THE WORK OF MARY haas has a special fascination for me, since she finished her doctoral dissertation on Tunica, an American Indian language, in 1935, the same year that I started my studies of linguistics (with the Summer Institute of Linguistics) and went to Mexico to study an Indian language (Mixtec). This was the explosive age of descriptive linguistics, which we shared and which was especially focused on American Indian languages. She studied with Sapir and some of the other leaders, as I did (I got my clue to the analysis of tone from Sapir at one of the early summer sessions of the Linguistic Society of America at the University of Michigan).
Haas's first article (on Nitinat spoken on Vancouver Island) was published in 1932 jointly with Morris Swadesh, her husband from 1931 to 1937, whose articles on phonemics in 1934 and 1937 were useful to me, supplementing work by Leonard Bloomfield and Edward Sapir. Haas's early concentration was on the description of American Indian languages of North America; later she and other descriptive linguists shifted their attention to the east during the war to help the U.S. Armed Forces understand languages that had not been well known to Americans.
I asked Paulette Hopple, who worked in Thailand for many years with the Summer Institute of Linguistics, for a comment on Haas. She replied, "I first met Mary Haas in 1979 at the Sino-Tibetan Conference in Paris, where we discussed numeral classifier systems in Mayan, Thai, and Burmese. Although delighted and awed by her knowledge and experience in linguistics, what intrigued me about Mary was her personal approachability and humility. She communicated personal interest, compassion, a gentle sense of humor, including an ability to laugh [at difficult] circumstances."
Haas was born in Richmond, Indiana, graduated there from high school and college, did graduate work (1930-31) in Chicago on comparative philology, and did her Ph.D. in linguistics (1931-35) on the American Indian language Tunica at Yale. After that, she carried on various research tasks on American Indian languages under the anthropology department at Yale and the American Philosophical Society, 1935-41; on Thai, 1941-45, under the American Council of Learned Societies; and research in connection with her appointments at the University of California, Berkeley, 1946-53. Along with her research, she had various regular university appointments at Berkeley: lecturer in Siamese (Thai) for an Army training program, 1943-44; lecturer in Siamese and linguistics, 1947-53; associate professor, 1953-57; professor, 1957-77; acting chairman of linguistics, 1956-57; and chairman, 1958-64. She had numerous short-term (e.g., summer or one-semester) appointments for lectures in anthropology, Thai, or linguistics in various places in the United States and Canada. She received honorary doctorates from Northwestern University (1975), University of Chicago (1976), Earlham College, Richmond, Indiana (1980), and Ohio State University (1980). She was a member of various professional societies, including the National Academy of Sciences (1978). In addition, she was vice-president of the Linguistic Society of America in 1956 and president in 1963.
Because of my own interest in phonology for the years 1935-50, I will start by discussing a few of the phonological issues that Haas faced in studying the North American language Tunica (an "isolate" with historical relationships not clear). The consonants of Tunica (Haas 1941, pp. 13-14) have one surprise: the voiced stops /b, d, g/ occur "only in a few isolated words (of foreign or probably foreign origin)"; but the voiceless fricatives have no voiced counterparts. Each syllable begins with a consonant, some end with a consonant, some clusters of two consonants come medially in a word, and some consonant clusters may be preceded by /n/. Vowels are normally short, unless in stressed syllables.
In Tunica (1941, pp. 19-20) stressed syllables with their pitch relations are also of interest to me. The first stressed syllable of a phrase may be stronger than the unstressed ones and is often (but not necessarily, and not with semantic implications) a bit higher in pitch. Stressed syllables as a whole, however, enter into various "phrasal pitch contours," or "melodies." In them, a final stressed syllable may be a bit higher than the penultimate one, or the final one may have a falling melody, or it may have a rising one, or a falling-rising one, or may be lower than the preceding syllable. Some monosyllabic prefixes (and some other forms) have special phonological rules (pp. 20-34), which I do not summarize here. The predicative word of a main clause (p. 89) will have high melody if it is indicative, low if quotative, rising if interrogative, and falling if imperative.
