Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, LINGUISTICS ( (c) Oxford University Press USA, 2019. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 17 September 2019

American Structuralism

Summary and Keywords

American structuralism is a label attached to a heterogeneous but distinctive style of language scholarship practiced in the United States, the heyday of which extended from around 1920 until the late 1950s. There is certainly diversity in the interests and intellectual stances of American structuralists. Nevertheless, some minimum common denominators stand out. American structuralists valued synchronic linguistic analysis, independent of—but not to the exclusion of—study of a language’s development over time; they looked for, and tried to articulate, systematic patterns in language data, attending in particular to the sound properties of language and to morphophonology; they identified their work as part of a science of language, rather than as philology or as a facet of literary studies, anthropology, or the study of particular languages. Some American structuralists tried to establish the identity or difference of linguistic units by studying their distribution with respect to other units, rather than by relying on identity or difference of meaning. Some (but not all) American structuralists avoided cross-linguistic generalizations, perceiving them as a threat to the hard-won notion of the integrity of individual languages; some (but not all) avoided attributing patterns they discovered in particular languages to cultural or psychological proclivities of speakers. A considerable amount of American structuralist research focused on indigenous languages of the Americas. One outstanding shared achievement of the group was the institutionalization of linguistics as an autonomous discipline in the United States, materialized by the founding of the Linguistic Society of America in 1924.

This composite picture of American structuralists needs to be balanced by recognition of their diversity. One important distinction is between the goals and orientations of foundational figures: Franz Boas (1858–1942), Edward Sapir (1884–1939), and Leonard Bloomfield (1887–1949). The influence of Boas, Sapir, and Bloomfield was strongly felt by the next generation of language scholars, who went on to appropriate, expand, modify, or otherwise retouch their ideas to produce what is called post-Bloomfieldian linguistics. Post-Bloomfieldian linguistics displays its own internal diversity, but still has enough coherence to put into relief the work of other language scholars who were close contemporaries to the post-Bloomfieldians, but who in various ways and for various reasons departed from them. American structuralism has at least this much heterogeneity.

This article illustrates the character of American structuralism in the first half of the 20th century. Analysis of a corpus of presidential addresses presented to the Linguistic Society of America by key American structuralists grounds the discussion, and provides a microcosm within which to observe some of its most salient features: both the shared preoccupations of American structuralists and evidence of the contributions of individual scholars to a significant collaborative project in the history of linguistics.

Keywords: mid-20th-century linguistics in the United States, American structuralism, descriptive linguistics, Linguistic Society of America

1. Diversity and Convergence in American Structuralism

This survey comprises three sub-sections, depicting, in order, the work of three groups of contributors to 20th-century structuralist linguistics in the United States, up to around 1960. First, there are three pioneering scholars: Franz Boas, Edward Sapir, and Leonard Bloomfield. Each of them differed in significant ways from the other two, but all three helped set the tone for the emergence of a version of structuralism native to the United States. Second are the so-called post-Bloomfieldian linguists who followed Boas, Sapir, and Bloomfield chronologically, but not always intellectually, as they built up a body of practices, assumptions, and goals that comprised American structuralism. Third are the language scholars whose work is recognizably structuralist, but who orbited outside post-Bloomfieldian linguistics without aspiring to consistency among themselves. To call them outliers exaggerates their positions, but will nevertheless serve as a label.

All such classifications of scholars into subgroups are artificial and not fully satisfactory. But despite their differences, there is a certain coherence to American structuralism, in part because of the centripetal institutional force of the Linguistic Society of America (LSA), through which American language scholars regularly came into contact with each other to challenge, debate, and exchange ideas about language (Gleason, 1988; Hill, 1991; Joos, 1986). Some American structuralists devoted a significant amount of professional energy to developing the LSA; others, less. But from the mid 1920s, the LSA and LSA-related institutions—the journal Language (Bloomfield, 1946, pp. 1–2), the summertime Linguistic Institutes (Falk, 2014; Hill, 1964; Murray, 1991), and (more loosely connected with the LSA) the Intensive Language Program (Cowan, 1991; Hall, 1991) and the linguistic atlas project (Joos, 1986, pp. 52–53, but cf. Murray, 1991)—were focal points for building up a sense that American structuralists belonged to an intellectual community, a community that variously extended to members of the outlier group.

1.1 Presidential Addresses to the LSA as a Record of American Structuralism

This exposition takes advantage of the central importance of the Linguistic Society of America to the development of American structuralism in first half of the 20th century, during which interval scholars from all three subgroups of American structuralists served as presidents of the LSA. The membership of the society elects a new president every year. The role is largely ceremonial, since the secretary in consultation with the executive committee makes most administrative and logistical decisions. However, a prominent responsibility for the outgoing president of the LSA is to present a plenary address at the annual meeting. The texts of many of the presidential addresses have been published, and serve as a corpus depicting the lifespan of American structuralism. As oral communications (later converted into written texts), they share the same rhetorical architecture: in a celebratory after-dinner setting at the climax of the Society’s annual meeting, a speaker who has been selected as the group’s leader by his or her peers assumes responsibility for a stimulating formal address. In published format, their average length is 14 pages, with a few exceptions on each side. The speaker’s choice of topic is apparently unconstrained, shaped only by the natural, sometimes competing, forces to build group solidarity while at the same time introducing enough novelty to set the speaker’s own idiosyncratic stamp on the exercise. In a first-person chronicle of the LSA, Hill (1991) mentioned, at least in passing, almost all of the presidential addresses presented between 1950 and 1968. He remarked that most of them “followed the usual practice of being working papers, subject to discussion” (p. 65), as opposed to (for example) personal memoirs or purely inspirational manifestos. Two presidents who opted out of this convention felt it necessary to call attention to that fact. Marckwardt (1964, p. 26) declared that his talk would depart from “a practice of long standing in the Linguistic Society, namely the presentation of what has generally been essentially a research paper as the presidential address.” It is significant that Marckwardt’s address was published in the LSA Bulletin, rather than in Language, as was usual. Twaddell (1972, p. 221) began his 1957 address—published belatedly in American Speech—by downplaying its importance as a “little intaglio” designed to balance the “thirty-nine substantial papers” presented elsewhere at the meeting.1

Analysis of the content of those LSA presidential addresses that may be attributed to American structuralists anchors this survey and illustrates what these scholars had in common, how they differed from each other, and what they contributed to the emergence of a distinctively American language science. Table 1 provides information about LSA presidential addresses, from its first meeting in 1925 until 1976.

Table 1. Presidential Addresses Presented at the Annual Meeting of the Linguistic Society of America, 1926–1976


LSA President

Title of Address

Publication of Address


Herman Collitz

“World languages”

Language, 2(1), 1–13 (1926)


Maurice Bloomfield

[President absent from Annual Meeting]



Carl D. Buck

“A project for a dictionary of selected synonyms in the principal Indo-European languages: A contribution to the history of ideas”

Likely related to Dictionary of selected synonyms in the principal Indo–European languages, Buck (1949)


Franz Boas

Classification of American Indian languages

Language, 5(1), 1–7 (1929)


Charles H. Grandgent

“Phonetic coincidences” [President absent; address “read by title only”]



Eduard Prokosch

“The limits of grammatical change in the Germanic strong verb”

[Text not located]


Edgar H. Sturtevant

“The ablative in Indo–European and Hittite”

Language, 8(1), 1–10 (1932)


George Melville Bolling

“On the dual in Homer: Homeric variants cited from Zenodotus and Aristophanes”

Language, 9(4), 298–308 (1933)


Edward Sapir

Consonantal symbolism in language

Likely related to Sapir, (1929), “A study in phonetic symbolism,” Journal of Experimental Psychology, 12(3), 225–239 or “Symbolism” (1934)


Franklin Edgerton

“Hybrid Sanskrit of the Buddhists”

Language, 13(2), 107–122 (1937)

Published as “Gerunds in Buddhist hybrid Sanskrit”


Leonard Bloomfield

Language or ideas?”

Language, 12(2), 89–95 (1936)


George T. Flom

“Breaking in Old Norse and Old English”

Language, 13(2), 123–136 (1937)


Carl D. Buck

“Some words for emotions and moral or aesthetic notions”

Likely an expansion of Buck (1949), 1084–1196


Louis H. Gray

“The Indo–European base-type *dō- and its extensions *doi̯- and *dou̯-”

The American Journal of Philology, 62(4), 476–484 (1941)


Charles C. Fries

On the development of the structural use of word–order in modern English

Language, 16(3), 199–208 (1940)


A. L. Kroeber

“Some relations of linguistics and ethnology”

Language, 17(4), 287–291 (1941)


Roland G. Kent

“Retrospect and prospect”

[Text not located]


Hans Kurath

[No Annual Meeting]



Fred N. Robinson

[No Annual Meeting]



Kemp Malone

“Some English etymologies”

Same-title text appears in Word, 26(3), 305–308 (1970); relationship to LSA address unclear


Yuen Ren Chao

The logical structure of Chinese words

Language, 22(1), 4–13 (1946)


E. Adelaide Hahn

“The development of parataxis into hypotaxis”

[Text not located]


Albrecht Goetze

“Akkadian and its dialects”

[Text not located]


Hayward Keniston

“Phonetic evidence and phonological change”

[Text not located]


Murray B. Emeneau

Language and non-linguistic patterns

Language, 26(2), 199–209 (1950)


Einar Haugen

Distribution and identity: An essay on metalinguistics

Language, 27(3), 211–222 (1951)

Published as “Directions in modern linguistics”


Joshua Whatmough

Religio grammatici

[Text not located]


George S. Lane

“The time has come, the walrus said”

[Text not located]


Bernard Bloch

“Theme and variations”

Deliberately unpublished (Hill, 1991, 77)


Charles F. Voegelin

Phonemicizing for dialect study, with reference to Hopi

Language, 32(1), 116–135 (1956)


Zellig Harris

Transformation in linguistic structure

Language, 33(3), 283–340 (1957)

Published as “Co–occurrence and transformation in linguistic structure”


Roman Jakobson

Metalanguage as a linguistic problem

Nyelvtudományi Közlemények, 78, 346–352 (1976)


W. Freeman Twaddell


American Speech, 27(3–4), 221–232 (1972)


Henry Hoenigswald

Some uses of nothing

Language, 35(3), 409–420 (1959)


Harry Hoijer

“Athapaskan consonants”

[Text not located]


George L. Trager

“Linguistics is linguistics”

Studies in Language (1963)

Published as “Occasional paper No. 10”


Kenneth L. Pike

Dimensions of grammatical constructions

Language, 38(3), 221–244 (1962)


Albert H. Marckwardt

“Opportunity and obligation”

Bulletin of the Linguistic Society of America, 37, 26–37 (1964)


Mary R. Haas

“Linguistics and history”

Published in Language, culture, and history: Essays, Haas (1978), 176–194


Charles Hockett

Sound change

Language 41(2), 185–204 (1965)


Yakov Malkiel

“Linguistics as a genetic science”

Language, 43(1), 223–245 (1967)


J Milton Cowan


Unpublished, Hockett (1995), 347


William G. Moulton

Structural dialectology.”

Language, 44(3), 451–466 (1968)


Eugene A. Nida

The science of translation

Language, 45(3), 483–498 (1969)

Published as “Science of translation”


Archibald A. Hill

Laymen, lexicographers, and linguists

Language, 46(2), 245–258 (1970)


Charles A. Ferguson

“Some requirements for a theory of language behavior”

[Text not located]


Eric P. Hamp

“Reconstruction, inheritance, diffusion, and change”

[Text not located]


Dwight L. Bolinger

“Truth is a linguistic question”

Language, 49(3), 539–550 (1973)


Winfred P. Lehmann


Language, 50(4), 622–629 (1974)


Morris Halle

Confessio grammatici

Language, 51(3), 525–535 (1975)


Thomas A. Sebeok

“The pertinence of Peirce to linguistics”

Related text published as Contributions to the doctrine of signs, Sebeok (1976)


Rulon S. Wells

“Suggested meaning”

[Text not located]

Note: To provide a sample of each author’s approach, the title of every address in bold typeface opens a link to a reproduction of the first page of its published version, for those texts that proved retrievable.

