The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 198
Biographical Memoirs: Volume 87

OCR for page 198
Biographical Memoirs: Volume 87 ZELLIG SABBATAI HARRIS October 27, 1909–May 22, 1992 BY W. C. WATT ZELLIG S. HARRIS DIED ON May 22, 1992, midway through his eighty-second year. (The delay in memorializing him in these pages is owing to happenstance.) He was one of the half-dozen linguists, since the beginning of the serious study of language a little after 1800, whom anyone conversant with the field would label a genius. He was the first (in 1947) to adumbrate the notion that linguistics could accept the responsibility of synthesizing or “generating” the sentences of a given language (say, English), as in an algorithm or computer program, from some explicit set of rules—and in so doing he exercised a deep and abiding influence on his best-known student, Noam Chomsky; on his many other students; and on all future researchers who yearn to understand language, surely our most distinctively human attribute. Indeed, it is impossible to imagine present-day linguistics, in either its aims or accomplishments, without taking his pioneering work into account, even though the field, as is ideally true of any science, in which progress is attained by later generations’ standing on the shoulders of earlier giants (as often as not after first stepping on their toes), has in part moved beyond his particular vision. Harris spent his entire scholarly life, until his retirement in 1979, at the University of Pennsylvania. He earned all of

OCR for page 198
Biographical Memoirs: Volume 87 his degrees at that institution (doctorate in 1934), and after joining its faculty became in 1946 the founding chair of what became two years later its Linguistics Department, reputedly the first in the world to be so named. He taught many, myself included, to probe deep and to respect the data; he was a merry but exacting taskmaster; he was venerated by all who knew him, surely, and by many was held in warm affection. He was quick; he was wise; he held scholarship to be a calling worthy of one’s best efforts and one (as will be seen in spades below) from which the personalities of its practitioners are best held apart. Oddly, perhaps, given his expressed wish to suppress personality in science, his own individual character was strongly expressed and strongly felt. Around such a person, inevitably, legends abound. One of them concerns his reclusiveness. Few of his students had ready access to him, and I was once importuned by one of them, after he’d spent a full year at Penn, at least to point Harris out (I was able to direct his attention to the receding taillights of his aging gray Mercedes as it vanished up Walnut Street); and in my day (1959-1963) he had appointed the formidable Miss Sparagna to serve, outside his office, as a sort of Cerberus. This she did with great relish. In fact, as time went on her blinds were often drawn and the lights turned off, lending further weight, there in the gloaming, to Harris’s inapproachability. Another legend concerns his lecturing habits: Some minutes before the time allotted to close a lecture he would sometimes pause, say “And that’s all,” and leave the room; and on occasion—still another legend—he would, on the first fall meeting of one of his courses, ask which of us were also registered for his other two and then, having discovered that we all were, announce that they would therefore be featly combined into one. Which was fine, since all of his courses, however titled, covered a vast domain. (In my Penn

OCR for page 198
Biographical Memoirs: Volume 87 graduate catalog for 1967, the year he awarded me my degree, Harris is listed as being sworn to teach “Formal Linguistics” and “Mathematical Systems in Linguistic Structure” in the fall semester and “Seminar in Linguistic Transformations” in the spring. His courses tended to merge one into the other; and the first and third of those just listed are specified in the catalog as “may be repeated for credit.” Which they were, and justly, since their contents overlapped and varied with Harris’s latest advances.) Harris was born in North Ossetia, now a constituent republic of the Russian Federation, but was taken by his family to Philadelphia when he was but four years old. (His middle name, “Sabbatai,” set beside his brother’s first name, “Tzvee,” appears to identify the family as followers of Sabbatai Tzvee or Tsvee (1626-1676), the “False Messiah of the Caucasus.”) To my ear he had virtually no foreign accent, sounding just like any native Philadelphian (meaning that he spoke one of the half-dozen or so equally distinctive Philadelphia dialects), except that his “filled pause,” as linguists term it, rather than the usual “uh,” was something like “eh” (linguistically, a simple long /ε:/ with a bit of nasalization and a hint of an “h” at the end). Unlike Chomsky he was no sailor, his physical activity being mostly confined to his working on a kibbutz in Israel many summers (his wife, Bruria Kaufman, was a professor at the Weizmann Institute there), in which purviews he was apparently known simply as “Carpenter Harris.” Prompting one to picture this great scholar, elegantly balding, slightly stooped and with thickish rimmed spectacles, astride a beam into which he was driving, with a framing hammer, a 10-penny nail. He was, as I understand it, a secular and indeed Socialist Zionist, committed to the independence of Israel (as who is not?) but less than pleased with that nation’s swerve toward theocracy.

