Peter Unger argues that fundamental claims in contemporary metaphysics, epistemology, and the philosophy of mind are just empty ideas.
During the middle of the twentieth century, perhaps until 1970, mainstream philosophers generally agreed that, by contrast with the natural sciences, intellectually responsible philosophy should offer no substantial thoughts about the general nature of concrete reality. This was conspicuously so during the heyday of “ordinary language philosophy”, though the period in question began well before that, and it continued for quite a while after. As many mainstream philosophers now would agree, those were the bad old days.
In the bad old days, it was widely thought that, unlike scientists, proper philosophers shouldn’t bother considering substantial ideas about concrete reality, except for those found in our fund of common sense and except for those concerning how certain people use certain words, or something similarly parochial. But, for a philosophy that aims to be highly relevant to concrete matters and issues, is there really any alternative? For my critique of recent philosophy to have much point, there had better be. Indeed, to be motivated to press on with any extended critique, we may want to observe a few philosophical thoughts that are each impressive exemplars of concretely substantial ideas. Consider these.
Entity Materialism: There’s at least one concrete entity. Any concrete entity is a purely material entity; that is, it’s fully physical. (So, if there are any minds or souls, then each will be something material.)
Entity Idealism: There’s at least one concrete entity. Any concrete entity is a purely mental immaterial entity; none is at all material, or physical. (So if there are any rocks, each will be something mental and immaterial.)
Entity Dualism: There are at least two distinct concrete entities. Any concrete entity is of either one or another of two radically different kinds. Some concrete entities are purely material, or fully physical; none of them is mental. Other concreta, each a mind, or a soul, are wholly and purely mental individuals; none is at all material, or physical.
As I trust, you’ll easily see each to be in strict conflict with the two others, with the conflicts concerning concretely substantial questions. At least on my way of taking things, these three apparently substantial ideas are nothing new: Thomas Hobbes was an Entity Materialist. Rene Descartes was an Entity Dualist, as was John Locke. George Berkeley was an Entity Idealist.
As those examples exemplify, before (the middle of) the twentieth century, many prominent philosophers actually endorsed substantial views regarding the nature of concrete reality.
In contrast with such earlier thinkers, many mid-twentieth century philosophers endorsed nothing even nearly as adventurous. Since about 1970, things have become a little bit better. For example, nowadays it’s fine for philosophers to endorse materialism and, among those working in metaphysics or the philosophy of mind, the vast majority do. Still, materialism is nothing new. And, in recent years, there have been precious few novel attempts to argue for the old view, maybe none of them very convincing or impressively illuminating.
In the heyday of ordinary language philosophy, as I’ve said, it was agreed that mere philosophers should steer clear of views about the nature of concrete reality. But then, what may such constrained philosophers place on offer? Toward quickly beginning to answer that, it’s useful to confront examples of clearly acceptable philosophical fare.
Easy to grasp, here’s one example:
“Someone will remember his old college days only if he actually went to college.” (If a person didn’t go to college, then, no matter how much he learned through his experience, he won’t remember his old college days.)
Here’s another example: “Taking it that you’re a normal human being, living a normal life, we may correctly suppose that you sometimes feel pain. But, there’s this strict limit on the pain you feel. It will always be only your own pain you’ll feel, never anyone else’s pain.”
And, here’s a third case: “Someone will perceive a dog that’s nearby him only if there really is a dog there.”
Well, I suppose that people are concrete individuals. That supposed, we may allow that our three illustrative ideas are thoughts about concrete individuals, in a perfectly ordinary sense of “about”. In the bargain, each will be about concrete reality. But, that means little. What means more, we’ll see, is this: None delineates a way for concrete reality to be which differs from other ways for it to be.
As I expect, you’ll have gotten the hang of what I’m after. So, you’ll have a fair grasp of what’s meant when saying that none of our three illustrative ideas is a concretely substantial thought. Rather, each is concretely insubstantial. That’s some helpful terminology. Let me introduce more nice terminology. As will be agreeable to say, a thought that doesn’t delineate some ways for concreta to be from other such ways is empty of import for, or as regards to, concrete reality. So, as we may then happily say, a concretely insubstantial idea will be a concretely empty idea.
