¹ Goodrich, Carter, ‘Indenture’, Encyclopaedia of the social sciences (New York, 1932), VII, 646Google Scholar; Morris, Richard B., Government and labor in early America (New York, 1946), p. 315Google Scholar; also see Gray, Lewis Cecil, History of agriculture in the southern United States to 1860 (New York, 1941), I, 348Google Scholar; Smith, Abbot Emerson, Colonists, in bondage (Chapel Hill, 1947), pp. 3–4Google Scholar; Hofstadter, Richard, America at 1750 (New York, 1973), p. 34Google Scholar.

² Smith, Abbot Emerson, ‘Indentured servants: new light on some of America's “first” families’, Journal of Economic History, II, 1 (05, 1942), 46Google Scholar; Campbell, Mildred, ‘Social origins of some early Americans’, in Smith, James Morton (ed.), Seventeenth century America (Chapel Hill, 1959), 76Google Scholar. For views consistent with Smith's, see Jernegan, Marcus, Laboring and dependent classes in colonial America, 1607–1783 (New York, 1965), p. 51Google Scholar, and Wertenbaker, Thomas, Patrician and plebean in Virginia (New York, 1959), pp. 162–3Google Scholar. Campbell's conclusion has been accepted by Craven, Wesley Frank, White, red, and black: the seventeenth-century Virginian (Charlottesville, 1971), p. 8Google Scholar.

³ For discussion, see David Galenson, ‘“Middiing people” or “common sort”?: the social origins of some early Americans reexamined’, William and Mary Quarterly, forthcoming.

⁴ The former is the contention of Campbell, ‘Social origins of some early Americans’, p. 71. The hypothesis she offers in support, that omissions were due to the specific types of printed form assigned to different servants, is inadequate; see Galenson, ‘“Middling people” or “common sort”?’. However this does not necessarily mean the argument is incorrect; given the importance of the issue, a direct test of it is desirable.

⁵ Schofield, R. S., ‘The measurement of literacy in pre-industrial England’, in Goody, Jack (ed.), Literacy in traditional societies (Cambridge, 1968), pp. 311–25Google Scholar; Stone, Lawrence, ‘Literacy and education in England 1640–1900’, Past and Present, no. 42 (02, 1969), p. 98Google Scholar. Schofield specifically suggested that: ‘For the early nineteenth century…a measure based on the ability to sign probably overestimates the number able to write, underestimates the number able to read at an elementary level, and gives a fair indication of the number able to read fluently’; ‘The measurement of literacy’, p. 324. An extended discussion of this relationship, with additional evidence for the first half of the nineteenth century, is Contained in R. S. Schofield, ‘Some dimensions of illiteracy in England, 1600–1800’ (unpublished paper), pp. 2–5. Based on an investigation of the history of school curricula, Schofield further concluded that ‘the structure of education was basically the same at least from the sixteenth century, and it is therefore probable that the relationships between ability to sign, ability to read, and ability to write remained the same throughout this period’; ‘The measurement of literacy’, p. 324. I am grateful to Roger Schofield for access to the unpublished paper cited above.

⁶ Stone, ‘;Literacy and education in England’, pp. 103–12; Schofield, R. S., ‘Dimensions of illiteracy, 1750–1850’, Explorations in Economic History, X, 4 (summer 1973), 449–51Google Scholar; Vann, Richard, ‘Literacy in seventeenth century England: some hearth-tax evidence’, Journal of Interdisciplinary History, V (1974), 292Google Scholar; Cressy, David, ‘Literacy in pre-industrial England’, Societas, IV, 3 (summer 1974), 234–40Google Scholar; Cressy, , ‘Levels of illiteracy in England, 1530–1730’, Historical Journal, XX, 1 (1977), 5–10Google Scholar.

⁷ Most of these are held at the Greater London Record Office, Middlesex Section, as ‘Plantation indentures’, MR/E; they have been abstracted by Nicholson, Cregoe D. P., Some early emigrants to America (Baltimore, 1965)Google Scholar. An additional set held elsewhere was abstracted by Wareing, John, ‘Some early emigrants to America, 1683–4: a supplementary list’, Genealogists' Magazine, XVIII, 5 (03 1976), 239–46Google Scholar.