With her Tunica (1941) Haas has some morphemic analysis of long words. This is interesting, since there are numerous words with up to six syllables in agglutinative arrangement. For example (p. 52), "The semelfactive paradigm consists of a causative stem plus the semelfactive forms of the causative auxiliary." For example:
For a full text with detailed analysis see pp. 135-43. Unfortunately, in her presentation it is often very difficult for the beginner to see where morphemes in a word or phrase begin or end.
Haas has a discussion of syntax (1941, pp. 89-134) with texts illustrated on pp. 135-43, and in 1950 with extensive texts (with notes giving morphemic analyses). She discusses (1941, pp. 90-91) simple versus compound and complex sentences (with compound ones having two or more main clauses and complex ones having a main clause plus one or more subordinate clauses of dependent, complementary, relative, or adverbial types). The following illustration is a simple sentence with just one clause:
Clauses (pp. 91-93) are of two principle types, main and subordinate. Subordinates are dependent (subordinate only to the main verb), complementary, relative, and adverbial. The dependent clauses have a subordinating postfix on the predicate (details on pp. 91-102; noun classification for gender and number, pp. 102-03; preverbs and postfixes, pp. 114-26; word classes in syntactic uses, pp. 126-34, including exclamatives and imitatives, p.134).
Texts (1950) include myths (solar, thunder, origins of corn or beans); tales (about eagles, owls, submarine people); animal stories; historical or pseudohistorical narratives (revenge, migrations, robberies); personal narratives (about families, or rabbits, or war); ethnological data (about food types, house construction, fire, fever remedies, shooting ghosts); and miscellaneous (one-eyed beings, water animals, woodpeckers, the ocean dried up). These occur in Tunica in English translation with footnote alternative literal translations or comments.
In 1941 Haas started fieldwork in the phonology and syntax of Thai (Siamese) at the University of Michigan because of the need for speakers of Asiatic languages as war developed. She had help in this from the American Council of Learned Societies. (She married one of the speakers of Thai, Heng R. Subhanka; they were divorced some years later.) While doing research on Thai (1942-43), she was concurrently an instructor in oriental languages at the University of Michigan. Moving to the University of California, Berkeley, she lectured on Thai for the Army Specialized Training Program. Her publications on Thai were considerable, for example, Spoken Thai, book I in 1945 and book II in 1948, with Heng R. Subhanka and the Thai-English Student's Dictionary (1964).
In 1958 Haas discussed the tones of four Thai dialects with tone patterns differing in their relation to consonants and to geographical occurrence. Thai itself has high, mid, low, rising, and falling tones (the dialect of Nakhonsit-hammarat has seven tones). Proto-Thai presumably had four tone categories, the first three of which were found only with a syllable having a long vowel, semivowel, or nasal, while the other occurred only with syllables having a final stop; and the initial consonant (voiceless versus voiced) conditioned the development of the different tones. For Thai, note:
Many Thai words are comprised of single syllables. Some samples were given above in the illustration of tones. Many more are given in Haas 1955. Some of these combine to make complex words. For example (p. 264):
Khwaam: the sense, substance (as of a letter), but in special usage often placed in front of a verb to form an abstract noun. May often be translated -ness, ity, -th, -tion, etc. Thus, khwaamklua is "fear."
However (1962, p. 49), "Since Thai is conventionally written without any spaces between words, the English-speaking student has no clue as to which elements form a semantic unit and which do not." As for syntax (1964, p. xx): "The typical sentence contains subject, verb, object, in that order, e.g.,
In 1969 Haas wrote a book on the prehistory of languages in relation to general principles. This includes phonological types of change, morphological reconstruction, problems of classification, and diffusion. Included also are some tables for Wiyot-Yorok-Algonquian-Gulf (p. 62), proto-Muskogean (p.42), Algonkian and Yurok cognates (p. 67), and pre-Muskogean and Tunica (pp. 63-64). As indicated above, Haas treats Tunica as an isolate without strongly provable relations to other languages, but she suggests that Tunica may be related to the pre-proto-Muskogean, based not on detailed lexical evidence, but (pp. 63-64) on some of its partially similar affix features.