Directly concerning matter more than mind, here’s a fourth illustrative example: “A rock that’s wholly composed of some matter won’t continue to exist if, all at once, all its matter ceases to exist.” (Now, perhaps some of a rock’s matter might cease to exist, and yet the rock will continue to exist. And, in a gradual and overlapping way, perhaps all the rock’s present matter may cease to exist, with the very same rock continuing to exist, eventually composed of entirely new matter. But, if all a rock’s matter suddenly ceases, the rock itself won’t continue to exist.)
Unlike our four illustrative ideas, many other ideas are concretely substantial. Assuming that our ordinary terms and common concepts are satisfiable, perhaps this may be one:
“Each of very many people remembers that he, or she, went to college.”
On that assumption, another will be the negation of the common-sense idea just displayed. The displayed common-sense thought differentiates among, or between, alternative ways for concrete reality to be. As we might say, the thought favors all the ways in which each of many folks remembers that she went to college and it disfavors the ways in which that doesn’t obtain.
With many empty ideas, it’s obvious that the thoughts are insubstantial. With many others, things aren’t so obvious. Often, this happens with wholly correct ideas. As we do well enough to suppose, these next two thoughts are correct ideas, though that’s not obvious:
“Someone can’t be happy that she went to college unless she knows she went to college.”
“Someone can’t assert that she went to college without representing herself as knowing she went to college.”
Though it’s unobvious that the ideas are correct, it’s obvious they’re concretely empty.
In the recent philosophical literature, we find many cases where it’s unobvious both whether a thought that’s on offer is correct and also whether the offered thought is concretely substantial.
One of the most widespread and influential of these cases concerns the offering of ideas to the effect that people can think about various concrete substances, like water, wood or gold, only if the people have had a suitable causal history, however appropriately complex and indirect it may be, involving the substances there thought concerns.
Most prominent here are thoughts flowing from various versions of Hilary Putnam’s Twin Earth examples – thought experiments which raise questions about the meaning of words when spoken on Earth or on a nearly identical world –and other far-fetched examples flowing from Putnam and his admirers. For a salient instance of such companion offerings, think of Donald Davidson’s example of his Swampman, a creature just like Davidson but created by a chance lightning strike in a swamp, and Davidson’s claim that, as he has no apt history, Swampman can’t, really, even think (about anything) at all. In a forthcoming book, I will argue that, first, and less importantly, all these offerings are incorrect ideas, even if their incorrectness may be quite unobvious. Secondly, and more importantly for our present work, I will there argue that they all represent just so many concretely empty ideas, and not any new philosophical thoughts that are concretely substantial, however unobvious this may be.
Another of the most widespread and influential of these cases concerns the offering of ideas to the effect that certain ordinary individuals, as with many ordinary tables and chairs, each has certain essential facts concerning its material origins.
Most prominent here are thoughts found in Saul Kripke’s enormously influential Naming and Necessity. Concerning a certain quite typical wooden table that was in his presence, Kripke there famously offered the thought that, for any candidate item to be that very table, the concrete candidate must be something first made of some certain wood, namely, (at least quite a lot of) just that wood that, in actual fact, first composed the table then present to Kripke. Continuing his thought, Kripke opined that the candidate could not have been something first made entirely of ice, and not of any (of that) wood at all. As with the case of Putnam, and Davidson, related just above, I think that first, and less importantly, these Kripkean offerings are also incorrect ideas, even if their incorrectness may be quite unobvious. Secondly, and more importantly for our present work, these Kripkean thoughts are also just so many concretely empty ideas, not any new philosophical thoughts that are concretely substantial.
But, perhaps I have too much gotten ahead of myself. So, let me return to present some far less tantalising fare, beginning with a rather fully articulated dialectic that, by steps, proceeds to engage in some pretty paradigmatic mainstream metaphysical ideas, quite fashionable to argue over, at least during the past couple of decades.
Toward this end, it may be helpful to proceed in three steps. First, I’ll sketch a make-believe philosophical debate. Second, I’ll sketch a closely related quasi-real-life debate. Third and again closely related, I’ll sketch a real-life contemporary debate that many mainstream philosophers have found tremendously engaging. Unless my writing is very ineffectual, it should strike you that there’s a deep similarity among the three disputes.
