Meaning, truth, language, reality
Ernie Lepore interviewed by Richard Marshall.
Enie Lepore has written a whole book on quotation marks and many think he has perverse intuitions when he rejects contextualism and semantic holism. He always thinks against the grain which led to Donald Davidson not speaking to him for five years after a brutal encounter. But he collaborates with Jerry Fodor, Hermen Cappelen, Kirk Ludwig and others all the time because he knows that the jig is up when you think you can’t be informed by someone else. He thinks the new generation of philosophers of language know more about language than their teachers and so are the start of a second renaissance. If he were a scotch he’d be Laphroaig, a rich jive that bites your throat. Fabaloosa.
3:AM: What made you become a philosopher? Was it a surprise to anyone?
Ernie Lepore: During my freshman year in college I promised my mom on her deathbed that I would become a lawyer. After a year of mourning I returned to college; the consensus was that philosophy was the best pre-law major. As it happened, UMass-Amherst had an excellent philosophy department, though not in the style I was vaguely familiar with from having read the fiction of Sartre and Camus in high school. Doubts about my prose brought me to ask where I might bone up on grammar; the answer was linguistics.
There I was eighteen years of age, taking metaphysics with Bruce Aune, modal logic with Ed Gettier, math logic with Terry Parsons, and “grammar” (semantics) with Barbara Hall Partee. It was not what I expected. To boot, many of the undergrads in Barbara’s class were English majors and I thought that was cheating; they already knew grammar. Little did I know! After the first year, I unofficially adopted both the philosophy and linguistics departments as my new family; it was a splendid arrangement. Did my newfound fascination with philosophy surprise anyone? Well, during regular visits to Amherst, I would suffocate my best friend from high school, Brian McLaughlin, with what I was learning about Quine and Skinner; he went back to Jersey, took a philosophy course with Chris Hill; and so, we both sort of backed into the field. I’m still surprised by grad students who tell me that they knew they wanted to become philosophers since high school. I can’t imagine either of us entertaining that idea while we were growing up and spending most of our time on the wrestling mat.
3:AM: Are you a philosopher of mind or a philosopher of language, or both? Or is the distinction no longer very useful?
EL: I’m a philosopher of language, not a philosopher of mind. Barry Loewer and Jerry Fodor pushed me into philosophy of mind in the late 1980s, convincing me it made little difference whether language or thought were being investigated, since both were representational. It took me longer than it should have to realise they snookered me. Once I did, I ran back to language. I bailed out right around the time that I was asked whether strong or weak, global or regional supervenience was guiding my judgments about the relationship between the mental and the physical. It wasn’t that philosophy of mind had become too scholastic; after all, I’ve written an entire book on quotation marks. (Fodor once chided me that the semi-colon was unsafe from my philosophical purview; and after he complained I shouldn’t be working on such a small topic as “that”, I turned to “the”; “a” has become a recent interest of mine.)
What drove me back to natural language were the robust intuitions I lacked about supervenience but that I had about linguistic usage. Colin McGinn once counselled me to always ask myself first, “What do I think about this subject?” That was sage advice. In the philosophy of mind, I simply didn’t have any intuitions, or last strong intuitions, about a wide range of subjects. Mind you, many philosophers would say (and have said) that my intuitions about language are perverse: those about context sensitivity are particularly so. It has come as a surprise to me to discover how deep my commitments to intuition run.
When I was younger I thought philosophers adopted certain positions just to be provocative. They wanted to see how far they could defend them. I now appreciate that given the perseverance it takes to endorse minimalism (or eliminativism or solipsism or anomalous monism or the language of thought or paraconsistent logics or a range of other “perverse” views), you’d better believe it; I believe minimalism. A more recent source of concern for me is consistency – my own, that is: I have nine books and nearly 150 published articles; that’s a lot potential inconsistency to worry about. Intuitions and pursuits in one area sometimes clash with those in another. A case in point is my work on quotation; I had published several articles on the topic before the book; and in those articles I advocated contextualism, until once in a seminar, Jason Stanley, John Hawthorne and Jerry Fodor all pressed the peculiarity of defending contextualism about quotation while rejecting it for gradable adjectives. That encounter caused a year’s delay on the book’s completion as well as a thorough rethinking of its subject.