OCR for page 198
Biographical Memoirs: Volume 87 Harris’s scholarly career seems to fall into several successive phases (though he might well have denied this, since they overlapped). During his early years (in the 1930s) he devoted himself to Semitics, having been a very early analyst of the then-new Ugaritic materials; at this point he was looked upon as a quite promising Semitist. Sometime around World War II he applied himself to more general problems in linguistics, the culmination of which was the completion in 1947, with long-delayed publication in 1951, of his magisterial Methods in Structural Linguistics (later reprinted in paperback as just Structural Linguistics), which became the standard text for the next decade and more, and which cognoscenti still regard as a classic. A little later he devoted himself to two other areas of research: computational linguistics, which was just becoming possible owing to the ready availability of computers (which had after all been invented at Penn, as ENIAC); and, above all, transformational analysis, which he had begun working on earlier in his career, an approach in which simpler sentences (“The archer shot the arrow”) can be “transformed” by general rules into more complex ones (“The arrow was shot by the archer”), and vice versa. His activity in the first of these two areas (he spearheaded development of the first truly functional computational syntactic analyzer [on a UNIVAC]) presumably arose quite naturally from his lifelong interest in analytic techniques. His work in the second eventuated from his interest in analyzing texts into simpler ones bearing the same information (a concern that never left him). And then toward the end of his life he developed a method of linguistic analysis that viewed sentences as being generated from a formally simple application of functors to their arguments (he used somewhat different terms). Such an analysis views all syntactic relationships in the same light, therefore sub-

OCR for page 198
Biographical Memoirs: Volume 87 suming the relationships traditionally termed “modification” (as when an adjective modifies a noun) and “predication” (in which a verb is predicated of its subject) or more broadly, “agreement” (as when “this” is pluralized to agree with “books” in “these books”) and “government” (as when “me” rather than “I” is mandated when the object of “kissed” in “Jenny kissed ___”: no “Jenny kissed I”). Then the adjective “tall,” for instance, as in the sentence “Tall men excel,” can be described as the sort of functor that turns a noun into another noun—as an FNN—while “excel” can be described as the sort of functor that turns a noun (“tall men”) into a sentence, that is, as an FNS. In “very tall men,” “very” would then be the sort of functor that turns an adjective—already an FNN—into another adjective (an “adjectival phrase”), and so it can be tagged as an FFNN(FNN). (The parentheses aren’t formally necessary, but they add perspicuity.) In this fashion then, from just the two primitives “N” and “S,” a fully developed functor/argument analysis can aim at providing an intriguing reformulation of all the traditional parts of speech and of the phrases, clauses, and sentences they occur in. Such a formal analysis is rather rigid, and perhaps overly limited, but it does suggest a different way of viewing sentences. Suppose, for example, that one should take up the examples of “He ran up a bill” and “He ran up a hill.” These are of course very different: One can ask (transformationally) “Up what hill did he run?” but hardly “Up what bill did he run?” Then “up” in the first sentence can be represented as the sort of functor that makes a verb (“ran”) into another verb (“ran up”)—as an FFNS(FNS)—and “up” in the second sentence as the sort of functor that makes a noun (“a hill”) into a sort of adverbial, hence perhaps as an FNF(FNS[FNS]). The resulting characterization may seem a little forced—and it is—but it does present an interesting new way of viewing syntax. In

