1. A question asked by Oliver, but answered by no one, during the discussion of a paper presented by Guthrie: ASAUK, , Proceedings of the First Conference…Held at Birmingham University in September, 1964 (London, 1965), 35.Google Scholar (The proceedings were published as a special issue of African Affairs

2. Johnston, Harry H., The Kilima-njaro Expedition; a Record of Scientific Exploration in Eastern Equatorial Africa (London, 1886), 480.Google Scholar As will appear below, Johnston's ideas exerted a considerable influence on the thinking of British historians during the 1950s.

3. Hair, P.E.H., “Temne and African Language Classification Before 1864”, Journal of African Languages, 4(1965), 54–55.Google Scholar

4. Such false dichotomies are a common pitfall, not only in the field of language classification. In the jargon of phylogenetic systematics (see below, note 11), “Hamitic” is (or seems to be) a paraphyletic grouping, and therefore invalid: a group is only valid so far as it can claim to be monophyletic.

5. Greenberg, Joseph H., “Studies in African Linguistic Classification I-VII,” Southwestern Journal of Anthropology 5(1949), 79-100, 190-98, 309–17CrossRefGoogle Scholar; 6(1950), 47-63, 143-60, 223-37, 388-98; 10(1954), 405-15; idem, Studies in African Linguistic Classification (New Haven, 1955); idem, Languages of Africa (Bloomington, 1963; 2nd ed., 1966).

6. Bantu was the topic chosen for the third article in the series: Greenberg, “Studies III” (Studies, 33-42). The assertion that Bantu was related genetically to Westermann1s “West Sudanese” had in fact been made before in a paper written in 1946: Greenberg, , “The Classification of African Languages,” American Anthropologist, 50(1948), 24–30.CrossRefGoogle Scholar Apparently Greenberg began by suspecting that Bantu might be coordinate with “West Sudanic” as a whole, but soon he saw that it was related more closely to “Benue-Cross” than to any of the other branches.

7. Greenberg, , “The Tonal System of Proto-Bantu,” Word, 4(1948), 196–208.CrossRefGoogle Scholar

8. Except for the controversy over the position of Ma'a. The most recent discussion I have seen of this perennial problem is Thomason, S.G., “Genetic Relationship and the Case of Ma'a (Mbugu),” Studies in African Linguistics, 14(1983), 195–231.Google Scholar

9. Guthrie, Malcolm, The Classification of the Bantu Languages (London, 1948).Google Scholar Greenberg wrote a review of this book for Word, 5(1949), 81–83.CrossRefGoogle Scholar

10. The name “Central” was subsequently dropped in favor of Benue-Congo (Greenberg, , Languages of Africa, 7Google Scholar). Apart from the inclusion of Bantu, Niger-Congo differed from Westermann's “West Sudanic” in three major respects: it included Fulani; it included a large number of languages (in the branch called Adamawa-Eastern) extending eastwards from Nigeria; it did not include Songhai.

11. Here and elsewhere, Greenberg anticipated the line of thinking which, among zoologists, has come to be known as cladism. Despite differences in terminology, the underlying ideas are just the same: that the only strategy which in principle leads to a unique solution is one designed to produce a genetic (phylogenetic) classification, and that the recognition of valid (monophyletic) groups depends on the identification of shared innovations (derived characters), as distinct from shared retentions (ancestral characters). For access to the literature on this subject see Wiley, E.O., Phylogenetics: The Theory and Practice of Phylogenetic Systematics (New York, 1981).Google Scholar

12. At one time, Greenberg was hoping to publish “a full discussion of the evidence regarding the northwest boundary of Bantu” (Languages of Africa, 41n34). He has not followed through with this plan, so far as I am aware.

13. The facts are complex, and a brief description risks descending into caricature. In classes 1/2, Bantu has the prefixes *mu/*ba-, as against *u-/ba- in the rest of Benue-Congo; in classes 3/4, *mu-/*mi- as against *u-/*i-; and in classes 5/6, *li-/*ma- as against *li-/*a-. All these differences were cited by Greenberg in 1949 (“Studies III,” 312-13; Studies, 36-37) and interpreted as innovations on the part of Bantu. Only the innovation affecting classes 3/4 was mentioned explicitly in 1963 (Languages of Africa 35), but the others were soon reinstated by Crabb, with the approval of Greenberg: Crabb, D.W., Ekoid Bantu Languages of Ogoja, Eastern Nigeria (Cambridge, 1965), 14.Google Scholar

14. Greenberg, , “Studies III,” 316Google Scholar (Studies, 40). This passage was retained in 1963 (Languages of Africa, 37) and reinforced by a similar statement on the preceding page.

