EARLY GREEK KINSHIP by EMILY KAREN VARTO BA(H), Queen’s University, 2002 MA, Dalhousie University, 2004 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Classics) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) December 2009 © Emily Karen Varto, 2009 Abstract Kinship is an important factor in modern explanations of social, political, and economic change in Early Greece (ca. 1000-450 BCE), particularly in social evolutionary schemes that see states develop from kinship-based clan societies. Following challenges to such schemes in several disciplines, including Classics, and following theoretical and methodological upheavals in anthropological kinship studies, our ideas and methodologies concerning families, descent groups, and kinship in Early Greece need to be reconsidered. In this dissertation, in order to avoid both applying typologies and employing universal biological kinship terminologies as points of analysis, a contextual methodology was developed to explore textual and archaeological evidence for ideas of kinship. Using this methodology, the expression and manifestation of kinship ideas were examined in Early Greek genealogical material, burial practices and patterns, and domestic architecture, taking each source individually to achieve a level of interpretative independence. Early Greek genealogies are usually linear and descendent-focused or tendrilled and ancestor-focused, and include sections of story-telling that are an integral part of the descent information. List-like genealogies are therefore not the standard structure for Early Greek genealogies and the few late extant examples may be associated with literary techniques or epigraphic traditions. The genealogies are mythico-historical and connected the legendary past with the present in the interests of individuals and states and were not charters determining status or membership in particular groups. Early Greek burial practices and patterns were informed by an idea of descent and an idea of households over a few generations, represented by small mixed burial groups. Residency patterns and changes in Early Greek domestic architecture suggest household units, some of which were participating and became successful in the domestic economy and in agricultural trade. A synthesis of the ii evidence reveals three broad overlapping Early Greek kinship ideas: blood and biology, generational households, and descent and ancestors. These ideas involve inheritance, ethnicity, success, wealth, and elitism. They therefore illuminate kinship’s role in social, political, and economic differentiation and power and resituate it in theorizing about the developing Greek polis. iii Table of Contents Abstract ................................................................................................................................ii Table of Contents................................................................................................................. iv List of Tables .....................................................................................................................viii List of Figures...................................................................................................................... ix List of Abbreviations...........................................................................................................xii Preface ................................................................................................................................ xv Acknowledgements............................................................................................................ xvi Chapter 1: Introduction ......................................................................................................... 1 Issues with Sources and Evidence ..................................................................................... 3 Quasi-Historical Trajectories ......................................................................................... 3 Homeric Society............................................................................................................ 4 Solonian Research ......................................................................................................... 5 Late Sources.................................................................................................................. 6 Theoretical Assumptions and Disciplinary Tendencies ...................................................... 7 After Evolutionary Typologies ...................................................................................... 7 Studying a Divided World ........................................................................................... 10 Women and the Family................................................................................................ 12 Primacy ....................................................................................................................... 13 The Aristotelian Model................................................................................................ 14 ‘Traditional’ Kinship Theory....................................................................................... 15 Kinship Theory ............................................................................................................... 16 Sources for Early Greek Kinship ..................................................................................... 24 Genealogy and Kinship ............................................................................................... 27 iv Archaeology and Kinship ............................................................................................ 28 Part I: Genealogy and Early Greek Kinship ........................................................................ 31 Chapter 2: Early Greek Genealogical Material .................................................................... 32 Sources of Early Greek Genealogical Material ................................................................ 44 Homeric Poetry ........................................................................................................... 44 Hesiod’s Theogony ...................................................................................................... 47 The Hesiodic Catalogues ............................................................................................. 48 Other Early Greek Poetry ............................................................................................ 49 Early Greek Prose Genealogists................................................................................... 53 Herodotus’ Histories ................................................................................................... 55 Epigraphical Sources ................................................................................................... 56 Structure and Scope in Early Greek Genealogical Material.............................................. 57 Genealogies and Family Trees ..................................................................................... 58 Branching and Grouping.............................................................................................. 63 Narrative Style and Story-Telling ................................................................................ 75 Chapter 3: List-Like Genealogies and Historiography ......................................................... 90 Genealogy, Chronology, and Hekataios........................................................................... 92 Herodotus’ List-Like Genealogies ................................................................................. 100 The Genealogy of Miltiades the Oikist .......................................................................... 107 The Genealogy of Heropythos ....................................................................................... 115 Chapter 4: The Types and Uses of Early Greek Genealogies ............................................. 120 Genealogy and the Individual ........................................................................................ 130 Genealogy and the State ................................................................................................ 150 Conclusion .................................................................................................................... 160 v Part II: Archaeology and Early Greek Kinship................................................................... 163 Chapter 5: Burials and Early Greek Kinship ...................................................................... 164 Osteology, Biological Kinship, and Burial Groups ........................................................ 166 Recognizing Intentional Burial Groups.......................................................................... 176 Ideas of Kinship in the Grouping of Burials................................................................... 182 Descent ..................................................................................................................... 183 Households and Generations of Households .............................................................. 193 Burying Men, Women, and Children ............................................................................. 213 Conclusion .................................................................................................................... 219 Chapter 6: Domestic Architecture ..................................................................................... 229 Evidence, Typologies, and Approaches ......................................................................... 232 House and Settlement .................................................................................................... 243 Spatial Relationships between Houses ....................................................................... 243 Kinship and House/Settlement Type .......................................................................... 246 Social Differentiation and Housing ............................................................................ 253 Internal Organization..................................................................................................... 257 Access Patterns and Functionality.............................................................................. 260 Finds and Features..................................................................................................... 265 Functional Specialization........................................................................................... 275 Conclusion .................................................................................................................... 292 Chapter 7: Early Greek Kinship Ideas: A Concluding Synthesis ........................................ 311 Blood and Biology ........................................................................................................ 312 Generational Households............................................................................................... 319 Descent and the Importance of Ancestors ...................................................................... 328 vi Final Conclusions.......................................................................................................... 337 Bibliography ..................................................................................................................... 339 Appendix 1: The Genealogy of Miltiades the Oikist .......................................................... 369 vii List of Tables Table 5.1. Cremation burials at Vroulia ............................................................................ 222 Table 5.2. Model of the life cycle of the ancient household ............................................... 223 Table 5.3. Adult burials in hypothetical generational groups of households ....................... 223 Table 6.1: Six types of Early Greek houses according to Nevett ........................................ 295 Table 6.2. Evidence associated with certain activities in Late Geometric II at Zagora....... 297 viii List of Figures Figure 2.1. The family tree of the Trojan royal family constructed from the information in the Library of Greek Mythology (adapted from Hard 1997, 22) ........................................... 88 Figure 2.2. The family tree of Aineias constructed from the information in Il. 20.200-41 ... 88 Figure 2.3. The outline of the genealogy of Aineias from Il. 20.200-41 .............................. 89 Figure 3.1. The Genealogy of Heropythos, Grave Stele from Chios, ca. 450 (SGDI 5656) following Wade-Gery’s transcription (1952, 8: fig. 8) ....................................................... 118 Figure 3.2. The Genealogy of Heropythos, Grave Stele from Chios, ca. 450 (SGDI 5656) (Wade-Gery 1952, 8: fig. 8) .............................................................................................. 119 Figure 5.1. Hypothesized family plots at the Pantanello Necropolis near Metaponto (Metaponto Necropoleis, 162, fig. 5A.7) ........................................................................... 224 Figure 5.2. The burials in hypothetical family group 10.3 at the Pantanello Necropolis near Metaponto (following Metaponto Necropoleis, 337, fig. 7.10)........................................... 225 Figure 5.3. Burial enclosure south of the Acropolis on Erechtheion Street (Meliades 1955, 41, image 2) ...................................................................................................................... 225 Figure 5.4. Geometric burials in the North Cemetery at Corinth (Corinth XIII, plan 2) ..... 226 Figure 5.5. Depth of cremation burials at Vroulia by number of cremations in each burial. 227 Figure 5.6: Tomb cluster 159-168 at the necropolis at Pithekoussai (Pithekoussai I, plan A II bis).................................................................................................................................... 228 Figure 6.1: Changes in the large central house at Kastanas (after Lang 1996, figs. 130-34) ......................................................................................................................................... 298 Figure 6.2. Site plan of Zagora (Zagora II, plan 1) ........................................................... 299 Figure 6.3. Zagora area H in late Geometric a) phase 1 and b) phase 2 (after Zagora II, plan 12b, c)............................................................................................................................... 300 ix Figure 6.4. Zagora area D in late Geometric phase 2 (Zagora II, plan 6)........................... 300 Figure 6.5. Detached houses at Nichoria (Nichoria III, fig. 2-10a,b) ................................. 301 Figure 6.6. Figure 6.6. Detached houses at Koukounaries (Schilardi 1983, fig. 3)............ 301 Figure 6.7. Detached houses at Emporio (Boardman 1967, fig. 4) .................................... 302 Figure 6.8. Compound at Oropos (Mazarakis-Ainian 2006, fig. 10.9)............................... 302 Figure 6.9. Agglomerated structures at Vrokastro: upper site (Sjögren 2003, fig. 60)........ 302 Figure 6.10. Agglomerated structures at Vrokastro: lower site (Sjögren 2003, fig. 59)...... 303 Figure 6.11. Agglomerated structures at Phaistos (Sjögren 2003, fig. 28) ......................... 303 Figure 6.12. Rows of houses at Vroulia (Lang 1996, fig. 64) ............................................ 304 Figure 6.13. Proposed Archaic plan of Euesperides (Gill and Flecks 2007, fig. 22.2)........ 304 Figure 6.14. Agglomerated structures at Prinias (Sjögren 2003, fig. 20) ........................... 305 Figure 6.15. House in block 18 at Megara Hyblaia in a) the seventh century and b) the sixth century (after Megara Hyblaea I, plans 11, 12) ................................................................. 306 Figure 6.16. Figure 6.16. Changes in a house at Miletos (Kalabaktepe) during the seventh century (Morris 1998, fig. 9) ............................................................................................. 306 Figure 6.17. Domestic architecture at Old Smyrna in area H a) from the last half of the seventh century and b) from the sixth century (after Akurgal 1983, fig. 19, fig. 30) ........... 306 Figure 6.18. House A at Onythe (Platon 1955, fig. 1) ....................................................... 308 Figure 6.19. Multiple room structure at Xobourgo (Kontoleon 1953, plate 1) ................... 308 Figure 6.20. Relief pithos from the multiple room structure at Xobourgo (Kontoleon 1953, fig. 9) ................................................................................................................................ 309 Figure 6.21. Changes from curvilinear to rectilinear houses at Eretria (Morris 1998, fig. 5) ......................................................................................................................................... 310 x Figure 6.22. House H/L, J, G at Thorikos (Thorikos III, plan 2)........................................ 310 xi List of Abbreviations The following list provides abbreviations used throughout for books and some periodicals. Other periodicals are abbreviated according to the guidelines of the American Journal of Archaeology. Periodicals AION ArchStAnt n.s. AION Annali dell’Instituto universitario orientale di Napoli: Sezione archeologia e storia antica, Nuova Serie Am Anthr American Anthropologist Am Ethnol American Ethnologist Am J Phys Anthropol American Journal of Physical Anthropology Annu Rev Anthropol Annual Review Anthropology AncSoc Ancient Society AP3A Archaeological Papers of the American Anthropology Association JFH Journal of Family History J Gerontol Journal of Gerontology JITE Journal of Institutional and Theoretical Economics J Radioanal Nucl Chem Journal of Radioanalytical and Nuclear Chemistry RivAntro Rivista di Antropologia Books, Collections, and Series Alt-Smyrna I Akurgal 1983. Athenian Agora VIII Brann 1962. xii Athenian Agora XIV Thompson and Wycherley 1972. Corinth XIII Blegen et al. 1964. EGF M. Davies 1988. EGM Fowler 2000. FrGrHist Jacoby 1923-58. IG Inscriptiones Graecae Inscr. Cret. Inscriptiones Creticae Kerameikos IX Knigge 1976. Kerameikos V Kübler 1954. Kerameikos VII Kübler 1976. Kerameikos VII.2 Kunze-Götte et al. 2000. Kerameikos XII Koenigs 1980. LSJ Liddel et al. 1996. Metaponto Necropoleis Carter et al. 1998. Milet VI Herrmann 1998. Nichoria III MacDonald et al. 1983. OCT Oxford Classical Texts PEG Bernabé 1987. Pfohl Pfohl 1967. Pithekoussai I Buchner and Ridgway 1993. SEG Supplementum Epigraphicum Graecum SGDI Collitz and Bechtel 1884-1915. xiii Thorikos I Mussche et al. 1965. Thorikos III Mussche et al. 1967. Zagora I Cambitogolou et al. 1971. Zagora II Cambitoglou et al. 1988. xiv Preface A note on spelling Greek names and places: I have chosen to use mostly Greek forms of transliteration over more traditional or Latinized forms. I do so for mythical figures, historical figures, and less well known historians (e.g., Achilleus over Achilles, Kimon over Cimon, Hekataios over Hecataeus), except in the case of very conventional names of well known historians, where I use the more standard Latinized forms (e.g., Herodotus over Herodotos and Thucydides over Thukydides). Place names generally remain in their more familiar English forms (e.g., Corinth over Korinth) for ease of reference and for consistency with the archaeological materials. There are some exceptions where common sense dictates the choice of a particular form. This system may appear somewhat inconsistent or arbitrary, but strict adherence to either style of transliteration would produce something quite foreign looking indeed and add an unnecessary degree of difficulty for the reader. xv Acknowledgements First and foremost, I would like to acknowledge the input, guidance, and support of my doctoral supervisor, Dr. Franco De Angelis. The importance of his advice and encouragement throughout this entire process cannot be overstated. Acknowledgement and gratitude are owed to my committee members, Dr. Hector Williams and Dr. Anthony Podlecki, for their invaluable comments and support of this work, as well as to Dr. Philip Harding for his careful reading of a large part of this project. I would also like to acknowledge Dr. Robert Fowler, who kindly provided some advice on the Greek mythographers. Any errors, omissions, or inconsistencies are, of course, my own and should in no way be ascribed to those who have so kindly offered their comments and advice. This research was made possible by funding from SSHRC through their Canada Graduate Scholarships Program and by funding from the University of British Columbia and from Green College. I would like to give thanks to the faculty and staff of the Department of Classics at Dalhousie University and of Corpus Christi College for providing encouraging working and teaching environments. Many thanks are also given to my fellow graduate students – friends and colleagues – for their invaluable help and support at many stages and in innumerable ways, academic and otherwise. xvi Dedication To my family, Jukka-Pekka Varto, Elizabeth Varto, and Hannah Varto, for whom kinship needs neither proof of existence nor explanation of importance. And to M.C., who shows me that definitions of kinship are all still relative! xvii Chapter 1: Introduction Although kinship plays a key role in all reconstructions of political and social organization and change (e.g., the development of the polis, the establishment of democratic ideas, the determination of class or citizenship) in Early Greece, we really know and have investigated very little about the Early Greek concept of kinship.1 Not much has been done specifically on kinship in this period. Pomeroy’s Families in Classical and Hellenistic Greece obviously does not go back that far.2 Lacey’s The Family in Classical Greece and Patterson’s The Family in Greek History both cover kinship in this period, but run into evidentiary problems, to be discussed below, but mostly connected with using primarily textual evidence. 3 A more recent volume on kinship and society in the Greek world, Parenté et société dans le monde Grece de l’Antiquité à l’âge moderne, the result of a conference on the topic of Greek kinship held in 2003, presents a number of excellent and theoretically informed articles concerning kinship from antiquity to the present; Early Greek kinship, however, is not specifically addressed. 4 Scholarship for this period often dances around the concept of kinship, talking about family, households, and descent groups or various concepts or things linked to kinship (e.g., clans, citizenship, inheritance, aristocracies), but not necessarily addressing the concept of kinship itself, which some consider to be at the heart of our most basic and earliest bonds as 1 I use the term Early Greece as a succinct way of expressing a combination of Dark Age, Geometric, and Archaic Greece, indicating a period ca. 1000-450 BCE. For the broadness of this term, see my discussion of my approach below, p. 26. 2 Pomeroy 1997. 3 Lacey 1968; Patterson 1998. 4 Bresson et al. 2006. 1 humans and others only Eurocentric fantasy and imposition. Classical scholarship has remained largely immune to such fundamental debates taking place in other disciplines. 5 And although classical scholarship has never had a good picture of family and kinship for this period, our understanding of the concept of kinship and the bonds it supposedly creates has direct bearing on several significant and contentious areas of study in Early Greek history. 1) For a long time, family or kinship was an important factor in theorizing about the development of states, particularly in social evolutionary schemes that saw states develop from previously kinship-based societies. The polis was seen to come about, in such schemes, through a breaking down, reforming, or subsuming of family ties. 2) Kinship and kinship bonds also lie at the heart of the tribalism, tribal models, and social evolutionary schemes that were once (up until the late 1970s) as a standard applied to Early Greece. Although such theories are no longer applied with regularity, some vestigial assumptions remain concerning the nature of kinship and kinship-based society in pre-Classical Greece. 3) Kinship is also often cited as a criterion for determining social status and explaining elitism, both of which are fundamental concepts in understanding social, economic, and political change in Early Greece, and in the formation of the polis and its character. 4) The debate over whether there was a fusion or separation of state and society in ancient Greece is largely concerned with whether the state controlled or interfered in society, and especially whether it interfered with what we might consider to be family and kinship matters. 5) How rights, citizenship, obligations and duties, inheritance, and identity (both personal and ethnic) were understood are concepts often linked with kinship in the scholarship. Kinship is essential to many of the 5 See the discussion of kinship theory below, pp. 15-16. 2 great questions of Early Greek scholarship, and yet the concept is largely ill-defined and untheorized in classical scholarship. 6 Early Greek scholarship itself has seen several fairly recent and significant changes: a fairly widespread abandonment of strict evolutionary schemes, a growing dissatisfaction with structuralist analysis, a recognition of the difficulties presented by poetic and late sources, and a greater than ever and increasing use of archaeological evidence and theory in historical studies. Kinship should be assessed or re-assessed in light of such challenges and new directions. The discussion to follow highlights those issues which have had the most impact on the study of kinship in Early Greece. They can generally be characterized in two ways: by the sources which were considered as evidence and the way in which they were used; and by the influential theoretical positions (both implicit and explicit) and disciplinary tendencies followed. Issues with Sources and Evidence Quasi-Historical Trajectories A major evidence or source-related problem is the quasi-historical trajectory which arises out of primarily using literary and textual sources for studying Early Greece. In reconstructing the Greek family and tracing its development from Homeric and Hesiodic society to the historical polis, some scholars end up drawing a progressive line from an earlier literary or semi-literary world of epic poetry to a later historical world. For example, Lacey’s The Family in Classical Greece has a separate section on the family in Homeric 6 Bresson et al. 2006 is a welcome development in this direction, particularly Bonnard’s article on kinship and filiation (Bonnard 2006). 3 society, which the author writes, “perished at some time between the 10th and 9th centuries and the 7th century” before the poleis evolved. 