⁸ All of these are now in the ‘Memoranda of agreement to serve in America and the West Indies’, Corporation of London Records Office. Abstracts are contained in Jack, and Kaminkow, Marion, A list of emigrants from England to America, 1718–1759 (Baltimore, 1964)Google Scholar, and Galenson, David, ‘Agreements to serve in America and the West Indies, 1727–31’, Genealogists' Magazine, XIX, 2 (06 1977), 40–4Google Scholar.

⁹ Servants judged by the magistrate to be at least 21 could be bound before one justice of the peace, while minors were to be bound before two; Public Record Office, P.C. 2/69, ‘Privy Council registers, 69’, pp. 595–96, or Salley, A. S. (ed.), Records in the British Public Record Office relating to South Carolina, 1663–1684 (Atlanta, 1928), 1, 235Google Scholar. For convenience, J.P.s may have been lenient in their judgments when they were sitting alone in informal sessions, and many servants within a few years of 21 might pematurely have been promoted to adulthood for the purposes of the registrations.

¹⁰ The act was 4 I, George c. 11 section v, in The statutes at large, from the first year of the reign of king George the first, to the ninth year of the reign of king George the second, V (London, 1786), 114Google Scholar.

¹¹ Cressy, ‘Levels of illiteracy in England’, Table 1, p. 5. The rate in London was apparently considerably higher, particularly late in the century; Cressy, ‘Literacy in pre-industrial England’, p. 233.

¹⁴ This is very close to the national rate for women found by Schofield for the mid-eighteenth century; ‘Dimensions of illiteracy’, figure 2, p. 445.

¹³ The number of observations is small, and this difference is not statistically significant at the 0.10 level. However, for discussion of a similar finding for the men in the sample, in which the increase is significant, see below. (Except where noted, significance tests for differences of proportions throughout this paper are based on one-tailed t-tests.)

¹⁴ Cressy, ‘Literacy in pre-industrial England’, Table III, p. 234; Cressy, ‘Levels of illiteracy in England’, Table I, p. 5; p. 8.

¹⁵ Cressy, ‘;Literacy in pre-industrial England’, p. 238; Cressy, ‘Levels of illiteracy in England’, Table I, p. 5; p. 8.

¹⁶ Cressy gives the figure cited for ‘tradesmen and craftsmen’ in the 1680s in ‘Literacy in pre-industrial England’, Table IV, p. 235, and for ‘tradesmen’ in ‘Levels of illiteracy in England’, Table 3, p. II; he gives the same rate as an average for the category ‘tradesmen and craftsmen’ for 1580–1700 in ibid., Table I, p. 5.

¹⁷ For a discussion of some of the factors relevant to such a comparison, see Cressy, ‘Levels of illiteracy in England’, p. 4.

¹⁸ Because of this temporal difference, some consideration might be given to its effect on the comparison; were significant changes occurring in the extent of literacy in England during the mid-eighteenth century? Schofield found male literacy to have been roughly stable during the second half of the century; ‘Dimensions of illiteracy’, p. 446. No comparable time series evidence is available for the English population in the first half of the century. Within the London registrations, there is some evidence of an upward trend in literacy over time. A comparison of the men registered in the two periods 1718–29 and 1730–40 indicates that the overall male literacy rate rose from 62% to 76%. To a large extent this increase was due to shifts in both the occupational and age distributions of the men registered over time. While the correlation between age and literacy for the men in the sample will be considered below, Table A presents a comparison of the literacy rates separately for the minors and adults in each of the eight occupational categories considered above in the two periods. The comparison reveals increases in the literacy rates of 12 of the 16 resulting groups, decreases for three, and no change for one. Only three of the changes are statistically significant at the 0.10 level or better; however all of these were increases in the respective rates. Since little is known of the degree to which the servants were representative of all Englishmen, no generalization is possible to changes in English literacy in the period. However, the relatively small changes in the rates in most groups indicate the absence of major changes in literacy over time for the men registered, and make aggregation for the entire period useful for comparison to Schofield's rates.

¹⁹ This and the following figures are taken from Schofield, ‘Dimensions of illiteracy’, Table I, p. 450. As was done above in using Cressy's rates, Schofield's ‘literacy’ rates have been transformed into ‘literacy’ rates by subtracting them from 100%. Statistically the information conveyed is precisely the same.

²⁰ The weights used in this and the calculation reported in the following sentence are the reported numbers of observations on which Schofield's rate for each of the respective groups is based.

