Defining terms

I left out truth in the particular place you mentioned because I said "at least".

For a belief to count as knowledge, it must be justified. It must also be true. These are necessary, not sufficient conditions. Also, it's often been pointed out that the justification has to be reality-tracking in a sense that is very difficult to nail down (read any encyclopedia entry on "Gettier cases" to get the idea).

But yes, for a belief to count as knowledge it must at the very least be justified. You can't justify a belief by saying: "It's in my holy book", or "I got it in a drug-induced trance" or "I had a mystical experience" or "it's church doctrine". Beliefs that happen to be true might be obtained in this way by one person or another, from time to time, but these are notoriously unreliable ways of forming true beliefs, so they can't be used in justification.

It's very misleading going around suggesting that these are "ways of knowing" in any sense comparable to science, historical scholarship, asking a trustworthy person about something they are likely to be correct about, seeing with your own senses, and so on. In particular, revelation is not a way of knowing comparable to science, etc.

It would be if it involved genuinely receiving information from a god, but if that were the case it would be reliable on all things, not just moral issues, or whatever you'd want to restrict its reliability to. Or, if not, you need to buy into some kind of controversial theory as to why the god speaks truth on some matters but not others (perhaps tailoring the message to the sophistication of the audience). But once you start saying that you are getting into theological waters where it's often better not to go if you're not a theologian. It's taking sides on a substantive theological issue.

Rosenau wrote:

To return, then, to ways of knowing, I'll define them as systematic methods of evaluating truth claims against new sources of knowledge...

And how, precisely, is "literature" in any way a "systematic method[] of evaluating truth claims against new sources of knowledge?" Will you be defining the word "literature" in such a way as to include only those works which are serious examinations of serious questions, and excluding all works which consist of stuff the author just made up for fun? Since it's inconvenient, we can ignore the fact that serious literature often disagrees with other serious literature regarding their truth claims.

But more importantly, Mr. Rosenau, you seem to be making the same mistake in these latest two posts that you were making four posts ago: confusing the conveyance of knowledge (or "delivery" of knowledge, to use Kazez's term) with the discernment of knowledge.

For example, the Bible may convey many truths, but it doesn't help in the matter of telling truth from falsehood (in fact, it muddies the waters by giving out self-contradictory advice on discriminating between the good and the bad).

Literature, art, golf, chess (etc.) may all convey truths, but they aren't tools for learning what's true and what's not. That is the goal of science, and that's what makes it a "way of knowing" while all those other human endeavors are not, no matter how many truths they actually do deliver, or how much meaning people derive from them.

That's why we don't teach literature in science classes, which you should already understand, given your job and your studies.

For a belief to count as knowledge, it must be justified. It must also be true. These are necessary, not sufficient conditions.

It's a matter of great debate in epistemology whether justification is a necessary condition for knowledge. It was once thought so, but now there's no consensus. According to "externalist" theories, it is not. A reliabilist, for example, says true beliefs are knowledge if they are merely the outcome of reliable belief formation processes. You don't have to be able to give the justification for a true belief, for that belief to qualify as knowledge. Everything Josh says in the post suggests he has reliabilist intuitions.

However you define "knowledge", I'm not sure why it matters whether people get capital "K" knowledge from reading fiction, or just some true beliefs. It's good to have true beliefs just so that our behavior meshes with reality. I don't think the value of having true beliefs always turns on their qualifying as knowledge.

I would still like a definition of 'science' and 'religion'.
I am at a loss how you can imagine them to be "less central to the issue"!

Jean, you're missing the point. Like it or not, most people think of "knowledge" as something stronger than just "belief". If Eugenie had said "there are other ways of believing" it would sound odd, but I'd have no great problem with it. But even if we adopted the highly controversial claim that knowledge means only "true belief", she said that such things as drug-induced trances and divine revelation are ways of "knowing". By any definition they are not.

They are not ways of believing things that are likely to be true. They are not ways of reliably forming true beliefs. This whole business about "other ways of knowing" tends to put supposed divine revelation, untested hunches, drug trances, mystical insight, etc., on the same footing as good, reliable means of looking for truths such as scientific testing, historical scholarship, reliable testimony, etc. It's a pernicious way of talking.

