762 LANGUAGE, VOLUME 54, NUMBER 3 (1978) These two books typify the two roads open to sociolinguistic study at present. We often hear it asserted that there can be no objectivity or freedom from politico-economic bias, and that all scholarship is, of necessity, politically oriented. Such assertions are false. The world around us exists in and by itself; and it can be studied scientifically, without bias, by methods which give the same results for all investigators. A politicized approach, as exemplified in Salvi's book, degrades objective science into demagogic rabble-rousing, with nothingbut harmresultingfor boththe pseudoscientist and the public whom he deceives. [Robert A. Hall Jr., Cornell University.] The dictionary of the Spanish of Texas (Spanish-English). Compiled by Roberto A. Galván and Richard V. Teschner. Silver Spring, MD: Institute of Modern Languages, 1975. Pp. vii, 102. This book was written as a supplementary dictionary to be used 'alongside' a full length lexicon of Texas Spanish, such as several which are listed in its bibliography (Appendix Q. It contains 7,000 entries, of which several thousand come from other published sources, and several thousand more from extensive field notes on Texas vocabulary and proverbs, collected by the senior compiler. Two types of entries from other sources have been excluded here: (1) those considered part of 'standard' Spanish, according to the criteria of inclusion in monolingual and bilingual dictionaries; and (2) entries already found in a number of widely circulated bilingual dictionaries. In addition, several dozen entries appearing in secondary sources, but unknown to Galván, were discarded after attempts were made to trace them in standard lexicons of Mexican Spanish. While the dictionary does not exclude all entries which might be used elsewhere in the Hispanic world, its strength lies in its authors' effort to verify each entry as Texas Spanish; thus, while this is not, strictly speaking, a dictionary of regionalisms, it is a contribution toward a much-needed description of Texas Spanish. It contains many taboo and slang words, English loans, and toponyms—updating work such as the Vocabulario español de Texas of Gilberto Cerda, Berta Cabaza, and Julieta Farias (1953). A major shortcoming of the present work results from the interaction of the criteria of entry exclusion described above with the attempt to include phonetic variants of Spanish words found in Texas. Phonetic variants are commonly included here without a definition, e.g. 'icir (var. of) decir'. In more than half the cases, no corresponding base entry (one with a definition) is included, presumably because it is 'standard' Spanish or is found in widely circulated bilingual dictionaries. But in some cases, this procedure causes confusion. Note, e.g., the following separate entries (p. 46): 'influencia (var. of) influenza'; 'inflencia (var. of) influencia'. There is no separate entry for influenza, nor are there definitions for influencia or inflencia; thus the user is left wondering whether speakers of Texas Spanish have two coincidentally similar words meaning 'a type offlu' and 'influence' (as do speakers of other dialects of Spanish), or whether inflencia is a further phonetic variant of influencia, and there is some quite different way to say 'influence' in Texas Spanish. A related problem is found in entries on pp. 1 and 43, respectively: 'abuelito -ta (slang) friend'; 'güelito -ta (var. of) abuelito -ta'. In this case, the question is whether the phonetic variant is a variant only of the slang expression, or whether it is also a variant of the standard abuelo and its diminutive, meaning 'grandfather' in Texas Spanish. A solution to such problems might have been an appendix devoted exclusively to phonetic variants in Texas Spanish, in which the rules ofentry selection were relaxed. This procedure would have been comparable to the list of proverbs in Appendix A, where the criterion of inclusion is simply that an entry be found in Texas Spanish. Such an appendix would also have complemented Appendix B, in which -ear and -iar verbs are conjugated in order to show how the e\i distinction is eliminated because of word-internal vowel sandhi, on the one hand, and hyper-correction on the other. [Sandra Pinkerton, Berkeley.] Les débuts de la lexicographie française: Estienne, Nicot...