Geoffrey Nunberg on Language

By Francisco Hulse

Our May speaker, Dr. Geoffrey Nunberg, is an adjunct full professor at UC Berkeley’s School of information and a researcher at Stanford’s Center for the Study of Language and Information, as well as a consulting professor in its Department of Linguistics. In addition to writing books and commentaries on language, he can be heard on NPR’s popular program “Fresh Air.” (Editor’s note: this transcript has been edited slightly in certain places to enhance clarity.)

Francisco Hulse: Your interests in the field of linguistics are many and varied. If you had to describe your main work and passion in a sentence or two, what would that description be?

Geoffrey Nunberg: In linguistic semantics, most of my work has to do with the phenomenon of polysemy: the way in which a single word can have a number of senses. “Newspaper,” for example, can mean a kind of publication, an instance of that publication, or the company that publishes it. It’s a systematic regularity that obtains across languages, unlike the accidental homonymy that makes “bank,” in English, for example, mean both “side of the river” and “financial institution.” So the patterns that underlie polysemy are, linguistically, of more interest than the accidental patterns that underlie homonymy. I’ve also worked a lot on the phenomenon of deixis: words whose references are determined by the context of utterances… Words like “I” and “you” and “here” and “now”, and demonstratives like “this” and “that” and “those” and so on. That’s a topic of considerable interest to both philosophers and linguists and I’ve done a lot of work in that area.

What is the most exciting work being done in the field of linguistics today?

There’s exciting work being done all over the field, but I think one area that’s particularly interesting is in corpus linguistics. It uses these enormous corpora (bodies of text) that are now online, whether historical or contemporary, where you can do statistical analyses of corpora and look at frequencies of this versus that and it really changes the way linguists do a certain amount of their research. Rather than having to say, “well, can I say such-and-such? Is it grammatical to say such-and-such?”, you can just go out there and see what people are saying.

How did people used to do this? Was there a hand-count method?

No, usually they would just consult their intuitions about whether such-and-such was grammatical or not … or they might dig out the odd example from a newspaper or a text, but that’s very difficult to do and you can’t do this kind of statistical classification. If I’m interested, for example, in the difference between “in the circumstances” and “under the circumstances,” to take one example that a colleague has worked on … I can look to see whether one is more common than the other, and whether the immediate context has an influence on that fact. Using those kind of data, I can make statistical generalizations about the use of this that will point to an analysis.

Before the existence of electronic corpora …

You either did laborious hand counts (provided that the phenomenon you were interested in was frequent enough!) or you just consulted your intuitions, or didn’t deal with the frequencies.

In the nineties, you wrote about the possibilities of books, and even libraries, surviving the age of electronic reproduction. Now, a decade later, how would you update your thoughts on those issues?

I don’t think things have changed much. It’s clear that paper books have a continuing role to play, and I think people are no longer quick to predict the disappearance of the book. In fact, e-books, for example, which were to have a very bright future, turn out to have had a relatively limited future. They’ve been available for ten or fifteen years, and people keep saying, “well, the technology will get better, and then we’ll do it” … and people don’t seem to want to read books on little hand-held readers. It’s very much a minority interest.

That’s true. You have iPods now that can hold several hours of video … text is much more compact, and yet …

And even with bigger and better readers—and there are some that give you more of the feel of the printed page—people just don’t want to go the trouble, or are just comfortable reading books. I mean, the book is not “broken” as a form. That said, there’s a lot more digitization of books going on, and at some point we can expect to have a very large part of the printed record available in electronic form, which is all to the best.

Any thoughts on digital rights management in that arena?

That’s all very complicated. Copyright issues and so on, the publishers are struggling to come to grips with that, and it may be a while before those issues are addressed. The publishers are very nervous about just letting this stuff out there in electronic form. On the other hand …

Who’s going to buy the book if they can read it online for free?


But then you’re back to the same question of why people don’t like e-books: personally, I don’t relish the idea of reading a whole damn book off the screen.

Right. I think that’s true. We’ll see what happens with that. I don’t have anything too strong to say about that.

Although Google is by no means the only organization digitizing the world’s books and printed information, it is perhaps the most well known. What is the state of affairs of this endeavor—by all involved in this effort—and what are the challenges that the digitizing institutions face?

People sometimes have the feeling it’s all been done already, but really, the surface has just been scratched. Google is doing it, in collaboration with a number of libraries; so is the Internet Archive, here in San Francisco (Brewster Kahle is the guy who does this — in collaboration with Yahoo, IBM, and some other people to do it); there are individual projects going on all over the place… I was just at Texas A&M for example, they have acquired a very important, small collection of illustrated versions of Cervantes, and if you go online, you can see all the illustrations over the years of Don Quixote, and they’ve digitized them, and they’re available. Now, that’s a very small collection; it’s a historical collection; it’s one library. But that’s how these things are coming on, the interesting things in particular in corpora.