Of course a novel may impart some information, etc. But writing a novel is not a "way of knowing". Reading a novel may give some moral insight, or it may do the exact opposite. I'm not denying that writing novels is interesting and fulfilling work (having written a few myself and even made some money from it). Nor am I denying that reading novels can be valuable, or that reading them critically can be valuable. But that was never what this argument was about. Josh took it off at 180 degrees. If people read the Bible critically in the same way that a literary critic reads a novel, I'd still be skeptical about using a phrase such as "other ways of knowing", but there'd be no great problem. The problem is that most Christians give the Bible, church tradition, etc., far greater epistemic and moral authority than critics or general readers accord to novels and novelists.

To return, then, to ways of knowing, I'll define them as systematic methods of evaluating truth claims against new sources of knowledge, whether those sources be experiential, introspective, logical, or rooted in testimony from others.

That seems very nearly post-modernist.

Here's what I see going on: you define a "way of knowing" as pretty much anything. It could be science. It could be fundamentalist religion. It could even be reading tarot cards. I don't see anything in this definition that concerns the correctness (or even degree of truthfulness) of the knowlege thus gained.

We non-accomondationists, however, are very concerned with the degree of truthfulness of knowledge. I (and, I presume, other non-accomodationists) would stipulate that the application of a "way of knowing" must result in actual, verifiably correct knowledge, or at least result in knowledge that is closer to being correct than what was known beforehand. Otherwise, we wouldn't consider it to be a way of knowing, but rather a way of believing.

Does that make sense?

I hope that posting these definitions produces a better dialogue, and I hope that Jerry Coyne follows suit and posts his own definitions. It seems like you two have been arguing past each other to a degree, and I think that these definitions will help alleviate that.

Jean, you're missing the point. Like it or not, most people think of "knowledge" as something stronger than just "belief". If Eugenie had said "there are other ways of believing" it would sound odd, but I'd have no great problem with it. But even if we adopted the highly controversial claim that knowledge means only "true belief", she said that such things as drug-induced trances and divine revelation are ways of "knowing". By any definition they are not.

I'm not saying any contemporary epistemologist claims that knowledge is just true belief. I'm saying there are views that get rid of justification as a condition on knowledge, and put something else in its place. For example, reliabilists hold that true belief can't be knowledge unless it's the output of a reliable process. That process might be opaque to the believer, so incapable of giving him "justification."

I wasn't addressing anything else you said--just this simple definitional matter. You said "for a belief to count as knowledge, it must be justified," and that's not the view of all epistemologists.

Jean, you're missing the point. Like it or not, most people think of "knowledge" as something stronger than just "belief". If Eugenie had said "there are other ways of believing" it would sound odd, but I'd have no great problem with it. But even if we adopted the highly controversial claim that knowledge means only "true belief", she said that such things as drug-induced trances and divine revelation are ways of "knowing". By any definition they are not.

I'm not claiming that any epistemologists define knowledge just as true belief. I'm claiming that some leave out justification as a condition of knowledge and replace it with something quite different. For example, reliabilists say a true belief is knowledge if it's the output of a reliable process, which may be opaque to the believer, and incapable of giving him justification. This might be helpful to Josh's point, but might not. I'm just commenting on terminology (since the post is after all about "defining terms.")

A am one of the non-accomdationists, as mentioned by Josh.
I certainly do not get all my information from science.
By the way, my definition of science is not of a collection of 'facts' but of a method - "the method we use to determine whether an idea about the natural world is incorrect".
The ideas can come from many sources, science, literature, movies, dreams, religious stories, etc.
There is lot limitation, whatsoever about the source.
While I can agree that each of these ideas received from these various sources may be defined as 'truth claims' I cannot see them as being classed as 'knowledge' without being subjected to verification - and in the case of ideas about the natural world or actions related to the natural world the verification process we use is the scientific method.
The story of Kekule's snake chasing his tail is a prime example. The reason we know of it is not because it was magical or contained such beautiful imagery. No, it is still talked about because it turned out to be true. It was verified using the scientific method. The daydream produced an idea that could be tested and it was found to conform to the natural world.
Science is perfectly open to new ideas like this that come from many sources. Even a revelation from 'God' can result in ideas that can be tested. For example the Mormon revelation that the Native Americans originate from a boatload of Israelites is testable by science.
The hypothesis that Native Americans are descended from Israelites is not a different way of knowing, it's simply yet another idea that science has shown to be incorrect. Is there another way of showing ideas about the world are incorrect?

