The Linguistics Wars

1.We will come back to Chomsky’s version later in the book, in Chapter 7. I should mention here, by the way, in the first of a great many endnotes, that the running commentary of elaboration, qualification, and citation, along with occasional expressions of personal whim, in these notes is parallel, rather than integral, to the story I tell. Feel free to join me here, but feel free to ignore me here as well, as suits your reading inclinations.

2.This is a précis of the version that he tells to John Goldsmith (Huck & Goldsmith 1995:107–19) which is in fairly close accord with the version he shared with me, on the phone, in person and in correspondence, paper and digital; see also his interviews with Francisco José Ruiz de Mendoza Ibañez (G. Lakoff 1997) and John Brockman (G. Lakoff 1999), as well as G. Lakoff 1989, for fuller accounts. In all of these accounts, Lakoff foregrounds a quasi-Lakatosian attitude toward scientific development, spelling out various value-commitments that we will see again.

3.Stabrius Eros is discussed in §XIII, Lucius Orbilius Pupillus in §IX, Lucius Crassicus in §XVIII.

Chapter 2

1.There is a small chance, given both who wrote this and what his ultimate verdict on Chomsky is in the book, that there is an element of satire here. But so many things are so howlingly inaccurate in Wolfe’s book, especially about Chomsky’s chronology, that this passage looks to be simple wild-eyed wrongness. Wolfe has Chomsky lecturing at Chicago and Yale in 1955, for instance, about the language organ, Deep Structure, Universal Grammar, the language acquisition device, and so on, all in terms that didn’t show up in Chomsky’s work until the 1960s (Wolfe 2016:87–88).

2.For a corrective to the allusive caricatures of the Bloomfieldians in this book—which I provide as a guide to the rhetorical climate of Chomsky’s rise, not as a realistic account of these linguists—see, for instance, Hymes and Fought 1981 [1974], Seuren 1998:178–219, Blevins 2013, Goldsmith & Laks 2019.

3.See R. A. Harris 2010 for an extended charting of how the Chomskyan revolution was enmeshed with psychology, computer science, and the general cognitive revolution. The first edition of this book has a chapter on the revolution that stands up moderately well, but when I wrote it there was relatively little scholarship on Chomsky’s rise. By 2010, that scholarship had grown and my treatment there is more nuanced and balanced.

4.See R. A. Harris 2010:247–49 for a mapping of Chomsky’s formalism against that of Alan Newell and Herbert Simon’s famous Logic Theory Machine, which was unveiled at the epochal 1956 Cambridge colloquium where Chomsky gave his equally famous “Three Models” argument and where Miller completed the foundational cognitive science triple-play with his “Magical Number Seven, Plus or Minus Two.”

5.For other typical applications or explications of Chomskyan linguistics for English studies, see Lees 1963 [1962], Hathaway 1962 & 1967, Ohmann 1964 & 1966, Levin 1963, 1965, & 1967, Thorne 1965, Rogovin 1965, Eschliman, Jones, & Burkett 1966, Hayes 1966, Hunt 1966, Thomas 1965, Roberts 1967, Lester 1967, Steinmann 1967, Auerbach et al. 1968, Beaver 1968, or anything from the literature on sentence-combining (beginning with Hunt 1967; see Stugrin 1979 for a relatively full treatment). Christensen’s two articles, “A Generative Rhetoric of the Sentence” (1967 [1963]) and “A Generative Rhetoric of the Paragraph” (1967 [1965]), though popular, are peripheral. He claims that the relevant term in his title is “not derived from generative grammar; I used it before I ever heard of Chomsky,” but Chomsky’s association with the term certainly gave Christensen’s work a cachet that helped it win some success. For some indication of the success of these efforts, see McCawley’s review (1976b [1967]:15–34) of Owen Thomas 1965.

6.Chomsky 1955d, Bar-Hillel 1954. Bar-Hillel later said that Chomsky’s defense of Harris had been “well taken” (Bar-Hillel 1963:542); see Tomalin 2006:125–39 for a detailed account of the two papers. Goldsmith & Laks 2007:282, in the context of this exchange, note the “stunning disconnect between Chomsky’s willingness to attack others’ models as hopelessly unrealistic, and therefore of no possible interest” and the defence he offers when his own “models diverge from reality;” namely, that it is just a matter of “sound scientific idealization.”

7.The term may have been forged by Bar-Hillel (1962 [1960]:555). Sampson argues that the credit for the hierarchy is only part Chomsky’s; in particular, that “the sound maths of the ‘Chomsky hierachy,’ ” so much more impressive than “the lamentable maths of The Logical Structure of Linguistic Theory,” were actually provided by Marcel-Paul Schützenberger. (2016:591, also 1979; Sampson is not the only linguist to complain about the quality of mathematics in Chomsky’s work—Pullum 2011).

9.All of these rules show up in Syntactic Structures but I have modified them somewhat here and subsequently, for simplification in later exposition. Some linguists will, I predict, be very unhappy about these modifications. It is fairly easy to poke holes in this system (and, in fact, in Chomsky’s original versions as well), but to the extent necessary, these holes can be patched with additional details.

10.Affix-hopping got this name a few years later, in the more insouciant attitude that soon arose in Transformational Grammar. Syntactic Structures just calls it “the Auxiliary Transformation.”

12.Chomsky cites Bloomfield here (1957a:81n6) to authorize “underlying form,” a notion of critical importance to Chomskyan linguistics, especially in the variety soon known as Deep Structure. Seuren 2009:108 traces the notion of “linguistic transformation” to Bloomfield’s use of underlying form. See Nevin 2009:469–75 for a response.

13.Don’t worry about little specifics, like how we get from the Tense of 1d to the Pres (Present; or, even more specifically, third-person, present, singular) of 13′a and on all the relevant trees. It’s actually standard operating procedure to leave many specifics unspecified, since the relevant mechanisms (a Phrase Structure Rule in this case) follow a familiar template and are easy enough to imagine.

14.John Viertel was writing a dissertation under Chomsky at the time (never completed), and working for him as a translator of Humboldt; the credential line of the article simply identifies him as “a member of the research staff at Massachusetts Institute of Technology” (Viertel 1964:65n1). Philosopher John Searle had an influential article on “Chomsky’s Revolution” (1972) which featured a version of the John-is-eager/easy-to-please argument. There was also a New Yorker article about Chomsky, and a subsequent book, both with the title John is Easy to Please (Mehta 1971a, 1971b).

15.There have been a variety of claims that Chomskyans wanted to keep the theory to themselves, actively shunning publication, alienating would-be supporters, creating an underground distribution circuit for their writings, and so on (see especially Nielsen 2010)—summarized and largely refuted in Newmeyer 2014. There are certainly coat hooks on which to hang such interpretations, like the long delay in publication of Chomsky’s opus, Logical Structure of Linguistic Theory, the standard operating procedure of aggression that characterized much of the MIT contingent, and the very active and shifting flow of quasi-publications (imagine pre-internet blogs, dependent on mimeograph machines, the postal service, and word of mouth). But (complementing Newmeyer’s evidence), Barbara Partee’s recollection puts this claim into perspective: “We [MIT] students were shocked [when we read Bach’s 1964 textbook] that someone not at MIT could even know about transformational grammar—we thought that we who were sitting at the feet of the master were the only ones privileged to receive the Word. When some of us met Emmon for the first time at a meeting, we asked him how he had found out enough about it to write a textbook without being there. He smiled indulgently and replied, ‘Well, I can read’ ” (2005:6).

16.The rest of the remark from Lees, after coat tails, is: “We had to. He is so smart that any idea you came up with he had already thought of, and thought over long and deeply. I wouldn’t be very happy if you published that. But it’s true, so I guess I’d be happy in another way. We all rode on his coat tails to prominence.”

18.The University of Minnesota Center for Cognitive Science link to the list is no longer active, but it can be found via the Wayback Machine: http://www.cogsci.umn.edu/OLD/calendar/past_events/millennium/final.html, and there are various other echoes in blogs and discussion lists. The list in The 100 Most Influential Books Ever Written, assembled by Seymour-Smith (1998), is a bit idiosyncratic, but Syntactic Structures is #82.

19.Program 143 of Firing Line (Season 4, Episode 13), entitled “Vietnam and the Intellectuals,” which aired on April 3, 1969. I acquired a transcript of the interview from the Hoover Institution Archives, but the show is available widely on the internet.

20.There is an almost shocking irony here, much commented on (see, for a multi-perspectival sampling, Newmeyer & Edmonds 1971, Newmeyer 1988:84–88, Harpham 1999:182–83, Schweizer 2006:21–24, Golumbia 2009:31–38, Knight 2016:28–42, 2020). A substantial portion of the funds for building up the MIT linguistics department came from the U.S. military—either in fairly direct ways, such as funding activity in the Research Laboratory of Electronics (RLE), or more indirectly, through various machine translation projects in Cambridge that hired MIT professors and students. The Cold War was particularly important for this funding. It is no coincidence that the Russian language, for instance, was a major focus of theoretical and applied work associated with the RLE, and the military investment in Chomskyan theory, whether at MIT or elsewhere, was expected to produce results for such military applications as encryption, machine translation, information retrieval, and command-and-control systems for jets and weapon delivery. Chris Knight documents the “high hopes [of the Pengagon] for precisely the new and unusual kind of linguistics being championed by Chomsky in the 1950s and 1960s” (2020:664; see also Newmeyer & Edmonds 1971:288–90). Virtually every paper or book coming from MIT faculty or students until the late 1960s, including Syntactic Structures, carried acknowledgments like “This work was supported in part by the U.S. Army (Signal Corps), the Air Force (Office of Scientific Research, Air Research and Development Command), and the Navy (Office of Naval Research)” (Chomsky 1957a:7). That is, the success of the MIT linguistics department (and, in fact, other Transformationally oriented departments, like UCLA’s), from which Chomsky gained much of his authority, comes from the establishment he excoriates. So much the better, many of us think. The funds might easily have been put to quite evil purposes, but Chomsky has latterly often cited them as an indication of his own, almost inevitable, complicity in American international oppression and war crimes. “We’re all compromised,” he said. “Look at me. I work at MIT, which has received millions from the Defense Department” (Branfman 2012). Knight has a theory that guilt over this complicity directly affected Chomsky’s linguistic theorizing, pushing it to levels of abstraction and abstruseness that made his work deliberately unusable by the military (Knight 2016 & 2018a, 2018b, 2018c, 2020). I reject that view (Harris 2018), as does Newmeyer (2018); see Chomsky’s 2019 response.

22.It is unclear, twenty-five years later, whether he has relinquished the spot—I could not find a reliable way to investigate this question—but I see no reason to doubt that he remains pretty high up on any such list .

