Simon Hornblower Greek Historiography Essay

The ancient Greek achievement in the writing of history set the standard for all time. The nature of that achievement, though, looks increasingly problematic in the light of recent work on such topics as literacy, orality, rhetoric, narrative technique, and the invention of tradition. In this book, eight ancient historians consider both the achievement and the problems invThe ancient Greek achievement in the writing of history set the standard for all time. The nature of that achievement, though, looks increasingly problematic in the light of recent work on such topics as literacy, orality, rhetoric, narrative technique, and the invention of tradition. In this book, eight ancient historians consider both the achievement and the problems involved in the study of the Greek historians. These essays reflect the best and most recent scholarship on the subject.

Paperback, 304 pages

Published October 31st 1996 by Clarendon Press (first published October 27th 1994)

Bryn Mawr Classical Review 2001.08.22

Simon Hornblower, Elaine Matthews (ed.), Greek Personal Names: Their Value as Evidence. Proceedings of the British Academy, 104.   Oxford:  Oxford University Press, 2001.  Pp. 184.  ISBN 0-19-726216-3.  $49.95.  

Contributors: Michael Crawford, Anna Morpurgo Davies, Laurent Dubois, Peter Fraser, Christian Habicht, Miltiades Hatzopoulos, Simon Hornblower, Denis Knoepfler, Elaine Matthews, Robert Parker

Reviewed by Stephen Lambert, Universities of Liverpool and Heidelberg (
Word count: 4197 words

This collection of essays, which makes a major contribution to historical onomastic studies, arises out of a colloquium held at the British Academy in July 1998 to celebrate the 80th birthday of Peter Fraser, the original mover of one of the most important projects currently underway in ancient Greek scholarship, the Lexicon of Greek Personal Names (LGPN), and co-editor, with Elaine Matthews, of three of the four volumes of the Lexicon to have appeared so far (vols. I, IIIA and IIIB; vol. II, Attica, with which your reviewer is most familiar from near daily use, was edited by Michael Osborne and Sean Byrne). As Matthews points out in her introduction, the volume also marks the half-way point in the Lexicon's publication programme, which has so far covered the Aegean Islands, Magna Graecia and most of the Greek mainland--cataloguing over 200,000 individuals, who bore nearly 30,000 names. The achievement, as many of the contributors point out, is already magnificent; the labours of the editors and their supporting researchers heroic; but the cataloguing of names is not an end in itself: "nous devons faire...l'histoire des noms, et même l'histoire par les noms," as Louis Robert once (rather obviously) pronounced.1 The essays in this book are all about "history of names and history by names" and they give us an impressive preview of the "new fields of enquiry, the opening of which, thanks to LGPN...we can witness, and perhaps participate in."2 While all the papers contain matter of interest, I was particularly struck by those of Morpurgo Davies, Parker and Hatzopoulos. The only area that I should like to have seen more fully explored in the book is what might be termed epigraphical onomastics. Though several of the contributors stress the extent to which LGPN (and therefore most of the discussion in this volume) is ultimately dependent on epigraphy, the extent and implications of the interface are not drawn out. I discuss this subject briefly below in an Appendix.

Greek onomastic lexicography did not, of course, begin with LGPN, and in the editors' introduction Matthews gives an informative brief history of it, from the pioneering work of Letronne through the various 19th cent. editions of LGPN's effective predecessor, "Pape-Benseler", to Fraser's initiation of LGPN in 1973 and work on the project to date (pp. 1-9). Hornblower supplies an overview of the papers in the volume (pp. 9-14).

One of the many themes that surface frequently in the book is the problem of what, if any, weight should be given to the literal meaning of a name (e.g. Theodotos, "given by god"). The meaning of a personal name, however, is its use to designate an individual and to avoid confusion in this review I shall describe a name's "literal meaning" as "literal connotation" (for that, if anything, is all that it is). A second, more implicit, theme, is the way that, in thinking about names, one finds oneself to-ing and fro-ing across a (sometimes rather rickety) bridge between thinking about a category of words and thinking about the historical realia which underlie, surround or can be inferred from them. All of the papers have linguistic and historical aspects, but some are more linguistic, some more historical in emphasis, and I shall discuss the contributions not quite in the order in which they occur in the book, but according to roughly where they are situated on this linguistic-to-historical spectrum.

