Linguistic atlas of the Gulf states, vol. 3: Technical index Ed. by Lee Pederson, Susan Leas McDaniel, Carol Adams, and Caisheng Liao (review)

BOOK NOTICES 659 Greet Cotton & John M. Sharp (two papers), Margaret Hidalgo, and Lenard Studerus, whose papers find features or processes in USSp linking it to other varieties of Spanish, especially Standard Mexican Spanish. In this group the most interesting papers, by Rogelio Reyes and George Green, examine Pachuco Caló in terms of its European (gypsy) origins and its relation to other varieties, e.g. Argentinian Lunfardo. Green, especially, cites a number of sound-change patterns, semantic shifts, and lexical processes that point to a set of 'universal attributes of oral Spanish' (244). Other studies examine features which distinguish USSp from other varieties. John M. LiPSKi discusses several that distinguish Central American from Mexican and Mexican-American , while Richard Otheguy and Ofelia Garcia examine English loanwords and caiques in Cuban-American. Halvor Clegg & Robert Smead's computerized analysis of border Spanish finds that variety to be differentiated by its influences from Pachuco Caló and English. The remaining papers treat a number of other topics. Gregory Denning's treatment of ojalá (que) variants in Kansas Spanish is the first analysis of Spanish in the U.S. Midwest. Jorge M. Guitart's report of English vowel perception by Hispanic children is significant for English (especially the low back vowels) and Spanish phonology alike. And Jacob OrnsteinGalicia 's call for a 'networking' approach to major language research problems is useful in a time of straitened budgets (133). This volume is important not only for Hispaniste but also for students of language use in the U.S.A. [Timothy C. Frazer. Western Illinois University.] Linguistic atlas of the Gulf states, vol. 3: Technical index. Ed. by Lee Pederson, Susan Leas McDaniel, Carol Adams, and Caisheng Liao. Athens: University of Georgia Press, 1989. Pp. xxiii, 435. $60.00. The third of a projected seven volumes of the Linguistic Atlas ofthe GulfStates (LAGS), the Technical Index ( TI) constitutes one component of a book analogue to the taped field interviews that constitute the complete data base to LAGS. Other analogues include an 'Automatic Atlas in Microform (AAM)', which is available on a single optical disk or several computer diskettes. and The Basic Materials, which have been published in microfiche. The AAM contains not only files of all edited data, but also sorting and mapping programs which allow 'virtually instantaneous analysis' (ix) of the LAGS evidence of regional dialect. Other components of the book analogue include the previously published Handbook (Vol. 1) and General Index or concordance (Vol. 2), as well as four cartographic volumes scheduled for publication. The LAGS data base has been organized into 390 lexical, 290 grammatical, 423 graphophonemic , and 74 systematic phonetics files which record the responses from 914 primary consultants . An urban supplement adds 201 lexical files, bringing the total number of responses to approximately 1 .2 million. To facilitate the analysis of this massive amount of data. Pederson et al. have encoded the data base as entries for microcomputer files; the TI provides a summary of these computer ASCII files. It is important to note that these various analogues are interdependent : thus the General Index and Technical Index summarize data presented in The Basic Materials, but do not give specific occurrences of the data. Rather, they serve as a guide to the materials presented in microfiche. And, as a guide, the 77 is suggestive rather than definitive. For example, the lexical files contain 159 lexical equivalents of salt pork. Presented in alphabetical order from Alabama ham to whiteback, each of which occurs only once as a consultant's response, the list reveals thatfatback (335 occurrences) is more than twice as common as each of the other responses that occur more than 100 times, salt pork ( 153 times), sowbelly (127). and salt meat (121). Whether these terms represent significant geographical or social distributions, however, cannot be determined from the data in the TI, for only frequency of occurrence is given. The TI presents phonological data in one of two forms, as 'Phonological Files' recorded between carets which indicate graphophonemic strings or as 'Systematic Phonetic Files' which give the phonetic features of fifteen stressed vowels. The Phonological Files are relatively transparent. The first such file, for one, contains the entries (wawn), (wo[n]>, (woe[n...