Moderation gobbled my first attempts. Let's try it without any html tags.

Russell, Nobody denies knowledge is more than belief....certainly not me. By all accounts, it's true belief. And by all accounts, it's even more--it's true belief with some sort of basis. I was merely pointing out that the "basis" part isn't universally thought to be "justification," if that means a consciously grasped set of evidence. As I said, there are reliabilists who drop the "justification" condition on knowledge. This is merely an attempt to be helpful in the endeavor to "define terms"--which I take it is the goal of this post.

And in case I'm going to be misunderstood again--reliabilists drop justification and put something else in its place! The SEP article Josh mentions probably goes into all that.

I don't follow this post.

First, we get an account of knowledge which leaves out truth. That's strange enough. (Maybe truth sneaks in under 'reliability', but then reliability gets inexplicably treated as an internalist component of knowledge! Or is truth supposed to be internalist now??)

But then we get the statement that knowledge is a truth claim! Where on earth did this come from? I mean, I thought truth was getting left out of knowledge. And whence 'claim'?: surely a person can know something without making any claims.

And then a monumental non sequitur: that truth claims are claims about experiences. It's as if the experiential stuff showing up in the original justification-heavy conception of knowledge has inexplicably migrated over to the recently introduced truth component. And isn't this just positivism? What exactly is wrong with making truth claims about entities which transcend experience? Don't people do this all the time? Any account of truth claims that ignores this phenomenon is, I would have thought, an unsuccessful account.

Jean, what you're saying about reliabilism isn't quite right. People disagree over whether reliabilists are providing a substitute for justification (as you suggest) or just an account of justification. For its part, the SEP article adopts the second approach: reliabilism is treated as providing a theory of justification, not as providing an account of knowledge that leaves out justification.

I said "there are reliabilists who drop the justification condition"). Yes, there are other options. The reliabilists I know best drop the justification condition. The whole matter is thoroughly covered in the SEP article called "reliabilism".

@ 12 (Dave2),

First, we get an account of knowledge which leaves out truth. That's strange enough.

An epistemology that gives primacy to knowledge, belief, justification, etc, over truth actually seems more useful, frankly. Any examiner of beliefs suffers from the same fundamental limitations as the original believers themselves: she does not have unmediated access to "truth", she has access to beliefs, evidence, and justifications (quite possibly more and better, but even so, not truth).

Therefore, regardless of the nature of truth and knowledge, what we actually witness are claims of belief and justification.

Jean, I was going off of this comment: "And in case I'm going to be misunderstood again--reliabilists drop justification and put something else in its place!"

I guess you were misunderstood again?

MPL, if a proposed account of knowledge ends up allowing for someone to 'know' something which is in fact false, then pretty obviously that's an unsuccessful account of knowledge. So leaving out a truth condition seems like a recipe for disaster.

Also, I'm not sure why you're looking for 'usefulness' in an account of knowledge. What people generally look for is a set of necessary and sufficient conditions for knowledge, regardless of whether it's easy to tell when those conditions are met. After all, there's no rule that says knowledge must be easily confirmed as such.

Dave2,

I'm looking for "usefulness" in an account of knowledge, because I believe that for the concept to be interesting (to me), it must be applicable at least some of the time. If knowledge is properly (whatever that means) justified true belief, then I am never entitled to say that I (or someone else) know something, only that I believe I know something, because strictly speaking, I only believe that I know something.

You're welcome to define "to know" in such a way that you can never use it, of course, but it sounds rather unfruitful to do so. I would rather discuss whatever it is that people mean when they say they know something, and when they should and shouldn't say so.