OCR for page 198
Biographical Memoirs: Volume 87 short, by presenting a rigid structure derived ultimately from combinatory logic, Harris yet again showed how one kind of analysis might illuminate others. Harris is sometimes thought of as having tried to furnish one or more “discovery procedures” that would permit the intending analyst to apply a sort of litmus test to whatever language he or she was studying, without further insight, whereupon the correct account of that language would emerge as if by magic. Not true. He always admitted that any initial linguistic analysis would depend on what are nowadays called the linguist’s “intuitions”; what he aimed to provide were checks on such analyses, what could be called “confirmation procedures.” All of his analytic methods were forthrightly stated to be aids to analysis but not infallible ones. This judgment holds for some of his early methodological insights, such as the one that showed how to extract, from a sequence of sound segments, such as from the consonant cluster /st/, a “phonemic long component” that those segments have in common. In the case of /st/, to continue our example, a “long component” that /s/ and /t/ have in common is “voicelessness” (the larynx isn’t vibrated when making either sound). Now suppose that “voicelessness” is represented by underlining any sequence thus characterized. Then we could write /st/ as /st/. This would be pointless, though because /st/ consists precisely of /s/ + /t/, both of which are defined as being voiceless. However, something has still been gained, since now we can write /st/ as its voiced counterpart, /zd/, if the latter is embellished by an underline—as /zd/—signifying that /st/ is /zd/’s voiceless counterpart. Seems simple, but such a move will strike the linguist as a gain, since it will have thus explicitly acknowledged the close relationship between /st/ and /zd/—they differ mainly in that /st/ is a “voiceless” variant of /zd/—and more im-

OCR for page 198
Biographical Memoirs: Volume 87 portantly it will have made way for a generalization to the effect that “/zd/ is always voiceless when word-initial.” This is easily captured by a rule, identifying the initiation of a new word as “#”, on the order of “#zd→#zd.” Such an expression is especially nice because so general a rule will automatically forbid any word (in English) from beginning with /zd/ itself, since any such initial cluster will be converted to /st/, thus capturing the fact that “stin” is a possible English word but “zdin” is not. (We ignore imports like Italiian “sdrucciolo” [Italian initial “sd” is pronounced as /zd/], a rhetorical term referring to poetic lines stressed and rhyming on the antepenult.) In this manner we could encapsulate simple facts about the English language, and facts well worth the capturing if we are to understand any language at all. (In this instance, the fact that we accept initial “st” in English but reject the initial “zd” that after all, since Italian-speakers have no trouble with it, is easily pronounced by the human mouth.) I note in passing that the posited paired relationship between /st/ and /zd/ is not discoverable “automatically,” since /st/ could also be paired with some other voiced cluster equally unable to occur initially in English, /dl/ for example (no “dlin”). But /st/ and /dl/ are related only at a rather more abstract level than /st/ and /zd/ are, a fact revealed by the sort of phonological analysis that can scarcely be thought of as “automatic.” For Harris, at least in the 1950s, such advances were made in the interest of achieving the simplest possible analytic account of the language. Later and presently, under the assumption that speakers’ brains attain to a maximum of static simplicity when representing the “head-grammar” that permits them to produce language, such advances are sometimes thought of as approaches to the cognitive grammar itself. For Harris this assumption, however inviting, could

OCR for page 198
Biographical Memoirs: Volume 87 not yet be firmly grounded in the psychological (much less neural) sciences. So Harris’s own prejudice remained always in the interests of analysis, despite the fact that a computational grammar, such as the version he pioneered in 1959, can be put to the proof only by generating sentences from it, and despite his having said as early as 1954 that a deep analytic grammar could be viewed as “a set of instructions which generates the sentence of a language.” Information of the sort conveyed by “phonemic long components” is nowadays couched rather differently, even though the information (if not its implications) remains much the same. For example, the “long components” just discussed would now be captured in the form of a rule for English declaring in essence that any sound of the set {s,z} must be voiceless when word-initial and preceding any sound of the set {t,d} (so the set {s,z} can only be realized as member /s/ in this position), plus a rule forcing any sound of the set {t,d} to be voiceless—hence, to be realized only as /t/—when between initial /s/ and anything else. Entailing, just as in Harris’s formulation, that “stin” and “strin” are possible English words while “zdin” and “zdrin” aren’t. (The characterization of the set {s,z} and of the set {t,d} would in later treatments be conveyed by stating each set’s members in terms of roughly the “distinctive features” they have in common.) In all, Harris’s notion of “phonemic long components” was an early and persuasive presentation of the idea that entities like /s/ and /z/ could (and should) be factored into “components”—now mostly called “distinctive features”—incorporable into general rules revealing of linguistic structure; and this notion, which dates from the completion of the MS of Methods in Structural Linguistics in 1947, was to prove quite influential elsewhere. It had a significant rebirth within anthropology, for instance, in the analysis of kin-terms due to Romney and D’Andrade. It

