranging, and it would take a week to respond to it properly. I think the topic of falsification has been done to death now, so I will just make a few remarks about naturalism and supernaturalism.

NBAA: {{My personal position is one of weak naturalism, and this is what I normally mean when I write "naturalism" alone.}}

JP: "That's fine. And I don't see it under (actual) debate anywhere nearby."

It is true that my definitions of weak and strong naturalism are somewhat idiosyncratic, my excuse being that I have never seen proper (non-tautological) definitions of naturalism or supernaturalism laid out anywhere.

However, I disagree that this discussion is not about weak naturalism. I think that, for naturalists in general, theism is a direct challenge to weak naturalism. If an interventionist God exists then miracles occur and weak naturalism is violated. If no such violations occur, then the best that theists can hope for is some form of deism.

JP: "Even weak naturalism could be falsified, though, by discovering the proposition necessarily entails logical inconsistency."

Weak naturalism is a synthetic claim. No logical inconsistency (or logical proof) is possible.

JP: "...although I reiterate that not all conclusions require experimentation. My belief that 2 oranges and 2 apples = 4 fruits does not depend on my never having caught them behaving differently."

The statement "2 oranges and 2 apples = 4 fruits" is analytic, and does not rely on experimental evidence for its verification (although some commentators, such as Quine, have argued otherwise). Apart from analytic statements, pretty much everything else that we reliably believe is based on observation or indoctrination.

JP: "Philosophical supernaturalism of the sort being proposed by, for instance, Christian theists, states (or proposes or concludes or hypothesizes, etc.) that there is another reality more fundamental than the evident Natural system, and distinct from it in its own existence, upon which the Natural system (and indeed everything in reality, if anything else than Nature and this Supernature does exist) depends for existence."

Being a scientist by training (and temperament) I find this definition rather hard to digest. What does it mean, in practical or observational terms, that "there is another reality more fundamental than the evident Natural system"? And what does it mean for the Natural system to depend on the Supernatural for its existence?

I have no doubt that there are theists (like you) who have adopted this position, and I am happy for you to label it "robust supernaturalism". But I doubt if there are any naturalists who would recognize in it the negation of their own position.

As far as I can make out, it is just one of the many possible ways in which strong naturalism could be violated. (And assuming it as the default definition of supernaturalism is a bit like assuming that Christianity is the default definition of theism.)

NBAA: {{However, any evidence that falsifies weak naturalism will also falsify strong naturalism.}}

JP: "Not at all. Discovering that the system behaves in erratic ways (falsifying (i)) neither falsifies the claim that no tampering from outside the system can be happening, nor establishes such tampering is happening. It certainly doesn't falsify the claim of robust phil-nat, that the system is the one and only (and thus by default most fundamental) reality."

At no point in my definition of strong naturalism did I refer to tampering "from outside the system". There are many ways that strong naturalism could be violated "from within": for example, by witchcraft. The two definitions I gave, of weak and strong naturalism, were worded so that weak naturalism was a subset of strong naturalism _by definition_.

Your comment might be more apposite if directed at a "robust phil-nat" supporter. But do you know of anyone who is actually a robust philosophical naturalist?

JP wrote: "By supernaturalism, I do not mean the negation of the proposition that the natural system will always of its own accord behave coherently. (Where would the 'super'-naturalism be in that??)"

Where indeed? It is evident that your idea of the supernatural is quite different from my own. My understanding is that supernaturalism (of whatever flavor) is just the negation of naturalism (of the same flavor). You pointed out that "extra-naturalism" would be a better word for this, and I am inclined to agree.

However, in an earlier post you wrote: "As usual, I will note that a discussion of naturalism/supernaturalism is _NOT_ (in principle, and should not be in practice) the same as a discussion of atheism/theism." This I read as an implicit endorsement of supernaturalism as the negation of naturalism.

Jason Pratt said...

NBAA (and afterward): {{I think the topic of falsification has been done to death now, so I will just make a few remarks about naturalism and supernaturalism.}}

Fair enough.


{{NBAA: My personal position is one of weak naturalism, and this is what I normally mean when I write "naturalism" alone.

JP: "That's fine. And I don't see it under (actual) debate anywhere nearby."

NBAA: It is true that my definitions of weak and strong naturalism are somewhat idiosyncratic...}}


Sorry. I thought the (i) and (ii) definitions were actually pretty good for what they did (just that (ii) didn't go far enough to be in direct comparison with _super_-naturalism per se.) What I meant when I said that I didn't see weak naturalism under actual debate anywhere nearby, is that I didn't see anyone trying to dispute against it.

Apparently, though, you did. The difference in perception (and concept?) here should be instructive.

{{However, I disagree that this discussion is not about weak naturalism. I think that, for naturalists in general, theism is a direct challenge to weak naturalism. If an interventionist God exists then miracles occur and weak naturalism is violated.}}

(ii) would certainly be mutually exclusive to this. But how would (i) per se be violated?--I mean if (i) is supposed to be something other than a polysyllabic variation of (ii).


Going back to your definition for (i) weak (or methodological) naturalism: "the world around us is governed by certain testable regularities (which we sometimes call scientific laws) which are never experimentally violated."

Maybe I misunderstood 'experimented' to mean something distinct from 'never violated _at all_'. Which would be tantamount to (ii). (In which case why bother giving (i) at all?)

When I think of 'testable regularities' never being 'experimentally violated', I think in terms of the system behaviors being coherently testable, with an expectation (for whatever reason) that experiments _in regard to this system_ aren't going to come up with purely arbitrary results from time to time. At least at the large scale. (If it comes to quantum behavior, that kind of weak naturalism might not apply.)

If I write a computer program, the system ought to be coherent and return testable regularities. A stably running system hardly means one that is invulnerable to input from outside the system, though. On the contrary, a _really_ well-designed system should run without problems integrating the input.

So, to take a very current example, I want Battlefield 2142 not to crash or throw up goofy space-time wonkiness when game leaders try to move the Titan ships around. (Watching people and bodies scoof through walls when a commander moves a Titan ship around, would be what I consider to be an 'experimental violation' of the system coherency.) But that should and in principle can be achievable, without making the system simply invulnerable to input from outside the system. Relatedly, having a non-crashing system which can cleanly incorporate outside input, does not in the least obviate tests of system behavior. If the system is just going about its normal business, what can we expect? This can be discovered and (to this or that degree) determined.

(In hindsight I know you weren't specifically mentioning outside interference; I'll discuss that presently.)

Furthermore, in the case of robust _super_-naturalism, the system properties and behaviors themselves depend upon continual upkeep by the supersystem anyway. It is _because of_ the supersystem that the system has behaviorial properties at all. Extra-naturalism might, in principle, involve arbitrary system crashes (so to speak); because the outside input has nothing inherently to do with the existence and properties of the system in the first place. But the properties of the system under supernaturalism are what they are _because of_ the supersystem in the first place. There's already an intimate continuing connection between them.

This in itself wouldn't exactly eliminate the possibility of a system crash from input, but it would mean that the relationship is already such that the notion of input is far from foreign to the system: its behaviors are what they are due to what could be considered a specially intimate kind of 'input' from the supersystem already.