"...this discussion takes off from a talk Eugenie Scott gave, a talk which I didn't attend"

Why don't you just ask Scott what she meant when she talked about "other ways of knowing"?

It seems like this exchange could be put back on the road towards productivity if someone would only _ask_ her. Why is that such a problem?

I shouldn't read your intellectual posts late at night, they don't make sense to me. I think there are plenty ways 'of knowing'

Russell: I appreciated your leaving "true" out for reasons that I explained. Attaching "truth" to the claim would, IMHO, make it impossible to know whether anything could be considered "knowledge," and the whole discussion would collapse. Reliability can be assessed in lots of ways, and as Jean notes, provides either an alternative to justification, or at least a particular way to determine if a bit of knowledge is reliable. That's why witnesses on the stand have their credibility and reliability attacked. The logic there is roughly: if you can trust this person on lots of other details about the circumstance in question, why not also the bits that can't be verified?

Thus, religious claims need not be reliable about scientific topics if those topics are not what you expect religion to tell you about. If the religious normative claims match other sources of moral and normative evidence you've accumulated, you might judge that religion to be reliable on such topics, and might deem knowledge gained from that religion as justified (though still subject to additional evaluation).

You say it's unfair to compare religion with "science, historical scholarship, asking a trustworthy person about something they are likely to be correct about, seeing with your own senses, and so on." But religious training starts by "asking a trustworthy person about something they are [considered] likely to be correct about," and proceeds to direct experience with religious practice. It sometimes leads to scholarship of one sort or another. So I think we're splitting hairs here.

DaveW: "Will you be defining the word "literature" in such a way as to include only those works which are serious examinations of serious questions, and excluding all works which consist of stuff the author just made up for fun? Since it's inconvenient, we can ignore the fact that serious literature often disagrees with other serious literature regarding their truth claims."

No and no. Stuff authors make up for fun contains truths to. We learn, for instance, what the author considers fun. Alice's adventures in Wonderland are rather different from John Ringo's adventures. Reading John Ringo's adventures (with the requisite "OH JOHN RINGO NO"s) teaches a great deal about what sort of person not to be in general.

And yeah, authors often disagree. Scientists often disagree. And yeah, religious people disagree, too. Each have means of resolving those disputes (any might resort to namecalling or bashing in of heads, but more acceptable modes include: write a different novel, or a critical review; conduct new research or reanalyze old data; provide scriptural basis, historical evidence, or examples from the lives of moral people, perhaps even oneself).

Russell replies to Jean: "Nor am I denying that reading novels can be valuable, or that reading them critically can be valuable. But that was never what this argument was about. Josh took it off at 180 degrees. If people read the Bible critically in the same way that a literary critic reads a novel, I'd still be skeptical about using a phrase such as "other ways of knowing", but there'd be no great problem. The problem is that most Christians give the Bible, church tradition, etc., far greater epistemic and moral authority than critics or general readers accord to novels and novelists."

First, those first group exist, so it's wrong to tar all religion just because some people take a different position. Second, I keep coming back to Huckleberry Finn, which novel strikes me as a powerful argument against racism and slavery, and one which at least makes it problematic to draw a bright line between morality drawn from literature and morality drawn from religious sources. I think people draw on a range of interactions with the world to inform their moral compass, and even if religious text and tradition represents the accumulated wisdom of several millenia of people like me, that's worth according a good deal of epistemic weight to it, and that's without even touching on supernatural interventions, etc.

JoshA: "We non-accomondationists, however, are very concerned with the degree of truthfulness of knowledge. I (and, I presume, other non-accomodationists) would stipulate that the application of a "way of knowing" must result in actual, verifiably correct knowledge, or at least result in knowledge that is closer to being correct than what was known beforehand. Otherwise, we wouldn't consider it to be a way of knowing, but rather a way of believing."