15. Greenberg, , “Studies VIII,” 415Google Scholar (Studies, 116). For this group the name Bantoid was afterwards introduced (Languages Of Africa, 9). Its membership has varied slightly in different versions of the classification, with changes in the status of Bute and Jarawa.

16. Though Greenberg used the word valley, rather than basin, he cannot have meant to imply a riverain location for Proto-Bantu. The distribution of the Bantoid languages appears to converge on a point well south of the Benue river, towards the Cameroun border. Greenberg, , “Bantu and its Closest Relatives” in Leben, W.R., ed., Papers From the Fifth Annual Conference on African Linguistics (Los Angeles, 1974), 117.Google Scholar

17. A glottochronological estimate of 2500 years for the age of Proto-Bantu seems only ever to have been mentioned on one occasion: Greenberg, , “Africa as a Linguistic Area” in Bascom, W.R. and Herskovits, M.J., eds., Continuity and Change in African Cultures (Chicago, 1959), 20.Google Scholar This should be read in the light of some guarded remarks on the value of such calculations in Greenberg, , Essays in Linguistics (Chicago, 1957), 54.Google Scholar Independently, Belgian linguists were also experimenting with glottochronology: Coupez, A., “Application de la lexicostatistique au mongo et au rwanda,” Aequatoria, 19(1956), 85–87Google Scholar; Meeussen, A.E., “Lexico-statistiek van het Bantoe: Bobangi en Zulu,” Kongo-Overzee, 22(1956), 86–89.Google Scholar News of this work was reported by Jan Vansina to the second SOAS conference, but it failed to arouse the attention of British historians.

18. Westermann, D., “African Linguistic Classification”, Africa, 22(1952), 250–56.CrossRefGoogle Scholar I have commented before on the importance of this article: Flight, C., “Trees and Traps: Strategies for the Classification of African Languages and Their Historical Significance”, HA, 8(1981), 52–53.Google Scholar

19. Tucker, A.N., “Philology and Africa”, BSOAS, 20(1957), 544n4.CrossRefGoogle Scholar This assessment is presented, not as a personal opinion, but as being “the general impression among Africanists of the British school.” Westermann had died in 1956.

20. Flight, , “Trees and traps,” 54–55.Google Scholar

21. Tucker, , “Philology and Africa,” 550–51Google Scholar; Flight, , “Trees and traps”, 65.Google Scholar

22. Tucker, , “Philology and Africa,” 547.Google Scholar

23. The history of the School during the first fifty years of its existence is recounted by Philips, C.H., The School of Oriental and African Studies, University of London, 1917-1967: An Introduction (London, 1967).Google Scholar

24. On the creation of the Department of Africa see ibid., 23-24.

25. Tucker, A.N., “Obituary: Ida Caroline Ward”, BSOAS, 13 (1950), 542–44.CrossRefGoogle Scholar

26. Carter, Hazel, “Obituary Notice: Malcolm Guthrie, 1903-1972,” Proceedings of the British Academy, 59(1973), 473–98Google Scholar; Arnott, D.W., “Obituary: Malcolm Guthrie,” BSOAS 36(1973), 629–37.CrossRefGoogle Scholar

27. Africa, 15(1945), 1CrossRefGoogle Scholar; 16(1946), 121-22.

28. Africa, 15(1945), 1, 213Google Scholar; 16(1946), 261; 18(1948), 129. The recommendations of the Handbook Sub-committee were published in Africa, 16(1946), 156–59.CrossRefGoogle Scholar

29. Africa, 18(1948), 308Google Scholar; 19(1949), 235; 20(1950), 243-44; 21(1951), 146-47.

30. This shift in Government policy was signaled by the report of the Scarbrough Commission, published in 1947; but the University of London, in anticipation of a favorable verdict, had already sanctioned a considerable increase in the size of the School for the quinquennium 1947-52 (Philips, School, 38-40).

