About Me

My photo
Australian philosopher, literary critic, legal scholar, and professional writer. Based in Newcastle, NSW. My latest books are THE TYRANNY OF OPINION: CONFORMITY AND THE FUTURE OF LIBERALISM (2019); AT THE DAWN OF A GREAT TRANSITION: THE QUESTION OF RADICAL ENHANCEMENT (2021); and HOW WE BECAME POST-LIBERAL: THE RISE AND FALL OF TOLERATION (2024).

Saturday, January 14, 2012

Currently reading - Plantinga v. Dennett

Yes, I've been reading this little book by Alvin Plantinga and Daniel Dennett, which includes and expands upon their debate at an American Philosophical Association conference a couple of years ago.

Plantinga has a new book coming out in which, apparently, he expands further on his views. I'll need to look at this, though he's already expressed the same views in various other places.

As far as I can make out, the debate is about two theses that Plantinga wants to argue for. Although they're ostensibly arguing about whether religion and science are compatible, they end up arguing different points, because Dennett actually agrees that religion and science are compatible in the sense defined by Plantinga. I probably do, too, given the very weak thesis that Plantinga puts forward.

It's difficult getting a handle on what, exactly, the first thesis is. Irrespective of how he words it, the idea seems to be just that evolutionary theory does not logically rule out the existence of God and some role played by God in the process. As far as Plantinga is concerned, evolutionary theory and religion are "compatible" even if the former renders the latter less plausible, as long as it doesn't logically rule it out.

Based on that understanding of "compatible" ... sure. If you are prepared to do enough work you can prevent almost any small set of claims from being ruled out, logically, by pretty much any findings whatsoever, whether those findings come from the sciences or the humanities.

First, you should define your claims as narrowly as required to make them a small target. You can also do other things. You can defend claims about historical events where the historical record is patchy, but jettison any claims where it is not and where the evidence is strongly against them. You can introduce ad hoc hypotheses to explain any results that seem to falsify your claims. If sufficiently desperate, you can even engage in selective forms of radical scepticism, refusing to accept even very plausible propositions that seem to be evidence against you.

If you're prepared to do all these things, you may end up with a bizarre view of the world, but you'll be able to avoid any outright logical inconsistency between your claims and whatever scientific or historical, etc., facts you are prepared to acknowledge as well established.

Thus, great resources are available to people to avoid their religious beliefs - or at least a core of them that they consider most important - ever being logically ruled out. Therefore, it's not surprising that Plantinga is able to demonstrate to his own satisfaction that there is no inconsistency in his view of the world ... even while holding on to a Christian worldview and some of the basics of evolutionary theory. The main point that Dennett seems to be making in response is, "So what? None of this shows that religion is at all plausible in the light of our modern, scientifically informed knowledge of the world." Indeed, he suggests, many ridiculous ideas could be "compatible" with science in the weak sense that science cannot logically rule them out once and for all to the satisfaction of somebody who is committed to defending them.

Religious believers may have sufficient resources - such as those I've sketched above - to avoid their main claims being logically ruled out by evolutionary theory or anything else. And yet, what you end up with as you conduct the defence of religious claims may be a system of beliefs sufficiently ad hoc and just plain weird that you can no longer, in good faith, believe in it as a whole ... while it is completely implausible to a well-informed outsider. If you want to describe this situation as "science and religion are compatible" then go for it. But don't be surprised if many people think that such "compatibility" is beside the point.

In fairness to Plantinga, he does seem to understand that the plausibility of religion is on the line - it increasingly becomes a focus of the debate - and he makes an attempt to defend it.

Let me also add that many statements that "science and religion are compatible" involve an attempt to insulate them from each other entirely, such as by claiming that religion makes claims only about a supernatural world, while science makes claims only about the natural world. Accordingly, they can never conflict.

Or they are insulated from each other by a theory that religion is restricted to matters of morality and meaning, and the like, while science has no authority in these areas. According to this argument, no scientific finding can ever make a legitimate religious claim less plausible.

These are much stronger claims about "compatibility" than anything Plantinga defends, and I think they are pretty clearly false.

That leaves me with Plantinga's other thesis to comment on. More about that later.


Ben said...