7 Similarly, Patterson’s The Family in Greek History traces the development of the polis-family relationship in a linear manner from Homeric and Hesiodic society to the historical polis.8 Not only is this strange history but it is nearly impossible in such projects to take into account the archaeological evidence which is the bulk of our evidence for Early Greece. This problem has meant that other evidentiary types have been under-utilized or have not been as influential as they should be. For example, the evidence of domestic architecture has been overlooked in understanding Early Greek kinship, although new theories and methodologies for studying it and using it to answer questions about social change have recently been proposed and employed to Early Greek housing by Nevett and Lang. 9 Homeric Society The quasi-historical trajectory also exemplifies our trouble with Homer. Much has been written on ‘Homeric Society’ and the usefulness of Homer as evidence for an historical society or reality. 10 Recent opinions on ‘Homeric Society’ range from completely denying historicity and arguing that the epic world is entirely fictional, to accepting and arguing that it 7 Lacey 1968, 51. 8 Patterson 1998. 9 Nevett 1999; 2007b; Lang 2005. See also Souvatzi 2008 for a theoretically informed methodology for studying neolithic Greek households. 10 Divergent opinions and approaches to ‘Homeric Society’ can be found in Snodgrass 1974; Morris 1986; 1997; Van Wees 1992; Murray 1993, 35-37; Seaford 1994; 2004; Osborne 1996, 147-160; Raaflaub 1997; 1998; Donlan 1997a; Finley 2002; Hall 2007a, 24-26; Ulf 2009b. 4 represents the society of a specific time. A middle ground is often claimed which treats it as a composite world made up of multiple fictional and real worlds and debate centres around how and if it is possible to read evidence out of such an amalgam. In the face of such divergent opinions and considerable critiques, studying ‘Homeric Society’ in isolation as a period or stage within an historical trajectory seems ineffective and fruitless in a historical study, unless absolute historicity can be explicitly proven. Much the same should also be said for Hesiod’s poetry and Hesiodic society.11 This is not to say that Homeric and Hesiodic poetry is useless as evidence, but that we should not study it without remembering the whole host of accompanying questions and debates. Nor should we study it in isolation, to the detriment of other evidentiary types. Solonian Research Many of the prevalent ideas about Early Greek kinship and affiliation, especially at Athens, have arisen out of the study of Solon, his poetry, laws, and reforms. Recent work on Solon, however, has identified serious source-related problems in all areas of research, including the authorship of the poetry, the veracity and provenance of the laws cited in fourth century legal speeches, the reality of land distribution and Solonian reforms, and the circumstances behind the writing of the poems and the political thought or legal spirit they espouse.12 A good amount of circular reasoning is also involved in Solon research since the 11 For discussion of Hesiodic society and history see Osborne 1996, 140-47, 156-60; C. G. Thomas 2005, 88127; Hall 2007a, 24-26. 12 Illustrative of such work is the collection Solon of Athens: New Historical and Philological Approaches (Blok and Lardinois 2006), which marks an important shift in Solonian research. The contributors, although often coming to different conclusions, raise important questions and apply healthy scepticism about the sources and what they indicate. Very intriguing is the use of archaeological approaches and questions, a rather new development in studies on Solon (e.g., Bintliff 2006; Forsdyke 2006; Ober 2006). 5 laws and constitutional reforms have often been interpreted within the framework of a tribal kinship-based society and its transition to a polis-centred society. But that evolutionary framework depends on interpreting Solonian laws and constitutional reforms as a move away from kinship-based power and society. This has begun to change with new assessments of the Solonian property classes and the ideology or political thought in Solon’s poetry. 13 As with other archaic poets, questions have arisen about authorship, intent, and audience in his poetry, involving and prompting further interest in ideologies and cultural ideas. In light of such work, earlier ideas about Solonian Athens and the influence of those ideas have to be reconsidered. Late Sources Recent work on Solon also reminds us that we should be very careful with the evidence of late sources for Early Greece. The laws and constitution of Solon are largely preserved for us through fourth century legal speeches, writers of the Roman period such as Cicero and Plutarch, and Aristotle’s Athenaion Politeia. Each of these sources merits careful criticism especially regarding their particular interest in appealing to Solon, who in later periods, as legendary sage, political reformer, and lawgiver, could be invoked in the name of many causes. 14 Re-evaluation of the late evidence has led scholars to seriously question and 13 E.g., Foxhall 1997; E. M. Harris 1997, 2006; Mitchell 1997; Bintliff 2006; Forsdyke 2006; Van Wees 2006; Ober 2006; Raaflaub 2006. 14 Studies on the late sources include e.g., Hansen 1989; de Blois 2006; Gehrke 2006. 6 in some cases completely re-assess the veracity of the Solonian laws and constitution as we have received them. 15 Another example of the problems with late sources is the association of the Homeric words genos (pl. genea) and phylon (pl. phyla) with the concepts of clan and tribe and the further association of these words with the classical Athenian civic divisions genē and phylê (pl. phylai) found in later sources. It has been argued and widely accepted that the late sixth to fourth century genē and phylê (pl. phylai) are not residual holdouts or remnants of a former tribal system, but rather fictions, involving kinship ideas, created for the political and social purposes and climate of those particular centuries. 16 It is essential to recognize and not be misled by the possible use and abuse of history, myth, and kinship ideas in later centuries and sources. Theoretical Assumptions and Disciplinary Tendencies After Evolutionary Typologies The evolutionist model of the development of the polis is based on seeing a move from a pre-state kinship-focused society to a state-focused society; therefore, theories of kinship are central to the model.17 Early Greece is seen to have had a tribal kinship-based society, sometimes called primitive or pre-state. Powerful family groups, usually labelled 15 E.g., Mossé 1979; Hansen 1989; Hölkeskamp 1992; Scafuro 2006; Blok 2006; cf. the ‘optimistic view’ of the authenticity of the laws and reforms in Rhodes 2006. 16 E.g., Bourriot 1976; Roussel 1976; Snodgrass 1980, 26; Donlan 1985; 2007; Sallares 1991, 197-202; Antonaccio 1997, 252-53; Hall 2007a, 124-25. 17 E.g., the evolutionary model can be found in Lacey 1968; Ehrenberg 1969; 1973; Jeffery 1976; Arnheim 1977; Forrest 1978; Littman 1979; Andrewes 1982; Frost 1994; Seaford 1994; 2004; Coldstream 2003. 7 clans or tribes (genea and phylai), led by a chief, controlled land and the population was indebted to them formally and informally. Land, and therefore political power and wealth, belonged to clans or powerful distinct kinship groups who feuded with one another and competed for power. But as the polis developed, the interests of the community overcame the interests of the family and the influence of family-based clans or tribes and family bonds had to be broken for the cohesion and security of the state because the respective interests of state and family were antithetical. Since the late 1970s, the evolutionist model has come under increasing attack especially for its reliance on the concepts of the genos and phylon and land ownership and property. 18 It is now uncommon to find it stated explicitly and fully formed in recent scholarship, although it does appear implicitly and thus remains somewhat pervasive. Some studies, where the idea of ancient genea and phyla as corporate descent groups have been rejected, aristocratic or elite families have been substituted for tribes or clans while the evolutionist construct remains intact. This is what I call a neo-evolutionist position, which recognizes the problems in accepting the genea and phyla as distinct corporate descent groups in Early Greece and questions the primary concentration in the evolutionist view on land ownership and property, but accepts the progression from pre-state kinship-focused society to a state society and employs typologies or stages of social and political 18 The works of Bourriot (1976) and Roussel (1976) represent a beginning in the challenge to the evolutionary view in classical history. See C. J. Smith 2006, 114-63, for a thorough survey of the historiography and evidence of the Attic genos and a rejection of its existence as a group before the sixth century (esp. 136), its continuation in Classical Athens as a relic of aristocratic society (esp. 136), and its direct correlation with the Roman gens (esp. 140). Further critiques and rejections of the evolutionary view include Snodgrass 1980, 2526; Donlan 1985; Humphreys 1986, 88-89; Morris 1987; Sallares 1991, 197-207; Antonaccio 1995, 252-53; Pomeroy 1995, 111; 1997, 102-3; J. K. Davies 1997; Patterson 1998; Hall 2007a, 123-25. 8 development to understand societies and political change. 19 Such progressive, linear schemes of political change and social development have met serious criticism in other disciplines, which indicates that it is probably time to re-evaluate the usefulness of both the evolutionary and neo-evolutionary model for studying Early Greece.20 Re-assessing kinship and our ideas about it in general can go a long way toward this end and properly situate its importance to society and state formation. Ideas concerning families, descent groups, and kinship in general have yet to be fully reconsidered in the wake of the challenges made by Bourriot and Roussel and in light of the questioning of evolutionist and neo-evolutionist schemes in other disciplines. As Pomeroy argues, there were important implications of the move away from the evolutionary model for women’s history and historiography which remained unaddressed. 21 The same is true for the history of the family and kinship. The study of kinship and kinship ideas in general has met with little interest after being excused as a structuring principle of Early Greek society. 22 Moreover the study of larger kinship groups has receded in favour of studying the household or smaller family unit.23 Meanwhile the study of social and political 19 E.g., Donlan 1985; 1997; 2007; Manville 1990. 20 See Yoffee 2005, 4-21, for a review of the scholarship and theorizing in anthropology and archaeology on the subject, and thorough and somewhat damning critique of neo-evolutionary theory and typologies. 21 Pomeroy 1995, 111. 22 A notable exception is Loraux (2000), who deals with Athenian ideas about kinship connected to the myth of autochthony. Donlan (1985) does re-assess early Greek ‘kinship’ terms, such as genos and phretra, but out of that assessment reconstructs a neo-evolutionary scheme in which society evolves from tribal to stratified and warrior-chiefs become a horizontal aristocratic class (308). The problem is to re-assess ‘kinship’ terms and ideas in a non-evolutionary framework. 23 E.g., Manville writes “Any discussion of Dark Ages culture must start with the oikos” (1990, 58). 9 change has turned to either aristocracies or elites and elite status with little critique of kinship ideas and their role in determining social status.24 A notable exception is Duplouy’s work on elitism in ancient Greece, which looks at the determination of elite status and suggests that political power and elitism may be less dependent on traditional kinship ideas of descent and membership in certain established families than on an individual’s ability to self-promote and to secure new kinship ties.25 For the most part, however, scholarship, even when not evolutionary, retains an understanding of kinship and descent largely informed by evolutionary theory and typologies or by the researcher’s cultural ideas about kinship. These persistent but largely implicit ideas or assumptions reveal the need to re-assess our ideas about kinship following the criticism and in many cases dismissal of the tribal model of kinship-focused society. Studying a Divided World With the rejection of strict evolutionary schemes (but not necessarily neoevolutionary ones) and a continually rising interest in women’s history, studies of ancient kinship have left the political history arena for that of social history. Work has turned, with good reason, away from large politicized descent groups and has come to consider smaller kinship groups, namely the oikos or household. The trouble with this shift, however, is the 24 Osborne (1996) and Morris (2002) write about an elite class and elite ideologies in the seventh and sixth centuries. Neither, however, specifically addresses what kinship ideas may lay behind the formation of an elite class or elite status. Hall (2007) writes about kinship terms in Homer, while criticizing the evolutionary model of a kinship-based society, and also about the problems in assuming there was an established aristocracy early on in epic poetry or in Dark Age Greece, but does not specifically connect the two issues. And so there is a place here to use new methodology and theory from kinship studies to understand the kinship ideas behind elite status. 25 Dupluoy 2006. 10 distancing of ideas of kinship or the family from political history and state formation. Part of this shift and its distancing effect can be attributed to structuralist approaches and their imposition of domains and dichotomies. In structuralist analysis, the family or household is placed in opposition to the state or community, conforming to the analytical binary pairs polis/oikos, public/private, even male/female. 26 Since kinship and state are seen to inhabit two separate, polar opposite domains, the result of employing such analysis is to divide the study of kinship from the study of the state except as they oppose one another. Approaching Greek society through this model is potentially exclusionary and therefore misleading, since it only leaves room for the incorporation of evidence which indicates opposition and none that does not.27 Moreover studies of structuralist domains and pairings have largely been concerned with and cemented in classical Athenian society, which in turn is often viewed as a prime example of a divided world perfect for structuralist analysis. The public/private dichotomy is then used to contrast a thoroughly divided society in which state and household are firmly set apart from one another with an earlier kinship-focused society.28 The resulting scheme of state formation, in which kinship and state are closely connected in earlier periods and then sharply opposed in the classical period, is a strange mix of evolutionary ideas and structuralist analysis. 26 E.g., Humphreys 1983; 2001; Katz 1981; Cohen 1996. Pomeroy, differing somewhat, argues, “The traditional dichotomy public/private used to describe Greek life is misleading. A tripartite division is more accurate: public, domestic/public, and domestic/private” (1997, 18-19); however, she still uses domains or spheres as her principle form of analysis. 27 For critiques of the use of analytical domains and dichotomies, especially those based in gender and ‘biology’, in anthropological kinship studies, see Collier and Yanagisako 1987; Yanagisako 1987; Comaroff 1987. 28 E.g., Katz, although writing a literary article on Iliad 6, contrasts the “divided world” of the classical period with that of the Homeric world in which “the dichotomization of roles, attributes, and spheres of activities is far less rigid, and the opposition between ‘public’ and ‘private’ domains is arguably non-existent” (1981, 19). 11 Women and the Family Influenced by structuralist domains and dichotomies, the scholarship on the social history of ancient Greece often links or equates the family with women and therefore the study of family with the study of women. 29 Linking the two can be sound, when it is not done automatically and when it results from the evidence. Equating, however, is a problem. Following the structuralist framework in which male equals public and female equals domestic, it promotes the assumption that family is primarily a female concern, diminishing or even excluding both male concern or involvement. It also either distances kinship ideas from the public sphere or distances kinship ideas or larger kinship groups from a smaller family unit. Moreover, while female activities in ancient Greece seem to have been largely, but not wholly, what we might characterize as domestic, to assign opposing structuralist domains to genders can be misleading. As Rawson writes, “The public-private opposition was not absolute. We therefore risk distorting women’s experience if we go too far in deemphasizing the public sphere.”30 She adds, “We should try to reconstruct women’s relationship to the city and the state.”31 As complement, I would add that we should also reconstruct kinship’s relationship to the city and the state, breaking down both the opposition between private and public spheres and the equation of the family with women. 29 See Pomeroy’s interesting comments on women’s history and family history, in which she writes that she assumed that she would be concentrating on women while studying the family, since women’s sphere was the family. Her assumption proved false in that “it was easy to lose track of women” (1997, 14) and she attributes this to male interest and bias in the sources. While I accept this argument (there is clearly bias in the sources), it is interesting that this does not lead her also to reconsider the usefulness of dichotomies and spheres of interest as analytical tools. 30 Rawson 1995, 13. 31 Rawson 1995, 13. 12 Primacy The idea of a primacy of either state or kinship groups can be found in both evolutionary and structuralist approaches. 32 Both involve a notion of competition between different types of human relationships. Both evolutionary and structuralist models set the interests of each party up as antithetical and explain political change through the breaking down of certain bonds (usually kinship) in favour of new bonds (those of the state).33 These models have meant studying societies as kinship-focused, i.e., as societies in which there is a primacy of kinship bonds over other bonds, dominated by unilineal descent groups before they become state-focused at the expense of those previously powerful descent groups.34 They, therefore, pit descent systems against societies of low and high levels of complexity and corporate descent groups against states and economic markets in complex societies.35 Although such evolutionary schemes have long drawn serious criticism in many disciplines, neo-evolutionary typologies and evolutionary ideas continue to be adopted in historical and 32 E.g., Manville writes that the oikos “provided the primary principle of classical society, and everything known about earlier time suggests that this had long been the case” (1990, 58); Patterson firmly rejects the evolutionary view of the family-state relationship and its primary focus on the clan, but argues that the primary focus is the oikos or household, emphasizing “the centrality of the household as the primary focus of both family loyalty and identity” (1998, 47). 33 E.g., in Seaford’s work (1994; 2004) a “contradiction between polis and household” (1994, xviii) and the need for the polis to triumph over kinship groups (the clan and the household) in order to become established are major themes. 34 The idea of kinship-based society is implicit, e.g., in Andrewes’ conception of the development of the Athenian polis: he sees a time in which kinship was “the basis for social and political organization” and from which remnants of clan power were held over into the non-kinship-based society of fourth century Athens (1982, 367-68). 35 For examples of such theories, see the discussion in Pasternak et al. 1997, 262-264. 13 anthropological studies of kinship and society, most often implicitly.36 For example, some scholars see the formation of kinship bonds as being in competition with other modes of human bonding. 37 As discussed above, this has been in the case in scholarship on Early Greek society and state formation. There are, however, serious problems in such analysis. 38 We need not assume that humans are incapable of belonging or being loyal to multiple social groups or networks at one time nor that the interests of those various social groups or networks are so antithetical that they cannot co-exist or both retain social power.39 It is therefore necessary to distance the study of kinship from ethnographic evolutionary models in which kinship is seen to characterize or constitute a whole society at a particular stage in its development.40 The Aristotelian Model Aristotle occupies an interesting place for historians as both colleague in political theory and ancient source. In his work we have the beginnings of the tradition of political 36 A criticism, for example, lies in the fact that, although descent groups are less common in what anthropologists have classified as simple and complex societies, they are not absent from them, as Pasternak et al. point out (1997, 264). 37 E.g., R. Parkin writes “human society has a long history of relationships formed in other ways than kinship, namely through informal associations or networks and formal contractual obligations. These, of course, are alternatives to kinship, and, while they may push back its boundaries, they have not so far been able to eradicate it entirely” (1997, 128). 38 For a critique of the idea of kinship-based societies as part of a wider critique of traditional anthropological kinship studies, see Schneider 1984, 57-65, 181-85. See also Fortes 1978, 14-16. 39 On social identity and social groups, see Hall 2007b, 338. 40 For a useful overview of the relationship between the concepts of evolution and kinship in anthropology, see Jamard 2006, 45-58. 14 theorizing of which we are inescapably a part. At the same time his work is that of the ancient world and is an historical source. Either way his work is as subject to the full array of criticism as any other ancient work or scholarly colleague. Thus we should not accept his theory or his model of the family as a formative part of the state on a temporal or cultural basis, i.e., that he shares a relatively similar time and culture with the society in question, as some have done. 41 Unless we can show both its applicability to the available evidence of kinship and its ability to address the questions we have of complex reality, we should, as J. K. Davies suggests, disengage ourselves from Aristotle’s model and its influence not only for state formation but also for Early Greek kinship.42 ‘Traditional’ Kinship Theory Although Humphreys and Cox have written on kinship structures in ancient Greece using anthropological theory, their methodology is that of more ‘traditional’ kinship studies, i.e., they follow a theoretical approach prevalent in anthropology before the ‘demise’ of kinship and kinship studies in the early 1980s. 43 This approach pre-supposes a cross- 41 E.g., Lacey 1968; Frost 1994; Sissa 1996; Patterson 1998. 42 J. K. Davies 1997, 26-27. 43 Humphreys 1977, 1983, 1986; Cox 1998. Humphreys’ work comes around the beginning of the ‘demise’ of kinship and does reflect the general move at that time toward the relativism that brought it on; however, it comes too early to have the benefit of the subsequent twenty-five years of theoretical debate and renewal in the field. Thus, although her work marks an important development in the anthropologically informed study of ancient Greece, the study of kinship needs updating. Cox was published in 1998 and, although she advocates for and uses anthropological methods (1998, xiv), does not address the methodological problems and theoretical debates that resulted from Schneider’s challenge nor subsequent approaches in the field. To illustrate, her bibliography lists few of the pre-eminent theorists or scholars in anthropological kinship studies following the ‘demise’ and none who have worked on or addressed that pivotal and crucial problem directly. Notable omissions pre-1998 include, e.g., Schneider himself, Collier, Yanagisako, Carsten, C. C. Harris, and R. Parkin. Cox’s work, however, is interesting and admirable for its adoption of Tilly’s goals of social history, namely the rather standard goal of reconstituting the experience of particular groups and ordinary people, followed up by 15 culturally applicable definition of kinship based on marriage and procreation, and thus on the supposed universality of biology in determining kinship. Such studies employ the terminology of ‘traditional’ kinship theory (e.g., consanguines, affines, patrilines, agnatic, etc...) as cross-cultural points of analysis and retain older kinship studies’ emphasis on marriage patterns and genealogy and descent mapped out in elaborate diagrams. Approximately thirty years ago this approach received significant criticism destructive enough to signal a ‘demise’ of the discipline, the most damning of which came from Schneider, who declared that there is no kinship.44 At greatest issue was the assumption of a universally applicable definition of kinship and hence the use of Eurocentric or Western notions of kinship as points of cross-cultural analysis or comparison. The field has since reinvented itself through intense theoretical argument and reflection. From that discussion and the subsequent renewal and re-direction of kinship studies in anthropology, there is much to be learned by Greek historians about our own ideas of, assumptions about, and approaches to kinship. Kinship Theory An important concern of this project has been to develop a contextual approach to studying kinship in Early Greece using current kinship theory from sociology and anthropology. As just discussed, a number of important shifts have taken place in kinship studies and theory in the past twenty-five years, which should be brought to bear on the study the important and meaningful goal of connecting such reconstitutions to larger social processes and change (Tilly 1987). 44 Schneider 1972, 1984. 16 of ancient kinship. 