25.Newmeyer (in 1980a:92 and 1986a:80) makes a similar point, also calling upon biblical polysemy.

26.Chomsky is quite clear from the beginning that kernel sentences are where the semantic action is (for instance, 1957a:92, 107–108), but I am ignoring several factors in my kernel-grows-up-to-become-Deep-Structure story, most notably that “understanding” a sentence in Syntactic Structures also involved the underlying phrase structure of the relevant related sentence, as well as the derivational history of the relevant derived sentence (captured in a representation called the transformation marker). My use of kernel in this section is therefore misleadingly broad for historical aficionados. Anyone who objects can mentally edit in “kernel + underlying phrase marker + transformation marker.” But Deep Structure became important to the theory precisely as the kernel became irrelevant, and there are very clear causal connections between their respective wax and wane, one semantic pivot displacing the other as a substantive theory of meaning entered Chomskyan linguistics. Seuren 1998:217–18 traces “deep grammar” to Hockett 1958:252; it is unlikely (though, who knows?) that this work had much direct influence on the Chomskyans, but Chomsky himself certainly read the book, citing it several times, and his famous choice of labels is a distinct echo of the Hockett chapter title, “Surface and Deep Grammar” (see especially Chomsky 1964d [1963]:30). My story also includes anachronisms for continuity with later discussions—for instance, Katz and Fodor did not introduce feature notation (the uses of pluses and minuses to describe attributes); also, trees have generally been simplified throughout. For instance, tense morphemes and Aux nodes are ignored.

27.This discussion involves my familiar anachronistic handling of history: strictly speaking, Katz & Fodor did not introduce [±] notation, using Porphyrian binary trees to represent the distribution of the semantic and syntactic features.

28.They are adopting homologous proposals from Lees 1968 [1960]:19 and Klima 1964 [1959] here.

30.See Newmeyer’s figure 3.1 (1980:65) for an even kludgier representation of Fillmore’s suggestions.

31.Curiously, Chomsky first introduced recursion as a necessary property of the language in a way that implicates the base in Syntactic Structures (1957a:18–21), in connection with finite-state grammars, before he rejects such devices as inadequate. Pullum 2011:288 points out that Generalized Transformations (or, at least, T_Conj) required something that we will later see being reviled by Chomsky, a transderivational constraint.

32.Actually, Katz and Fodor’s equation was “linguistic description minus grammar equals semantics,” where grammar meant phonology and syntax. I’ve translated the equation into more specific terms, to avoid confusion over grammar, which includes semantics in this book, and in linguistics generally from about 1965 (see Chomsky 1965 [1964]:16ff).

33.McEnery & Wilson 2001:11 do a more formal corpus analysis of Chomsky’s claim and find a wide range of counterexamples to Chomsky’s perform intuition.

34.See “New Grammar, The” (1965) for the details and quotations in this section, and the photograph. Shugrue 1966 and Pooley 1968 give more comprehensive accounts of these curricular moves, detailing the federally-funded developments in California, the District of Columbia, Florida, Illinois, Indiana, Minnesota, Nebraska, New York, Pennsylvania, Ohio, Oregon, and Wisconsin, as well as related research at numerous universities, and various other initiatives, proposals and studies. The sixteen-school study cited by Newsweek is reported in Bateman & Zidonis 1966; see also Zidonis 1965 & 1967.

35.As above, many scholars have commented on the boost that Cold War military funding gave to Chomskyan linguistics, but the educational side of this boost to its popularity often goes unnoticed. The 1957 success of Sputnik threw the United States into a panic over almost every cultural and social factor that might be measured, and literacy was high among them. A 1965 assessment of U.S. education noted that in the late 1950s,

educators and laymen alike awakened to the sharp realization that we must put forth more vigorous effort if we were to preserve and improve the American way of life. In all aspects of national endeavor pressures were felt to produce more and more and to do it faster and faster. In reading, pressure to produce higher competency in a shorter time became apparent. (Smith 1965:312)

In English studies, Transformational Grammar quickly became a front runner among remedies for the perceived literacy gap (Sledd 1966:166–67), with major funding for experimentation, pilot studies, and curricular development through the 1960s.

Chapter 3

1.Warning: This note is for historical aficionados only. The discussion here, and, in fact, this book as a whole, ignores the Generative-Semantics-like proposals by people outside the immediate transformational community that sprung from MIT’s loins. So, for instance, Martin Kay delivered a paper in 1967 sketching a model which mapped semantic representations, expressed in symbolic logic, onto phrase markers; which was encased in an argument from “intellectual hygiene” akin to Postal’s (1972a [1969]) “Best Theory” argument; which included “rhetorical predicates” that paralleled Ross’s performatives; and which showed a concern over the competence-performance distinction that surfaced later in Generative Semantics (Kay 1970 [1967]). This paper appears to have had no influence on Generative Semantics whatsoever. It was delivered at a major conference (the Tenth International Congress of Linguists), at which Ross (and, perhaps, a few others in our dramatis personae) was present, but its omission from the Generative Semanticists’ references is not especially unusual. Transformational Grammarians of the period were very parochial and quite indifferent to proposals outside their immediate framework. Similarly, Petr Sgall also “explicitly propose[d] that the basis of a generative grammar should be constructed out of a set of semantic concepts” at a 1964 conference in Magdeburg, Germany (see Brekle 1969b:84–85, who mistakenly calls Sgall the earliest to make such a proposal, indicating just how obscure Lakoff 1976a [1963] was). He, too, made no impression on the group under discussion; very likely, they knew nothing of the proposal at all. Some independent developments in West Germany also bloomed in Generativesemantische ways in the mid-1960s (Vater 1971:13; of that group, Brekle, at least, was influenced by Sgall). Several linguists at Warsaw University—most notably, Andrzej Bogusławski and Anna Wierzbicka—began exploring semantic primitives in the early-to-mid-1960s, and Wierzbicka visited MIT in the 1966–67 academic year, when she urged the emerging Generative Semanticists to go more deeply, more quickly, than they were prepared at the time to go. A few of her papers made some impression on the underground circuit (Wierzbicka 1976 [1967]; 1972:166–90, 203–20), and her work was generally well received by Generative Semanticists, but she was never an active member of the school; indeed, she found Generative Semantics semantically rather halfhearted. As she recalls, she pushed Ross and Lakoff to accept “that Deep Structure was both semantic and universal. Both Haj and George listened to me politely but told me that Bogusławski and I were too radical and that they wouldn’t want to go that far themselves . . . You can imagine my shock when one day [they] announced [in their jointly taught class] that they were abandoning Chomsky’s position on the relationship between syntax and semantics, and rejecting a syntactic “Deep Structure” in favour of a deeper, universal, semantic Deep Structure. . . . They didn’t refer to our conversations at all” (Newmeyer 2014:257, quoting personal communication with Wierzbicka from 2012; Newmeyer’s interpolations). Wallace Chafe proposed obliterating the syntax-semantics distinction in generative grammar and dispensing with Deep Structure in a 1967 Language paper and a review of Katz (1966) the same year (Chafe 1967a, 1967b; see also Chafe 1970a, 1970b; Langacker 1972). He was an outsider at the time, though he became a sympathizer of the general program of Generative Semantics, while rejecting many of its specific proposals.

If I am reading Seuren right (“It was proposed in Generative Semantics, after 1964, that syntactic transformations should be semantically invariant,” 2009:110), he appears to be saying that Katz and Postal created Generative Semantics when they proposed their semantic-transparency principle. Both Katz (categorically) and Postal (temporally) would disavow that association.

I also make very little attempt, except in passing, to follow the developments of Generative-Semantics-like models influenced by the conceptions of the Lakoffs, Ross, McCawley, and Postal. Liefrink 1973, for instance, developed his interesting Semantico-Syntax in the wake of those Generative Semanticists, but had virtually no interchange with them. Similarly Sgall’s later work was influenced by their proposals, particularly through his commerce with German linguists working directly in the Generative Semantics framework (see, especially, Sgall, Hajičová, & Benešová 1973), and Eugene Nida claims “a dependence upon the generative-semantics approach” for his componential analysis (1975:7–8), especially under the influence of Chafe’s work. The one lesson we can take from all of these overlaps of approach privileging meaning is simply how compelling that position is—as a “natural” position, but also and especially in the wake of Aspects.

2.Most of these quotes and paraphrases are from discussions, letters, or email. Newmeyer’s comments, however, are also in print (1980a:93f; 1986a:82f), as are McCawley’s (1976a:159), and Ross gives the lion’s share of credit to Postal in the acknowledgments to one paper (1974b:122).

3.See Lakoff’s memoir, “The way we were; or, the real actual truth about generative semantics” (1989), for many insights into the movement overall and some details about her own involvement. Bucholtz and Hall also rightly point out that Robin Lakoff’s pioneering work in feminist linguistics (coinciding with the devolution of Generative Semantics), develops naturally from the pragmatic and sociolinguistic concerns that she developed in Generative Semantics. “Post-Chomskyan” is from Green (1970:153).

4.Nor was Case Grammar formulated in any direct alliance with Generative Semantics. “When I wrote ‘The Case for Case,’ ” he later noted, “the Generative Semantics position had not yet been formulated . . . in my hearing except in its preliminary version as Abstract Syntax” (Fillmore 1977:62).

5.Lakoff actually dates Generative Semantics even earlier, to his 1962 undergraduate thesis (Huck & Goldsmith 1995:107). I won’t pursue those origins here, since (1) Lakoff’s undergraduate thesis had a vanishingly small sociological effect, whereas his 1963 paper saw at least modest circulation among relevant linguists, (2) the story I’m telling in this book is of rhetorical currents, not of isolated personal development, and (3) there are some clear lines connecting that paper to the later 1967 joint letter with Ross, which is widely recognized as the manifesto for Generative Semantics, and the trigger for a sociological phenomenon in linguistics. But (4) it is hard not to notice the parallel with Chomsky, who often says of his own undergraduate thesis that it was “the first ‘generative grammar’ ” (1979 [1976]:112).

6.MITRE is professedly “just a name,” though it is always given in all-caps, and has often been assumed to stand for Massachusetts Institute of Technology Research & Engineering (mitre.org/news/media-resources).

7.More accurately, he softened the distinction on principled grounds by translating it into feature notation—Ross 1969b [1967].

9.For the agitated grammarians in the crowd, yes, I do know that the verb in my examples here is not show but one of the special compounds known as two-word verbs—namely, show off.

10.There are some anachronisms and simplifications in this discussion—in particular, I am taking the decompositional analysis further than Lakoff does in his thesis, and using abstract verbs rather than feature bundles, but this stretching simply goes in the direction these analyses later followed in Generative Semantics.

