William Cowan

In Search of New England's Native Past: Selected Essays by Gordon M. Day. Ed. by Michael K. Foster and William Cowan. (Amherst: University of Massachusetts Press, 1998. xvi, 328 pp. Cloth,

This volume highlights the work of the late Gordon M. Day, renowned for his groundbreaking research on the history and culture of the Western Abenakis and their Indian neighbors. Where previous historians had tended to portray northern New England as an area largely devoid of aboriginal peoples, Day established beyond all doubt the presence of Abenakj settlements along the eastern shore of Lake Champlain as well as the upperreaches of the Connecticut and Merrimack rivers. For nearly three decades, Day focused his work on the community of Saint Francis, or Odanak, in Quebec, to which Abenaki refugees from interior New England had fled, beginning in the mid-seventeenth century and continuing into the nineteenth. Drawing on the methods of several disciplines, including ethnology, linguistics, and ethnohistory, he synthesized data from fragmentary historical records, oral traditions, and place names to reconstruct a world assumed to be lost.

70.00, ISBN 1-55849-150-3. Paper,

0. Introduction. A number of Algonquian languages are reported to use pitch as a phonemically distinct element in lexical and inflectional formations and to be, in effect, tone languages. Among them are the Plains languages: Blackfoot, Arapaho, in various dialects, and Cheyenne, also in various dialects (cf. Frantz 1972). In a footnote to his article, Frantz notes that two Eastern Algonquian languages, Penobscot and Abnaki, have been reported to have tone or pitch accent (but these reports are either unpublished or consist of remarks made in passing). A third language reported to have a possibly very restricted use of distinctive pitch is Montagnais, a Central Algonquian language closely related to Cree and spoken in Quebec. This article deals with some observations about the historical development of the feature of pitch in various dialects of Montagnais. As in Cheyenne, it appears that vowel length is associated with high pitch and vowel shortness with low; however, in at least one type of inflection, the development of high pitch is associated with the presence of a subphonemic glottal stop.

19.95, ISBN 1-55849-151-1.)

1. The earliest reliable record I know of Pequot-an Eastern Algonquian language formerly spoken in Connecticut-is that made in 1762 by Ezra Stiles, a Rhode Island clergyman who later became president of Yale University. This is a manuscript vocabulary list taken from a Pequot Indian at Groton, Connecticut, of some 159 forms, the last 50 of which are a selection of the numbers from 11 to 1000, consisting largely of combinations and repetitions. Also included are a few connected sentences and, on a separate sheet, a truncated free paraphrase of the Lords prayer. Like many records of these languages dating from the 18th century, this list poses some problems of interpretation. Stiles was an intelligent and educated man, familiar with languages other than English, and we must assume that he attempted to give as accurate and consistent an orthographic form as he could to the material he was faced with-this in

The Development of Suprasegmental Inflections in Montagnais

STORY, GILLIAN L. 1984. Babine and Carrier Phonology: An Historically Oriented Study. Summer Institute of Linguistics Publications in Linguistics, no. 70. Dallas: Summer Institute of Linguistics and the University of Texas at Arlington. WALKER, RICHARD. 1979. Central Carrier phonemics. Contributions to Canadian Linguistics, by E. P. Hamp et. al., pp. 93-107. National Museum of Man, Mercury Series, Canadian Ethnology Service Paper no. 50. Ottawa: National Museum of Man, National Museums of Canada.

Pequot from Stiles to Speck.

