BMCR 2000.09.06

Between Geography and History: Hellenistic Constructions of the Roman World

, Between geography and history : Hellenistic constructions of the Roman world. Oxford: Clarendon Press, 1999. xi, 407 pages ; 23 cm. ISBN 0199240035 hb. $85.00.

This is a fresh, lively and original study of three writers of antiquity who dealt with geographical as well as historical themes: Polybius, Posidonius, and Strabo.1 Clarke (henceforth C.) argues, on the one hand, that their representation of geographical space is more complicated than commentators on the history of geographical ideas have realized and, on the other hand, that classical scholars have tended to define geography too narrowly when assessing the contributions of these writers. The loss of texts is a problem in the case of all three authors. C. concentrates on the problem in connection with Posidonius, whose Histories and On Ocean have not survived, but also notes it in connection with Polybius, whose History has survived only partially, and Strabo, whose Geography survives largely intact, but not his History. Lost works can be reconstructed to some extent through references made by later authors (‘fragments’). An appreciation of the geographical and historical ideas of Polybius, Posidonius and Strabo is thus tied up with the basis on which such references are interpreted, and it becomes clear that the ‘constructions’ of C.’s title are on more than one level. By arguing from what survives of each author’s work and applying it to what does not survive and by looking at all three authors in conjunction and in the broader context of Hellenistic historiography, C. establishes a platform from which to assess these authors on their own terms. She argues that the categorization of aspects of their work as ‘geographical’ too often reflects definitions used by modern scholars, rather than necessarily those used by the ancient authors themselves.

Classicists will find C.’s detailed approach extremely gratifying. She makes frequent reference to the text or, in the case of lost works, ‘fragments’ of her three authors, and also to other writers of the same period, particularly the first century BC, whose works were universal in scope and reflect the contemporary expansion of Roman power. C.’s book is also aimed at non-classicists. She starts the book with a discussion about the nature of history and geography which draws on the modern debate, and states: ‘…a study of the ancient sources in the light of the arguments of modern geographers can contribute to refining some of their assertions, with the result that both the ancient and the modern studies are enriched’ (4). C. is always careful to provide quotations in English, with the Greek generally provided in parentheses or footnotes. However, C. assumes some familiarity with the texts under discussion.2 She also assumes familiarity with the somewhat peculiar terminology used in the study of lost texts and their ‘fragments’.3

Chapter 1 is a wide-ranging and discursive look at the general question of how geography as a discipline is differentiated from history. C. adumbrates many themes which recur in her later discussions of the individual authors. C. takes two models from the modern debate, in which the contrast between geography and history is seen as the contrast ‘space : time’ and ‘present : past’. She looks at the value of the two models in relation to the modern disciplines, in relation to how ancient authors saw their subject matter, and as a way for classicists to look at ancient writers. She notes the importance of differentiating between space and time as abstract or experienced, continuous or discrete. Two particular points which she brings up in this chapter, and to which she makes frequent reference in later chapters, are the misapprehension that abstract space could not be conceptualized in antiquity, against which she holds up the example of Polybius ( e.g. 80, 103 with n. 59, 114), and the concept of ‘place’ as ‘lived-in space’, which she finds exemplified in Strabo’s work ( e.g. 217, 281). C. traces the development of interest in geography by classicists. She praises Jacoby for refusing to draw a sharp distinction between history and geography in his collection of the fragments of Greek historians, in contrast to the earlier practice of Müller: C.’s study of the organizational principles of their respective works (56-72) stands in its own right as a valuable discussion about the problems of organizing (and hence of using) fragment collections. C. defends her own view that much Hellenistic history was Herodotean in nature. She chooses to treat her three authors together because ‘they all wrote during the protracted period through which a single power, Rome, was gradually transforming the world…’ (73).