I could tell similar sagas about my work on Davidson as well as my current book in progress on the semantic/pragmatic divide. Unfortunately, we sometimes set out to theorise about a subject on the basis of too small a portion of the relevant data. That reminds me of a BBC radio interview I once did with Michael Dummett and Myles Byrnyeat. The host began by saying, rather loudly, “The cat is on the mat”. When it became evident we had no idea what he was getting at, he added he had been told this was the most important sentence in philosophy. Dummett corrected him saying that that sentence had been replaced by “Snow is white”; of course, were he speaking now he’d have to say that both sentences have been replaced with “It’s raining”. We’ve come a long way!
3:AM: Semantic holism has been a terribly influential approach in philosophy of language and has been contrived and defended by some of the biggest names in modern philosophy hasn’t it? Quine, Davidson, Lewis, Dennett, Block, Brandom and Churchland et al. And Wittgenstein of the PI. Can you say what question meaning holism is trying to answer and how it goes about doing so?
EL: When Fodor showed up at Rutgers in the late 80s, we co-taught a graduate seminar on holism vs. atomism. He was puzzled how I could reconcile my endorsement of Davidson’s holism – the idea that the contents of words in some sense depend on one another – with my criticisms of conceptual role semantics – a rather old idea dating back at least to Ferdinand de Saussure that the meaning of a word is a function of its inferential relationships to other words. Fodor thought the best (only?) defense of holism would rely on inferential (conceptual) role semantics. (Fodor and I agreed from the start that wherever I used “word” he would say “concept” and vice versa; I don’t think natural language has interested him much since his days with Katz in the 60s.)
I thought Davidson had independent arguments for holism; and so, Fodor agreed to co-teach the seminar. It came as an embarrassment – a game changer intellectually – to discover that my “arguments” in defense of Davidson’s holism were mere restatements of the thesis; Fodor and other colleagues, e.g., Stich, McLaughlin, Loewer, Klein, and Matthews, pressed me on this weekly. (It’s worth noting that the seminar never got to atomism; a number of philosophers over the years have inferred that I had gone Fodorian with the holism book. I doubt there exist any good arguments for atomism either.) I found myself by the end of the term mumbling something like “Holism is a metaphor” (or a picture or a way of life or some other nonsense).
3:AM: Interestingly Donald Davidson was one of the targets of your attack on holism, but you have returned to him again and again. A reviewer of your first book about his work with Kirk Ludwig praised you for laying out the complexity of this difficult philosopher but nevertheless he considered it ‘400 pages of rough sledding.’ Is your fertility partly explained by this oppositional stance to Davidson’s work?
EP: One consequence of publishing the holism book was that Davidson did not speak to me for five years beginning with a rather hostile Pacific APA symposium Fodor and I did with him on holism and radical interpretation. That encounter was brutal. Though Fodor and I intentionally moderated our tone and hedged our critical comments, Donald would hear none of what we had to say.
After my five-year exile, Davidson invited me to a workshop on him in Girona. He would speak in the morning and an invited participant would guide the afternoon discussion. I was assigned the follow-up discussion of holism. Quine, Davidson and Dreben complained that I had not fully appreciated the role of indeterminacy in defending holism; and that some of my arguments against holism presumed a relational account of belief. That was the substance of their objections. I recall phoning Jerry from Spain to tell him we remained on terra firma.
For the rest of his life, Davidson always referred to the holism book as Fodor’s book. He intended this as a compliment. When he died, Marcia Cavell, his widow, gave me what she assumed was a copy of the book I must have given Davidson. When I asked why she thought it was my copy, she replied that there was a dedication to Donald on the first page. I couldn’t imagine I had done that. When I looked at the book later I found an inscription in Davidson’s own handwriting, “To Don, from the authors, with contempt”. After Girona, back in the fold, I began several efforts to implement the Davidson semantic program for a few largely ignored linguistic fragments, for example, quotation and (with Kirk Ludwig) complex demonstratives like “that man”.