Just getting these Christian philosophers to show up with a case is 95% of the battle. Then when they're there, pointing out how weak their case is that they were so apparently embarrassed to present is that last little bit. Religion...why you still with us?

Eamon Knight said...

Thanks for this, Russell. My feeling is that, in gnu circles, compatibilism often gets glossed over and dismissed as false a little too quickly. As you point out, theology can always be gerrymandered to be, at least, non-contradictory with scientific knowledge.

ISTM the the real incompatibility between science and religion is in their respective epistemologies. The former is based a fairly rigorous standard of empirical observation. And the latter has....what, exactly? (That's a rhetorical question).

Jean Kazez said...

I think you haven't gotten to the main course yet. Pantinga has something very surprising to say (he gave a talk based on the new book at SMU last year)--that rational belief in evolutionary theory REQUIRES religion, and that it's actually atheism, not religion, that's incompatible with rational acceptance of evolution. Happy reading;-)

Eamon Knight said...

@Jean Kazez: Hasn't he been using that argument for like, about 10 years now? Or has he got a new take on it that answers the objections?

Tom Clark said...

Eamon, Jean,

Yes, Plantinga's basic argument over the years has been the EAAN, the evolutionary argument against naturalism, see http://www.naturalism.org/plantinga.htm

According to EEAN, we can't possibly track truth or be rational unless we have some supernatural epistemic capacity bequeathed us by God during (guided) evolution. But we are rational, therefore evolution was guided and naturalism is false.

But it isn't naturalism that shows evolution to be unguided, it's science. As Russell and Dennett and others have pointed out, there are no good *scientific* grounds for supposing anything supernatural played a role in evolution. And we have perfectly defensible naturalistic accounts of truth-tracking and rationality.

Dave said...

As I read it, Plantinga thinks natuaralism says our individual observations are faulty and cannot yield true beliefs. It may be true that sometimes our faculties fail. See the gorilla experiment. But it is not our individual ability that sustains the claim of naturalism.

Science is a collective, self correcting enterprise (50% of the observers do see the gorilla). And it is this self correcting enterprise that confirms naturalism.

So Plantinga is wrong to jump from individual error-prone observations to the conclusion that god did it. Science is what concludes that evolution is true and requires only naturalism, on which, one would expect individual error-prone observations given the jury rigged nature of evolution. No god required.

Charles Sullivan said...

Plantinga does something similar is his response to the problem of evil. He simply claims that the existence of both the Omni-God and evil are not logically inconsistent, because God may have had his reasons (of which we know nothing) for allowing evil.

He refrains from engaging in a genuine theodicy because he thinks we cannot know the mind of God.

Rachels mentions this in Problems from Philosophy (3rd ed.), chapter 3, p. 29.

Russell Blackford said...

Yes, Jean - that's the second thesis that Plantinga defends. More on that later!

Russell Blackford said...

Charles, I talk about this in my own essay in 50 Voices of Disbelief. There may be some sense in which the existence of evil (suffering, etc.) is not inconsistent with the existence of an omnibenevolent God. We might say that it is epistemically possible that there's a solution the problem that we haven't thought of, or that it is logically possible that a loving God would have reasons for allowing evil, or some such thing. It does end up becoming a matter of what is plausible as the theists describe their God, offer whatever explanations of suffering, etc., that they can, and we contemplate the enormity of it all. Hence, my essay is entitled "Unbelievable!" not "Logically Impossible!"

Among the things that make Plantinga's views unbelievable to an outsider - or to an insider who has any real doubt and distance from her beliefs - are his resort to evil beings such as Satan to explain why evil existed before any acts of human free will (bracket off whether we do have free will in any meaningful sense) and his claim that theists (but evidently not atheists) perceive God through a special sensus divinitatus.

These claims are not logically impossible in themselves. Nor do I know any way of ruling them out with evidence. However, they are highly implausible, and any theology that resorts to them will seem unbelievable to most of us.

Eamon Knight said...

We might say that it is epistemically possible that there's a solution the problem that we haven't thought of,

There could always be "something we haven't thought of" that solves any difficulty in any theory, or that disproves any view we currently hold. But relying on such a slim reed more than minimally and temporarily is some combination of special pleading, wishful thinking and desperation.