{{If no such violations occur, then the best that theists can hope for is some form of deism.}}

I agree with that, btw; although I will also note again that the reason deism tends to drift over into cosmological dualism (with God and Nature existing independently of each other), is precisely because a lack-of-input would include lack-of-upkeep for the natural system (not to say lack of creative action to begin with.)


{{Weak naturalism is a synthetic claim. No logical inconsistency (or logical proof) is possible.}}

I'll pass on this, since at the moment I can't tell what your distinction is supposed to be between (i) and (ii).


{{The statement "2 oranges and 2 apples = 4 fruits" is analytic, and does not rely on experimental evidence for its verification}}

I agree. Thus my point, by implication: the distinction between philosophical claims 'strong naturalism' and 'bare deism' is something that doesn't require experimentation for their distinction. (You could still be agnostic about which one is true, but in principle it might be possible to figure out something of their relative truth by other than experimental analysis.)


{{JP: "Philosophical supernaturalism of the sort being proposed by, for instance, Christian theists, states (or proposes or concludes or hypothesizes, etc.) that there is another reality more fundamental than the evident Natural system, and distinct from it in its own existence, upon which the Natural system (and indeed everything in reality, if anything else than Nature and this Supernature does exist) depends for existence."

NBAA: "Being a scientist by training (and temperament) I find this definition rather hard to digest. What does it mean, in practical or observational terms[?]"}}

It's a conceptual statement first. Metaphysics is about working out conceptual relationships. (This is why pure mathematics has not infrequently been considered metaphysical. Mathematics can be 'applied', of course, but the application depends on the concepts having been worked out first.)

If you're asking what practical difference it would make for supernaturalism to be true, I'm not entirely sure it would make much practical difference simply in and of itself. It would be like saying string theory is true. So what? It may be of interest to some specialists, but the overwhelming majority of us can go about our business without being anything other than agnostic (in various ways!) about that question.

Supernaturalism compared to extra-naturalism makes a conceptual difference at least to the notion of 'violation' of natural law, though: miracles might happen in either case, but the quality could be rather different.

To take a more colorful example: even Richard Dawkins has a tendency to speak as if there is a real and important distinction between natural and artistic behavior. But on the grounds of his philosophical naturalism, such a distinction can only be an perceptual illusion on our part. On the grounds of supernaturalism, however, this distinction could be real in a way qualitatively different from the distinction between physics and chemistry, yet still be far from simply discontinuous with natural system behavior.

Mr. D makes for some fun and interesting examples of principle application, now that I think of it. {g} He clearly has some intuitions that he considers to be both true and important; but they routinely land him in amazing self-contradictions when he tries to account for them under naturalism.


{{I have no doubt that there are theists (like you) who have adopted this position, and I am happy for you to label it "robust supernaturalism". But I doubt if there are any naturalists who would recognize in it the negation of their own position.}}

If I walked up to Dawkins (since I was just speaking of him) and asked whether Nature depends for its existence on anything else more fundamental than itself, I guarantee he would say 'NO!', and then launch into a riposte about how putting the question of self-existence back one stage doesn't help or solve anything. And he's not even specially apt (ept?) on the topic. (His most recent book notwithstanding. {wry g})

Victor routinely quotes from naturalists whose definitions of naturalism involve just this sort of concept: that Nature doesn't depend upon something else for its existence.

The whole foofaraw about what the Big Bang does and does not entail (which personally I don't think naturalists have anything to worry about, btw, but still...) is utterly predicated on whether this means Nature relies for its existence upon something else. It has virtually nothing to do with the question of whether Nature can be _affected by_ a causation that is anything other than natural causation.

{{(And assuming it as the default definition of supernaturalism is a bit like assuming that Christianity is the default definition of theism.)}}

I didn't assume that this is 'the' 'default' definition of supernaturalism. I specifically identified it with the kind of supernaturalism being routinely discussed on this site, which is the kind held by the type of Christianity (though not exclusive to that type) which is typically under debate here and elsewhere in our culture.

{{((ii) strong (or philosophical) naturalism: there exist no entities or forces which are (even in principle) capable of violating the scientific laws postulated by weak naturalism [...] I am agnostic about strong naturalism, mainly because I can't think of any way to distinguish experimentally between strong naturalism (no law-violating entities exist) and bare deism (such entities do exist, but they choose not to violate the laws). [...] At no point in my definition of strong naturalism did I refer to tampering "from outside the system".There are many ways that strong naturalism could be violated "from within": for example, by witchcraft.}}

Or by free will?

Classically witchcraft has involved either appeal to powers not of this system of reality; or else has proceeded on the notion that the behaviors of the natural system are simply not exhaustible by what we now would call scientific investigation. In the latter case the behaviors were nominally 'natural' to begin with, meaning there was no 'violation' 'from within' or otherwise. (Which is why most Wiccans, btw, believe Nature is the most fundamental reality, or at worst is one of multiple independently self-existant realities.) In the former case, the 'violation' was made with outside help, so to speak.


In any case, the notion of tampering from-and-only-from within system properties is incoherent. The tampering would either have to be done using system properties--which means nothing is actually happening other than system properties, whatever appearances may seem--or else would be done using something other than properties of that system. This is extra-naturalism at least.

You have nothing to contrast (ii) against, therefore, except a position that is at least extra-natural. Positing not-natural-but-also-natural-and-only-natural behaviors as a contrast won't work; and positing behaviors that only _appear_ to involve something other than system behaviors isn't a real contrast of position. It's simply x=x; not x instead of not-x.

{{The two definitions I gave, of weak and strong naturalism, were worded so that weak naturalism was a subset of strong naturalism _by definition_.}}

Either you were trying to word (i) to exclude anything but (ii), or you were not. If the former--which now appears to be the case--then you might as well have stopped with (ii), since any difference involved between the two positions will be trivial at best; besides which it would mean you aren't really agnostic about (ii). If it was the latter, then it shouldn't be of concern that discovering the system behaves in erratic ways would not (as I was reassuring) in any way falsify (ii).


{{But do you know of anyone who is actually a robust philosophical naturalist?}}

In the words of Paul Draper (quoted on the main page of infidels.org, aka the Secular Web), naturalism is "the hypothesis that the physical universe is a 'closed system' in the sense that nothing that is *neither* a part *nor* a product of it can affect it."

Many of the guest sceptics on this site hail from the SecWeb; and the position is pretty common in the field, actually, which is why the site admins use it. (I will add that it is common for naturalists to put 'method nat' somewhat stronger than you've done it, too: i.e. that science and history _should presume_, for the purpose of promoting successful investigation, that _all_ causes are natural causes.)


Now, a naturalist wouldn't have to _strictly_ be robust in the sense of claiming Nature is _definitely_ the one and only Independent Fact. But Nature has to be at least _an_ Independent Fact, self-existent, with no interference from other IFs (if there are any). Which means for all practical purposes that the position is robust. (And besides, I find most philosophical naturalists denying such things as cosmological dualism, too.)


NBAA: "I disagree [that supernaturalism is at least potentially symmetrically parallel to naturalism in regard to falsification by propositional analysis--something I, JP, said before NBAA gave (i) and (ii), btw.]. Naturalism and supernaturalism are contingent statements about the way the world works. (Here I mean weak naturalism and its negation [i.e. distinct from (ii)].)"

JP: "By supernaturalism, I do not mean the negation of the proposition that the natural system will always of its own accord behave coherently. (Where would the 'super'-naturalism be in that??)"