12.Brief histories of the proposals that brought linguistics to this inventory (with some slight variations) and its relation to symbolic logic are given in McCawley 1976b [1968]:136–39, Newmeyer 1980a:148–50 & 1986a:100–101, and R. Lakoff 1989:946–53. Additionally, not every abstract syntactician was happy with only three base categories; Ross, for instance, retained the VP in much of his work.

13.I have reduced the bracketing somewhat, for expository reasons. A more fully bracketed string for 18 would represent all the lexical categories, as in (((the)_Det(man)_N)_NP ((love)_V((the)_Det(platypus)_N)_NP)_VP)_S

14.The influence of Harris in this area is substantial. He had ensured that Chomsky read widely and studied deeply in modern logic. Harris and Chomsky both stayed away from a direct incorporation of logical syntax and semantics into linguistics, but they borrowed from the mechanisms and terminology for their models eagerly. For attitudes about logical syntax, see especially Chomsky’s 1955b response to Bar-Hillel in Language, occasioned by the latter’s 1954 criticism of Harris. For faith in logical mechanisms, see Harris 1951a [1947]:18. See Tomalin 2006 for extensive analysis of the role of symbolic logic in Harris and Chomsky up to and including Syntactic Structures.

15.“Concerning the Base Component of a Transformational Grammar” (McCawley 1976b [1967]:35–58); see Anderson 1976 [1966]:l14ff, Lancelot et al. 1976 [1968]:258, and Bach 1968 [1967]:114 for commentary.

16.Greenberg showed up only sporadically in discussions of the universal base (as in Bach 1967; G. Lakoff 1968b [1966]), but R. Lakoff 1989:950 suggests he had a good deal to do with the Generative Semanticists’ notions about constituent order.

17.See Newmeyer 1990:170 & 1991 [1989]:208–14 et passim for a fuller discussion.

18.Robin Lakoff 1968:170 and McCawley 1976b [1967]:84 both credit Ross with the proposal, and it was one of the highlights of the elaborate Floyd talks he gave on the early lecture circuit. In addition to Lakoff and McCawley, Boyd & Thorne 1969 and Sadock 1969 also beat Ross to print with underlying performative treatments. Sadock’s work was clearly inspired by Ross, but Boyd & Thorne’s proposal appears to have been independent. Their notes acknowledge the comments of Ross on earlier versions of their proposal, but don’t indicate any influence beyond useful criticism. In any case, Ross does not mention their work in his paper and it had virtually no impact on the performative debates, which is unfortunate because it is in some ways more subtle, cutting finer semantic distinctions. The paper, published in England, appears to have been a casualty of American chauvinism.

19.Ross says only that the underlying abstract verb is “like say” (1970b [1968]:238), and that it is associated with certain syntactic features, like [+performative] and [+declarative]. Since nothing of consequence hangs on representing the abstract verb as say or declare, as most representations of his argument do, I have opted for the more congruous tell, a convention later adopted by Ross (1975:249n19).

20.According to Ross (1975b:71), Predicate-raising began life as Lakoff’s transformation, Plugging-in; presumably this is from unpublished work which generalized inchoative and causative transformations.

Chapter 4

1.Bolinger’s comments here are spliced from two sources; some come from letters to me, and some from his “First Person, Not Singular” memoir (1991 [1974]:29). Newmeyer (1980a:93 & 1986a:82) characterizes the Lakoff-Ross seminars as devoted to challenging analyses then favored by Chomsky, which both Lakoff and Ross deny. There is evidence on both sides. Some participants recall antagonism toward Chomsky, others recall reverence; the two attitudes are not exclusive and were probably equally well represented. Indeed, even relatively late in the Wars, Generative Semantics papers could cite Syntactic Structures or Aspects like scripture and dismiss Chomsky’s contemporary work rather curtly.

2.Chomsky’s leave was for the fall term of 1966. He returned to teach two courses in the spring term of 1967, “Structure of English I,” on phonology, and “Intellectuals and Social Change,” which indicates how his political concerns were increasingly taking his energies. The lectures Jackendoff cites (the “Remarks” lectures) took place in fall 1967.

3.In the three-essay book featuring “Remarks,” Lees is mentioned eleven times, Lakoff thirty-six, with Ross getting eighteen, McCawley twenty-one, and Postal twenty-two.

4.Chomsky also mentions Paul Chapin even more briefly (1972a [1967]:17n), who presents fascinating case. He was the most obvious ‘target’ in an essay proposing lexicalism in explicit opposition to transformationalism. The thesis statement of Chapin’s dissertation, under Chomsky, is “to show that this proposal [to “account for derivational relationships by an enriched theory of the lexicon”] (which I will call the lexicalist hypothesis) is untenable” (1967: 13). Chapin’s thesis was accepted in August 1967, citing the origin of this hypothesis as Chomsky’s “lectures and private conversations” with him. Chomsky’s “Remarks” paper was mimeographed for his fall-term students (making the rounds much more widely). Yet Chapin, who apparently had naming rights for the hypothesis Chomsky was advancing and who argued extensively for the transformationalist position gets only the most glancing notice.

6.The reference to Time concerns a Lakoff recollection I had just related to Postal: Lakoff, Chomsky, and Ross had only one meeting upon Chomsky’s return, which was interrupted by a call from Time magazine which took up most of the scheduled time; subsequent meetings were cancelled. Chomsky getting the call is certainly easy enough to believe. Demands for his time were extremely pressing (see Shenker 1971:105, for instance, who discusses these demands, and a few pages later remarks that “characteristically, Chomsky had to interrupt our conversation to attend a meeting of dissident professors at MIT. His colleague Morris Halle . . . replaced him on the dilapidated chair.”

7.Chomsky may have adapted the term from Lakoff, or Lakoff may have picked it up from Chomsky’s lectures, or it may have just arose in the debates spontaneously to them and others. But Chomsky’s first attested use is in “Deep Structure, Surface Structure, and Semantic Interpretation,” circulated from 1968, published in 1972 (1972a).

8.Chomsky’s use of determine in such contexts earned him some rebukes from Generative Semanticists about his finesse with terminology. See Lakoff (1971b:236–37) and McCawley (1982b [1973]:66).

9.The title under which Chomsky’s paper was read is in some ways just as value-laden as Postal’s, but also, after the comma, refreshingly temporal for a guy who regularly talks and writes as if his own intellectual history is the simple, inevitable progress from clear early insights, “Credo, 1969.”

10.The pagination for the quotations from “Some Empirical Issues” (1972b [1969]) is: “uninteresting” (137), “vacuous” (133), “totally obscure” (148n22), “no substance” (146), “permitting any rule imaginable” (141), “at best dubious rules” (152), “a terminological proposal of an extremely unclear sort” (137), “not only unmotivated but in fact unacceptable” (150), “is probably correct, in essence,” (151) “more natural” (187), “somewhat more careful” (188), “well-supported” (165), “to be preferred” (196), “again to be preferred” (197), “more restrictive, hence preferable” (197).

12.I apologize to readers who find McCawley’s pseudonyms offensive because of their obvious measure of anti-Asian racism, however unintended by McCawley, or by me in my recirculation of them. Likewise, for readers who find them offensive because of the crudity of their phonological evocations of English terms, I also apologize (there’s more crudity of that kind coming as well). On both counts, it simply would not be reflective of the ethos, with its melange of political allusion, scatology, and counter-culture nose-thumbing, not to mention systemic racism, to leave these references out, and McCawley, an inveterate free-speech advocate, would curse me from the beyond if I omitted them on the basis of any discomfort they might cause.

13.Chomsky & Halle (1965); see, especially, 106, 108, 119, 126, 127, 128n25, 129n26 133n27, 136 for the more obvious Householder insults.

15.In a decades-later attempt at mitigation, the introduction to the third edition (by James McGilvray) says, somewhat startlingly about a book subtitled A Chapter in the History of Rationalist Thought, that the book “does not pretend to be a work in intellectual history” (in Chomsky 2009b [1966]:2), and quotes Chomsky as saying “one might say that I’m looking at history . . . from the point of view of . . . an art lover who wants to look at the 17th century to find in it things that are of particular value and that obtain part of their value . . . because of the perspective with which he approaches them” (in Chomsky 2008 [1966]:109n2). But Chomsky has only addressed critics of the book in casual circumstances (e.g., Parret 1974 [1972–1973]:32; Chomsky [and Ronat] 1979 [1976]:77–79; Barsky 1997a:105), and has done so in rather scurrilous ways. For instance, in Barsky 1997a, Chomsky dismisses the criticism as the “absurdity and falsification” typical of his critics, proudly declaiming that he has “never bothered to respond, because [warning: the following analogic directions may induce vertigo] my contempt for the intellectual world reaches such heights that I have no interest in pursuing them in their gutters” (1997a:105, quoting a letter dated March 31, 1995), though he builds an elaborate and viciously distorted attack on one critic, Hans Aarsleff (see R. A. Harris 1998 for documentation of the bilious mischaracterizations, and of Barsky’s credulity in repeating them). Chomsky did, however, feel the need to inoculate his book against the charge that it is historically inaccurate, since he gives Barsky a defense, which Barsky quotes and paraphrases as:

The term “Cartesian” is not used [by Chomsky] according to its generally accepted definition. Chomsky extends that definition to encompass, as he puts it, “a certain collection of ideas which were not expressed by Descartes, [were] rejected by followers of Descartes, and many first expressed by anti-Cartesians” 1997a:106, quoting the same March 31, 1995 letter)

This defense actually brings us back to Robin Lakoff’s question, why would one use the term Cartesian for a collection of non-Cartesian, anti-Cartesian, and rejected-by-Cartesians ideas?

16.The paper has been republished with a brief but illuminating postscript (in Schiller et al., 1988:25–45).

17.See Postal’s account of this event, including a clinical itemization of Chomsky’s brutal incivility (Huck & Goldsmith 1995:134, 161–62n6). Ross reports that Chomsky later made him a quasi-apology, saying that he “just couldn’t take” what the Generative Semanticists as a group were saying (Huck & Goldsmith 1995:125).

Chapter 5

1.At a theory-comparison conference in Milwaukee (Moravcsik & Wirth, 1980), for instance, a mimeograph paper made the rounds, concerning specious evaluation arguments such as “the claim that ‘Your theory is a notational variant of mine, and it’s wrong’ ” (Lawler 1980:59n14, quoting or paraphrasing K. Whistler et al.).

2.See Huck & Goldsmith’s fascinating account of the correspondence, publications, and overall argumentation (1995:60–70), an episode they characterize as “comical” except for the fact that “it bore serious scientific consequences” (158n4).