Two warnings are in order. First, it is understood that no single text captures the full complexity of a scholar’s thought, much less of its development over time, just as no single photograph captures the full complexity of a person’s appearance. Nevertheless, a series of individual portraits, considered together, may represent the public face of a group, and may indicate something about what makes individual members distinctive. Second, because American structuralism was a composite phenomenon without strict boundaries, it is fruitless to try to dichotomously classify authors or texts as either structuralist or non-structuralist. However, it is still possible to identify which ones among the 30-some retrievable entries in Table 1 instantiated relatively more, versus relatively fewer, features of American structuralism, understanding from the outset that not all readers will agree with the author’s judgments.

Surveying Table 1 with this in mind, one fact stands out. Since the convention has been to elect a president at the apex of his or her career—or somewhat past that point—presidential addresses have tended to recollect the past rather than anticipate the future. Therefore the preponderance of addresses until the mid 1940s seemed to look backward to (pre-structuralist) philological or historical-comparative issues. A parallel lag re-occurs in the 1960s, when presidential addresses took for granted the preoccupations and methods of structuralists, persisting into the interval after American structuralism had passed its prime. Entries in Table 1 that appear in unmarked typeface indicate addresses that either looked back to historical-comparative linguistics, looked forward beyond American structuralism to what followed, or otherwise took up topics belonging to other streams of research. (Also unmarked are texts that could not be located, or that were not actually presented, for example during the war years of 1942 and 1943.) Entries in Table 1 appearing in bold typeface are of greatest interest and are included below in the discussion of American structuralism. Their dates run from the mid-1940s to the mid-1970s, an interval that—due to the conservatism of LSA electoral practices—is lagging relative to the conventionally acknowledged heyday of American structuralism, from 1920 to 1960. To provide a sample of each author’s approach, the title of every address in bold typeface in Table 1 opens a link to a reproduction of the first page of its published version, for those texts that proved retrievable.2

2. “Structuralism” and “American Structuralism”

The first task is difficult: to acquire a beginning sense of what is meant by the term American structuralism. In general, structuralism in linguistics looks back to the distinction, traditionally attributed to Ferdinand de Saussure (1857–1913), between diachronic study of language (which considers that understanding of a language is most accessible through study of its development over time) versus synchronic study of language (which treats languages as coherent, analyzable systems at every stage of their histories). Structuralists followed Saussure in accepting the validity of synchronic analysis, although not all accepted the full scope of Saussure’s other proposals. Various styles of structuralist linguistics emerged in the early 20th century in Europe: in Copenhagen, Hjelmslev’s glossematics; the Firthian “London School”; the Prague School; the Geneva School (see Koerner & Asher, 1995, pp. 221–368).

In the United States at the end of the 1800s, the training and orientation of the dominant American linguist, William Dwight Whitney (1827–1894), prioritized diachronic study of language in the style of pre-structuralists. Despite Whitney’s extensive historical research on Sanskrit, he also wrote about language in a more modern vein, for example in promoting the idea that language is a social institution (rather than a natural phenomenon)—an idea that Saussure famously admired, and that supported synchronic linguistics (Joseph, 2002).

This is the conventional understanding of the referent of ‘structuralism’. It is important to acknowledge, however, that not all parties adopt it. Specifically, some contemporary linguists limit the term to pre-generative linguistic theory (Newmeyer, 1986, pp. 3–15). Writing from this point of view, Noam Chomsky (b. 1928) asserted (inter alia) that grammar resides in speakers’ abstract competence, not their performance; therefore, analysis of observable linguistic data is fundamentally inadequate. This directly challenges the working habits of American structuralists, who took for granted the bedrock value of collecting and classifying language data. Generative linguistics separates structuralism from contemporary generativism, using these terms to mark what it conceives as a sharp break in the history of linguistics in the middle of the 20th century.

Another difficulty resides in the modifier “American” in the expression American structuralism. By convention, the term refers to structuralism as it developed in the United States, putting aside the full referent of “America,” which covers a far larger region. Even accepting a constricted geographical range for the term, there is the correlative difficulty of determining who counts as an American. To take three of the first important figures, Boas was born, educated, and worked extensively outside the United States; Sapir was born in what is now Poland and worked in Canada for 15 years; Bloomfield spent an important, formative, year studying in Germany. Among the post-Bloomfieldians and members of the outlier group, Charles Fries and Archibald Hill alone stand out as having neither been born, nor (at least briefly) educated, nor employed outside the United States. Yet all these scholars can be identified as American structuralists. The expressions American linguistics (Austerlitz, 1975) and the Americanist tradition (Hymes, 1976) add another wrinkle. These terms label research by scholars, regardless of their nationalities or theoretical stances, who study indigenous American languages. This text accepts imprecision on this score, treating language scholars with structuralist orientations, but with diverse personal affiliations to the United States and interests in diverse languages, as equally eligible for the status of American structuralist.

A more intractable difficulty is identifying scholars whose intellectual orientations do or don’t meet the minimum definition of American structuralism. William Haas (1978, p. 294) insists that “Modern linguistics is structural linguistics” (emphasis in the original), so that, although generative linguistics represents itself has having replaced structuralism, according to Haas, what Chomskyan generative grammar abandoned was “only a particular version of structural linguistics.” In contrast, Anderson (1985, p. 278) restricts American structuralism to the post-Bloomfieldians; Murray (1994, p. 57) excludes Boas but includes Sapir and Bloomfield; Joseph (2002, p. 166) considers American structuralism to have been initiated in 1942 by Roman Jakobson. Amid the complexity of these overlapping distinctions, it is reassuring that, regardless of how historiographers divide individuals into teams, they all seem to work from the same roster of players.

A final difficulty with the term American structuralism is that it competes in semantic space with at least two other terms. Some American structuralists—notably Bloomfield (e.g., 1925, pp. 117–119)—favored the term descriptive linguistics as a characterization of his approach. Bolinger (1968, pp. 189–207) called Bloomfield, the post-Bloomfieldians, and the Prague School structuralists, and opposed them on the one hand to the descriptivists Boas and Sapir, and the missionary linguists (including Kenneth Pike), and on the other hand to generative grammarians, who practice formal linguistics. Lane (1945) called the work of all these scholars descriptive linguistics, distinguishing what he saw as their shared approach from historical linguistics. Like Bolinger, Hymes and Fought (1975/1981, pp. 8–11) contrasted descriptive, structural, and formal linguistics. But Hymes and Fought find much overlap across them, with all three differing from historical and prescriptivist linguistics.

There is also an additional term, distributionalist, associated in particular with the post-Bloomfieldians. Distributionalism names one of the key assertions of post-Bloomfieldians: that the patterns of occurrence of comparable linguistic units can adequately define the relationships of those units to each other, without reference to divergence or convergence in meaning (e.g., Harris, 1951). Distributionalists imposed “the austere pressure of semantic abstinence” (W. Haas, 1978, p. 295) on language science so as to expel what they viewed as uncontrollable and distorting subjectivity. However, not all structuralists were descriptivists, and not all descriptivists were distributionalists, so that the scope of the last group is narrower. In this text, the term American structuralist refers broadly to scholarship carried out by U.S.-affiliated linguists in the first half of the 20th century, who presupposed the soundness of a synchronic approach to language; who collected language data, then sought out systematic patterns in it; and who considered their work to contribute to a science of language. It was characteristic, but not distinctive, of American structuralists to be members of the LSA, to participate in LSA-sponsored activities, and to read and publish in the LSA’s journal Language—some more avidly than others.

3. Boas, Sapir, Bloomfield

3.1 Franz Boas

Franz Boas (1858–1942) was born in Germany. He traveled to Baffin Island in what is now the Canadian territory of Nunavut in 1883 with a doctorate in physics, an interest in geography, and an assignment to study the relationship between an extreme climate and the culture of peoples living there. Boas soon became absorbed in studying the languages and society of the Inuit people. From 1886 onward, he lived in the United States and continued his ethnographic work between trips to the Pacific Northwest, where he elicited language data from speakers of the Wakashan, Chemakum, Sahaptian, Salishan, Chinookan, and Siouan languages, among others. Boas held positions briefly at the American Museum of Natural History in New York, and for years on the faculty at Columbia University. He is often depicted as the founder of American anthropological linguistics, although as many have pointed out (e.g., Silverstein, 2015, p. 84; Stocking, 1974, p. 455), he had no formal training in linguistics; moreover, this view downplays his disciplinary ancestors in the United States (Darnell, 1998). Even if Boas was more of a predecessor to American structuralism than a founder, few dispute the historical importance of his study of Native American languages, or his role in establishing some of the key presuppositions and methods of American structuralism. Boas was also an influential figure in the professionalization of the study of Native American languages through his cooperation with the Bureau of American Ethnology (later absorbed into the Smithsonian Institution), his museum work, his leadership roles in various professional societies and journals, and above all through the dispersal of his Columbia University graduate students throughout the United States, from among whom he winnowed for close mentorship those who met his high standards (Stocking, 1974).

Many of the hallmarks of Boas’s work that most influenced the development of American structuralism are evident in his 1928 LSA presidential address. For example, Boas famously insisted that study of Native American languages required direct face-to-face fieldwork in which a core goal was to collect oral narratives, those of high cultural value (myths, folk tales) as well as informally communicated ethnographic or autobiographical narratives. He transcribed those narratives, translated the transcriptions word-by-word, and scrutinized them with the same care that philologists directed to ancient Indo-European scriptures. In this way, Boas implicitly rejected the exclusion of unwritten languages as “primitive” and therefore outside the scope of linguistic research. Boas’s presidential address displays this orientation. Basing his claims on examples from genealogically unrelated Pacific Northwest and Plains Indian languages, the address argues for hybridization (1929, p. 2) at the level of morphology: that is, that grammatical processes—not only lexical items or phonetic features—can diffuse across unrelated languages when speakers are in close contact with each other. Boas closed by citing analogous evidence in African, New Guinean, and Melanesian languages, and by calling for this insight to be turned back onto study of the European languages.

American structuralists accepted and built on Boas’s treatment of languages without literary traditions as equally rich and worthy of analysis as the Indo-European languages of scholarship, a stance labeled linguistic relativism (Haas, 1976, p. 60). Boasian linguistic relativism became a cornerstone of American structuralism. In a classic essay, Boas (1911) set forth this principle and, moreover, showed that his interest in Native American languages was not restricted to their sound properties, but rather extended to semantic categories and grammatical processes as well. The 1928 presidential address reprised both these themes. Above all it demonstrated Boas’s commitment to linguistic relativism in that he resisted identifying the categories and features of Native American languages with those of the familiar Indo-European languages. Instead, he showed how the corpus of texts he had elicited from untutored speakers of diverse Native American languages could enlarge our understanding of language contact and language change. Boas’s working style and his respect for the integrity of all human languages had a lasting influence on American structuralism.