NBAA: "Where indeed? It is evident that your idea of the supernatural is quite different from my own. My understanding is that supernaturalism (of whatever flavor) is just the negation of naturalism (of the same flavor)."

In which case, wouldn't it make more sense to call it '_anti_-naturalism'? 'Super' is not usually considered a contrapositive or negation term. (Neither is 'extra', btw.)

It ought to be linguistically obvious, if nothing else, that 'super-' or even 'extra-'naturalism involves claiming there is something more in reality than naturalism alone is claiming. If anything, naturalism is the negation of something being positively claimed in supernaturalism. (I don't mean that against naturalism in any way; it just seems common sense.)


{{You pointed out that "extra-naturalism" would be a better word for this, and I am inclined to agree.}}

Actually, I pointed out that "extra-naturalism" would be a better word for cases where one system is producing effects in another system without one system being dependent for its existence upon the other. That was specifically in connection with (ii); not with (i), which I took to be a statement about coherent system behavior.

There is a major conceptual difference between the notion that a system will always behave coherently of its own accord; and the notion that _there is nothing_ that can even in principle introduce behaviors into the system (from wherever) that are not what the system is coherently producing. If you didn't mean to introduce that kind of conceptual difference in (i) and (ii), then I need to know in what way you weren't just giving two ways of saying the same thing.


Jason Pratt

Anonymous said...

Interesting discussion so far. I'm not sure why there has been this seperation between supernaturalism and theism. Surely, there is no logical necessity to connect the two, but for all practical matters, in the West at least, they are inseperably bound. Theists are always supernaturalists. Are there any atheistic supernaturalists?
Sam

Anonymous said...

"A stably running system hardly means one that is invulnerable to input from outside the system, though. On the contrary, a _really_ well-designed system should run without problems integrating the input."

The problem isn't stability, but coherency, Jason. In your Battlefield 1942 analogy, any intervention from outside is obeying the exact same physical laws that the game obeys. So one can easily explain the intervention. I've yet to see a coeherent explanation for any intervention by God. Probably because this intervention is explicitly postulated as a violation of physical laws?

Anonymous said...

"There is a major conceptual difference between the notion that a system will always behave coherently of its own accord; and the notion that _there is nothing_ that can even in principle introduce behaviors into the system (from wherever) that are not what the system is coherently producing. "

The real problem as far as I can see is coming up with a coherent account of how a supernatural being like the Christian God can intervene in the universe.
How can a being that exists outside of space and time think and act in a manner that would enable It to cause physical effects?
Sam

Anonymous said...

(NBAA)

NBAA wrote: "My understanding is that supernaturalism... is just the negation of naturalism..."

JP wrote: "In which case, wouldn't it make more sense to call it '_anti_-naturalism'?"

NBAA: I am happy for you to call the negation of naturalism anti-naturalism, extra-naturalism or supernaturalism as you will, provided that everyone understands the usage. (However, I note that you did not explain why you were prepared to endorse the naturalism/supernaturalism dichotomy in an earlier post.)

In order to clarify matters (at least in my own mind), I will start from the basics, and try wherever possible not to use "ism" labels.

1. You defined "robust philosophical supernaturalism" (RPS) to be the belief that (i) there is another reality more fundamental than the evident Natural system, and (ii) the Natural system depends on this other reality for its existence.

My claim was (and is) that no Naturalist would define his or her Naturalism as the negation of this position. Yes, Richard Dawkins would probably say "No" if asked whether Nature depends on any reality more fundamental than itself, as would Paul Draper. (I would be reluctant to say "No" myself, mainly because I have no idea what the observational implications of RPS are.) But just because someone disagrees with a statement S does not mean that his or her philosophical position is not-S. It is usually not-S plus lots more.

So it is with Naturalism. The negation of RPS is the belief that either (i) there is no reality more fundamental than Nature, or (ii) Nature does not depend on any other realities for its existence, even if there are some. An adherent of not-RPS (and nothing more) is free to believe in all sorts of things that no modern Naturalist would believe in. This adherent could believe in the paranormal or in the modern Wiccan belief system, and still consistently believe there is no reality more fundamental than Nature. Or he/she could believe in the "otherworld" of classical witchcraft or a separate spirit world, and still consistently believe that Nature does not depend on these other worlds for its existence.

No modern Naturalist that I know of believes in the paranormal or Wicca or classical witchcraft or the spirit world. People of this persuasion are spiritualists or mystics, not Naturalists. (Although it is true that when Naturalism was first defined in the 17th century it was generally identified with belief in the occult.) So RPS rules in only a tiny part of what Naturalism rules out, and one is not the negation of the other.

2. What exactly is Naturalism then? I have defined weak naturalism (WN) to be the belief that the world is governed by scientific laws that are never experimentally violated. (And yes, "never experimentally violated" _is_ distinct from "never violated at all". It means "never violated in closely controlled experimental conditions".) And I have defined strong naturalism (SN) to be the belief that no entities or forces exist that could even in principle violate the laws.

You claim that SN "didn't go far enough to be in direct comparison with _super_-naturalism per se." I hope it is clear now that SN is considerably stronger than not-RPS, and so goes quite a bit further. Also, my definition of SN very deliberately makes no reference to "other realities". In defence of this I quote Wikipedia's definition of "metaphysical naturalism":

"Metaphysical naturalism is any worldview in which nature is all there is and all things supernatural (which stipulatively includes as well as spirits and souls, non-natural values, and universals as they are commonly conceived) do not exist. It is often simply referred to as naturalism, and occasionally as philosophical naturalism or ontological naturalism..."

This definition (like most I have seen) is somewhat unsatisfactory because of its tautological reference to "nature". But there is no mention anywhere in it of "other realities" (although you might argue that it is implied), and certainly not of any "realities" that Nature depends on for its existence.

3. I said that I am agnostic about SN because it seems experimentally indistinguishable from deism. In fact, I should have added that SN is experimentally indistinguishable from WN as well. SN is just WN with much stronger epistemic claims.

In connection with this, you state that "You could still be agnostic about which one is true [SN or deism], but in principle it might be possible to figure out something of their relative truth by other than experimental analysis". My only response is this: If you know of a way to distinguish contingent claims by anything other than experimental means, you owe it to the world to reveal this wonderful secret.

4. I am indebted to you for your example about the computer program. It provides a neat way of clarifying the distinctions between the various propositions under discussion here. So suppose that the world is just an elaborate computer program which every so often (perhaps each time the spin of an electron is measured) pauses to wait for input. This input could be entered in a variety of ways. Perhaps an external operator ("God") enters "spin up" or "spin down" to fix the result of the measurement. Or maybe a random number generator within the program does the same thing. A third option is that the computer hosting the program is linked to an external random number generator which plays the role of the (absent) external operator.

Now, from the point of the view of the "people" inside the program, the distinction between input "from within" the program and input "from without" is completely meaningless. How could they distinguish between a random number generator embedded within the program and one linked externally? And why would they care? Now I know that theists _do_ seem to care about whether there is "external" intervention in the world or not, but Naturalists generally regard the distinction as an artificial one. What is far more important for Naturalists is the distinction between "God the operator" and a random number generator (whether internal or external).