OCR for page 198
Biographical Memoirs: Volume 87 remains to be determined whether the notion of “phonemic long component,” in which for instance initial /st/ is recognized as a unitary consonant-cluster having “voicelessness” extending over the whole, is more or less “cognitively real” than the notion of segmental-phoneme rules like those that convert first initial {s,z} into /s/ before {t,d} and the {t,d} into /t/ after initial /s/. It might well have surprised Harris to learn that his analytic devices might turn out to be superior in cognitive reality, should they do so; but science, of course, consists in large part of surprises. As noted just above, some “analytic” approaches to language are, at least in implication, also “synthetic,” in that by comprehensively revealing the inner attributes of language one may show how to synthesize or “generate” sentences adhering to that analysis. Certainly Harris’s “transformational analysis” is of this sort. Once active sentences containing transitive verbs have been shown to be systematically related to passive sentences bearing the same information, as “The boy broke the toy” is related to “The toy was broken by the boy” (e.g., by a simple formula [omitting tense and ignoring many problems and complexities] on the order of “N1 V N2 ↔ N2 be Ven by N1”) then one has indicated how passive sentences might be generated from their active counterparts. Still, some of Harris’s analytic techniques are much less easily thus characterized. For instance, he once proposed a technique for determining morpheme-boundaries by examining the phoneme-sequences composing them (a “morpheme” is roughly a minimum stretch of meaningful language: the word “meaningful” consists of the three morphemes “mean,” “-ing,” and “-ful”). Any such technique could be “generative” only in the sense that an auditor might use it, unconsciously perhaps, to aid in the “re-generation” of a sentence he has heard; surely no speakers use it to generate their own sentences, since of course for them their

OCR for page 198
Biographical Memoirs: Volume 87 morpheme boundaries are determined by what they want to say. As to the question sometimes asked by observers outside of linguistics why “information-preserving” alterations of sentences are by almost any analyst assigned a privileged position, the answer is that while indeed speakers can produce for any sentence of the form “The cat sat on the mat” another sentence of the form “The mat sat on the cat,” with no preservation of the original’s information at all, such an ability is universally consigned to the periphery, on the thesis that our defining use of language is to convey information, not to play with it. Naturally there’s still room for disputation regarding what is the nature of the “information” that language seems designed to convey: Is it really only “truth-value,” or is more involved (e.g., point of view, point of emphasis, point of interest, or news to the auditor)? Under this rubric does “It was the cat that sat on the mat”—which stresses that it was the cat rather than some other creature, perhaps in denying an auditor’s contrary assertion—really only “preserve the information” of its possibly underlying transform “The cat sat on the mat”? In this case, surely not. But such questions are still, as in Harris’s own time, being debated in the scholarly literature, albeit with ever-increasing sophistication. Harris’s continuing main concern with matters of linguistic analysis of productions gleaned from actual speakers and writers could be characterized in modern terms as a concern with “performance” over “competence” (i.e., with physical evidence of what speakers and writers do over hypotheses [or mere conjectures] of what they must have inside their crania to enable them to do it). (He may have maintained some mistrust of factors lying beyond the analyst’s access as available to 1930s informant techniques: “Can you say this? How about this?”) Are transformations revealing

OCR for page 198
Biographical Memoirs: Volume 87 of speakers’ understanding and generating of speech? Are functors on arguments revealing of different aspects of that speech? Is a treatment of the transitional probabilities obtaining among the phonemes of a language’s sentences (roughly their distinctive individual sounds) revealing of basic elements such as boundaries, such as might be used by auditors to understand speech? Then let all these methods be brought into play, that each may disclose a different aspect of language in all its performative complexity with all, perhaps, in the aggregate, revealing language as a whole. This is a fascinating view of the relation between language and its analysts, and a challenging one; and it appears to have been a view that for a while, at least, was almost Harris’s alone. No longer: For nowadays some linguists hold that at least at first we parse simple sentences in a Harrisian (surfacebased) manner, and then if those sentences don’t compute to sense once that has happened, we apply a deeper parse, in a Chomskian manner, say, to understand them. Like all of my fellow students, I think, I revered Zellig Harris as mentor and as resident genius; like more than a few, I had a warm affection for him, in my own case as a sort of intellectual father. This affection was only increased by my personal interactions with him, not just in his office, once regularly admitted, but also, more casually, on the streets of Philadelphia. An accident of residence—his apartment and mine were only a block or so apart—led me often to find myself afoot behind him crossing the Walnut Street Viaduct to the Penn Campus, he not infrequently in his greenish outdoorsman’s jacket, with wooden toggles in the stead of buttons and, armed against Philadelphia’s blustery weather, a prominent hood. As he marched along across the Schuylkill he would sometimes reach into the side pockets of this capacious garment and fish out various pieces of paper on which, presumably, he had written notes to him-