What’s harder is the great body of material that’s under copyright. That’s coming on only slowly. So what you have, for example, if you look at Project Gutenberg, there are a lot of classical texts there, but they’re almost all taken from these really bad 19th-century editions that are out of copyright. And the cleaned-up editions that you buy from Penguin or Modern Library or whatever, with the proper apparatus of footnotes, and so on and so forth, they’re not available. They’re still under copyright. So if you go online to look for A Sentimental Education, for example, you’ll get a bad 19th century edition of the book. That’s a problem. Only a tiny fraction of what’s out there has been digitized. And it’ll be a long time. Also, the cost of doing these things is very expensive. It might cost $50 or a $100 to digitize a book, just in the terms of the time and the equipment and so on. Well, multiply that by the hundreds of millions of books that are out there, of different titles, and that’s a lot, too. And it costs more if they’re older books, as well. Because they’re rare, you have to take more care with them; they’re not standardized; it may involve more hand correction of the images … The thing is, it has to be done well the first time, nobody’s going to go back … A lot of the earlier digitization was done by the French National Library at screen resolutions that weren’t so great, or without hand-correcting certain things … If you don’t do that right the first time, it’s just prohibitively expensive to go back and do it again.

Do they then apply an optical character recognition software to it?

Sometimes they do; other times it’s a question of working with software that flattens the page, so to speak. You have an image of a page that’s curved, and there’s software that can actually flatten it. But that often has to be done by hand: sometimes the machine tries to do it automatically, but sometimes you really have to do that by hand. Sometimes it’s just matter of making sure that the image was properly captured, not blurred. Go to these digitized books, you often see there’s a blurred page. That’s true with the digitized newspapers, for example.

Because to make it searchable, you have to have OCR to turn it into ASCII characters.

… you have to have OCR. OCR is pretty good. I’ve just been looking at Time Magazine’s archives, because I’m doing a piece on Time tomorrow, for Fresh Air, so I was looking at back issues of Time. It’s well digitized: well imaged. It’s not like an old book: the pages are flat and they did a careful job in scanning them. And the OCR is pretty good: it probably gets more than 99%, maybe 99.5% accuracy, but …

Do they go through and correct it afterwards?

Well, some of it, I suppose they do; lots of it they don’t. You keep seeing cases where they haven’t: where an O is rendered as an E, or an E as an O, or something. You know, 99.9% accuracy means that there will be a mistake on every page. So you get a sense of how hard it is do to well.

When might an information consumer be able to access any book, on demand?

It depends. I think it will be a very long time before you can all books currently under copyright, just because of the legal issues.

Ten years? A hundred years?

Because it’s a legal question, it’s not possible to answer it. It’s not a technological question. It’s a question of the economics and the law, mostly. If people suspended copyright law and congress voted a couple of billion dollars to do it, it could be within a relatively short period of time, but neither of those is likely to happen. So it’s a question of when all these people get around to doing it and how they get around the copyright problem, and then what it will mean to access it. It may be that you’ll be able to access it in the clumsy restricted way that you get now from Amazon or Google, where half the pages aren’t available, and you have to look at them in this reader, and you can’t print the pages, and so forth. So we can access a lot of stuff at Google Books now: I do it, but it’s a pain, because you have to do it in their reader, and there are pages that are missing, and you can’t do all the searches you want to do on it, and so on. And it’s done deliberately, to get around copyright. You can see a little section, but you can’t read the whole thing. You wouldn’t want to try to read a whole book in Google Book’s reader or Amazon’s reader. So what it means to “get any book online”, that’s hard to know, too. But it’s a ways away. Probably even now, the total number of books that have been digitized is a small fraction of 1% of what’s out there.

You wrote “The Persistence of English,” as the introductory essay to The Norton Anthology of English Literature. Give us an update on what you believe is the “persistence of English” in the modern, globalized world.

The status of a language doesn’t change much in 8 or 10 or even 25 or 30 years … My sense is that a couple of things are clear: first of all, English, the dominance of English, is probably exaggerated. The web, though it looked like it would be a force for the globalization of English, is actually becoming much more interesting as a way of spreading the influence or maintaining the influence of smaller languages: simply because it isn’t subject to the economies of scale that other forms of publication are. So, for example, if you just take news: before the web, it was really only Anglophones, and really only Americans, who could routinely access “their” news in all foreign capitals. Even major languages like Spanish and French and German, you couldn’t do that. Now, of course, you can do that for Czech or Greek, or much less influential, much less important, you know, smaller, national languages, such as Danish, for example. Similarly with cultural products, like music and, ultimately, film and television.

Cultural products that you can stick in a wire and send someplace.

Right. So, for example, I think I saw the number somewhere, there are about a couple of thousand movie screens in France … If they’re having to decide whether they’re going to show a certain French film director or Steven Spielberg … If you’re showing Spielberg, you can’t show a French film, right?

Right, it’s a zero sum.

If the radio station is playing The Rolling Stones, it can’t be playing Gainsbourg. There isn’t that forced choice anymore. One language can’t crowd out another on the web. So if, for instance, a French film or an Italian film doesn’t get distribution in the United States, which is increasingly the case, it can at least be seen on DVD or over the web and so on, so electronic media, in a certain sense, levels the playing field among languages. Beyond that, it’s hard to know. Will English, or one form of English, continue to be the dominant language for international business, trade, commerce, politics, science? Sure! And it will be English: this form of written English. Will English remain coherent in the sense of coherence among its local varieties in different places? Less clear.

It’s not coherent now!