Fair enough, and I don't think we disagree that much. I think truth is something you approach asymptotically, and that no belief can be guaranteed to be entirely true, but yeah, we ought to get closer to the truth rather than farther from it. The truth exists, however imperfectly we perceive it. Science is nice because we can all look at the evidence together and agree what it actually shows, even if we wind up interpreting it differently. And if we do differ in our interpretations, we've got a shared reality in which we can devise new tests to help us choose between our options. Literature, religion, philosophy, etc. are harder to tie to a shared reality. Some people shop around and find themselves in a church that "just feels right." I don't know what that means, but I can't very well deny that they feel that way, even if the place makes my skin crawl. So does my feeling falsify someone else's feeling? Are we dealing with different truths on a religious level? Does that discussion need to move to a different level of abstraction? Is it all meaningless and should we go get a beer? If we're committed to the notion of objective truths, we want to skip the penultimate choice at least, and the first seems authoritarian. So I prefer to broaden the discussion and look for some different level of commonality, in hopes that we're narrowing in on a similar truth by different paths (if you'll allow a metaphor).

So my quibble, I guess, is with the idea that anything not verifiable is not knowledge, unless you use a definition of verification big enough to drive a truck through. It sounds like you're applying standards of science outside its realm, then declaring that nonscience things aren't knowledge simply because they aren't science, which feels circular to me.

Dave2: "But then we get the statement that knowledge is a truth claim! Where on earth did this come from? I mean, I thought truth was getting left out of knowledge. And whence 'claim'?"

"Claim" because "truth" isn't something we can know absolutely. We can make a claim about what we think the truth is, of course, and when we do so with some sort of reliable basis (I'm not an epistemologist, and I know I'm mangling terminology, sorry), some sort of justification, we have knowledge. We'll continue to evaluate that claim in light of new evidence, updating it or scrapping it as appropriate, hopefully approaching the truth ever more closely.

Joshua Rosenau wrote:

Stuff authors make up for fun contains truths to. We learn, for instance, what the author considers fun.

But we don't learn that through any "systematic method[] of evaluating truth claims against new sources of knowledge." The authors simply tell us what they consider to be fun through demonstration or explanation (or a mix).

And again, containing truths (excuse me, truth claims) is different from discerning them, which is what science does. My high school yearbook contained many truth claims, but it (by itself) would be utterly worthless in determining the truth value of any of those claims.

Dave W.: "The authors simply tell us what they consider to be fun through demonstration or explanation (or a mix)."

Unless the narrator is unreliable. Reading literature sometimes means getting past the surface meanings and contending with contradictions in the text. Cf. "negative capability." And in those cases, I think we are discerning truths, perhaps things that the author didn't intend to include, or didn't even consciously intend. That's how it's possible to read Shakespeare through the lens of queer theory or Freudian psychology and discover something new. I think you're being too reductive about the meaning of a (literary) text.

Josh, thanks for your thoughts. But there is still a problem with your account, I think, because I don't think it ends up being compatible with accommodationism. I tried to explain my troubles in a previous post yesterday. I'll try to rephrase the worries here.

We're getting caught up in truth-claims. But if we want to be talking about ways of knowing that are unique to a religion, then we first have to be referring to claims that are true through the text. Unfortunately, texts themselves don't tend to give us a hint about how we're supposed to interpret them (i.e., literally, allegorically, etc.) When it comes to conversation with real people, we can figure out whether or not a person is being non-literal (i.e, sarcastic) by certain hints when we are charitable interpreters in the most robust sense. But when it comes to millenia-old texts, even when we are charitable there are not very many hints for whether or not some interpretation is going to introduce false positives or eliminate correct readings.

When interpretation is that pliable, it is on the same plane as Harry Frankfurt's sense of bullshit: there's no clue as to what kinds of truth are supposed to make sense of this stuff. As a consequence, we're stuck with the same problem as the ad hoc vampire examples. They themselves may lead us into knowing something that is new and true, but they are not true through the text. Unless other people independently affirm my interpretations, the interpretations are true because I'm plastering my own true beliefs onto the text. This kind of inspired revelation is not itself religious. (Though I suppose we can sometimes call it "spiritual" if we enjoy that kind of talk. If we don't, then "clever" is a close synonym.)