3.2 Edward Sapir

Edward Sapir (1884–1939) was a student of Boas at Columbia University. He inherited many but not all of his mentor’s methods and preoccupations, and he surpassed Boas in linguistic expertise. Sapir’s first degree was in Germanic languages from Columbia, where he acquired the techniques of comparative-historical linguistics. But after taking a course with Boas, he “was sidetracked . . . into the more compelling and urgent task of recording American Indian languages” (Darnell, 1990, p. 130). Sapir traveled to Oregon for his first fieldwork experience, on Takelma, but eventually studied numerous indigenous languages in western and southwestern areas of the United States. He had unusually keen analytic abilities and a sensitive ear for the sound features of languages. He also was a famously engaging writer. He wrote a classic textbook introduction to linguistics, Language (Sapir, 1921), that has retained its freshness and clarity of expression for almost a century. Sapir directed the anthropology division of the Canadian National Museum in Ottawa for 15 years and taught at the University of Chicago and Yale before dying at the age of 55. Near the end of his career, his attention turned back to Indo-European philology and to Hebrew, so that he produced a few papers in those fields. Sapir’s wide humanistic interests also extended to psychology and culture, personality, psychoanalysis, poetry, and music.

Sapir’s influence on the study of language matched the breadth of his interests. He elicited and transcribed oral texts in the Boasian fashion, then applied to the transcriptions the analytic techniques that philologists developed in working with ancient Indo-European materials. He aimed for a much fuller linguistic analysis than was the norm among Boas’s other students, or even Boas himself. Sapir also overshot Boas in boldly reconstructing genetic relationships among Native American languages. Some but not all of his proposals have stood the test of time. Sapir’s student, Benjamin Lee Whorf (1897–1941), developed his interest in language and culture to articulate what is now sometimes called the “Sapir-Whorf hypothesis,” essentially a late 1930s version of long-standing beliefs that language plays a role in conditioning speakers’ perceptions and cultural practices (Joseph, 2002, pp. 71–105). Sapir’s writings often commented on the intuitions of individual speakers about their own language and probed their conscious and unconscious linguistic subjectivity. He mined that stream of data for access to a language’s structural patterns. Silverstein (1986, p. 83) linked this feature of Sapir’s “cosmographical-historicist linguistics” to a Boasian drive to understand language and culture from a native speaker’s perspective, although here as in other domains, Sapir seems to start with Boas then surpass him.

Unfortunately, Sapir’s 1933 LSA presidential address, “Consonantal Symbolism in Language,” has apparently not been published in a form recognizable from its title. Weinstein (1986, p. 492) cited the lecture notes of New Zealand ethnologist Ernest Beaglehole (then a post-doctoral student), which indicate that Sapir was teaching about symbolism at Yale in 1933. There are two texts that preserve some of his thinking on the topic, Sapir (1929) and Sapir (1934); however, it is unclear what relationship to the presidential address either text might have had. Sapir (1934) is an encyclopedia entry for “Symbolism.” Weinstein (1986) discusses this text and Sapir’s several efforts to define types of symbolism in language and culture, in their conscious and unconscious aspects. All this seems rather far from the concerns of American structuralists, and hence, unlikely material for a presidential address to the LSA.

The second text, which may or may not be related to Sapir’s address to the LSA, is a 1929 journal article, “A Study in Phonetic Symbolism.” Published four years before his presidency, Sapir introduced it as a preliminary report of research still ongoing. The article recounts an empirical study of the intuitions of 500 Chicago residents aged 11 through adulthood, plus 7 speakers of Chinese. The participants were presented, in writing, with contrasting pairs of single English vowels or consonants, including pairs of vowels contrasting in height or degree of roundedness, and pairs of consonants differing in voice or manner of articulation. In response to a series of oral stimuli containing the paired contrasts, participants marked which member of the pair they thought labeled a physically larger versus smaller referent. In a follow-up experiment with only adults, Sapir employed nonce words, that is, novel words invented for the purpose of the research itself. He assigned arbitrary meanings to half of them, asking participants to guess the meanings of phonetically related invented words: “If mīla means ‘brook’,” what does mala mean?” Speakers identified /i/ with smaller and /a/ with larger referents, so that if mīla meant ‘brook’, mala meant ‘water running through a ravine’ (p. 237) or “|‘a lake’, or ‘a lake . . . [with] greater depth of color’|” (p. 238). He concluded that “English-speaking society does, for some reason or another, feel that of these two vowels, a, by and large, is possessed of greater potential magnitude symbolism than the contrasted vowel i” (p. 231). Sapir reported that this generalization did not, however, predict speakers’ intuitions about a versus e (i.e., /ɛ/), which finding he attributed to the fact that /ɛ/ is not a vowel native to English, and therefore speakers tend to project a different identification onto it (p. 234). Sapir therefore found in these data evidence that speakers have a “definite symbolic feeling-significance that seemed to have little relation to the associative values of actual words” (p. 228).

Unfortunately, the 1929 article does not report Sapir’s experimental results with consonants. We do not know whether Sapir’s presidential address in 1933, on consonantal symbolism, recapitulated, elaborated on, or even had any connection with these findings. If it did extend the 1929 article, Sapir would have displayed to his colleagues many of his trademark interests: the search for patterns in language, in this instance, sound patterns; his meticulous collection of data and his imaginative but disciplined interpretation of them; and his respect for individual speakers’ intuitions, and for a speech community’s unconscious shared “feeling” for the complexion of their language.

3.3 Leonard Bloomfield

Leonard Bloomfield (1887–1949) and Sapir were near-contemporaries. Both men started out with training in Germanic and Indo-European philology; both carried out fieldwork on unwritten Native American languages (albeit not exclusively) and applied comparative-historical techniques to their analyses; both went on to teach at the University of Chicago and then at Yale; both served as presidents of the LSA. Bloomfield and Sapir both took for granted the value of synchronic analysis and sought out a language’s phonological and morphological patterns. However, Bloomfield’s and Sapir’s contributions to American structuralism contrast sharply.

Bloomfield went to Germany in 1913 for post-graduate training. On his return, he moved from the University of Illinois, to Ohio State, to Chicago, and finally (after Sapir’s death) to Yale. From his direct exposure to the Neogrammarians, he absorbed the principle of the exceptionlessness of phonetic laws (Fries, 1961). From the Boasian tradition (without having been a direct student of Boas), Bloomfield embraced linguistic relativism, the value of collecting oral narratives in a fieldwork setting, and an aversion to a priori cross-linguistic generalizations about linguistic categories and features. In addition to his work on Indo-European historical-comparative linguistics (on which topic his publications cluster in the first half of his career), the languages Bloomfield is most associated with are Algonquian, especially Menominee and Fox, and Tagalog. Bloomfield paid close attention to epistemological questions: he insisted on identifying linguistics as a science of language (Bloomfield, 1925, 1926, 1930, 1935), and from his first major review of the field (Bloomfield, 1914/1983, p. v) onward, pointedly referred to the discipline as “the scientific study of language” or “our science.” Identifying linguistics as a science meant, to Bloomfield, that claims made about language must be empirically verifiable and constructed out of a fully explicit tissue of definitions (e.g., of ‘utterance’; ‘bound form’; ‘zero element’) and assumptions (e.g., “In a construction, a phoneme may alternate with another phoneme according to accompanying phonemes”; “Among persons, linguistic change is uniform in ratio with the amount of communication between them”) (1926). Securing recognition for linguistics as a science was also a motif of Bloomfield’s institutional activities, including his leadership in founding the LSA, through which he hoped to establish an independent identity for the field separate from anthropology, philology, or literary studies.

The sojourn at Ohio State was important to Bloomfield’s intellectual history because there he met the behavioral psychologist Alfred P. Weiss (1879–1931). Weiss influenced him to try to develop a conceptualization of human language as observable verbal behavior, without relying on mentalistic notions such as ‘mind’, ‘thought’, ‘will’, ‘image’, ‘consciousness’, and so forth. This challenge attracted Bloomfield because, although the independence of linguistics meant to him that linguistics could not privilege any specific school of psychology, he felt that linguistics based on observable data rather than mental constructs reinforced the scientific basis of the field. Bloomfield’s investment in behaviorism has probably been exaggerated and even distorted (Hymes & Fought, 1975/1981, pp. 103–111). Nevertheless, the austerely ‘mechanistic’ nature of Bloomfield’s descriptive linguistics (to use a term to which his approach aspired [1933, pp. 142–144]) throws into relief Sapir’s descriptive linguistics, which passed freely from similarly austere observation and recording of linguistic data, into analysis of those data with respect to the speakers’ culture and their feeling for their language.

Bloomfield’s 1935 presidential address to the LSA was entitled “Language or Ideas?” (1936). It presented a succinct argument for depicting language in mechanistic, anti-mentalist, terms, which he located as an underdog position relative to the popular terminology of “mentalism and animism” (p. 89). Bloomfield called his position the “hypothesis of physicalism” (p. 93), which he felt was eventually bound to succeed, but which in the meantime had to face pervasive misrepresentation and prejudice. Bloomfield writes approvingly of the Vienna Circle’s logical positivism, since (in his view) its members had independently arrived at the same physicalism as had Weiss, replacing “traditional but useless and confusing” (p. 94) mentalistic terms with descriptions of observable behavior (see also Hiż & Swiggers, 1990). It is significant that Bloomfield’s illustrations of such successful replacement come from the histories of physics and geometry.

There is no direct reference in Bloomfield’s presidential address to his fieldwork with Native American languages, to his extensive bibliography in descriptive linguistics, or to his historical-comparative work. At face value, “Language or Ideas?” captures one important facet of Bloomfield’s contribution to American structuralism, his promotion of the hypothesis of physicalism. Implicit in the address may have been Bloomfield’s most important contribution, namely his investment in creating a science of language in which it would make sense (for example) to develop a way of analyzing the object of interest—human language—on the model of physics or mathematics. Joos (1966, p. 31) characterized Bloomfield as having “infected people with a high and serious respect for the evidence and for responsible statement about it.” In “Language or Ideas?” Bloomfield displayed what it meant to him to make responsible statement[s] about evidence within a frankly mechanistic linguistics, a trait that he valued more than any particular psychological theory.

4. The “Post-Bloomfieldians”

The heterogeneous influences of Boas, Sapir, and Bloomfield on the next generation of linguists had in common the shared commitment of these three thinkers to the synchronic description of languages, even languages without literary traditions (without extinguishing the value that each scholar placed on diachronic studies). In this sense, Boas, Sapir, and Bloomfield can be identified as American structuralists. All three aimed to articulate the systematic patterning of sounds and forms in languages. Boas set a tableau with analysis of Native American languages at the center, and elicited data to which he applied his text-based methods. Among other accomplishments, Sapir and Bloomfield elaborated on different facets of Boas’s tableau and raised its level of linguistic sophistication (Voegelin & Voegelin, 1963).

Sapir’s students recall him as a charismatic and supportive teacher. He influenced a generation of linguists in the United States. Despite his short life, seven of his direct students went on the serve as presidents of the LSA: Emeneau (President in 1949), Voegelin (in 1954), Hoijer (in 1959), Trager (in 1960), Pike (in 1961), Haas (in 1963), and Hockett (in 1964). Sapir’s students also achieved prominence in linguistics or anthropology according to other measures: Benjamin Lee Whorf, Morris Swadesh, Stanley Newman, Fang-Kuei Li. In contrast, Bloomfield trained few doctoral students (Bloch, 1949, p. 91). His greatest successes as a teacher probably took place at the LSA’s Linguistic Institutes in 1938, 1939, and 1940, when his presence was felt strongly—especially in 1938, when Bloomfield’s lecture series was apparently attended by almost the entire enrollment of the Institute (Report, 1940, p. 87). However, he did not have much verve or stage presence. Rather, Bloomfield’s influence was most felt through his writings, including his professional correspondence, and the reviews he wrote of publications by other scholars—and, indirectly, through the fruits of his institution-building activities (Murray, 1991).