It may be less so. So it’s hard to know. You’re looking at periods of 50 or 100 or a couple hundred years to talk about relative dominance … assuming no nuclear [annihilation]… so you’re looking at a very long period. I haven’t changed my opinion since then. I think English will remain dominant in that form for those purposes. But the web actually helps, as I say, minority languages, the smaller languages.

Do you speak other languages?

French, Italian, some German, un poco de español.

I have more questions about the other languages you speak, a little later on, but I’ll come back to those … Are translators more valuable today, in a world of globalization? If so, how?

Sure, translators are crucial! One of the things that’s become clear, that should have been clear all along, but wasn’t, is that the translator’s job can’t be automated. At least in important contexts. Translation software is enormously useful, and enormously useful to translators as well or to consumers of translation. In the European Union, for instance, they use it … If you get a letter in Bulgarian, run it through an automatic translator, to see whether it’s worth really translating it. But you’re not going to get automatic translation to even do what seem routine jobs like translating meat-packing regulations from Dutch to Italian. So the need for translators increases. Partly because every time the European Union adds another country, you get all these new pairwise combinations. Or if you make Catalan an official language, then you get I-don’t-know-how-many more pairwise combinations. So globalization and trade just increase the need for translators: of laws, of regulations, of commercial documents, of scientific texts … and of course, the need for literary translators is as it always was. People continue to write literature in their native languages.

What are the major issues—linguistic, cultural, political—that translators face?

Well, they’re the same issues they’ve always faced, really. That’s too general a question. What can I say? Translators have to deal with all that stuff; that always happens. It hasn’t changed much. There are more these legal questions of how official a translation … you know, there are term banks, [others] …

It’s really more of a question aimed at a translator, which you aren’t.

These problems for technical translators, there are special issues involving “simplified English,” or in term banks, what do I call a “dialogue box” in Russian or something like that, and there’s somebody that’s made a list of all those things, and consistency and whatever, but … I know one guy, I don’t know what he’s doing now, he’s spending a lot of time just … he was trying to build software that could determine … you have a bunch of say, Xerox copier manuals. And Xerox, as he pointed out, never translates anything into one language. If they translate into one language, they translate it into ten or eleven, or whatever. So you have all these varieties of all these manuals, online, and so on, and you merely want to say, well, here’s a paragraph of such-and-such copier manual in English. Find me the translations of this in all these things. So it isn’t a question of translating: somebody’s already hand-translated; it’s a question of saying “this paragraph is a translation of that paragraph,” and that’s hard to do, for lots of reasons. Even if you were to isolate it, sometimes the paragraphs aren’t the same … That turns out to be really hard to do, and useful. Because you don’t rewrite a whole documentation every time you have a new machine. The thing on how to push the button is still how to push the button. If you’ve translated that, you may wind up translating it five times, because you didn’t know … it wasn’t worth your time to find the translation of that paragraph…

Because this year, when they put out the model XZMQR5, it’s an update of XZMQR3.

So there’s lots of issues like that, but that’s technical…

The big thing that a translator would answer on a question like that would be talking about things like client education. I just read a very interesting article that had to do with the next question, on the role of the translator as a copywriter, and how somebody hands you a product slogan, advertising copy, and it’s a minimum number of words, and it comes in under a minimum charge if you’re just charging them by the word … If Nike wants their slogan “Just Do It” translated —

— they should pay a lot for that.

Exactly. What does the term “localization” mean to you? Are translators becoming “localizers”?

That’s not my area of specialization. I know what it means and I know … again, it’s too complicated. Yes, translators have always been localizers. It’s always a translator’s job not simply to render the text, but to render it relative to the social, cultural, economic, legal circumstances of the setting where it’s going to be used, and translators have always had to do that. I mean, “localization” is just a name for something that good translators were always doing, anyway. The members of this society [NCTA] know much more about that and have much more experience with that than I do.

What is the state of machine translation? What can we expect from it, and when?

It’s enormously better than it used to be, that’s the first thing, and the second thing is, it’s very hard to tell. The reason is because getting from … just pick an arbitrary number: 15% to 30% is a huge leap, but it still falls so far short of where you want it to be, your understanding of where it should be, that you’re not going to notice it, perhaps, as an improvement. I asked a friend of mine, who is the dean of this field, once, “if you asked people working in machine translation how long it will be until we have perfect, idiomatic machine translation of text …?”, “they would all say about 25 years. And that’s been a constant since 1969.” He gives wonderful examples of this, he has lots of examples of why it’s so hard to do it well: the different verbs in German that all mean “remove,” depending on whether you’re pulling something up, or out, or over, or sideways, and so on and so forth. You can only know if you understand the organization of automobile motors, say. “Remove the fan belt” is not the same as “remove the carburetor” is not the same as … because the gestures are different. He has lots of amusing examples of that sort.

How would you describe the role of linguistics in the Iraq war? That is, have the coalition forces done a good job in communicating with the people in whose country they are fighting?

That’s a silly question! [HEAVY IRONIC TONE] Yes! They’ve done an excellent job, which is why we’re winning! [Resumes normal tone.] No, that’s a set-up.

Can you envision, in the future, a truly international language, on the model of Esperanto? Or is that language simply English?