C. deals with her first author, Polybius, in Chapter 2. As she states, Polybius’ History is relatively well-known. It is, however, a fragmentary work beyond the first five books, even though extensive excerpts have aided its reconstruction, and much of the valuable discussion on fragmentary texts which C. provides later in connection with Posidonius (130-185) is relevant to Polybius, too. In particular, C. argues against the traditional allocation of ‘geographical’ fragments to a particular ‘geographical’ book (assumed to be Book 34) of Polybius’ History (104). This is a perfectly valid argument but one which is sometimes obscured when C. cites fragments of Polybius according to this traditional allocation.4 She treats, for example, a passage of Strabo as Polybius 34.12.12, even where she argues that the information to which Strabo refers was not necessarily presented by Polybius in a ‘geographical’ book (105); and a passage from Geminus as Polybius 34.1.7, even where she argues that Geminus is referring to a work by Polybius other than the History (78). Sometimes, C. uses a clearer form of citation, which helps her argument: ‘Among the fragments assigned to Book 34 is one from Strabo in which Polybius is cited as having written…(34.1.3-6)’ (92-93); ‘Fragments assigned to Book 34 testify to Polybius’ interest in astrology and cosmology. He is said to have written a work entitled ‘On the parts of the globe under the celestial quarter’…(34.1.7…)’ (112).5

C.’s argument that there is no need to suppose the existence of a separate ‘geographical’ book in the History is part of a larger argument, that geography and history were fused in Polybius’ work to reflect their inseparability in reality (79-97). C. argues that Polybius was able to conceptualize abstract space separately, for example by the use of geometric figures (97-114), but that he subordinated space to time in his work, and she sees the explanation as lying in the nature of his ‘universalism’. Polybius took as his subject the extension of Roman power over almost the whole world (114-28).6 C. argues that Polybius’ view of the convergence of world affairs from 220 BC onwards means that ‘spatial separation no longer gave rise to different histories, so space could not be the primary matrix against which Polybius’ account was written; rather it was subordinate to time’ (119). C.’s point about how closely geography and history are related in Polybius’ account is good, and it is true that there is no need to assume that ‘geographical’ fragments should necessarily be relegated to a ‘geographical’ book. However, Strabo’s specification of Ephorus and Polybius as among those ‘in their universal writing of history separately laying out the topography of the continents’ (8.1.1) suggests that, somewhere in the History, Polybius isolated at least this one aspect of geography in such a way that it was considered ‘separate’.

In Chapter 3, C. deals with Posidonius and his On Ocean (139-54) and Histories (154-70, with 132-36). C. emphasizes the possibility that these two works were broad in nature and stresses the consequent difficulty in identifying to which of these works ‘unplaced’ fragments (references to Posidonius without mention of a particular work) should be allocated (171-85). She prefers Jacoby’s method of allocating ambiguous fragments into a broad ethnographical category over Kidd’s more detailed classification, which C. finds misleading (171). Her analysis of how given fragments are differently categorized by Jacoby and Kidd well illustrates her general point that modern scholars bring their own ideas of what constitutes ‘geography’ and ‘history’ to their study of lost works. C.’s preference for Jacoby’s inclusion of an ethnographical category of fragments which ‘allowed for considerable generic flexibility between the works’ presumably explains her decision to cite fragments according to the numbering allocated by Jacoby rather than Kidd (139).7

C.’s own approach is first to analyze the ‘one securely attributed extant fragment’ (139) of On Ocean,8 Strabo 2.2.1-2.3.8. She argues that it reveals an interest in ethnography and in the use of travellers’ tales to establish the circumnavigability of Libya; she notes Strabo’s characterization of the work as geographical in a way that is partly ‘fitting’ and partly ‘more mathematical’; and she concludes that the work is more wide-ranging and less narrowly scientific than is often supposed from the fragments containing discussions of tidal theory which are attributed to it (145-52, 171-79). C. supports the view that a Periplus or Periegesis attributed to Posidonius by Pliny should be identified with On Ocean (152). She makes the further suggestion that On Ocean‘came in the form of a real or imaginary journey around the outer edge of the continents …’ (153). This further suggestion seems to me unlikely. Even in a periplus text, the significance of the Ocean is likely to have been as a boundary. The theory of a circumfluent Ocean acting as a boundary to the ‘inhabited world’ was traditionally based on arguments about the behaviour of known seas, precisely the subject matter of many Posidonian fragments.