Obviously this isn't good enough to discuss ways of knowing that we can attribute to religion. So we can add in social mediators, like religious communities, and look for cues in their conventional or inevitable interpretations of the texts. i.e., we don't take the Torah on its own, we also consider the Talmud.

Fundamentalists are literal interpreters of texts, which means they have a "real world" sense of truth; and that's wrong and superstitious, everyone agrees. Non-fundamentalists are allegorical interpreters of texts, which means they have an "imaginary world" sense of truth, in such a way that the imaginary world is supposed to tell us something about our own world, usually about moral conduct in our world. But allegories can be false, too, if they produce conduct that is immoral, either by design or by inevitable idiosyncratic charitable interpretations of the text.

Hence the rub. Nobody is let off the hook when we all agree to condemn fundamentalism. The critique must go further. What's instructive is that a pluralistic "ways of knowing" account is what sets the stage for the case against accommodationism.

Josh Rosenau wrote:

That's how it's possible to read Shakespeare through the lens of queer theory or Freudian psychology and discover something new. I think you're being too reductive about the meaning of a (literary) text.

And I think that since different "lenses" offer different "truths," you've shown that those "truths" don't actually come from the literature, but from the reader.

Creationists claim these days that they are looking at the same evidence as science does, and simply interpreting it differently, when what they're really doing is throwing the rules of science out the window in order to arrive at preconceived conclusions.

Since there seems to be no systematic method of interpreting literature such that every reasonable reader comes to the same conclusion regarding the meaning of a work, everyone is free to bring whichever "lens" they want and walk away with whatever meaning they please. If anything, the "shared text" is thus a mirror of each reader's individual biases, and literature is an exercise in introspection by the readers where (for example) the Harry Potter stories become a scathing condemnation of U.S. foreign policy or a tribute to French wine-making or whatever else one might want to see in them. These "truths" are awfully maleable.

Dave W.: "These 'truths' are awfully malleable."

Or at least personal, and therefore somewhat subjective. That's what make these sorts of things different from scientific truth claims.

Indeed, if "truth is the intersection of multiple independent lies," then the contradictions and disagreements are less important than areas of agreement.

Ben Nelson: I've never been clear on what "accommodationism" means. It's isn't my phrase and I feel no need to defend it. Yes, some religious claims are wrong. It's surely wrong to impose such claims on another, and that, to me, imposes a limit on how broadly it's worth expanding a critique of religion. I don't see the use in trying to argue people out of beliefs that result in good behavior (or which don't result in bad behavior). I certainly don't see the point in antagonizing people who are my allies on issues that matter to me, just because we disagree about abstract notions that don't actually change how we interact with one another or with the world we share.

If that be accommodationism, I don't see where you've made the case against it.

Joshua Rosenau wrote:

Dave W.: "These 'truths' are awfully malleable."

Or at least personal, and therefore somewhat subjective. That's what make these sorts of things different from scientific truth claims.

Or wholly subjective, in which case they're not so much truths as they are desires. People want lots of different stuff to be true, but as you've agreed, that doesn't make 'em true. And it certainly doesn't make 'em "knowledge," nor does it make literature a "way of knowing."

Indeed, if "truth is the intersection of multiple independent lies," then the contradictions and disagreements are less important than areas of agreement.

What "areas of agreement?" Besides - and I may be misremembering - weren't you the one saying that consensus isn't knowledge?

I don't remember saying that "consensus isn't knowledge," and I have to think about it a bit to see whether I agree about that. Consensus certainly plays a significant role in knowledge evaluation in science, but I don't know if that generalizes, or if it counts as knowledge on its own.

I think that a truth claim can be subjective. It seems like "eating a pomegranate will make one happy" is not objectively testable, but it's certainly knowledge of a sort. True for me, maybe not for you. You might say that it's a desire, not truth, but it's a belief that I can justify from personal experience and which, to the extent anything can be said to be true, is true for me. Justified true belief, ahoy!