No, I can’t envision it. I think that since the late 19th century, it’s become clear how complicated languages are, and how difficult it is to design one that’s an improvement over any individual language. All the idiosyncrasies of particular languages just come with language.

So if you build another Esperanto, it’s going to pick up idiosyncrasies of its own?

Yeah, there’s a reason that languages have irregularities for example, they develop because they facilitate other features.

Give me an example.

For example, pronunciation. You’ll have irregularities in form, sometimes, just because they facilitate pronunciation. They reduce certain clusters for example, consonants. Why, in Italian, do you have two forms of the definite article: “il” and “lo”? The latter, used before S plus a consonant? And the answer is that those clusters of S plus a consonant have always been problematic given the syllable structure of Romance. In Spanish, an E was added before them, in French, they became an é: étude. In Italian, they remained as they were, but a special form of the definite article was devised … So irregularities of that sort or complications are constantly arising because each language is sort of …

… coming up with a different strategy?

There are lots of problems. You pay in one area to simplify in another, and so on. So if a language like that were genuinely used by large numbers of people, it would a) become localized as each [unintelligible] to its means and b) develop irregularities and inconsistencies and so on the way languages always do, and c) it would have a point of view. The idea that you can have a neutral-point-of-view language is absurd. Every language has a point of view. Do you chop up reality this way or that way? Are you going to have two verbs for “to know,” or one verb?

One of my favorites is how, contrasting English and Spanish, with their division of body parts: in Spanish, these [waving hands] are fingers, and those [waving feet] are fingers. But these [holding ear] are outer ears, and those [sticking finger in ear] are inner ears! Orejasand oídos. These are the garden-variety words, but [waving hands and feet again] dedosand dedos.

One thing I always have trouble with in foreign languages, even, say, Italian and French, which I speak relatively well, is the word for “face.” Because you have, for instance, in Italian, “faca”, “vizo”, “visto”, “muzo”… I guess that’s it… There’s another one I can’t think of now… But then they all have different senses, depending on whether it’s the whole face, or just the features, depending on whether you mean it metaphorically or not, I mean, I know which word to use in a given … there’s a whole large set of contexts, but I don’t, maybe, have a general theory that predicts everything for me, so … do you register reality differently?

… and in Hebrew, it’s plural! Another one of those weird options that they have plural for, and we don’t. Our myopia in America toward English-only limits our ability to understand other cultures. How do other cultures’ knowledge of English (limited though it might be) affect their perceptions of the U.S.?

I probably would put that the other way … I mean, I would say that their perceptions of the U.S. affect their knowledge of English. Why do they learn English? Because it’s the language of Shakespeare, as some people like to say? Does it represent a judgment: “I prefer Shakespeare to Cervantes or Voltaire or Goethe?” No! Obviously, it’s because of the enormous cultural, political, economic and scientific influence of the language, so, because they appreciate that, and because they want to travel in the world, and not just the English-speaking world, although in the English-speaking world and America in particular, but if you’re an Italian, you can go to Germany, English is going to be more helpful than …

… not more helpful than German …

… but, then if you want to go to Czechoslovakia the next week, as a practical matter, English is the language of travel.

Right. You get two for one.

And there’s a kind of … American cultural products, particularly film and music, have been … or Anglo-American cultural products, particularly film and music, have been enormously influential, partly because of the general impact of English and partly because the culture, with it’s appearance of democracy and its early discovery of popular culture, I mean, it’s been absorbing influences like black culture, and so on, it’s been very influential and that’s another thing … that increases their desire to learn English, so that they can listen to the music and watch the movies. It stands to reason that no language can be as influential for English-speakers as English is for anybody else. That’s almost just a mathematical principle. So for English speakers, there’s less incentive to learn foreign languages for those reasons. For me, as someone who has worked against “English-only,”—I’ve exaggerated, you know, on behalf of these issues—I often point to the economic importance of speaking foreign languages … Actually, it isn’t that great. And I’ve talked to people at the conference board and they say “look, apart from a few cases like China, we really don’t hire … if American corporations don’t see a need to hire people who speak the local [language] … don’t see that much of an advantage in speaking [it] … and there are certain industries where that’s important.

What do they do? They hire interpreters and translators?

You hire interpreters and translators, you hire local people, or you hire … or your people are trained in those languages if they go there for a short term. The reason China was an issue, and it may have been less of an issue now, is that they had problems hiring [unintelligible]

Hiring Chinese?

Hiring Chinese in mainland China, for a number of years. Now, I think that’s changed. But you had a big market and for various political reasons, it was difficult to hire Chinese, and who was working for whom, and so on and so forth, and the government was intrusive … But that was a very special case, and there are a few others like that … If you have an American who grows up speaking Arabic, that’s one thing, but it’s very hard to learn most of those languages, most of those languages are very hard to learn. So the economic reasons for learning other languages really aren’t as strong as some of us advocates have made them out to be, frankly. And the cultural reasons are the cultural reasons. I mean, I think Americans should know Spanish and I think it’s pretty good if you live in this part of the country or the northeast, or whatever, it’s just [unintelligible] but there isn’t the strong incentive that speakers of other, for instance, other European languages, have. It just isn’t there.

It’s a little bit of a segue to the next question: What is your opinion of the push in some quarters to make English the official language of the United States?