In one case (the random number generator) we expect to be able to analyse the results of experiments to calculate the statistical properties of the generator, and then use these to make (statistical) predictions about future measurements. In the other case ("God") we expect the results to be arbitrary and capricious, or at the very least beyond our understanding.

Nor is it necessary for the "God" option to be an "external" one. "God" could have inserted a number generator into the program with no discernible statistical properties. (There are uncountably more sequences of numbers which do not converge towards a mean than there are sequences which do have a limiting mean.) Alternatively, "God" (being both lazy and omniscient) could have predicted exactly which spins He was going to choose in advance, and simply entered the sequence of choices directly into the number generator when writing the program. In either case the world would be able to run smoothly without any "external" intervention, and yet Naturalism would be falsified, as there would be no discernible laws governing measurements of electron spin.

I hope this example clarifies the meaning of (weak) naturalism. The task of science (guided by WN) is not to distinguish behavior that is "inherent" to the world from input that is "foreign" to it, nor to test whether there is some "intimate continuing connection" between the world and "another reality". It is simply to test each and every phenomenon we see for statistical or other regularities. And there is no _a priori_ reason why "the system should... return testable regularities".

Why should there be a statistical law governing electron spin? Why should the planets and Sun have formed in predictable ways instead of popping into existence by divine fiat? Why should humans and animals have evolved by continuing speciation from (usually) simpler ancestors rather than appearing in puffs of separate creation? 500 years ago such regularities would have been scoffed at by everyone.

5. JP wrote: "...even Richard Dawkins has a tendency to speak as if there is a real and important distinction between natural and artistic behavior. But on the grounds of his philosophical naturalism, such a distinction can only be an [sic] perceptual illusion on our part."

Far be it from me to jump to the defence of Richard Dawkins, whose rhetoric is often too extreme and inflammatory for my taste. However, "Mr D" is obliglated to use the English language in the same ways as everyone else, and it is rather flimsy to take him to task for speaking "as if there is a real and important distinction between natural and artistic behavior." It's a bit like accusing an atheist of hyprocrisy for saying "Good God!", or an astronomer of geocentrism for saying "The Sun rose this morning."

In any case, where in the definition of philosophical naturalism (in _any_ of the definitions of phil-nat that have been raised here) does it say that a "distinction between natural and artistic behavior... can only be a perceptual illusion"? Smells suspiciously like a straw man to me.

Jason Pratt said...

Necessarily long post arriving. {g} (I would have replied sooner, but the thread went off the bottom of the main page, making it more difficult for me to keep track of it. I doubt I'll be returning again, unless Victor sets up a new post for the discussion to continue in; though he may possibly redate this one later, too, so...)



Sam (Anon) writes: {{I'm not sure why there has been this separation between supernaturalism and theism. Surely, there is no logical necessity to connect the two, but for all practical matters, in the West at least, they are inseparably bound. Theists are always supernaturalists. Are there any atheistic supernaturalists?}}

Depends on whether there are any atheists proposing that the evident system of Nature depends for its existence on something that is distinctly different in essence from the system. I think there are astrophysicists and quantum physicists here and there who are proposing exactly this--and not in religious terms. They may not call it supernaturalism, being physicists instead of metaphysicists (there would be a terminology gap {s}) but it's the same thing I'm talking about in principle.

And the reason I make sure to keep the difference between the two concepts in the account, is to avoid introducing category error by eliding between the concepts. This occasionally works in favor of my opponents, btw, by nixing overly enthusiastic apologetics from my side of the aisle. Too bad for my side. {shrug}{s}


A(nother?)non (Sam?) also wrote: {{The problem isn't stability, but coherency, Jason. In your Battlefield 1942 analogy, any intervention from outside is obeying the exact same physical laws that the game obeys. So one can easily explain the intervention. I've yet to see a coherent explanation for any intervention by God. }}

The question of stability _could_ be a problem; I frequently detect an underlying expectation to the effect that the introduction of input from outside the system would simply throw the system out of whack. However, I entirely accept your coherency criticism, too, in the technical sense (which seems implied by the content of your crit) of the question of the feasibility of a juncture between two essentially different systems of reality--which, I fully admit (and thought about mentioning in my previous comment, but it was long already and I expected an opponent would draw the distinction anyway {g!} {bow!}), is not an issue in the BF2142 example--and I agree that it is not an issue for precisely the reason you gave (setting aside the otherwise-far-from-unimportant question of whether active intention is in fact _only_ a product of "the exact same physical laws the game obeys".)


Having said that: it should have been noticed that continuously throughout my comments (meaning it has been hardly buried in a spot mention here and there), I have been talking about ontological dependency and independence. (Independency? {s?}) This was absolutely the distinction I was drawing between an extra-natural situation and a _super_-natural situation, for example. Nor is this notion in the least foreign to the discussion; because the particular largescale and widespread system of belief being opposed by philosophical naturalists (of whatever stripe), has _always_ maintained that the evident system of Nature depends for its existence on God Who is distinctly different in essence from the natural 'kosmos' (a Greek word that itself is used in regard to a designed work of art, btw.) This is what traditional mainstream advocates of the Big Three Theisms (Judaism, Christianity, Islam) have always been claiming; and it is the position being taken up and defended for belief by Victor on this site (as well as by most if not all of the rest of us theists, myself included. As Sam, assuming for convenience he was the 2nd anon, has rightly pointed out, most of us, in the West anyway, are _supernaturalistic_ theists.)

A supernaturalistic theist does not _only_ say God is distinctly different in essence from Nature. We also say God _created_ Nature. That's why we're supernaturalistic theists _instead of being_ God/Nature cosmological dualists (where both God and Nature exist independently of each other, and so by corollary one entity cannot affect the other by sheer application of power. Meaning, not incidentally, no miracles; unless perhaps Nature personally gives permission. {g})

To ignore or discount this concept in a dispute with supernaturalists (theistic _or_ otherwise) about the question of (technical) coherency in _affecting_ the natural system by supernatural causation, is simply to be aiming off in a direction different from your target. If Nature is _effected_ (i.e. produced) by a supernatural system, then on the face of it there is already an _effective_ conjunction between the two systems; thus if the independent system goes on (within the 'natural history' of the dependent system) to introduce _further_ effects, then in principle there wouldn't be any more incoherency in doing this than in producing the natural system in the first place.

(I anticipate a further objection accepting this rebuttal and so taking the question of coherence back one stage to the creation or production of the dependent system itself. I welcome the objection, if so; but for sake of abbreviating a little I will not go on to address it here.)


A very large percentage of us supernaturalists (and this could logically apply to supernaturalistic atheists just as easily, too) also insist (implicitly or explicitly) as a logical corollary to the production of a dependent system by an independent system (ultimately recusing in principle back to _the_ Independent Fact, of course), that the independent system is _continuously upkeeping_ the dependent system. This would especially have to be true when discussing (as of course we ultimately are) the relation of proposed system derivation from the Independent Fact itself. A novel's story may continue existing independently of me, even though in some real sense I am the creator of it (speaking as a novelist {g}); but I am not _the_ foundational system of _all_ reality. The novel (or computer program) can go on to exist independently of me to any of various degrees, all the way out to having nothing ever again to do with me at all (aside from its creation by me being a fact of history); but it will still be dependent upon something else for its existence, even if not upon me per se. Its ability to continue existence without my continual upkeep, indicates a shared level of existence between us; ultimately that it and I are both dependent for our own existence upon something more fundamental than either.