That literary truth claims work this way is certainly part of Shelley's critical approach to the question of "what quality went to form a Man of Achievement especially in literature." Negative capability is a trait he ascribes to Shakespeare and others, "when man is capable of being in uncertainties, Mysteries, doubts without any irritable reaching after fact & reason." It's surely a different thing than science, but I think it gets at truths about our world in a profound and important way.

Joshua Rosenau wrote:

I think that a truth claim can be subjective. It seems like "eating a pomegranate will make one happy" is not objectively testable, but it's certainly knowledge of a sort.

Well, that represents a fundamental disagreement between us, then.

"Eating a pomegranate will make one happy" is clearly a generalized truth claim about our "shared reality," and not something simply subjective like, "eating a pomegranate will make me happy." To create an objective test for the generalized version, one simply needs to clinically define "happy" and then feed people items in a double-blind fashion, and see whether pomegranates really do induce happiness, like an antidepressant alleviates depression. While such a test may not be feasible, that doesn't mean it's outside the realm of science. And just because people will (inevitably) argue about the definition of "happy" that's been used, that doesn't mean it's outside of science, either (people argue about the definition of "science," so should we say that that means nothing is scientifically determinable?).

Second, if "mind" is nothing more than the physical operation of a brain, then it is logically plausible that statements like "eating a pomegranate will make me happy" will, some day, be objectively verifiable. Clearly such statements are untestable right now, but that doesn't make them unscientific in principle. Besides, given a more-personal definition of "happy" which includes nothing but externally verifiable cues (like smiling, a relaxed posture, etc.), a metric boatload of surveillance cameras and a whole lot of time, it's quite likely that even today, we could determine whether eating pomegranates makes some hypothetical individual "happy" or not.

As I've said before, it seems like you want to wall off certain areas of the natural world from science. The emotional effects of fruit are a part of the natural world, regardless of our ability to study them with ease or robustness at this moment. In my opinion, whether pomegranates make a single person happy (or 64% of eaters happy) is definitely knowledge, and scientific knowledge, to boot. It's simply knowledge that we don't have right now, except anecdotally.

But even then, we can form informal empirical tests about it. If I tell you "I love chocolate," but then order some other flavor every time you see me in an ice cream parlor, you're going to think that I lied for some bizarre reason or that there's something else going on (like an allergy). That's not knowledge "of a sort," it's plain-old empirical knowledge gained through observation and hypothesis testing.

The pomegranate example was a bad one. I started off with "Eating a pomegranate will make me happy," then tweaked it to fix a problem I saw with that example, and didn't think through the new example. I'll go back to an old stand-by: "Charles Dickens is the best writer in the English language." I happen to disagree with that statement, others agree. There is no objective standard for what constitutes the "best writer," or even perfect agreement about what constitutes good writing. Yes, you could impose a standard, but again the problem is how to decide which standard is "better"? What metric does one use to evaluate metrics for good writing? What metric is best to evaluate that metric? Etc.

I note only in passing that one can love a given food but not eat it. Dogs seem to love chocolate and anti-freeze, yet rarely eat either more than once. Chocolate allergy is rare to non-existent, but caffeine (and possibly theobromine, I don't know) can be dangerous for people with heart problems. So simple empirical observation may not be enough. Even fMRI may not be adequate, as that involves various assumptions about what it means to love something, and whether we want to define love in terms of what parts of the brain lights up.

The reason I didn't use "Pomegranates make me happy," though, is precisely that my personal evaluation can be tested. You can ask me, observe me, test me, etc. But as a general statement, is there any empirical truth the to the claim that "pomegranates are yummy" or is it merely a matter of subjective personal preference? Is it true for me and false for someone else?

Josh, to my mind accommodationism involves the claim that religion and science are epistemically compatible. The upshot of the accommodationist argument is to foster a kind of "hands-off" attitude towards religion which explicitly condemns being an activist atheist.

I don't pretend to know anything about your views on activist atheism. If you think atheists should stop criticizing religion in public, then you're an accommodationist. If not, not.

However, the upshot of this stuff about being a pluralist about ways of knowing seems like you're apt to come to the conclusion that religion and science are compatible in an epistemic sense. But my point is that this pluralist stuff about "ways of knowing" actually leads in the opposite direction: both to epistemic incompatibility and to anti-accommodationism (whether you as a person happen to be an accommodationist or not).