In a manual published in 1945 (and reprinted in 1978), Haas and Subhanka wrote that "Prosecution of the war created the need for these materials to teach spoken language." (The material was especially indebted to Henry Lee Smith , Jr., of the Language Section in the Education Branch and liaison with the Intensive Language Program, through J. Milton Cowan.) The sections include basic sentences, new words (and "how to take apart the words and phrases . . . and to make new words and phrases on the same model"), hints on pronunciation, and "a number of new ways of saying things." Part one (in book one) includes "Getting around," "Buying things," "Meeting people," "Family and friends," "What do you do for a living?" and 'Review." "Part two" includes "How do you like the weather?" "Getting a room in a hotel," "Getting dressed," "Let's go eat," "A shopping trip," and "Review." Part three (in Book Two): "On the train," "At the beach," "Let's go to the game," "Making a call," "At the play," and "Review." Part 4: "Getting a passport," "At the university," "Going to the doctor's," "The bank and the post office," "Home and neighbors," and "Review." Part Five: "Geography," "Agriculture," "Industry," "Government," "The country and its people," and "Review." The materials are on phonograph records for practice hearing.
In a more theoretical article Haas (1953) discusses the important relation of linguistics to language teaching. She mentions the work of Boas, Sapir, and Bloomfield, and how the earlier descriptive work is still needed, but it needs approaches to teaching applications (as Bloomfield tried to show). Reading is not enough. Memorization of paradigms is not enough; conversational teaching is needed. For many students beginning the study of a foreign language, these approaches should best precede detailed analytical work. She mentions also some materials of personal interest to me (e.g., the work of Charles C. Fries and the English Language Institute at the University of Michigan, where I worked for a time on the intonation of American English, which she also refers to).
In a short, excellent, and easy-to-understand earlier article (1943), Haas gives instructions to help the beginning student learn any language. The linguist can learn a language quickly (p. 202) by working with "a native speaker, whom he treats not as a teacher but purely as a source of information." But the linguist must also teach the student the techniques of eliciting information, and analyzing and organizing the speaker's data, including appropriate phonetics. And (p. 205) "many of the fundamental features of the [analytical] method were first developed in the study of American Indian languages, which often present unusually difficult phonetic and grammatical systems." The students may either "participate with the linguist in the analysis of the language to be learned" or (p. 206) the linguist analyzes in advance and acts "as a model for imitation," teaching "instead of only guiding."
Haas's students (pp. 206-207) in Thai are "first of all taught to use a phonemic notation" so that "they may concentrate on the pronunciation from the very beginning," and they can use it to "carry on all their work. . . . This practice reflects one of the basic assumptions of our method: SPEAKING MUST COME BEFORE READING." About half of the student's time goes into "more traditional" work with "grammatical discussion, word study, translation of Thai into English and English into Thai," and drill on "troublesome points of grammar." The other half goes into drill "to develop good pronunciation and, later on, the ability to converse in Thai." The drill consists of three kinds: (1) exercises of imitation to train students to imitate exactly "so that they may be understood"; (2) exercises of dictation so that students may write down what they hear, improve perception, and "record new words even before they have learned the traditional native alphabet"; and (3) exercises of recognition and response to "train them to understand and answer what they hear, so that they may gain experience in the actual use of the language as a means of social intercourse."
The dictation exercises are "intended to improve the student's ability to hear, not his ability to spell." And in early stages the informant dictates only words that the class have already studied; then "short sentences containing familiar words"; then "sentences containing new words"; and finally "whole passages with old and new words mingled." Later, the student learns how to correct mistakes "by comparing the troublesome feature of a new word with a similar feature of some word already known." And when the tone of a word is not clearly heard, it is studied by comparing it with a word "whose tone is known to him." He does the same (p. 208) for vowels, aspiration, etc. After the student "has learned several hundred words and has acquired reasonable facility in conversation through the use of the phonemic writing alone, then--but not until then--he begins to learn the traditional system of writing." Haas informs us (p. 208) that her experience tells her that students like to learn a foreign language this way. "It gives them a sense of reality and the assurance that they are actually on the way."
FOR THIS MEMOIR I have drawn heavily on Haas's curriculum vitae provided by the National Academy of Sciences, and I am grateful for a recent obituary by Golla, which includes a Haas bibliography of about 130 items (V. Golla. Mary R. Haas (obituary). Language 73(4):826-37).