31. The Director at the time, (Sir) R.L. Turner, would later allude to this difficult passage in the history of the Department by way of the following comparison: “The British General commanding the brigade in which my unit, a battalion of Gurkhas, served beside three British battalions, wrote at the end of the war in 1918 that whatever was happening elsewhere he never had to worry about the 2/3 Gurkhas. Very soon after Guthrie took charge… that was the way I felt about the Department of Africa” (Carter, , “Guthrie,” 484Google Scholar).

32. Like the other institutions comprising the University of London, the School appointed its own lecturers, who could then apply, in due course, for recognition by the Univesity. Professorships and Readerships were University appointments, though each was tenable in some specified Department. Guthrie's chair was the first University post to be established in the Department of Africa. Ward had been made Professor in 1944, but only by title.

33. On Guthrie's reluctance to abandon the Baptist ministry as his full-time vocation see Arnott, , “Guthrie,” 630.Google Scholar It was several years before he ceased using the title “Reverend”.

34. The account which follows is derived almost entirely from the annual reports of the Department, contained in the annual reports of the School, for the period 1950-60. Where specific references are needed, they are given in shortened form, as SOAS, Report (1951). The reference in full would be SOAS, Report of the Governing Body, Statements of Accounts and Departmental Reports for the Session 1960-51 (London, 1951).Google Scholar Presumably the reports submitted by the Department of Africa were all compiled by Guthrie, with one possible exception. Since Guthrie was away on study leave for most of the session 1956-57, the report for this year may have been drafted by Jack Berry as Acting Head.

35. Guthrie's assessment of the Treasury studentship scheme, as it affected the Department of Africa, can be found in SOAS, Report (1955), 100.Google Scholar See also Philips, , School, 47–48.Google Scholar

36. SOAS, Report (1956), 109.Google Scholar For a hair-raising account of what it was like to be supervised by Guthrie see Carter, , “Guthrie,” 496–97.Google Scholar

37. Arnott, , “Guthrie,” 636.Google Scholar

38. SOAS, Report (1954), 92Google Scholar; Report (1956), 106.Google Scholar

39. Guthrie, , “Obituary: Wilfred Howell Whiteley,” 36(1973), 119–25.Google Scholar Whiteley had been associated with the Department throughout the 1950s; he was formally attached to it as a visitor in 1956-57 (SOAS, Report [1957], 100Google Scholar).

40. Tucker may have been responsible for instigating a seminar on the “Hamitic” languages: this began meeting in 1952 (SOAS, Report [1953], 87–88Google Scholar) but seems to have petered out within a year or two.

41. Perhaps it needs to be pointed out that African studies were still considered a minor, marginal extension of Orientalism. Contemporary reports of the 23rd International Congress of Orientalists, which met in Cambridge in 1954, make much of the fact that this had been the first such Congress to include a separate section devoted to Africa (SOAS, Report [1955], 105Google Scholar; Africa, 25 [1955], 97).Google Scholar At the 25th Congress held in Moscow in 1960 the African section voted to secede (Africa, 31 [1961], 79–80Google Scholar); and the first International Congress of Africanists convened in Ghana in December 1962 (Africa, 33 [1963], 148–50CrossRefGoogle Scholar).

42. SOAS, Report (1951), 83Google Scholar; Report (1952), 86Google Scholar; Report (1953), 90Google Scholar; Report (1954), 96.Google Scholar

43. Guthrie, , “African Languages” in Mohrmann, C., Norman, F., and Sommerfelt, A., eds., Trends in Modern Linguistics (Utrecht/ Antwerp, 1963), 60.Google Scholar

44. The original plan had been for part 1 of the Handbook covering the Bantu and Khoisan languages, to be edited by Guthrie, (Africa, 18[1948], 129).Google Scholar By March 1951, however, he had apparently withdrawn from the assignment: reports on the progress of the Handbook during this period avoid any mention of part 1 (Africa, 21 [1951], 232Google Scholar). Two years later, it was announced that the Bantu section - now being called part 4 - would be taken in hand by Bryan as soon as she had finished with part 3 (Africa, 23[1953], 345Google Scholar). She was to have access to “much useful material assembled by Professor Guthrie.”

