What’s the best English translation of Herodotus’ Histories?

The Italian proverb Traduttore, Traditore reminds us that there is no perfect translation into English of a Greek text written two and a half millennia ago. The question about the ‘best translation of Herodotus’ requires us to ask: best for whom? And best for what purpose? Do we mean, best for the student of ancient history or archaeology? Many classrooms are grateful for the Purvis/Strassler Landmark edition that offers helpful running headers, notes, dates, locales, summaries, figures, maps, and appendices. Or do we rather mean, best in that it tries to emulate the effect of Herodotus’ own often sonorous syntax and diction, like the nineteenth-century translation of Macaulay, or to reproduce his distinctive dialect, even to the point of including egregious archaisms, as Powell did? Do we privilege a translation that features the voice and cadences of a skilled teller of tales, like the twentieth-century translation of David Grene? Do we simply look for the Herodotus who is most timely, accessible, and compelling for the early twenty-first century general reader, perhaps enjoying the recent Holland translation?

The choices each translator makes in representing Herodotus’ diction and narrative cadences depend in large part on his or her understanding of Herodotus’ own goals and gifts. Was Herodotus an ordinary fifth-century thinker but consummate narrator, who strung together accounts he thought his various audiences would like and sometimes improved them for the sake of the narrative effect? Or was he an innovative ‘pre-socratic’ empiricist, striving to record different individual voices and Weltanschauungen – a genius who invented a new way of recording and integrating into a single rich whole many reports of distant peoples and places and fading memories of important past events? Subtler issues will play a part as well: how significant does a given translator think are Herodotus’ (rare) expressions of religious piety or the degree to which dreams, omens, and the like are featured? Or his relatively frequent inclusions of women as important historical actors? Or his flashes of slyly-expressed irony? How these sorts of questions are answered in the translator’s mind will affect the diction and syntax of the resulting translation. And, in turn, the reader’s own feelings about such questions – or, indeed, about the larger question of what ‘history’ is, or should be! – will affect his or her sense of a given translation’s worth.

Our own preferences include the following. Ideally, a readable English translation should avoid both archaizing and the temptation to naturalize the text so that the ancient author’s own expressive voice becomes unrecognizable. Among the most valuable additional criteria are arguably the following:

1) Recurring words that serve as unifying markers for Herodotus’ narrative would be translated by the same English words (or roots) whenever possible.  The most obvious example is represented by words of the θῶμα (thôma, ‘wonder’) family.  It is not clear to us, for example, why Holland goes with ‘astounding’ in the first sentence and at 1.23 (Arion), but translates ‘greatest wonder’ (by far the best choice) the celebratory phrase Herodotus applies to the Assyrian boats in 1.194. By the same token, Herodotus is careful in distributing the words τύραννος (tyrannos, ‘tyrant’), βασιλεύς (basileus, ‘king’), δεσπότης (despotes, ‘despot’), μούναρχος (mounarchos, ‘sole ruler’) throughout the text: a translation should respect his choices.

2) Introductory and concluding sentences both mark the rhythm of the narrative and reveal the organizing/interpreting voice of the narrator and should be kept, even the least expressive of these (e.g. of the ‘this is what he did’ or ‘this is what happened’ type). Waterfield almost entirely eliminates them, thereby disguising the oral flavor (this does not mean ‘oral nature’) of Herodotus’ style: compare e.g. his translation of 3.41 (on Polycrates’ decision to sacrifice his ring) with the greater faithfulness of Grene in the same passage.

3) Generally speaking, a translation should keep as close as possible to the formulation of the original. Grene’s transformation of the third into the first person (‘I, Herodotus . . .’) in the first sentence is unnecessarily disconcerting.

Many other observations could be made, and we look forward to reading some of them on this website. It’s a fascinating question, how best to translate Herodotus, and by now it should be clear that our general answer is, ‘well, it depends … .’

In finish, we thought it might be helpful to list English translators of Herodotus, as found in D. Lateiner’s 2005 edition of Macaulay’s Herodotus (revised), pp. 503-504. Only those from Lateiner’s list who translated Herodotus’ complete text are included here: Isaac Littlebury (1709, 1737); William Beloe (1791); P. E. Laurent (1827); Isaac Taylor (1829); Henry Carey (1847); George Rawlinson (1858-60, 1862, 1875); G. C. Macaulay (1890); A. D. Godley (1920-1924); J. E. Powell (1949); Aubrey de Selincourt (1952, rev. 1972, rev. by John Marincola 1996); Harry Carter (1958); Kenneth Cavander (1959); David Grene (1987); Robin Waterfield (1998). Lateiner additionally briefly annotates each entry with his judgment on the translation’s quality. Translations that have occurred since Lateiner’s list came out include those of Andrea Purvis, ed. by R. Strassler (2007); Tom Holland, ed. by P. Cartledge (2014); Pamela Mensch (2014); and James Romm (2014). Peter Green’s posthumous translation is expected this year (2025).

At the risk of seeming to waffle from the outset, in my view the designation of any translation of Herodotus as “the best” depends on who’s reading it and for what purposes. The last quarter-century has witnessed the appearance of several laudable translations, complete with scholarly commentary as needed for the contextualization and comprehension of the Histories. In addition to John Marincola’s revised, annotated edition of the de Selincourt translation (London 1996), these include the translations by Robin Waterfield (commentary by Carolyn Dewald, Oxford 1998), Andrea Purvis (in Robert Strassler’s Landmark edition, New York 2007), Tom Holland (commentary by Paul Cartledge, London 2014), and Pamela Mensch (commentary by James Romm, Indianapolis 2014).         

The translations themselves vary in their balance between fidelity to Herodotus’ Greek and use of modern English idiom (Holland’s rendition is especially lively and engaging). For my own purposes and those of my (former) students, the Strassler Landmark edition has proved most useful because of the sheer amount of editorial assistance given the reader, including an introduction by Rosalind Thomas; footnotes rather than endnotes, as well as “side notes” that summarize the content of each chapter of the text; numerous maps keyed to the translation (an indispensable feature for the geographically challenged, like myself); a near-alphabet of appendices by experts on important Herodotean topics; and a truly encyclopedic index of 100 pages, an indispensable aide-mémoire for any reader of Herodotus, no matter how experienced.