Seeing On Ocean as a periplus text opens up the possibility that some of the unplaced, but generically ambiguous, fragments might be assigned to it, which in turn allows us to see it as less ‘geographical’ and more ‘historical’.9 Conversely, C. argues that the Histories may have contained more ‘geographical’ information than is sometimes assumed (154-169, 132-34, Appendices A and B). She challenges the assumption that the Histories were written according to a strict chronological pattern, takes exception to the practice of emending book numbers in citations of the Histories according to preconceived ideas of how the work was arranged, and suggests that regional considerations played at least some role in the organization of the Histories. Moving away from a generically strict characterization of the Histories has implications for the assignment of unplaced fragments (179-85), and C. rejects the separate monographs that have been hypothesized to accommodate awkward fragments (175, 177, cf. 155). C. concludes her treatment of Posidonius with a discussion of his ‘universalism’, which might explain and reinforce the overlapping nature she has suggested for On Ocean and the Histories (185-92). She sees their overlap as part of Stoic thought, where all processes are inter-related and under the sole direction of fate (189). However, she also sees both works as fitted to ‘the requirements and preoccupations of an age in which Rome’s world dominion was becoming firmly established’ (191, cf. 137).

C.’s third and main author is Strabo, to whom Chapters 4-6 are devoted.10 Despite stating ‘I shall conclude by arguing that both Strabo’s Geography and his History fell between geography and history’ (195), C. concentrates on the Geography and does not address in any detail the question of the nature of his lost History. This is a shame, since C.’s views would have been welcome. Nevertheless, these three chapters provide fertile ground for a new way of looking at the Geography. C. devotes Chapter 4 to a study of Strabo’s spatial conceptions. She argues that Strabo’s clock-wise description of the world around the Mediterranean is influenced by the periplus format (197-210); that he is little influenced by the scientific study of, for example, zones and latitude, which are dealt with only in the first two books; and that this is consistent with his professed intention to concentrate on the ‘inhabited world’ (207-209). In addition, C. sees Strabo as using a spatial model that might be conceived of as circular, to the extent that it had a periphery and was centred, although not literally, on Rome, Italy and the Mediterranean (210-28). C.’s view on the centrality of Rome is illuminating, but I question her views on the centrality of the Mediterranean (e.g. 198, 206, 216). What Strabo calls the ‘inner sea’ or ‘our sea’ comprised the Black Sea as well as the Mediterranean, and surely it is this whole system, conceived of as one gulf, that is central to Strabo’s exposition.11 As well as looking at spatial models for Strabo’s narrative, C. also looks for evidence of the author’s own focus (228-44). She finds it in Strabo’s use of the expression ‘in my time’ which she sees as locating the author among the intellectuals of the Greek East (242-43): the temporal aspects of this phrase are discussed in the next chapter.

Chapter 5, ‘Strabo and Time’, is particularly thoughtful. Time is, indeed, a very important element in Strabo’s account, and one which has not hitherto been given the attention it deserves. C. seeks to reconcile Strabo’s professed intention to concentrate on the present with the reality that much of his material concerns the past. She sees the answer in Strabo’s interest in the current identity of places, a current identity which is itself the product of past events and the way that places have been transformed by them. First, C. analyzes Strabo’s use of temporal markers (252-60). She sees the main temporal markers as the Trojan war, the return of the Heracleidae, and the battle of Actium, and notes that Strabo also uses vague terms like ‘long ago’, ‘previously’, ‘now’, ‘later’, which do not provide precise dates. C. concludes that Strabo is not concerned with the construction of a coherent temporal system which could be universally applied (260). C. does note that Strabo sometimes uses the succession of empires as a variation on the theme. I would counter that Strabo uses temporal markers such as ‘in the time of the kings’ and ‘up until the Romans’ more often that C. suggests and that the precise dates they imply vary according to the place under discussion. Perhaps Strabo’s attitude to time might be seen as the converse of his approach to place: his concern with creating a present identity for place out of past events, rather than through an abstract spatial system, results in his use of temporal indicators for the past which depend on place.