Our make-believe debate has two thinkers agreeing that, in a certain spatial region and at a certain time, there are six perfectly fundamental physical concreta. As agreed, they are arranged hexagonally. Now, one of our thinkers holds that, in addition to the six elementary particles, there’s a certain further thing in the region, namely, a hexagonal arrangement. Perhaps, she may be a quite commonsensical thinker. Her opponent, apparently an ontologically stricter thinker, holds that, then and there, all that exists are the six particles (arranged hexagonally).
The dispute between these two make-believe thinkers concerns only an empty issue, no concretely substantial question. Let me amplify. One of the parties to our dispute, apparently quite commonsensical, accepts some such empty thought as this idea:
HEXAGONAL ARRANGEMENTS: If there are, at a certain time in a certain spatial region, six particles arranged hexagonally, there will be, then and there, a hexagonal arrangement (of the particles).
Rejecting that, the other party accepts, instead, some relevantly opposite empty idea, a thought to be denied, of course, by our first party:
NO HEXAGONAL ARRANGEMENTS: Even if there are, at a certain time in a certain spatial region, six particles arranged hexagonally, there (still) won’t be, then and there, any hexagonal arrangement (of the particles).
Barring a poor choice of words, both HEXAGONAL ARRANGEMENTS and NO HEXAGONAL ARRANGEMENTS fail to delineate any way for concrete reality to be, differentiating it from others. Here the same, both are empty ideas.
Next, there’s our quasi-real-life dispute. As before, both parties agree on quite a lot: First, in a certain spatial region and at a certain time, there are enormously many elementary particles. And, second, these particles are arranged tablewise. What’s this? Don’t worry. To be sure, “tablewise” is something of a barbarism. But, for some philosophers, the barbarism may have some utility. Anyway, when agreeing on how it is that our particles are arranged, our present debaters mostly agree about just this: First, the aforesaid particles are related to each other in a way that, then and there, is as conducive as can be for the constitution of a table by just those very particles. (Given the propensities of the basic particles, and other physical factors, this will mean a certain mutual cohesiveness for our particles, and other “nice physical features”, too.) For example, regarding our many currently considered particles, they must be tellingly separate from other particles, as well as any other disqualifying concreta. Perhaps, each aforesaid particle is nearer all the others than any is to any other particle, or any other disqualifier.
Well, then, given all their shared common ground, what’s debated by our quasi-real-life disputants? No surprise, here our story goes like this: On one side of this second dispute, there’s a quite commonsensical thinker. She holds that, in our considered spatial region at our contemplated time, there is, in addition to all those particles arranged tablewise, a tablewise arrangement (of the particles). As before, her opponent is, apparently, an ontologically stricter thinker. He holds that, in our region and at our time, all that really exists are the enormously many particles themselves (arranged tablewise).
The quasi-make-believe debate concerns only a concretely empty issue. One party accepts an empty idea relevantly paralleling HEXAGONAL ARRANGEMENTS:
TABLEWISE ARRANGEMENTS: If there are, at a certain time in a certain spatial region, very many particles arranged tablewise, there will be, then and there, a tablewise arrangement (of particles).
Her opponent denies that, and he accepts this, instead:
NO TABLEWISE ARRANGEMENTS: Even if there are, at a certain time in a certain spatial region, very many particles arranged tablewise, there (still) won’t be, then and there, any tablewise arrangement (of particles).
The dispute between these two parties will be, of course, just as concretely empty a debate as that between those who affirm HEXAGONAL ARRANGEMENTS and those others who affirm, instead, NO HEXAGONAL ARRANGEMENTS.
Third and finally, here’s a real-life metaphysical debate, salient during the last few decades of mainstream philosophy. In this third debate, the two parties agree on just what was agreed by the disputants in the quasi-real-life debate. With no difference as regards agreement, what’s newly debated here?
On one side, there’s a commonsensical thinker. She holds that, in our considered space at our contemplated time, there exists a table, and not only many particles arranged tablewise. (Typically, she might say that the particles in focus serve to compose the table in question, which ordinary thing thus may well be called a composite table. But, as concerns the crux of the current discussion, she needn’t say anything more.) As regards her opponent, we may say the same as just before: He is, apparently, an ontologically stricter thinker. He holds that, in our region and at our time, all that really exists are the enormously many particles themselves (arranged tablewise). (The thought of a composite table, he might say, is “an ontologically profligate idea fostered by a certain way of talking”. Or, he might say nothing further at all.)