I’ve written and worked against that. It’s a terrible mistake for a number of reasons. I’ve called it a bad cure for a non-existent disease. It isn’t going to help anybody learn English. People don’t learn English so that they can apply for a driver’s license or read Department of Agriculture pamphlets. It isn’t a problem. Hispanics—and it’s aimed at Hispanics—Hispanics in America want to learn English and the second generation of Hispanics does learn English. The first generation has a lot of trouble with it for lots of reasons: it’s hard to learn another language when you’re older; they’re usually living in monolingual communities, Spanish-speaking communities; they’re working at jobs where they have very little access …

… they’re often working two jobs!

… there are long waiting lists to get into English-as-a-foreign-language courses … but the second generation just learns it. It’s going to cease to be a problem within a period of time. Even if immigration continues, it’s … People may remain bilingual longer than they have, historically, but, there’s a way in which people confuse … there’s a systematic ambiguity with the word “bilingualism” that proponents of things like “English-only” exploit: where they talk about America becoming a bilingual country. Now, that can mean two things: we can maybe become bilingual in the sense in which French-Canadians are, or the Swiss are, in the sense that everybody speaks two languages. What’s wrong with that?

Nothing at all!

But they use it to mean a country divided into two monolingual communities.

Right, like countries that have minority languages within them, like France, with Provençal

Well, not even that. I guess the model would be more something like …

Spain with Basque?

Catalan and Basque and so on … But they see it as politically divisive, and it isn’t. Multilingualism in a population is not divisive.

I don’t know if completely agree with that!

No, not by itself.

No, not by itself.

In fact, multilingualism by itself, if you look at the surveys, is a very poor predictor of political instability. Now, religious difference, if it correlates with multilingualism, yeah! I mean, you have cases like the former Yugoslavia, where you have these languages that basically have been mutually intelligible, historically, but they’re written with different alphabets and they correspond to religious differences and so on and so forth and then you see, what are by the standards of English, just separate dialects, becoming the loci for intense struggles about multilingualism and so on, but it’s all bogus, because the languages really aren’t that different, it’s just that they become the loci for these national aspirations.

Do you have an example for one of those?

Yeah, that’s the former Yugoslavia. We used to talk about Serbo-Croatian, right?

Right. And one is written in Cyrillic, and the other’s written in Roman?

Right: Serbian is written in Cyrillic, Croatian is written in Roman. The vocabularies are different, but actually, they’re different, not in the way that say, English and American, British and American are different, but what tends to happen is that with British and American, you have common abstract vocabulary, and then we’ll have different words for “truck” or “sweater” or vegetables, and so on. There, it will be the opposite: it’s the abstract vocabulary, because it’s based on religious and philosophical traditions that will be different, and the words for “sweater” and “truck” and “rutabaga” will be the same.

How bizarre!

There again, it’s a bogus … the idea that multilingualism underlies the …

… no, it’s the religious tensions that …

… yeah, and similarly in Sri Lanka, these are always independently religious or ethnic struggles that are maybe manifested through language, but language itself is very rarely a predictor of this.

Although with the ethnic split being a possible predictor, the “English-only” types I’m sure would point to the fact that those folks there are coming from a Hispanic tradition, they’re largely Catholic …

They’re Catholic, it’s a different culture, they’ve got the cultura de la mordida [custom of bribery], as they always say, and so on down the line, and those are cultural differences. Those cultural differences don’t …

… amount to much?

Well, they don’t evaporate; they are what they are, right? They don’t evaporate when people learn English, unless they become assimilated. They can learn English very well and still remain as they were. And the whole argument underestimates the allure of American language and culture. Here you have people in France passing laws to keep American movies, American language and so on out, right? And you have these other people coming here, and people say, “we need laws to make them speak English.” They don’t need legal incentive to do that!

What are your current projects?

I’m working on this book on assholes, on the notion of assholes, which I think is an interesting problem.

Say 20 words about that.

It’s really about civility in America. In one way, it’s sort of a proof of a methodology. The question is, what can you learn, by looking at language, about a culture? When people look at that, when people look at cultural history, they’ll often look at words that have an independent interest to historians, to political scientists, to economists, to [unintelligible] look at the history of the word, if you’re a literary historian, “originality,” or something, or for a political scientist, look at the history of the word “liberty” and what did it mean in the 17th century … Now, that’s interesting but that’s by way of using language to try to get a hook on other problems that arise independently in the field. What I’m interested in is not different, really, but it’s sort of looking at language … looking at things that you can only understand from the lens of language or where language gives you unique perspective on social change or social mentalities and for that reason it’s very often most interesting to look at the kinds of words that don’t figure in these discussions. So I’m looking at the word asshole, for instance. Around the middle of the 20th century, this word emerges as the standard term in colloquial English for someone who’s irritating, inconsiderate, uncivil, arrogant, and stands at the center of our, so-to-speak, folk theory of civility. You know, what makes somebody an asshole? That’s where that theory lives. Not in the terms of what we think the term civility means, or politeness. Those are words that you … that come top-down, but this is the one that comes bottom-up. And it is, moreover, a word that is interestingly banned from public discourse because of its origin.

Sure. It’s gutter vocabulary.