Since we are concerned with discussing _the_ IF (as well as with identifying and relating to the IF as God--i.e. as theists instead of as atheists, though again a supernaturalistic atheist could make the same logical insistence I'm currently discussing without being a theist at all), we therefore recognize that a proposal of Nature existing independently of Supernature 'after' Nature's creation by the supernatural system, would imply that in discussing this 'supernatural' entity or system, even though it _would_ admittedly still be supernatural to the natural system, we would _not_ yet in fact be discussing _the_ final Independent Fact of all reality. We would only be talking about the Qabbalic Metatron, or the Platonic Dynad, or the Arian Christ; not yet God Himself. (A supernaturalistic atheist would put that another way, of course.)


This notion of the _dependency_ of a system (which, I remind, is a notion I have been discussing all along) thus answers the immediate question of coherence (leaving off until some other time a discussion of the coherence of creation in the first place): there cannot be some intrinsic 'big ugly ditch' between an operation of the IF and operations of a DF. There is already an intimate connection of _effecting_. If Nature was not "violated" by being created as Nature in the first place, it would not necessarily be "violated" by subsequent operations from the IF within the natural system. And if those of us among supernaturalism (theistic or otherwise) are correct (even against other supernaturalists, theistic or otherwise--and historically there have been such proponents; not incidentally a school of thought that did lead eventually to cosmological dualism via nominal and then minimal deism) about the necessity of _continuing_ dependence for existence upon the IF--and especially upon the continuing _behavior_ of the IF (notice the theistic flavor of that, btw)...

...then there is even less reason to talk about the "violation" of the natural system by any input from the supernatural; and there is (if anything) even _more_ coherence between the natural and the supernatural.

(In passing, I pause to draw notice to the traditional language and insistence of both transcendence _and_ immanence in God's relation to Nature, among at least two of the Big Three Theisms, i.e. Judaism and Christianity. I expect, though I don't know for sure, that there should be some significant strand of this in Islam, too; but I also know their theology of radical transcendence tends to move away from any immanence _also_ being true, so it wouldn't surprise me to find this element far more suppressed in their culture. As a matter of practical history, this separation is theologically connected to the fact that they have been, if anything, even more combatively 'us vs them' than Christians and Jews have been in our own turns. I am not surprised to find orthodox Christian universalists, such as among the EOx, or myself for instance; but I would be _very_ surprised to find an 'orthodox' Muslim universalist.)



Sam: {{How can a being that exists outside of space and time think and act in a manner that would enable It to cause physical effects?}}

I quote this here only to point out that this same question would be applicable to creation in the first place (which is where I expect the crit to eventually go, though you presented this question in the immediate context of _intervention_ instead of _creation_.)

I do have some notion of an answer to this, btw. (I'm just not addressing it until the ground of complaint is formally moved a notch to that topic.)


Moving on to NBAA (here and afterward): {{I note that you did not explain why you were prepared to endorse the naturalism/supernaturalism dichotomy in an earlier post.}}

Sorry--not everything can be discussed even in a long post. {s}

I never did claim that naturalism and supernaturalism were dichomatic in a merely opposite fashion. They _are_ mutually exclusive options, but not all mutually exclusive options are simply negatives of each other. (This would be another reason why I make a point of routinely distinguishing the topical category of theism/atheism from supernaturalism/naturalism, btw. It's hard not to talk of theism/atheism without making atheism primarily the negation of the other; but naturalism can be claimed in a primarily positive fashion.)

If a negative description between the two must be given, then of course by topic supernaturalism necessarily proposes something more in content than naturalism does. i.e. naturalism (in this sense) would be the negation of a positive claim of extra reality being made under supernaturalism. (Which connects with the Draper definition being used over at the Secular Web, btw.) The negation by topic goes only one way. (Unless one considers a positive claim to be a negative-negative claim, but that would seem to be overly baroque. {g})

I don't usually try to present philosophical naturalism _primarily_ in terms of being a negation to supernaturalism, though. Instead I usually try to present it in its own positive fashion (e.g. 'the evident system of Nature is the single Independent Fact of reality'), with any negative statements being provided as corollaries to that position (e.g. 'meaning there is nothing upon which Nature depends for its existence, thus necessarily excluding supernaturalism, nor anything that can be even possibly recognizable as existing in ontological equality with it, i.e. cosmological dualism is either false or absolutely irrelevant.')

This sort of definition would include fundamental reliance upon physical laws (using 'law' as a handy abstract description for regular physical behaviors); but isn't necessarily married to physical law per se (since for all we know scientists may discover natural behaviors someday which don't strictly count as physical. As non-eliminative materialists tend to frequently propose, for instance.) Thus I avoid hinging the definition on a potential vulnerability for the philosophical naturalist position. (i.e. I do this to protect against a potential perceived defeater for phil-nat. Philosophical naturalism is a metaphysical claim, the important principles of which I recognize would _not_ be defeated if physics turned out not to be the most fundamental natural operation.)


{{My claim was (and is) that no Naturalist would define his or her Naturalism as the negation of this position.}}

I try not to do that _primarily_ either. Nevertheless, it's very far from being irrelevant to the current discussion, once supernaturalism is introduced for comparison.

{{But just because someone disagrees with a statement S does not mean that his or her philosophical position is not-S. It is usually not-S plus lots more.}}

This is in reply to my somewhat wry observation "In which case, wouldn't it make more sense to call [supernaturalism] '_anti_-naturalism'?"

That observation, however, was _NOT_ given by me to something _I_ was saying about supernaturalism. It was given to something _you_ were saying about what _you_ consider supernaturalism to be.

Reposting the two comments again: _you_ wrote: "My understanding is that supernaturalism... is just the negation of naturalism..." After my wry comment on putting supernaturalism this way, you then wrote: "just because someone disagrees with a statement S does not mean that his or her philosophical position is not-S. It is usually not-S plus lots more. [...] So RPS rules in only a tiny part of what Naturalism rules out, and one is not the negation of the other." Uh... yeah. Re-read your own understanding of supernaturalism again, please. {g} _That's_ why I was wondering why _you_ would even call it supernaturalism at all, instead of anti-naturalism. (I know, and explained at great length, why _I_ would and do call the position _super_-naturalism; in distinction from extra-naturalism for instance. It has nothing to do with being "just the negation of naturalism".)

On the other hand, your SN definition is also _robustly and altogether negative_. ("I have defined strong naturalism (SN) to be the belief that no entities or forces exist that could even in principle violate the laws.") You made no statement of its _positive_ tenets (if any) at all, in your own definition. (The positive tenets were made in your WN. SN, by your definition, adds only denials of existence.)


{{An adherent of not-RPS (and nothing more) is free to believe in all sorts of things that no modern Naturalist would believe in. This adherent could believe in the paranormal or in the modern Wiccan belief system}}

As I think I pointed out myself.

{{Or he/she could believe in the "otherworld" of classical witchcraft or a separate spirit world, and still consistently believe that Nature does not depend on these other worlds for its existence.}}

Actually, I have very consistently insisted that a philosophical naturalist _per se_ would _not_ believe in such separately existent realities. This is because I have consistently been operating on a _positive_ notion of naturalism, not just naturalism as a negation of supernaturalism.