OCR for page 198
Biographical Memoirs: Volume 87 self on this or that linguistic point. He’d examine these without breaking pace and then, about half the time, toss them into the river. They fluttered down like the inscribed leaves that, in legend at least, some Chinese poets, uncaring of posterity, used to toss into the nearest creek. Sometimes I wondered how many potential dissertations, by us his epigones, floated down those then-noisome waters, to be swept eventually into the Delaware and then out to the Atlantic. Some of these notes, though, have survived, for I still have a few in my possession, since he at one point delegated me to be his inquirer into the tangled web of English adverbs and so passed on to me his jottings thereanent. They make interesting reading. First, they contain brilliant aperçus—if not yet analyses as such—and secondly they consist in large part of stray scraps of paper saved from the Schuylkill. They comprise as follows: (1) a note on formal University of Pennsylvania letterhead addressed to “Dear Watt”; (2) six notes on 8 1/2 × 11 brown-flecked blue-lined notepaper; (3) two notes on 5 1/2 × 7 notepaper; (4) one note on a different 5 × 8 notepaper; (5) twenty-nine notes on 3 × 5 sheets torn from some tablet; (6) twenty-three 4 × 6 sheets excised from some other tablet; and lastly (7) two notes on the reverse (flap side) of two University of Pennsylvania envelopes, one small and one letter-size. They constitute a set of casual records of a superbly talented linguist’s cogitations on language—probably, given the ordinary evanescence of such things in the destructive course of time, among the best we’ll ever have. As to their contents, they were, as just noted, mere notes, except for the letter addressed to me. “How many differences,” he asks me, “can you get between time-point adverbs (yesterday, at 10 a.m.) and time-aspect (recently, frequently, generally)?” And then on one of the lesser sheets he queries

OCR for page 198
Biographical Memoirs: Volume 87 the possible difference between “He only slept an hour” and “He slept only an hour.” On another, he wonders about “She is often tired” in relation to both “However often she is tired” and “However she is often tired.” On still a third, about the possible (to me, dubious) relation between “Their names were linked romantically” and “Their arms were linked in a romantic way.” On a fourth, about the best comparative analysis of similar adverbial locutions such as those contained in “The deer was killed with a blowgun,” “… near the brook,” “… by moonlight,” and “… at nine.” And so on. He was, in other words, searching through his interior sense of the English language, unrelentingly and unflinchingly, for thorny problems demanding respectful and hopefully explanatory solutions. A model, surely, for any intending language analyst. (And, by no coincidence, the exploratory model to which all contemporary linguists adhere, be their purview widened to take in many other languages or even such dialects as that found on one wharf of some obscure Sardinian village where a distinctive version of Catalan is spoken.) We come now to a still more revealing incident in Harris’s life, and for that matter my own, that has not hitherto been disclosed to the public eye. In 1969, having become aware that on October 12 of that year (Julian calendar) Harris would celebrate his sixtieth birthday, I conceived the notion that the occasion mustn’t pass unremarked, and gained the assurance of Mouton & Company, in the Netherlands, that that concern would publish a Festschrift should I be able to garner the requisite number and quality of participants. Accordingly, I wrote some of the prominent linguists of the day, therefore including a good few of Harris’s onetime students, asking if they’d be interested in contributing. The response was overwhelming, and the Festschrift, to be entitled with maximum simplicity “To Honor Zellig Harris at 60,”