45 Discussions and themes particularly important to this study are: the history and development of kinship studies and its classical connections; the difficulties of defining and studying kinship across culture and time; the benefits and challenges associated with contextual approaches; and the relationship between kinship and society. Some attention must be given to the history and development of the terminology and concepts in kinship studies, in particular, to their connection to the study of the Classical world. Many of the working definitions and ideas in kinship studies today are derived from Eurocentric models, which in turn are linked to nineteenth-century views of the Classical world, ancient Rome in particular. For example, Stone gives the Roman gens as a prime example of a patrilineal descent group, referencing L. H. Morgan’s Ancient Society from 1877.46 Without turning to recent, even twentieth-century scholarship on Early Rome and the gens, she accepts, as is, L. H. Morgan’s nineteenth-century ethnographic analysis of the Roman gens as a “named, exogamous, highly corporate group with land and property rights held in common, and with religious and political significance.” 47 Stone also accepts L. H. Morgan’s evolutionary, idealized, and epitomizing scheme of Roman history, in which the gentes (as clans) lose their corporate nature as Rome becomes a state and as land and property become held individually, so that by the later Republic only a sense of tribal identity remained as a holdover of the former corporate gentes.48 The acceptance of such schemes of 45 Conversely, recent shifts and ideas in ancient history should also be brought to bear on anthropological and sociological kinship studies, especially re-examinations the classical ideas and kinship systems which, in their nineteenth-century idealized forms, are so foundational to the field of kinship studies. But that is a matter for a different study. 46 L. H. Morgan 1877, 285-308, esp. 292-93; Stone 2006, 76-78. 47 Stone 2006, 77. 48 Stone 2006, 76-77. 17 classical history, formed in the very different academic and cultural climate of nineteenthcentury ethnography, seemingly without reservation, is extremely curious in light of the criticisms made in anthropology for applying such models to other cultures. It may be that the idea exists that such nineteenth-century models, while being too Eurocentric or culturally specific to be applicable to other societies and cultures, are accurate and appropriate representations of the classical societies which were their prototypes. L. H. Morgan and colleagues such as Maine, McLennan, Bachofen, and Fustel de Coulanges, who were working to identify and characterize stages in the progressive evolution of human society, were very much interested in the classical world and its systems of kinship, especially ancient Rome and its gens, which were for them the patrilineal society and the patrilineal descent group par excellence.49 They drew their terminology from ancient Rome and the contemporary understanding of the Roman gens became particularly influential in developing evolutionist ethnographic models of the stages in the progression of human society. Although such models have long been questioned and/or rejected across the humanities and social sciences, terminology and concepts from these early works in ethnography remain (e.g., pater, mater, patriarchy, matrilineal).50 Anthropological models of tribal societies were fundamentally shaped by early ideas of the classical world. Since this study questions the theory that Early Greek society was focused on and characterized by kinship and descent groups along the order of the Roman gens or the Greek genē or genos, it 49 See Service 1986, 113-32. Some of the consequences of this for understanding ancient Greek kinship have been recognized, e.g., by Bourriot (1976, 29-198), Roussel (1976, 17-25, 169-71), Humphreys with Momigliano (Humphreys 1983, 131-43), Sallares (1991, 197-201); Patterson (1998, 1-43), and C. J. Smith (2006, 65-113, 141). 50 For an overview of the standard Latin kinship terminology used in anthropology, see, e.g., R. Parkin 1997, 14-36. 18 will be necessary to be wary of terminology and concepts influenced, at their origin, by idealized classical models and to be aware of the circular arguments such terminology and concepts can provoke. As C. J. Smith writes, “Taking a rigid definition of the gens and applying it as an archetype (explicitly or otherwise) to other societies has been the baneful characteristic of a century and more of classical scholarship.” 51 We must also consider if universally or cross-culturally applicable terminology for family, household, and kinship is possible. This is what was at the heart of the so-called ‘demise’ of kinship in anthropology in 1970s and 1980s, in which the extreme relativist view, attributed largely to Schneider, claimed that there is no ‘kinship’, i.e., no comparable, universal institution or concept in human societies that can be called ‘kinship’, and therefore there is nothing to be studied, no field of kinship studies.52 Challenging the traditional viewpoint that all human societies have kinship because they recognize and elevate bonds created through the biological universals of sex and reproduction, Schneider casts serious doubt on the universality of biology and genealogy in determining relationships and thus also on the universality of the concept of kinship itself. After Schneider, defining and studying kinship has largely been caught in a struggle between his relativism in which there is no universally present concept of ‘kinship’ and a desire or need to continue studying family and kin relations as something that, outside of theory and the academy, people recognize is there 51 C. J. Smith 2006, 141. 52 Schneider 1972, 1984. Although Schneider is the usual representative of the critique of traditional kinship theory, others scholars similarly questioned the categories used in kinship stuides as part of a larger movement in anthropological kinship studies (see Franklin and MacKinnon 2001, 2-4). On the demise, see C. C. Harris 1990, 34-35; Holy 1996, 3-8; R. Parkin 1997, 153-59; Stone 2006, 19-22. See Lamphere 2001, on what she labels the ‘transformation’ of kinship studies following the ‘demise’. 19 and is an important part of human existence. The challenge is to study kinship crossculturally, while recognizing that it cannot be cross-culturally defined. So what do we do? If we become too relativist, we risk having nothing to study. If we ignore Schneider’s critique, we risk mistakenly imposing our concept of kinship on other cultures or picking apples and oranges to compare. Some scholars have returned to a more traditional universal definition, which is connected to biology, although they recognize that kinship can also be determined by social factors and that kinship in general is a matter of purely social definition.53 Others have accepted Schneider’s challenge and through them kinship studies re-emerged from its ‘demise’ more attuned to cultural differences and transformed in focus.54 Collier and Yanagisako, for example, advocate that gender and kinship can be studied together as mutually constituted in social systems and that both are determined culturally, removing biological fact from both concepts.55 Carsten has tried to resurrect kinship studies in light of the lack of a cross-cultural definition of kinship by studying ‘relatedness’ instead and beginning with a given culture’s conceptions of ‘relatedness’. 56 This approach has problems in that it only really renames the initial difficulty and becomes overly broad in scope encompassing all human relationships.57 Thus 53 For a more traditional definition with a recognition of social factors, see, e.g., R. Parkin 1997, 3, 6, 32: “All human societies have kinship, that is, they all impose some privileged cultural order over the biological universals of sexual relations and continuous human reproduction through birth” (R. Parkin 1997, 3). 54 For movement towards a non-biological definition, see, e.g., Collier and Yanagisako 1987; Carsten 1995. See also, articles in Carsten’s edited volume Cultures of Relatedness (2000) and in Franklin and McKinnon’s edited volume Relative Values: Reconfiguring Kinship Studies (2001). 55 Collier and Yanagisako 1987; Yanagisako 1987. 56 Carsten 1995, 2000. 57 For criticisms of Carsten’s semantic switch, see Holy 1996, 168. 20 ‘relatedness’ leaves us with the same problem of whether there are certain human relationships that are distinguishable from others and able be classified apart from others and whether such relationships can be called kinship. It does, however, have the appeal of breaking free from some of the Eurocentric or Western ideas associated with kinship, if only by a semantic substitution. Another change that took place in the 1990s as a part of the renewal of kinship studies was to view kinship as a process, i.e., as created or emerging through various actions over time or lapsing with action or inaction. 58 While such an approach still does not provide a cross-cultural definition for kinship, it does present a way of thinking about kinship that is free of the constraints of biology and genealogy, but need not be divorced from them. Although no ‘solution’ has been found to the problem of how to study kinship universally, the field of kinship studies in anthropology and sociology is nevertheless thriving through methodological discussions and new areas of inquiry. 59 Since the mid-1980s, there have been calls for more contextual approaches to kinship and family. 60 It is not that cross-cultural comparisons are not useful or interesting, but rather that universal definitions and categories of kinship should not be assumed. The questioning of the universal role of biology in determining kinship reminds us that cultures may not have the same technological ability, respect, or taste for such scientific ‘facts’ and may determine 58 E.g., Cowan et al. adopt a definition for family or kin in which “people’s being family or kin to one another constitutes a special kind of personal and collective project -kinship involves a set of task as well as relationships” (1993, xi). Similarly, Stone writes, “Kinship relations in general entail the idea of rights and obligations” (2006, 5). 59 E.g., the collections of articles embracing new approachs and theory edited by Carsten (2000), Stone (2001), Franklin and McKinnon (2001). 60 E.g., Collier and Yanagisako 1987; Cowan et al. 1993, ix, xi; Stone 2006. 21 kin, kinship relationships, and descent in ways other than through blood and procreation. 61 Thus it is neither possible nor appropriate to construct definitions that are cross-culturally (or cross-temporally) applicable, even if the concepts are seemingly biologically or genealogically derived. And so terms rooted in Western or Eurocentric ideas of kinship such as patrilinear, consanguines, or cognatic descent, for example, are useful only as descriptors and not as definitions or points and tools for analysis or comparison. 62 Contextual approaches can also treat gender as culturally determined rather than universally present through biological fact.63 In this view, then, there can be no universal structuralist dichotomies and domains such as male equals public and female equals domestic. Either dichotomies exist but are not universal and their character is culturally determined or else dichotomies do not exist in reality and are only a theoretical product of structural analysis. Either position, however, suggests that automatically identifying women with the domestic sphere and domestic interests or with nature or any other supposedly universal pigeonhole should be avoided. Instead we should determine spheres and interests contextually along with our terminology and concepts. One way to do this is to focus on kinship as relationships involving actions and obligations and rights and privileges, instead of focusing on kinship as comprised of concrete separate entities in a system. 64 This means concentrating less on descent or cognatic groups 61 See the discussions in Collier and Yanagisako 1987, 27-35; Franklin and MacKinnon 2001, 10-15. 62 This was a major component of Schneider’s critique of kinship studies (1984, 196-97). See also, R. Parkin 1997, 7-8. 63 E.g., Collier and Yanagisako 1987; Yanagisako 1987; Comaroff 1987. 64 In this, I adopt more of a relational approach than a substantive approach, although it also involves a processual idea of kinship. See Parkin (1997, 138-39) on the substantive and relational debate. 22 and more on the whole class or category of relationships which may be based on indigenous ideas of kinship. Such an approach looks at the ways in which relatedness was expressed and thought what was important about kinship and its expression. This avoids the use of traditional kinship methodology and points of analysis and comparison, and attempts to understand kinship culturally and contextually. It still, however, allows for an understanding and importance of biologically and culturally determined kinship. It is an attempt to identify and study the ideas people had and expressed concerning their own relatedness. Such an approach is predicated on the idea that there is at least some loose category of human relationships to be studied across culture and time, which we might call kinship, but it need not rest on a universally applicable definition or concept of kinship. Following this kinship ideas approach, I have set out in this project to explore how Early Greeks conceived of and expressed kinship and what was important in the expression of kinship. To investigate this question, I accept kinship as a certain type or classification of relatedness or human bonds, which can contain notions of obligation, privilege, and affection often based on, but not limited to, procreation and marriage. While this is a very loose working definition of kinship, this looseness allows the components of the working definition to be filled out and characterized contextually by Early Greek concepts and terminology. Therefore, where categories or terminologies are established in order to proceed beyond the paralyzing effects of extreme relativism and move on to analysis (e.g., households, generational households, descent), I do not accept their traditional definitions from kinship studies, but have attempted to understand them contextually. 65 Given the lack of a theoretical universal definition or concept of kinship resulting from the challenge to 65 I do, however, use kinship terminology occasionally where necessary as practical descriptors. 23 traditional approaches to and definitions of kinship based on biology, the best method for this project has been to recognize the profound limitations of earlier conventions, adopt an approach sympathetic to Schneider’s critique, and consider the matter of kinship contextually through kinship ideas. Sources for Early Greek Kinship The evidence of Early Greek kinship cannot tell us much about biological relatedness. Although efforts in physical anthropology are interesting, especially DNA research, maybe even promising, there are so few sites on which this work has been done, and those sites have so few answers. 66 The evidence also cannot give us actual family tree structures for Early Greece. The closest we have to the relationships of real people are those family trees compiled by J. K. Davies and Toepffner, which are later and only Attic.67 Previous studies of Early Greek kinship, as has been mentioned, have been few. Lacey’s is highly positivist in its acceptance of Homeric and Hesiodic society at face value, picking through the poetry for oblique references to kinship and accepting these out-of-context phrases as a direct reflection of Early Greek reality. Patterson’s volume, while useful in its rejection of evolutionary typologies and models, contains only literary evidence for this early period and moves from the Homeric and Hesiodic worlds into the realities of the Archaic and Classical poleis. The realities of Early Greece cannot be read so directly out of poetic worlds. Poetic worlds and the kinship references they contain are the products of expression, and ought to be considered as such. 66 This interesting area of recent development in Classical archaeology is covered in detail below, p. 166-76. 67 Toepffner 1889; J. K. Davies 1971. 24 The evidence for family or kinship in this early period seems limited indeed when we put such provisos on the textual evidence, but that is the case for all questions involving Early Greece. I think what is required is a less positivist methodology than has been employed in the past and an acceptance of the material as it is, while maintaining an optimism with less explored evidentiary types. This is where the approach of kinship ideas proves useful: the evidence, textual and material, can tell us about expressions of kinship and the ideas informing those expressions. In this study, therefore, I have intended to make a departure from previous methods of studying kinship by looking at Early Greek kinship ideas and considering both textual and material evidence. It has also been my intention to investigate each evidentiary type in its own regard, i.e., with neither providing the interpretative framework for the other. For this reason the project follows what is perhaps a more traditional disciplinary divide between textual and archaeological approaches to ancient history before combining the evidence in a final synthesis. Part 1 considers textual evidence, looking at Early Greek genealogical material and genealogical thinking and the ideas of kinship expressed therein. Part 2 turns to Early Greek burials and domestic architecture, to examine ideas of kinship expressed in the archaeological record. The project culminates in a synthesis of kinship ideas drawn together from each evidentiary type. Although the aim of this project is ultimately a discussion of kinship ideas using text and archaeology, I felt it important, at least initially, to investigate each type of evidence in its own right, without interpreting one through the other and take on the challenges each presented without one type overshadowing or depending on the other, as can be a danger in incorporating both evidentiary types in ancient history. In this, the project corresponds to a degree with Kosso’s model of epistemic independence, in which distinct pieces of evidence 25 with different sources of transmission and with different justifications for use report on the same topic. 68 Given that Greek material culture is generally not datable without reference somewhere down the line to textual evidence linked into our absolute chronological system, complete independence is not really possible. I have, however, adopted approaches to the material culture which do not hinge on scenarios derived from textual evidence, and likewise for the textual evidence. So, although the divide in this project falls along evidentiary types, which Kosso warns is not enough to determine evidentiary independence, the transmission and justification for the use of each type of evidence are different enough to provide at least some level of independence between the two parts of this project.69 Before moving on to a brief discussion of the specific sources for Early Greek kinship used in this project, a few things should be said about its scope and scale. The term Early Greece, as indicated above, is a succinct way of expressing Dark Age, Geometric, and Archaic Greece. It indicates a lengthy span of time and serves to link rather than sever these periods whose overlapping, shared boundaries are difficult and contentious to determine. This is not to say that the periods of pre-classical Greece should be viewed together as an unchanging monolithic block or as a long period of small gradual changes, but that it is important in a project such as this, which is very much concerned with ideas, to consider as much evidence as possible and over periods of change. This project also considers Early Greek culture on a large geographic scale, including settlements across the Mediterranean and beyond, and across the whole spectrum of society, encompassing whatever conceptions of kinship arise from the evidence, such as it is, 68 Kosso 1995. 69 Kosso rightly argues that independence of evidence cuts across disciplinary lines, not along them (1995). 26 wherever it is. Therefore, I do not abide by culturally arbitrary geographical boundaries, but rather try to consider the concept of cultural identity in determining areas in which to look for Early Greek kinship ideas. Such a broad approach allows for a greater evidence base with which to investigate wide-ranging and big picture questions about Early Greek society. This is not to say that we can expect to see the same things happening across this world and across different socio-economic classes and social strata, but that we can see differences and similarities, and trends and innovations better with a broader temporal and geographic perspective. Genealogy and Kinship For textual evidence, I have chosen to examine Early Greek genealogical material, partly because a comprehensive study of the material has not been done, but mostly, as is appropriate to the approach of this study, because genealogies were a context in which kinship was actively and purposefully expressed and through them we can see something of what was important to be said about or through kinship. I take Early Greek genealogical material to be extended expressions of descent or ancestry beyond a simple statement of relatedness. This means genealogy not as genealogical charts or family trees reconstructed by prosopographers, but genealogy as Early Greeks expressed it, which is really quite different. My study of genealogy is therefore not an exercise of modern prosopography, but an examination of Early Greek story-telling and mythico-history. Part 1, Genealogy and Early Greek Kinship, begins with a chapter that considers definitions of ‘genealogy’ and examines genealogical material produced by Early Greeks, rather than tables, diagrams, or charts of ancestors and/or descendents created by modern scholars. Such material can be found in Homeric and Hesiodic poetry, archaic poetry, the 27 Early Greek mythographers, Herodotus, and inscriptions. This chapter also examines the forms that recounting ancestors or descendents took in Early Greek literature and the scope of genealogies. Chapter 3 looks at the relationship between genealogy and the origins of history writing and the place of written genealogies in Early Greek historiography. In it, I examine the reputed genealogy of Hekataios reported in Herodotus, and three sources of listlike genealogical material: the genealogies actually recounted in Herodotus (of the Spartan kings and Alexander of Macedon), the difficult and much-debated genealogy of Miltiades the Oikist (also known as the Philaid genealogy), and the genealogy of Heropythos (a lone inscription). Chapter 4, using the results from the preceding chapters, considers the categorization of Early Greek genealogies. It then investigates the purpose of expressing ancestry or descent information and ultimately the use of genealogies by various parties in the Early Greek world. Archaeology and Kinship For material evidence, I turn to Early Greek burials and domestic architecture. In this project, I take the position that the archaeological study of the household and of burial practices can help answer important questions about social organization and change and that it need not necessarily be informed by predetermined models of social organization and change derived from the literary evidence, teleological schemes, or anthropological typologies. 70 Since understanding kinship in Early Greece has been plagued by typologies, I have tried in this study to import as little background as possible about society and kinship from textual sources, and let the material evidence inform the framework as much as 70 Similarly, on the potential of studying households archaeologically, see Souvatzi 2008, 2-3. 28 possible. I have done this with an eye to combining the evidence of each type later as at least relatively independent threads in a synthesis. In approaching the material evidence, I have opted for a people-centred approach. Although I use the word reflect when considering the material culture and cultural practices, I recognize, as Hodder and Hutson have argued, that material culture is not simply a passive reflection.71 After all, what would kinship and kinship ideas be without people to be related to one another and think about how they are related? With Early Greek kinship ideas, however, we are in a realm of ideologies, representations, and conceptions that are not our own, and in investigating ideas in such a world, we have to look for their expression, whether in words or in things, which brings in the concept of reflection. It is not so much that material culture is doing the reflecting, but that the cultural practices or actions performed by agents or individuals to create material culture can reflect for us the ideas that inform them as well as contribute to their expression. But that is not the end of it; the expression of kinship ideas through actions which may manifest in the material record is also part of the creation and reaffirmation of those kinship ideas, all of which are shaped by the individual, the culture, and the past.72 Kinship ideas and acting upon them could be considered to be part of the individual, the culture, and the past that Hodder sets between behaviour and material culture. Indeed, how Early Greeks were related to their past and how that past was related to the present were major parts of their thinking about kinship and actions concerning kinship (both informing them and being informed by them) and as such, these questions have become major themes of this project. 71 Hodder and Hutson 2003, 6-10, 99-105. 72 See Hodder and Hutson 2003, 14-15, on behaviour and material culture. 29 Part 2, Archaeology and Early Greek Kinship, is concerned with the archaeology of kinship ideas. Chapter 5 considers the evidence of kinship in burials and burial practices. Possibilities for osteological research for the study of biological relatedness are reviewed in light of the nature of the science used and state of the discipline. Most of the chapter, however, considers burial practices and behaviours, and the kinship ideas that inform them and are reflected in them. Particularly important to our understanding of kinship ideas in burial practices is the grouping of burials in multiple inhumations, enclosures, plots, or clusters and whether or not grouping can be attributed to kinship and what such practices might tell us about how kinship was perceived and expressed. I also look at the kinship ideas that may be reflected in and reaffirmed through differentiation among burials based on age and gender. Chapter 6 is a study of domestic architecture as the physical space of the household and kinship. Instead of looking at typologies and classifications of houses, I consider what we may learn about kinship and the household, which I take as a kinship group, from the relationship between houses and settlements, from access patterns and room functionality, and from changes in each of these things. The concluding chapter is a synthesis, a tying together of the kinship ideas from the evidence examined in detail in the preceding chapters. It takes the findings of those chapters and brings the evidentiary types of the two parts together to explore broad and overlapping kinship ideas from the Early Greek world. 