45. Guthrie, , “African languages”, 61.Google Scholar

46. Thus Parsons saw it as Guthrie's “greatest contribution to posterity” that he had introduced a “truly scientific approach” to the comparative study of African languages. Parsons, F.W., “Is Hausa Really a Chadic Language? Some Problems of Comparative Phonology,” African Language Studies, 11(1970), 272.Google Scholar

47. SOAS, Report (1956), 107Google Scholar; Report (1958), 107.Google Scholar

48. Greenberg passed through London in 1954 on his way to Nigeria. His name appears in a list of visitors included in the annual report of the Department of Phonetics and Linguistics (SOAS, Report [1955], 110Google Scholar); he also visited the Department of AFrica.

49. Guthrie, , “Contributions from comparative Bantu Studies to the Prehistory of Africa” in Dalby, David, ed., Language and History in Africa (London, 1970), 21–22.Google Scholar

50. Guthrie, , “Languages and History,” JAH, 5(1964), 135–36CrossRefGoogle Scholar, reviewing Greenberg, Languages of Africa

51. Berry, Jack, BSOAS, 18(1956), 395CrossRefGoogle Scholar, reviewing Greenberg, Studies. From the points of detail specifically criticized, it seems clear that Berry had consulted with some of his colleagues–Westphal, Tucker, Parsons–in writing this review.

52. Philips, , School, 45-46, 49-51, 53–57.Google Scholar

53. SOAS, Report (1951), 92Google Scholar; Report (1955), 113.Google Scholar

54. Hamilton, R. A., ed., History and Archaeology in Africa: Report of a Conference Held in July 1953 at the School of Oriental and African Studies (London, 1955).Google Scholar The foreword is by Philips, head of the Department of History at the time, who became Director of the School in 1957.

55. Jones, D. H., ed., History and Archaeology in Africa: Second Conference Held in July 1957 at the School of Oriental and African Studies (London, 1959).Google Scholar See also Africa, 28(1958), 57–58.CrossRefGoogle Scholar

56. Johnston returned to the subject at intervals over the next forty years. On the development of his thinking see Vansina, Jan, “Bantu in the Crystal Ball I” HA 6(1979), 304–13.Google Scholar

57. Oliver, R. A., Sir Harry Johnston and the Scramble for Africa (London, 1957).Google Scholar The book is principally concerned with Johnston's political career; his linguistic research is mentioned only briefly (21-23, 80-82, 354-56).

58. Johnston, , Kilima-njaro, 478–88.Google Scholar

59. For a much more elaborate attempt to map the Bantu expansion in space and time see Johnston, H.H., “A Survey of the Ethnography of Africa,” Journal of the Royal Anthropological Institute, 43(1913), 390–94.Google Scholar

60. It might be argued from this that Johnston had gone some way towards anticipating Greenberg; but his later publications weaken the case.

61. On the Maasai expansion see Johnston, , Kilima-njaro, 405–08.Google Scholar

62. Hamilton, , History and Archaeology, 15.Google Scholar

63. SOAS/ICS, African Hsitory Seminar, paper AH/58/5: R.A. Oliver, “The Bantu.” This was the text of a lecture given in Ghana during the autumn term of 1957: the dating is significant in relation to the development of Guthrie's ideas (see notes 68-69 below). I am grateful to Professor Oliver for letting me have a copy of this paper, and for allowing me to cite it here.

64. Especially perhaps by Childe, V.G., New Light on the Most Ancient East (4th ed.: London, 1952).Google Scholar Until shortly before his death in 1957, Childe had been Director of the University of London's Institute of Archaeology.

65. SOAS/ICS, African History Seminar, paper AH/58/4: C. Wrigley, “Speculations on the Economic Prehistory of Africa.” The paper is dated November 1958. I am grateful to Mr. Wrigley for allowing me to cite it here, and to quote certain passages. The text was greatly revised before being published as Wrigley, C.C., “Speculations on the Economic Prehistory of Africa”, JAH, 1(1960), 189–203.CrossRefGoogle Scholar

66. Jones, , History and Archaeology, 27Google Scholar; Africa 27(1957), 403Google Scholar; Barendsen, G.W., Deevey, E.S., and Gralenski, L.J., “Yale Natural Radiocarbon Measurements III,” Science, 126 (1957), 916.CrossRefGoogle ScholarPubMed The inference turns out to have been correct, though it was hardly justified by the original series of dates.