While I'm at it, the 'modern Wiccan belief system' would include (in some variants) _theistic_ naturalism, i.e. pantheism. Yes, that would still be philosophical naturalism by (rigorous use of) the definition used in metaphysics. Yes, "modern Naturalists" would not believe in that. This isn't because it isn't physics-based, though--I think most pantheists nowadays have no problem with things being physics-based (or anyway on whatever sub-physical behaviors 'physics' operations are based on, per QM). It's because "modern (i.e. western) Naturalists" have an atheistic notion of the natural world: the most fundamental behaviors are non-sentient. (e.g. the most basic particles, and in fact _all_ particles up through highly complex biological systems at least, only react and counterreact with each other; they don't initiate intentional actions.)

The reference to the Wiccans, therefore, might be a version of the elision I've been constantly cautioning against: some of them are atheistic, but most of them are theistic in some way. And _are also_ frequently philosophical naturalists (though with some strong contingents of God/Nature or rather God/Goddess cosmological dualism among them, too.) Unless she's specifically atheistic, a Wiccan is going to think (assuming she has any specificity on the matter at all--they're notoriously loose on metaphysical reasoning {g}) that the fundamental behavior of reality is sentient, even if she's a philosophical naturalist (by metaphysical reckoning).


{{And I have defined strong naturalism (SN) to be the belief that no entities or forces exist that could even in principle violate the laws.}}

I'm now starting to wonder whether, despite your original remark about understanding and approving of the distinction between naturalism and atheism--which I took you at your word about--you aren't trying to position 'strong naturalism' as being necessarily atheistic after all.

Let us consider the test case of pantheism. By the metaphysical definition I am using, this would be philosophical naturalism, but not atheism.

Would it be philosophical naturalism under your definition of 'strong naturalism', though?

If you say yes, then by the terms of your definition this would be because the "spiritual" operations of nature, while still (per postulation) being real enough to count as theism, nevertheless do not violate scientific laws.

Is this well and good for your definition? But if it is, then so would be practically all the other things you just mentioned that "no modern Naturalist would believe"! (Apart from a 'separate spirit world', which I take to be categorically a proposition of cos-du, not naturalism, anyway; the key word being 'separate'.) Why would they not believe such things, then? (And why would I be sniped for proposing a definition of naturalism that would include things they wouldn't believe?--which I certainly do, even though not quite on the ground you were giving.) The only obvious answer I can think of, is because those things taste a little too much like _theism_. Which pantheism _definitely_ would!

On the other hand, if pantheism is excluded from your definition because the sentient operation of the natural system would involve 'scientific laws' being "violated" after all; then your definition is tacitly including a built-in exclusion of theism.

Now, it is certainly true that in our western society (as Sam himself mentioned earlier) philosophical naturalists are usually going to be atheists, too. But that is something of an accident of history: western science is based (in that regard) upon a Judeo-Christian expectation that Nature will operate in a non-sentient fashion. Not surprisingly then, our naturalism, once we begin going that way, is likely to be atheistic. But naturalism doesn't _have to be_ atheistic.

Whatever else your SN excludes, it also excludes any possibly effective recognition (if not the reality outright) of systems 'outside' of Nature (metaphorically speaking).

When I cast around in comparing philosophical systems, though, I see, broadly speaking, that there is another set of beliefs which feature that exclusion; or which (as I try to do when describing the belief) put the matter more positively (e.g. that there is only one reality.) That other set of beliefs is (again broadly speaking) theistic. Which contrasts easily and fairly clearly with modern 'western' _atheistic_ naturalism. The category distinction then _for_ naturalism is what popular-western and pantheistic naturalism share, which is certainly not a proposal about God: on _that_ they are opposite. i.e., one is atheistic and the other is theistic naturalism.

Thus I reach _again_ the philosophical definition of naturalism, distinct from the question of a/theism--and I work pretty hard to make sure the distinction actually _is_ kept distinct in category when discussing the claims. (Yet _not_ a simple dichomatic negation of theism as atheism is.) I do this to help avoid category error in the discussion.


Now, are you, in your definition of SN, being careful not to elide between categories? Even you recognize and have said that naturalism and atheism are _not_ the same claim. But if that's true, then your SN definition should allow for pantheism, too. But if _that's_ true, then it ought to even more easily allow for a number of things which you do rightly note (as a matter of historical fact) that "modern Naturalists" would not accept.


{{I hope it is clear now that SN is considerably stronger than not-RPS, and so goes quite a bit further.}}

Indeed--at the moment it looks like it goes 'quite a bit further' by being tacitly (though not unexpectedly) atheistic! {g}

Because if it's going "further" by further _exclusion_ (and the form of your definition is completely exclusory), then I'm not sure what else is being more primarily excluded. Only things that are either theistic in property-claim, or might happen to taste that way (so to speak), so far as I can make out. (There is a reason why I briefly asked "Like free will?" when you first wrote your SN definition. The sort of things you're talking about "modern" Naturalists _not_ believing in--again a completely exclusory reference--all involve something like free will somewhere; and free will has traditionally been held to be a problem for naturalism even by some prominent modern naturalists.)

{{Also, my definition of SN very deliberately makes no reference to "other realities".}}

And yet, no modern Naturalist believes in other realities, do they? Wouldn't that seem to point to an important element being left out of your definition?

But it could be fairly answered that your definition of SN involves excluding _even more_ than 'other realities' (or rather any effective operation thereof within Nature at all). The _even more_ being _excluded_ by your definition, is (by property)...?

{{In defense of [my, i.e. NBAA's, definition of SN very deliberately making no reference to other realities] I quote Wikipedia's definition of "metaphysical naturalism":}}

Um. Did you actually read it? It doesn't even get past the first sentence before making very deliberate and explicit reference to other realities, specifically the supernatural. Indeed, its definition is practically similar to what I've been saying all along:

"Metaphysical naturalism is any worldview in which nature is all there is [i.e. is the one and only IF, dependent upon nothing else for existence, and exclusive of such things as cosmological dualism] and all things supernatural (which stipulatively includes [a list of things that, by terms of this definition, would seem to tacitly or explicitly involve admitting that nature is _not_ all there is]) do not exist."

(At the moment it looks as though you simply scanned through the definition, noticed it did not happen to include the two English words "other realities", and so figured that it must not be talking about them.)

{{This definition (like most I have seen) is somewhat unsatisfactory because of its tautological reference to "nature".}}

That 'tautological reference to "nature"' is precisely what I've been mentioning as being critical to the definition of naturalism per se (distinct from atheism). i.e. the definition is positively worded to exclude, by corollary, other realities than nature. A system that is _everything_ and yet which is _exclusive of_ something, is a reference to a single-essence reality over against postulations that something exists distinctly different from nature. (This is also what keeps the definition from being actually tautological. Nature isn’t simply everything whatever everything happens to be.)


{{SN is just WN with much stronger epistemic claims.}}

If that's true, then they are in fact identical as claims, with SN being only more strongly worded.

I say this, because I have to suppose the 'epistemic' claims you're talking about are being read into the definition by the SN advocate (thus you're talking about subjective force of belief being the only difference between WN and SN). Nothing in your actual SN definition has anything to do with an epistemic claim at all, much less one vs. something else than an epistemic claim in WM.

