1. Thanks are due the West European Studies Program at Indiana University, which paid for research assistance and for computer time. Justice, David supervised much of the collecting and grouping of data, and David Kelly wrote up a very thorough initial report on the results. I owe much to these two former students.Google Scholar

2. In the case of the (as yet incomplete) Neue deutsche Biographic, vols. I–IV (Berlin, 1953–1964), all pertinent entries through Grasman are covered. The Dictionnaire de biographie française, vols. I–X (Paris, 1933–1961, also complete only through Dallière) treats many more persons per letter than its German counterpart, and many of its descriptions are very brief, so that coverage was restricted to entries at least 25 lines in length. In the case of the Dictionary of National Biography (London, 1917, with supplements 1927, 1937, 1949, 1958, 1959) and in that of the Dictionary of American Biography (New York, 1928, with supplements 1944, 1958), entries considered had to be at least a page in length; they were taken in alphabetical order from the original sets and the various supplements, until total samples of 1001 had been obtained. (The intention had been to stop at 1000.) No attempt was made to exclude foreigners, and the British encyclopedia covers Scotland and Ireland as well.Google Scholar

3. Biographical dictionaries cannot be written by one or two people. They also tend to draw upon such publicly available materials as newspaper obituaries, older encyclopedias and published biographies. To that extent, they are collective enterprises. The generalities their editors themselves advance about the choices they make, on the other hand, can not be taken very seriously.Google Scholar

4. The occupational categories used in Table 1 are groupings of initially more detailed breakdowns. ‘Writers, Arts, Private Scholars’ covers all full-time writers outside the sciences, journalists, editors and publishers, along with the creative and performing arts, and a few private scholars or intellectually active “rentiers.” ‘Academics’ includes academic secondary school teachers, librarians and archivists. ‘Military, Landowners’ are predominantly officers, though a few are owners, managers or lease-holders of large estates. ‘High Officials’ includes judges and diplomats. ‘Politicians’ covers full-time elected politicians, whether national or local, along with leaders of political parties and labor unions. ‘Medical and other Liberal Professions’ are chiefly medical men. ‘Entrepreneurs, Technical Professions’ are owners, managers and (rarely) executive employees of large enterprises, along with somewhat smaller contingents of scientists, engineers, inventors, explorers, architects and technicians who were not explicitly linked with business or the academic world.Google Scholar

5. When each of the four samples was divided into two chronological halves, the academics were everywhere found to have increased their representation during the second half of the period. Other results obtained by way of chronological breakdowns, however, were not statistically or analytically significant.Google Scholar

6. Within the category ‘Writers, Arts, Private Scholars,’ journalists and publicists were a good deal more numerous in the American than in the French sample.Google Scholar

7. All but the last three occupational groupings are identical with those in Table 1. Yet it must be remembered that the higher ranks (i.e. senior university professors among ‘Academics’ and officers above the rank of colonel within the military) inevitably predominate among the elites described in Table 1, whereas the lower-ranking and more numerous groups (i.e. secondary teachers among ‘Academics’) must be considered typical when dealing with the fathers. A few skilled workers are counted with the artisans, and ‘Lower Officials’ include teachers below the secondary level.Google Scholar

8. Examples are: “did his law in Paris,” “took up medical practice after completing his studies,” and “studied at Heidelberg and Berlin.” Google Scholar

9. The specific percentages for the European countries may be found in Tables 3–5 below; but the American figures are cited only in the text.Google Scholar

10. Percentages by column are superfluous for two reasons. First, it is possible to obtain information directly about the social origins of students at various European institutions of higher education. Second, percentages by column for particular institutions would deviate from the norm for all institutions exactly where the percentages by row for particular occupations deviate from the norm for the sample as a whole. Of course one cannot read percentages by row as though they were percentages by column; one cannot forget, for example, that the German technical institutes accounted for many more academics and secondary teachers, in absolute terms, than high officials.Google Scholar

11. Here in the following tables, university-level education is broadly defined to include attendance at technical institutes and similar institutions even before they were officially considered of university level. ‘Type Not Known’ covers men for whom the sources indicated university-level education without specifying the type; ‘University Law, Humanities, Theology’ includes the social sciences as well; ‘University Medicine, Science’ also includes pharmacy, dentistry, and mathematics, along with a handful of cases of university study of “business.” ‘Technical Institutes’ covers “polytechnical” schools along with their heirs, the technische Hochschulen, even before the latter were granted university status in 1899. ‘Professional Academies’ includes art, music, and other professional and technical academies (mining, agriculture, forestry) other than the technical institutes.Google Scholar

12. The classification of types of university-level education was difficult for France. Around four-fifths of cases under ‘University Law, Letters’ are law students, just as about four-fifths of those under ‘University Medicine, Science’ are actually in medicine or (in a few cases only) pharmacy. Thus university students in the humanities and sciences properly speaking are very poorly represented indeed, and about many of them one actually knows only that they were tested and certified (licenciés, agrégés) by the faculties of letters and of sciences. Students who attended the Ecole Normale and similar institutions while also earning faculty certification were included under the institutions they actually attended. ‘Ecole Polytechnique’ covers all those who ever enrolled there, whether or not they afterwards went on to more specialized schools (écoles d'application, Saint-Cyr). ‘Ecole Normale’ means Ecole Normale Supérieure (Paris) alone. ‘Other Grandes Ecoles’ are a mixture, unfortunately. The classification is dominated by the higher art and music schools (the Ecole des Beaux-Arts and the Paris Conservatory, with 125 cases jointly). At the same time, it includes the Ecole des Chartes (55 cases), other high-level scholarly institutions in Paris (15 cases, including the Ecole Pratique des Hautes Etudes, the Ecole des Langues Orientales, the College de France), the Ecole Libre des Sciences Politiques (Sciences Po, 11 cases), the Ecole Centrale des Arts et Manufactures (Ecole Centrale, 29 cases), and a whole cluster of less distinguished technical-professional and higher commercial schools (43 cases, including Arts et Métiers, Agronomique, Forestière, Vétérinaire, and Normale Cluny). Taken together, these institutions account for 16.5% of specified university-level educations in the French sample, and of that 16.5 one should assign about 7.4 to the art schools, 4.2 to the philological and scholarly institutions, 0.7 to Sciences Po, 1.7 to Centrale, and 2.5 to the remaining technical-vocational group. ‘Religious’ covers theological faculties, the Catholic Institute, a large share of higher seminaires (grandes seminaires) and related institutions that appeared to be clearly above the level of the lower seminaires (petits seminaires), and ‘Military’ is made up chiefly of Saint-Cyr, and the Navale, Ecole.Google Scholar

13. ‘No University Level Education’ is defined as in Tables 3 and 4, but those few individuals who would have been listed as ‘Type Not Known’ in the German and French samples were grouped with the ‘Other Universities’ in the case of the United Kingdom. This category of ‘Other Universities’ includes all universities (even outside Great Britain) other than Oxford and Cambridge. No university-level educations were assumed, in the case of the United Kingdom, on the basis of profession alone. The slight difference between the approaches to the British and the Continental samples is substantially justified, in that no profession as certainly required university-level certification in the United Kingdom as some did in Germany and France during the nineteenth century.Google Scholar