67. Jones, , History and Archaeology, 40Google Scholar

68. SOAS, Report (1957), 23, 96.Google Scholar One publication which arose directly out of this tour is Guthrie, , “Teke Radical Structure and Common Bantu,” African Languages Studies 1(1960), 1–15.Google Scholar Here the results were still being presented in a strictly synchronic framework, without the slightest hint that they would need to be reinterpreted later in diachronic terms.

69. In a previous paper I described at some length the development of Guthrie's historical interpretation: Flight, , “Malcolm Guthrie and the Reconstruction of Bantu Prehistory”, HA, 7(1980), 81–118.Google Scholar However, I would no longer wish to suggest that he went through a “crisis of confidence” in 1947. His only problem then, I think, was to limit the volume of data with which he would have to cope, by means of some sampling strategy; and that problem was solved by his Classification, of which we know that the draft was complete by 1946 (Africa, 16[1946], 261).Google Scholar As I read the evidence now, Guthrie's anxiety that his work might all seem futile began festering only later, perhaps under the pressure of Oliver's “importunities”, and did not come to a head until 1957/58. I thus regard his “two-stage method” (note 72) as merely a retrospective rationalization. In short, he managed to convince himself that what he happened to have done was exactly what he ought to have done.

70. Oliver, , Johnston, xGoogle Scholar, 82n.

71. SOAS, Report (1958), 23, 120.Google Scholar In Oliver's absence the African History Seminar had been organized jointly with the ICS and the same arrangement continued after his return.

72. Guthrie, , “A Two-Stage Method of Comparative Bantu Study,” African Language Studies, 3(1962), 1.Google Scholar

73. Guthrie, , “Comparative Bantu: A Preview,” Journal of African Languages, 4(1965), 41.Google Scholar

74. The occurrence was thought worth mentioning in the annual report by both the Departments concerned: SOAS, Report (1959), 107, 117.Google Scholar

75. For a more detailed account, explaining where Guthrie went wrong, see Flight, , “Guthrie,” 88–102.Google Scholar

76. SOAS/ICS, African History Seminar, “Note of meeting held on 28 January 1959”. Guthrie was also present at several of the subsequent meetings, intervening frequently in the discussion. The “Notes” which summarized the proceedings– thanks to which we can follow the evolution of the argument step by step–were compiled by Roy Bridges, a graduate student from King's College, London. I am grateful to Dr. Bridges for elucidating certain passages.

77. At this stage Guthrie seems not to have been using the term Proto-Bantu.

78. On this point see SOAS/ICS, African History Seminar, “Note of Meeting Held on 21 January 1959”. Arkell does not appear to have attended the meetings which dealt with the Bantu problem.

79. SOAS/ICS, African History Seminar, “Note of Meeting Held on 4 February 1959”.

80. SOAS/ICS, African History Seminar, “Note of Meeting Held on 18 February 1959”.

81. The difficulties which were thought to be involved in this hypothesis are indicated by Wrigley, “Speculations,” 201.

82. Guthrie, “Two-Stage Method”.

83. See, for example, Guthrie, , “Bantu Origins: A Tentative New Hypothesis”, Journal of African Languages, 1(1962), 11n2.Google Scholar

84. Cf. Guthrie, , “Bantu origins,” 21Google Scholarn4.

85. For a parenthetical reference to this dilemma see Wrigley, C.C., “Population in African History”, JAH, 20(1979), 129.CrossRefGoogle Scholar

86. Wrigley, , “Speculations,” 201.Google Scholar

87. Ibid., 198. The wording compares very closely with that of a similar sentence in Oliver, R.A. and Fage, J.D., A Short History of Africa (Harmondsworth, 1962), 32.Google Scholar

88. SOAS, Report (1959), 115Google Scholar; Report (1960), 24.Google Scholar

89. SOAS/ICS, African History Seminar, paper AH/59/c: R.A. Oliver, “Cultivated Plants of Indonesian Origin in Africa.” I am grateful to Professor Oliver for permission to cite this paper, and to quote some passages from it. The book in question was Murdock, G.P., Africa: Its Peoples and Their Culture History (New York, 1959)Google Scholar, reviewed at length by Fage, J.D., JAH, 2(1961), 299–309.CrossRefGoogle Scholar

90. See also Wrigley, , “Speculations,” 197–99.Google Scholar Like Oliver, Wrigley objected to Murdock's hypothesis on the grounds that it seemed impossibly far-fetched. He also cited botanical opinion favoring an African origin for several tropical crops, especially some of the principal varieties of yam.