(ii) reads: there exist no entities or forces capable of violating the laws postulated by weak naturalism. (With the postulation of WN not being epistemic in topic, either.) That is purely an ontological claim. And a totally negative one at that. (It is (i) which makes the positive claim: the world around us is governed by certain testable regularities. Which, btw, is kind of a category error in itself, though I've let it pass for now. The regularities per se are not what is doing, metaphorically speaking, the governing. The regularities are what actually happens. The regularities themselves depend on the intrinsic and ultimately fundamental properties of something else--something else only 'natural' per naturalism.)


{{If you know of a way to distinguish contingent claims by anything other than experimental means, you owe it to the world to reveal this wonderful secret.}}

Philosophers have been distinguishing (and judging between) claims for a few thousand years now, without needing experimental means. Science is a very specialized form of logic (involving experimental means); admittedly with wide application, but we don't need it to solve every problem. Ruling out the truth of a proposition (as given) on grounds of logical incoherency doesn't _require_ experimental means in the least, for instance. This is not remotely news to a lot of the world; and I find it hard to believe that deductive logic is a wonderful secret to you.

For example, if by 'contingent' you mean (in this case) that neither one of the propositions you mentioned (SN or deism) is necessarily true (and I'm making a bit of a guess there, since several subtly different definitions of contingency are used in the field, but my point would still be the same if you meant something else), then I absolutely guarantee you didn't come up with _that_ knowledge (that the two claims are 'contingent') by inference from data acquired by testable experimentation.

You yourself have made a claim of exclusivity regarding the truth of the propositions: they are 'contingent' claims and not 'not-contingent'. How did you figure this out? The same (general) process would be precisely what I meant when I wrote, "in principle it might be possible to figure out something of their relative truth by other than experimental analysis."


{{So suppose that the world is just an elaborate computer program which every so often (perhaps each time the spin of an electron is measured) pauses to wait for input.}}

Or even goes about its systemic business without pausing for input. (If, on the other hand, a pause is necessary for your example, that might be worth further commenting on.)

Option 3, btw, would be more-or-less tantamount to what I've been calling supernaturalistic atheism (if the external random number generator was the final ground of all reality.)

{{Now, from the point of the view of the "people" inside the program, the distinction between input "from within" the program and input "from without" is completely meaningless.}}

Not at all; unless you are also postulating (for no immediately obvious reason) that the people inside the program wouldn't even use the mutually exclusive terms or think about the concepts.

And if some of those results involve properties not otherwise recognized to be fundamental behaviors of their system, nor derivable in principle from those behaviors, then the results wouldn't be conceptually indistinguishable, either. (They might even be experimentally distinguishable.)

{{How could they distinguish between a random number generator embedded within the program and one linked externally?}}

That would admittedly be more difficult than distinguishing intentional input. I am presuming that, like us, they _are_ in fact capable of conceptually distinguishing non-intentive and intentive input, though. If I thought the comments you've been writing were fundamentally indistinguishable in property from random quantum flux (exhibited at the macro-scale in regularly averaged patterns), I would simply ignore you and go play a skirmish mission against the computer on Warhammer 40K. As I have had occasion to quip to one of the (other {g}) prolific commenters here on the site (since even before there _were_ such things as webjournals) an illusion of conscious will is still only an illusion; and I don't debate things with my little cousin's Furbee.


{{Now I know that theists _do_ seem to care about whether there is "external" intervention in the world or not}}

True, as it happens, here in the West theists usually will be found to care about external intervention, but that isn't what theists care about _as theists_, whether here or elsewhere. We care about intentionality as theists. We care about external intervention as supernaturalists. Different categories. (And while our imaginary computer people might not care much, except as a technical matter, for whether a mere random number generator was internal or external to their system of reality, they might care a lot about whether there was a real intentionality involved, whether intrinsically internal to their system or external to it.)


{{What is far more important for Naturalists is the distinction between "God the operator" and a random number generator (whether internal or external).}}

What you have just said is that "Naturalists" primarily care about the theism vs atheism question. In which case, you'd be clearer if you said "atheists" instead of "naturalists". (This is more evidence that you are, in fact, eliding between the categories after all, and that your SN claim is really only atheism renamed--maybe even the WN claim, too. Despite the fact that you yourself began by agreeing there was a _distinction_ in category between naturalism/supernaturalism claims and atheism/theism claims.)

{{In the other case ("God") we expect the results to be arbitrary and capricious, or at the very least beyond our understanding. }}

Why? Are your own actions _only_ arbitrary and capricious or at the very least beyond your understanding? Do you consider the behavior of other persons to _only_ fit into those categories?

Granted, if you analyze a person's behavior as though he is _not_ a person but only a reactive illusion of personhood, then you're likely to run into some errors of prediction; but even then you may also get a 'distantly close' result of expectation, so to speak.

Meanwhile, you are testifying to the strength of the notion, that the behavior of persons you actually believe to exist, such as yourself, are not the behaviors we should expect to develop in an fundamentally atheistic reality. (Meaning you _do_ in fact have routine evidence of something that doesn't fit even in principle into your SN definition.)

{{Nor is it necessary for the "God" option to be an "external" one.}}

{sigh} As if I haven't been saying this constantly myself.

Though the example you then gave ("'God' could have inserted a number generator") is in fact, by your own chosen wording, an external God option.

So was your second example, too. ("'God'... could have... simply entered the sequence of choices directly into the number generator when writing the program.") That means God is still external to the program and putting things into it, and indeed creating it. He isn't the program itself.

{{In either case the world would be able to run smoothly without any "external" intervention, and yet Naturalism would be falsified}}

Your SN would be falsified because God was involved. The notion of philosophical naturalism _I_ was using would be falsified because your examples do in fact both involve definite input from a source external to the system. (Though also keeping in mind Sam's correct observation that for the analogy to work better God has to be different in essence from the essence of the system; which in the case of humans and computers is at least partly false: humans are at least partly of the same essence as the computer system, though it would be begging the question of the discussion one way or another to be more definite than that.)

{{The task of science (guided by WN) is not to distinguish behavior that is "inherent" to the world from input that is "foreign" to it}}}

Yet scientists _can_ identify (in principle) behavior within a system that isn't generated by the system. Otherwise forensic pathologists, to give one example, would be flat out of a job. (So would people who work in computer security.)


{{there is no _a priori_ reason why "the system should... return testable regularities".}}

Actually, I agreed that from a merely scientific perspective, there could be no a priori reason to expect that.

{{Why should the planets and Sun have formed in predictable ways instead of popping into existence by divine fiat?}}

Or even by merely natural fiat? (As subatomic particles apparently do, and would have to be doing per naturalism.)

{{500 years ago such regularities would have been scoffed at by everyone.}}

In 1507?? You do realize that things like medicine, agriculture, cooking, engineering and metallurgy (etc. etc. etc.) absolutely rely on and expect such natural regularities, and that such disciplines had been widely practiced with wild success for thousands of years prior to the High Middle Ages, right?

In any case, if science's job is "simply to test phenomena for statistical or other regularities" (i.e. only to discover whether such regularities occur), then science's job is _not_ to answer the question of _why_ the regularities occur. By your own definition, _you_ ought to be scoffing at answering that set of questions, insofar as you are a scientist.