That said, the first two papers are also the most linguistic in emphasis. The first, by Morpurgo Davies (pp. 15-39), seeks to explain how and why, linguistically, (Greek) personal names differ from other lexical items. The key difference, she finds, is that they have an "intentionality" which other lexical items lack, i.e. they are deliberately and consciously given. She draws two consequences: names may preserve archaic (e.g. dialectal) features after those features have disappeared in other areas of the language; and, more generally, names can be highly revealing as to the "cohesion and cultural continuity of a specific community". With the help of LGPN she proceeds to investigate the differing reactions to the coming of koine in Arcadia and Cyprus: "conservative" Arcadia did not alter its name forms into koine; "innovative" Cyprus did when writing alphabetically, but did not when using the local syllabic script. The explanation: Arcadia was a fundamentally homogeneous linguistic area; on Cyprus the presence of at least three different and non-mutually intelligible languages (Greek, Phoenician and Eteocypriot) in which onomastic alteration from one to another was normal created a disposition to linguistic adaptability.

Morpurgo Davies' observation of the "intentionality" of Greek personal naming seems to me an important one, with far reaching implications which will require close attention in future historical onomastic studies. Her Arcadia/Cyprus example is illuminating, though the dual script on Cyprus means that we are not dealing with strictly comparable regions and the "conservative" vs. "innovative" dichotomy is therefore not ultimately very convincing. My main regret, however, is that she stays safely (but unadventurously) on the linguistic side of the linguistic-to-historical bridge. She not only describes the onomastic phenomena linguistically, she also (perhaps less convincingly) explains them in entirely linguistic terms. We are left in the dark about what interplay there may have been (theoretically and/or specifically in the cases she treats) between linguistic and other aspects of "cohesion and cultural continuity" and how that might have affected onomastic usage.

In a much more narrowly focused, technically linguistic, but nonetheless interesting, paper (pp. 41-52), Dubois discusses names with the element (h)ippo-. Some are "irrational compounds", arising e.g. from use of -ippos in effect as a suffix without its literal connotation, "horse" (e.g. Nikandrippos). Where there is semantic content it may be "having" (Leukippos), or "being like" (Aristippos). Most commonly (h)ippo- names had a verbal first element, which, as Dubois engagingly explains (and this is the best part of the paper), usually expressed phases in the breeding, training etc. of war horses such as selection (e.g. Hippokritos), taming (Damasippos), training (Strepsippos), yoking (Zeuxippos), unyoking (Lysippos). Dubois argues that the inverse of the last, Hippolytos, eponym of the Euripidean play, should mean "man whose horses are unyoked" (he compares βουλυτός, "time when horses are unyoked"). The clarity of the argument leaves something to be desired at crucial points (pp. 48-51), making it difficult to judge whether it is technically sound. But my main concern is broader. The paper would have been more successful if the author had reflected on what exactly his discussion reveals about this name. Unlike an ordinary word such as βουλυτός, Dubois' analysis tells us (or might tell us) not about the name's meaning (which is the designation of an individual), merely its literal connotation. But in a case like this, where the literal connotation of the name is not immediately obvious, that connotation would seem to be affected (if not determined) by Morpurgo Davies' "intentionality". Even in the case of an ordinary historical individual this "intentionality" is not straightforward (e.g. is the agent the inventor of the name or the parent(s)?); but with "Hippolytos" the intentionality is hugely complex, for the intention of the notional inventor of the name, of the parent of the historical individual, if any, who may have been the basis of the Hippolytos myth and cult, of the persons who founded or were involved in the development of that myth/cult, as well as that of Euripides himself, would all seem to be relevant. Dubois may tell us about the first of these; but the other intentions may not have been the same as this notional initial one; and the literal connotation of the name may therefore also have varied. If Euripides, for example, intended "Hippolytos" to connote literally "loosed by horses" (as Theodotos literally connotes "given by god"), or if he intended no literal connotation at all (and the irrational hippo- names and the absence of obvious allusion to the name's literal connotation in the play show that this is a very strong possibility) it does not seem that we could say that he would be "wrong". There needs to be awareness that, in relation to the literal connotation of a name, philological analysis of the sort applied by Dubois may not give us a (or the) "right answer".

