Surely no one involved with comparative Dravidian, or Altaic, or Elamite, or even simply with historical linguistics in general, would question the exhortations to linguistic virtue laid down by K. V. Zvelebil (BSOAS, XLVIII, 1, 1985, 116–20): every scholar in the world who is interested in these questions at all will agree that, touching upon the genetic affiliation of Tamil and Japanese, ‘hypotheses have emerged that deserve more careful and sympathetic attention’ (p. 116), and also that ‘it would be premature, sweepingly to dismiss … a hypothesis [of non-accidental connexion between Japanese and Tamil (Dravidian)] as impossible and fantastic’ (p. 117). But over and above these unexceptionable generalities, Zvelebil's note reflects his tacit assumption that the Japanese linguistic forms and English glosses in Ōno Susumu's Sound correspondences between Tamil and Japanese (Tokyo, 1980) are valid, and reliable as evidence for the comparativist. Unfortunately, this is not true.