{{However, "Mr D" is obligated to use the English language in the same ways as everyone else}}

I recommend being cautious in defending him on this. {g}

At the bottom of page 60 of his 1996 revision (though it wasn't much revised, even by his own admission in the preface) of _The Blind Watchmaker_, Mr. D has been going through a sequence where he has been designing and predicating evolving forms of computer graphics; and in order to (properly, btw) qualify that his experimental example cannot be considered natural selection, he distinguishes it by calling it _artificial_ selection instead: by which he explicitly means that his program gets the results it does because he chooses the shapes, and even (an unusual admission for naturalistic proponent of evolutionary theory, and much to his credit) that if we used a pattern recognition computer program instead of a 'by-eye' method, it would still be us setting the parameters, and so in principle would not be bringing us close to natural selection.

But then there's a problem. If he holds to this, then he has to be implicitly but strongly affirming that the behavior he is engaging in is in fact qualitatively and significantly different from _natural_ behavior. Which of course he disaffirms very strongly elsewhere. But if it isn't different from natural behavior after all, then whence the distinction?

He himself appears to recognize this problem at some level during the example itself, because up to the middle of the same page he can be found going into panegyrics about these little monsters and their many sequels being explicitly "undesigned and unpredictable"!--when he himself knows and understands very well, and says as much shortly afterward, that they are no such thing. But his philosophical naturalism leads him to speak of these things as being that way anyway, when he isn't paying sufficient attention.

Again, he contrasts his own behavior with natural selection at the top of p 62 where he notes that by and large the selection criterion is nonrandom death. And of course, this is a blind, non-sentient process. So what exactly is Mr. Dawkins distinguishing between when he declares one set of selection 'artificial' and the other 'natural'? He himself is supposed to be totally a part of the blind interlockings of nature! But he makes the distinction anyway, as if he were not. How did he come to be separate, or partly separate, from nature? And if he isn't, why the distinction at all?

When an author says there is nothing but the blind interworkings of nature, and then goes on to describe his own behavior as artificial _distinct from_ natural, then I think it's fair to suspect him of applying a double-standard somewhere. More specifically, he recognizes intentionality to be something other than natural behavior, but is committed to _atheistic_ naturalism. So intentional behavior must be natural and only natural; but it must also be something distinctly different from natural behavior, and not only in a way that happens to be a largescale composite exhibition of smallscale behaviors. (He knows and accepts very well the principle of property transfer; the fact that atoms are non-sentient means beavers are non-sentient, too, appearances of ‘design’ notwithstanding. Design isn’t _really_ design--except when it’s _his_ design.)

The double standard occurs because he has to make the former assertion in order to keep up his atheism and naturalism together; but he has to make the latter position in order to be regarded with any seriousness personally on any topic even for purely practical purposes (such as appealing to a right to be fairly compensated for his work in writing TBW, for instance.)

Consequently, he flip-flops around a lot. {s}


Jason Pratt

Bilbo said...

Too many comments to read through them all.

I think a claim that "x is best explained supernaturally," can be falsified by offering a simpler account that adequately explains x naturally.

unkleE said...

Like Bilbo, I gave up about halfway down the comments. But I had two brief points about practical falsifiability (as opposed to theoretical):

1. Vic and others have argued many times that the Hume objection to a supernatural explanation for a miracle (or any other natural phenomenon) makes it practically impossible for disbelief to be falsified to a Humean.

2. If theism is not falsifiable, why did John Loftus and others to convert to atheism? Surely their change of mind shows that theism is indeed practically falsifiable, as John himself agrees? It's just that others of us conclude differently.

Gregory said...

The principle of "falsifiability" is, itself, unfalsifiable. If one were to succeed in falsifying that principle, then, as a matter of course, one would have merely proved it's truth. In which case, it would be impossible to disprove. Therefore, the falsifiability principle is riddled with self-referential stultification.

Flew's thesis is the secular version of the Ontological argument, in reverse. It's an attempt to disprove God via an a priori maxim. Perhaps it would be better to say that it attempts to define God out of existence.

Furthermore, I cannot falsify the laws of thought (i.e. A is A, A cannot be non-A, Either A or non-A), because any attempt to do so will, necessarily, implicate their existence.

What this inevitably leads to is, as Chisolm's Gifford Lectures have called it, the "problem of the criterion". For, while we can speak of a "meta" ethical justification of virtue, we cannot speak of a "meta" epistemic justification of knowledge, since any sort of "justification" scheme would already be implied in the knowing process.

Epistemic principles cannot, in the final analysis, be proven. Rather, these principles are the things by which we are able to proceed with "proof" in the first place.

I think this situation is conducive to theism, not because epistemic principles are "provable", but that they are actually undeniable. They are undeniable because they are not grounded in, or reducible to, a region of the material world. Epistemic principles are not spatio-temporal entities. If that were the case, then matters of "reason" would simply be reduced to being a particular vector of space; in which case, the "mental" is synonymous with whichever particular vector of space "reason" happened to be. So, we could apprehend an argument if, and only if, we happened to share the same locale as "reason"....whatever that could mean. And if we shared that same locale, then we would be the same thing!!

It seems to me that naturalism involves highly counter-intuitive notions of "mind", "reason" and "freewill", such that naturalism ends up being more miraculous and incredible than supernaturalism.

Anonymous said...

Here's a supernatural claim that is falsifiable.

"God has put large elephants in the Lattie F. Coor building in ASU."

William said...

Yes, as Platinga says in his Stanford Encyclopedia article:

"And while the proposition living things have been designed by an intelligent being is not falsifiable in isolation, the proposition an intelligent being has designed and created 800 lb. rabbits that live in Cleveland is clearly falsifiable (and false). "

J said...

The principle of "falsifiability" is, itself, unfalsifiable. If one were to succeed in falsifying that principle, then, as a matter of course, one would have merely proved it's truth.


Believers continually insist that any inductive criteria of Truth cannot itself be shown to be true, but they rarely take the time to think through their trivial analysis.

Falsifiability should be read as a strict sort of verificationism, or really evidentialism. Empirical research does often result in a modification of some particular fact claim, whether in terms of the physical sciences, or investigative journalism, or even history. No sane person could dispute that.

So, examining the "research process" we see something like evidentialism (or even falsificationism at times). That adequately describes the research process. Perhaps some scientist at times stumbles upon an empirical finding, or follows a hunch. But normal science (and research) does depend on finding facts, and at times revising/disposing with previous claims (ie falsification).

Obviously, ancient religious texts do not allow for this revisionism--that was Popper's point; dogma is not fact-based, but...dogma. The Book of Rev. claims that a jezebel rode on the back of a 7 headed beast (or virgin birth, or....resurrection itself). This is not ordinary knowledge; it's not part of the uniformity of experience. That doesn't falsify it per se, but given our normal experience (ie the uniformity of experience as Hume said) we have no good grounds to accept it,any more than we should accept reports of...Chupacabras; it's far more reasonable (given problems of testimony, etc) to doubt it, than to accept. Miraculous claims are NOT falsifiable (then, neither is ancient history itself--or even say history 50 years ago), but shown to be highly improbable. Hume in ways still a better guide than Popper's strict either/or.

(and given Hume's denial of scriptural inerrancy, it would be difficult to call him a believer in any traditional sense--also reinforced by his comments near the end of life: "the soul dies with the body", said DH)