C. goes on to provide a valuable analysis of the different types of ‘time’ in the Geography. She looks at ‘internal textual time’, e.g., promises by Strabo that he will return to a topic later, or that he will deal with matters in order (260-61);12 his system of literary time (262-64); and the ‘real, historical past’ (264-81). C. then deals with the ‘present’ in the Geography (281-93). She looks at the expressions ‘shortly before us’ (283) and ‘recently’ (284), and argues that composition appears to have been a gradual and long-lasting process, or that ‘even if the work was finally put together in a relatively short period, the author speaks as though events from the whole of the mid-first century BC onwards formed the back-drop for his composition’ (285). She then states that the ‘first century BC and the early first century AD must be considered Strabo’s “present”‘ (285) and looks at his treatment of events within that period, from the activities of Lucullus and Pompey in the Greek East to events under Tiberius (285-89). It is the ‘present’ thus broadly and vaguely defined which C. sees as lying behind the expression ‘in our times’ which, she argues, is not a precise temporal indicator and has no bearing on Strabo’s own dates or the composition of the Geography (289-93).13 Sometimes, however, she leans towards equating the ‘present’ with Strabo’s own lifetime: ‘the life of the author, which defines the present in the Geography…’ (293); he ‘regarded the whole of his lifetime as his “present”‘ (307). C.’s definition of the ‘present’ explains her treatment of Strabo’s ‘present’ as Augustan (250) and his work as reflecting the first century BC (114 n. 82), while at the same time seeing the Geography‘as a work of the late Augustan/early Tiberian period’ (193 n. 1, 217).

In Chapter 6, C. argues that Strabo’s interest in the transformation of the world leads to a spatial arrangement, as each place is affected individually rather than synchronically, but that the work is nevertheless universal in that each place is related to Rome, which now dominates nearly the whole world.14 She analyzes the sources which Strabo mentions for various areas and explains Strabo’s apparent unwillingness to use regional histories from the Fertile Crescent as the result of the difficulty in assimilating them to the temporal markers of the Graeco-Roman tradition (315-325, and Appendix C). Thus, there is an internal coherence to the Geography, despite its lack of overall chronological system, and it is more than a sum of individual regional accounts. Strabo’s interest in past moments of transformation is related to the transformation of his own world. It is worth looking back here at C.’s earlier comment that Polybius was writing in ‘a period of Roman expansion, resulting in a very different kind of spatial conception from that of Strabo’s Geography, which describes a relatively stable world’ (78, cf. 217); and at her suggestion that Posidonius’ Histories perhaps formed ‘a link between the historical dynamism of the expanding world of Polybius and Strabo’s descriptions of the lands and peoples which comprised the newly established world of Rome’ (192). I would have liked to know whether C. might relate Strabo’s transformed world to his decision to call his work ‘geographical’ and separate it from his History, and what the implications might be for the structure of that lost work, but it would be churlish in a book that offers so much to complain about what is left out. I note C.’s own remarks in the ‘Epilogue’ (337-45) that, although her project has run out of time and space, this ‘does not imply that the issues are closed. Rather, I hope to have raised more questions than could be answered…’ (344).

C.’s book concludes with three ‘Appendices’ and an ‘Index of passages’ (395-99). The latter is essential in a work of this kind, but seems to be seriously incomplete. I noted the omission of several of Posidonius’ fragments cited by C., e.g. on pp. 172 n. 65, 173, 174 with n. 72, 175 and 180. Also, note that fragments of Polybius and Posidonius are listed under those authors, not under the authors from whom the fragments are taken. As for errors in the main body of the book, I noticed a couple of misrepresentations of Strabo’s self-referential statements. C. states (316) that Strabo ‘visited’ Alexandria, whereas Strabo’s words indicate that he stayed there ‘for a long time’ (2.3.5). C. states (234) ‘Strabo’s maternal grandfather, Dorilaus, … caused fifteen Mithridatic garrisons to revolt to Lucullus so as to be on the winning side … (12.3.33)’. Strabo had two ancestors called Dorylaus: one was his great great grandfather; the other was a cousin of Strabo’s great grandfather Lagetas.15 Neither is to be confused with the grandfather who took 15 garrisons over to the Romans, who is not named; nor is it certain (textual readings vary) that he was Strabo’s maternal grandfather. Also, note that the date of 7 BC for the annexation of Amaseia has been much disputed since it was given by Pais (234). Finally, the assumed rate of travel of ’50 stades’ in a day (200) should read ‘500 stades’.


1. C.’s ‘main’ (2, 339) author is Strabo, and her book is one of a triad published on Strabo in the space of just over a year. The other two are Johannes Engels, Augusteische Oikumenegeographie und Universalhistorie im Werk Strabons von Amaseia (Stuttgart: Steiner, 1999), and Daniela Dueck, Strabo of Amasia. A Greek Man of Letters in Augustan Rome (Routledge, 2000).