If Morpurgo Davies ultimately stays safely on the linguistic side of the rickety bridge, Dubois finishes with a bold attempt to drive his horse across it at a single leap, goading it with a spur given him by Morpurgo Davies, i.e. her observation of the remarkable absence of (h)ippo- names from Mycenean in contrast to later Greek. This will need more sophisticated treatment than Dubois' assumption (unsupported by archaeological or historical analysis or consideration of horse-use in Homer and the early poets) that it reflects increasing horse-use in the Dark Ages. The relationship between onomastic patterns and the realia, as several contributors stress, is rarely this simple.

This problem of literal connotation surfaces again in two papers which are positioned further towards the historical end of the linguistic-historical spectrum, both handling specific classes of names: Parker (pp. 53-79) on the extremely common theophoric names (e.g. Apollonios, Theodotos); Fraser himself (pp. 149-57) more briefly on the much less common use of ethnics as personal names (e.g. Thessalos).

Parker supplies a useful table of Attic theophorics on p. 58, albeit unfortunately limited, as is his paper generally, to evidence allegedly occurring BC--"allegedly" because most Attic funerary monuments around the turn of the era, a major source of names, have yet to be validly dated; --"unfortunately" inter alia because, as Morpurgo Davies clearly shows, and as Parker himself notes in other connections (see below), name-giving may be conservative and later name use is therefore relevant to analysis of earlier periods. Despite this minor reservation, the paper seemed to me one of the best in the book, deeply researched, well judged and frequently original. Parker notes that LGPN confirms the old observation that literal theonymy (i.e. personal names which are the same as those of gods) is extremely rare BC (in fact it is also uncommon AD). Other gods (and "god" itself, cult titles, festival names etc.) can form names in various ways (with some significant exceptions--e.g. underworld powers). He proceeds to analyse the implications for cultic realia, treading with appropriate caution (and generally successfully) across the rickety bridge and considering e.g. how far persons named for a deity had a continuing relationship to the deity's cult (answer: only occasionally); how far weight should be given to the significance of endings such as -doros, -dotos (not much); how far names were given in response to prayer to the deity in question (probably quite often, but explicit evidence very slight). Then he considers distribution: theophorics are more commonly borne by men than women; by slaves than free persons; are more common in some regions than others; more common later than earlier. Here Parker is much more cautious in suggesting explanations; more work is clearly needed. Certainly, the explanation he offers for the bias to men--essentially that religion was a macho, men-only affair--is one of the few significant points on which the paper left me unconvinced. Religion was one area of life in which, generally, women played a significant role (including, often uniquely, publicly, as priestesses etc.); and the explanation seems to me to sit ill with the popularity of theophorics for slaves. Other possible factors need to be considered, e.g. the impact of the respective roles of husband and wife in name-choosing (an important, if obscure, aspect of onomastic "intentionality"); and the likelihood (I should say) that female offspring were not desirable to parents such that the birth of a girl might be regarded as in answer to a prayer.

Occasionally, by virtue of their tendency to retain archaic features, theophoric names may even demonstrate the existence of an otherwise unattested god (Mandros) and (if cautiously interpreted) may tell us something about the place of origin, geographical distribution and popularity of cults. Here Parker shows clearly how while cultic realia can influence the onomastic phenomena so can many other factors (he has specially interesting comments on Hekate and the Mother). Much firmer is the evidence names can supply for the terminus ante of the existence of a cult (Parker has original observations e.g. on Isis, Serapis, Ammon, Men, Pan and Bendis). Finally Parker raises, but unfortunately has not space to pursue, the important issue of the light changes in patterns of theophoric name usage may shed on changes in religious attitudes.

