or even primarily what is now extant or canonical is to distort. As a result, we include the
obvious major authors, but also all the lost, fragmentary, obscure, and anonymous scientific
writers whom we can. However, we include only authors known to have made some written
or conceptual contribution, but not practitioners, such as architects and physicians, however
famous or accomplished. Likewise, teachers, however important they may have been to
their students, or for the transmission of ideas, we include only if they themselves made a
contribution. Despite the plausible presumption that most of the adherents of the schools of
Plato, Aristotle, Ze ̄no ̄n, or Epicurus who made any contribution would also have made
some contribution to natural science, we include only those who are explicitly attested to
have done so. Anonymous and pseudonymous works receive a separate entry, in order to
give them their due prominence – for example, the Hippokratic and Aristotelian corpora
are each divided into about a dozen entries, whereas many other anonymi each receive
individual entries.
We begin with He ̄siod and Homer and Thale ̄s, who represent key parts of the origin
of Greek scientific thought, even if we can no longer or not yet assert with confidence
any theory of that origin. (It may be that we should go further back: Arnott 1996.) The
encyclopedia extends to ca 650 , rather than (say) to 529 (closing of the Academy), 410
(sack of Rome), or even 313 (edict of Milan), on the grounds that the whole “late antique”
period represents a gradual transition from Mediterranean antiquity to the medieval or
Byzantine periods. (Moreover, the century 650– 750 is remarkably weak in science,
whereas adequately to study science after 650 requires great familiarity with many
languages, such as Arabic, Aramaic, Armenian, Coptic, etc., which the editors alas do not
possess.) We include the authors of the “early Byzantine” period (330– 650 ) in order to
ensure that the encyclopedia offer a synoptic view of ancient Greek science. Preferring
errors of inclusion to those of exclusion, we also include authors of unknown or uncertain
date, so long as there is a reasonable chance that they are prior to our terminal date.
Although nearly all of the entries concern works written in Greek, we include around 200
entries on authors or works in other languages, that in each case are based upon the Greek
scientific tradition; we thus exclude the copious works of Chinese science, as well as most
Babylonian, or Egyptian, or Sanskrit works. Classicists and historians are familiar with the
Latin reception and transformation of Greek science; less familiar but of equal interest are
a number of other scientific traditions also influenced by the Greek. Readers should consult
the relevant indices for works whose language or author was Armenian, Celtic, Gothic,
Egyptian, Persian, Sanskrit, or one of the Semitic languages (Arabic, Aramaic including
Mandaic and Syriac, Babylonian, Hebrew, or Punic), as well as the index of well over a
hundred Latin authors and works.
II. Names. Because the book primarily contains Greek scientists, Greek names are trans-
literated without prior Latinization. Moreover, names of other traditions (including Latin)
are also directly transliterated, all according to the conventions of scholarship in those fields.
Direct transliteration is no more arbitrary than any other system, more accurately preserves
pronunciation and etymology, and more clearly signals culture. Of course, no system is
wholly consistent: the standard “Anglo-Latinate” rendering of Greek names has Plato and
Hero for Plato ̄n and He ̄ro ̄n, but Theon (not Theo) and Cleon (not Cleo), and oscillates over
Dio or Dion; finally, even in that system, Nike ̄ is never Nice. Even in Latin, if one writes
“Pompey” and “Pliny” and “Livy” and “Antony”, why then does one not also write “Tully”
(as indeed in 18th c. English) or even “Porcy”? Moreover, even in English, we do not write
INTRODUCTION