This real-life dispute is a concretely empty debate. One party accepts some such thought as:
COMPOSITE TABLES: If there are, at a certain time in a certain spatial region, very many particles arranged tablewise, there is, then and there, a composite table.
Rejecting COMPOSITE TABLES, the other party accepts some such opposite thought as:
NO COMPOSITE TABLES: Even if there are, at a certain time in a certain spatial region, very many particles arranged tablewise, there (still) won’t be, then and there, any composite table.
Neither thought delineates any way for concrete reality to be, as against any other way.
At this juncture, I should say something about how our newly drawn distinction, between the concretely empty and the concretely substantial, compares with certain philosophically familiar distinctions. First, I’ll say something about how it compares with the philosophically familiar distinction between thoughts that are strictly necessary and (merely) contingent thoughts. This will be pretty simple stuff. To begin it, we may reconsider the illustrative thoughts that figured saliently at the start. In particular, return to consider this familiar concretely empty idea: “Someone won’t ever remember his old college days – indeed, he can’t possibly ever do so – unless he’s someone who’s been to college”. Now, quite as well as this thought “about remembering” is a concretely empty thought, it’s also a strictly necessary thought. And, doubtless, it’s a strictly necessary truth, not any falsehood.
Of course, many thoughts I call “concretely empty” aren’t characteristically philosophical empty thoughts. Consider the thought that someone is a husband only if the person is married to another person. While certainly concretely empty, that’s not any philosophical idea at all. Anyway, even as many characteristically philosophical empty thoughts are necessarily correct ideas, so are many nonphilosophical empty thoughts. Even as that’s humdrum, so is this other mere reminder: A denial of a strictly necessary truth, or a negation of it, will be a thought that’s necessarily false, or strictly impossible. Accordingly, it’s strictly impossible for someone who’s never been to college to remember her old college days.
Along with others, we may ask: What’s the source of that strict impossibility? Though it may concern nothing deep, I’m happy to suppose this standard line may be on track: The thought that someone remembers something to be so essentially features the concept of someone’s remembering something to be so. And, it’s essential to this concept that it apply only where the allegedly remembered item actually is so. Well, with this line in mind, we may allow that very many correct empty thoughts are not only necessarily true, but they’re also conceptually true. And, it may be that the reason these thoughts are necessarily true is that they’re conceptually true.
We turn to another familiar distinction, that between analytic thoughts and synthetic thoughts. For this distinction to apply widely, we need a broad notion of notion of analyticity. So, our notion will have these correct empty thoughts be analytically correct:
“If a certain one material object is more massive than a second, and the second is more massive than a third, then the first is more massive than the third material object.”
“If a man wrote a book, then a book was written by a man.”
Now, while the first of the two just might be a conceptual truth, turning on certain conceptual features of MORE MASSIVE THAN, it’s likely, I think, that the latter is best treated as merely grammatically analytic, not a thought that’s conceptually analytic. But, perhaps that’s not so. Either way, there’s nothing in this last bit that’s crucial for to our inquiry.
Of course, I can’t give a helpfully adequate articulation of what it is for something to be analytic. But, then, neither can I, or anyone else, I think, give a helpfully adequate articulation, either, of what it is for something to be conceptual, or linguistic, or experiential. At all events, most sensible inquirers are able to distinguish, with great frequency and reliability, between the experiential and the nonexperiential – and the analytic and the nonanalytic. As I believe, most reading these sentences are well able to make all those discriminations, not just those between the experiential and the nonexperiential, but also those between the analytic and the nonanalytic. To help make sure that this is so, I’ll provide quite a few examples of thoughts that are analytic, or analytically correct ideas. Indeed, I have already done so, with at least these recalled exemplars:
“Someone will (and can) remember his old college days only if he actually went to college.”
“It will always be (and it must be) only your own pain you’ll feel, never anyone else’s pain.”
“Someone will (and can) perceive a dog that’s nearby him only if there really is a dog there.”
“A rock that’s wholly composed of some matter won’t (and can’t) continue to exist if, all at once, all its matter ceases to exist.”
“Someone can’t be happy that she went to college unless she knows she went to college.”
“Someone can’t assert that she went to college without representing herself as knowing she went to college.”
“Someone is (and can be) a husband only if the person is married to another person.”