3.Dougherty, by the way, does get closer to the real problems of McCawley’s original argument—as, in fact, McCawley does, both in his correspondence with Chomsky and in a review article of the book in which Chomsky published his straw reconstruction (1982b [1973]:39–41). See McCawley (1976b [1967]:121–32; 1972:535–38; 1982b [1973]:39–41; 1988.2:536–41). Lakoff gives a fairly clear presentation of McCawley’s argument (1971b:273–77), but ends up only with a lukewarm endorsement. Lakoff also offers some Hallean-type arguments with other constructions somewhat later (1972b:547–59). See also Green (1974 [1971]:7, 27–28), who indicates that even if Generative Semanticists were not willing to acknowledge the full force of McCawley’s argument, they felt it at least indicated something was wrong with Deep Structure.

4.Chomsky’s remark was in a letter to McCawley (December 20, 1967) over Deep Structure; he makes a similar observation in Logical Structure (1975a [1955]:93), though obviously not about Deep Structure.

5.See Postal 1971b [1969]—which Ross still calls “the best articulation of what Generative Semantics is about . . . a beautiful argument;” G. Lakoff 1971b:246–52; Postal 1988a [1969]:85—86; Binnick 1971.

7.McCawley also had an elegant way around this violation in terms of logical form (1976b [1967]:108–9).

8.Shallow Structure was apparently Postal’s term for the level after the application of all cyclic transformations and before any post-cyclic transformations (Perlmutter 1968:239n17).

9.The fleshing-out comes especially in two important books. Semantic Interpretation in Generative Grammar (1972), and

Syntax (1977), and an important paper, “Morphological and Semantic Regularities in the Lexicon” (1975).

10.The only possible contender in Chomsky’s oeuvre as an articulation of the Extended Standard Theory is his Studies on Semantics in Generative Grammar (1972b), but it just collects his anti-Generative Semantics trilogy.

11.His first major work, for instance, adopts

-notation for some Phrase Structure rules, but seems embarrassed to extend the conventions to other rules (e.g., 1972:6ff). In particular, he uses

-syntax only for noun phrases, as had Chomsky—a move that must have reinforced the impression that the notation was an ad hoc convenience Chomsky had drafted to get himself out of the explanatory pickle he was in for having abandoned a transformational account of derived nominals.

12.Or, maybe Jackendoff had one semantic representation with four parts, which amounts to the same thing. See Jackendoff (1972:3) for the clearest account of his architecture.

14.The term “wild cards” in connection with Chomsky’s shell-and-pea data game comes from Lakoff’s (1967) response to “Remarks on Nominalization.” Postal’s comments here have been edited slightly. A longer passage containing these remarks is quoted in my dissertation (Harris 1990:381–82).

15.Chomsky’s old high school chum, Hilary Putnam, had anticipated this result very early in the ascendancy of Transformational Grammar, noting at a symposium on the mathematics of language that featured Chomsky’s proposals, “It is easy to show that any recursively enumerable set of sentences could be generated by a transformational grammar in Chomsky’s sense” (1975 [1961]:105), but linguists were not yet ready to listen.

16.The applicability of the Peters-Ritchie results needs some qualification. The results concern what is known as weak generative capacity, the ability of a rule system to specify strings of symbols. Chomsky argues in Aspects (1965 [1964]:60–62) and elsewhere that this is not the goal of linguistic theory, which should concern itself with strong generative capacity—the assignment of structural descriptions to strings of symbols (for instance, assigning Phrase Structure trees to English sentences). Additionally, he argues in Aspects that “the real problem [of grammatical theory] is almost always to restrict the range of possible hypotheses by adding additional structure to the notion ‘generative grammar’ ” (1965 [1964]:35), where he is also talking of (among other criteria) strong generative capacity. That is, restrictiveness for Chomsky does not dovetail with Peters and Ritchie’s work in any direct way. Nonetheless, the results and Chomsky’s call for restrictiveness (which was far more urgent in “Some Empirical Issues” than in Aspects, or anywhere else) were almost always mentioned in the same breath. Newmeyer, for instance, the most astute and thorough historian of Chomskyan linguistics, says that

The Peters-Ritchie findings served as silent witness to almost all of the significant work in syntax in the 1970s [where significant means significant in the Interpretivist framework]. There was hardly a paper written that did not appeal to the increased restrictiveness of the theory that followed as a consequence of the adoption of the proposals in its pages. Constraint after constraint was put forward to limit the power of the grammar. (Newmeyer 1980a:176)

19.The want to/wanna data was actually first noticed by Larry Horn, whom Lakoff cites in his global rules paper, along with related data suggested by Sadock and Fillmore (1970b:632).

20.McCawley (1982b [1973]:29), however, observes that Chomsky had no compunction about appealing to transderivational constraints as needed—in particular, that his “rules of analogy” ploy in “Remarks” was an appeal to a transderivational rule—but that he avoided the term like poison. McCawley makes a similar point with respect to the use of global rules in Chomsky & Halle (1968); see McCawley (1974c:73).

21.Actually, the term in Chomsky is usually grammaticalness (e.g., 1965 [1964]:3, for acceptability, see, e.g., 1965 [1964]:11).

23.The archeology necessary to trace these all out is beyond me, but the double asterisk seems to have begun with Lakoff’s dissertation (1970a [1965]:62), and the double shriek seems to be from Janet Dean Fodor, not a Generative Semanticist (1976 [1970]:143).

24.See, as you might have guessed, Jackendoff (1972:120ff) for a quite detailed argument supporting this syntactic- to-semantic reanalysis.

25.For a fascinating history of empiricism and rationalism in linguistics, see Chater et al. 2015.

27.Both observations are true, of course—again I recommend Tomalin (2006) for following these threads of influence—but the extraordinarily dependent and the two taken . . . directlys wilfully ignore the very original shapes those influences took in Chomsky’s work.

Chapter 6

1.There are several important caveats to make, however. First, there were many linguists who bought into Generative Semantics—who, for instance, published dyed-in-the-wool Generative Semantics papers like “On the Alleged Boundary between Syntax and Semantics” (Newmeyer 1970), and “On the Syntax and Semantics of the Atomic Predicate cause” (Dowty 1972)—but whose work shows only faint traces of this style. Second, there are linguists who employed this style only in very restricted subsets of their writing. Postal, for instance, tended to write in a formal, almost Chomskyan style, but frequently juxtaposed offbeat sample sentences to this prose, and his publications carried a much more serious tone than his samizdat papers. Third, some Interpretive Semanticists, like Jackendoff and Akmajian, also used elements of this style—considerably more elements, for instance, than Newmeyer or Dowty. And, most importantly, a great deal of work which showcases this style is to be found in the various proceedings of the Chicago Linguistic Society (CLS); that is, one of its principal aims is to be effective when delivered orally. Indeed, this chapter might, not unreasonably, have been entitled “The CLS Ethos.” But the identification of Generative Semantics with the CLS in this period was an extremely close one. The society was the organizational center of Generative Semantics, publishing at least half the movement’s papers in its proceedings, serving as a swap meet for the latest ideas, and throwing the best parties.

2.For readers to whom this allusion is obscure: there was an elaborate and bizarre rumor, characteristic of the paranoia and mysticism of the period, that Beatle Paul McCartney was dead, that there had been a massive cover-up of this by evil-minded executives (aka “The Establishment”), but that the genuine remaining Beatles were trying to get the truth to the faithful in the form of cryptic lyrics, symbolic album covers, and, strangest of all, by encoding messages that could only be discerned by playing certain sections of their records backwards.

3.The k-spelling of Amerika was distinctive of counterculture orthography, overlapping with African-American dissident usage. The Oxford English dictionary, dating it to 1969, defines it as “Alteration of the name of America . . . after German Amerika and Russian Amerika, to express associations of fascism and authoritarianism.”

4.To keep this paragraph from getting too bogged down by parenthetical references, I’ve saved them all for here: Flip, Ross recalls, was coined by Postal in his early 1960s classes at MIT, showing up in print in Lakoff’s thesis: Slifting (Ross 1973a); Sluicing (Ross 1970a [1967]:252); Stuffing (Ross 1972a:162); Irving (Morgan, ms., cited in Horn 1970:325); Ludwig (Neubauer 1970:403); Richard (Rogers 1971); Apparel Pronoun Deletion (Grinder 1970:300); Euphemistic Genital Deletion (as above, cited in Borkin 1972:26; see also 1984 [1974]:105–106, McCawley 1973c:236).

5.There is actually some serious linguistic work going on in Left Field as well, but the work was deliberately linked to themes of scatolinguistics, pornolinguistics, whimsy, parody, burlesque, and so on (using some of the editors’ selected labels for the content).

6.The politics/linguistics segregation in Chomsky’s life and work has been widely noted. Paul Robinson calls this “The Chomsky Problem,” and Chomsky himself “a disturbingly divided intellectual” (1979:3). Chris Knight has pursued this theme in great detail, seeing a “veritable firewall” Chomsky has built “to keep these two constituencies of his apart” (2016:134; see also 2018a, 2018b, 2020).

7.All quotations in this paragraph are from the preface to Borkin et al. 1968. The first quotation is a paraphrase of Burt’s position, presumably by Borkin. The others are quotations directly from Burt.

Chapter 7

1.Newmeyer (1979), however, draws exactly the opposite conclusion about the book. He apparently admires it a great deal but finds the framework it articulates seriously wanting.

2.There are many discussions of Chomsky’s Cartesianism and the issues surrounding it. The most thorough are Kretzmann 1975 and Behme 2014a; see R. A. Harris 1998 for a very unsympathetic account of Chomsky’s accusations against Aarsleff.

3.The first appearance of the term, natural logic, invoking the idea that logic should be an empirical pursuit, appears to be in G. Lakoff (1971b:277), but tendencies in this direction date back at least to McCawley (1976b [1967]:106). So natural logic was not solely Lakoff’s project, though he was on this front, as on most fronts, the most vociferous. At one conference, for instance, McCawley made some modest suggestions in the hope of stimulating logicians “to study the logical properties of items [they] generally ignore (e.g., is it valid to argue ‘goddamn all imperialist butchers; Nixon is an imperialist butcher; therefore, goddamn Nixon’?)” (1976b [1972]:319). At the same conference, Lakoff delivered a long, forceful sermon (1972c) to the effect that logicians had painted themselves into a corner by allowing their formalisms to shrink the subject matter of logic until it contained only the narrowest subset of facts about human reason and its vehicle, natural language; Generative Semantics, he told them, would be their salvation.

5.Postal’s influence extends to the titles as well. His Anarchy Notes use titles like “Horrors of Identity,” “Temporal Monstrosities,” and “Coordinate Mind Snappers” (1976 [1967–70]:203–204).

6.Lakoff has been quite explicit about the commitments he and the others held, and how he perceived them to shift. See especially his discussion with John Goldsmith (Huck & Goldsmith 1995:107–19), as well as some remarks to Francisco José Ruiz De Mendoza Ibáñez (Lakoff [with Mendoza Ibáñez] 1997:41–42).