It’s considered uncivil. So to accuse someone of being an asshole is, at once, to accuse that person of incivility and then to be guilty of something that other people would consider an act of incivility in itself. So, I’m sort of interested in that. People have written a lot about where civility is [unintelligible] America as a sort of central problem in a certain kind of sociological talk about the modern world and America … the downfall of civility and community and so on and so forth. And I’m trying to look at it from the [point of view of the] … people … If you ask people about civility in America, they’ll give you these bromides that are what they think they ought to say. If you look at the way they use a word like “asshole,” you’ll learn much more. So I’m interested in that, and actually the way that the notion plays in political discourse, and so it grew out of my interest in political discourse: this idea of “asshole,” the characterization of others as assholes. As I say, it’s something I’ve just begun to work on, so it will be a while before I can give you [unintelligible]

Any other projects that you’d like to mention?

No. I’m teaching at the School of Information, courses in the history of information and the use of information technologies and so on. I kind of do that in the fall, but right now I’m off and trying to write. And doing my normal public-radio and other writing.

Which did you learn first: French or Italian?

I learned French in 8th, 9th grade, I guess, or started [unintelligible] and Italian I learned fairly late. I didn’t learn Italian until my early 30s. But I feel more at home now speaking Italian than speaking French, though I still read and write French better than I do Italian.

That segues into my next question: When did you learn Italian? You said in your early 30s?

Yeah, I just went there with a Fulbright and I sat down with a grammar book and learned a bit and then si put adentro, as the Italians say, you dive into it. Putare is to dive … er, no, to throw. You know, to throw yourself into it.

How old were you when you felt fluent?

Well, after a couple of years. I feel fluent now. I was talking Italian the other day … Every once in a while, I have to grope for a word, or go around the barn to find a way to say what I want to say … I say, “shit, I don’t know how to say that”, but, no, I feel fluent in Italian. I feel a little rusty in French, but I can still … it’s a question of what the purposes are. If I had to give a lecture tomorrow, I would be sitting down, looking words up …

… and you’d make sure to have a written script?

Yeah, but if it’s a question of just having a conversation, or something, or going to dinner [unintelligible] I can do that.

Did you live in Italy? How long?

I lived in Italy a total of maybe two years … not a long time. I lived there for a year in 1981 I guess, and other visits of several months and I was there for ’99 for almost a year [unintelligible].

Same questions for French …

French? I lived in France for about 3 years, on and off … two and a half years.

And you started learning when you were…

Well, I started learning in school, but I didn’t really speak French until I got there. I could …

… ask for directions?

Well, I took it through high school and college, but you … it’s astonishing how little of a language you learn after all those years.

Same question about fluency. When did you… how old were you when you felt fluent?

When I lived there, after I’d lived there for a year or two.

Here we get into the personal questions: You were married to a French woman for … how many years?

For ten years.

What language(s) did you speak to each other in?

Mostly English, unless we were in France. When we were in France, we spoke French more, but she’s completely bilingual. She lived in the states when she was a young girl, so she’s utterly bilingual, or as close to bilingual as you can be. Usually we spoke English if it was just the two of us.

Were you or was she resistant to speaking one or another of your common languages?

No. She speaks the language of the context she’s in and the switch just goes automatically, the way real bilinguals do.

So if you speak to her in French, she answers you in French?

Yeah, or it depends where we are, and so on and so forth. But there’s a switch in there: it isn’t conscious. I don’t think she very often consciously chooses … I heard her once translating [sic] for an architect who was giving a talk in Berkeley—Ricardo Bovián[?]—he’s actually a Spanish architect but he works in France. He was presenting his works and she was translating [sic] to English, and at one point, he said something in French, “cet œuvre a été construit en 1970”, and she says, “cet œuvre a été construit en 1970”. She was really not even aware of that, which is striking to me, because I can’t imagine not realizing that I’m not speaking [unintelligible]

As a person who works as an interpreter, that’s happened to me on one occasion or another, where I was interpreting in one direction or another and I just repeated instead of interpreting. That’s funny. Did you both speak French and English to Sophie [your daughter], or did your wife speak only French to her, leaving you to speak English to her?

No, she heard both and we didn’t really make an effort, we both spoke French. When we were in France, we spoke French, mostly, to Sophie, and as I say, among us, we used French more, just because it was in the … you know, we were in France. And Sophie grew up hearing both languages. She was dominant-French when we came back to the States in 1994, I guess. She was 5 or 6 … but very quickly acquired English. And now is certainly dominant-English, although her phonology is native in French.

Her writing?

Yeah, she’s studying French. [Tape ran out. GN kept talking, off tape about Sophie’s experiences at a French school in the U.S.] She came home the first day … They couldn’t decide what level to put her in when she started at the French-American School, because she was rattling away, but couldn’t read or write a word. They finally put her in an advanced beginners class, which she had some problems with at first, because she hadn’t ever written the language. But she came home that first day and she said, “Dad, you know vingt [20], has a G in it?”

Name (and explain) a grammatical feature or phenomenon of a language: specifically, a feature that you are still trying to wrap your head around.

Which language?

Any language. Pick a language. Something that you say, “in such-and-such a language, they do things this way, and isn’t that bizarre?”