²¹ That the farmers and labourers among the indentured servants in both periods may actually have been more highly literate on average than other Englishmen of the same occupations suggests that the more skilled men from these groups might have been selected by merchants to be indentured; it is also possible that it was the more educated from these groups who were aware of the opportunities offered by the indenture system and chose to emigrate.

²² The difference is significant at the 0.01 level.

²³ The differences are statistically significant at the 0.01 level for three of the categories, at 0.05 for two, and 0.10 for one, while two fail to be significant at the 0.10 level.

²⁴ An interesting implication is the possibility that, though boys generally began to train for skilled trades at about 14, they did not begin to be identified with the trade until they had had as many as three or more years of training.

²⁵ This is suggested, for example, by Plumb, J. H., ‘How freedom took root in slavery’, New York Review of Books, XXII, 19 (27 11 1975), 4Google Scholar.

²⁶ The use of the terms ‘formal’ and ‘informal education’ throughout this discussion follows the usage of Schofield, ‘The measurement of literacy’, p. 315.

²⁷ The normal school-leaving age in this period was not above 15; Stone, ‘Literacy and education in England’, p. 99; Cressy, ‘Levels of illiteracy in England’, p. 11; Schofield, R. S., ‘Age-specific mobility in an eighteenth century rural English parish’, Annales de Démographie Historique (1970), p. 266Google Scholar. The youngest servants included in the two categories of husbandmen and labourers were 15, and there was only one of that age in each. The minor-adult differentials might therefore be estimates of the numbers of men who learned to sign after the school-leaving age, and may constitute lower-bound estimates of the number who learned to sign through means other than formal instruction in school. As noted above, Schofield's investigation of the relationship between the abilities of signing, reading, and writing for the period before the nineteenth century was based on the structure of formal education, specifically the order in which the skills were normally taught in schools; ‘The measurement of literacy’, p. 324. Schofield recognized the problem posed by informal education, as neither its structure nor its extent had been systematically investigated; ibid. p. 315. If it was widespread, and its order of instruction of literary skills often significantly different from that of the schools, it could upset the general validity of the relationships suggested by Schofield in ibid. p. 324. Neuberg, Victor E. has suggested that the ability to read was more common than the ability to write among the eighteenth-century poor, but as he notes quantitative evidence is lacking; ‘Literacy in eighteenth century England: a caveat’, Local Population Studies, no. 2 (Spring, 1969), pp. 44–6Google Scholar. The minor-adult differentials of Table 3 may give some indication of the extent of informal education in teaching the ability to sign to men from their mid-teens on.

²⁸ Evidence of variation by county for the mid-seventeenth century is given by Schofield,. ‘Some dimensions of illiteracy in England’, p. 12 and Table 5. These differences were persistent; for an indication of their magnitude in the first half of the nineteenth century, see Schofield, ‘Dimensions of illiteracy’, p. 444.

²⁹ Cressy, ‘Literacy in pre-industrial England’, pp. 235, 238; Schofield, ‘Dimensions of illiteracy’, p. 444.

³⁰ Vann, Richard T., ‘Quakerism and social structur e in the Interregnum’, Past and Present, no. 43 (05 1969), 74CrossRefGoogle Scholar.

³¹ On the basis of two-tailed t-tests, the rate for the tradesmen and craftsmen is significantly higher than that of those without additions at the 0.01 level, while the difference between the rates for those without additions and for the husbandmen and labourers fails to be significant at the 0.10 level.

³² Again using two-tailed t-tests, the rate for the tradesmen and craftsmen is significantly higher than that for those without additions at the 0.01 level, while that of the husbandmen and labourers is significantly greater than that of the men without additions at the 0.05 level.

³³ On the basis of two-tailed t-tests, the literacy rate of the men with additions is significantly higher than that of the others at the 0.05 level in the Middlesex sample, and at the 0.01 level in the London sample.

³ For an indirect indication of the economic productivity of servants with and without additions in the eighteenth-century London registrations which points toward the same conclusion as this evidence of literacy, see Galenson, David, ‘Immigration and the colonial labor system: an analysis of the length of indenture’, Explorations in Economic History, XIV, 4 (10 1977), 369, 371–2Google Scholar.

³⁵ For the basis of this conclusion, see Galenson, ‘“Middling people” or “ common sort”?’.