4.1 Who Were the “Post-Bloomfieldians”?

Although Sapir had a higher profile as a teacher, a significant portion of American linguists from the 1930s through the 1950s whose work is recognizably structuralist adopted, or have been labeled with the term ‘post-Bloomfieldians’ (sometimes ‘neo-Bloomfieldians’ or simply ‘Bloomfieldians’). The terms post-Sapirian and neo-Sapirian also appear, often with reference to Sapir’s role in the development of the Sapir-Whorf hypothesis (e.g., Hill & Mannheim, 1992), but not always (e.g., Malkiel, 1982, on post-Sapirian notions of ‘drift’; Schwegler, 1990 on neo-Sapirian language typology). But American structuralism advanced much more fully under Bloomfield’s name rather than Sapir’s. Among reasons for this fact are Sapir’s early death; his students’ geographical dispersion; and the expansion of linguists’ attention beyond Indo-European and Native American languages (Murray, 1994, pp. 175–176). Perhaps most of all, Bloomfield (1926, 1933), not Sapir, provided attractive and powerful tools with which linguists could clarify what their responsibilities were and how to go about carrying them out. In the words of Voegelin and Voegelin (1963, p. 20), members of the LSA “began to realize that Bloomfield had given them a wholly American and wholly explicit linguistic theory.”

A lot has been written about the propriety of the terms (post-/neo-) Bloomfieldian. Murray’s (1994, pp. 113–134) chapter title sums up much of the discussion by asking, point-blank, “Was Bloomfield a Bloomfieldian?” A consensus seems to be that, unsurprisingly, not all of Bloomfield’s linguistics work was carried forward intact; moreover, some parts of it either were distorted in the course of its transmission to a new generation, or simply were developed in directions that Bloomfield might not have found congenial.

Various taxonomies of post-Bloomfieldians have been proposed. Koerner (2002) separated them into two cohorts based on the extent to which they were willing to admit reference to meaning into linguistic analyses (see Section 4.2.2). A core group “tried to minimize the reference to semantic criteria” (p. 88): George L. Trager (1906–1992), Bernard Bloch (1907–1965), Zellig S. Harris (1909–1992), and Henry Lee Smith, Jr. (1913–1972). Murray (1994, p. 158) would add to this group Martin Joos (1907–1978) and (unexpectedly) Rulon S. Wells (1919–2008). Koerner (2002, pp. 82–87), contrasted this cohort with “moderate Bloomfieldians,” who were willing to employ semantic information in their analyses, if sometimes in a hedged manner, including Archibald A. Hill (1902–1992), Robert A. Hall, Jr. (1911–1997), Eugene A. Nida (1914–2011), Charles F. Hockett (1916–2000), and Henry Allan Gleason (1917–2007). Hymes and Fought (1975/1981, p. 129) would add Charles C. Fries (1887–1967). Murray (1994) examined data on the acknowledgement and co-authorship practices of post-Bloomfieldians but did not find a strict separation between these two cohorts. He did, however, add to the total group the names of Carl F. Voegelin (1906–1986) and Kenneth Pike (1912–2000). One might further widen the circle to include Yuen Ren Chao (1892–1982), Murray B. Emeneau (1904–2005), Einar Haugen (1906–1994), W. Freeman Twaddell (1906–1982), Henry Hoenigswald (1915–2003), Harry Hoijer (1904–1976), Albert H. Marckwardt (1903–1975), Mary R. Haas (1910–1996), J. Milton Cowan (1907–1993), William G. Moulton (1914–2000), Morris Swadesh (1909–1967), and Benjamin Lee Whorf (1897–1941). Roman Jakobson (1896–1982) took up a place outside post-Bloomfieldian linguistics but was nonetheless an important influence on mid-20th century study of language in the United States. All of these scholars were active during the relevant period and played roles in the development of American structuralism. Hymes and Fought (1975/1981, pp. 127–157) discuss these and other classificatory schemes at length and with great sensitivity. Falk (2003, pp. 129–135) synthesized inventories of American descriptivists/structuralists, some compiled by participants and some compiled retroactively by historiographers.

Focusing on the intersection of any of these candidate members of the cohort of post-Bloomfieldians and the list of LSA presidents, they can be separated into two groups (regardless of any role attributed to meaning). One group includes Fries, Chao, Emeneau, Haugen, Bloch, Voegelin, Harris, Twaddell, Hoijer, Trager, Marckwardt, Haas, Cowan, Hockett, Moulton, Hill, Wells, addressed in section 4. The other group, addressed in section 5, comprises Nida, Pike, and Jakobson, with an aside about two outlier figures who never served as presidents of the LSA, Swadesh and Whorf. It is impossible here to examine individually the works and intellectual biographies of each of these scholars. Instead, the goal of this section is to characterize some of what are purported to be the main tenets of post-Bloomfieldian linguistics, drawing for illustration and counter-illustration on presidential addresses presented to the LSA. Unfortunately, the addresses by Bloch, Hoijer, Cowan, and Wells proved irretrievable, so their important voices are not heard.3 Section 5 extends the characterization to outliers, including outliers to the post-Bloomfieldian group and others who contributed to American structuralism in the relevant interval, privileging those who left behind records of their views in the form of presidential addresses to the LSA.

4.2 Core Characteristics of the Post-Bloomfieldians

4.2.1 Identification of Linguistics as a Science; Anti-Mentalism

One famous hallmark of American structuralism derives from Bloomfield’s promotion of a mechanistic or physicalist approach to the study of language, often cited as his anti-mentalist stance. To Bloomfield, this was a cornerstone of his drive to conceive of linguistics as a science. Almost all presidential addresses in this corpus by post-Bloomfieldians cite Bloomfield in some capacity, and all adopt concepts and styles of argumentation that align them with scientific study of language. For example, Emeneau (1950, p. 199) announces flat-out that his goal is to build on Bloomfield’s 1926 “Postulates”; Voegelin (1956, p. 120) defines his assumptions about what constitute a “scientific analysis of natural languages”; Trager (1963, p. 5) takes it as his starting point that linguistics is a science. Even Haas (1978, p. 176), for whom Sapir was an inspiration more than Bloomfield, concedes that “most linguists are as anxious as ever to lay claim to being scientists.” Post-Bloomfieldians seem to take for granted the scientific basis of linguistics, to the extent that they labor over the definitions of terms (Chao, 1946, p. 5, ‘word’; Harris, 1957, p. 285, ‘co-occurrence’; Moulton, 1968, p. 451, ‘structural dialectology’); traffic in ‘hypotheses’ (Trager, 1963, pp. 17–18; Hockett, 1965, p. 185); predict features of unattested data (Haugen, 1951, p. 222); value economy in weighing the virtues of rival proposals (Twaddell, 1972, p. 232); or question the value of economy (Haugen, 1951, p. 222). It is fair to say that all these presidential addresses approach the collection and assessment of language data with the tools of scientific analysis as they were conceptualized in the mid 1900s. To this extent American structuralists, and in particular post-Bloomfieldians, unquestionably identified linguistics as a science.

However, the reputation of the post-Bloomfieldians as wedded to a narrowly mechanistic, physicalist, or anti-mentalist orientation seems under-attested in these texts. Hockett (1965, p. 194) did allude to Bloomfield’s anti-mentalism, but only in the context of recounting stages in the development of linguistics, not as a position that he and his peers had adopted. In fact, there is evidence that post-Bloomfieldians accepted as a matter of course terms and concepts incompatible with Bloomfieldian behaviorism. Chao (1946, p. 11) contrasted word-conscious English-speakers to character-conscious Chinese-speakers. Moulton (1968, pp. 462–465) discussed at length the ineluctable subjectivity of fieldworkers and of informants alike, how it affects collected data, and how a linguist’s analysis must accommodate the subjectivity of both parties. Hill (1970, pp. 253–257) treated speakers’ intuition about language delicately, but ended up investing value in it. M. Haas stayed true to her teacher Sapir in advocating for a “dynamic model” because “language is a social phenomenon” (1978, p. 216). She depicted Sapir as “always fascinated with the ‘why’,|. . . not only the historical ‘why’, but the psychological ‘why’”(p. 214)—an explicitly mentalistic stance, which Sapir passed on to at least some of his students who are now identified as post-Bloomfieldians.

These data point out that anti-mentalism was not widespread among American structuralists. For the scholars cited here, the scientific status of linguistics did not seem to require ruling out mentalistic concepts and practices, as it did to Bloomfield. In this sense, our sample of addresses to the LSA is consistent with a conclusion reached by Hymes and Fought (1975/1981, pp. 161–165), namely that thoroughgoing anti-mentalism was not necessarily a pillar of post-Bloomfieldian linguistics, despite its reputation to the contrary. In fact, Hockett concluded a summary of Bloomfield’s adoption of anti-mentalism by evincing mild skepticism about its adequacy:

In the 1920’s, Bloomfield came to see that the relationship of psychological ‘explanation’ to the linguistic phenomenon was something like that of the House of Lords to the House of Commons: when it agrees, superfluous; when it disagrees, obnoxious. He therefore recommended a policy known first as ‘mechanism’ or ‘antimentalism’, later as ‘physicalism’, which . . . amounted to the emancipation of linguistics from irrelevant psychologizing. . . . Bloomfield himself was not able to pursue this policy with complete success.

(Hockett, 1965, p. 196)

Hockett did not elaborate on how anti-mentalism came up short for Bloomfield, but the remark adds to the sense that the post-Bloomfieldians did not invest in it deeply.

4.2.2 Avoidance of Reliance on Meaning

Another trait purportedly passed from Bloomfield’s linguistics to post-Bloomfieldian linguistics is reluctance to rely on the meanings of utterances as tools in their analysis. In Bloomfield’s actual writings, he treats meaning thoroughly (1925, in which the word meaning appears multiple times on almost every page; 1933, pp. 139–157), if somewhat skeptically, on the grounds that semantics is an underdeveloped part of the study of language, a problem for future generations of linguists to deal with (1933, p. 140). Like anti-mentalism, avoiding reliance on meaning was part of Bloomfield’s campaign to conceive of linguistics as a science, because he considered meanings intractably subjective and therefore hard to integrate into a science of language.

Some post-Bloomfieldians seized upon this reluctance and magnified it into a trademark methodological practice. The claim was that the subjectivity entailed in using semantics to determine the identity or difference of two linguistic forms could be subverted by comparing not their meanings, but their distributions (Bloch & Trager, 1942, pp. 68–70; Trager & Smith, 1951, pp. 53–55). That is to say, /d/ and /t/ represent separate phonemes in the English pair ‘den’ / ‘ten’ not because ‘den’ is where bears hibernate and ‘ten’ is a numeral indicating one full unit in addition to nine, but because initial /d/ is voiced and unaspirated, while initial /t/ is devoiced and aspirated; or because ‘den’ and ‘ten’ belong to different form-classes: ‘den’ is the head of a noun phrase, while ‘ten’ appears in the (not fully specified) pre-nominal environment of determiners, quantifiers, and attributes. Replacement of meaning with formal distribution as an analytic tool is variously attested among the post-Bloomfieldians. Recall that Koerner (2002) sorted the post-Bloomfieldians into two cohorts according to the degree of their commitment to this principle. It is difficult to imagine how study of language could succeed under strict adherence to distributionalism. However, the corporate reputation of the group is skewed in that direction, so that Bolinger (1973, p. 540) could quip that “Those of us who trekked across the semantic desert of the forties and fifties could hardly have been blamed for feeling that life had lost all meaning, except perhaps differential meaning’.”