I’m trying to think of examples that aren’t too…

… prosaic?

No, too technical, or too complicated. I mean, I was thinking about the partitive in Italian … so if I ask you in Italian, how many students are ready? You say “ne sono pronti dieci,” “of them, are ready, ten”!

That doesn’t seem that weird…

If I say, “cuanti toi studenti son inteligenti?”, you don’t say “ne sono inteligenti dieci.”

“Of them, are intelligent ten”?

Because you only do that with temporary states, not permanent states! So the permanence of the adjective affects the use of the partitive. It’s one of these connections that you see … In French, there’s a similar phenomenon. There’s a construction that you don’t learn much. You hear people use it all the time; you don’t learn it in school. If you call a restaurant, you say “est-ce que vous avez une table de libre a dix heure ?” “A table of ready at 10 o’clock?”! But you don’t say “est-ce que vous avez une table de grande a dix heure ?” [“Do you have a big table at 10 o’clock?”] Now, why that “de,” that construction, should depend on the stativeness of the adjective? So there are lots of things like that. There’s clearly an explanation, but it’s not obvious at all. And there are numerous things like that in English. I’ll give you another example in English. You can make a plural noun out of a nationality-denoting adjective, in many cases [without adding anything]. “He is French. The French do such-and-such.” “He is Japanese. The Japanese do such-and-such.” But “He is German. The German [sic] do such-and-such.” You don’t say that! What’s the generality?

What’s the rule, then?

I’ll let you figure it out … “The English, the French, the Dutch, the Japanese,” but not “the German, the Russian, the Bulgarian, the Greek, the Israeli” …

Those all take plurals [with an S]. Gosh, what is the rule?

It’s a productive rule. Does it end with a /sh/, a /ch/, a /z/, or a /s/? That is, does it end with a fricative consonant in that part of the mouth? Go figure!

So a fricative consonant makes …

A lingual fricative or an affricate. So, a /sh/ … there is none with a /j/, but it would … a /ch/, a /z/, or a /s/. So not an F or a B, but …

… any other …

Well, I don’t know… would you say “the Wolof”? Yeah, so maybe it is a fricative … There aren’t too many nationality names that end in F … So it may be just a fricative. “Wolof” is the interesting case. Can you say “the Wolof live in this part of the world”? Yeah, so maybe it’s just a fricative or an affricate. So if there were a language called Goov, could you say “the Goov”? I think you probably could. So it’s just whether it ends in a fricative. But those are these weird … I don’t know, I don’t know why I picked those … There are examples, but there are just these little goofy things, and we say “why is that”? First of all, it takes a long time to figure out what’s going on, and then you say, “well, why would that be?”

That makes a good segue to this question: Why is the last line of this excerpt from A. A. Milne’s poem “Nursery Chairs” so funny?

I’m a great big lion in my cage,

And I often frighten Nanny with a roar.

Then I hold her very tight, and

Tell her not to be so frightened–

And she doesn’t be so frightened any more.

In other words, why does the verb “be” bridle at taking the auxiliary verb “to do”?

“Be”, in most dialects, is the only main verb that cannot take “do” as an auxiliary.

It’s just a rule?

Well, it’s more complicated, because “be” is a very special verb, usually, but it’s not just … but “be” is the only verb that doesn’t take … “have” can function either as a main verb or an auxiliary: have goes both ways. You can say “she doesn’t have a bike” or “she hasn’t a bike.” But you can say “doesn’t have” or “hasn’t got.” So “be” is the one that just doesn’t .. it’s a main verb, but it just doesn’t behave like a main verb. It doesn’t occur in the progressive …

“He’s being a jerk.” “He’s being an asshole.”

No, it isn’t an irregularity in the progressive; there’s another irregularity in the progressive … I can’t remember … there’s another one that involves the progressive … There’s all kinds of oddities surrounding “be.” But it’s just the only main verb … if you analyze it as a main verb, and give it the same structure as “look” (“she is frightened,” “she looks frightened”), then you should be able to say “she doesn’t look frightened,” “she doesn’t seem frightened,” “she doesn’t feel frightened,” “she doesn’t be frightened.” It’s a regularization that children invariably apply to language. And it’s the source of linguistic change. That’s why irregular verbs become regular, or regular verbs become irregular, by process of analogy. An analogy is always at work in language, too, trying to smooth out these paradigms, even as it creates other problems elsewhere. It’s always a child’s …

… yeah, that’s the idea of the poem … Have you caught George W. saying “nuclear” rather than “nucular” when he’s not talking about weapons or power plants (e.g. “nuclear family”) [something you suspected he might do when not posturing for Bubba cred]?

No, I haven’t caught him using that.

Do you use speech recognition software?


What do you think of Steven Pinker’s books for the general public?

I like Steven. I like the book. I think he’s a little simplistic when it comes to dealing with prescriptive grammar. He takes this very strong naturalistic view of linguistics that since language is a natural phenomenon, you’re just meddling if you try to tell people what’s right and what’s wrong. I think that’s overdone. “Nature” is a complicated word, and there are parts of language that are subject to … Language is, among other things, not just the expression of this cognitive apparatus, but also the expression of social beliefs, and so on. Sometimes you can criticize a usage because you think that it betrays a point of view that is not a point of view that you want people to have, or an attitude, and in that case, you’re perfectly within your reason to criticize that usage, so: apart from that, that end of the book discussion …

… you’re talking about a particular book now?