30 Part I: Genealogy and Early Greek Kinship 31 Chapter 2: Early Greek Genealogical Material Early Greek genealogical material is usually studied for its place in the development of written history, its association with oral and literate culture, or its use in the expression of ethnic identity. 73 Associated with such important themes, genealogy is often assigned an important place in ancient Greek culture; however, no systematic study exists surveying ancient Greek genealogical material and its characteristics, structures, contexts, and purposes.74 Recently, conferences on ancient genealogy have resulted in collections of short limited studies on a variety of topics involving ancient Greek genealogy. 75 While many of these studies offer interesting views and cover many aspects of Greek genealogy, as a whole they are no substitute for a systematic study of Greek genealogical material. Some important work on genealogy has been done by scholars studying the Hesiodic Catalogue of Women. Both West and Fowler take a comparative approach to genealogy in order to interpret the Catalogue, turning to examples from anthropological studies of a multiple number of societies.76 Similarly, R. Thomas also uses ideas from anthropological studies of genealogy to inform her work on Greek genealogies. Such cross-cultural analysis has provided us with some good points to consider and has helped us to recognize, label, 73 Historiography: e.g., Fornara 1983, 4-12; Jacob 1994; Möller 1996; Couloubaritsis 1998, 33-91; Bertelli 2001. Orality and literacy: R. Thomas 1989, 173-95. Ethnicity and identity: Hall 1997; Fowler 1998. 74 An extended discussion of Greek genealogy appears in R. Thomas and is important and informative (1989, 173-95); however, it is not a thorough survey of genealogical material. Couloubaritsis discusses genealogy at some length but is distinctly interested in the development of philosophy and the history of ideas and does not produce a survey of genealogical material (1998, 33-91). Hall’s important coverage of the role of genealogy in constructing ethnicity, although thorough, is very much focussed on genealogies connected with the Argolid, the subject of his case study (1997, 67-110). 75 CIERGA (Kernos 19) 2006; Auger and Saïd 1998. 76 West 1985, 11-30; Fowler 1998. 32 understand, and explain phenomena or practices we see in the Greek material, like telescoping or fluidity, segmentation, and filiation. 77 Cross-cultural examples, however, cannot fill in the gaps in our material. We cannot judge from what occurs or occurred in other societies, what we should see concerning genealogy in ancient Greek society. We cannot import a definition of genealogy, what it should and should not be, into the Greek world. Doing so could lead us to expect something quite different from what the material actually is, and in some cases lead us to make assumptions about structure, purpose, context, and material. For example, an association of genealogies with tribes, descent groups, or even aristocratic families should not be assumed for the Greek world unless it can be shown in the genealogical material itself or by the use and context of the genealogical material, regardless of the role genealogies may play in connection with such groups in societies labelled ‘tribal’ by anthropologists. Neither should list-like structures or models, perhaps familiar to Western scholars from biblical genealogies (x begat y, begat z, etc...), be held as the expected standard of genealogical material. In part 1, I am interested in the genealogical material of the Early Greeks, i.e., how Early Greeks did genealogy. The reason for this is that, ultimately, I am interested in what kinship ideas may be found in the genealogical material to add and/or to compare to kinship ideas from other sources and evidence types in the conclusions of this work. In order to do so, however, Early Greek genealogical material must be studied in its own right, without the expectations or assumptions imported from other societies, including our own. I take this approach, in line with the broader methodology of this study, in order to understand kinship as indigenously as possible, and therefore not as universally based on blood and procreation. 77 See R. Thomas 1989, 158; Fowler 1998. 33 Thus, in dealing with genealogy, I will focus on what and how the Early Greeks γενεαλογοῦσιν, i.e., do genealogy. In this respect, most extended works that focus on ancient Greek genealogy are methodologically not about Greek genealogies at all. For example, Toepffer’s Attische Genealogie, Broadbent’s Studies in Greek Genealogy, J. K. Davies’ Athenian Propertied Families are studies of modern genealogical methods applied to Greek material and not studies of actual Greek genealogies.78 If we use Broadbent’s methodology for illustration, we can see that there are two fundamental problems concerning context and the cross-cultural study of kinship in such an approach. The first is that the author compiles the genealogies herself from Greek materials with modern techniques. For example, Broadbent compiles the “genealogy of the local Epidaurian gentry” from forty separate inscriptions, rather than studying the inscriptions themselves as genealogies or representative of genealogical practice. 79 Broadbent’s genealogies, therefore, are made through modern study and compilation, which produce elaborate family tree diagrams, results which we will see are quite foreign to the ancient Greek world. To claim that such diagrams are representative of Greek ideas of family, chronology, and history is to be led astray by over-confidence in the universality of biological and affinal kinship and modern methods of illustrating the world. For example, in her section on Hellanikos on the Queens of Troy, Broadbent compiles the genealogy of the Queens of Troy by piecing together what she believes to be Hellanikos’ version of the genealogy. This is to look for our version of a genealogy in the fragments, 78 Toepffer 1889; Broadbent 1968; J. K. Davies 1971. 79 Broadbent 1968, 18-23. While such a practice may be useful or necessary for prosopographical studies (e.g., Toepffer 1889; J. K. Davies 1971), it is not as useful for understanding how Early Greeks thought about kinship and did genealogy. 34 rather than to read Greek genealogy out of the fragments. The problem is that Hellanikos’ version would never have looked like a family tree diagram nor do the fragments of Hellanikos enable us to say that he thought of and presented this lineage altogether as a whole in a focused genealogy. We cannot reconstruct Hellanikos’ work on the basis of what we want or think a genealogy ought to look like. Broadbent says that her genealogies do not correspond with modern genealogies, but she does so only in regard to their level of completion.80 She attributes the lack of completion to the limited survival of the evidence, and yet, that she expects that there was more evidence reveals an assumption that genealogy ought to include everything we want it to or need it to in order to make a complete diagram. 81 Moreover, the very fact that Broadbent’s genealogies can be so readily compared to modern genealogical diagrams as to reveal missing elements, shows the modernity of their composition. The much-studied genealogy of Miltiades the Oikist (usually called the Philaid genealogy) in Pherekydes (FGrHist 3 F2) is a prime example of the folly of applying modern genealogical thinking to ancient evidence and a good illustration of how reconstructed modern genealogies do not give us an accurate picture of what ancient Greeks thought about genealogy and the expression of kinship connections. 82 The genealogy given by Pherekydes 80 Broadbent 1968, 21. 81 I cannot help but wonder what an ancient Greek might have felt was missing in modern genealogies, maybe background information, story-telling, an oral quality, pizzazz. 82 The association of genealogies with aristocratic families and lengthy inherited genealogical traditions, has led to this genealogy’s commonly being called the Philaid genealogy, named for the family, which allegedly would have preserved their early history and taken their name from their earliest ancestor. The name Philaidai, however, is not attested in the ancient evidence concerning this genealogy or the family of Miltiades the Oikist, but is merely what modern scholars have dubbed it (see Parker 1996, 316-17 on the nature and historicity of the Philaidai). Despite the change in scholarly opinion about the context of this genealogy, to be discussed below, the name has stuck. A descriptive name, even just as shorthand, seems to make the most sense, and I have thus 35 is vastly different from the genealogical information about the family of Miltiades the Oikist or Elder, Miltiades the Younger, and Kimon provided by Herodotus in his Histories. The discrepancy has caused many emendations to the very corrupt text of Pherekydes by editors attempting to bring the genealogy in line with the information in Herodotus 6.34-35. For example, Jacoby adds the name of the tyrant Kypselos in as the son of Hippokleides and the father of Miltiades the Oikist.83 Wade-Gery likewise adds Kypselos, saying that his name “must have dropped out”, but sets him up as a brother of Teisander.84 Wade-Gery also drops the first Miltiades (the one before the Oikist) as a meaningless duplication. Scholarship has now turned away from the desire to emend the text to correspond with Herodotus’ information toward explorations and explanations of the differences with an eye to ancient propaganda, the development of literary techniques, the existence of multiple traditions, or the pursuit of social recognition.85 The original academic impulse, however, inspired by more modern Western genealogical thinking, was to reconstruct a precise and accurate family tree that was not initially present in the ancient material. The second fundamental problem with Broadbent’s methodology is that she divides genealogies into four types based upon differences in their material. These types are genealogies that are historical, forged or made-up, mythic or fictitious, and historiographic chosen to label it the genealogy of Miltiades the Oikist, who comes at the end of the quotation of Pherekydes. It is not that the name Philaidai is inaccurate to describe the descendents of Philaios, it is just that in this study, I want to avoid any connotations of lengthy inherited family traditions and historical kinship groups associated with that name. 83 See Jacoby’s edition FGrHist 3 F2 and his commentary, FGrHist commentary 1a, p. 388. 84 For the details of Wade-Gery’s emendations see Wade-Gery 1952, 88-89, 93-94. 85 E.g., Viviers 1987; R. Thomas 1989; Möller 1996; Duplouy 2006. 36 sociological. 86 To draw such divisions in the material is to assume that there was a difference between history and myth and reason and aetiology in the ancient world. It is also to assume that kinship is either true or false and that those creating genealogies recognized their work as one or the other. There was such a strong connection drawn between distant heroic ancestors and contemporary historical individuals in the ancient Greek world, as R. Thomas has shown, so as to make distinctions between history and myth moot.87 Genealogy is an expression of kinship and of the past, whether that past is understood mythologically or historically. So the question of type belongs more to the way the past is understood, i.e., whether through mythology or history. Such categories, however, seem more fitting to the discussions of later Greek historians and modern scholars than to Early Greek poetry and even prose. The problem of when history was born or developed in the Greek world, however, is a tricky one here, especially since the scope of this project includes key writers in the development of historiography such as Hekataios, Hellanikos, and Herodotus. However, it is fair to say that the understanding of the past in each of these writers, although they may have worked with literate methods of historiography, contains elements which are sufficiently mythological so as to distance them from a modern understanding of the past. That Hekataios writes, in a much-quoted phrase, οἱ γὰρ Ἑλλήνων λόγοι πολλοί τε καὶ γελοῖοι (the stories of the Greeks are many and ridiculous) (FGrHist 1 F1), does not make him our colleague. While this statement may represent a profound development in critical historiography (being selective - maybe even snobbish? - with one’s sources), Hekataios does not follow through with a rejection or a 86 Broadbent 1968, 18. 87 R. Thomas 1989, 157-58. 37 rationalization of myth. He only claims to write those things which seem to him to be true: τάδε γράφω, ὥς μοι δοκεῖ ἀληθέα (FGrHist 1 F1). This does not imply the wholesale rejection of myth and myth-telling but rather a selection of material, a rejection of certain myths or parts of myths. And as we can see from the surviving fragments of his works, Hekataios’ selection process was certainly not based upon a truth that is opposed to myth. The fragments reveal many mythical and fantastical stories. The story of the origins of wine, for example, from a vine root birthed by a dog (FGrHist 1 F15) should tell us that the author who selected this story has a very different idea of what is ridiculous or not than a modern historian.88 Furthermore, Hekataios’ effort to represent what he understands to be true is probably something to which epic poets and myth-tellers in general could and would also lay claim, the difference lies in how that truth is determined and not in truth-telling itself. 89 He did not have a monopoly on truth-telling. After all to tell those things that are true - ἀληθέα - is to tell those things that are ἀ-ληθέα: do not escape notice, are not obscured, are not unknown or forgotten.90 Such things are definitely in the realm of Greek epic and praise poetry. The difference for poets, myth-tellers, and Hekataios is not between true and mythical, but rather between true and false. While for a modern historian mythical and false may amount to very much the same thing, for Hekataios and the makers of Early Greek genealogies they did not. 88 See Dowden 1992, 42-44 for more examples of ‘ridiculous’ myths in Hekataios’ works. 89 On Hekataios and truth or rationalizations, see Fowler 1996, 71-72; 2001, 101; Marincola 2001, 15-16. 90 The first meaning of ἀληθής given by the LSJ is “unconcealed” and from there we get “true” and “real” as opposed to “false” and “apparent”. It is derived from the alpha privative plus λήθω = λανθάνω meaning “to escape notice, be obscured, be unknown or be forgotten” (LSJ, s.v. “ἀληθής”; Chantraine, s.v. “λανθάνω”). 38 It is important to understand that mythological genealogies (i.e., genealogies based upon a mythological understanding of the past) are not necessarily made up or fictional. Such genealogies are what Malkin labels historicizing myths, i.e., myths that narrate the past and are told as history. 91 They are also examples of what Gehrke similarly calls ‘intentional history”, the amalgamation of myth and history that comprised the social knowledge of the past.92 While genealogies may be based on convenient and deliberate invention or adaptation, those inventions or adaptations reflect, inform, and are informed by beliefs and traditions, especially those about how the world came to be as it is. As Luraghi argues, for mythico-historical traditions to be accepted, they could not be arbitrary, but rather had to be both functional as well as plausible.93 The idea of plausibility that comes alongside the intentional aspect of these traditions is not plausibility that differentiates between what seems likely or not given reality (as opposed to fantasy), but what seems likely or not given the present state of affairs. It is a plausibility that is rooted in the way things are; the mythical past must make sense in and of the present. It is whether or not such beliefs or traditions are similar or dissimilar to our own that determines for us (not for Early Greeks) their character as mythological or historical in our eyes, and not whether or not they are fictional. Thus mythological genealogies may not be defined as those that are made up or fictional, but rather as those based on a mythological understanding of the past, which may represent actual belief and tradition as well as deliberate adaptation or invention. 91 On historicizing myths and the transformation of myth into history and history into myth, see Malkin 1994, 3- 6. 92 On “intentional history” as social knowledge and the amalgamation of myth and history, see Gehrke 2001, passim, esp. 297-8. On mythico-historical tradition with reference to the Atthis, see Harding 2007, 183; 2008, 3. 93 Luraghi 2008, 46-48. 39 In this work, I will refrain from dividing genealogies along the lines of mythological and historical, accepting that understanding the past in the Early Greek world involved no such firm distinction. But while I will not categorize ancient genealogies in this way, the historical/mythical distinction is nonetheless important in analysis, particularly in regard to the material and where context and use are at issue. If we are to understand the genealogies as ‘intentional history’ or ‘mythical history’, we need to know precisely who is involved and connected with whom. Therefore, I will use the terms ‘historical’ and ‘mythical’ to denote things or people that we consider to be historical or mythical, so as to not neglect or negate the ancient mythico-historical tradition. If genealogies should not be divided along mythical and historical lines, may genealogies be divided in other ways, by different criteria? Hall, in his work Ethnic Identity in Greek Antiquity, sees a marked difference between family genealogies and ethnic genealogies. “Family genealogies” were those which allowed “individuals to trace their lineage back to three-dimensional characters” (perhaps for the purpose of competition and co-operation among elite families); whereas “ethnic genealogies” were “the instrument by which whole social collectivities could situate themselves in space and time, reaffirming their identity through appeals to eponymous ancestors.”94 Hall sees a difference in terms of material (known, established ancestors or largely unknown, eponymous ancestors) and purpose (for familial interests or ethnic interests). Such a division involving material and purpose, however, is something that must result from a thorough study of the material as a 94 Hall 1997, 41. Hall does not appear to agree wholly with the premise, which he says is “normally assumed”, that family genealogies developed originally among elite families in an arena of competition and co-operation, only that he sees a difference in material and purpose between such genealogies, and those belonging to ethnicities. Therefore, I represent the statement about elite families in parentheses and qualified by ‘perhaps’. 40 whole prior to categorization. Just what information is contained in the genealogies? To whom and for whom were they being performed, written, read, or otherwise disseminated? For the initial purposes of methodology, I would prefer not to make pre-emptive categories of genealogies, but rather, to do so in concluding, to see if differences in information and structure present themselves in combination with audience and purpose to create typological divisions in the genealogical material. This chapter, therefore, looks at all genealogical material regardless of its structure, audience, purpose, and material. The connecting thread throughout the material will be that it is genealogical. But what does that entail? Our first extant evidence of the combining of the Greek words γενεά (origin, birth, stock) and λόγος (tale, account) is the word γενεαλογέω, the ancient Greek verb that describes the action of doing genealogy. It first appears in our extant literature in Herodotus, where its contexts may suggest a complete reckoning of one’s ancestry in a linear fashion, counting each generation. 95 Thus, the LSJ defines it as “to trace a pedigree.”96 It is later used to describe the work of Hekataios, Akousilaos, and Pherekydes, writers largely contemporary to Herodotus, along with the words γενεαλογία (genealogy) and γενεαλογός (genealogist).97 Although it remains unknown just how early these words were applied to these authors and works, it is not, however, unreasonable, given the use of the related verb in Herodotus, to suggest that the writers of the age were aware of them. 95 The verb appears seven times (Hdt. 2.91.22; 2.143.1, 4, and 11; 2.146.15; 3.75.6; 6.54.1) plus two more times with the prefix ἀντε- (Hdt 2.143.12, 14). 96 LSJ, s.v. “γενεαλογέω.” 97 See the discussion on the Early Greek prose genealogists below for examples, p. 54. 41 Although our first extant evidence for the word for doing genealogy thus appears in the fifth century BCE, Greek interest in recounting ancestry and descent extends for centuries in either direction, taking on various forms and characters, and thus making our definition based on Herodotus alone inadequate to encompass the full act of genealogy-making in the Greek world. The first problem with the LSJ’s definition is the word ‘trace’. ‘Trace’ implies a studied and complete counting of each generation from descendant A to ancestor B, as we see in Herodotus. But ‘trace’ is inadequate to describe the reckoning of ancestry that occurs in all of the sources.98 Our evidence simply does not support the notion of completeness, as if each stage were worked through and plotted. As R. Thomas has pointed out, we only have three extant full genealogies involving historical families (those whom we believe not to be mythical), that is, where the line is complete from contemporary subject to founding ancestor.99 Other extant genealogies involving people we know to be historical (i.e., not mythological) are subject to what R. Thomas identifies as telescoping, wherein there is a substantial gap in the genealogy between the generations of the recent past and those of the distant past. The middle is left out. R. Thomas concludes that the interest in genealogy was not so much in linking all the descendents or ancestors in a long lineage, but to connect recent generations, that is, those close in time to the creation of the genealogy, to the distant heroic past. The second problem with the LSJ’s definition is the word ‘pedigree’. ‘Pedigree’ suggests a particular lineage of an individual or family, which again is not always the case. 98 This is not to say that the word “trace” cannot be used to describe the movement in some genealogies (and it will be used in this study), but rather that “to do genealogy” is not necessarily “to trace”. 99 R. Thomas 1989, 159. 42 Hall, for example, studies “ethnic genealogies” which involve whole ethnic groups and ancestors who are eponymous to locations in various regions. 100 Many genealogies seem to be concerned not with individuals but with wider ethnic groups. Some genealogies also involve an ancestor and then branch out from there, not aiming at any particular individual or family. 101 Moreover, the word ‘pedigree’ is a loaded term in modern English, antiquated, and even inappropriate to use in reference to human beings. It seems more appropriate to the class-conscious, elite Victorian milieu in which the lexicon was compiled. 102 In the use of the word pedigree, the lexicon reveals it age. It is preferable to approach the act of making genealogies more neutrally as “the recounting of ancestry and/or descent” -not necessarily complete nor linear nor focused on an individual lineage. This does not mean that genealogical material only contains ancestry and/or descent statements or information, but that ancestry and/or descent figure largely in the information being conveyed. 103 Such recounting may take on several different forms from rather bare lists of names and relationships to elaborate stories. Genealogy is always, however, more than a statement of relatedness, e.g., a patronymic. It is an extended narrative or presentation of relationships. It need not be long nor complete, but it must move beyond simply stating relatedness with one or more individuals. 100 Hall 1997, 40-51. 101 Various structures and directions found in the genealogical material will be discussed at length below, p. 57- 87. 102 In the second edition of the LSJ, dating to 1845, the entry for γενεαλογέω in Herodotus uses much the same language as the newest edition of the LSJ. It reads, “to trace ancestry, make a pedigree” (LSJ, 2nd ed., s.v. “γενεαλογέω”). 103 What else may be considered part of genealogical material besides statements of ancestry or descent will be discussed further on. 43 Following this definition, I have chosen to include as early genealogical material all Early Greek works or passages for which we have evidence of the recounting of ancestry and/or descent beyond a statement of relatedness and for which we have sufficient reason to place its composition around or before the middle of the fifth century BCE. The scope of this study, therefore, starts with our earliest extant Greek literature, i.e., Homer and Hesiod, and goes beyond the mid-fifth century just to include Herodotus and his contemporaries, straddling the Archaic and Classical periods. I have chosen to extend the scope of this study down into the Classical period because of the intense intellectual changes in historiography from the sixth through fifth centuries and the important, perhaps related, changes in Greek genealogy-making. The mid-fifth century is not a firm cut-off point at a specific year, especially given that Herodotus’ Histories were published after the middle of the century, but rather a loose range of years in which to draw the scope of the study to a reasonable close. Every study needs a limit to its scope; however, to be too rigid about dates here would be to lose important evidence about the development of genealogy and genealogical thinking. Any changes we may see in genealogy-making (e.g., in structure, production, purpose, and ideas) may inform us about changes in what kinship was and how it was conceived of at different times within our period of interest. Early Greek genealogical material is the product of a tradition of recounting ancestry and/or descent in different structures and styles and for various purposes. Sources of Early Greek Genealogical Material Homeric Poetry There are eight passages in Homeric poetry that can be clearly described as genealogical. These are the genealogies of Krethen and Orsilochos (Il. 5.541-49); Glaukos (Il. 6.144-211); 44 Idomeneus (Il. 13.445-54); Diomedes (Il. 14.109-27); Aineias (Il. 20.200-41); Achilleus (Il. 21.182-91); Theoklymenos (Od. 15.223-57), and Telemachos (Od. 16.112-21). Each of these passages establishes identity and character, on or off the battlefield, through the recounting of ancestors. Upon encountering a stranger, Homeric characters often ask not only for a name, but a location and parentage, e.g., τίς πόθεν εἰς ἀνδρῶν; πόθι τοι πόλις ἠδὲ τοκῆες; (What man are you? From where? Where is your city? And your parents?) (Od. 15.264).104 Or Homeric characters may respond automatically with such information, as in the exchange between Diomedes and Glaukos: Diomedes asks, τίς δὲ σύ ἐσσι, φέριστε, καταθνητῶν ἀνθρώπων; (Who among mortal men are you, dear friend?) (Il. 6. 