Davidson appreciated, encouraged and liked the results; this led to more work on Davidson’s semantic program, eventually culminating in Kirk Ludwig’s and my fat book on truth theoretic semantics. It also led to an invitation to teach at Berkeley in the winter of 1995. There I met Herman Cappelen at Berkeley; he was my TA in philosophy of language. The following year he moved to NYC and we began a collaboration, which ultimately climaxed in Insensitive Semantics.
It started out harmlessly enough; Herman recommended a paper by his tutor from Balliol, Ian Rumfitt, on the semantics of indexicals. Rumfitt argued that each speaker must have a different semantic theory since distinct utterances of (e.g.) “I am Norwegian,” can differ in truth conditions, depending on who’s talking. I thought the view idiosyncratic, but after a little research discovered it was ubiquitous; its impetus was the seemingly innocent thesis that a semantic theory should capture what speakers say when they speak, as well as the correct intuition that what I say when I utter “I’m Norwegian” differs from what Herman says when he utters it. Our attack on this thesis became the basis for what we call speech act pluralism – the view that we say indefinitely many things when we utter a single sentence.
This latter view, interestingly enough, precedes our development of semantic minimalism. Once what’s said is separated from semantic content, we need to rethink what the semantic content of an expression is. Minimalism, our answer, was consonant with Davidson’s and Fodor’s disquotational stories, I am returning to this important question in current research.
3:AM: In your first book with Ludwig, you focus largely on issues in Davidson’s ‘Truth and Meaning’, don’t you? So you look at Davidson’s theory of trying to run a theory of meaning off the application of a theory of truth, Tarski’s famous Convention T. You argue that as he constructs it it fails. You look at Davidson’s notion of radical interpretation and interpretative charity which as you see it Davidson hoped would add a constraint to the theory of truth and help fix meaning. You think this fails. And you look at Davidson’s thought that because of his linking a theory of truth to meaning, plus radical interpretation and charity we could commit to certain positions in metaphysics and epistemology – one in particular is striking that there couldn’t be massive error in our beliefs. Again, you are unconvinced. Of course you can’t go through all your reasoning here but perhaps you could sketch how Davidson goes wrong in these three areas and why despite this you think his truth conditional semantics approach is still fertile?
EL: We don’t advocate truth conditional semantics for natural language if by that you mean an absolute truth theory as a theory of meaning. We’re convinced by the standard arguments; in particular, by Foster’s argument that sentences of the form “S is true iff p” affirm only a material bi-conditional, which is too weak to underwrite an attribution of meaning to S. But, as we’ve argued several times in print, an appropriately constrained truth theory can deliver a correct meaning theory. If “S is true iff p” is derived from interpretive axioms by means of rules that preserve meaning, then the sentence used on its right hand side will give the meaning (without saying it does) of the sentence described on the left hand side.
Such canonically derived material conditionals provide us with all the information we need to assign meaning. It’s my contention that nothing I’ve just said is incompatible with Davidson’s views. The literature is littered with sophomoric critiques of Davidson’s philosophy of language, advanced by authors who have read one article and on that basis decided they’ve understood the motivations for the larger program. They were wrong. Davidson was never a meaning eliminativist (like Quine); nor a meaning reductionist nor an extensionalist. If someone can give an argument to the contrary, please do.
3:AM: An appeal of Davidson’s attempt to get meaning out of truth is that it has the attraction of trying to explain a lot using very little. In this respect it seems in the same ballpark as what Fodor was doing with his LOT and Chomsky with his UG. In ‘Meaning Mind and Matter’ with Barry Loewer you say that this approach can explain, ‘one central ability: the ability to acquire justifiable beliefs about the world and about each other’s beliefs through our linguistic abilities.’ Can you explain how it does this? Does this approach imply that a non-linguistic creature does not have beliefs, or that we do it this way, but that lions have other options?
EL: I steer clear of discussions about animal thought. My intuition is that they obviously do think, and since so many arguments to the contrary rely on an unfounded holism, I tend not to be seriously engaged with this debate, so, I’m not currently on top of it. On the role of a meaning theory in explaining belief formation, however, I’ve had much to say over the years.