7.Cognitive Grammar comes from some papers he developed with Henry Thompson in the mid-1970s (G. Lakoff & Thompson 1975a; 1975b); Experiential Linguistics, from Lakoff (1977a). Newmeyer notes Lakoff’s fondness for new labels—“By coming out almost yearly with a newly named theory . . . Lakoff has not presented himself to the linguistic world as a consistent theoretician” (1980a:172)—though two of the labels Newmeyer cites (Global Transderivational Well-Formedness Grammar and Dual-Hierarchy Grammar) are from underground sources. The name Cognitive Grammar was later adopted by Ron Langacker for a different and much more fully specified theory (e.g., 1987, 1990, 1991), and, as we will see, Lakoff eventually settled under the umbrella of Cognitive Linguistics.

8.More precisely, the name was Move-σ, where σ was a variable that stood for any constituent, but the σ was dropped after a while by Chomsky and I have followed my selective-anachronism practice here in the hope of avoiding unnecessary complications.

9.The name change was half-hearted at best and slow to propagate, though eventually it took root, sort of. Chomsky himself still regularly used the term Deep Structure for a long while after his formal repudiation—rather perversely as a gloss for D-Structure (e.g., 1981a [1979]:5, 18, 362; 1982b:5, 2002:113), and his followers often adopted such locutions as “D(eep)-Structure” (May 1985:3).

11.This elevation of frankness was another counterculture trail. See, for instance, Rubin’s (1971:124–26) discussion of the hypocrisy of the Chicago police testifying in court, who insisted they could not repeat some of the defendants’ words in front of women jury members, but who boasted to one another in the back room about how they busted the heads of “the fucking little fagots”; Yippies, on the other hand, don’t hide anything.

12.Chomsky’s close ally, Robert Freidin (2010 [1997]:262–265) gives a version of the dispute much in line with Chomsky’s. Feargal Murphy (1997) endorses and slightly extends Barsky’s representation of Chomsky’s account. D. Terence Langendoen (1995:586n10) endorses at least some points of Chomsky’s version. But detailed studies of the clash (Huck & Goldsmith 1995, 1998; Koerner 2003b:105-130; R. Lakoff 1989; Newmeyer 1980a, 1986a, 1996; Seuren 1998) offer accounts that do not fit at all well with Chomsky’s. Pullum (1997a) specifically contests Chomsky’s account in Barsky; Pullum & Postal (1997) respond to Murphy (1997); Barsky (1997b) responds to Pullum & Postal, and back to Pullum (1997b).

For my money, the most damning evidence against Chomsky’s version is in the account he gives to Grewendorf (Chomsky 1994), since he offers this self-representation immediately after another one, the familiarly ludicrous account that, despite having a PhD from one of the most distinguished institutions in the United States, under one of the most acclaimed and respected linguists in the world, he was “largely self-taught (including linguistics),” he had “no serious professional qualifications in any field,” and he got his job at MIT because “no one cared much about credentials” (not because, for instance, on top of his obvious documented credentials, and his prestigious fellowship at Harvard, and his publications, and the active support of Roman Jakobson, and the pipeline his supervisor had to the RLE, that his good friend Morris Halle was making the decision). See also Barsky (1998:86). Even without the counterevidence of so much testimony and literature standing against Chomsky’s account, why on earth would one believe a story that sits exactly next to another one so blatantly untruthful?

14.Everyone in the period would have recognized, for instance, that a comment like “the appeal to meaning within the theory of linguistic form has all too often served simply as a way of side-stepping serious inquiry” (1976:21–22) was directed at Generative Semantics; we hear that G. Lakoff has advanced one argument that is “hopelessly in error” (49n28), Postal another that is “not . . . at all convincing . . . [and] surely incorrect” (50n40). Ross, it should be noted, is singled out for his “very important study” (i.e., his thesis, 1986 [1967]).

15.Chomsky insists that he never tries to persuade anyone, and that no one else should either: “I don’t have any theory of rhetoric, but what I have in the back of my mind is that one should not try to persuade. . . . To the extent that I can monitor my own rhetorical activities, which is probably not a lot, I try to refrain from efforts to bring people to reach my conclusions. . . . So I think the best rhetoric is the least rhetoric” (2003 [1991]:376).

Chapter 8

1.See also Jackendoff’s confession in Foundations of Language: “The reaction in the wider community was one of disillusionment, above all at the bad behavior displayed by both sides in the dispute (including the present author)” (Jackendoff 2002:74).

2.The quotations here are, in order, from Seuren 2005:521, 2005:521, 2004:71, and 2005:525. Seuren’s bitterness is so unrestrained that he could not even forbear from a rather personal attack on McCawley in a review of the tribute volume after his death (!), Polymorphous Linguistics: Jim McCawley’s Legacy (Mufwene et al., 2005). Seuren acknowledges McCawley’s creativity during the early Generative Semantics period, but condemns him thereafter for “picking his earnings, so to speak, from whatever sources were available, while trying, as far as possible, to keep up his flamboyant academic and personal style (which, one must say, did grow a little tired in the end);” for “intellectual weakness;” for “political views [that] were as monorail and dogmatic as they were quixotic and unrealistic;” and for a “Freudian twist of mind” that led him to embrace “a philosophy of science . . . that justif[ied] the destruction of the school of thought he had played such a central role in founding” (2005:525).

3.Heny briefly entertains such an alternate history, suggesting that if LSLT had been published prior to 1965, “we could have avoided the entire Generative Semantics period, and the subsequent, bewildering, piecemeal abandonment of one transformation after another, which has characterized much of the work in the Extended Standard Theory” (1979:318).

5.This is not to say, however, that no promises were addressed and none fulfilled. McCawley’s work from the early 1970s right up to his death was largely an attempt to make good on promises he made in the late 1960s. For instance, his abstract analyses of the 1970s addressed the simplicity promises of early Generative Semantics; Everything That Linguists Have Always Wanted to Know about Logic (1981, 1993) is the most detailed exercise in natural logic, by far, and The Syntactic Phenomena of English (1988, 1998) is the most comprehensive application of Generative Semantics principles to linguistic description. Levi’s The Syntax and Semantics of Complex Nominals (1978) satisfies much of the promise of a transformational analysis of nominalization that Generative Semanticists kept alleging was possible in response to Chomsky’s “Remarks.” G. Lakoff’s Women, Fire, and Dangerous Things (1987) addresses many of his late Generative Semantics promises, and all of his work has remained spiritually in sync with the later Generative Semantics framework’s commitment to fuzziness, general cognition, and the “Gestaltist” nature of language use. But, where these projects were not too little, they were certainly too late, and explicitly disengaged from the Generative Semantics brand in most cases.

7.“We prefer the term ‘Montague Semantics,’ ” Dowty, Wall, and Peters say, “inasmuch as a grammar, as conceived of in current linguistics, would contain at least a phonological component, a morphological component, and other subsystems which are either lacking entirely or present only in a very rudimentary state in [Montague’s] system” (1981:ix).

8.On Chomsky’s disinterest, see his remarks in 1982a [1979–1980]:70, though Partee’s (2005:18n45) observations suggest that this disinterest was not accompanied by hostility. See Janssen 2006 and 2011 for concise accounts of Montague’s heritage.

9.See especially Rogers et al. 1977 and Cole’s two collections (1978, 1981). Of the ten articles in Rogers et al., only three were not by card-carrying Generative Semanticists, all of them philosophers (Harman, Searle, Stalnaker); of the twelve articles in Pragmatics (Cole 1978), seven are Generative Semanticists (four are by philosophers, including Grice and Kaplan, with the remaining one by Givón, who had an early affiliation with Generative Semantics, but, by 1969 had turned his back on the dispute); Radical Pragmatics (Cole 1981), includes work by Sadock, Davison, Prince, Green, and Morgan, as well as Grice, Bach, and Fillmore. There is no presence in any of them by linguists of the Interpretivist camp.

10.By the time Form and Function was published, a lightly edited version of Borkin’s dissertation (Borkin 1984 [1974])—“a very typical example of the kind of work that was being produced within the framework of generative semantics . . . right down to the naughty examples and political references to Nixon” (Hudson 1985:263)—G. Lakoff and Ross were aware that her work involves figurative language in some way, saying in their preface that “the examples [Borkin] discusses . . . are known in the tradition of classical rhetoric as ‘metonymy’ ” (viii). Lakoff and Johnson use the term metonymy for these sorts of examples as well, regarding that trope unfortunately to include synecdoche as a special case (1980:40), and R. Lakoff (2006) offers an interesting reason for treating at least some examples of metonymy in Generative terms: “In construing the process not as a rhetorical trope but a grammatical process, the generative semanticists were making an important statement about the nature and scope of syntax, suggesting that the process of beheading (or metonymy) was predictable and rule-governed, that there were environments in which it was applicable and those in which it was not, that its scope of applicability could be rigorously stated”—adding, however, “that was never actually done.” I can find nothing from the period that discusses these phenomena in figurative terms. She also offers a curious argument that the phenomenon is not metonymy at all (rather than, say, a metonymical process governed by rules). One particular case of the Generative Semantics rule called Casteration, she says, “has been treated by some scholars as metonymy: reginam ineo standing for lectus/cubiculum reginae ineo ‘I go into the queen[‘s bed (chamber)].’ But in . . . its sexual sense, ineo is most commonly used of animals: Taurus vaccam init ‘The bull enters the cow.’ It is hard to imagine this as metonymy: cows do not wait for bulls in bedchambers.” No, but this is a clear example of synecdoche (whole = bull, part = penis) with metaphorical (zoomorphism) overtones (I = bull/animal). Two scholars who did recognize the intersection of rhetoric and pragmatics fairly early in the emergence of pragmatics out of the anomalies this sort of work posed to Chomskyan linguistics, Dan Sperber and Deirdre Wilson, have an important paper in Cole (1981), “Irony and the use-mention distinction,” in which they make one of the earliest claims that “rhetoric, like linguistics, is a branch of cognitive psychology” (1981:297; see also Sperber 1975, translated as Sperber 2007).

11.For her politeness article, see R. Lakoff 1973a; also 1977. Her Language and Woman’s Place is R. Lakoff 1975; see also 1973c, 2004; as well as the work of her most influential student, Deborah Tannen (e.g., 1996).

12.You probably don’t need a key, but I’m nothing if not redundant: Transformational Grammar, Transformational Generative Grammar, Standard Theory, Generative Semantics, Relational Grammar, and Extended Standard Theory.