123). And Glaukos answers, in an apparent formula (repeated by Aineias at Il. 20.213-14), εἰ δ’ ἐθέλεις καὶ ταῦτα δαήμεναι, ὄφρ’ ἐῢ εἰδῇς / ἡμετέρην γενεήν, πολλοὶ δέ μιν ἄνδρες ἴσασιν (If you wish to learn these things, so that you may know my lineage well; many men know it) (Il. 6.150-51). Glaukos then proceeds to give his full genealogy. Such is the context of Homeric genealogies: they appear, given either by the hero or the poet, in response to questions of identity or character. Compared to the other sources of genealogical material, Homeric poetry presents a rare opportunity to study complete extant genealogies in context. However, this context must be handled with care. When it comes to Homer, we should not claim that Homeric examples represent some aspect of reality, transposing the amalgamated and literary world of epic poetry onto the real Early Greek world. For example, we cannot assume that historical men had pedigrees equivalent to their Homeric counterparts, as Lacey does when, following a 104 References to and quotations from the Iliad and Odyssey are from Monro and Allen 1920 and Allen 1917 respectively. Translations are my own. 45 discussion of Homeric genealogies, he writes “The assertion of a claim to status by pedigree is the likeliest explanation of the growth of catalogue poetry (in which genealogies played a large part), as hereditary leaders of the aristocratic age in Greece sought to establish their claims to rule by prerogative of descent from the ruling gods.”105 The problem here is that Lacey takes Homeric society at face value as “the aristocratic age” of Greek history, assigning it a time period between the tenth and seventh centuries BCE, and therefore, can project Homeric genealogical practices onto historical ones, as if what happens in Homer is directly indicative of what was happening in reality. 106 We should not simply accept Homeric genealogies as directly representative of historical genealogical practices involving historical people in Early Greece, but rather as representative of genealogical thinking and an interest in descent. Homeric genealogies cannot be used as direct evidence that historical men had genealogies like those of the Homeric heroes. We can, however, look at the genealogies in Homer within their own literary context. How were they told? What was their structure and form? What did they include, stress, or leave out? We can look at their use and purpose in the world of epic poetry. We can look at the myths they tell and the ideas of kinship the express. We can study them in their own right. A more nuanced understanding of their relationship to the realities of the Early Greek world can thus be developed on the basis of those investigations concerning structure, scope, purpose, myth, and kinship ideas. This approach must also hold true for other Early Greek genealogical material, as we are dealing with literary and mythical worlds that were not a- 105 Lacey 1968, 37-38. 106 Lacey 1968, 33-50, 51. 46 historical to their audiences, but which, as we will see, were rarely connected to worlds which we would consider historical. Hesiod’s Theogony Hesiod’s Theogony is arguably one large genealogy of the gods. In the prologue, the poet calls on the Muses, κλείετε δ’ ἀθανάτων ἱερὸν γένος αἰὲν ἐόντων (Make known the holy lineage of immortals who exist forever) (Theog. 102).107 Then what follows, from line 116 on, is a recounting of the descent relationships of the gods alongside what West calls the “Succession Myth.”108 This myth is the story of the rulers of the gods, from Ouranos to Zeus, and answers Hesiod’s second request of the Muses to tell how the powers and riches of the gods are divided and how they come to take hold of Mount Olympus (Theog. 112-13). These two requests are very closely connected. The myth of succession and the recounting of the descent of the gods are closely bound together and both are part of the same story: the story of how things came to be as they are. As West’s comparative overview of world theogonic poetry shows, the genealogies of gods, heroes, and humans play a major role in peoples’ stories of the creation of the world and their mythological explanations of how things come to be as they are.109 The Greek Hesiodic Theogony shares this context of genealogy. Although West, Thalmann, and Hamilton each differ on the extent to which they 107 References to and quotations from the Theogony are from Merkelbach and West 1970. Translations are my own. 108 West 1966, 18-19. 109 See West 1966, 1-16 on the theogonic literature of several civilizations, i.e. literature treating “the origin of the world and the gods, and the events which led to the establishment of the present order” (1966, 1). From West’s overview, we can see that genealogies of gods, heroes, and humans feature prominently in the theogonic literature of many cultures, e.g., Hebrew, Persian, Indian, Germanic, Norse, Ancient English, and Japanese. 47 think the Theogony is genealogical and on genealogy’s relationship to the succession myth, each sees genealogy as a major structuring element of the poem. 110 In the discussion of the story-telling character of Early Greek genealogy below, I will argue that in both its structure and the information it presents, the Theogony is very similar to other Early Greek genealogical material and thus is well described as a genealogy. Although the end of the Theogony as we have received it is spurious, following West’s assessment of the final point of preservation of Hesiod’s work at line 900, it is nonetheless a piece of Greek genealogical material.111 If it is datable to before the mid-fifth century, as West has it, as a revised ending and connector to the Catalogue of Women, it belongs in a study of Early Greek genealogy. 112 So this study will include the end of the Theogony as genealogical material in its own right. The Hesiodic Catalogues The fragmentary Catalogue of Women is also a work of genealogical poetry and its status as such is not usually disputed, as is that of the Theogony. Also called the Ehoiai or 110 West 1966, 31-39; Thalmann 1984, 40; Hamilton 1989, 15. Hamilton, although he does not accept that the whole poem is genealogical, does recognize that genealogy is a large part of the ‘program’ of the Theogony as outlined in the proem. 111 West 1966, 398. There is disagreement on this end point (see Hamilton 1989, 96-99 for an assessment of the major arguments); however, given that our earliest suspicion about the text occurs beginning at line 901, it may be safest to consider anything after line 900 to be in doubt. In any case, this project’s focus on ideas and its inclusive methodology, looking at Early Greek genealogical material as a whole, allows for the study of multiple works despite unclear authorship. Therefore, we may include the information and be careful about authorship when discussing specific points of structure. 112 West gives it and the Catalogue of Women a probable date in the sixth century on the basis of the editorial activity of the time seen in other examples of poetry composed as continuations of pre-existing poems (1966, 49). 48 shortened to the Catalogue, it was attributed falsely in the ancient world to Hesiod and, although it acts as a continuation of the Theogony, it belongs to a later period, probably the sixth century.113 Another catalogue called the Megalai Ehoiai or Great Ehoiai was also attributed to Hesiod; however, it survives in an even more fragmentary state than the Catalogue of Women.114 The fragmentary state of the Hesiodic catalogues, even that of the Catalogue of Women which is better preserved than its cousin through quotation and on papyrus, renders it difficult to be confident in their structure. Matters of structure and the order of the fragments will have to be treated with caution and careful attention will have to paid to whether evidence is found in quotation, paraphrase, or on papyri fragments and how soundly it is attributed to the work in question. We can tell, however, from the content of the Hesiodic catalogues, that, despite their fragmentary state, they are works interested in descent and ancestry relationships involving gods and heroes, and thus are genealogical. Other Early Greek Poetry We have evidence that other poets in Early Greece wrote works of genealogy or at least works with passages of genealogical material. Eumelos of Corinth, paraphrased largely in Pausanias and the scholia on Apollonius of Rhodes, is attributed with genealogical poetry 113 For the Catalogue’s date and faulty attribution to Hesiod see West 1985, 131-136, 125-130. The name Ehoiai comes from the structuring of the poems around the recurring connecting formula: ἠ’ οἵη... like the one who... Fragments associated with Hesiod (those of the Catalogue of Women and the Great Ehoiai) are quoted and referenced from Most’s collection of Hesiodic fragments (2007) and are cited in the following manner: Hes. frag. 1 (Most). Translations are my own. 114 The Great Ehoiai is preserved largely through quotations in Pausanias and the scholia on Pindar and Apollonius of Rhodes (see Hes. frags. 185-201 (Most)). Hesiod is also associated with the Melampodia, a poem on seers, which may or may not have been genealogical in character and structure (see Huxley 1969, 54-9; West 1985, 3-4). Given the relatively few fragments (nine altogether see Hes. frags. 206-15 [Most]) and its unknown structure and genre, it cannot be considered among the genealogical material in this study. 49 which helped to situate Corinth and the Corinthians in the world of epic poetry and mythology.115 Although a few remaining lines of poetry survive, most of the material is found in prose in the scholia or Pausanias and this material is given the name Korinthiaka. Huxley dates the poetry of Eumelos to the late eighth century and treats the Korinthiaka as a version of the Eumelos’ real poetry rendered into prose at a much later date, possibly the first part of the fourth century BCE.116 Thus he is unaffected in his analysis by the prose material’s later date, treating the material as the body of work of one man in the eighth century. While this is an approach to a troublesome collection of fragments, paraphrases, and testimonia attributed to Eumelos that allows us to move forward, it is far from resting on solid ground. Therefore, for the purposes of this study, I will accept that there probably was a poet called Eumelos working in the late eighth century and that he may have written genealogical poetry or at least passages of genealogical material; however, I will refrain from placing too much emphasis or importance on the work on Eumelos alone in argumentation. Kinaithon of Lakedaimon and Asios of Samos were also credited in the ancient world with writing works of genealogical poetry.117 Pausanias, for example, cites the genealogies of both of these poets, along with the Catalogue of Women (the Ehoiai) and the Naupaktia, as 115 See Huxley 1969, 60-79. The collected testimonia and fragments of Eumelos are in Jacoby FGrHist 451; M. Davies EGF, 95-103; Bernabé PEG, 106-114; Fowler EGM, 105-109. In this study the testimonia and fragments of Eumelos are quoted and referenced from M. Davies’ collection and are cited in the following manner: EGF Eumelos T1. The translations are my own. 116 Huxley 1969, 62-63. Huxley arrives at the date of Eumelos’ poetry from a quotation of it in Pausanias (4.33.2), which he relates to the time of the first war between the Spartans and Messenians ca. 730 BCE. Bernabé also accepts this date (PEG, 108). 117 The collected testimonia and fragments of Kinaithon are in M. Davies EGF, 92-93 and Bernabé PEG, 115117. Those of Asios are in M. Davies EGF, 88-91 and Bernabé PEG, 127-131. In this study the testimonia and fragments of Kinaithon and Asios are quoted and referenced from M. Davies and cited in the following manner: EGF Kinaithon F1 and EGF Asios F1. The translations are my own. 50 sources for his own genealogies (4.2.1).118 He also cites their works individually in connection with specific pieces of genealogical data.119 Modern scholars tentatively assign Kinaithon and Asios dates in the seventh to sixth centuries and the sixth century respectively. 120 While we have very few fragments of both of these authors, even fewer of which appear to be direct quotations, they do provide some evidence and examples of genealogy making in verse outside of the Theogony and Catalogue of Women. As fragmentary material, the works of Kinaithon and Asios will be treated in the same manner as the those of the early mythographers to be discussed below. The genealogical information about the Spartan royal lines embedded in the wider narrative of Pausanias (3.1.1-10.5) may also have its source in the genealogical poetry or at least tradition of the sixth century. Pausanias begins with the earliest eponymous royals followed by the return of the Herakleidai and continues recounting the sons and successions of the two Herakleidai royal lines after the twin sons of Aristodemos. Pausanias himself does not mention a particular poet or source for the Spartan royal genealogy, but rather attributes the genealogy to Lakedaimonian tradition: ὡς δὲ αὐτοὶ Λακεδαιμόνιοι λέγουσι... (As the Lakedaimonians themselves say...) (Paus. 3.1.1).121 As Hall points out, the late date of Pausanias’ writing need not mean that the genealogical myth which he represents is likewise late.122 If we can accept the existence of the poets Eumelos, Kinaithon, and Asios, to whom 118 Paus. 4.2.1 = EGF Kinaithon F5 = EGF Asios F12. 119 See EGF Kinaithon F2 and 4; EGF Asios F1, 3-8, and 11. 120 See Huxley 1969, 85-98. 121 References to and quotations from Pausanias are from Jones and Ormerod 1918-35. Translations are my own. 122 Hall 1997, 79. 51 Pausanias attributes his genealogical information for other cities or ethnic groups, can we not accept the existence of an earlier Lakedaimonian tradition? The problem is in dating it. Calame dates it internally from the mythological information, connecting the genealogy’s representation of space with the “spatial situation” brought about by Sparta’s expansion and consolidation of power in the Peloponnese in the late sixth to early fifth century.123 Even if the genealogy of the early Spartan royal line appeared, became codified, or was solidified later than this date, if we can judge by Calame’s observations about its spatial representation of Spartan power, it would probably still fall into the period of interest here, i.e., probably earlier than the mid-fifth century while Sparta was still at its most influential and powerful in the Peloponnese. Furthermore, the comparanda of Pausanias’ other similarly paraphrased genealogies, those of Kinaithon and Asios in particular, may also suggest an early date for the genealogy. However, the dating of the ‘Lakedaimonian tradition’ is far from certain, the word ‘tradition’ meaning it could be difficult to nail all the pieces down to one particular time or origin. Therefore, because its dating is not entirely secure, I will apply the same approach to the Spartan genealogy in Pausanias as to the work of Eumelos and refrain from placing too much emphasis or importance on the genealogy alone. As with Eumelos, however, it may be added to the weight of other evidence with a note of caution. The poetry of Pindar also contains some genealogical information, but whether such information can properly be called genealogical material is tricky. 124 There are statements of relatedness between contemporary figures and those of a distant, legendary past but no 123 Calame 1988, 176-78. 124 References to and quotations from Pindar are from Race 1997, vols. 1-2. Translations are my own. For discussion of genealogical themes in Pindar, see Suárez de la Torre 2006. 52 intervening links supplied and no extended descent or ancestry details. 125 It could, however, be argued that the information supplied about the distant, legendary figures is told as a genealogical story, recounting something about the descent and ancestry of those legendary figures in the story, something which we will see is a major element of Early Greek genealogy alongside descent and ancestry information. Thus, there are some passages in Pindar’s poetry that should be considered at least to be related to genealogical material and part of genealogical thinking, if not genealogical material proper. So-called genealogies of ethical concepts or abstractions also appear in Archaic Greek elegiac and lyric poetry, but such ‘genealogies’ do not usually extend beyond a simple statement of relatedness. 126 However, similar applications of genealogical metaphors or kinship metaphors to ethical concepts and abstractions can be seen in the Theogony, and so accordingly elegiac and lyric poetry will come up again in connection to genealogical metaphors. Early Greek Prose Genealogists The Early Greek mythographers wrote works of mythography, among which genealogy seems to have been a major interest if not a full-fledged genre. Those early mythographers who are credited with works of genealogy and who fall into the time frame presented (i.e., whose works date to approximately the mid-fifth century and earlier) are Hekataios of Miletus (FGrHist 1), Akousilaos of Argos (FGrHist 2), Pherekydes of Athens 125 E.g., Ol. 2.35-48; Ol. 6.24-25 and 28-73; Ol. 7.20-38 and 92-94; Pyth. 4.247-62; Nem. 11.33-42; Isthm. 3.1317b. 126 See Abel 1943. 53 (FGrHist 3), Hellanikos of Lesbos (FGrHist 4), and Damastes of Sigeum (FGrHist 5).127 These men are among the earliest prose writers in ancient Greece and the first authors for whom we have evidence of works of genealogy written in prose. Ancient testimonia and introductions to the quotations or paraphrases place these writers within a tradition of genealogical writing. They do so by writing about authors making genealogies, e.g., Herodotus on Hekataios (Hdt. 2.143 = FGrHist 1 T4); by referring to works as genealogies, e.g., Akousilaos (FGrHist 2 F3): Ἀκουσίλαος ἐν τρίτῳ Γενεαλογιῶν... (in the third book of his Genealogies...); by naming works after mythical families or dynasties, e.g., Hellanikos (FGrHist 4 F5): Ἐλλάνικος δ’ ἐν Φορωνίδι... (Hellanikos in the Phoronis...); or by calling authors genealogists, e.g., Pherekydes (FGrHist 3 T7): Φερεκύδην τὸν Ἀθηναῖον, γενεαλόγων οὐδενὸς δεύτερον (Pherekydes the Athenian, second to none of the genealogists). Judging from the fragments themselves, we can see that many deal very clearly with descent and/or ancestry, but many do not. Many appear to be concerned with myths and stories. So how do we decide what is genealogical material? As we will see myths and stories were a large part of what genealogy-making in Early Greece was all about,128 and so we could go wrong by discounting the fragments that do not directly address ancestry and/or descent. That would be, in effect, to distil the information into a false purity not intended nor 127 Fragments of their works are collected in Jacoby’s Die Fragmente der griechischen Historiker (1923-58) and more recently in Fowler’s Ancient Greek Mythography (2000). In this study the testimonia and fragments of the prose genealogists are quoted from Fowler, unless absent from Fowler, in which case they are quoted from Jacoby. They are cited in the following manner: FGrHist 1 F1. Where Fowler’s numbering is different, it is indicated in accompanying parentheses as per (EGM Hekataios F1). The translations of the fragments of the Early Greek mythographers are my own. 128 The story-telling character of Early Greek genealogy will be discussed below, p. 75-87. 54 perhaps even imagined by its presenters. The methodology of this study steers the study of genealogy in the works of the mythographers towards reading genealogy out of the fragments rather than looking for and plucking out what may look like genealogy to us.129 If we do so, we would run the risk of imposing on the material a pre-determined structure, type of information, and overall look and feel. In specific matters of structure, we should be careful to focus on those fragments which appear to contain actual quotations of the mythographers rather than pre-amble or paraphrases. Damastes’ floruit (ca. 440-30) may be a little late for this study’s period of interest but I have chosen to include his work and that of Hellanikos, whose lifespan and therefore perhaps much of his career seem to encompass almost the whole fifth century. I have done so because the early prose genealogists seem to form a particular tradition that is deemed by many to hold a specific spot in the development of historiography, and it is important to consider as much of that tradition as possible without going too far beyond the scope of the study. Herodotus’ Histories There are three clearly genealogical passages in Herodotus’ Histories, which recount the genealogies of the Spartan kings Leonidas and Leotychidas (Hdt. 7.204 and 8.131) and that of Alexander of Macedon (Hdt. 8.139).130 The genealogy of Pausanias (Hdt. 9.64), the Spartan commander at Plataia, is truncated after two generations and we are told by Herodotus that the rest of his lineage is recorded in the lineage of Leonidas (Hdt. 7.204) since 129 As Broadbent does in her study of Hellanikos on the Queens of Troy (1968, 27-39). 130 References to and quotations from Herodotus are from Hude 1927. Translations are my own. 55 they are the same. This sort of cross-referencing is unusual among the genealogical material and merits consideration and so the passage will be included in this study despite its apparent brevity. Another unusual thing about the genealogies in Herodotus is that they run like bare lists of paternal relationships, something which will have to be discussed in comparison with the rest of the genealogical material and especially with the work of Herodotus’ contemporary prose genealogists. Epigraphical Sources Inscriptions from the ancient Greek world involving kinship are usually simple statements of relatedness that rarely go beyond one generation rather than extended presentations of ancestry and/or descent.131 A small group of genealogical inscriptions (those that recount ancestry and/or descent beyond two generations) do survive but they are usually much later than the period of interest in this study. The inscriptions used by Broadbent, for example, in reconstructing the genealogy of “Local Epidaurian Gentry” date from between the third century BCE and the second century CE.132 A few more genealogical inscriptions from Miletus, Crete, and Cyrene date to the Hellenistic and Imperial periods.133 There is one 131 E.g., the inscriptions on epitaphs, stelae, vases, and monuments recording immediate family members (wives, husbands, sons, daughters) studied by Pomeroy, most of which date to the Classical period and after (1997, 126-40). 132 Broadbent 1968, 18-23. 133 These are: the genealogy of Antigonos in a second century BCE inscription from Miletus (Milet no. 422 see Hermann 1987, 183-189; Chaniotis 1987); the genealogy of Eteanor in a second century BCE inscription from Crete (Inscr. Cret. III S.56 iii 8 see Chaniotis 1987, 42-43), the genealogy of Klearchos in an inscription from Cyrene (SGDI 4859; Masson 1974), the date of which is debatable. It is dated to the late third century BCE by Chaniotis following Letronne (Chaniotis 1987, 43) but also to the first or second century CE as suggested by Masson following Fraser (Masson 1974, 266n17). 56 inscription, SGDI 5656, that does involve extended descent relationships and that is thought to come from approximately the middle to late fifth century and thus lands within the scope of this study. 134 It is a grave stele from Chios listing the ancestors of one Heropythos (figs. 3.1 and 3.2).135 It is unique in its early date and, like the genealogies in Herodotus, it has a list-like structure that is unusual for Early Greek genealogical material. It may have more in common with the later epigraphic examples than with contemporary or earlier genealogical practices, something which requires further inquiry below. Structure and Scope in Early Greek Genealogical Material Early Greek genealogies take on various structures and have vastly different scopes, from highly detailed genealogies with several branches to linear genealogies with one clear descendant as subject and no branches. However, they have at least one structural feature in common: Greek genealogies do not express family trees. Although family trees are a familiar model of kinship and a useful visual aid for making sense of information, they are a model and an interpretative structure foreign to the Early Greek world. They do not tell us how Early Greeks thought of kinship and we can go wrong by distilling or distorting the information through interpretation. 134 Wade-Gery and Chaniotis date it to ca. 450 (Wade-Gery 1952, 8-9; Chaniotis 1987, 43); Jeffery dates it uncertainly to ca. 475 (Jeffery 1990, 344). 135 Wade-Gery 1952, 8-9, fig. 1. Modern commentators take Wade-Gery’s transcription as correct over that of Corritz and Bechtel (SGDI 5656), which leaves out a generation (see Chaniotis 1987, 43; R. Thomas 1989, 156n1; Jeffery 1990, 344). 57 Genealogies and Family Trees The family tree and the Greek genealogy are two very different ways of conceiving of and presenting supposedly the same information. It is not just a difference between visual and oral or textual information. The difference lies in what connections or relationships are drawn within the information presented. Greek genealogies connect individuals through kinship primarily vertically through time in a narrative focusing mainly on descent. They do not make connections both laterally and vertically to equal measure creating a web or tree, as family trees do.136 To illustrate the significance of the difference, consider the following example involving two family trees (figs. 2.1 and 2.2) and an outline of the related genealogy from Homer (fig. 2.3). Figure 2.1 is a family tree of Aineias or the Trojan royal line adapted from the appendices of an edition of Apollodorus’ Library.137 It was compiled by the translator from the information in the Library. Figure 2.2 is a family tree of Aineias based on the information from the hero’s genealogy in Book 20 of the Iliad (Il. 20.200-41). Notice that several details from the fuller family tree from the edition of Apollodorus are missing from that created from the information in Homer. Absent are the women apart from the divine Aphrodite, the other children of Priam, and several siblings, especially in the earlier generations. It is not that the poet knows less of the mythological family or details than the writer of the Library does. The poet has the rest of the information or at least knows more 136 Duplouy similarly writes “Une généalogie antique n’a rien d’un arbre généalogique” (2006, 60) [An ancient genealogy is not a family tree]. He makes the argument, however, from ancient genealogy’s lack of precision as well as its different purpose. I make mine from the profoundly different ways ancient genealogy and modern genealogy present information. 