Benjamin: Thanks for the thoughtful reply.

To clarify, is the only reason that someone might want atheists not to criticize religion that they believe religion and science are epistemically compatible?

This seems like the inescapable implication of your first two paragraphs, but maybe I'm missing something. If so, I find that assumption problematic. Published statements by Genie Scott, Chris Mooney, Barbara Forrest, and me all make clear that we don't think religion is epistemically compatible with science (or at least that we do not think it necessary to invoke the supernatural to explain anything, and that certain religious claims are not compatible with science). Thus, if that's the definition, the people presented as the most vocal accommodationists do not meet the first definition you offered for the term.

Is it possible, do you think, that we might have some reason to think attacking religion qua religion is a bad idea other than thinking that religion qua religion is inherently compatible with science? Might those reasons even go beyond political expedience?

To your last paragraph, I confess that I still don't get it. I would see your point if we all agreed that science was omnicompetent, because then we'd be in a position to say that any claim from nonscientific epistemologies must be verifiable against science, and if they agree with science they're fine but also superfluous and if they disagree then they're wrong and should go away.

But what if science isn't omnicompetent, as philosophers of science since Popper seem generally to agree? What if there are questions that can't be answered within science but which are answerable in other epistemological frameworks? People seem to disagree wildly about whether this sort of non-overlap (or minimal overlap) constitutes compatibility or not, but one need not have an answer in mind to pose the question about epistemologies in general or in the specific case of two particular epistemologies.

(I've switched from "ways of knowing" to "epistemologies" because I think it's a little clearer and because I've been able to confirm that Eugenie Scott regards the two as synonyms.)

Josh, just to clarify, I don't think that epistemic compatibility must by force of necessity appear in the arguments of the accommodationists. It's just that every so often they seem to be flirted with.

Mooney/Kirshenbaum is a difficult case because they explicitly resist clarification. (I think it's more respectful to include Kirshenbaum as part of the relevant criticism, since she was a co-author. Even if I disagree with them, it doesn't feel right to talk as if she were at the sidelines. Still, from here on in I'll refer to Mooney in the singular, since you did, with the unstated understanding that what I say about him applies just as well to her.)

So on the one hand, he does of course say that science is in touch with reality, religion not. But on the other hand, he seems to think that epistemic compatibility is available in practice: after all, there are all these clever scientists who are religious, and surely they're not mad. I, and many others, have found that this argument falls terrifically short of the mark, because it doesn't touch questions like "fine, but what if these scientists are tired of truth on Sundays?" For an interesting approach to the question of epistemic compatibility, you need an argument to the effect that they're compatible in principle. That they're real and honest parts of the whole truth project.

And when we look at it, this turn of argument is the most ironic part of the story so far, I think. For Dawkins is actually saying that religion and science are epistemically compatible in principle -- it's just that religion doesn't survive the encounter! That's effectively what he means by the 'god hypothesis'. Meanwhile, Mooney attacks Dawkins' argument against the god hypothesis on what he thinks are scientific grounds (i.e., by invoking the methodological/metaphysical naturalism distinction). This is tantamount to saying, "They're incommensurable", as if the former did not and cannot progressively reveal to us the demonstrably false claims in the latter -- which is the only way that he could be objecting to Dawkins's point. In other words, if we took Mooney seriously, then we'd have to say that he endorses epistemic incompatibility in principle. Either way, the epistemic line of inquiry does not come out in favor of any kind of happy compatibility.

Then I enter in and say, "Ways of knowing? Sure, let's have lots!", echoing you. (As a shorthand, I called this position "pluralism" in my final paragraph, if that helps to make it more intelligible.) This would seem to be a way to bring life back into the epistemic compatibility argument. Maybe religion helps us know in this way, and not that way.

But as I say, I think this line of argument fails, too. Pick whatever way of knowing you like, and Abrahamic religions come out false. It makes for bad therapy, bad morals, bad ontology. Just about the only true quality it has is the truth of its success as ideology.