14.We’ll only see a few of these below, and far too briefly for all of them, but here’s the key: Lexical-Functional Grammar, Daughter-Dependency Grammar, Word Grammar, Arc-Pair Grammar, Government-Binding Theory, Principles and Parameters, Minimalist Program, and Generalized Phrase Structure Grammar; all the offshoots of GPSG can be found in Pullum (1991:9), except for ETCPSG, which I am hoping you caught as my joke, for ETCetera Phrase Structure Grammar; if you didn’t catch the joke, and possibly even if you did, forgive me.

15.Chomsky never seems to have used REST—maybe that’s one joke he did get—continuing to prefer EST through the 1970s. But it was routinely used by others, even by initiates, and historically usually has the rough life-dates of 1973–1979, from “Conditions on transformations” (Chomsky 1973a [1971]) to Lectures on Government and Binding (Chomsky 1981a [1979]), with perhaps Chomsky & Lasnik 1977 as the signal document.

16.Part of the problem, too, was not with the type of rules, so much as with Chomsky’s specific construal of Phrase Structure Rules, particularly with respect to English syntax. As Harman put it, even before Aspects: “[Chomsky’s] critique of phrase structure consists in the construction of a formal model of phrase-structure theory and the demonstration that this is inadequate as a complete theory of grammar. The defense of phrase structure consists in repudiating the formal model. There are good reasons for repudiating the model of constituent structure, as Chomsky defines it, and for replacing it with a model which obviates the original criticisms” (Harman 1963:610–11).

17.Johnson & Postal (1980:15–19) sketch out some of the currents leading to Relational Grammar, including Ross’s argument that rule application was controlled by a hierarchy of grammatical relations (1974b: 106–10); some of Lakoff’s unpublished work from the period (partially reaching publication in 1977a) argued for eliminating derivations in a way that put more freight on grammatical relations (albeit not as primitives, but as syntactic products of semantic and pragmatic factors, rather than as Phrase Structure configurations); G. Lakoff & Thompson endorse RG, see 1975a. Edward Keenan and Bernard Comrie’s arguments for an Accessibility (or Case) Hierarchy, which circulated from 1972 but weren’t published in a formal version until 1977, were extraordinarily influential. Ross, for instance, leveraged his observations on them (Ross 1974b:106), and they shaped Lakoff’s work in the period (G. Lakoff 1977a:262). Socio-rhetorically, Relational Grammar gained considerable momentum after a very well attended Linguistic Institute session that Postal and Perlmutter team-taught in Amherst, 1974 (see especially Perlmutter 1980; Perlmutter and Rosen 1984; Blake 1990). Generative Semantics refugees who found their way to Relational Grammar (other than Postal), include Matthew Dryer, Peter Cole, Donald Frantz, and Jerry Morgan.

18.The Lexicalist developments building from “Remarks” were substantial—effectively, shifting the transformational burden entirely, or almost entirely, to “the lexicon”; that is, to base-component rules that increased lexical inventories in systematic ways. The most notable Superlexicalist outside LFG was Michael Brame, whose BGS (1978, 1979), Base-Generated Syntax, shared a number of central features with LFG but never achieved the same level of articulation, nor attracted significant numbers of adherents.

19.Missing from this sketch is Harman 1963, an article that argues for eliminating transformations by enriching phrase structure and was an early harbinger, if not a direct influence, on this trend of downloading transformational duties onto Phrase Structure, and that became something of a touchstone for GPSG (Gazdar & Pullum 1982:2; Gazdar et al. 1985:18). As always, of course, there is a Chomskyan crystal ball in the story: he projects an extension of the phrase structure component in Syntactic Structures, concomitantly diminishing the role of transformations (1957a:41–42n6; see also Chomsky 1955a:VI–229, 1975a [1955, 1956]:190). He calls such a move “ill-advised,” remarking “that it can only lead to the development of ad hoc and fruitless elaborations” (1957a:41–42n6), resulting in “very inelegant characterization[s]” (1955a:VI–230, 1975a [1955, 1956]:190), and his reaction to Harman 1963 was a characteristically withering dismissal, test-marketing an early version of his notational-variants charge (1966b [1964]:41–45), adding irrelevance, obscurity, and baselessness into the bargain, even implicating baboons; all the same, the Phrase-Structure-/Transformational-Rule tradeoff shows up in Chomsky’s earliest work.

20.Jan Koster, a scholar who was visiting MIT at the time and who soon became an important force in the development and propagation of Chomsky’s work—someone very sympathetic to his technical work and on “his side” with respect to Bresnan’s position—recalls how disagreement “gradually poisoned the atmosphere between Chomsky and Bresnan in [the] fall of 1976. At first, she was still coming to his weekly lecture on Thursdays, but Chomsky was so aggressive about her position (never let her finish a sentence) that she eventually didn’t come anymore, in the longer run inspiring her to go her own way.” That’s only one view—maybe Bresnan was asking long questions irrelevant to the course, so Chomsky needed to cut her short to do the job he was paid for; who knows?—but Bresnan left MIT in 1979. It is notable that neither Chomsky’s nor Bresnan’s published criticisms of each other’s linguistics had anything like the acid tones of the Generative Semantics dispute. In Chomsky (1977), for instance, he examines in considerable detail and rejects one of her arguments with no harsher pronouncement than “I am not at all convinced” (1977:118). It is also notable, with respect to her famous supervisor, that Bresnan says “Chomsky was my doctoral advisor, [but] my mentor was Morris Halle,” whom she credits as well for fostering the relationship with Kaplan that led to LFG (Bresnan 2011).

21.These three—Role and Reference Grammar, Co-Representational Grammar, and Equation Grammar—were picked pretty much at random. Okay, not at random at all; for their alphabetic colligations, to help you in your games of Linguistic Scrabble. The other frameworks at the Milwaukee conference also included some that predated Aspects (such as Tagmemics and Stratificational Grammar), some quite orthogonal to formal modeling (Functional Grammar and Functional Syntax), “Trace Theory” (a version of REST featuring its most prominent device), and perhaps the most traditionally rooted generative approach, despite its label, McCawley’s Unsyntax.

22.Haspelmath’s (2008:83) Table 1 lists these and other parameters with citations for their initial proposals and some related discussion and/or modification.

23.I don’t mean that it was incapable of providing more insights, just that it didn’t provide many more insights with the kind of influence earlier research had. Postal’s work, which remains in the RG mold, has certainly continued to provide new insights, particularly the picture of grammatical objects he presents in Edge-based Clausal Syntax (Postal 2011), but they are not perceived as coming from an RG paradigm, just from an individual RG linguist in a post-RG ecosystem. Indeed, the book’s foreword, by Chris Collins, goes out of its way to argue that Postal’s work could (maybe should) be largely absorbed into Chomsky’s program.

24.Given Lakoff’s propensities, his paper probably showed up somewhere, or parts of it, in some form, but it was not published anywhere in connection with the conference, and can only be glimpsed through Lawler’s 1980 critique, which seems more a commentary on Lakoff’s approach to linguistics than on the properties of some specific model. Cognitive Grammar, as a label, came to be associated with Ronald Langacker, and Lakoff is widely seen as one of the leading theorists who ushered the broader framework called Cognitive Linguistics onto the scene.

25.When their book came out, the next year, Lakoff and Johnson arrayed such data as this, some of it identical to the textbook, as their newly discovered ‘conceptual metaphor,’ ideas are food.

26.See Ricoeur’s historically detailed Rule of Metaphor: The Creation of Meaning in Language (whose title might easily have been Lakoff and Johnson’s), where one finds such observations as “The functioning of metaphor is to be detected in ordinary usage” (1975:92) and such representative historical quotations as Dumarsais’s (1730) “There are more tropes used in the marketplace in a single day than in the entire Aeneid” (1975:72) and, most tellingly, I. A. Richards’ remark that metaphor is “the omnipresent principle of language” (1975:92).

27.See Jakobson’s important paper on metaphor and metonymy as the realizations of cognitive poles organizing language and thought, which he called respectively Similarity and Contiguity (1990b [1956]).

28.These examples are just called “metonymy” in Metaphors We Live By (1980:35), and sometimes (in a slight improvement) “metonymic concepts.” In fact, Lakoff and Johnson’s phrasing is weirdly proprietary: “This is a case of what we will call metonymy,” they say, as if no one else had thought to use that term of such data (see also Lakoff 1987:19). The locution conceptual metonymy has now become common in Cognitive Linguistics, but it does not appear, so far as I could find, until Lakoff (1987:511), where he also says “Metonymy is one of the basic characteristics of cognition” (77), again confusing cause for effect. What I am calling correlation here, by the way (see also Dancygier and Sweetser 2014:5), has a long history in association psychology, where it is usually called contiguity (to signal a relationship of proximity in space or time between the relevant notions); see, for instance, Hume (1993 [1748]:14), where analogy similitude, etc., show up as resemblance.

Chapter 9

2.A synchronic slice of the Flea Circus, from the early 1990s, is available as McCawley 1991; see Meyer 1993.

3.See, in particular, McCawley 1985. At his death, he was working on a book to be entitled, A Linguist’s Guide to the Philosophy of Science (McCawley c.1996). There was some talk that it might be completed by colleagues and released posthumously, but nothing has come of that idea.

4.As a small sample of his personal reach (where personal is deeply permeated with the intellectual), see Mufwene et al. 1999, a thirty-thousand-word wall of tributes. As for his publications, the incomplete CV on his website (the last dated items are from 1996) lists well over two hundred items; see also Mufwene, Francis, & Wheeler (2005:xvii–xxx), a bibliography assembled for a tribute volume to McCawley.

5.His three collections of articles (1976b [1964–1971], 1979 [1965–1967], 1982b [1973–1979]), as well as his Eater’s Guide (1984), are all available in the ubiquitous second-hand book market, and well worth the effort to track down.

6.George Lakoff has a very important book in literary studies, influential in the development of Cognitive Poetics (Lakoff & Turner 1989), but not the same kind of deep literary interests as Ross.

7.See Ross (2000), George Lakoff (1987:582–85, 1989, 1997, 1999), Robin Lakoff (1989, 2014), and McCawley (1980a, 2009 [1994]). Postal has talked and corresponded about it with me, and with Huck and Goldsmith (1995:126–42), but mostly as a courtesy to others, as well as to illustrate early manifestations of something he continues to care about, Chomsky’s perfidy.

8.As advocates, in 2009, Postal could only cite works by Katz and himself, one of them written with another scholar, and one book by another scholar (Katz 1981, 1984, 1998; Katz & Postal 1991; Langendoen & Postal 1984; Carr 1990), for a position over thirty years old. There are few vocal opponents—Linguistic Platonism is treated mostly by silence in linguistics—but Jackendoff articulates some objections, chiefly that it “disconnects generative linguistics from all sources of evidence based on processing, acquisition, genetics, and brain damage” (Jackendoff 2002:297; see also Chomsky 1986:33–4; more neutrally, Stainton 2014).