137 Hard 1997, 22. 58 details than are presented in the genealogy, as we can clearly tell from the rest of the epic. The poet of the Iliad knows, just as the writer of the Library does, that Paris and many others are among the children of Priam, but only Hektor is mentioned. The Homeric genealogy does not recount the whole or even most of the family tree, nor does it try to as we can see from the outline of the genealogy itself (fig. 2.3). The poet picks and chooses the information to be conveyed as part of his epic technique. In the genealogies, the poet expands at key places, but seldom includes siblings. His interest is generally much more linear. Before the genealogy begins there is a small amount of preamble, which is largely heroic posturing, in which Aineias acknowledges his opponent, Achilleus, and his divine parentage. Then he boasts about his own semi-divine parentage from Anchises and Aphrodite. There is a bit more posturing and then the beginning of the genealogy proper is signalled by the apparent stock phrase: εἰ δ’ ἐθέλεις καὶ ταῦτα δαήμεναι, ὄφρ’ ἐῢ εἰδῇς / ἡμετέρην γενεήν, πολλοὶ δέ μιν ἄνδρες ἴσασιν (If you wish to learn these things, so that you may know my lineage well; many men know it) (Il. 20.213-14).138 Aineias starts by going back to the beginning of his lineage, back to Zeus, and from there he moves forward through time recounting sons and sometimes giving the story of their deeds, accomplishments, or adventures, e.g., the story of Erichthonios and his horses takes up several lines of poetry. At the generation of the sons of Tros he begins to recount the descendents of two branches: that of Ilos down to Hektor and that of Assarakos down to himself. Genealogical passages in Homer follow this general outline: they start with the subject, followed sometimes with a statement about his father, then they jump back in time to 138 The exact phrase also signals the start of Glaukos’ genealogy (Il. 6.150-51). 59 the earliest ancestor (divine or human) and from there move forward in time toward the present and the subject.139 Aineias’ genealogy, like the other seven Homeric genealogies, also focuses on one line of descent, that of the subject, and does so paternally, linking generations only by the relationship from father to son. The mother’s lineage is never traced beyond the occasional reference to her father. There is very little branching off of the main line of descent in order to fully describe more familial connections. Very rarely do the genealogies branch off to include the descendents of the siblings in a given generation. Although the siblings of a generation are usually listed and their stories sometimes elaborated upon, their descendents and their stories generally are not. There are only three places in all of the generations reckoned in the Homeric genealogies where branching does occur.140 One of which we have seen in the genealogy of Aineias. The genealogy of Aineias diverges at the point where Aineias’ lineage and that of Hektor diverge, at the sons of Tros: Ilos, Assarakos, and Ganymedes (Il. 20.231-40) (see fig. 2.3). After a short story about Ganymedes’ fate as the wine-pourer of Zeus, the genealogy follows the descendents of Ilos: Laomedon and his sons including Priam. Then the genealogy follows the descendents of Assarakos: Kapys, his son Anchises, and his son Aineias. At this point Aineias, the speaker, connects himself with Hektor son of Priam: αὐτὰρ ἔµ’ Ἀγχίσης, Πρίαμος δὲ τέχ’ Ἕκτορα δῖον. / ταύτης τοι γενεῆς τε καὶ 139 This is true for all of the genealogies in Homer, except the short genealogy of Achilleus (Il. 21.182-91), which jumps in time from the subject to the god as earliest ancestor and then returns to the subject and moves back in time generation by generation from the subject to the god. This is discussed below. 140 The other instances of branching occur in the genealogies of Glaukos and Theoklymenos are discussed below. 60 αἵματος εὔχομαι εἶναι. (And Anchises bore me, and Priam bore brilliant Hektor. / I claim to be from this generation and blood) (Il. 20.240-41). Although the genealogy branches to connect contemporary figures in a statement of kinship, it only connects two figures. The genealogy does not include other contemporary relatives, like Paris and Cassandra and the other children of Priam, nor does it include the children of Priam’s siblings. 141 Aineias’ genealogy branches only as much as is needed to establish his relationship in blood with Hektor the brilliant and so enhance Aineias’ reputation.142 Therefore, the branching is limited to two narrow lineages and does not represent a broader family tree nor link contemporary kin as a group. The genealogy of Glaukos branches briefly from the main line of descent after the generation of Bellerophon’s children (Il. 6.198-99), relaying the information that Bellerophon’s daughter, Laodameia, bore the hero Sarpedon to Zeus. The genealogy then resumes the story of Bellerophon before moving on along Glaukos’ line of descent. It is very limited branching, lasting only one generation and taking up only two lines of poetry in the 141 This is unlike the genealogical material in the Theogony or the Catalogue of Women, which do include many if not all offspring in each generation. 142 It has been argued by several scholars that the prominence of Aineias in Book 20 of the Iliad, including the telling of his genealogy, may be a result of the influence of an historical family, the Aineidai of the Troad, looking to connect themselves with their eponymous (so they would have claimed) heroic ancestor (e.g., Jacoby, Wilamovitz, Malten; see P. M. Smith 1981 for an outline of the support, largely German, for this idea). P. M. Smith’s reassessment of the independent evidence into the existence of such a family leaves the theory wanting. He concludes that there is no reason to suggest from the independent evidence that the role of Aineias, his aristeia and genealogy, in Iliad 20 is related to the patronage of a family or a civic tradition (P. M. Smith 1981, 58). Lenz, furthermore, regardless of his belief or not in the actual existence of a family called the Aineidai, argues for the integrity of the genealogy and aristeia of Aineias in Iliad 20 and therefore sees no need to explain the role of Aineias through patronage. I would add the following question to the criticism of the theory: Would we even be doubting the heroics of Aineias if it were not for the special status of the hero resulting from the later Roman claim to and exploitation of Aineias? There is no need to look at this genealogy differently from the rest. 61 long and central telling of Bellerophon’s story. But it ends with a famous heroic figure, to whom Glaukos is now connected through ancestry. The third instance of branching in Homeric genealogies occurs in the genealogy of Theoklymenos. The lineage branches at Mantios and Antiphates, the sons of Melampous the founding ancestor (Od. 15.242-49). First the genealogy relates the descendents of Antiphates down two generations to Amphiaraos, whose story is elaborated upon briefly, and down one further generation to his sons. Next the genealogy picks back up with Mantios and his descendents down to Theoklymenos, the subject of the genealogy. Again the branching is very limited, stopping after only three generations. It is not clear whether the third generation is contemporary with that of Theoklymenos, but that lack of clarity, in and of itself, tells us that making contemporary connections is not what is important here. What is important is the story of Amphiaraos, a past hero, a great warrior, beloved by the gods, who is related in some way by blood to the subject of the genealogy. In all three cases of branching in Homeric genealogies, the branch ends in a famous heroic individual to whom and to whose reputation the subject would wish to be connected through blood. With very limited branching that only occurs to connect two individuals at most and no expression of maternal descent relationships, Homeric genealogies do not resemble nor represent family trees, neither in structure nor in the scope of their material. They encompass neither the entire nor even a large part of the family and the breadth of familial connections.143 Therefore they do not connect multiple individuals laterally by generation or 143 Cf. the branching structures of the Catalogue of Women and the Theogony discussed next. 62 vertically through time in kinship groups.144 The genealogies neither contain enough information nor present the information they do contain in such a fashion as to make expressing kinship groups their aim. The closest we come to expressions of lateral kinship are like that of Aineias about Hektor, which connects only two heroes for a very specific purpose and not a whole family. Branching and Grouping In other genealogical material, however, there is more significant branching and more information about siblings as well as expressions of maternal descent relationships than in Homeric genealogies, but still not a family tree nor a connecting of contemporary individuals in a group. In recounting the family of the gods, the Theogony branches significantly, presumably lists all children and siblings, and presents descent largely maternally, i.e., it mostly organizes the information by the mother and not by the father, although this is not always the case and the father is usually mentioned. This makes the genealogy what I would describe as largely maternally organized but not matriarchal nor matrilinear, as West asserts.145 Maternal organization (of the genealogical material) is a more neutral term involving only the structuring of information, whereas matriarchal and matrilinear are cultural and/or socio-political terms involving the organization of kinship and society. Such concepts require evidence beyond that of the Theogony and even poetry, requiring investigation into the wider socio-political climate of Early Greece. It is, however, 144 Donlan also has noted that there is a marked absence of kinship groups beyond the oikos in Homeric poetry (2007). He considers this absence to be somewhat problematic, however, since, as he writes, “Certainly Homer’s and Hesiod’s contemporaries belonged to phratriai and phulai” (2007, 32). 145 West 1966, 34-35, 39. 63 interesting that both the Theogony and the Catalogue of Women are structured largely maternally and the possible significance of this with respect to kinship and society will be addressed below. But for now, I would prefer to use the term maternally organized to distance the organization of material from socio-political organization. Besides the largely maternal structure, West discerns six other principles regarding the arrangement and presentation of the genealogical information in the Theogony: the order of the genealogies is basically chronological and progresses collaterally detailing each generation before moving on to the next; if a branch is close to its end, it is often traced to that end without waiting until the next generation; related sections are made adjacent where possible creating a chiasmus; other families (sets of offspring) appear in the same order as the parents were listed, with the exception of the Titans; the last god listed is sometimes the youngest; and at the end of the Theogony, there are various combinations of mortals and gods in families and descent is no longer matrilinear. Hamilton rejects these principles, citing examples where they break down. 146 It is not necessary, however, to interpret the structure of the Theogony, or other genealogical material, as following a rigid set of principles or consistent rules. Why can a poet not change his method or structure as suits his need, desire, or artistic inclination? We should probably take West’s principles not as strict and proscriptive (as if Hesiod followed a set of pre-determined rules to the letter) but as loose and generally descriptive. As West’s first principle states, the general overall progression of the genealogical information is chronological, moving forward in time, and generational, in that it focuses largely on one generation at a time. This means at every stage that there is significant 146 Hamilton 1989, 7-8. 64 reckoning of children and jumping from one branch of the family of the gods to another. This is largely done, as West observes, in the order that the siblings were first listed and by connecting the branches with a device like ἣ δέ or a goddess’ name plus δέ. The poem, therefore, does not seek to connect these individuals, gods or otherwise, in broader statements of kinship, i.e., the broader, non-immediate descent relationships are not dwelt upon, e.g., cousin-ships, uncle-ships, and even sibling-ships. What is important is where each individual comes from and the links in the chain, not the lateral relationships. There are several places in the poetry, however, where a lineage is traced beyond the current generation, seemingly deviating from the overall structure of the genealogy by generation.147 West attributes this to unmanageability, in that the poem’s genealogical information would become unwieldy if the generational structure were followed too strictly. 148 While this is true, there is more to it. It might not just become unmanageable but also unpleasant and overly mechanical. We may be better off interpreting the structure of the genealogies through the stories the poet wishes to tell. 149 These relatively small sections that 147 For West’s description of the examples, see West 1966, 38. The branch from Medusa to Geryon (Theog. 278-94) involves the stories of Pegasus, Chrysaor and his son Geryon and how he was killed by Herakles and the context of his death in Herakles’ labours. The branch with Echidna and her children and grandchildren (Theog. 295-336) runs as the story of the births and deaths of Echidna’s monstrous children and their children. The branch reckoning the descendents of Krieos and Eurybie (Hes. Theog. 375-403) culminates with the duties, loyalty to Zeus, and subsequent honours of Kratos and Bios and their mother (not a member of the direct lineage) Styx. Finally the branch from Asteria to her daughter Hekate (Theog. 409-452) culminates in a relatively long exposition on Hekate’s duties and honours. The one branch that does not also tell a story is that of the children of Night and Erebus (incidentally also her father) (Theog. 124-25). 148 West 1966, 38. 149 For a similar approach to the relationship between stories and organization in which the stories come before chronological scheme, see Harding 2008, 3-4 on the Atthis. The key to interpretation seems to lie not so much in the organization of the matieral by chronological scheme, but in the material itself, the stories the Atthidographers intended to tell gathered from traditional tales, communal memory, physical remains, and documentary evidence. 65 follow branches down outside of the overall generational structure may have more to do with story-telling than pure mechanical necessity. Four out of the five examples of such branches that West gives either culminate in or involve extended stories about members of the lineage and the one that does not involve an extended story branches so unobtrusively that all but the keenest listener or reader would allow it to pass by without noticing the supposed deviation from the overall generational structure of the poem. Most listeners too, one would assume, would also let it go by without comment or objection. That such stories occur seemingly not in line with the overall progression of the genealogy should not be terribly troubling. These offshoots are Hesiod telling a story. They are part of the narrative structure and character of Early Greek genealogies. They are after all narratives and are neither so formulaic, nor mechanical, nor dogmatic in structure that they could not adapt to suit information, stories, purpose, and even cultural aesthetics. The structure of the Catalogue of Women is similar in some ways to that of the Theogony. It also seems to have been maternally organized and attempts to recount all family members in the genealogy. However, while the Theogony recounts one whole related family (that of the gods) with two common ancestors (Gaia and Ouranos) largely generation by generation, the Catalogue of Women recounts several different mythological families seemingly unrelated at their origins or else only loosely connected laterally and not necessarily through any expression of kinship. West identifies these ‘great genealogies’ in the surviving fragments as those of the descendents of Deukalion, Io (branching into the Belidai and Agenoridai), Pelasgos, Arkas, Atlas, and Pelops.150 There were probably others, such as the descendents of Erechtheus from the Athenian autochthony myth, but, given the 150 West 1985, 43-44. 66 fragmentary state of the material, the other families of the Catalogue are difficult to discern as clearly as those mentioned above. As West’s study of the papyri shows, from their internal structure, namely the transitions and progressions within the fragments, and from their mythological ‘spread’, i.e., their sustained interest in the same story, location, or family line, the poet of the Catalogue proceeded systematically genealogy by genealogy. Given the poem’s stated purpose in the proem to recount those mortal women who lay with immortal gods and begot the children, these largely independent genealogies may have been linked, as West suggests, through the ἠ’ οἵη... (like the one who...) formula.151 If this is the case, the great genealogies were connected through similarity and not through kinship. It is important to note, however, that the ἠ’ οἵη formula only appears among the surviving remains of the Catalogue within the genealogies themselves, introducing or re-introducing branches within the great genealogies.152 Given that we have no papyrus fragments or quotations showing the transitions between the great genealogies, this fact is not as troubling as it may seem. Moreover, the number of instances of the formula is relatively small, just twelve (three of which are just possible reconstructions of the text), we have a limited number of quotations and papyrus fragments, and capturing the formula involves a lucky convergence of a quotation or papyrus fragment and the right place in the text. All this may suggest that we have a very small sample of the actual number of instances of the ἠ’ οἵη formula. But even if the great genealogies were not connected through this device, there is still no indication 151 See West’s suggested reconstructions of the transitions between the great genealogies, in which he insists on the use of the formula for introducing each great genealogy (West 1985, 56, 76, 92-93, 94, 100-1, 104, 109). 152 Hes. frags. 19.3; 23.5; possibly 47.1; 60.7: 69.2, 94; 124; possibly 136.9; 138.8; 158; 164 (Most). And possibly frag. 94.2 (West). It also appears, just once, among the fragments of the Great Ehoiai: Hes. frag. 191a (Most). 67 that they were connected through any statement of kinship. They are quite separate entities in that regard. That being said, however, although the great genealogies of the catalogue are separate blocks, they do sometimes overlap in material mentioning the same individual in two genealogies. This is to be somewhat expected given the mythological stories that accompany these names; they are stories of intermarriage, battles, rapes, which occur between members of different families. These great genealogies are also quite large encompassing many generations and branches within them. Their internal structure seems to be largely based on the branches, following each down for several generations and then returning (jumping back many generations) to cover another branch often connecting it to the narrative with the ἠ’ οἵη device. 153 This means that even the branches of the same genealogy are often connected by the same statements of similarity that may connect the great genealogies and not by statements of kinship.154 Where the formula is not used to connect branches, other forms of transitions are used which also do not express lateral kinship connections. For example, Hes. fragment 35.16-17 (Most), a papyrus fragment, shows the transition between the descendents of Neleus and the descendents of his brother Pelias: αὕτη μὲν γενεὴ Νηλῆος [ αὐτὰρ ὅ γ’ αὐτοῦ µ[ίμνεν ἐν εὐρυχόρωι Ἰαωλκῶι σκῆπτρον ἔχων [Πελίης 153 For specific examples, see note 152 above. 154 West suggests that the origins of the ἠ’ οἵη formula lie in a tradition of simple catalogue making (perhaps from north-western Greece and the western Peloponnese) with a “radically different system of arrangement” than we see in the Theogony and that the poet of the Catalogue of Women combined that system with that of the Theogony (1985, 167). Whatever its origins, however, the fact remains that the formula is based upon connecting people through similarity in story or situation, and we may see from its use a lack of interest on the part of the poet in drawing lateral kinship connections. 68 (This is the lineage of Neleus [ But he [remained] there [in broad Iolcus Pelias,] having the sceptre [ ) The transition does not draw an explicit connection between Neleus and Pelias as brothers, although we learn that they are the sons of Tyro by Poseidon and settle in different cities from another papyrus fragment (Hes. frag. 31 (Most)), which we know from the evidence of the papyri comes before fragment 35.155 There is no attempt by the poet to connect either Neleus and Pelias nor their respective descendents in a statement of kinship. The point to be drawn from this, from the junctures of the Catalogue both between great genealogies and between branches within the genealogies, is that the genealogical information is not expressed in such a way as to emphasize lateral or web-like kinship connections between lineages, branches, and individuals. There appears to be no interest on the part of the poet in making or reinforcing kinship groups. Within the branches themselves the progress tends to be somewhat generational, listing the children of a couple (divine or otherwise) and then following the lineage down. The result is quite a complex structure, which West characterizes as a ‘middle’ course, a combination of the horizontal approach, going generation by generation, and the vertical approach, following each lineage straight down before beginning the next. 156 It is important to note that the connections between the branches are again usually made through an expression of similarity (the ἠ’ οἵη formula) and not through kinship. Nor, like the Theogony, are connections of kinship beyond descent (e.g., sibling-ship or cousin-ship) drawn between individuals to make them a group. 155 See West 1985, 37. 156 West 1985, 46. 69 The maternal organization of the material in the Catalogue is striking, although it is also present in the Theogony, as already discussed. But maternal organization does not necessarily mean that we are dealing with a matriarchal society or matrilinear descent to the exclusion or even detriment of patrilinear descent. It is merely the way the information is organized not the necessarily the society. To get from maternal organization of information to matriarchy requires more steps. That maternal organization is present, however, is important, just as the lack of expressions of kinship connections between multiple individuals is important. It is what is expressed or not that is the key. That wider kinship connections are not expressed reveals only that they appear to not be of interest to genealogy-making, whether the presence of wider kinship connections in genealogies or the lack thereof has any bearing on the existence in reality of kinship groups based upon them is a matter for further argument.157 That descent data are organized maternally shows only that maternal descent relationships are understood to carry some significance in terms of kinship and that paternal descent is not the only form of descent, as one might believe from looking at the Homeric heroic genealogies which generally exclude maternal descent information. That the Catalogue of Women is structured around women does not suggest that matriliny occurred in Early Greece. As in the Theogony, fathers are usually supplied, often in the context of the sexual act that brought about the offspring or in the context of the birth itself. 158 Males appear so often in these contexts, that it is difficult to deny that they play a very important part indeed in descent ideas. Moreover, the women presented with the ἠ’ οἵη formula, as West himself argues, appear at the ruptures between the great genealogies and 157 This notion is investigated below, p.124. 158 E.g., Hes. frags. 7, 10.6-7, 10.20-24, 10.31-34 (Most), etc… 70 also between the branches within the great genealogies. Fowler, thus, argues that the structure of the Catalogue presents women as the glue between the men, who are the building blocks.159 While it is clear that some of the women, that is those specific women of Catalogue of Women referenced by the ἠ’ οἵη formula, are the glue that binds the blocks, it is not clear that all of the women in the catalogue are glue and that the blocks are necessarily male. The ἠ’ οἵη women are the glue that binds the genealogical blocks and branches consisting of both male and female members. When it comes to other sources of Early Greek genealogical material, structure is not so readily analysed because of the fragmentary state of much of the material. We know that there were other works of genealogical poetry (e.g., those of Eumelos, Kinaithon, and Asios and the almost entirely lost Great Ehoiai), but because there is so little of them extant, we can know very little about their internal structure. Among the works of the prose genealogists, however, there are more surviving fragments, enough to be able to make some claims about structure with varying degrees of caution. Since fragments of these prose genealogists survive mostly through quotations and paraphrases by later authors, the material has been selected and plucked out of context and it is up to modern editors to put them back into context as best they can. Luckily not all of the fragments are entirely devoid of their original context. Some fragments come with references to book numbers, and, although these cannot all be assumed to be accurate, there are enough of them to group together and from that grouping give clues about structure. By matching the material associated with these book numbers with the material of other fragments involving the same individuals, the members of the same families, or the same myths, editors like Jacoby and Fowler have been 159 Fowler 1998, 5-6. 