Distributionalism, and distributionalist-like leanings, are attested in presidential addresses to the LSA. As early as 1939, Fries’s discussion of how rigidity of word order waxed as inflectional complexity waned in the transition between Old to Modern English shows the influence of distributionalism. Fries conceptualizes this transition through the Bloomfieldian terms ‘taxeme of selection’ and ‘taxeme of order’ and tries at length, with modest success, to minimize reliance on semantics in defining ‘subject’, ‘direct object’, ‘object complement’, and so forth. Chao (1946) defined ‘word’ in Chinese using distributional criteria. Harris (1957) worked through examples of how ‘pronoun’, ‘modifier’, ‘direct object’, and so forth can be identified by inspection of their co-occurrence sets.

On the other hand, Haugen (1951, pp. 214–215) made a point of comparing the conceptualization of meaning in the work of Danish structuralist Louis Hjelmslev (1899–1965) to that of post-Bloomfieldian Martin Joos. Haugen concluded that “the analysis is the same” (p. 215), because both posit divisions of language into experience and meaning versus content/expression or sound. To Haugen, the difference between Hjelmslev and Joos is only terminological. Hill (1970, p. 253) made a point of identifying himself as a neo-Bloomfieldian, then went on to comment that even “in the days before the great transformation” (presumably, before the advent of [transformational] generative grammar), he felt uncomfortable assuming that “difference in context or environment . . . prove[s] difference of identity.” This may be a display of Hill’s membership in Koerner’s ‘moderate Bloomfieldian’ cohort—or it may indicate that the wave of distributionalism had crested by 1969 and was beginning to break.

4.2.3 Ban on Level-Mixing

Certain post-Bloomfieldians—for example Trager and Smith (1951, p. 54)—recommended that analysis of a language move from the identification and classification of phonemes to analysis at the ‘morphemic’ level—using the preferred term for where phonology interacts with morphology—and then on to syntax, after which point the clear path forward dissipates (into, for example, discourse analysis, or for some, the dark jungle of “meaning” [Gleason, 1988, unpublished manuscript, p. 44]). Trager and Smith warned that “This is not to say that in the actual procedure of analyzing a language there is not a constant going back and forth between phonology and morphemics” (p. 54); in any case, Trager and Smith’s plan mimics the organization of many long-standing pre- and non-Bloomfieldian grammars. Nevertheless, a formal ban on “level-mixing” has been attributed to post-Bloomfieldian linguistics. Gleason (1988, pp. 27–28) presented what he called stadial progression somewhat equivocally: as a convention that did not actually hold in the elicitation of grammars in a fieldwork setting, but that served as an organizational principle for the communication of research results. That is, Chapter 1 of a post-Bloomfieldian descriptive tract presented the phonology of the target language; Chapter 2, its morphology, in the course of which the analyst could freely refer to phonology; Chapter 3, syntax, which could presuppose Chapters 1 and 2. But Gleason does not offer a motivation for this practice. It was “simply an inflexible rule of the game” (p. 28).

With the exception of Trager (1963), which expands on level-ordering to propose 27 nested levels of analysis, the concept does not have an obvious presence in this corpus of presidential addresses. This does not necessarily mean that level-ordering was not an important consideration to the authors, or absent from their concerns. But it is notable that, aside from Trager (1963), the concept appears only sparsely. In this context, Emeneau’s (1950) presentation is striking. He proposed a version of distributionalism that admits the classification of some linguistic forms (e.g., numerals, kinship terms) not solely on the basis of morphology or syntax, but on the basis of patterns of organization in the speakers’ culture. Violations of the ban on level-ordering is usually depicted (e.g., by Voegelin & Voegelin 1963, p. 19) as the intrusion of morphology into phonemics; what Emeneau proposed is a much more audacious violation.

4.2.4 Absence of Theory

An additional purported attribute of post-Bloomfieldian linguistics is that it failed to develop a theory aimed at explaining the facts of language, but instead was satisfied with the lower goal of producing descriptive grammars in the style of “a sober taxonomy”(Joos, 1957/1966, p. 96, emphasis in the original) of elements of the target language.

It is certainly true that American structuralists in general felt the urgency of recording the many Native American languages that were then on the verge of extinction. From Boas forward, they prioritized the collection of data and its analysis. They also were slow to generalize their findings cross-linguistically. To do so might divert them from treating each language as a discrete, coherent system and risk imposing an a priori template onto the structure of all languages (Bloomfield, 1933, p. 20)—a move incompatible with the science of language that American structuralists were trying to build up.

This outlook, however, does not constitute failure to pursue a theory of language. In an insightful, carefully supported argument, Hymes and Fought (1975/1981, pp. 165–186) developed the point fully. In brief, Hymes and Fought assert that, using the terminology of ‘system’, ‘models’, ‘postulates’, and the close specification of terms, post-Bloomfieldians developed what they considered to be linguistic theory; there is evidence that post-Bloomfieldians explicitly rejected limiting their goal to assembling of a corpus of data; and they recognized that formal procedures for eliciting and handling data (or a body of informal practices and methods) are important instruments that sometimes precede theory and sometimes interact with it—but they did not mistake methods or procedures for theory itself.

Each of Hymes and Fought’s points—post-Bloomfieldian ‘models’; transcendence of data-collection; recognition of the distinction between methods or procedures and theory—are inscribed in the corpus of presidential addresses. For example, Voegelin (1956) is an extended comparison of conventions for phonemicizing languages, in which Voegelin brings data from Hopi to bear on the question of how to evaluate phonetic accuracy against morphophonetic fidelity—which is not a question about data per se, but about how to fashion a theory of language. Twaddell 1957 address (published 1972) raised a similar dilemma in recounting the history of attempts to deal with juncture in English affricates. Adjudicating between proposals means weighing faithfulness to phonetic detail against simplicity and compactness of statement: not the data, but theory, helps make this decision. Moulton (1968) likewise pointed out how “a structural approach to dialectology can contribute to linguistic theory” (p. 457) by bringing language variation to bear on phonemic analysis.

Moreover, post-Bloomfieldians did not simply collect and organize data, and they distinguished theories from methods. Inspired by an article by Harris (1946—which in turn cites Bloomfield, e.g., 1933, p. 161), Wells (1947) developed a technique called immediate constituent analysis (ICA). By repeatedly applying distributionalist criteria, ICA provided a means of analyzing the internal structure of words and of strings of words into a series of hierarchically arrayed constituents. Harris had presented his contribution as a “procedure” (p. 161). Wells develops it and introduces ICA as a “unified systematic theory” (p. 81); Hockett (1950, p. 56) insisted that it is “not an analytic technique, but a hypothesis about the nature of talking and hearing language.” Under any of these terms, ICA is not the activity of collecting and labeling data; rather, it presupposes and builds upon an interlocking family of assumptions and hypotheses about the nature of language. Other examples include Hockett’s (1954) “two models of grammatical description” (item-and-arrangement versus item-and-process), and Harris’s (1951) detailed exposition of descriptive procedures or structural methods, which he explicitly labeled as “not a plan for obtaining data for field work” (p. 1). Trager (1963) presented a direct argument for theory-building, in which he clearly distinguished theories from methods and from procedures, even as he called for development of all three:

We must clearly decide and state what we are doing. We must know what we are looking for. We must have a general theory, and special theories stemming from it. We must work out a methodology. We must develop procedures for using the methodology effectively. We must, in short, have a . . . usable frame of reference.

(Trager, 1963, p. 18)

Absence of theory misfires as a characterization of post-Bloomfieldian linguistics: these addresses to the LSA extensively state, challenge, support, refine, re-formulate, and extend theories of language. It may well be the case that what specifically counted as theory to the most prominent successor of the post-Bloomfieldians, (transformational) generative grammar, was not a major preoccupation to American structuralists in the first half of the 1900s (Hymes & Fought, 1975/1981, p. 170). But that is an unremarkable example of the development of a discipline over time.

4.3 Post-Bloomfieldian Projects

In addition to their intellectual contributions to the study of language, American structuralists, including the post-Bloomfieldian generation, left considerable material and institutional residue of their work. The foundation of the LSA was central. Its sponsorship of the summer Linguistic Institute, first held in 1928, then after 1936 annually, biennially, or at longer intervals as conditions permitted, continues the LSA’s educational mission (Falk, 2014; Hill, 1964; Report, 1940). First-hand accounts of the meetings of the Linguistic Institute (Gleason, 1988; Hill, 1991; Joos, 1986) emphasize the efficacy of Linguistic Institutes in building social solidarity and disciplinary cohesion among attendees, and in advancing the spread of new ideas. Those accounts also attest to the high level of commitment of the organizers, host universities, and the membership of the LSA during the post-Bloomfieldian decades. The Linguistic Institutes were collaboratively conceived, planned, and run, to the benefit of the American linguists at large.4

In addition, two projects were undertaken with the involvement of LSA members during the 1930s through 1950s, even though neither was formally within the purview of the Society. From the very beginning of the LSA, members had evinced interest in undertaking a survey of American dialectology (Murray, 1994, p. 142). After a few false starts, the American Council of Learned Societies agreed to sponsor a linguistic atlas of the United States, although many meetings were held at Linguistic Institutes under the aegis of the LSA. The Modern Language Association and the American Dialect Society also supported the project at various stages (Joos, 1986, pp. 43–44). Nevertheless, the Linguistic Atlas of the United States and Canada took a long time to get on its feet, and even longer to get through the stages of planning, development of field techniques, data collection, analysis, and editing to see the daylight of publication. LSA members (and presidents) Hans Kurath and Charles Fries played leading roles. The scope of the total project eventually proved overwhelming, so that it was broken down by region (Linguistic Atlas of the Pacific Northwest; . . . of the Upper Midwest, and so on), with many of the findings now available online at the Linguistic Atlas Projects. The first completed volume, the Linguistic Atlas of New England (Kurath, 1939–1943) set the tone. Its production involved as fieldworkers post-Bloomfieldians Martin Joos and Bernard Bloch (and for the Linguistic Atlas of the North Central States, Albert H. Marckwardt). The dialect atlas captured between hard covers one of the interests of American structuralists, and it provided some with training in field techniques.

A second project, spanning the years 1939 to 1945, comprised two related initiatives of American structuralists in the field of language teaching. J Milton Cowan (1975) describes his involvement in what became the Intensive Language Program (ILP), organized by the American Council of Learned Societies. Henry Lee Smith Jr., a former student of Sapir’s and then officer in the U.S. Air Force, was also involved as an organizer. Starting in 1939, the goal was to apply to modern foreign languages of strategic national importance the techniques of elicitation and analysis that had developed in American structuralists’ fieldwork with Native American languages, and to develop appropriate teaching materials. Bloomfield (1942) published through the LSA a pamphlet, designed for both teachers and students, that summarized how to elicit and analyze a foreign language from a native-speaking informant. Not surprisingly it took for granted Bloomfieldian anti-mentalism and focused on the distribution of forms (Moulton, 1961). Mary Haas was one of the first linguists to be recruited, to teach Thai in collaboration with a native-speaking informant. Eventually, in the urgent “hothouse atmosphere of the wartime work” (Joos, 1957/1966, p. 108), a large number of American structuralists became involved: Hockett, Sebeok, Hall, Moulton, Hoenigswald, Bloch, Haugen, Swadesh, Trager, Chao, Harris, Emeneau, Newman (Cowan, 1975, p. 32; Murray, 1994, pp. 147–148).

The Intensive Language Program was underway and producing promising results when the U.S. government decided to expand it into the Army Specialized Training Program (ASTP). The goal of this project was to prepare materials and methods for teaching 1,500 enlistees one of 13 under-studied languages of Asia, Eastern Europe, and the Middle East. A subset of American structuralists, mostly overlapping with ILP personnel, worked in New York City alongside native speakers under high pressure to produce grammars and dictionaries. In the end, ten times more learners than had originally been anticipated—“15,000 of the best minds in our armed Forces” (Cowan, 1975, p. 33)—underwent language training through the ASTP. Unfortunately, the program overshot its own organization and collapsed without making much of a dent in American monolingualism before it was abruptly abandoned with the end of World War II. However, the ASTP galvanized American structuralists, led them to work out applications of their research into language, and enhanced the coherence of the group. Hall characterized it “as a gathering-point for American descriptivists and as a proving ground to test and demonstrate their newly developed capabilities, from which they and their influence were diffused to the rest of the country once the war was over” (1991, p. 164).