9.After leaving IBM, Postal joined the linguistics faculty at New York University for about fifteen years, but in a semi-retired capacity, with one or two small-enrollment classes per year; the official designation was “Research Professor.” He took on only one doctoral student there.

10.These three quotations are from, respectively, Swain 1984:205, Bigand, Lalitte & Dowling 2009:185, and Pearce & Rohrmeierb 2012:469. Not all the initial reviews were welcoming (see especially Longuet- Higgins 1983, as well as Peel & Slawson 1984, to which Lerdahl & Jackendoff 1985 respond), but subsequent scholarship has been almost universally appreciative (see, in particular, the special issue of Music Perception, 26.3).

11.Goldsmith’s Autosegmental Phonology program (1976, 1990), which elevated phonological representations to the rich tree-structure expressiveness that characterized syntactic and semantic representations of the period, grew out of an increased attention to suprasegmental phonology (sound patterns larger than a segment or two—like, say, the upturned question intonation in English), and such phenomena as vowel harmony and tone. Influential earlier work includes Williams 1976 [1971] and Leben 1973. While it soon came to replace the flat representations of the Chomsky-Halle program, it is important to note that Autosegmental Phonology was not intended as a replacement for generative phonology at all. Rather, it was a proposed enrichment sparked by the need to handle phenomena which gave generative phonology trouble; Halle was Goldsmith’s supervisor, Chomsky was on the committee, and as one would expect from their reputations as kind and committed mentors, both come in for hearty praise and gratitude in Goldsmith’s acknowledgments.

13.Though not everything, apparently, can be let go. In one paper, he is careful to remove some alleged priority assigned to Lakoff (2007b:367n19).

14.There’s more. Pinker responds to Lakoff, Lakoff to Pinker, reproduced variously across the internet; the exchange, oddly reworked (and imperfectly edited) as a set of letters, was then published again, in Public Policy Research (Pinker and Lakoff 2007). Among the weirdest moments is when Pinker, whose two entries in the exchange heap insults on Lakoff, accuses him of “a number of ad hominem speculations about what is wrong with me.” For an overview of the exchange, the first two decidedly tilted against Lakoff, see “Chris” 2006, Lehrer 2006, and Harryman 2006; the aggregate comments on all of them batter both Pinker and Lakoff, laud both Lakoff and Pinker. A much more measured account by a linguist of Lakoff’s approach to political marketing is Nunberg (2006).

15.Newmeyer replies to this frequent charge, specifically citing Huck & Goldsmith’s version, not so much by refuting the hypocritical-appropriation charge as by mitigating it: “The fact that a core hypothesis of generative semantics [i.e., Quantifier Lowering] . . . turned up in a rather different form as an auxiliary hypothesis of GB [is just a function of the fact that] generative semantics and its rivals share parentage and a host of common assumptions” (Newmeyer 1996:132). A common heritage would seem to make the citational oversights (which Newmeyer does denounce) all the more flagrantly discourteous, however, and this charge is largely about discourtesy. On a more general note, salvaging is a widely established practice in science. When two programs clash, the victorious one frequently, and often covertly, incorporates solutions, data, and methods from the defeated one. Even in relatively uncontentious circumstances, proposals are up for grabs in science. See Pullum 1991 [1983]:14, for instance, on the rapid coöpting of material from Relational Grammar in the 1970s.

16.Aside from the marked ellipsis, I have made some small alterations to the quotation—eliminating two footnotes and changing the Postal citation to match my bibliography.

17.See also Bach (1977:140–41), Ruwet (1991:xxi–xxii), Huck & Goldsmith (1995, 1998), and Gazdar’s wonderfully snide “As has often been remarked, the ‘Aspects’ view of language leads inexorably to Generative Semantics. What is surprising about REST [the Revised Extended Standard Theory] is that it has taken Chomsky so long to get there” (1982:472).

18.G. Lakoff had expressed these principles a year earlier as characteristic of Generative Semantics (G. Lakoff 1989:55, 59).

19.See Langacker 1987:217–18, from which the quotation comes. Despite the motional implications of the word, a trajector can be part of a static relation, as well as a dynamic relation.

20.This one is slightly trickier than I am making out, and if you are attentive, you’ve noticed that the TR and LM have changed places from the ICM for Sentence 7. That’s because Sentence 16 is a passive. It is about the song; was overlooked serves as the ground that situates the song.

21.My treatment of these examples diverges from Brugman’s and Lakoff’s in some respects, most significantly in the analogic data. The ICMs of Figure 9.4, as well, take some poetic license for cohesiveness. There is no reason that the lunch of 18, for instance, or the song in 15 or 16, should resemble a hill. In 15 and 18, a plane would be more neutral, so long as passage over topography is indicated. In 16 a simple geometric form, like a box, would be a more reasonable representation, since the song is reified as some encapsulated entity, not as part of some continuous topography.

23.Fillmore’s work, from Case Grammar on, had been influential computationally, so when Fillmore officially retired from Berkeley in 1995, he moved to the International Computer Science Institute (ICSI), to develop lexicons that are both reflective of human usage and machine readable, the FrameNet project. If you want to see representations of Semantic Frames, that’s where to go (framenet.icsi.berkeley.edu).

Chapter 10

1.Newmeyer (1980a:206–7) catalogs this debris somewhat for the first shifts, and Wasow (1985) makes some similar remarks. Today such a catalog would be very much longer.

2.The “Biolinguistics Manifesto,” which inaugurates the journal Biolinguistics, for instance, begins “Exactly fifty years ago Noam Chomsky published Syntactic Structures (Chomsky 1957), a slim volume that conveyed some essential results of his then-unpublished Logical Structure of Linguistic Theory (Chomsky 1955/1975)”; anchoring all things biolinguistic in the originary generative texts. See also Di Sciullo and Jenkins (2016) for a particularly exemplary representation of trademarking, which starts with the statement “Biolinguistics is the study of the biology of language,” in tones that suggest it will chart a new field for its readers, but then becomes all Chomsky all the time (over fifty invocations or citations, to nearly twenty articles). Derek Bickerton, who has written far more of direct concern to the evolution of language, over a far longer period, is introduced just to be set aside. James Hurford, who has several important books on the evolution of language, is mentioned just to be set aside. Pinker and Bloom (1990), widely regarded as having initiated the recent boom in evolutionary linguistics, is unmentioned. Talmy Givón, who wrote a book entitled Bio-Linguistics, is unmentioned. Jackendoff, who has some cogent counterarguments to Chomsky’s evolutionary theory, is introduced to be refuted. The article ends with a quotation from Chomsky. I don’t want to suggest that the 800-pound gorilla should not get a prominent place in almost any article in linguistics, but it is indisputable that the article is almost exclusively about his work, not about some general field studying the biology of language.

3.The few-Harvard-graduate-students origin story is a common topos in the Biolinguistics literature (e.g., Jenkins 2000:1; Sperlich 2006:35; Di Sciullo & Boeckx 2011:1; Berwick & Chomsky 2016:94). Descartes usually comes in as well, as the great ancestor of that small Harvard collective. McGilvray’s introduction to the third edition of Cartesian Linguistics, for instance, is obsessed with biology, though biology is barely noticed in the text he is introducing (see especially Chomsky 2009b:4–5); see also Chomsky 2005:5, Berwick & Chomsky 2016:88; and especially Boeckx 2011a, which coins the oxymoronic phrase, Cartesian biolinguistics. As Chomsky himself notes elsewhere, for “Descartes the mind is not part of the biological world” (Chomsky 1980b [1978]: 30).

4.Most of my characterization of Gopnik’s work comes from her more formal paper a year after the Nature letter, with her colleague, Martha Crago (Gopnik & Crago 1991).

5.See Harpaz 2002 for a brief discussion with a nice collection of citations and links to some of the primary materials for the first decade of the exchange. Harpaz strongly favors the broad-pattern interpretation.

7.I don’t mean that the claim Chomsky has long been interested in the biology of language is baseless. There are currents in many of his books that engage biology in various ways, such as his thematic analogy that grammar/I-language is a mental organ on parallel with digestion, the recurrent argument that humans grow languages, not learn them, the strong intimations in the Principles and Parameters program of genetic correlates, and so on. The entire language-instinct position is inherently biological. Nor has Chomsky shied away from biological/evolutionary issues when they have crossed his bow (see especially Piattelli-Palmarini 1980); he has rarely shied away from any issues. But it is only in the 2000s, with papers like the Hauser, Chomsky, & Fitch (2002) article that Chomsky offers specific and concerted claims about biology, in parallel to the way, say, Chomsky & Miller (1957, 1958, 1963), and Miller & Chomsky (1963) made specific and concerted claims about the cognitive implications of Transformational Grammar. The Biolinguistics label must largely be retrofitted on Chomsky’s career before 2000. Or, to use Hauser, Chomsky, & Fitch’s conjugal imagery, while the generative/biolinguistic program began “more than 50 years ago,” it is only in the 2000s that it was “consummated” (2003:1570). Berwick and Chomsky explain the dearth of reference to evolution in twentieth century Chomskyan linguistics by saying that, although “the topic was much discussed from the early 1950s . . . there could be few substantive conclusions, and hence there are few references” (2016:97). In other words, there was a research program but it had nothing to say for forty years.

8.My explication may not have done this diagram full justice, since there are some elements I don’t understand. In particular, I’m unsure why the pen-holding hands and the alphabetic characters are in the Conceptual-Intentional segment of the graphic, rather than in the Sensory-Motor segment, and I don’t know why memory is not more proximal to the Conceptual-Intentional segment.

10.Chomsky seems to be using copy in a loosely generic way here, to indicate all instances of a given word or phrase, since it is not the copy generated by Merge that is deleted. It is the original.

11.Tree 10-1 is familiarly abbreviated, though less so than many of the earlier trees and diagrams, because Merge trees are pretty sparse, minimalist, when the labeling projections are bypassed. Most notable, perhaps, is the lack of depiction for morphological Merge beyond cable-knit: technically the Tree should include subtrees for wear+s, argue+s, and sweater+s.

12.There are scattered exceptions to this in the literature of Construction Grammar which features the most ecumenical research affiliated with Cognitive Linguistics.

13.Chomsky apparently believes there can be no such thing as a protolanguage, despite its wide acceptance among other theorists. Derek Bickerton is the scholar most closely associated with this proposal, and he reports on an exchange with Chomsky after the publication of Hauser, Chomsky, & Fitch (2003) in which Chomsky asserts that it is “a logical truism” that protolanguage must either include Merge or be finite. Bickerton wrote back that there were contemporary systems, namely pidgin languages, that have a concatenation operation (word1 + word2 + word3, etc., with no bound but also no hierarchical structure), not a set-theoretic operation, which Chomsky (promptly, as always) labeled an error (Bickerton 2009:187–88).