91. Oliver's memorandum was prepared for a meeting which dealt with the question of Indonesian influence in Africa. This was a problem which seemed important at the time–even before the publication of Murdock's book–and which often became entangled with discussion of the Bantu expansion. Apart from Guthrie's work the musicological theories of the Rev. A.M. Jones were the only contributions proffered to historians by the Department of Africa.

92. Guthrie, , “Preview,” 41.Google Scholar It was in the School's annual report for 1959-60 that Guthrie first referred to the fact that he had developed “a new hypothesis about Bantu origins:” SOAS, Report (1960), 100.Google Scholar

93. Guthrie, , “Problèmes de génétique linguistique: la question du bantu commun,” Travaux: de l'Institut de Linguistique, 4(1961), 92.Google Scholar

94. Fage, J.D., “Conference on African history and archaeology,” Africa, 31(1961), 379–80Google Scholar; Gray, Richard, “A Report on the Conference,” JAH, 3(1962), 175–91CrossRefGoogle Scholar, with a list of participants and papers submitted, 367-74.

95. Gray, , “Conference,” 183Google Scholar, summarizing an intervention by W.B. Morgan (Department of Geography, University of Birmingham).

96. Ibid., 182, summarizing an intervention by A.H. Christie (Department of South-East Asia, SOAS).

97. Ibid., 184-85. Wrigley's paper was published in JAH (note 108).

98. SOAS, Third Conference on African History and Archaeology, ppaer 61/71: Guthrie, “Some Developments in the Prehistory of the Bantu Languages.” This is a completely different paper from the one published later in JAH under the same title (note 103).

99. Gray, , “Conference,” 185Google Scholar, summarizing Guthrie's verbal presentation.

100. Guthrie, , “Some developments,” 3.Google Scholar The term Proto-Bantu seems first to have been used by Guthrie in this paper.

101. Gray, , “Conference,” 185–87Google Scholar, summarizing the discussion provoked by Guthrie's paper.

102. Ibid., 185-86; Guthrie, , “Bantu Origins,” 18.Google Scholar

103. Gray, , “Conference,” 186Google Scholar; Guthrie, , “Bantu origins,” 20–21Google Scholar; idem., “Some Developments in the Prehistory of the Bantu Languages”, JAH, 3(1962), 281-82.

104. Guthrie, , “Some Developments,” 273.Google Scholar

105. Gray, , “Conference,” 186.Google Scholar

106. Guthrie, , Some Developments,” 281Google Scholar; idem., “Bantu origins,” 11.

107. Guthrie, , “Some developments,” 4.Google Scholar

108. As Wrigley pointed out, to say that Greenberg's theories had not been verified was not equivalent to saying that they had been falsified. Wrigley, C.C., “Linguistic Clues to African History,” JAH, 3(1962), 269.CrossRefGoogle Scholar

109 Welmers, W.E., “Note on the Classification of African-Languages,” Linguistic Repoter, 1(1959)Google Scholar, Supplement 1, 6.

110. Oliver, and Fage, , Short History, 29–33.Google Scholar This section underwent some changes in the second (1966) and third (1970) editions; in the fourth (1972) it was quite extensively revised.

111. Ibid., 32. The “Iron Age hunters and fishermen” did not survive beyond the second edition. In 1970 “Iron Age” became “Late Stone Age,” in 1972 “hunters” became “vegeculturalists”.

112. Ibid. Despite its hesitancy this sentence was retained in 1972.

113. Ibid., 30n1, citing the occurrence of bored stones, supposed to be weights for digging sticks, “as far back as the fifth millennium B.C.” This footnote disappeared in 1972, when the paragraph to which it had referred was completely rewritten.