Marckwardt’s 1962 presidential address (published, 1964) comes closest among texts in our corpus to offering a primary source of reflection on the institutional accomplishments of American structuralists in the first half of the 1900s. He reviewed the success of the society, the Linguistic Institutes, the ILP, and the growth of graduate education. But Marckwardt was not satisfied by what the profession had achieved to date. His theme is that linguists need to much more fully engage in public outreach, especially in language education and foreign-language teaching. His talk is evidence that at least some American structuralists felt that their advanced training in language entailed a sense of responsibility to the public.

Our corpus could, but doesn’t, contain direct attestations of applications of American structuralism. Cowan, president of the LSA in 1966, was an organizational leader behind the ILP and the ASTP. Unfortunately, his address was never printed and has not survived (Hockett, 1995, p. 347). Kurath’s presidency occurred in 1942, when the Society did not meet. But even without the witness of these texts, it is clear that American structuralists launched ambitious ventures, constructed valuable intellectual institutions, and applied their skills to real-world problems.

5. Outliers Within American Structuralism

Several detailed, first-hand narratives of American structuralism, produced by core participants (Gleason, 1988; Hall, 1975; Hill, 1991; Joos, 1986), converge on a representation of it that emphasizes its cohesion. Competition and differences of opinion certainly burst forth into outright controversy among American structuralists, but the overall atmosphere of camaraderie, even “clubbiness,” in the face of a shared project is unmistakable. However, there were also contemporary linguists who for various reasons had one foot outside the group—and whose points of view are worth attending to. Contributors to American structuralism who require treatment outside the discussion above include missionary linguists; European-American structuralists; sui generis figures; and, harder to characterize, those whose contributions went unnoticed.

5.1 Missionary Linguists

One semi-distinctive group of American structuralists belongs among the post-Bloomfieldians, by dates and by their inheritance of the linguistics of Boas, Sapir, and Bloomfield. However, the intellectual differences between core Bloomfieldians such as Bloch, Trager, or Harris and this group are somewhat wider than that between core Bloomfieldians and Koerner’s “moderate Bloomfieldians” such as Hill, Hall, or Hockett. This group comprises scholars who took up the study of language motivated by their vocations as Christian missionaries. Two LSA presidents in the relevant interval represent missionary-linguists: Pike (president, 1961) and Nida (1968).

Pike was one of the very first recruits, in 1935, of a Christian evangelical group now known as SIL International. The goal of the group is to translate the Bible so that all people can encounter it in their native language. SIL trains missionaries in linguistic field methods, then sends them out in teams to immerse themselves first-hand in the language and culture they are assigned to, which often entails developing an orthography for a previously unwritten language. Pike’s first placement was in Mexico, where he studied Mixtec. Eventually he also did fieldwork in Peru and Ecuador, and advised SIL-affiliated linguists worldwide.

Pike published prolifically, developing among other proposals a distinctive linguistic theory, called tagmemics. The name derived from Bloomfield’s (1933, pp. 166–167) introduction of the term tageme to label the smallest meaningful grammatical unit. Although Pike studied with Charles Fries, corresponded with Bloomfield, enjoyed friendships with many of the post-Bloomfieldians, attended the same Linguistic Institutes as they did, and read and published in the same journals, he did not share their struggles with meaning or level-ordering, or their reticence to vault forward into syntax before exhausting phonology and morphophonemics. Pike’s (1967) work made an ambitious attempt not only to provide a framework for describing every dimension of human language but also to extend the same structure-analytic tools (hierarchies, immediate constituents, features, levels, segmentation, classes, and slots) to the description of gesture, human behavior, and even large-scale cultural constructions such as church services or football games. His 1961 presidential address works through one part of this massive undertaking, by proposing a “dimensional analysis” that extends to syntax the kind of feature analysis used in phonology, within a system of “formal interlocking of structural levels” (p. 223).

The extent to which Pike and other missionary-linguists’ intellectual lives were shaped by the exigencies of Bible translation needs further study. Eugene Nida prepared for missionary fieldwork with the same group as Pike, but in 1943, switched to translation studies under the aegis of the American Bible Society. His career was taken up with articulating a theory of translation that he called “dynamic equivalence,” which aimed to translate the intended meaning behind the source text rather than re-creating it word-for-word. In Nida’s 1968 presidential address, he presupposed a science of translation, but otherwise argued for dynamic equivalence in terms alien to those of the post-Bloomfieldians, including discussion of restructuring and kernel versus surface structures (pp. 484–485). Nida also plunged without inhibition into semantics to explain the complexities of words like ‘dancer’ or ‘apostle’, the complexities of word connotations, and to speculate about semantic universals (pp. 485–492). Nida was not as socially or intellectually integrated as Pike into the post-Bloomfieldians, but his work on translation is still part of the landscape of American structuralism.

5.2 “European-American Structuralists”

More tangentially related to the post-Bloomfieldians was a group of linguists whose claims to being American were more attenuated than their credentials as structuralists. The group comprised European language scholars who had already completed their professional training before being displaced by World War II. Some among them returned to Europe after the war (functional linguist André Martinet, Italian specialist Giuliano Bonfante). Some stayed (philologist Yakov Malkiel, historical linguist Henry Hoenigswald, Romance language scholars Leo Spitzer and Henry and Renée Kahane) and, to various extents, adopted or resisted American structuralism. Most prominent among the latter group was Roman Jakobson, who came to New York from Russia via Prague, where he had lived for almost 20 years. In Prague, Jakobson helped develop a distinctive “Prague School” of structural functionalism, which examined how meaning, discourse principles, or information structure shape grammar, and which played up the intersection of literature with linguistics. Some American structuralists were put off by Jakobson’s intellectual outlook and his over-the-top self-confidence (Hall, 1975), but he eventually established a following and moved forward with his many interests: Slavic philology, sound structure (especially distinctive features), markedness, poetics, semiotics, psycholinguistics. Jakobson kept his distance intellectually from the anti-mentalist, mechanistic, strain of post-Bloomfieldian linguistics, but he eventually won the friendship of many American structuralists and remained based in the United States for the rest of his life. His 1956 presidential address gives a glimpse into Jakobsonian linguistics and its exuberant, unretouched, functionalism. Jakobson’s only citation of an American structuralist was to remark in passing that his metalinguistic analysis “permit[s] us to overcome Leonard Bloomfield’s forebodings in his endeavors to incorporate meaning into the science of language” (1976, p. 350). Jakobson was a force in American linguistics from the 1940s, even positioned as he was, as an outlier to the dominant group.

Among other émigré linguists during the period in focus, Hoenigswald was elected to the LSA presidency in 1958, and Malkiel in 1965. Hoenigswald emigrated from Germany and participated in the wartime ILP by authoring a textbook on Hindustani. He went on to make significant contributions to Indo-European comparative-historical linguistics, the topic of his presidential address. Although Malkiel emigrated from Russia to the United States soon after he completed his doctorate and at a younger age than Jakobson, like Hoenigswald he retained his philological orientation relatively undisturbed by being physically displaced into the milieu of American structuralism. Malkiel’s presidential address concerns the role of etymology in comparative-historical linguistics. The papers in Hoenigswald (1979) survey a range of assessments of the interaction between European and American linguistics in the relevant period—or of their independence.

5.3 Sui Generis Figures

In addition, there are two important American structuralists who in different ways took up positions outside the group’s central orbit. Both were students of Sapir. Benjamin Lee Whorf’s (1897–1941) name is routinely invoked in discussions of the interrelationship of language and culture, in the guise of the Sapir-Whorf, sometimes simply Whorfian, hypothesis. Whorf started out self-taught in linguistics. Later, he enrolled in a PhD program at Yale where he studied with Sapir without pursuing a degree, although he taught Sapir’s classes as his mentor’s health failed. Whorf’s first fieldwork was on Aztec and Mayan writing systems, which he undertook by taking leave of his full-time job with the Hartford Fire Insurance Company. He did extensive work on Hopi as well. Whorf’s reputation on the topic of language and culture, or linguistic relativity (in its milder version) versus linguistic determinism (in its rigorous, less compromising version), derives from three articles he wrote for a non-specialist readership, excerpted in Whorf (1956). The interpretation of these texts has given rise to an extensive hermeneutical industry (Koerner, 1995; Lee, 1996). Among the questions still being probed are whether the features of particular languages differentially privilege different ways of perceiving the environment or lead to different cognitive styles. This wing of Whorf’s work stands out as an original contribution to American structuralism that retains a post-Sapirian orientation against the backdrop of post-Bloomfieldian ‘mechanistic’ linguistics. Whorf never completed doctoral work, never held a regular academic appointment, and died young. He was not elected president of the LSA.

Morris Swadesh (1909–1967) worked on many different Native American languages, especially those spoken in what is now Mexico. He had a rocky professional life. Swadesh participated in the ILP as an officer in the U.S. Army Signal Corps, and had a series of short appointments, sometimes prematurely truncated, at the City College of New York, Yale, the University of Wisconsin, and several universities and institutes in Mexico City. He produced many valuable descriptive grammars of endangered (and now extinct) languages, while also maintaining an interest in theoretical work and, like Sapir, in the application of comparative-historical linguistics to the unwritten languages of the Americas. In particular, from around 1950, Swadesh was associated with the emergence of glottochronology, a technique for estimating the time-depth of separation of two genetically related languages by examining sets of cognates (Swadesh, 1952). Glottochronology (sometimes conceived more broadly as lexicostatistics) raises many controversial issues, including the extent to which lexical borrowing skews glottochronological estimates and the validity of its assumption that languages change at a steady rate. Swadesh retained his commitment to the technique, applied it to diverse language families, and doggedly chipped away at critiques of it. He published widely, participated in the LSA and in general professional debates. Swadesh was never elected to the presidency of the LSA, however—perhaps partly due to resistance to glottochronology, or to his lack of a stable academic position in the United States. Nevertheless, both Swadesh and Whorf are distinctive parts of the mosaic of American structuralism.

5.4 Invisible Contributors

This account closes by recognizing some who are not acknowledged in the history of American structuralism, but who built it up—or who could have done so. Principal among them are women scholars (Falk, 1994, 1999). With the salient exception of Mary Haas, none of the leaders of American structuralism discussed here was female. In the first 60 years of the LSA, only two women served as president: classicist E. Adelaide Hahn in 1946, and Haas in 1963. Few of the extensive first-hand chronicles of the emergence of American structuralism (e.g., Cowan, 1991; Gleason, 1988; Hall, 1975; Hill, 1991; Joos, 1986), seem to notice a gap between the participation of men versus women (but cf. Abramson, 1975, p. 39). Granted that fewer women than men received doctoral training in linguistics in those years (Thomas, 2014), it is still painful to come across Joos’s (1986, p. 9) offhand comment about the “routine ignoring of all female scholars” in the 1920s and 1930s. In the pages of the LSA Bulletin and other texts documenting attendance or presentations at conferences, what few women’s names do appear are uniformly segregated, in an apparently genial spirit, by the terms of address used: women are ‘Miss (rarely ‘Mrs.’) LastName’; men are the unmarked case, ‘FirstName LastName’.