nonnumerical measurements or measurements that have only some numerical properties. I refer the interested reader to the inside front cover of Siegels book on statistics,l on which there is a taxonomic guide to nonparametric tests. Siegels text is also recommended as a follow-up to Anshens introduction. Anshens discussion of errors is also too lean. He highlights the risk of a Type II error, that is, the risk of overlooking, or not being able to detect, a real phenomenon. Decreasing the probability of a Type II error is desirable. The risk of a Type I error, however, may be more important in some studies than the risk of a Type II error. A Type I error is made when a phenomenon is incorrectly detected, that is, an observed effect is really due to chance. For example, a Type I error should be avoided in testing the effectiveness of a new drug, especially when alternative drugs are available. If a Type I error is committed, the new drug is said to be effective when it actually is not, and the patient may not be given other drugs which actually are effective. An investigator should be aware of both types of error, although, in a particular study, he or she may regard one type of error as more important to avoid than the other. There are several ways in which an investigator can decrease the probability of a Type II error. Anshen discusses two important ones: choosing an appropriate sample size and adjusting the criterion for significance. Adjusting the criterion for significance to decrease the probability of a Type II error, however, usually increases the probability of a Type I error. Thus, it is particularly important to be aware of the relative importance of the two types of error, and the trade-offs between them, in a given study. Two other omissions are minor, but nevertheless frustrating for a beginner in statistics. There should be answers given to the problems at the back of the book, and the tables at the back should be given more detailed captions; 1 Sidney Siegel, Nonparametric Statistics for the Behavioral Sciences (New York: McGrawHill, 1956). the tables are indecipherable to someone without any background in statistics. Anshen has accomplished a difficult task. He has distilled most of the basics about statistics into a short, readable book designed especially for linguists. It is a good introduction that should be supplemented by additional

A Double Obviative in Montagnais

comparative investigations. The organization of the material, using the roots as the basic criterion, is ideal for such purposes, especially for research within the Mayan linguistic family. In this field one of the main tasks is a reconstruction of Proto-Maya, while attempting also to establish a list of its metaphors. Such investigations which are of prime importance with regard to deciphering Maya script could make greater progress, if, for the remaining Mayan languages, there were dictionaries as well organized and as thorough as this great work of scholarship.1

Eleventh Algonquian Conference

conservative (particularly in contrast with the Southern New England languages of the archaic coastal subgroup). More important, perhaps, Powhatan also retains certain ancient vowel alternations (pp. 429-30). Following a discussion of consonant clusters (pp. 427-29; also p. 298), the remainder of this section is devoted to the recalcitrant and diverse sequences of vowels and semivowels. (7) The heart of Sieberts monograph is the lexicon which provides the basic documentation for most of the discussion. There are 263 entries each of which consists of three parts: (a) the reconstituted Powhatan form; (b) the Strachey and Smith records; and (c) ProtoAlgonquian and Proto-Eastern-Algonquian reconstructions and comparative evidence. I will bypass a detailed analysis of the reconstitutions and reconstructions, and will simply say that the former are convincing and that the documentation which accompanies the latter is unexcelled in depth and detail. Sieberts Powhatan lexicon is a major original step toward a future Proto-Algonquian dictionary. (8) In addition to its strictly linguistic value, the lexicon contains a wealth of material on related topics. In a masterly foray into semantic and cultural analysis, Siebert disentangles three numeral systems reflected in the attested Algonquian languages (pp. 303-11). The quinary system has elementary roots for the numerals from one to five and uses compounds of these roots with a complex final to express those from six to ten. The Proto-Algonquian final *-asyeka consists of a prefinal *-aOby the side, of the row of the second hand and an abstract final *-yeka-for example, *nyiswi two, *nyiswasyeka seven. The decimal system uses a slightly different abstract final, *-yekwi, as in *pesyekwi one; note the Proto-Algonquian root *pe0single. The decimal system is less fully attested than the quinary one. The descriptive system, which is substituted in some cases, shows differences in digital identification. For instance, Proto-Algonquian *peyakwi little finger (cf. the root *peyakwalone, by itself) is reflected by Cree peyak one, and as payak ten in the languages of southern New England. Siebert not only provides an ingenious analysis of a refractory linguistic problem of long standing; he also formulates a social hypothesis (p. 309) based on sexual dichotomization (e.g., in the division of labor and in residential separation) to account for the coexistence of three systems of approximately equal antiquity. A full and explicit summary and a new classification of the Eastern Algonquian languages complete Sieberts Powhatan monograph, which is a rare example of methodological and substantive excellence. Algonquianists and general linguists alike can look forward to further masterpieces.