Fraser's paper explores, with a light touch, the relationship between ethnics used as personal names and the place to which the ethnic refers. He is cautious about the weight to be assigned to the literal connotation of a name, stressing that we can see only the external face of name-giving, not normally the internal (i.e. what was in the mind of individual name-givers) and noting that, in Britain for example, names such as "Ireland" or "English" (like "Brown" or "White") lack their literal connotations. Nevertheless, while reluctant to generalise, Fraser thinks it likely that, with ethnic-names, "the ethnic or topical link made a greater impact on the mind of the ancient Greek than it does with us." There is good evidence that a slave might very commonly be named from the place where he came from, while the normal avoidance of own-state-ethnics as names for citizens suggests that such names did indeed maintain their original connotation in some way, though Fraser stresses (surely correctly) the range of possible relationships to a city which the name might reflect. He includes a table of some examples of ethnics used as names culled from LGPN I-IV (not complete) and finishes with some very brief (descriptive) remarks on literal "toponymy" (my word, not his, designating, by analogy with "theonymy", personal names which correspond to place names) and ktetics (Attikos etc.) as names. The overall direction of Fraser's argument seems to me right, though I have some reservations, of which I mention two. English (and other modern European) surnames are not a good parallel to Greek names. I would suggest that they are more lacking in literal connotation largely because they are automatically inherited and there is not the crucial element of "intentionality" stressed by Morpurgo Davies. Nor can I follow Fraser on "Athenaios...frequent in Athens as elsewhere, but, of course, not an ethnic but a theophoric name". This name was in fact extremely rarely borne by Athenians before the 2nd cent. BC and becomes common only in the imperial period. I should say this is because, while formally also a theophoric, the ethnic connotation was primary and the name was therefore avoided in the earlier period. I suspect that the emergence of this name, as also in the imperial period of "Attikos," reflects the increasing ethnic diversity of "Athenians" and the declining importance of Athenian citizenship (which was opened to foreigners precisely from the late 2nd cent.), such that it became no longer incongruous to allude to (perhaps even assert) this ethnic origin in a personal name used in Attica. "Attikos/Atticus" also raises another issue requiring systematic analysis, of obvious importance in relation to wider debates about Roman cultural influence: the impact of Roman names and naming practice on Greek.

Starting from the observations that names in Oropo- occur only in Oropos and the area of Eretria immediately opposite it and that the personal name Oropodoros suggests that "Oropos" should designate a "god or a hero...or more likely a river," Knoepfler (pp. 81-98) argues that the river in question was the Asopos, which name, by assumption of various linguistic shifts, can be argued to be equivalent to "Oropos". The argument, to my mind, is brilliantly convincing; it would have been no less so had the presentation been rather less discursive (in a postscript, p. 98, Knoepfler refers to recent work by L. Del Barrio Vega which tends to similar conclusions).