“If a certain one material object is more massive than a second, and the second is more massive than a third, then the first is (and must be) more massive than the third material object.”
Quite likely, providing that list was just so much overkill. At all events, we do well to proceed.
It’s been my contention that, with hardly any exceptions, when mainstream philosophers have offered concretely empty thoughts that are correct ideas – and they haven’t just missed the mark – those are analytically correct ideas. If it was a necessary truth that such a thinker offered, and she didn’t miss her mark, then the necessity of her thought is due to its t being an analytic idea.
Prominent mainstream philosophers have, in recent decades, offered many ideas as being deep philosophical truths, concerning, perhaps, deeply necessary features of concrete reality. As I’ve come to believe, most of these offerings weren’t truths of any sort, at all. Suppose that’s so. But, still, what about the relatively rare truths that were thus offered? Well, as I’ve observed just above, the situation will then be this: In almost all the cases, not only are the correct thoughts concretely empty propositions, but, what’s more, their emptiness derives from their analyticity. As we may then well say, the few truths newly offered were, almost without exception, just so many analytically empty ideas. Just so, the very few prominent proposals that were correct, or are correct, are, almost without exception, on a par with, for instance, the thought that someone who asserts that something is so represents herself as knowing that the thing is so.
Except when it’s trafficked in quite parochial thoughts, mainstream philosophy has, in recent decades, offered hardly anything by way of concretely substantial ideas. Does this mean that, in an important way, mainstream philosophy is, or it has been, quite deficient? Yes; of course, it does. Let me amplify.
Suppose that, as many believe, all of reality comprises not only all concrete reality, but abstract reality, as well. And, suppose that, far from comprising just something scant or marginal, abstract reality comprises quite a lot of all reality. Finally for now, suppose that, although it says little or nothing about concrete reality, (pure) mathematics offers us a great deal about abstract reality; which offerings include many necessary truths about abstracta. Now, given these assumptions, which many have found plausible, it may be that mathematics offers us many ideas, indeed, many correct ideas, each of which is a substantial thought about abstract reality, in an importantly apt sense of “substantial”. And, hence, it may be that, about all reality, mathematics offers quite a lot that’s substantial, in a certain reasonable sense of that term.
Well, even if all its offerings should be concretely empty ideas, mightn’t mainstream philosophy be quite like pure mathematics, offering us many ideas, even many correct thoughts, concerning abstract reality, each a strictly necessary truth about what may be a vast realm of reality? And, in that case, mightn’t mainstream philosophy be perfectly adequate philosophy, not in the least bit deficient?
The previous paragraph comprises two questions. The answers to those questions are, respectively, “No” and “No”. Let’s start by seeing why there’s a negative answer to the second and last question: (Supposing it offers many strictly necessary truths about a vast realm of abstract reality) mightn’t mainstream philosophy be perfectly adequate philosophy?
In many ways, philosophy differs from mathematics. In line with that, there’s this: Throughout most of its history, philosophy has aspired to offer substantial thoughts about concrete reality, whatever it may have said concerning abstracta. And, with fair frequency, it has succeeded in this, and not just by offering quite parochial ideas. Thus, philosophy has offered us various sorts of idealism, and various sorts of materialism, and various sorts dualism, to take just one way of philosophically slicing a big concrete cake. Now, it must be admitted, of course, that many philosophers who have been successful in this concretely substantial regard have also been concerned to offer important ideas as to abstracta. But, of course, that doesn’t cast even the slightest doubt on whether it’s proper for philosophy to offer (nonparochial) substantial ideas about concrete reality. So, even if mainstream philosophy should offer many important thoughts about abstract reality, it may still be terribly deficient philosophy, at least compared with much philosophy saliently offered in earlier eras. Quite unlike with pure mathematics, it may be important for philosophy to offer quite a few substantial ideas about concrete reality – and many of them not just some pretty parochial thoughts. So, if recent mainstream philosophy offers little or nothing substantial about concrete reality, beyond what’s quite parochial, then this recent philosophy has been terribly deficient – not by the standards proper for, and to, mathematics, of course, but by the standards proper for, and to, philosophy.