OCR for page 198
Biographical Memoirs: Volume 87 was thereby set in motion. The 31 who agreed to submit tributary articles ranged widely over the fields to which Harris had made major contributions, but they were naturally concentrated in linguistics and its computational applications. Listed alphabetically, they comprise: Yehoshua Bar-Hillel, Dwight Bolinger, William Bright, I. D. J. Bross, A. F. Brown, Paul G. Chapin, Noam Chomsky, Charles A. Ferguson, Bruce Fraser, Lila Gleitman, Henry Hiż, Carleton Hodge, Henry M. Hoenigswald, Fred W. Householder, Dell Hymes, Ray Jackendoff, Aravind K. Joshi, Sheldon Klein, Susumu Kuno, George Lakoff, Robin Lakoff, Leigh Lisker, Yakov Malkiel, Christine A. Montgomery, David Perlmutter, John Robert (“Haj”) Ross, Naomi Sager, Arthur Schwartz, Carlota S. Smith, Zeno Vendler, and C. F. Voegelin, plus myself. (A few of these acceptances were tentative, and there were others whom I solicited but who pled a supervening and perhaps subsequent commitment.) The promised participants in the projected volume included, then, a representative selection of his onetime students (Noam Chomsky chief among them), plus a few others, among them the most respected names in the scholarly world of the study of language in its various aspects. Readers need not cudgel their wits for memory of this volume, for it never appeared. Harris aborted it. He learned of the planned Festschrift, just in advance of his returning to the States and there receiving my letter apprising him of it (these things are supposed to be a surprise, after all), while passing through the Netherlands offices of Mouton & Company. His refusal of the intended honor was at first acerbic. “Dear Watt,” he wrote me on October 20, 1969, in his tiny longhand, I am sorry to intervene in your actions, but I am writing in a matter in which I have human rights. It has come to my attention that you and

OCR for page 198
Biographical Memoirs: Volume 87 Mouton are planning a Festschrift for me. Such a publication would be a deep personal affront to me and to my sense of values. I have managed to live this long with the principle that scientists can be people who do the best work they can for the sake of knowledge and of its human value. Any special—and unavoidably invidious—recognition of their work, such as honors, prizes, and Festschriften, is abhorrent to me, and would violate what I feel is a human right and dignity. Therefore, I ask you to withdraw this activity…. Many years ago, during Bloomfield’s lifetime, I had to get a similar project stopped for Bloomfield’s sake, and I am sorry that now I have to do it for myself. I am sure, however, that you will understand me, and will respect my principles even if they may seem excessive. With best regards, Zellig Harris P.S. I have just seen your letter [a greeting to him announcing the occasion], after writing the above. Thanks for writing me, & I will answer your letter tonight, although the above (for which I apologize again) will indicate how I feel in the matter. Yours, Z. S. H. Here, beyond cavil, was a response to a prospective honor—and one granted to few—from an honorable man. Moreover, one couched in such a way as to cause me, the offender, the least pain, partly by basing his declining the proposed honor on his having scuttled a similar tribute to Leonard Bloomfield, one of the earlier gods of linguistics. My reaction, besides of course immediately resolving to cancel the projected Festschrift and to write its promised participants to that effect, was also to arrive at a new respect for the opinions that Harris had just evinced and to conclude that, in his sense, Festschriften are indeed an abomination of a sort. Having canceled the Festschrift, I so informed Harris. Before he could receive my notification of withdrawal he wrote me again, as promised, in a way still more indicative of what he held to be “human values” and also of his sensi-

OCR for page 198
Biographical Memoirs: Volume 87 tivities to a very junior colleague (whom, after all, he might rightly have suspected of an activity not wholly divorced from self-aggrandizement). His second letter, also dated October 20: “Dear Watt,” he wrote (I should explain that, just as in California nobody has a surname, in Philadelphia, except in the case of strangers or when extreme deference is due, a man is typically not addressed by any other), “Thank you for your kind letter, and I would never have been able to write as I did yesterday had I seen [it] first—though it may be just as well for my earlier letter represents my feelings…. Small as the whole issue is, I think you too see that there are values involved. As for me, anything that I could have gotten from the Festschrift, I think I have gotten from the tone of your letter, for which I thank you.” And then after Harris had received my notice that the abominable Festschrift had indeed been aborted there followed still a third letter, which he concluded by saying, more broadly and more personally, that “anyway, it is good sometimes to air one’s feelings about the culture we live in (I don’t mean only ours, or only now—the others are even worse)…” A final note on the aborted Festschrift may serve to deflate a certain rumor. Some believe that Noam Chomsky, Harris’s best student, had a violent falling-away from his mentor and that there was “bad blood” between them. There’s good reason to doubt this. Chomsky’s letter to me, accepting my invitation to contribute to the Festschrift, betrayed not the slightest hint of such a rift. He wrote that he’d be “very pleased” to contribute and offered moreover to go over my list of proposed contributors to “see if any other suggestions come to mind.” This was after all in 1969, well after the publication of Syntactic Structures and after he had completed, under Harris’s direction, his iconoclastic dissertation at Penn. I may mention that during my own tenure at Penn, Chomsky was a sometime visitor at Harris’s