2. There is now an English translation of the fragments of Posidonius: I.G. Kidd. Posidonius. Vol. 3. The Translation of the Fragments (C.U.P., 1999). Note that C. uses fragment numbers assigned by Jacoby, rather than by Kidd. For C.’s own translations of Fragments 1-27, 38 and 61, see Appendix B (347-373).

3. For example, Athenaeus is a later writer who cites Posidonius. Athenaeus’ citations are treated as fragments of Posidonius. Thus C. (132) speaks of Athenaeus as a ‘source’ for Posidonius’ Histories.

4. The traditional allocation is that adopted by Büttner-Wobst. Note the comment of Walbank, A Historical Commentary on Polybius, vol 2 (1967), v: ‘It has now become standard practice to quote according to Büttner-Wobst and there is clearly something to be said for some rule of this kind, provided that Büttner-Wobst’s arrangement does not as a result acquire the sanctity of holy writ’.

5.. C. elsewhere represents the name differently as ‘On the inhabiting of the equatorial region’ (78); elsewhere, differently again, as ‘ On the Habitation of the Equatorial Region‘ (144). Note that even parts of works could have names or be referred to by descriptions of their contents. See C.’s note on Arrian (156 n. 49); and Strabo apparently refers to separate parts of his History by descriptive names, although these references are sometimes taken to indicate different works (11.9.3, 2.1.9).

6. C. generally uses ‘universal’, ‘world’, and ‘global’ as pertaining to the ‘inhabited world’, the world as it was conceptualized by Hellenistic authors, e.g. when she notes, against Polybius’ claim, that ‘Rome had not conquered the entire world’ (118). This leaves the problem of how to differentiate these terms when applied to the entire sphere of the earth and the celestial sphere. C. acknowledges that Polybius’ geographical conceptions were universal ‘in the true sense of the word’ (118); and sometimes she appears to use ‘global’ in this wider sense (e.g. 147).

7. In Jacoby’s collection of fragments, Posidonius is FGrH 87, not FGrH 91, as C. erroneously states in the ‘Index of passages’ s.v. Posidonius (396). Similarly, the testimonium on Posidonius cited as FGrH 91 T 2 (144 n. 25) would have been better given as FGrH 87 T 12b. FGrH 91 is Strabo.

8. The notion of a ‘securely attributed’ fragment begs the question of exactly what the fragment is securely attributed to, and this is a question of how we render the names used by ancient authors in referring to other works. Strabo’s reference at 2.2.1 may be to what Posidonius says ‘in On Ocean‘ or ‘in the [books, volumes, discussions?] about Ocean’. Cf. Strabo’s forms of reference to Posidonius at 1.3.12, 1.1.9 and 2.3.3.

9. Might On Ocean be referred to as a ‘history’? I note C.’s own comment on ‘histories’ as meaning essentially ‘pieces of research’ (170). Athenaeus refers (9.401A) to a statement of Posidonius ‘in the History‘, which C. herself contrasts with Athenaeus’ other references to the ‘ Histories (plural)’ (372 n. 8). We could compare Strabo’s statement (11.1.6) that Posidonius wrote a ‘history about him/it’, or perhaps a ‘ History about him/it’, where the pronoun might refer to Ocean. C. herself prefers to see it as referring to Pompey (155).

10. C. published some of the material from these chapters separately: ‘In Search of the Author of Strabo’s Geography‘, JRS 87 (1997) 92-110.

11. The same objection applies to C.’s use of ‘the Mediterranean’ in connection with Polybius (113).

12. These and similar remarks by Strabo are treated by Daniela Dueck, ‘The Date and Method of Composition of Strabo’s Geography‘, Hermes 127.4 (1999) 467-78, esp. 469-70, as indicative of composition over a relatively short period of time. Note also Dueck’s remarks on the use of ‘recently’ (469).

13. In ‘The Expression “Our Times” in Strabo’s Geography‘, CP (1997) 235-46, I argue that the expression has no direct bearing on Strabo’s birth-date, but that Strabo considered ‘our times’ as beginning in the mid-60’s BC with the Roman conquest of Pontus and Pompey’s reorganization of Asia.

14. Polybius, too, made this claim for Rome, although C. points out that this is something of a conceit (121).

15. For Strabo’s family tree, see Pothecary, ‘Strabo: his Name and its Meaning’, Mnemosyne 52.6 (1999) 691-704, esp. 699-701, 704.