It began in my third year of graduate school; I found myself attending courses on Frege, model theory, lexical semantics, language acquisition (with Kennith MacCorquodale), and psycho-linguistics (with J. J. Jenkins). I was puzzled why all these courses were said to be concerned with semantics; as far as I could tell, they had little to do with each other. This put me in a funk over what was semantics all about, and what counted as getting it right. Reading Field’s ‘Tarski’s Theory of Truth‘ the following summer, I was baffled by his portrayal of semantics as a threat to physicalism. I then had an epiphany. A role Field (and, as I realised later, Davidson as well) assigned to a theory of meaning was to account for transitions from heard utterances to beliefs about the world. If you utter “It’s raining” and I find you reliable, ceteris paribus, I will come to believe that it’s raining, whereas someone who doesn’t understand English won’t. Semanticists are (or should be) asking which knowledge underwrites this linguistic/non-linguistic transition. This presents a way for evaluating alternative semantic frameworks, one from which my dissertation grew as well as probably a half-dozen of my first publications.
I concluded that many of the then popular frameworks were ill-suited to answer this question, and so, were quite useless. Davidson himself hedged on my question; sometimes he spoke of linguistic knowledge, but more often he spoke of what a speaker could know that would enable her to make these transitions. I believe, and have argued as much, that the stronger approach is needed. My stand on the psychological reality of semantics theories, linguistic communication, and convention perhaps reveal a commitment to a deeper connection between mind and language than I suggested earlier.
3:AM: Meaning holism still presses its claims and in ‘The Compositionality Papers’ you and Fodor continue to push back. Could you say what is so important for this dispute?
EL: Compositionality is important. Punkt! Most of us got into semantics worrying about productivity – i.e., about how a finite being can understand indefinitely many sentences? Comprehending how this is possible obviously invokes compositionality – the meaning of a complex is built up out of the meanings of its parts. Closely related but distinct motivations include systematicity and innocence. How can you understand a complex without understanding its parts (assuming the complex is not idiomatic)? How can you understand closely related but non-synonymous sentences without compositionality; for example, since you can’t understand “The cook questioned the plumber” without also understanding “The plumber questioned the cook”, mustn’t we suppose that the meaning-contribution of an expression is constant throughout its occurrences?
3:AM: In that collection, prototypes as well as supervaluated approaches to vagueness get slapped down don’t they? Why are they in trouble?
EL: Our paper critical of supervaluations as a technique for explaining vagueness has caused me personally a great deal of consternation. I am not someone who stubbornly endorses a position regardless of its criticisms. I’ve changed my mind about many issues over the years, even in print. Yet Fodor and I remain convinced that we were right in that paper, even though no one else is. Close friends, and even others equally skeptical about supervaluations, tell us we got it wrong. We listen and invariably conclude that their criticisms place them in one of several niches we attack. It’s been frustrating.
Prototypes (stereotypes, conceptual or interferential roles) come in for a shellacking in our essays. Our argument is embarrassingly simple: meaning is compositional; prototypes aren’t. So, prototypes aren’t meanings. I could belabor the sallies to and fro between us and our critics but our papers are so transparent that I doubt I could add anything. Brandom is a (non-solipsistic) meaning/conceptual role theorist; we argue he is therefore stuck with a variety of familiar unhappy consequences: e.g., if any change of mind results in a change in every belief; and, no two people can assign the same content to their sentences (and thoughts), so disagreement is unlikely.
Churchland, facing these problems, invoked similarity of meaning to try to avoid it. We argued that sense could not be made out of similarity of meaning without a prior robust notion of meaning simpliciter. Brandom took a different tack, embracing meaning holism but saying that because speakers share words conversation can still flow. To be honest, we’ve never understood this reply.
3:AM The other book with Cappelen is Language Turned On Itself. It’s great because it looks at something that sounds as if its subject is something terribly technical (metalinguistic discourse) but is really about something we all do from the very start and keep doing forever – quotation. This is a great example of something dead familiar that philosophy opens up and shows how strange it is. Quotation seems to behave in ways many theories of language can’t accommodate, doesn’t it? So can you say why this is a subject that raises serious philosophical issues for anyone wanting to understand language and minds and perhaps give us a taste of the intriguing nature of this phenomenon?