If Plantinga seriously pushes that sort of thing, he's...pathetic.

Jason Streitfeld said...

Unfortunately, there's a strong tendency for many philosophers (some very prominent ones, like David Chalmers) to completely detach epistemological or phenomenological content from observable behavior. This makes it extremely difficult, perhaps impossible, to account for true beliefs in terms of evolutionary theory. However, these philosophers are not against evolutionary theory. Chalmers has even expressed disapproval at the way his own arguments have appropriated in religious arguments.

I've always held that beliefs and knowledge are primarily, maybe even entirely, understandable through observable behavior. This is just the simplest way of thinking about it, in my opinion. So it's easy for me to reject common arguments against physicalism (like the knowledge and compatability arguments) as well as Plantinga's argument against naturalism. I've blogged about these issues in some depth before, including a post about Plantinga's argument.

Jambe said...

I was recently (for the second time in a relatively short span) placed in a position that made me consider very carefully what I take for granted and what I'm thankful for.

One thing I'm very thankful for is the sort of considered discussion of philosophy and science I regularly find here (and at most of the "new atheist" hotspots generally).

So thanks, Russell. You're a trouper. I don't have the patience to mull over most apologetic philosophy, but I do gladly read thoughtful analysis of it. Thanks again.

Anonymous said...

Religionists who want a created universe to be compatible with evolution just have to make sure their claims contain very little information. If you have an omnipotent, omniscient god, and he wants humans to evolve, he doesn't have to "guide" evolution in an active interventionist sense, he just has to choose an evolutionary history that will result in humans from amongst all the other possible evolutionary histories. Within that history, natural selection will be seen to be doing its thing. Alternatively, religionists could embrace the multiverse concept and say that God simply let every history unfold, in the knowledge that this one would be amongst them. Of course, none of these claims make any attempt to explain how God did these things, or even what he is, in any remotely detailed sense. But that's religion for you - it shouldn't be confused with knowledge :-)

Richard Wein said...

Good point, Jason.

We can see beliefs as arrangements of brain matter that affect behaviour. There's no reason why such arrangements can't accumulate information which tracks reality, and no reason why a mechanism for accumulating such information can't evolve. Reality-tracking brain states allow more effective behaviour, and therefore such a mechanism can be selected for.

Plantinga considers several possible naturalistic accounts of beliefs, but I think he's caused a lot of confusion by the fact that in his popular presentations he deals with only one or two of them, not clearly identifying them, and often not including the most relevant one. For example, when I heard him in a radio interview he seemed to consider only the possibility that beliefs are an epiphenomenon, having no causal correlation with behaviour. But he didn't identify this assumption or mention that he considers other possibilities elsewhere.

Fitelson and Sober (http://fitelson.org/plant.pdf) indicate that Plantinga addresses 5 possibilities (section 1.3) of which the most relevant is this: "(v) beliefs cause adaptive behaviour". It's in addressing this possibility that Plantinga deploys his argument that false beliefs are just as likely to cause adaptive behaviour as true ones. A caveman may run away from a sabre-toothed tiger because he thinks it's part of a game, rather than because he knows the tiger is dangerous. True, sometimes people may do the right thing for the wrong reasons. But in general we should expect reality-tracking beliefs to be better than non-reality-tracking beliefs in making us effective in dealing with reality.

Verbose Stoic said...


"I've always held that beliefs and knowledge are primarily, maybe even entirely, understandable through observable behavior. This is just the simplest way of thinking about it, in my opinion."

The problem is that you run into this odd situation: I can act as if I have a belief or phenomenal experience that I do not, in fact, have. If you consider only third-person observable behaviour to indicate what my beliefs are, then you run into the nasty problem that because I'm essentially lying to you you would conclude that I have a belief that I do not have, and you would make that conclusion despite my legitimately knowing otherwise. You can argue that your observations include the first-person ... but then you walk right back into the problem that you are claiming you avoid by limiting it to observable behaviour.

Verbose Stoic said...