In addition to women who were not included in the development of American structuralism, there were women who did contribute, but invisibly so. LSA secretary Hill (1991, p. 53) tipped his hat to his wife’s hours of organizational and administrative work on the behalf of the society. His predecessor, J Milton Cowan, seemed to take for granted that his own wife’s talents as a bookkeeper were at the disposal of the LSA (1975, p. 29). It was not until 1995 that Hockett (1995, p. 344) noticed the systematic, silent, contribution that many women married to linguists made to the institutional, if not the intellectual, life of American structuralism, attributing it ruefully to “the temper of the times.”

One unexpected exception to the “routine ignoring of all female scholars” is materialized in Pike’s presidential address, published in 1962 in Language. In 23 pages, Pike cites the scholarship of 15 men and 10 women. The closest relevant comparison can be made to Hockett’s 1965 presidential address, also published in Language. In 19 pages, Hockett cites the scholarship of 80 men and 2 women. Pike’s citation practices are unusual in the context of American structuralism in general, but consistent with a trend among affiliates of SIL International to recognize and incorporate the work of linguists who happened to be women—a striking, paradoxical, attribute of a socially conservative missionary group (Thomas, 2009, 2019).

American structuralism was no doubt impoverished by its almost exclusively male participants, but perhaps even more impoverished by its ethnic and racial homogeneity. In a discussion in the 1950s, of the society’s practices of assessing abstracts submitted for presentation at the annual meeting, Hill (1975, p. 36) weighed the pros and cons of blind review. If submissions continued to be read without blinding, Hill wrote that “There is, of course, the disadvantage that newcomers, those with un-WASP-ish names, and women may be slighted.” It seems likely that all three of those groups were disadvantaged on this score as on others, a bias that was part of the cultural backdrop to American structuralism. Although the social and ethnic homogeneity of American structuralism may not be its most salient feature, its recognition reminds us that history must be periodically re-written, because each generation reconceives what is remarkable about the past.

Further Reading

Bloch, B., & Trager, G. L. (1942). Outline of linguistic analysis. Baltimore: Linguistic Society of America.Find this resource:

    Gleason, H. A. (1955). An introduction to descriptive linguistics. New York: Holt.Find this resource:

      Hockett, C. F. (1958). A course in modern linguistics. New York: Macmillan. doi:10.1111/j.1467-1770.1958.tb00870.xFind this resource:

        Hockett, C. F. (1970). A Leonard Bloomfield anthology. Bloomington: Indiana University Press.Find this resource:

          Mandelbaum, D. G. (1949). Selected writings of Edward Sapir in language, culture, and personality. Berkeley, CA: University of California Press. doi:10.2307/2103561Find this resource:

            Pike, K. L. (1947). Phonemics: A technique for reducing languages to writing. Ann Arbor: University of Michigan Press.Find this resource:


              Abramson, A. S. (1975). The future: Golden reminiscences. LSA Bulletin, 64, 25–39.Find this resource:

                Anderson, S. R. (1985). Phonology in the twentieth century: Theories of rules and theories of representations. Chicago: Chicago University Press.Find this resource:

                  Austerlitz, R. (Ed.). (1975). The scope of American linguistics. Lisse, Netherlands: Peter de Ridder. doi:10.1515/9783110857610Find this resource:

                    Bloomfield, L. (1983). An introduction to the study of language (new ed.). Amsterdam: John Benjamins. (Originally published in 1914.) doi:10.1075/cipl.3Find this resource:

                      Bloomfield, L. (1925). Why a linguistic society? Language, 1(1), 1–5.Find this resource:

                        Bloomfield, L. (1926). A set of postulates for the science of language. Language 2(3), 153–164. doi:10.2307/408741Find this resource:

                          Bloomfield, L. (1930). Linguistics as a science. Studies in Philology, 27(4), 553–557.Find this resource:

                            Bloomfield, L. (1933). Language. New York: Henry Holt.Find this resource:

                              Bloomfield, L. (1935). Linguistic aspects of science. Philosophy of Science, 2(4), 499–517. doi:10.1086/286392Find this resource:

                                Bloomfield, L. (1936). Language or ideas? Language, 12(2), 89–95. doi:10.2307/408751Find this resource:

                                  Bloomfield, L. (1942). Outline guide for the practical study of foreign languages. Baltimore: Linguistic Society of America.Find this resource:

                                    Bloomfield, L. (1946). Twenty-one years of the Linguistic Society. Language, 22(1), 1–3.Find this resource:

                                      Bloch, B. (1949). Leonard Bloomfield [Obituary]. Language, 25, 87–94.Find this resource:

                                        Boas, F. (1911). Introduction. In F. Boas, Handbook of American Indian languages (Part 1, pp. 1–83). Washington, DC: Government Printing Office.Find this resource:

                                          Boas, F. (1929). The classification of American Indian languages. Language, 5(1), 1–7. doi:10.2307/408990Find this resource:

                                            Bolinger, D. (1968). Aspects of language. New York: Harcourt, Brace, and World.Find this resource:

                                              Bolinger, D. (1973). Truth is a linguistic question. Language, 49(3), 539–550. doi:10.2307/412350Find this resource:

                                                Buck, C. D. (1949). Dictionary of selected synonyms in the principal Indo–European languages. Chicago: University of Chicago Press.Find this resource:

                                                  Chao, Y. R. (1946). The logical structure of Chinese words. Language, 22(1), 4–13. doi:10.2307/410329Find this resource:

                                                    Cowan, J. M. (1975). Peace and war: Golden reminiscences. LSA Bulletin, 64, 28–34.Find this resource:

                                                      Cowan, J. M. (1991). American linguistics in peace and at war. In K. Koerner (Ed.), First person singular II: Autobiographies by North American scholars in the language sciences (pp. 67–82). Amsterdam: John Benjamins. doi:10.1075/sihols.61.05cowFind this resource:

                                                        Darnell, R. (1990). Franz Boas, Edward Sapir, and the Americanist text tradition. In F. P. Dinneen, S. J. Koerner, & E. F. K. Koerner (Eds.), North American contributions to the history of linguistics (pp. 129–144). Amsterdam: John Benjamins. doi:10.1075/hl.17.1-2.11darFind this resource:

                                                          Darnell, R. (1998). And along came Boas: Continuity and revolution in Americanist anthropology. Amsterdam: John Benjamins. doi:10.1075/sihols.86Find this resource:

                                                            Emeneau, M. B. (1950). Language and non-linguistic patterns. Language, 26(2), 199–209. doi:10.2307/410057Find this resource:

                                                              Falk, J. S. (1994). The women foundation members of the Linguistic Society of America. Language, 70(3), 455–490. doi:10.2307/416482Find this resource:

                                                                Falk, J. S. (1999). Women, language and linguistics: Three American stories from the first half of the twentieth century. New York: Routledge. doi:10.4324/9780203212646Find this resource:

                                                                  Falk, J. S. (2014). The LSA Linguistic Institutes. Linguistic Society of America 90th Anniversary Celebration. Minneapolis, MN, January 2014.

                                                                  Falk, J. S. (2003). Turn to the history of linguistics: Noam Chomsky and Charles Hockett in the 1960s. Historiographia Linguistica, 30(1–2), 129–185. doi:10.1075/hl.30.1.05falFind this resource:

                                                                    Fries, C. C. (1940). On the development of the structural use of word-order in modern English. Language, 16(3), 199–208. doi:10.2307/409055Find this resource:

                                                                      Fries, C. C. (1961). The Bloomfield “school”. In C. Mohrmann, A. Sommerfelt, & J. Whatmough (Eds.), Trends in European and American linguistics, 1930–1960 (pp. 196–224). Utrecht, The Netherlands: Spectrum.Find this resource:

                                                                        Gleason, H. A. (1988). Theories in conflict: North American linguistics in the fifties and sixties. Unpublished manuscript, Willowdale, Ontario.Find this resource:

                                                                          Haas, M. R. (1978). Linguistics and history. In A. S. Dil (Ed.), Language, culture, and history: Essays by Mary R. Haas (pp. 207–219). Stanford, CA: Stanford University Press.Find this resource:

                                                                            Haas, M. R. (1976). Boas, Sapir, and Bloomfield. In W. L. Chafe (Ed.), American Indian languages and American linguistics (pp. 59–69). Lisse, The Netherlands: Peter de Ridder.Find this resource:

                                                                              Haas, W. (1978). Linguistics 1930–1980. Journal of Linguistics, 14(2), 293–308. doi:10.1017/S0022226700005910Find this resource:

                                                                                Hall, R. A., Jr. (1975). Stormy petrel in linguistics. Ithaca, NY: Spoken Language Services.Find this resource:

                                                                                  Hall, R. A., Jr. (1991). 165 Broadway: A crucial node in American structural linguistics. Historiographia Linguistica, 18(1), 153–166. doi:10.1075/hl.18.1.05halFind this resource:

                                                                                    Harris, Z. S. (1951). Methods in structural linguistics. Chicago: University of Chicago Press.Find this resource:

                                                                                      Harris, Z. S. (1957). Co-occurrence and transformation in linguistic structure. Language, 33(3), 283–340. doi:10.2307/411155Find this resource:

                                                                                        Harris, Z. S. (1946). From morpheme to utterance. Language, 22(3), 161–183. doi:10.2307/410205Find this resource:

                                                                                          Harris, Z. S. (1986). Language and information: The Bampton Lectures at Columbia University.Find this resource:

                                                                                            Harris, Z. S. (1988). Language and information. New York: Columbia University Press.Find this resource:

                                                                                              Haugen, E. (1951). Directions in modern linguistics. Language, 27(3), 211–222. doi:10.2307/409751Find this resource:

                                                                                                Hill, A. A. (1964). History of the Linguistic Institute. ACLS Newsletter, 15(3), 1–12.Find this resource:

                                                                                                  Hill, A. A. (1970). Laymen, lexicographers, and linguists. Language, 46(2), 245–258. doi:10.2307/412277Find this resource:

                                                                                                    Hill, A. A. (1975). The middle years: Golden reminiscences. LSA Bulletin, 64, 34–39.Find this resource:

                                                                                                      Hill, A. A. (1991). The Linguistic Society of America and North American linguistics, 1950–1968. Historiographia Linguistica, 18(1), 49–152. doi:10.1075/hl.18.1.04hilFind this resource:

                                                                                                        Hill, J., & Mannheim, B. (1992). Language and world view. Annual Review of Anthropology, 21, 381–406. doi:10.1146/ this resource:

                                                                                                          Hiż, H., & Swiggers, P. (1990). Bloomfield, the logical positivist. Semiotica, 79(3–4), 257–270. doi:10.1515/semi.1990.79.3-4.257Find this resource:

                                                                                                            Hockett, C. F. (1950). Which approach in linguistics is “scientific”? Studies in Language, 8(3), 53–57.Find this resource:

                                                                                                              Hockett, C. F. (1954). Two models of grammatical description. Word, 10(2–3), 210–234. doi:10.1080/00437956.1954.11659524Find this resource:

                                                                                                                Hockett, C. F. (1965). Sound change. Language, 41(2), 185–204. doi:10.2307/411873Find this resource:

                                                                                                                  Hockett, C. F. (Ed.). (1970). A Leonard Bloomfield anthology (pp. 544–547). Bloomington: Indiana University Press. (Originally published in 1950, in E. H. Sturtevant. Leonard Bloomfield. The American Philosophical Society yearbook 1949, pp. 302–305.)Find this resource:

                                                                                                                    Hockett, C. F. (1993). George Leonard Trager [Obituary]. Language, 69(4), 778–788.Find this resource:

                                                                                                                      Hockett, C. F. (1995). J Milton Cowan [Obituary]. Language, 71(2), 341–348.Find this resource:

                                                                                                                        Hoenigswald, H. M. (1959). Some uses of nothing. Language, 35(3), 409–420. doi:10.2307/411228Find this resource:

                                                                                                                          Hoenigswald, H. M. (Ed.). (1979). The European background of American linguistics: Papers of the Third Golden Anniversary Symposium of the Linguistic Society of America. Dordrecht, The Netherlands: Foris. doi:10.1515/9783110870350Find this resource:

                                                                                                                            Hymes, D. (1976). The Americanist tradition. In W. L. Chafe (Ed.), American Indian languages and American linguistics (pp. 11–28). Lisse, The Netherlands: Peter de Ridder.Find this resource:

                                                                                                                              Hymes, D., & Fought, J. (1975/1981). American structuralism. The Hague: Mouton.Find this resource:

                                                                                                                                Jakobson, R. (1976). Metalanguage as a linguistic problem. Nyelvtudományi Közlemények, 78, 346–352.Find this resource:

                                                                                                                                  Joos, M. (1986). Notes on the development of the Linguistic Society of America. Ithaca, NY: Linguistica.Find this resource:

                                                                                                                                    Joos, M. (Ed.). (1966). Readings in linguistics: The development of descriptive linguistics in America since 1925 (4th ed.). Washington, DC: American Council of Learned Societies. Original work published 1957.Find this resource:

                                                                                                                                      Joseph, J. E. (2002). From Whitney to Chomsky: Essays in the history of American linguistics. Amsterdam: John Benjamins. doi:10.1075/sihols.103Find this resource:

                                                                                                                                        Koerner, E. F. K. (2002). American structuralist linguistics and the “problem of meaning”. Toward a history of American linguistics (pp. 75–104). London: Routledge (Original work published 1970.)Find this resource:

                                                                                                                                          Koerner, E. F. K. (2002). Toward a history of American linguistics (pp. 75–104). London: Routledge. (Originally published in 1970 as E. F. K. Koerner, Bloomfieldian linguistics and the problem of ‘meaning’: A chapter in the history and study of language. Jahrbuch für Amerikastudien / German Yearbook of American Studies, 15, 162–183.)Find this resource:

                                                                                                                                            Koerner, E. F. K. (1995). The “Sapir-Whorf Hypothesis”: A historico-bibliographical essay. In Professing linguistic historiography (pp. 203–240). Amsterdam: John Benjamins. doi:10.1075/sihols.79Find this resource:

                                                                                                                                              Koerner, E. F. K., & Asher, R. E. (Eds.). (1995). Concise history of the language sciences. Oxford: Elsevier doi:10.1075/hl.23.1-2.23linFind this resource:

                                                                                                                                                Kurath, H. (1939–1943). Linguistic atlas of New England (Vols. 1–3). Providence, RI: Brown University for the American Council of Learned Societies.Find this resource:

                                                                                                                                                  Lane, G. S. (1945). Changes in emphasis in linguistics with particular reference to Paul and Bloomfield. Studies in Philology, 42(3), 465–483.Find this resource:

                                                                                                                                                    Lee, P. (1996). The Whorf theory complex: A critical reconstruction. Amsterdam: John Benjamins. doi:10.1075/sihols.81Find this resource:

                                                                                                                                                      Malkiel, Y. (1982). Mutual attraction, typological convergence, multiple borrowing, parallel independent development, spiral-shaped curve, periodicity, advanced state, slope, drift, drag, slant. Romance Philology, 35(3), 479–484.Find this resource:

                                                                                                                                                        Malkiel, Y. (1967). Linguistics as a genetic science. Language, 43(1), 223–245. doi:10.2307/411396Find this resource:

                                                                                                                                                          Marckwardt, A. H. (1964). Opportunity and obligation. LSA Bulletin, 37, 26–37.Find this resource:

                                                                                                                                                            Moulton, W. G. (1961). Linguistics and language teaching in the United States, 1940–1960. In C. Mohrmann, A. Sommerfelt, & J. Whatmough (Eds.), Trends in European and American linguistics, 1930–1960 (pp. 82–109). Utrecht, The Netherlands: Spectrum. doi:10.1515/iral.1963.1.1.21Find this resource:

                                                                                                                                                              Moulton, W. G. (1968). Structural dialectology. Language, 44(3), 451–466. doi:10.2307/411713Find this resource:

                                                                                                                                                                Murray, S. O. (1991). The first quarter century of the Linguistic Society of America, 1924–1949. Historiographia Linguistica, 18(1), 1–48. doi:10.1075/hl.18.1.03murFind this resource:

                                                                                                                                                                  Murray, S. O. (1994). Theory groups and the study of language in North America: A social history. Amsterdam: John Benjamins. doi:10.1075/sihols.69Find this resource:

                                                                                                                                                                    Newmeyer, F. J. (1986). Linguistic theory in America (2nd ed.). Orlando, FL: Academic Press.Find this resource:

                                                                                                                                                                      Nida, E. A. (1969). Science of translation. Language, 45(3), 483–498. doi:10.2307/411434Find this resource:

                                                                                                                                                                        Pike, K. L. (1962). Dimensions of grammatical constructions. Language, 38(3), 221–244. doi:10.2307/410782Find this resource:

                                                                                                                                                                          Pike, K. L. (1967). Language in relation to a unified theory of the structure of human behavior. The Hague: Mouton. doi:10.1037/14786-000Find this resource:

                                                                                                                                                                            Report of the Special Committee on the Linguistic Institute. (1940). Language, 16(1), 83–101.Find this resource:

                                                                                                                                                                              Sapir, E. (1921). Language: An introduction to the study of speech. New York: Harcourt, Brace and World.Find this resource:

                                                                                                                                                                                Sapir, E. (1929). A study in phonetic symbolism. Journal of Experimental Psychology, 12, 225–239. doi:10.1037/h0070931Find this resource:

                                                                                                                                                                                  Sapir, E. (1934). Symbolism. Encyclopaedia of the Social Sciences (Vol. 14), 492–495. New York: Macmillan.Find this resource:

                                                                                                                                                                                    Schwegler, A. (1990). Analyticitiy and syntheticity: A diachronic perspective with special reference to Romance languages. Berlin: Walter de Gruyter.Find this resource:

                                                                                                                                                                                      Sebeok, T. A. (1976). Contributions to the doctrine of signs. Bloomington, IN: Indiana University Press.Find this resource:

                                                                                                                                                                                        Silverstein, M. (1986). The diachrony of Sapir’s synchronic linguistic description; or, Sapir’s “cosmographical” linguistics. In W. Cowan, M. K. Foster, & E. F. K. Koerner (Eds.), New perspectives in language, culture, and personality (pp. 67–106). Amsterdam: John Benjamins. doi:10.1075/sihols.41.06silFind this resource:

                                                                                                                                                                                          Silverstein, M. (2015). From Baffin Island to Boasian induction: How anthropology and linguistics got into their interlinear groove. In R. Darnell, M. Hamilton, R. L. A. Hancock, & J. Smith (Eds.), The Franz Boas papers (Vol. 1, pp. 83–127). Lincoln, NE: University of Nebraska Press.Find this resource:

                                                                                                                                                                                            Stocking, G. W., Jr. (1974). The Boas plan for the study of American Indian languages. In D. Hymes (Ed.), Studies in the history of linguistics: Traditions and paradigms (pp. 454–484). Bloomington, IN: Indiana University Press.Find this resource:

                                                                                                                                                                                              Swadesh, M. (1952). Lexico-statistic dating of prehistoric ethnic contacts: With special reference to North American Indians and Eskimos. Proceedings of the American Philosophical Society, 96(4), 452–463.Find this resource:

                                                                                                                                                                                                Thomas, M. (2009). Gender and the language scholarship of the Summer Institute of Linguistics in the context of mid-twentieth century American linguistics. In G. Hassler (Ed.), History of linguistics 2008: Selected papers from the XI International Conference on the History of the Language Sciences (ICHOLS XI), Potsdam, 28 August–2 September 2008 (pp. 389–397). Amsterdam: John Benjamins Press.Find this resource:

                                                                                                                                                                                                  Thomas, M. (2014). How did (American) women linguists become women in linguistics? Linguistic Society of America 90th Anniversary Celebration, Minneapolis, MN, January 2014. 2014 Annual Meeting 90th Anniversary Presentation Materials. Linguistic Society of America.Find this resource:

                                                                                                                                                                                                    Thomas, M. (2019). American women’s contributions to the history of linguistics. To appear in H. Sanson & W. Ayres-Bennett (Eds.), Distant and neglected voices. Oxford: Oxford University Press.Find this resource:

                                                                                                                                                                                                      Trager, G. L. (1963). Linguistics is linguistics: Studies in Linguistics: Occasional Papers, 10.Find this resource:

                                                                                                                                                                                                        Trager, G. L., & Smith, H. L. (1951). An outline of English structure. Normal, OK: Oklahoma University Press.Find this resource:

                                                                                                                                                                                                          Twaddell, W. F. (1972). /č/? American Speech, 27(3–4), 221–232. (Originally published 1957.)Find this resource:

                                                                                                                                                                                                            Voegelin, C. F. (1956). Phonemicizing for dialect study, with reference to Hopi. Language, 32(1), 116–135. doi:10.2307/410660Find this resource:

                                                                                                                                                                                                              Voegelin, C. F., & Voegelin, F. M. (1963). On the history of structuralizing in 20th century America. Anthropological Linguistics, 5(1), 12–37.Find this resource:

                                                                                                                                                                                                                Weinstein, E. A. (1986). Sapir’s classification of symbolism. In W. Cowan, M. K. Foster, & E. F. K. Koerner (Eds.), New perspectives in language, culture, and personality (pp. 483–493). Amsterdam: John Benjamins. doi:10.1075/sihols.41.19weiFind this resource:

                                                                                                                                                                                                                  Wells, R. S. (1947). Immediate constituents. Language, 23(2), 81–117. doi:10.2307/410382Find this resource:

                                                                                                                                                                                                                    Whorf, B. L. (1956). Language, thought, and reality: Selected writings of Benjamin Lee Whorf (J. B. Carroll, Ed.). Cambridge, MA: MIT Press.Find this resource:


                                                                                                                                                                                                                      (1.) Trager’s presidential address of 1960 (published 1963 in Studies in Linguistics), gives a different characterization of the exercise. In the Preface, Trager writes that “Presidential addresses [to the LSA] . . . are often general, discursive, admonitory, retrospective, and even lightly humorous, though they may also be highly technical or specialized” (p. 4). Hockett (1993, p. 785) seems to differ, and wrote that “it is difficult to understand why, for his Presidential Address to the LSA in 1960, George [Trager] chose such a nontechnical, presumably popularizing style” granted the high rigor of his work in general. Trager (1963) and Hockett (1993) both note that the editor of Language (Bernard Bloch) declined to publish Trager’s text. In my own reading, “general, discursive, admonitory, retrospective” presidential addresses were few.

                                                                                                                                                                                                                      (2.) A few audio recordings by LSA presidents exist (e.g., Zellig Harris’s 1986 Bampton lectures at Columbia University), although none contain presidential addresses from the relevant historical period.

                                                                                                                                                                                                                      (3.) It is regrettable in particular that Bernard Bloch’s presidential address of 1953 is not part of the corpus examined here. Not only was Bloch an important post-Bloomfieldian, but, as editor of Language from 1940 to 1965, he exerted considerable influence over the development of American structuralism. Hill (1991, p. 77) reported that Bloch withheld publication of his presidential address in Language, deliberately breaching convention, so as to set a precedent for non-publication of future presidential addresses.

                                                                                                                                                                                                                      (4.) The LSA’s flagship journal Language is an outstanding, invaluable, accomplishment initiated by the founders of the Society. However, it is not discussed here because, by design and from its inception, Language is not limited to publishing the work of American structuralists. Its successes are the successes of the discipline worldwide.