EL: Quotation is special; I can recognise a word as a name or a predicate and never figure out what it names or predicates, but I can’t recognise an expression as a quotation without knowing what it quotes. To explain why this is so, we supposed that quotations include the very items they quote as constituents. It’s as if a name carried its referent around with it. This idea has many interesting consequences; one is that infinitely many expressions in English cannot be spoken; infinitely many others cannot be written, and so on, for each distinct manner of articulation language affords us. Since we can quote graphemes and pronunciations, and since what we quote must be a constituent of the quotation expression we use, it follows that if we quote a grapheme, for example, we can’t say it out loud; and vice versa for the quotation of a sound.
In various published pieces on poetry and metaphor, I exploit my theory of quotation to explain, e.g., the heresy of paraphrase, arduously defended by the New Critics and their guiding light T.S. Eliot. The thesis that poetry resists translation (and paraphrase), I argue, is best explained by treating poems, in effect, as quoting their own articulation – by which I mean the vehicles by which the poem is presented; which can run the gamut from its typography, its font, its rhyme, its meter, and sundry other poetic effects. If I’m right, then, of course, a poem cannot be translated since its translation (or paraphrase) will result automatically in a different presentations, and therefore in a poem with partly different subject matter.
3:AM: I think quotation hadn’t really been the subject of great philosophical reflection until you two waded in. Another subject you’ve turned to is words. You ask: ‘What is a word?’ So how should we start trying to answer that question?
EL: I got into the topic of what a word is through worrying about quotation individuation. Here’s an important distinction, one I just alluded to, that philosophers completely ignore. Suppose there are two people who are equally proficient with English but only one can read and write it, while the other can only speak it. The two thereby cannot communicate even though they share a language. This simple thought experiment makes clear we must distinguish between the vehicles with which we articulate language and the expressions these vehicles articulate.
This immediately entails that Davidson and others were wrong to identify an expression with its shape or sound: doing so conflates words with their vehicles of articulation. My work on word individuation starts there. When I looked around, I was stunned by how little has been written on words.
The main paper was Kaplan’s. Around this time, John Hawthorne helped me appreciate I (as well as Kaplan) was engaged in a familiar metaphysical project. He charted out the possible moves, and we collaborated on figuring out what was wrong with each; in the end, we concluded that our linguistic practices don’t settle on exhaustive necessary and sufficient conditions for word individuation. When our interests are historical, intuitions go one way; when they are about communication, they go a different way. We are waiting for replies.
3:AM: It’s striking how much collaboration figures in your works. Your collaborators read like a premier division team in the field – including Fodor, Hawthorne, Borg, Cappelen, Loewer, Ludwig, McLaughlin, Stone. Is it your preferred mode of operation? Is collaboration a reflection of the subject matter or more a reflection of the way you personally like to work?
EL: I believe my collaborations emulate the collaborative nature of cognitive science: good work happens by engaging with others. Still, each collaboration differs from every other; sometimes a philosophical problem spans area boundaries and ultimately requires insights from others disciplines both in an out of philosophy. In these cases, I require a collaborator who knows something I don’t know. This happens more often than I’d like to admit. And even the style of collaboration varies; sometimes I write the first draft; sometimes a collaborator does. (If you entertaining philosophical collaboration, here is some advice: write the first draft for yourself. The person who does this dictates the direction of the rest of the investigation.) With Fodor, we wrote every word together, side by side, fighting over the keyboard, often with Jerry’s cat Mr. James, all fifty pounds of him, refusing to move.
3:AM: Looking at the current situation in your field, are we closer now than when you started to understanding the relationship of mind and language? Have certain ideas been killed for good and are others beginning to seem like secure?
EL: I believe the philosophy of language is going through a second renaissance. In the 90s, I kept hearing it was dead; that it had peaked in the 70s with Montague, Davidson, Kripke, Lewis, et al but that was all over. Yet I kept discovering philosophers advancing contextualist theses to dissolve one ancient philosophical conundrum or another. Suspicious from the beginning, I wondered how generations of philosophers could have missed so much context sensitivity. No one missed the context sensitivity of “I”, “now” and “here”; can it really be that different with “know” and “heap” and “penguin”? More interestingly, the very same epistemologists, metaphysicians, logicians, ethicists, and aestheticians who told me philosophy of language was dead were now advocating a rather liberal position on the scope of context sensitivity in natural language.