71 able to tentatively suggest some order for the fragments. For example, by comparing the material in those fragments of Hekataios said to have come from Book 1 of his Genealogies (sometimes called Histories) (FGrHist 1 F1-F5), Fowler suggests that fragments containing information about the Deukalionidai (FGrHist 1 F13-F16) and the myth of the Argonauts (FGrHist 1 F17-F18 and EGM Hekataios F18A) also come from Book 1.160 Such schemes both support and rely upon the supposition that the genealogies were structured very much in the same way as the Catalogue of Women, by great genealogies. In constructing such schemes, fragments with no attested book numbers must first be grouped by their material thematically into either families or myths known from the mythical tradition surviving in other sources (genealogical or otherwise). This apparently circularity need not deter us, however, for a few reasons: 1) the scheme suits the evidence well in that the framework provided by those fragments with numbers, albeit loose, allows for and in some cases hints at such a scheme and the other fragments slot in well, 2) a similar structure is well attested in the Catalogue of Women and so there is at least one genealogical precedent and maybe even a tradition, 161 3) the titles given to the works of Hellanikos suggest that his genealogies were written or at least disseminated as separate works, one for each mythical family or local tradition, and so his works at least appear to have been divided along the lines of great genealogies.162 Unlike Hellanikos and Damastes, however, Hekataios, Pherekydes, and 160 EGM, pp. 128, 129. See also Fowler 2006, 33, for genealogy as structure in Hekataios’ mythographical works. 161 It has also been suggested that the Library of Greek Mythology may have been modelled upon the works of Akousilaos and Pherekydes (West 1985, 45-46), and thus it may also present some clues as to the structure of the works of the prose genealogists. 162 E.g., the Asopidai, the Phoronidai, the Deukalioneia, and the Atlantika of Hellanikos. 72 Akousilaos each appear to have written their genealogies as a large singular work, collecting and presenting the traditions of several areas of the Greek world through a number of great genealogies, much as the Catalogue of Women does. Akousilaos, in fact, is accused in later antiquity of merely putting Hesiod’s works into prose and publishing it as his own, further adding to the case that the prose genealogists structured their works in a similar fashion.163 If we can accept that the Greek prose genealogists structured their works in great genealogies, and I believe we should, two major questions remain about structure. Question one: how were those great genealogies related to one another and brought together into one larger work? The fragments of any given prose genealogist give us only a spotty picture of how the material in the different great genealogies was related.164 The fragments do not abut one another directly nor do they overlap in such a way as to reveal transitions. Moreover, given the extremely limited papyrus fragments of the works of the prose genealogists, work such as that done by West on transitions in the Catalogue of Women is impossible here.165 Thus how exactly Hekataios, Akousilaos, and Pherekydes transitioned between the great genealogies and so connected their material remains unknown. What we can tell from the 163 This charge is leveled against both Eumelos and Akousilaos by Clement of Alexandria (Strom. 6.26.8 = FGrHist 2 T5). How fair an assessment it is, however, is questionable, given both Clement’s negative attitude (vitriol?) towards plagiarism among Greek writers in general and the statement from Josephus that Akousilaos amended the works of Hesiod (Joseph. Ap. 1.16 = FGrHist 2 T6). 164 The related question of just how much the prose genealogists tried to synthesize the material of various traditions to create cohesion between the great genealogies is debatable and difficult to determine given the spottiness of the evidence. 165 The actual works of Hekataios, Akousilaos, Pherekydes, Hellanikos, and Damastes have not survived independently in papyri even in fragmentary form. The best evidence of the actual works (i.e. not paraphrases) from papyri consists of two direct quotations: FGrHist 2 F22 is a fragment from a papyrus with what appears to be a relatively long (approximately 27 lines) direct quote from Akousilaos and FGrHist 4 F19b appears to be a quotation from Hellanikos in the margin of papyrus. 73 fragments is that the great genealogies in Hekataios, Akousilaos, and Pherekydes, like those in the Catalogue of Women, do not seem to coincide with the books in a work, i.e., the great genealogies seem to have overlapped books. Question two: how were the great genealogies structured internally? The structure and approach to the material in the great genealogies in the works of the prose genealogists appear to be very similar to the structure and approach in the Catalogue of Women. We must judge the internal structure of each great genealogy in the works of Hekataios, Pherekydes, and Akousilaos and of each independent genealogy of Hellanikos and Damastes, again from the grouping of fragments by book number and related mythological and genealogical information. The informational spread of the genealogies, which includes members of different branches of the same family and their stories, suggests that there was probably significant branching within the great genealogies. For example, the fragments concerning the Agenoridai in Pherekydes (FGrHist 3 F85-F97) involve the sons of Agenor (e.g., Kilix and Phoinix) and well as Kadmos and the whole Theban saga. It is likely that the prose genealogists took up a middle approach to dealing with genealogical information, much like the poet of the Catalogue of Women, i.e., they combined a horizontal approach, going generation by generation, and a vertical approach, following each lineage or branch straight down before beginning the next. Greek genealogies are not family trees, nor is it very useful to construct family tree diagrams from genealogical materials in order to understand the kinship relationships expressed in them. That practice gives us a false reading on what kinship connections were important in genealogies and how they were expressed. Studying the genealogies as a whole package yields better results about the expression of kinship. In Homeric poetry, the genealogies are geared towards the subject, an individual descendent. They branch very 74 little. When they do so, a story or important ancestor is involved, the branch lasts for a very limited number of generations, and there is always a return to the main line of descent leading down to the subject. Other examples of Early Greek genealogical material are structured around great genealogies. These recount the ancestry/descent relationships of mythical families. They are ancestor-focused in that, unlike Homeric genealogies, they are not geared toward an individual subject, but start with a common ancestor and then branch out and down, with seemingly no one particular descendent in sight. Even though they are thus organized by family, these genealogies do not show an interest in drawing lateral or web-like connections between members to form a cohesive group. In the final chapter of part 2, I will put forward some explanations for these differences in structure and contextualize them alongside reasons for genealogy-making in the Early Greek world. Before that, however, a look into the important style and story-telling elements of both sets of genealogies is required. 166 Narrative Style and Story-Telling Early Greek genealogies are not family trees, but neither do they usually assume the form of a straightforward list of ancestors or descendents generation by generation (father to son or otherwise). For example, they do not follow the bare formulaic pattern: x, son of y, son of z; or the pattern: x, from whom y was born, from whom z was born. Instead, ancestry and descent relationships are usually associated with and given alongside myths and stories. 166 Another set of Early Greek genealogical sources that are fully extant and their structures able to be studied are the genealogies in Herodotus (Hdt. 7.204, 8.131, 8.139, and 9.64) and the genealogy of Heropythos on the grave stele from Chios (SGDI 5656). Their unusually barren, list-like structure will be studied further below, because first I must establish, in the following section, why their structure is so unusual among Early Greek genealogical material. 75 The result is genealogies that appear to be series of stories pertaining to important ancestors with sections featuring descent information in between, which can be list-like. Thus, in a given genealogy, in the course of spelling out several generations, the level of detail swells at key generations making the genealogy a collection of stories and ancestors or descendents and not simply a list of ancestors and descendents connected formulaically. We can see this combination of descent information and story-telling in the genealogy of Aineias. In figure 2.3, the outline of the genealogy of Aineias, the sections of story-telling are rendered in italics. In forty-two lines of poetry altogether, fifteen lines are preamble and positioning before the genealogy proper begins (although in them we do learn of Aineias’ parentage) and sixteen are given over to story-telling.167 This leaves just eleven lines dedicated to recounting descent relationships in Aineias’ genealogy. These eleven lines are divided among four sections, none of which looks excessively formulaic. Two (Il. 20.215 and 219) consist of one line of poetry each and give just one piece of descent information each and do not appear list-like. The other sections of descent information consist of three and six lines (Il. 20.230-32 and 236-241), but also do not appear to be excessively list-like. They lack a strict formulaic or repetitive structure and language. The father-son descent relationships are expressed by the same term - τίκτω - but with different syntax. We also see the addition of epithets and short asides, e.g., ἀµύμονα Λαομέδοντα (blameless Laomedon) (Il. 20.236) or Ἱκετάονά τ’, ὄζον Ἄρηος (Hiketaon, scion of Ares) (Il. 20.238). This results in a repetitive character with respect to some terminology and content (father begets son) but not with respect to structure and style. Aineias’ genealogy is a 167 Three lines each are given to the stories of Dardanos and Ganymedes and ten to that of Erichthonios. 76 combination of story-telling with the telling of descent relationships that have some very limited characteristics of a list. The element of story-telling in Homeric (and Hesiodic) genealogies has also been observed by R. Thomas and by Graf. R. Thomas treats the stories as something separate from or added onto the genealogies, seeing them as elaborations upon the bare-bones of genealogy. 168 This division between the recounting of ancestry and/or descent and the stories seems artificial, given that such stories, as we will see, are present in nearly all of our Early Greek genealogical material. The recounting of descent and ancestry relationships seems very rarely to come without embellishment and elaboration in the Early Greek world. Thus, to separate the two is to separate mistakenly into two practices what is only one. Storytelling is a part of the Early Greek recounting of ancestry and/or descent. As Graf writes, “Genealogy may appear to have been just a chain of names and not a form of mythical narration. Yet nearly every name entails a story.”169 Graf treats the stories and descent relationships as more closely connected, seeing in their combination the chronological systemization of myth and mythical data. Whether or not a genealogical scheme was imposed upon mythical material and the result was this combined structure of stories and descent/ancestry relationships, whatever its origins, considering the two elements together seems essential to understanding the Early Greek genealogical tradition as it was. That nearly every name entails a story, as Graf writes, is, however, not quite accurate. Not every name gets a story, not even most names, only a select few. Some names appear only as connectors and are often simple eponyms drawn in to link generations or to explain 168 R. Thomas 1989, 174. 169 Graf 1993, 127. 77 topographical names, for example, as Graf points out, the names of Ilos or Tros in the genealogy of Aineias (Il. 20.230 and 232).170 Some names, however, get special treatment with sometimes very elaborate stories of their wealth, adventures, deeds etc... like that of Erichthonios and his wealth and famous horses (Il. 20.219-29). In Homeric poetry, such genealogical stories celebrate key figures, putting the spotlight on the most famous and accomplished of the hero’s descendents, connecting the hero, not only with his ancestors, but with his greatest and most renowned ancestors and their deeds and greatness. The genealogy of Glaukos, for example, encompasses 68 lines of poetry, 47 of which are dedicated to the story of Bellerophon, from his rise to great success, his entrapment by a scorned woman, to his battle with the Chimaira, to his falling out with the gods (Il. 6.156-202). The story is central to the genealogy, as Erichthonios’ is to that of Aineias, Tydeus’ is to that of Diomedes (Il.14.119-25), and Melampous’ is to that of Theoklymenos (Od. 15.226-42). Sometimes the story is shorter, taking up only two to four lines, for example, the story of Dardanos, who founded Dardania before Troy existed, in the genealogy of Aineias (Il. 20.216-18) or that of Kleitos, whom Dawn carried away to live among the immortals, in the genealogy of Theoklymenos (Od. 15.250-51). The genealogies also often relate the stories of earlier relatives that are not in the direct line of descent, the stories of siblings of those in the lineage, for example Ganymedes in the genealogy of Aineias, Kleitos in the genealogy of Theoklymenos, or Amphiaraos also in the genealogy of Theoklymenos, whose story occurs in one of the rare instances of branching discussed above. Although the shorter genealogies (those of Krethon and Orsilochos, Idomeneus, Achilleus, and Telemachos, each less than 10 lines long) do not contain extended stories, they do not read like lists. The short genealogy of 170 Graf 1993, 126. 78 Achilleus culminates in a celebration of Zeus’ strength (Il. 21.192-199), and so the genealogical material seems to build up to an elaboration on a very important ancestor. Moreover, the shorter genealogies are littered, as are the longer ones, with the small details appropriate to epic poetry and a narrative style, i.e., epithets, set phrases, short descriptions, and not so orderly recounting of information. Both the stories and this narrative style lend Homeric genealogy a story-telling character. This kind of story-telling combined with recounting descent relationships that we see in Homeric genealogies is characteristic of most of Early Greek genealogy. It is evident throughout the Catalogue of Women and our remaining examples of genealogical poetry and prose, even in their fragmentary state. At first glance, however, it may seem that the authors of poetic and prose genealogies were mostly concerned with ancestry and descent information, since many of the fragments deal solely with descent and ancestry relationships. But this is an illusion. Most of those fragments dealing with just ancestry or descent are selections of material paraphrased or summarized by the citing author, and so may not be indicative of the style, structure, and entire scope of the original. Therefore, how much the genealogists used a narrative style and told stories and how much they plainly listed descent and/or ancestry is something to be considered by careful examination of their fragments according to type. Thus in order to sort out the nature of the material and the balance between story-telling and listing descent and/ancestry, I will look at the fragments of the Catalogue of Women and the prose and poetic genealogies by the following types: papyrus fragments, direct quotations by citing authors, and paraphrases by citing authors. Papyrus fragments are very useful for determining the balance between the telling of stories and descent/ancestry relationships in the fragmentary genealogies. They generally preserve larger amounts of text than quotations and their material has not been selected and 79 plucked out of context by an author for a particular reason or purpose. Unfortunately all of the examples of papyrus fragments preserving Early Greek genealogical material belong to the Catalogue of Women. Therefore the range of evidence is limited in scope; nevertheless we may add it to the overall picture. The papyrus fragments of the Catalogue of Women contain mythological stories and short sections listing descent relationships. In many cases we see both together in one fragment. For example, Hesiod fragment 31 (Most), preserved primarily on three Oxyrhynchus papyri with a little help from a scholium on Apollonius of Rhodes, has thirtysix lines of extant text. It begins with the brothers Neleus and Pelias, and then recounts the children of Neleus, the last of whom to be listed is Periklymenos, upon whose adventures the poet then elaborates. We are told of the gift of shape-shifting he received from Poseidon, of his prowess in defending his city Pylos, his shape-shifting as he fought, and of his final defeat by Herakles at the will of Athene. This example contains lists of descendents and descent relationships culminating in the extended story of Periklymenos. Another example, this time of fragments of a single papyrus, showing both story-telling and descent relationships combined are the fragments from POxy 1359: Hesiod fragments 117, 120, and 121 (Most). Hesiod fragment 117 (Most) captures a portion of the recounting of the descendents of Arkas, coming in at Auge, daughter of Aleus (grandson of Arkas) and her son by Herakles, Telephus, who is specifically referred to as a descendent of Arkas: Ἀρκασίδην (Hes. frag. 117.8 (Most)). Then the story of Telephus is told, during which the papyrus becomes too fragmentary to read. Two more fragments of the same papyrus are found in Hesiod frags. 120 and 121 (Most), whose material are closely connected, both belonging to the recounting of the descendents of Atlas. In Hesiod fragment 120 we get the descent relationships from Amyklas down to Hyakinthos, who was killed accidentally by Apollo’s 80 discus, after or during which story the fragment ends. Hesiod fragment 121 picks up at the sons of Elektra by Zeus, one of whom, Eetion, we are told, once slept with Demeter and was killed by Zeus, and continues down to Elektra’s grandsons, Erichthonios and Ilos, whereupon the fragment ends. In each of these examples, we see not only both stories and descent relationships, but a combining of the two elements to create a genealogical narrative, in which the recounting of descent and ancestry involves both stories and relationships. It is more difficult to see such direct evidence of story-telling in passages of genealogical material quoted by later authors. This is because they usually represent a very small segment of the original text (generally one to three lines) and they are selected to make very specific points for various purposes, which run from illustrating the ancestry of a particular figure to comparing mythical information in different authors to exemplifying grammatical or semantic practice. For example, Herodian quotes Hekataios to illustrate the placement of accents in disyllabic words ending in -κος (FGrHist 1 F16); whereas, EGF Asios F1 contains a quotation of Asios’ poetry by Pausanias, who compares its information with that in Homer: καὶ ἐπὶ τούτωι πεποίηκεν Ἄσιος ὁ Ἀμφιπτολέμου· “Αντιόπη δ’ἔτεκε Ζῆθον †καὶ Ἀμφίονα δῖον† Ἀσωποῦ κούρη ποταμοῦ βαθυδινήεντος Ζηνί τε κυσαμένη καὶ Ἐπωπέι ποιμένι λαῶν.” (Concerning this, Asios son of Amphiptolemos says in his poem: “Antiope, daughter of Asopos, the swift-eddying river, bore Zethon †and god-like Amphion†, impregnated by Zeus and by Epopeos, shepherd of peoples.”) Both of these quotations, although chosen by the quoting authors for different reasons and despite their relatively small size, reveal a style of prose and poetry that is not particularly list-like. Thus, from such quotations we can see that the genealogists in question (Hekataios 81 and Asios) use a narrative style, with epithets and small descriptions, and do not put their information in any particular prescribed order, such as in a formulaic list. Other direct quotations, of the prose genealogists in particular, however, do appear either to be quite list-like or contain list-like sections preceding or following stories. Consider, for example, Hellanikos fragment 4 (FGrHist 4 F4), quoted by Dionysius of Halicarnassus in a discussion on the origins of the Tyrrhenian race, which contains a very list-like section followed by a story. 171 The list-like section outlines formulaically, in a line from father to son, the descendents of Phrastor, son of Pelasgos and Menippe, down three generations to Nanas: τοῦ Πελασγοῦ [τοῦ Βασιλέως αὐτῶν (sc. τῶν Πελασγῶν)] καὶ Μενίππης τῆς Πηνειοῦ ἐγένετο Φράστωρ, τοῦ δὲ Ἀµύντωρ, τοῦ δὲ Τευταμίδης, τοῦ δὲ Νανᾶς (Phrastor was born of Pelasgos, [their (sc. the Pelasgians’) king], and Menippe, the daughter of Peneios, from him Amyntor, from him Teutamides, from him Nanas) (FGrHist 4 F4.1-3). The quoted fragment continues with the story of how, during Nanas’ reign, the Pelasgians were driven out by the Greeks and eventually settled Tyrrhenia. Other fragments with quoted material reveal similar movements from lists to stories or vice versa. Pherekydes fragments 20 and 66 (FGrHist 3 F20, F66) move from more narrative sections into short list-like sections. Akousilaos fragments 3 and 44 (FGrHist 2 F3, F44) consist of very short sections of list-like material. Given the comparanda in other prose genealogists, Homer, and the Catalogue of Women, and the story-telling elements and character in other fragments of Akousilaos, it is reasonable to argue that Akousilaos followed the seemingly customary genealogical practice of combining descent information with storytelling and that what is quoted in these two fragments are list-like sections such as we see 171 FGrHist 4 F4 = Dion. Hal. Ant. Rom. 1.28.3 82 before or after stories in other genealogical material. 172 Direct quotations that contain both list-like sections and story-telling, show that the genealogists were not in fact creating straightforward lists of descendents or ancestors, but rather telling stories of descent or ancestry, following descent information with stories and stories with descent information. Pherekydes fragment 2 (FGrHist 3 F2; app. 1) presents an interesting challenge. It tells the genealogy of Miltiades the Oikist of the Chersonese and the Elder, (sometimes called the Philaid genealogy) in a lengthy and very formulaic list of descendents from Philaios, son of the Salaminian Aias, to Miltiades the Oikist. Near the end of the genealogy, it begins to impart more information about a few figures, who are believed to be historical. The genealogy is thus not entirely list-like, but because of its formularity and length it does seem to have a much more list-like character than other fragments of the prose genealogists, Pherekydes’ other fragments included. This genealogy, moreover, is unique for its completeness and historicism, i.e., its complete reckoning from mythical figures down to historical figures, and has been much discussed in scholarship for these features as well as for its discrepancies with the information in Herodotus about the family of Miltiades the Oikist. Given its uniqueness and the complexities of the issues, one of which is its transmission and the very corrupt state of the text, and scholarship surrounding it, this genealogy and the fragment to which it belongs will be discussed in much greater detail below alongside other unusual list-like genealogies. It is important for now, however, to acknowledge that in 172 There is always, however, the possibility that some things could have been lost in transmission and that any one of these quotations, especially those of Akousilaos with no stories, could in fact be a list-like paraphrase of information and not direct quotation. That the authenticity of all of the quotations must be doubted, however, seems unlikely and perhaps excessively pessimistic. Moreover, there is precedent for small somewhat list-like sections in Homeric genealogies and the Catalogue of Women. Nevertheless, this idea will be revisited below in the discussion of Pherekydes’ genealogy of Miltiades the Oikist (FGrHist 3 F2), in which we are dealing with not only a quotation, but a quotation of a quotation. 83 Pherekydes fragment 2 we appear to have a lengthy list-like genealogy or, at least, a lengthy list-like portion of a genealogy, but that it is highly unusual among the fragments of the early prose genealogists and may perhaps be better understood through other comparanda. From direct quotations, like that of Asios in Pausanias discussed above, we can also see that later authors drew upon the early genealogists for mythological details, which suggests that stories containing mythical information beyond descent and ancestry were a part of genealogy-making. However, when it comes to citing works, for their mythological information especially, later authors more often paraphrase or summarize their sources than quote them directly. They select the information they need for their particular point or purpose. For example, we can compare the papyrus fragments containing the lineage of Nereus and its culmination in the story of Periklymenos and his death (Hes. frags. 