Habicht's concise paper (pp. 119-127) investigates a rather different sort of onomastic "ethnicity" from Fraser's, the occurrence of characteristically non-Athenian names amongst the Athenian citizen body. He argues, with examples (many original) that this could arise: (a) by xenia between an Athenian and a non-Athenian family (e.g. Alcibiades); (b) before Pericles' citizenship law, and again after about 125, when the law was no longer enforced, by marriage between an Athenian and a foreign woman (e.g. Cleisthenes); (c) by naming after a foreign celebrity (e.g. Kroisos); (d) by naturalisation at Athens of men of non-Athenian origin (e.g. the 4th cent. Athenian sculptor Sthennis). Habicht has done a considerable service by bringing order and shape to this issue. I am not persuaded, however, by his contention that, in the late classical and Hellenistic periods, (a) and (b) were no longer applicable and that one should interpret "foreign names" as resulting from (c) or (more often) (d). The reason Habicht gives for believing in the continuing application of Pericles' citizenship law in hellenistic Athens is that foreign wives scarcely appear on funerary monuments and never as mothers of citizen sons at this period. This is not convincingly argued. It is possible, for example, that foreign wives with citizen sons might not have advertised their foreign origin on a funerary monument (or any other inscription) in Attica. When she married she might (formally) cease to be, as it were, "Aspasia, [daughter of x] of Miletos", and would become "Aspasia wife of x of deme y"; in other words she would cease to be identifiable in the epigraphic record as of foreign origin. Whether Pericles' law remained in force in hellenistic Athens, therefore, remains uncertain. Habicht may also be right that formal xenia was a mainly archaic phenomenon, but there were continuing international contacts that might have given rise to a man's naming a son for (say) a friend, commercial partner or family connection in another Greek city. To Habicht "it seems obvious" that the occurrence of Samian names among Athenian cleruchs on Samos arose from Samians becoming naturalised at Athens and later returning to Samos as cleruchs; but it is also possible that these were Athenians whose names reflect pre-existing family or other connections with Samos which inclined them to participate in the cleruchy when the opportunity arose. Moreover, Habicht overlooks at least one potential route by which foreign names might have entered the citizen body, adoption. Our evidence suggests that, in classical Athens, only Athenians could be adopted (though in fact, even at this period, it is not certain that this was an absolute rule applied strictly in every phratry); but, as with Pericles' citizenship law, we do not know how far this rule continued to apply after the legal revision of Demetrios of Phaleron (or, more generally, in hellenistic Athens); and, if it ceased to be, that, would again be undetectable in the epigraphic record. For these and other reasons I am unconvinced by many of Habicht's specific attempts to explain the appearance of foreign names as a consequence of naturalisation.

Perhaps the most striking use of the evidence of names to carry out not only "histoire", but even "préhistoire" par noms, is that of Hatzopoulos (pp. 99-117), who (often summarising other recent work) uses onomastic evidence compiled in preparation for the projected LGPN IV to explore the origin and diffusion of Macedonians and Macedonian influence. Drawing on epigraphic evidence, he shows, for example, how Macedonian onomastics can be used to illuminate the Macedonian impact in areas such as Anthemous, Morrylos, Amphipolis, Kassandreia and Philippi. Then, arguing from other recent work, in particular a recent (posthumously published) paper of Masson,3 he argues that certain "Macedonian" names, while Greek in origin, reflect a "wavering between voiced and unvoiced consonants" also characteristic of the Perrhaibian Tripolis. This area also shares other onomastic features with Macedonia, thus confirming the (late 8th cent., in Hatzopoulos' view) evidence of Hesiod, Cat. of Women fr. 7, that Macedonians at that time dwelled in the area of Pieria and Olympos. Macedonia experts will find much food for thought (and perhaps argument) in this bold and engaging paper.

Hornblower's paper shows how analysis of names of persons mentioned by historians can be used as a check on the historian's reliability. He gives several convincing examples of correspondences between names used by historians and external onomastic evidence, confirming the name's authenticity for the individual, or the region, in question, but does not neglect to address the possibility of "plausible onomastic fiction" (i.e. authors using names characteristic of a region for men who are nonetheless fictional creations). He concludes that whether such fiction is likely depends on whether the historian had a plausible motive for it. He finds no such motive in the case of Thucydides. This analysis will, I think, meet with a fair level of assent, though the intellectual level sometimes dips. At one point (p. 133), for example, Hornblower describes an elderly American scholar as a "positivist", a fashionable "hoorah"--or more often "boo-word" about as content-free (and unhelpful to rational discourse) as "fascist" when chanted by anarchist protesters at an international financial conference. It (and many other "-ist" labels) have no place in a serious academic publication unless the author has defined precisely what intellectual attitude it is supposed to designate and has reflected on how this is relevant in the context of his argument.