Yeah, The Language Instinct. When he writes about prescriptive grammar, I take exception to some of the things he says. But I think he’s a wonderful, very smart … a very good linguist and he’s a smart popularizer of his stuff. I think you learn a lot of linguistics in an engaging way from those books, so I like them.

Do you feel like a different person when you speak a different language? Do people who know you in more than one language say that you are different or come across differently in different languages?

Oh, yeah. My voice changes. I don’t know about you, but you find that you’re … When I speak Italian, for example, here’s my normal voice timbre, and when I speak Italian, it’s more this kind of voice timbre, I pull my … there’s more glottal noise, “cuimbi cuando parlo italiano che cuesta vocce que esto exagerando adesso, pero che una vocce che non coresponde a la vocce que uso normalmente cuando parlo ingles” and so you find your voice changing, you’re using gestures, your body changes, in some sense, and you have a different sensibility, in some sense. I like to say that my ex-wife was a lot nicer when she spoke a language that contained the word “nice” than when she didn’t. I’m sure that you’ve seen this too, with your life and your families and so on. So do you speak Spanish and English [to one another] or just only English really?

Who, Merav [my wife] and I? No, we speak almost exclusively Spanish.

Oh, really? Okay.

With my folks, I speak almost exclusively Spanish. With my siblings, however, out of the hearing of my parents, I speak almost exclusively English.

So yeah, sure, you change. There’s a certain sensibility that goes with speaking a language. French tends to be a little more mesquin, a little nastier. But that may be one’s own personal experience. You also associate a language with the context in which you learn it, and the sensibility with which you learn it. Hearth languages, languages you learn from your parents … people who speak, for instance, one language at home and another at work, which is very often the case in America, will just naturally assume different personalities with connection to the languages because there’s one personality for home and one personality for work. But I don’t think their personality is inherent in the language; it’s more just a question of the context you associate it with.

Merav always says that I’m very different in different languages and yet…

Well, you’re your mother’s son in Spanish and your father’s son in English.

Except that my father also spoke Spanish to me.

Comun que.

Let’s go back to this question: What question should we have asked you, that we didn’t? Please answer it.

That’s the kind of question you hate … Nah, I’ll let you know at the meeting. Ah, “have you ever done translation?”

Okay, have you ever done any translation?

Rarely. Occasionally, I’ve translated things like some art catalogues and essays for artist friends.

Which direction?

Into English. I’ve done interpreting from Italian and French to English on occasion, which is just crazy-making. I can’t imagine how people do that. It’s just harrowing! In a public setting, anyway. I’ve done some translations of catalogues and things like that. Because the words have a valence, in the original context. It’s the same with everything, but with art, in particular these words that have a certain valence, historical valence, that’s very hard to render … It’s an enjoyable, if maddening task. I think people should have to do it to … Even when I have to occasionally translate a passage of French, [unintelligible] it’s very hard to do. You realize that you’ve read it and understood it, but then when you come back to translate it … There’s this word “or,” it’s this word that you never translate.

You do a zero translation?

Well, it’s just … “tous les hommes sont mortels, or Socrate est homme, donc Socrate est mortel” [all men are mortal, Socrates is a man, therefore Socrates is mortal].

Okay, is it “and” in that context?

No, it’s like a new thing is being introduced, a new topic is being introduced, a new premise is being introduced … I don’t know.


No, it’s not “additionally”. What would you say? You would just say, “all men are mortal; Socrates is a man.” There’s no particle you would use there. Things like in German, like “ja” and “doch” and so on. “Ich bin ja fertig.”

“I am finished with it”! Is it emphatic?

Noooo, it’s like … There are just these things that you don’t quite know how to render … It’s little things. A friend of mine always says you’re unsure if you’re going from English to Spanish, [the thing that you don’t know how to say is] “by all means.” Now, you can say it “certainly,” or “of course,” and so on, in that sense.

Desde luego.”

“May I take another cup of coffee?”

¡Cómo no!”

Yeah, “how not?” or “certainly!”… “By all means” combines the politeness with the… whatever.

With the permission.

“If you’re in Barcelona, by all means, give me a call.”

No dudes [don’t hesitate].

It’s not that you can’t in every case render the propositional content, it’s just that “by all means” has a common sense that’s very hard to …

You’d use different phrases to translate it depending on the context.

It’s not as if you can’t translate it, you can’t find a way to translate it, it’s just that it has a value that comes of the … what’s another one in … Italian? “Pure” … you look it up, it’s one of these words that when you look up in the dictionary [unintelligible] “yet,” “still,” “however,” “also,” “in addition,” and it’s all of those and none of them. “Ansi” is another one. “On the contrary,” “rather.” “Was it difficult to do?” “Ansi.” “No, it was easy.” “Bring me 3 books, ansi, 4.” I can’t think of the range of uses; you can always find a way to translate it, but it’s ansi! Like “au contraire,” but not quite.

It’s a little bit like Hebrew’s davka, which I’ve also heard is …

Anyway, every language has these things that are untranslatable.