La Nouvelle-Orleans: Croissance demographique, Integrations urbaine et sociale (1803-1860), by Maijorie Bourdelais. Bern, Peter Lang Publishing, 2012. xxvi, 472 pp. $112.95 US (paper). The dissertation that Marjorie Bourdelais defended in 2007 at the EHESS in Paris, now published by Peter Lang press as the sixteenth volume in its series Population, Family, and Society, challenges the prevailing wisdom regarding how long it took for English to replace French as the dominant language in New Orleans. Taking exception with the notion that arrival en masse of Irish and German immigrants marginalized the Gallic community after 1840, she argues: Acquise par les Etats-Unis en 1803, la Nouvelle-Orleans presente la particularite de demeurer majoritairement francophone pendant un demi-siecle (p. 128), or until 1860 in her conclusion (p. 366). In support of this thesis, the author presents numerous tables and graphs based on an echantillon exhaustif ' of 72,203 heads of households enumerated in the six decennial federal censuses from 1810 to 1860. Although the conclusion refers to identification of the linguistic origin of the totalite des chefs de menages neo-orleanais (quel que soit leur sexe, age. origine, couleur de peau) (p. 363), she claims elsewhere to have found evidence in other records of the linguistic of three-fourths of them. Their classification as francophone, anglophone, germanophone, hispano-italophone, or other is based on the primary language in the country of origin for immigrants or, for the first generation born in Louisiana, of their parents. For a second database of 1,063 marriages of persons whose names began with the letter B in three sets of years (1806-10, 1836-40, and 1848-50), she rightly remarks that covering all years from 1803 to 1860 would have been hors de portee d'un travail individual (p. 132), leading one to wonder how she created the census database of linguistic for all heads of households with 70 times as many records. Despite description of her sample as a recherche patronymique, suggesting inference of language from surnames, use of birthplace is a much more robust method of determining linguistic origins for which she is to be commended. It is nonetheless subject to bias resulting from richer genealogical sources for Catholics than for non-Catholics in antebellum New Orleans. No attempt is made to measure the odds of finding information on birthplace for different nationalities, or of the relative frequency of last names starting with B (in the 1860 census sample cited below, for example, 15% of francophones, 8% of anglophones, 10% of germanophones). Comparison of Bourdelais's statistics for the city as a whole with IPUMS census samples is one way to test their credibility, allowing for margins of error of 7% in the 1850 and 5% in the 1860 extracts for New Orleans. Since the census reported birthplace in these years, the percentage of cases where linguistic cannot be determined should be lower than in preceding years; yet it is slightly higher (p. 375), another reason for concern about methodology. The percentage of household heads in each linguistic group should be greater by however many native Louisianians had parents of that group. 1850 1850 1860 1860 IPUMS BOURDELAIS IPUMS BOURDELAIS Francophones 21% 49% 14% 39% Angophones 40% 38% 39% 41% Germanophones 23% 11% 21% 17% Hispano-Italophones 3% 2% 5% 2% Others 3% 0 2% 0 Louisianians 11% 20% Total 100% 100% 100% 100% In 1850, even if all Louisianians were francophones, the latter were still only 32% of household heads in the IPUMS sample. …