Crawford's short paper (pp. 145-48) is about naming practice in a fictional context, showing how names in three stories preserved by Phlegon of Tralles can be used to trace the story's historical-literary or regional origin.

The book includes a helpful index of names and a good general index, but no index locorum.


There is only one aspect of the impact of LGPN which I should like to have seen discussed more fully in this book, namely its impact on epigraphy. As several contributors point out, epigraphists have contributed the immense majority of the data (even for the literature-rich Attica volume well over 90% by my rough calculation) contained in LGPN. But this is a two-way process. The onomastic problem that interests an epigraphist, and it is one he faces more or less every day, is that of reading, restoring and articulating fragmentary names in inscriptions. His capacity to do this rationally is being immensely facilitated by the appearance of LGPN. I discuss elsewhere4 some epigraphical-onomastic problems and issues and present around 80 new name amendments which arise wholly or in part from the application of the implications of data contained in LGPN volumes to about 10% of the persons listed in LGPN II. This suggests (and I have probably missed quite a few potential amendments) that LGPN II has the potential to facilitate its own amendment to the extent of about 1.5% of its content (i.e. 800 of 60,000 entries). When this is combined with the results of other vigorous research currently underway in Attic onomastics (e.g. John Traill's, Persons of Ancient Athens) and epigraphy (e.g. publication of new inscriptions and the work now underway for IG ii3, initial results of which suggest that its onomastic harvest will be considerable), it can be seen that LGPN II (and to a perhaps lesser extent--because data are more thinly spread and therefore less rich in implication--the other volumes) contains within itself the seeds of its own, fairly swift, obsolescence. In that regard the decision to post addenda and corrigenda on the LGPN website ( is very welcome. (I understand that, for Attica, a new, extended set, collected by Sean Byrne, will shortly be made available). The question of a second edition of the Attica volume, however, will before long have to be faced. When it is, questions of format change will arise. The general scheme of the individual entries, I think, is wonderfully concise and requires little adjustment. I also generally agree with LGPN II's conservative policy with regard to identifications (though adjustments are continuously becoming necessary in individual cases). I do, however, have three suggestions for improvements:

(a) persons from homonymous Attic demes are currently all listed together under the same deme heading. It would considerably enhance the usefulness of this volume and reduce its potential to mislead if the homonymous demes were disaggregated;

(b) part names are currently omitted. This is justifiable in most cases, but is unfortunate in respect of unique (in the volume) name components, especially where such components preserve the initial letter of a name. It is of interest, for example, that there was an Athenian uniquely (in any published LGPN volume) named Ameinoph[-,5 for this confirms that, like the related common names in Kalli- and Aristo-, the more unusual Ameino- names sometimes took terminations in ph-. But, on current rules, this man would not be listed in LGPN II, since there are several possible ph- terminations and his name can not therefore securely be restored. That is a pity;

(c) In the pre-LGPN era it was often very difficult to make good restorations of personal names in inscriptions. Because of this, many (as can now be seen) irrational name restorations have worked their way into the Attic onomasticon. It would helpful if, before a second edition is published, a review could be carried out to ensure that, where a restored name is listed in LGPN II, it is listed under the epigraphically and onomastically most likely restoration(s).


1.   Quoted by Hatzopoulos, p. 99. Robert presented "histoire déspar noms" as a preferable alternative to "faire catalogues de noms", but, as Hatzopoulos points out, and as the abundant use of LGPN by every one of the papers in the volume demonstrates, the former can not be done without the latter.
2.   A slight adaptation of Morpurgo Davies, p. 39.
3.  ZPE 123 (1998), 117-120.
4.   In the proceedings of a conference in memory of Adolf Wilhelm held in Athens in 2000, due to appear shortly. Some examples of the amendments at ZPE 135 (2001), 51-62.
5.   His existence is revealed in my paper in the conference proceedings referred to in n. 4.


Leave a Reply

Your email address will not be published. Required fields are marked *