Fashions in Herodotean translation change with the social conventions of the translators. We no longer, like Rawlinson in 1882, expurgate the Histories in pursuit of ‘purity of thought’, or replace the names of Greek deities by their Roman equivalents, as he did, to avoid ‘harshness and repulsiveness’ [sic]; we are more likely to euphemise Herodotus' racial terms out of respect for political correctness. The habit of treating his use of Ionic dialect as a mark of archaic quaintness (rather than as the vehicle of cutting-edge scientific investigation), and of then attempting to reproduce that quaintness in English, has likewise, fortunately, been abandoned. But one characteristic that has stamped all literary versions of Herodotus remains as dominant as ever, and that is a resolute indifference to the demands of his enchanting style, the strung-along εἰρομένη λέξις disliked by Aristotle (Rhet. 1409^a 29–32), which does so much to make him, more than any other ancient author, sound as though he is speaking directly to his readers rather than writing for them.

There have been various reasons for this. The current argument would seem to be that the hypothetical general reader is accustomed to short simple sentences, and must not be scared off by long ones. Here the standard was set half a century ago by A. de Sélincourt's durable Penguin version, of which D. Grene well remarked, in the introduction to his own version, that it ‘sounds exactly as though new-minted by a twentieth-century journalist’. The danger attendant on simplification is always a lapse into populist or bureaucratic cliché.

This is the tradition in which M. works; and, given its limitations, she has done much better by Herodotus than might have been expected. Her prose is never dull, and on the whole avoids clichés. She does break up his longer sentences, but not radically; she avoids glossing his text, leaving that job, where needed, to R.'s always concisely helpful notes; and, while working from Hude's text, she is ready, on occasion, to jettison some of the otiose emendations that survive there (e.g. ἑπτά for δυῶν at 8.131.3). Though she can, rarely, lapse into archaisms – her ‘showing-forth’ in the proem for ἀπόδεξις is uncharacteristic – her Englishing of Herodotus is, overall, clear, straightforward and accurate, without any attempts at improving his style. It would be hard to produce a version of Herodotus that was not, for the most part, vivid narrative, and this M. does exceptionally well. But in her quiet understated way she also gives the reader at least some sense of her author's quasi-conversational rhetoric, and this is much rarer. Hackett has a tradition of valuable translations – most notably S. Lattimore's Thucydides – and the M.–R. Herodotus is at least as good as any of the competitive versions that have appeared in the last year or two.