We can hypothesize from the previous discussion that the role of English in research may be considerably inflated. e In fact, several early small-scale studies bear this out: Throgmartin (1980) produced English percentages in the 40% range for social sciences, and Velhoand Krige (1984) showed a clear preference for publication in Portuguese among Brazilian agricultural researchers. A complete bibliography on schistosomiasis, a tropical disease, by Warren and Newhill (1978) revealed an English language percentage of only 45%. 48 These studies would seem to indicate that a more accurate percentage for English would be around 50% rather than around 80%. underrepresented. Indeed, Najjar (1988) showed that no Arabic language science journal was consistently covered by the Science Citation Index in the mid-1980s.