14.James Hurford makes just this connection: “Our combinatorial ability is as impressive as it is because we have massive stores of constructions to combine” (2012:537).

15.In Foundations, Jackendoff notes that his project “begins to converge with the Minimalist Program” over the similarities between Merge and his UNIFY PIECES—also “the only rule of grammar” and also a procedure that puts words and “treelets” together. Indeed, the “major difference” between his syntactic model and the MP is the latter’s inclusion of Move, which the MP has since abandoned (Jackendoff 2002:180, 180n11).

16.This anecdote is from the web essay collection assembled for Chomsky’s 90^th birthday, but it seems to have almost disappeared from the web. As we go to press Newmeyer’s essay is only available in a pdf version at boskovic.linguistics.uconn.edu/wp-content/uploads/sites/2801/2020/11/Schiffmann-ed.-2018-Revolutionary-New-Ideas-Appear-Infrequently-18-12-07.pdf (accessed March 5 2021).

18.Just so “his first wife” doesn’t convey an erroneous sense of his home life, or suggest that this remark is from a bitter, sniping, ex-wife: Carol Chomsky was a gifted and accomplished Harvard psycholinguist who passed away at seventy-eight, in 2008, after a marriage of nearly sixty years with Noam, which Hughes (2006 [2001]:94–98) characterizes as an affectionate, intellectual partnership; the remark is in keeping with that portrait. See Marquard (2008) for another tribute to Carol Chomsky. Noam Chomsky remarried in 2014, to Valeria Wasserman.

19.Everett is often called “the sole authority on Pirahã” (or some similar phrase), though his ex-wife Keren Everett, who continues to work with them, is also an expert, and Barkham (2008) mentions “two aging missionaries” who speak the language.

20.References to this article are based on a translation kindly provided by Gloria Arze-Bravo.

21.This is not the case in the political arena, where Chomsky’s supporters frequently address accusations of dishonesty and malice, usually by undermining the accuser in some way. But in academics (linguistics, but also philosophy, psychology, English studies, and so on), the response from Chomsky’s supporters is overwhelmingly just silence. Even a book entitled Chomsky and His Critics (Antony & Hornstein 2003) does not broach the subject of his honesty. Neil Smith, who is among Chomsky’s strongest proponents, is the only one whom I know to have given the accusations any airtime at all (Smith 2004:4), for which he is to be commended.

22.The two interpolations are mine. I have invisibly elided Boden’s internal references, which make no sense in isolation, but she documents both Chomsky’s accusations and the relevant points in her own texts where she either does something Chomsky accuses her of not doing or does something very different from what Chomsky accuses her of doing.

23.I remain flabbergasted at the energy Chomsky could pour into correspondence with someone so marginal (sixty-four pages of single-spaced typing, and I mean typing, with a typewriter, before he got a computer; no cutting and pasting for him)—one letter clocking in at nearly fifty pages on its own—to a recent graduate in a field for which he has very little regard). At Chomsky’s very reasonable request, I agreed not to quote from those letters. They are personal letters in which he is extraordinarily candid, and it was very helpful for me to get such a frank, carefully itemized account of his views. But I can at least offer a reflected view of his comments by quoting from my replies. With respect to one passage that he objected to, I wrote, “The most negative responses I have received [from others about this passage] concern me playing advocate for you here. You apparently take it as an attack; most strangely, you then defend yourself against my charges largely by making precisely the points I have already made in the text, but somehow you have completely missed there. . . . [Contrary to what you state, I] don’t say that you aren’t allowed to incorporate Ross’s work because it is part of GS. I don’t say that Lexicalism was crackpot revisionism. I don’t say that you should never have changed your mind about anything. I don’t say that you wrestled with Lakoff for a microphone. I don’t say that Peters and Ritchie influenced the way you do grammar. I don’t say that you said ‘See how dumb McCawley is.’ I don’t say that ‘Remarks’ never had any impact. I suppose it’s best [to] take these up in turn, below, but I must say that the frequency of such misreadings gives me some sympathy for the people who accuse you of deliberately distorting their arguments. I can’t believe it is deliberate, but it is certainly distortion” (Harris to Chomsky, c. August 10, 1991).

24.It’s a good and interesting book, Decoding Chomsky (Knight 2016), which everyone who is interested in Chomsky’s impact on contemporary culture should read, but I don’t think very much of its deliberate unusability theory. See the exchange in OpenDemocracy (Knight 2018a, Newmeyer 2018, R. A. Harris 2018a, Knight 2018b, Golumbia 2018, with other related articles and sundry commentary).

25.The exchange then, is Searle 2002a, Bromberger 2002, and Chomsky 20002b, with Searle replying to each of them (2000b, 2000c, respectively).

26.These remarks come in an interview with James McGilvray, who says, “Chomsky is not claiming in this discussion, and never has claimed, that there is a homunculus” (Chomsky 2012b:258), but does not offer a satisfactory account of what, then, Chomsky is claiming, saying only that people use this kind of folk psychology all the time. McGilvray may be right. He was there, after all, and I wasn’t. Nor has Chomsky made similar claims elsewhere of which I know. But, really, would anyone be surprised if Chomsky meant this literally? Is there anything he might say that would seem outlandish after the career of astonishing assertions and arguments he has advanced?

27.The works to which Chomsky refers us in this quotation as exemplifying early work in which grammaticality and well-formedness were uncharacterized and played no role, Logical Structure and Aspects, appear to support precisely the opposite conclusion; nor does Chomsky give any pages where his claims might be tested. Logical Structure, for instance, includes the following, rather precise, characterization:

Def. 6.

A well-formed statement (WFS) in G is any statement of the form:

a —>b in the env[ironment] c--d,

where a, b, c, and d are WFEs of orders k₁, k₂, k₃, k₄, respectively

(Chomsky 1955a:III–94).

This study . . . will be concerned with the syntactic component of a generative grammar, that is, with the rules that specify the well-formed strings of minimal syntactically functioning units (formatives) and assign structural information of various kinds both to these strings and to strings that deviate from well-formedness in certain respects (1965 [1964]:1).

We might note, too, that neither of these books qualify as informal expositions; Logical Structure, in particular, is the High Formal Tome that stands behind all other early-theory texts; should any relaxation of expression creep into the discussion, one always knew one could turn to it for the rigorous gospel. The words well-formed and grammatical suffuse the literature of Transformational Generative Grammar, where they are effectively synonymous, and the precise characterization of grammaticality was one of Chomsky’s most prominent calling cards, resonating with everybody to whom the appeal of science and precision was compelling, but especially to the computer scientists and mathematical psychologists who propelled the cognitive revolution.

28.Technically, the answer is prolepsis, the rhetorical move one makes in anticipating a counterargument and addressing it before it is voiced. He wouldn’t have to look very far for someone who might offer such a counterargument, however. As Pullum points out, someone with the name Chomsky in 1966 berated another linguist (R. M. W. Dixon) for his rejection of grammaticality (Pullum 1996:139; see Chomsky 1966b [1964]:32n7).

29.He comes close in a footnote, saying that Aspects may have been hasty in “implicitly reject[ing]” a quasi-lexicalist account of nominalizations, but the “transformationalist position” is identified in varying degrees with Lees, Lakoff, and Chapin, rather than with Aspects. Chapin may be notable here with respect to private meanings. One of Chomsky’s objections to my account of “Remarks” is that I deliberately ignored how much more he associated Chapin with transformationalism than Lakoff. There is always a chance he did so in the lectures, I suppose, but the paper barely mentions Chapin. Its argument is with Lakoff.

30.Chomsky’s remarks are on video, as part of panel entitled The Golden Age: A Look at the Original Roots of Artificial Intelligence, Cognitive Science, and Neuroscience (http://mit150.mit.edu/symposia/brains-minds-machines.html). The video is worth watching, if for no other reason than to see Marvin Minsky’s contemptuous disinterest as Chomsky speaks (fidgeting, looking away, examining his own fingernails in an exaggerated display of boredom; the question was also asked of him, but Chomsky’s reply consumes all the time and he never gets to answer). But there is also a transcript of the relevant portion (http://languagelog.ldc.upenn.edu/myl/PinkerChomskyMIT.html). Norvig’s blog (http://norvig.com/chomsky.html) is clearly articulated and includes a nice exchange with Barbara Partee about reducing divisions and supporting collaborative research. It is probably worth noting, too, that Norvig comes out of the Berkeley Cognitive Science community (thanking George Lakoff, Paul Kay, and Charles Fillmore in his dissertation; Fillmore was on his committee; he co-published with Lakoff; see Norvig 1987; Norvig & Lakoff 1987).

31.It is curious that twenty-five years remains the metric, as the claim appears at least as early as 2008, with the duration unchanged a decade later. The original source of this topos may be Newmeyer, who says of Chomsky’s program, much earlier yet, that “more has been learned about language in the last twenty years than in the preceding 2,000 years” (1980:250). For Robert De Beaugrande, this claim calls “to mind Firth’s words: ‘to dismiss two thousand years of linguistic study in Asia and Europe’ ‘is just plain stupid’ ” (De Beaugrande 1991:369n2).

32.The same trend is apparent in his political writings—new editions of Manufacturing Consent, 9–11, Turning the Tide, Rogue States, even “The Responsibility of Intellectuals, Redux” (respectively, Herman & Chomsky 2002, Chomsky 2011a, 2015d, 2015e, & 2011b [reprinted in 2016b])—though the impulse here seems to be less about legacy, more about Chomsky’s relentless drive to keep the political topics he values on the radar. The perspective is presentist in these volumes, not historical. His prefaces, afterwords, and new chapters don’t look back for retrospective precedents; rather, they are updates.

33.I have somewhat arbitrarily changed Basic Principle to Basic Property in this quotation, and used Basic Property throughout my discussions of the minimalist program and Biolinguistics. Chomsky—sigh—uses both terms in different writings but the latter is somewhat more recent (used exclusively in Chomsky 2016a and Berwick & Chomsky 2016).

34.McEnery & Hardie 2013, “The history of corpus linguistics,” as one clear example, has more citations to Chomsky for his opposition to corpora than to Firth (an intellectual source) or John Sinclair (the most influential practitioner), with an entire section dedicated to “The Chomskyan rejection of corpus data” (McEnery & Hardie 2013:732–34).

35.There are many good accounts of the Copernican revolution. This version is an epitome of Kuhn 1957. Danielson and Graney 2014 have a particularly nice account of Brahe’s role.

36.See Whitaker & Jarema 2017 for some speculation about the cause of the rift and some charting of the bitterness.