American structuralist phonology NEO OR POSTBLOOMFIELDIAN LINGUISTICS BETWEEN

American structuralist phonology NEO- OR POST-BLOOMFIELDIAN LINGUISTICS (BETWEEN BLOOMFIELD’S 1933 LANGUAGE AND THE LATE 1950 s)

intro • no central figure, but a variety of individuals who developed issues whose roots can be found in Bloomfield’s earlier statements • the first generation to be employed as professional linguists • linguistics claims to have a uniquely scientific approach to the study of language, based on the empiricist, logical-positivists views of contemporary philosophy of science

post-Bloomfieldians • Bernard Bloch, Charles Hockett, George Trager, Henry Lee Smith, Archibald Hill, Martin Joos, Rulon Wells • solidarity and community of interests; they felt themselves to be a group of crusaders with a common mission • Zellig Harris: supporter of rigorous and purely distributional methods; personally outside

successors of Sapir • Morris Swadesh, Mary Haas, Stanley Newman, Carl Voegelin • attacked Sapir’s mentalism

critics of Am. structuralism • Eugene Nida and Kenneth Pike, Charles Fries • developed practical methods for investigating unfamiliar lgs • they were especially against the (Bloomfieldian) rejection of meaning • unfortunately, the lack of prestige of religiously motivated fieldwork (Nida and Pike)

The American structuralist view of language (1) • objection against a deductive explanatory system based on language universals (like that of the Prague School) • universals should (if at all) be discovered as purely inductive generalizations • a consequence of Boas’s “languages could differ from each other without limit and in unpredictable ways”

The American structuralist view of language (2) • Bloomfield’s rejection of semantics • linguistics must proceed without reference to the mind • no theorizing about the psychological implementation of linguistic structure • Bloomfield’s rejection of theoretical status for phonetics within linguistics • Trubetzkoy’s “phonetics is to linguistics as numismatics is to economics” • linguistics thus cut off at both sides (semantics and phonetics) • [interestingly, Hockett 1955 and Pike 1943 - treatments of phonetics, Joos 1948 acoustic analysis in phonetics]

The American structuralist view of language (3) • the only acceptable ‘general principles’ were inductive generalizations based on corpus of descriptions • structural descriptive • extensive fieldwork, description for its own sake • prohibition against ‘mixing levels’ (i. e. allowing considerations from higher levels of structure to play a role in determining the relation between phonemic and phonetic representations) • phonological representation could only have properties recoverable from the bare signal itself

The phoneme (1) • Swadesh 1934: the inductive procedure for discovering the phonemes of a lg begins with phonetic facts • first, abstract from free variation • if complementary distribution, then subtypes of the same phoneme, acc. to phonetic similarity • completely specified basic variant theory • phonemes should be organized into a system so as to maximize ‘pattern congruity’ (e. g. symmetry along the phonetic dimensions of contrast; segments that share distributional properties are assumed to be similar)

The phoneme (2) • Twaddell 1935: challenged both conceptions of the phoneme common at the time (i. e. Sapir & B. de Courtenay’s psychological view vs. Jones’ physicalist view) • phonemes are fictitious units to express contrasts in a lg as a system for transcribing utterances in that lg • no mental reality (it’s rather informant’s failure to make a distinction which a trained phonetician would make): what you see is all you get

The phoneme (3) • Bloch 1941: phonemic overlapping • partial overlapping, like flap assigned to /t/ in butter, and to /r/ in throw - no confusion • complete overlapping, like final vless obstruents in German, Russian, Polish; such cases must logically be excluded - a necessary condition on phonemic analysis regardless of the coherence of the resulting description; later called biuniqueness (Harris 1944) • also, thus, restriction that no grammatical facts (morph. or syntactic) can be used in phonological analysis (‘mixing levels’ prohibited)

The phoneme (4) • Pike 1947, 1952: grammatical prerequisites to phonemic analysis - thus against the above prohibition, the role of grammatical boundaries • others did not accept Pike’s view, so an alternative: juncture phonemes /+/ (distinctive conditioning effect on other phonemes)

American structuralist morphophonemics (1) • relatedness between forms which is not deducible from the facts of pronunciation alone must lie outside of phonology • morphophonology or morphophonemics (acc. to Swadesh 1934): Ø the study of the phonemic structure of morphemes (morpheme-structure rules in generative grammar) Ø the study of interchange between phonemes as a morphological process

American structuralist morphophonemics (2) • tendency to eliminate dynamic process-like statements deriving forms from one another or from more basic forms, in favour of static descriptive accounts characterizing the range of occurring forms • the method for ‘morpheme alternants’ (Harris 1942): “every sequence of phonemes which has meaning & which is not composed of smaller sequences having meaning is a morpheme” • e. g. knife-knives /nai. F/- a morphoph. symbol /F/ is an abbreviation for a set of alternating phonemes and for the alternation itself: phoneme /v/ before plural /z/, phoneme /f/ elsewhere (parallel to morphophonemes posited by Trubetzkoy)

Rule interactions and the nature of descriptions • Bloomfield: rules of a morphophonemic analysis are applied in a sequence (descriptive order), each applying to the result of the previous one • Sapir: sequence predictable from general principles; rules having access to organic and inorganic elements (later: nonderived & derived forms) • only the forms are ‘real’; regularities are sth the linguist finds • a language = a hierarchy of inventories and not a structured cognitive system