Now, let’s address the other question: if pure mathematics offers us ever so many necessary truths, then will all these truths, or almost all, be just so many analytically correct ideas? That is very hard to believe. More plausibly, those correct propositions will be correct for some other reason, having nothing to do with any analyticity. As many have said, perhaps following Kant, they will be synthetic necessary propositions. But, as I’ve been arguing, and as I do far more in the book towards showing, the few correct offerings of recent mainstream philosophers have been, almost without exception, just so many, or so few, analytically correct ideas. So, quite fundamentally, virtually all (correct) recent mainstream philosophy is quite unlike mathematics. Of course, everyone knew that anyway – but it seems worth remarking here.
It should be clear what my target is, and what it is not: My target is most of what has been called the core of analytic philosophy: metaphysics, the fundamentals of philosophy of mind, the fundamentals of philosophy of language, and quite a bit of epistemology. That, and not very much more. Just so, almost all of ethics is beyond the scope of my discussion, as another big bit of epistemology, as well as many more marginal philosophical areas.
As just indicated, nowhere in this work will I say a very great deal about anything that’s deeply normative, or fully evaluative, or anything of the ilk. Or, at least, I won’t say all that much about claims that are obviously, or explicitly, or paradigmatically of any such forms or sorts. Now, on the most robust reckoning of what it is for a claim to be concerned with concrete reality, such claims aren’t concerned with concrete reality. So, most robustly, what is their concern? Is it certain abstracta? That boggles the mind. And, placing that aside, what sort of abstracta are they to be? Quite unlike numbers, if there are any, or transcendental universals, if any there are, these normative abstracta must be, somehow or other, the source of some especially important normative, or evaluative, force or ground. How should that happen? As it seems, purported answers will be little more than puff.
Along with many others, I have doubts as to whether any purely normative or evaluative claim is ever any truth at all, much less any truth robustly involved with concrete reality. Here I mean, very unsurprisingly, for paradigmatic cases to include any claim as to someone’s being justified in something, whether the something is the person’s believing a certain thing, or her doing a certain thing, or whether her wanting or valuing or feeling whatever it is that she wants or values or feels: Placing aside whichever of these purely normative or evaluative claims may be, just perhaps, an offering that’s merely analytic, and so merely analytically empty, what about the rest of these claims, all those that are, supposedly, very far from being just so empty. Are they some synthetic apriori truths – somewhat like, perhaps, the thought that nothing can possibly be, all at once, perfectly red all over, and through and through, and also perfectly green all over, and through and through. It certainly doesn’t seem so. Indeed, any purely evaluative offering I’ve ever seen seems, in all plausibly relevant ways, entirely unlike that thought about red, and green, and their (putative) mutual exclusion. So, though my project’s targets are limited, and they certainly don’t include all recent philosophical offerings, this is not something that, as I see it, will do much to reduce what interest there may in the work.
Summing up, I’ll say this: What has pervaded so much of recent mainstream philosophy are ideas that, if correct, are just analytically correct. So, even if there should be some synthetic necessities that strongly bear on concrete reality, which I very much doubt, these recent proposals won’t be among them. Rather, insofar as these proposals are ever correct ideas, that’s just because they’re analytically correct. Again, it’s for their being, at best, just some analytically correct ideas that these recent proposals are empty of import for concrete reality.
With that said, perhaps this may be said, just as well: for future mainstream philosophy to be more robust than it has been recently, and it is currently, there must be, in the future, quite a few proposals that aren’t just so many analytic thoughts (and, presumably, that, aren’t strict necessities, at all). In broad, deep and philosophically interesting ways, these proposals will delineate some ways for concrete reality to be from other ways. As such, they’ll be concretely substantial proposals, certainly in my sense of the terms and, I think, in any reasonable sense of them. None of this precludes philosophers continuing on with what, in recent years, so many have been doing. Far from it, I think it fine for that to continue. What I’m proposing is that, in addition to all that, there should be offered, as well, a fairly wide variety of aptly broad and deep concretely substantial thoughts, each a happily synthetic idea.
Peter Unger is Professor of Philosophy at New York University. He has written extensively in epistemology, ethics, metaphysics and the philosophy of mind. He is the author many books, including Ignorance: A Case for Scepticism; Philosophical Relativity; and Living High and Letting Die: Our Illusion of Innocence. This essay is based on ideas explored in his 2014 book, Beyond Emptiness.
Subscribe to The Philosophers' Magazine for exclusive content and access to 20 years of back issues.