I think the comparison to the Problem of Evil really highlights why those sorts of arguments are being used. Both evolution and the Problem of Evil have been used as attempts to DISPROVE the existence of God, to make it so that it can be said that we know that God doesn't exist or at least should not believe that God does not exist. But then as disproofs all that's necessary is indeed to show that the disproof doesn't work, and that it is possible -- well, reasonably so -- for God to exist in light of that evidence and argument. If it is possible for your premises to be true and your conclusion false, it doesn't work as a disproof -- at least not to the level of knowledge -- and then we have to talk about what we can choose to believe and how that should work. But it's quite hard to demand that someone must give up a belief they hold if you don't know that it's false.

The same sort of strategy is used against people who oppose evolution. Take, for example, irreducible complexity. Almost all of the counters to irreducible complexity claims do not, in fact, actually show what evolution actually did. All they do is say that evolution COULD produce that complexity by small hops. And this is a valid argument because all the defender needs to do is show that it's possible for evolution to produce, say, an eye, not that it actually did it. The same, then, can apply to the disproofs of the Problem of Evil and evolution.

If you treat the two theories from a neutral starting ground, then you might be able to make a claim based on neutral plausibility. But I deny that any such neutral starting ground exists. We always determine plausibility based on our own beliefs and what we consider to be the case. Given that, I ask if the disproofs are plausible enough to force me to abandon an existing belief, and I don't see that they are.

Jason Streitfeld said...

Verbose Stoic, I think that's the objection to behaviorism that David Chalmers primarily relies on. In any case, I don't think it's a forceful objection.

First, the very fact that people can mislead people about their beliefs through their behavior is evidence in support of behaviorism: It implies that people understand beliefs (at least to some extent) on the basis of observable behavior. The question then is whether that is the full extent of our understanding of beliefs, or if there is something more. I think Ockham stands in favor of behaviorism.

Second, the objection relies on an impoverished version of behaviorism, where mental contents are supposed to have a one-to-one relationship to observable behaviors. The more robust forms of behaviorism aren't like that. The same behavior is not necessarily evidence of the same belief.

I'm a fan of Ryle's view, where attributions of belief and knowledge are "inference tickets": they give us license to make certain predictions/explanations of observed behavior without directly corresponding to any particular physical state or event (neurological, behavior, whatever.) This is the basic idea behind Dennett's "intentional stance," even though he tries (or tried) to distance himself a little from Ryle. The Rylean view requires us to be pragmatists (perhaps pragmatic realists, as opposed to naive realists) about mental contents. We shouldn't reduce mental contents to brain states, but we shouldn't reduce them to specific behaviors, either. They aren't things at all, apart from the predictive frameworks in which we attribute them. As such, they relate directly to observable behavior, and cannot be understood in any other terms.

As for subjectivity and first-person access, Ryle's view (and Dennett's too, I think) is that our occasionally having private access to "the truth" about our beliefs is contingent, a matter of how our bodies are built, and not a theoretical necessity built into the nature of mental contents. Thus it is not surprising that other people are often in a better position to determine our beliefs than we are.

Unfortunately, I haven't read the Plantinga/Dennett exchange, nor have I listened to the original debate. I'm curious to know if Dennett's taken up the sort of argument I'm advocating.

Jason Streitfeld said...

I forgot to comment on Plantinga's first thesis, and Russell's reaction to it: I think it's conceding way too much to claim that God is a logical possibility. What, exactly, is "God" supposed to mean there? I'm not saying no coherent answer to that question is possible, but I wouldn't assume that a coherent answer is forthcoming. More, I don't think that any coherent notion of God is capable of doing the work required of it by the world's major religions. I rather think that theism is at root, and by necessity, unintelligible.

Bubblecar said...

"I rather think that theism is at root, and by necessity, unintelligible."

I agree, Jason. As I argued above (I'm the Anonymous up there^) creationist claims can be "compatible with science" only by being virtually devoid of any information. The believers never make any attempt to define God beyond "some kind of omniscient, omnipotent, immaterial being who created the universe" and provide us with no scientifically meaningful definitions of these characteristics, let alone descriptions of how they work. It's not enough to say this immaterial god (whatever that's supposed to be) is "all powerful" and can thus create a material universe - there must be *some way* that he did this, and asking for a reasonably detailed description of it is just asking for some way to rationally and/or scientifically assess this claim.