OCR for page 198
Biographical Memoirs: Volume 87 seminars, and certainly no animus was evident, to me, on either side. In fact, I don’t think either Harris or Chomsky (or for that matter anyone else who had to any degree absorbed Harris’s gentle nature) would be capable of the kind of pettiness it would have taken to sour their relations to the degree some have postulated. I may also mention that in Harris’s introduction to his monumental Methods in Structural Linguistics, dated 1947, he gave due credit for “much-needed assistance with the manuscript” to one “N. Chomsky.” “N. Chomsky” was then 19. There is after all a human quality fitly called “nobility of character.” It isn’t found only at the great universities—I came upon it one desert morning in a radiator-repair shop in Kingman, Arizona—but when encountered in a university setting, it’s likely to affect a great many lives, and for the better. Zellig Harris, besides being a towering figure in linguistics—and one many of whose insights and discoveries will be of perennial relevance—had, to my eye, that quality and in spades. It was manifested in many ways, and not least in his modesty: “I’d give it all up if I could write one good sonata,” he once confided to his onetime student A. F. Brown, of his scholarly achievements and their resulting renown, as “Pete” Brown told me one day. “Leges sine moribus vanae” has long been Penn’s motto: roughly, “Laws without morality are useless.” Though from Horace (xxiv in Book III of his Odes—there really was a time when a university’s motto was likely to be drawn from Horace or Virgil) it well conveys Harris’s own voice, in his time there, and as it continues in many of us who had the privilege of studying under him. In those, as in many others who knew him, surely, his spirit still lives. At Penn, as I’ve noted, he was a rather elusive figure, and once when he remarked in class that any of us with questions were welcome to seek him out of his office, a student (Tolly Holt) called out, “But you’re

OCR for page 198
Biographical Memoirs: Volume 87 always absent!” Without a moment’s hesitation Harris replied, “That’s false. I’m always present somewhere.” And now we know where.

OCR for page 198
Biographical Memoirs: Volume 87 SELECTED BIBLIOGRAPHY The following is an abridgment of E. F. K. Koerner’s compilation published in The Legacy of Zellig Harris, vol. 1, edited by Bruce E. Nevin. Amsterdam: John Benjamins, 2002. 1933 Acrophony and vowellessness in the creation of the alphabet. J. Am. Orient. Soc. 53:387. 1935 A Hurrian affricate or sibilant in Ras Shamra. J. Am. Orient. Soc. 55:95-100. 1936 A Grammar of the Phoenician Language. New Haven, Conn.: American Oriental Society. 1939 Development of the Canaanite Dialects: An Investigation in Linguistic Theory. New Haven, Conn.: American Oriental Society. 1942 Morpheme alternants in linguistic analysis. Language 18:169-180. 1944 Simultaneous components in phonology. Language 20:181-205. 1945 Discontinuous morphemes. Language 21:121-127. 1946 From morpheme to utterance. Language 22:161-183. 1951 Methods in Structural Linguistics. Chicago: University of Chicago Press. 1952 Discourse analysis. Language 28:1-30.

OCR for page 198
Biographical Memoirs: Volume 87 1954 Transfer grammar. Int. J. Am. Linguist. 20:259-270. 1955 From phoneme to morpheme. Language 31:190-222. 1957 Co-occurrence and transformation in linguistic structure. Language 33:283-340. 1964 Transformations in linguistic structure. Proc. Am. Philos. Soc. 108:418-422. 1965 Transformational theory. Language 41:363-401. 1968 Mathematical Structures of Language. New York: John Wiley. 1970 Papers in Structural and Transformational Linguistics. (Ed., H. Hiż.) Dordrecht: D. Reidel. 1976 A theory of language structure. Am. Philos. Q. 13:237-255. 1978 Grammar on mathematical principles. J. Linguist. 14:1-20. 1982 A Grammar of English on Mathematical Principles. New York: John Wiley. 1988 Language and Information. New York: Columbia University Press. 1989 With M. Gottfried, T. Ryckman, P. Mattick Jr., A. Daladier, T. N. Harris, and S. Harris. The Form of Information in Science: Analysis of an Immunology Sublanguage. Dordrecht: Kluwer Academic.

OCR for page 198
Biographical Memoirs: Volume 87 1991 A Theory of Language and Information: A Mathematical Approach. Oxford: Clarendon Press. 1997 The Transformation of Capitalist Society. Baltimore: Rowman and Littlefield.