That’s philosophy of language, alive and flourishing, after all. What separates the 70s from the present is that young philosophers of language are better informed than their teachers about the nature of language. It will no longer suffice to permit a small range of belief sentences to command an entire philosophy of language. This doesn’t mean philosophers should become second-rate linguists. Philosophers have interests in language that linguists don’t, and shouldn’t, but we can still learn from them. Ignore them at your own peril.
In 1997, after becoming Director of Cognitive Science at Rutgers, I sponsored several small inter-disciplinary workshops in semantics and pragmatics. There have now been nearly twenty of them. They are no longer small. The last one included nearly seventy participants. They bring together young and old philosophers, linguists and computer scientists. Some tend to be more philosophical than others, some more linguistic, but the one constant is that no participant is advocating isolationism. When philosophers advocate a priori that their research cannot be informed by anyone else, the jig is up.
3:AM: And finally, are there five books you could recommend for readers here at 3:AM Magazine (other than your own which we’ll be dashing away to read straight after this) that will help them understand more about this area of philosophy?
EL: Different books have had impact on me at different times in my career. Probably the first to have a great impact on me was Quine’s Word and Object; at one point I knew it so well that you could read a passage and I could tell you which page you were reading from. Hey, I was a graduate student. Sadly, it may not have stood the test of time; or perhaps it’s just my presentational skills. I taught it in a graduate seminar two years ago, and Gil Harman and I worked through it in our NEH Summer Seminar two summers ago; the reaction both times was “This is just behaviorism”. I’m still puzzled.
Interestingly, both seminars reacted differently to Davidson’s work; though largely critical, they recognised Davidson’s work as important and philosophical. Davidson’s writings were hard to come by when I was an undergraduate and a graduate student. They were scattered in relatively unknown and not easily accessible journals. This added to their mystique. I recall the excitement when ‘Radical Interpretation’ or ‘Belief and the Basis of Meaning’ surfaced. It wasn’t until 1980 and 1984 when OUP published the first two volumes of Davidson’s collected papers that I had books by him. I must have taught those collections a dozen times over the years. Trying to pull together a cohesive project or program has taken up a substantial chunk of my professional life. I think it was worth it; over and over again, I came to realise that some idea of Davidson’s that I found too unclear to defend or too peripheral to worry about took would up taking center stage. This is certainly true, e.g., of his event based semantics and his paratactic accounts of direct and indirect quotation.
Kripke’s Naming and Necessity is an obvious choice; anyone who doesn’t list this book is either insincere or just not thinking. I say this as someone weaned on the opposition – Davidson and Quine. Their books are also significant but if you are looking to engage a young student; I doubt there’s a better book than this one.
Dummett’s first book on Frege I remember devouring as a young faculty member at Notre Dame. I couldn’t put it down. I found it mesmerising. I had no one to discuss it with at the time; but my engagement with the book was enough for me.
More recently, Lewis’ Convention has captivated me. It took me a long time to appreciate its significance. Philosophers have a reputation for not answering the questions they ask, but Lewis answers his questions and his answers are correct. I once asked Fodor why Lewis’ book isn’t more widely discussed; he replied, “What are we going to say, ‘Lewis got it right’?” Under Lewis’ influence I’ve begun to see that if you intend to separate semantics from pragmatics along familiar Gricean linguistic vs. psychological lines, Lewis’ work will make you rethink hard about where you draw that line.
There is a lot more conventionality in linguistic practice than philosophers typically acknowledge; for example, Kripke (following Grice) tells us there are psychological reasons to infer a speaker disagrees with his interlocutor when he replies to her utterance of “He put a red handkerchief on the table” with “It looked red”. But the intonation pattern with which one utters this second sentence conventionally determines whether there is disagreement or not. Psychology has nothing to do with it. Reading Lewis helped me see why this must be so.
ABOUT THE INTERVIEWER
Richard Marshall is still biding his time.
First published in 3:AM Magazine: Monday, October 29th, 2012.