Those believers who say these things are simply "unknown" or "unknowable" are really saying: "we can offer no information content for the basic terms (God, Creation, immaterial etc) we're using, and so might as well be talking about nothing." What these believers actually "believe" is very much less than they've somehow convinced themselves.

Legal Eagle said...

I have just been reading 'The Passion of the Western Mind' by Richard Tarnas, and it beats me how anyone who is religious can reconcile those beliefs with scientific belief. The only way one could possibly manage it is to say that science is about empirically verifiable evidence and religion is about non-empirically verifiable faith (an inner subjective belief) - in which case it has to inhabit a separate part of the brain...and therefore never the twain shall meet. Or, as others above have commented, the concept of "God" is so watered down as to be meaningless anyway.

Jason Streitfeld said...

I just found an interesting response to Plantinga's EAAN: http://www.infidels.org/library/modern/wesley_robbins/contraplantinga.html

It's very similar to my own.

Verbose Stoic said...


On your first point, no, the misleading example doesn't support behaviourism because no one is denying that beliefs play a causal role in our behaviour, or that if I want to try to get at what someone else believes I'm limited to looking at their third-person observable behaviour. That's not where the clash between the theories is. The clash between the theories is over what it means to be a belief, and the third-person observable behaviour or a set of inferences about what behaviour you might see are not what it means to be a belief, or else you run into the real objection which is that you are committed to saying that what you observe as the belief even in misleading cases is what they really believe, misleading notwithstanding, which is a bit odd.

On the second point, you can add as much into behaviour as you want and I'll still be able to fake having a belief that I don't have, so it doesn't save you unless you include things that will leave the door open for the kinds of moves you say you've escaped with that view.

Note that even people like myself and Plantinga can adopt the intentional stance as a useful stance, and that you can also take a functionalist position on belief which defines it by the role it plays without having to get down to physical states or be strictly behaviourist. However, that functionality will have to include the first-person stuff to actually accurately describe things like misleading.

Finally, no one needs to hold that one has perfect access to their own beliefs. Even conceding that, we can still say that sometimes we can get beliefs from the first-person view that you can't get from the third-person view.

Jason Streitfeld said...

The issue is the degree to which our understanding of beliefs is a matter of observable behavior. If you concede that our understanding of other people's beliefs is limited to observable behavior, then you grant that, for the vast majority of cases, we are all behaviorists. The question Ockham raises is, why claim that our understanding of our own beliefs is any different?

Maybe because of private access?
I noted earlier that there are cases where we have private access to our own beliefs. We certainly can mislead people about our beliefs. But that doesn't mean our understanding of our own beliefs is of a different sort than our understanding of other people's beliefs. It just means that we often have better information about our selves than what is available to other people. But how do we get that information? From observing our behavior.

We have access to our own behaviors which nobody else has. This is a matter physiology--as I wrote earlier, it's because of the way our bodies are built, and not an essential feature of mental contents.

So how do I account for the first-person point of view? Well, I think the logic of first-person language is complex, involving (at the very least) issues about social identity, agency and responsibility, as well as epistemological and phenomenological issues. I don't have a simple account of first-person discourse on hand to offer. But I think it's clear enough that we do have privileged access to certain of our behaviors, and this allows us to often (though also often not) form uniquely accurate predictions about ourselves. We can explain why it is possible to mislead people about our beliefs, and we can do so in terms of observable behavior.

By the way, I'm not advocating a form of functionalism. In my view (and Ryle's view), beliefs are not things which could play a functional role in any sort of process. Beliefs aren't things in that sense.

Andgnat said...

I am currently reading Hitchens "God Is Not Great: How Religion Poisons Everything". He makes some very salient points against the marriage of science and religion. While conceding that the origins of science were founded in religion, Hitchens goes on to elaborate on how the church then chastised and punished those that studied science if the material contradicted church idealogy.

The contradictions are clearly and logically explained by Hitchens and seem to also dismiss Platingas fence sitting mentioned here.

Jason Streitfeld said...

I should also mention that, while it is obvious that we can fool others about some of our beliefs some of the time, I don't think we can fool people about all of our beliefs all of the time. That's the difference that undermines Plantinga's argument.

Verbose Stoic said...


"The issue is the degree to which our understanding of beliefs is a matter of observable behavior."

No, actually, it isn't, at least not for me and I suspect not for Chalmers and anyone else you're calling out on this. It's about what beliefs really are, not about how we get access to them.

Returning, though, to discussions more directly relevant to Plantinga, recall that your initial comment was that your model allowed you to avoid the sorts of misunderstandings that he uses to build his case. I don't think that you can build a model that works to explain misleading and doesn't leave him the same sort of moves. Let's divide up "observable behaviour" into two main types: publicly observable and privately observable. Now, if you claim that we observe our own behaviour only through publicly observable behaviour, you can't explain misleading without taking the odd position that if the public observable behaviour says "Believes X" then that's what they believe regardless of the privately observable behaviour (inner speech, deliberation, introspection, etc, etc). You seem to be trying to avoid that by claiming that you can include privately observable behaviour in your behaviourism. Putting aside that it seems like you then don't really have any difference from the models you're rejecting, this move is what allows Plantinga to make his move. After all, it allows for having two different sets of beliefs -- identified by differing private behaviour -- that are identical at the level of public behaviour. Evolution can only select on the basis of public behaviour, and thus can't select between them. Add the not-so-odd assumption that at most one of them can be right, and you've opened the door to a set of false beliefs being selected for by evolution because they happen to mimic the same behaviour as the true set would. At which point you've hit Plantinga's starting point.

Verbose Stoic said...


"I should also mention that, while it is obvious that we can fool others about some of our beliefs some of the time, I don't think we can fool people about all of our beliefs all of the time. That's the difference that undermines Plantinga's argument."

1) In principle, you absolutely can, as long as you know how to act in all cases.

2) Plantinga's argument doesn't rely on fooling, but on building a consistent set of false beliefs that happen to produce the right public behaviours most of the time. CBT suggests that a lot of mental problems are produced by just those sorts of consistent belief systems that break down in specific cases. If those cases are never hit, evolution could not filter those belief systems out since they'd be successful.

Jason Streitfeld said...

Verbose Stoic,

As a matter of fact, I think the main problem with EAAN is entirely to do with the extent to which our beliefs are understandable in terms of observable behavior. That's the argument I laid out against Plantinga on my blog. (See earlier post for link.) You disagree, which is fine, but I don't know why you disagree. Again, see the link to my blog post for my argument. I'll be happy to try to explain myself further, if you still want me to.

You argue that if I allow for a distinction between privately and publicly observable behavior, then my position is no different from Plantinga's. I do acknowledge such a distinction, but it is not the same as Plantinga's. In my view, the public/private distinction is a practical one. There are practical limitations on what is publicly observable. Since there is nothing about our private behavior which it inherently private, private behavior can still be selected for.

As for being able to mislead people about all of your beliefs all of the time: No, I really don't think so. I'm well aware that Plantinga's argument is not about fooling people, but about having a set of completely false beliefs. My point was that, if you understand why you cannot mislead everybody about all of your beliefs all of the time, then you will understand what is wrong with Plantinga's argument. It is not possible to live and communicate in a community with a shared language without sharing at least some beliefs, and without those shared beliefs being known.

Jason Streitfeld said...

Also, Verbose Stoic, you might be forgetting about Quine's thesis about the indeterminacy of translation. Broadly put, it states that the same physical evidence can be interpreted equally well by two or more incompatible explanatory frameworks. I mention this because you suggest (though don't quite formulate) such a hypothesis, but you add "the not-so-odd assumption that at most one of them can be right." But Quine's point--and, indeed, one of the basic ideas of pragmatism--is that there is no sense in claiming that at most one can be right. If they both are equally good at explaining the evidence, then they are both equally right.

You might say this is absurd, because it means that a person from the "outside", as it were, is in just as good of a position to interpret your behavior as you are. But remember the stipulation: the two interpreters make equally good predictions about your physical behavior. So, yes, if another person can predict your own behavior just as well as you can, then what basis do you have for claiming that they are wrong? You might say that you have been fooling them, but that is something they will have predicted, too, isn't it?

Perhaps it is impossible to imagine being able to predict people's behavior that well, but the theoretical point remains, and it is a great example of pragmatic reasoning.