31 and 33 (Most)) with a paraphrase of the same material by a scholiast on the Iliad (Hes. frag. 32 (Most)). From approximately twenty-five lines of narrative in the papyrus, the scholiast condenses the story down to one concise sentence: καὶ δὴ γενόμενον αὐτὸν μέλισσαν καὶ στάντα ἐπὶ τοῦ Ἡρακλέους ἅρματος Ἀθηνᾶ δείξασα Ἡρακλεῖ ἐποίησεν ἀναιρεθῆναι... ἱστορεῖ Ἡσίοδος ἐν Καταλόγοις (And when he [sc. Periklymenos] became a bee and stood upon the chariot of Herakles, Athene, having revealed him to Herakles, caused him to be killed... Hesiod tells the story in the Catalogues) (Hes. frag. 32 (Most)). The scholiast picks and relates only the information he needs to make his point. Among the fragments of genealogical material, descent or ancestry information alone can be plucked out and summarized as necessary, as in Pherekydes fragment 53: Φερεκύδης δέ φησιν αὐτὸν Ὠκεανοῦ καὶ Γῆς (Pherekydes says that he (sc. Triptolemos) was born from Okeanos and Gaia) (FGrHist 3 F53); or Akousilaos fragment 42: Ἀκουσίλαος Φόρκυνος καὶ Ἑκάτης τὴν Σκύλλαν λέγει (Akousilaos says that Skylla was born from Phorkys and 84 Hekate) (FGrHist 2 F42). Such distilling, paraphrasing, and summarizing could mislead us into thinking that Early Greek genealogies were sparse affairs, recounting only descent and ancestry, when the picture actually appears to be quite the opposite. Information other than strictly descent or ancestry information is also paraphrased. For example, Pausanias writes, οἶδα δὲ Ἡσίοδον ποιήσαντα ἐν Καταλόγῳ Γυναικῶν Ἰφιγένειαν οὐκ ἀποθανεῖν, γνώµῃ δὲ Ἀρτέμιδος Ἑκάτην εἶναι (I know that Hesiod in the Catalogue of Women said that Iphigenia did not die, but is Hekate by the will of Artemis) (Hes. frag. 20a (Most)). From such examples we can see that details beyond ancestry and descent are part of Early Greek genealogies. A particular way of citing mythological material in the scholia further suggests that story-telling was a major part of the work of the prose genealogists. The scholia often use the formula ἡ ἱστορία παρὰ... (the story according to...) to cite the paraphrasing of mythical stories from other sources. Among Early Greek genealogical material, the works of the prose genealogists especially receive this treatment. For example, in a scholiast presents the story according to Akousilaos of Aphrodite and Anchises, their love, and Aphrodite’s planning of the Trojan War to benefit their children (FGrHist 2 F39). Likewise another scholiast presents the story according to Pherekydes of Sisyphus, his transgression, and his infamous punishment (FGrHist 3 F119). That the scholiasts went to the prose genealogists and cited them as sources for stories about certain figures or events, shows that there were not only many stories in their genealogies for the scholiasts to draw from, but also that they were considered sources of information beyond descent and ancestry. Stories must have formed a significant part of the works referred to, just as we see in the papyrus fragments in the Catalogue of Women and in Homeric genealogies. 85 Extended stories about figures in a given genealogy also feature prominently in the Theogony. In terms of structure and branching, the Theogony is composed very much in the same vein as other Early Greek genealogical material. The same is true in terms of narrative style and also story-telling. In two prominent analyses of the structure of the Theogony (those of West and Hamilton), the stories are called and treated as digressions.173 They are seen as something separate from, but closely accompanying genealogy which is considered to be only the recounting of the descent relationships of the gods. For example, West writes, “If the Succession Myth is the backbone of the Theogony, the genealogies are its flesh and blood.”174 Consider also West’s synopsis of the Theogony in which he places the ‘Genealogies’ in one column and the ‘Myths and Digressions’ in another.175 This approach of separating the elements of the Theogony, while visually appealing, may lead us to think that stories and myths are something quite different from recounting descent and/or ancestry in the Early Greek world. This is not so. I hesitate to call the stories in the Theogony digressions. They are no mere offshoots of tangential or trivial material; they are as much a part of genealogy-making as recounting sons and daughters. That stories are interwoven with the recounting of descent relationships in the Theogony, or elsewhere, is not an issue or problem and the threads need not be unwoven nor should they be. By taking descent and ancestry apart from story-telling we unravel the fabric to look only at the threads and fail to see the fabric. In doing so, we impose upon the genealogical material foreign notions of 173 West 1966; Hamilton 1989. 174 West 1966, 31. 175 West 1966, 16-18. 86 genealogy-making, and miss the way of doing it in the Early Greek world. To make genealogies was to tell stories. Stories and a narrative style run throughout almost all of extant Early Greek genealogical material, defining and determining nature and structure.176 In the final chapter of part 1, we will see that story-telling and narrative style are intimately connected to the purpose and context of genealogy in the Early Greek world. First, however, we must deal with a few exceptional genealogies, those, few in number, which exhibit neither a narrative style nor story-telling. 176 The stories most probably come before the genealogies and the genealogical ordering of information. Why else would genealogies with heroes and divinities be told, except if the figures were somehow already significant? The stories, relatively short compared to epic story-telling, are drawn or linked from what is obviously a wider body of mythical material. Similarly, see Harding 2007, 181-82; 2008, 2-4, on the stories of the Atthis coming before the chronological scheme. 87 Figure 2.1. The family tree of the Trojan royal family constructed from the information in the Library of Greek Mythology (adapted from Hard 1997, 22) Figure 2.2. The family tree of Aineias constructed from the information in Il. 20.200-41 88 Figure 2.3. The outline of the genealogy of Aineias from Il. 20.200-41 (Segments of story-telling are rendered in italics. Segments of descent information are rendered in bold). 89 Chapter 3: List-Like Genealogies and Historiography The combination of stories with the recounting of descent and/or ancestry characterizes Early Greek genealogy with very few late exceptions, where the genealogical material is presented as a long list with little to no information other than descent relationships given. 177 These are: the genealogies of the Spartan kings Leonidas and Leotychidas in Herodotus (Hdt. 7.204 and 8.131), the genealogy of Alexander of Macedon also in Herodotus (Hdt. 8.139), the genealogy of Miltiades the Oikist (often called the Philaid genealogy) given by Pherekydes (FGrHist 3 F2; app. 1), and the genealogy of Heropythos from an inscription on a tombstone from Chios (SGDI 5656) (figs. 3.1 and 3.2). All can be dated approximately to the mid-fifth century, late in the period of interest for this study. Since all extant instances of list-like genealogies appear in this later period, they may represent a change in the practice or climate of genealogy-making in the fifth century. In belonging to the mid-fifth century, these genealogies also belong to a period that saw the beginnings of Greek historiography, defined by Bertelli as: “the definition of a standard of analysis for the transmitted material; source criticism and the search for ‘rational’ explanation; and … a chronological backbone to order the events.”178 Whether or not all of the above criteria of historiographical practice apply (and shortly I argue that they do not all apply) to works of the early genealogists, the critical (as defined by cultural and political 177 As discussed above, other list-like sections appear in the fragments of the prose genealogists, but these are usually either paraphrases and not direct quotations or small sections of list-like material, and so they cannot be taken as indicative of the structure of the actual genealogical material. 178 Bertelli 2001, 94. Although I adopt Bertelli’s definition of Greek historiography, I do not agree with his conclusion that Hekataios meets all three requirements. As discussed below, I do not accept that Hekataios’ genealogy as reported by Herodotus is evidence of the creation of chronological genealogy and the application of a “chronological backbone to order events” in Hekataios’ works. 90 environments and ideologies) collection, manipulation, and presentation of information from sources occurred in the creation of genealogies, whether from traditional myths, local memories, or even documents. This is not to say that all genealogy-making in this period was subject to such methods; the point is rather that such methods could be used in genealogy-making and were being applied to source material in other genres by this time. It remains, therefore, to be investigated, whether these list-like genealogies may owe their unusual or perhaps novel structure and style to application of the techniques of distillation and compilation to the creation of genealogies in the fifth century. We have seen how later authors paraphrase earlier genealogical works, selecting the details they require for their purposes, and in doing so distil the information into something that little resembles the structure and character of the original. When authors select for the purpose of illustrating descent or ancestry relationships, the result is a fragment with only that information intact. The same process of selection and compilation from more detailed sources, which could also have been genealogical, likely took place in the construction of the list-like genealogies of the mid-fifth century. The corollary of this argument is that genealogy did not necessarily play the role in historiographical development that scholars have more traditionally assigned it, that of chronological impetus, example, or tool, organizing information by generation and therefore time.179 Instead of affecting history-writing by its methods, genealogy-making seems rather to have been affected by the literary techniques of history-writing. As Mitchel writes, 179 The development of chronology from genealogy and the association with a chronological genealogical system of Hekataios, see e.g., Meyer 1892, 153-88; Jacoby 1949, 199; Grant 1970, 18-20; Fornara 1983, 4-7; Luce 1997, 10-12. For various challenges to this view and a variety of conclusions see: Pearson 1939, 96-106, esp. 105-6; Mitchel 1956, 49-52; R. Thomas 1989; Möller 1996; Bertelli 2001. 91 “chronology is patently an outgrowth not of genealogy but of historiography”180 and, as such, may have had an influence on the development of genealogy rather than vice versa. And so, I will begin with the genealogy supposedly made by Hekataios of his own descent, about which Herodotus writes in Book 2 and which lies at the heart of the historiographical connection drawn by some scholars between genealogy and chronology. Then I will consider each of the list-like genealogies of the fifth century in turn, since each is different from the others in transmission and context. Although they share approximately the same time-frame and list-like character, their structures and their relationships to other genealogies and historiographical developments are different and must be explained in different ways. Genealogy, Chronology, and Hekataios According to Herodotus, Hekataios made his own genealogy, going back sixteen generations to god (Hdt. 2.143). The genealogy does not exist for us and we have no evidence of it in the fragments of Hekataios. That it ever did exist is a matter on which we must trust Herodotus or sources. This genealogy sits at the root of the theory that genealogical thinking imposed chronological thinking on the study of and writing about the past. This theory, however, lies on the implicit assumptions about Early Greek genealogies that it was fundamentally chronological and counted generations in a linear fashion and that it only involved descent or ancestry relationships, and the resulting idea that the genealogy 180 Mitchel 1956, 49. 92 supposedly composed by Hekataios was linear, list-like, and only involved descent or ancestry relationships.181 We have little to no evidence to suggest that genealogy in the Early Greek world was inherently chronological. Indeed our evidence suggests that Early Greek genealogy was not and could not have been overly concerned with chronology or chronological thinking about the past. First, as R. Thomas has pointed out, there are only three extant ‘complete’ genealogies, i.e., genealogies which completely recount descent, generation by generation, from an earlier mythic period to the contemporary present. All of these complete genealogies date to around the mid-fifth century, very late in the sample of genealogical material we have been looking at and are a departure in genealogy making from earlier and even contemporary genealogical material. The date of our extant complete genealogies involving historical figures comes too late to have played a role in a development of chronology from earlier genealogies.182 It may be that earlier ‘complete’ oral genealogies provided a model for chronology, but there is, of course, no evidence of such genealogies so it is difficult to say for certain. It does, however, seem a stretch to argue, in the absence of evidence, that the impetus behind the oral material would have been so very different than that behind the preserved written material as to produce such a different product. The stories associated with 181 For the assumption that Hekataios’ genealogy was complete, list-like, and appeared in his Genealogies, see Bertelli 2001, 91-92, where it is supposed that the genealogy looked like those of Heropythos and the genealogy of Miltiades the Oikist in Pherekydes. 182 R. Thomas 1989, 159. These are the aforementioned rather list-like genealogies of Miltiades the Oikist by Pherekydes (FGrHist 3 F2) and Heropythos from Chios (SGDI 5656), and the genealogy of Hippokrates by Pherekydes (FGrHist 3 F59), although whether this last genealogy extended all the way down to Hippokrates is uncertain (FGrHist commentary 1a, 409-10; R. Thomas 1989, 159n6) and it only survives as a description in Soranus’ Life of Hippokrates (Vit. Hippoc. 1). The list-like genealogies in Herodotus could also be considered in this list. It depends, however, on one’s interpretation of them as king-lists or as genealogies. Their relationship to historiography and chronology will be discussed in detail below. 93 the genealogical figures were, after all, of the utmost importance to genealogical thinking. The impetus was to connect the distant past and its illustrious or divine figures with the present. In other surviving ‘genealogies’ in which historical figures claim divine or legendary ancestry, there is a substantial gap between the recent and the distant past as contemporary and recent figures are connected only with a distant legendary ancestor and the middle links in the chain are left out. This is what R. Thomas calls telescoping, the connection of the present with the distant past while ignoring, not caring about, or simply not knowing what comes in between. 183 The resulting gap in genealogical information not only makes it impossible to reckon time by generations, but perhaps more importantly indicates a nonchronological mindset about and purpose for genealogy in Early Greece. 184 Early Greek genealogical thinking involving roughly contemporary figures (historical to us) privileged the latest descendent and earliest ancestor and largely disregarded the intervening links so as to produce most often telescoping connections between the distant past and the present and not exact and thorough genealogical charters accounting for all family members.185 Genealogies of that type certainly would have been useful chronologically and could have provided an impetus to chronological thinking through genealogy, but we do not have evidence of such genealogical thinking. Even our complete genealogies involving historical figures do not 183 R. Thomas 1989, 158. 184 Similarly, R. Thomas 1989, 157-59; Möller 1996, 19-20. Cf. Jacob 1994, 170-71. 185 Recent studies have drawn attention to the lack of completeness, exactitude, consistency, and a charter-like character in Greek genealogies: e.g., R. Thomas 1989, 157-59; Möller 1996, 21; Fowler 1998, 4; Duplouy 2006, 60. Jacob, however, sees this lack as characteristic only of poetic genealogies, but not of prose genealogies, which (he argues) sought to fix those ‘problems’ and could through literate methods (1994, 182-84). 94 conform to such a standard of exactitude, clarity, and breadth. What we do have most often involving historical individuals and mythical figures are statements of relationships between distant ancestor and contemporary descendent, which are not very useful as chronological tools. Genealogies wholly involving mythical or legendary figures also do not provide evidence of chronological thinking in Early Greek genealogy-making. Although they are ‘complete’, in that they usually outline all generations between two given points (whether it be between the heroic subject of a genealogy and his distant, usually divine, ancestor or between the earliest ancestor in great genealogy and any given end point among its branches), such genealogies do not usually connect with the present. The point is not that genealogies involving mythical figures were necessarily conceived of differently from those involving historical figures or that they did not involve events that were considered to be true or historical, e.g., the Trojan War. The point is that Early Greek genealogies involved figures of the distant past (mythical to us) and were not usually extended out of the distant past, beyond the age of heroes, into contemporary time (historical to us). The first evidence we have for such a genealogical practice comes in the mid-fifth century with the three complete genealogies identified by R. Thomas and the genealogies in Herodotus. As Tosetti has shown, the end points of genealogies of legendary figures are connected to current interests and propaganda; however, they are not brought down out of mythical time, out of the age of heroes to connect with people contemporary to the genealogy. 186 Consider, for example, the end of the Catalogue of Women. It ends with Zeus’ plan for destruction, namely the Trojan 186 Tosetti 2006, 113-30. 95 War, at the close of the age of heroes. We do not hear of the generations following those heroes. Möller’s assessment of the fundamental concern of Greek genealogies with the distant legendary past is important and compelling: “Alle Genealogen konzentierten sich im wesentlichen auf die Vergangenheit, die wir mythisch nennen, indem sie z. B. die Familien der Aiakiden, der Herakliden oder der Deukaloniden beschrieben, weshalb sie auch Mythographen genannt werden.”187 Even those few genealogies that do involve historical figures draw heavily upon the mythical period. We are not dealing with a genealogical tradition that is overly concerned with genealogies of the contemporary period, but with genealogies in the contemporary period, as aetiologies, propaganda, or the basis of social or political claims. 188 Early Greek genealogies, with their fundamental interest in the distant and legendary past, show no sense of an overall generational or temporal framework operating within the genealogies by which the information has been organized or which could act as a paradigm of chronological order.189 Chronological ordering is not inherent to genealogy-making. While genealogy-making may have set information from myth into a kinship order, an extended framework of parents and offspring, it does not follow that such a framework is necessarily chronological or arithmetical. Similarly, Möller in arguing against the direct 187 Möller 1996, 19. “All genealogies concentrated fundamentally on the past, which we call mythical, in that they write about, e.g., the families of the Aiakidai, the Herakleidai, or the Deukalionidai, wherefore they are also called mythographies.” 188 The contexts and uses of genealogies will be discussed in detail in Chapter 4. 189 The argument (see, e.g., Jacob 1994; Carrière 1998) that the prose genealogists cleaned up inconsistent mythical poetic genealogies or traditions and set them in temporal order is difficult to prove and has yet to be clearly shown in the evidence of the fragmentary historians. 96 shaping of annalistic models of history from genealogical writing, notes that genealogies only give rough generations and not precise years or arithmetical schemes. She adds that annalistic models divide events from names and reorganize them with the possibility of adding more events, unlike genealogies that connect names and events in the story of a family. 190 Neither the structure nor the character of Early Greek genealogies suggest that a generational or temporal framework, let alone an arithmetical framework, is at play or that chronological thinking is part and parcel of genealogical thinking. As argued in the preceding chapter, Early Greek genealogies are not strictly linear and they have a narrative style and story-telling character. They generally fall into two structural types. The first type, exemplified by Homeric genealogies, is rather linear, progressing from the earliest ancestor toward an individual descendent, the subject of the genealogy. This type has limited branching and when it does branch, it returns to the main line of descent, which leads down to the subject. The second type is ancestor-focused in that it is not geared toward an individual subject, but starts with a common ancestor and then branches out and down, with no one particular descendent in sight. Neither type is list-like. Instead, both types are full of stories and are generally narrative in style and language. The stories are an integral part of the practice of genealogy-making in the Early Greek world. The presence of these stories, sometimes very elaborate and comprising of more than half of a genealogy, does not 190 Möller 2001, 251. Even if later chronographers, such as Kastor of Rhodes and Eusebius, looked to the structure of mythographers, it does not mean that chronology grew from genealogy, but rather that works of chronology could use genealogical information. For further reservations about the development of chronography from genealogy, see Mosshammer 1979, 101-5. 97 particularly evoke chronology and the chronological ordering of information about the past. Instead, it seems to indicate a narrative and story-telling character. What we have, up until the mid-fifth century, are genealogies made up of stories and the recounting of descent/ancestry relationships of the legendary distant past that are unbound by the dictates of chronological structuring and historical time-keeping. The interest, fundamental focus, structure, and character of Early Greek genealogical material does not suggest that chronological order was inspired by Early Greek genealogy. It seems, instead, that literary methods of historiography, including chronological ordering, had an effect on Greek genealogy. Consider the late date of all of our list-like genealogies: they are all mid-fifth century, roughly contemporary to Herodotus. This is clearly after literary historical methods have begun to be applied to thinking and writing about the past. Indeed, it is not until Herodotus, our ‘father of history’, that we see strictly list-like genealogies, which could have been used chronologically or carried an inherent notion of chronology. However, the first evidence we have of a clearly chronological scheme comes with Hellanikos and his priestesses of Hera at Argos and victors at the Karneia. 191 Any idea that Early Greek genealogies were essentially list-like, chronological, complete, and could be connected to historical events plays into an assumption that the genealogy of Hekataios was linear and list-like and therefore did not branch, have a narrative style, and tell stories. If this was so, it would have been unlike any other genealogy that we have extant until the middle of the fifth century.192 The fact is that if Hekataios did write his 191 See Pearson 1939, 105-6, 209-33; Jacoby 1949, 199-200; Mitchel 1956, 69; Möller 1996, 26-27; Marincola 2001, 17-18. 192 In Herodotus’ story, it is the Egyptians who impose a linear counting system upon Hekataios’ genealogy, but even in doing so they must prove to Herodotus through explanation that their system of reckoning is correct and unbroken, that in each generation son succeeded father. 98 own genealogy, Herodotus does not tell us what it was like beyond that a divinity was involved and that it went back sixteen generations. That it went back a certain number of generations does not necessarily mean that that is all it did. How was it structured? Did it branch at key generations? In what style was it written? Did it tell stories? Would a storyteller or mythographer, such as Hekataios, leave out interesting or heroic details about his own ancestors? We do not know what kind of genealogy Hekataios made. If Hekataios made it along the lines of the great genealogies in his work called Genealogies, for which we have fragments surviving as evidence, a linear list-like genealogy would not have been likely at all. It would have been ancestor-focused, contained branches, and involved story-telling. Or he may have composed his genealogy like one of the personal heroic genealogies in Homer, more linear, with story-telling and maybe a little branching, but moving toward an individual subject, himself. Both of these options, for which Hekataios had ample precedents, seem more likely than a bare-bones, list-like, chronologically-minded genealogy, for which he would have no precedents of which we are aware.193 Either Hekataios is the first and makes a novel list-like genealogy unlike anything previously seen or he composed something in the vein of Early Greek genealogy. I suspect the latter. List-like genealogies were probably not the brain child of Hekataios nor a sudden unprovoked development in genealogical writing. Nor were they the impetus to impose chronology on writing about the past. They were probably the result of the development of literary techniques of historywriting, which brings us to Herodotus. 193 All instances of list-like genealogies appear in the mid-fifth century, and we know that if Hekataios wrote a genealogy, he would have done it before Herodotus wrote his histories. 99 Herodotus’ List-Like Genealogies The list-like genealogies in Herodotus (those of the Spartan kings and of Alexander of Macedon) represent a development in the fifth century in genealogy-making, as the products of the application of literary historiographical techniques of compilation and distillation in the creation of genealogies. Therefore, they cannot be taken as evidence that list-like genealogies influenced literary techniques through an innate sense of chronology, but rather that the use of literary techniques created something better suited to chronological organization. Any discussion of these genealogies, however, must inevitably tackle the problem of whether they are genealogies or if they are king-lists. There is disagreement in the scholarship on this question. Henige argues that the lists are genealogical, because Herodotus never purports that they are king-lists.194 Further arguments supporting this position, as pointed out by Cartledge, include that a number of known rulers are absent from the lists and that Herodotus cross-references the προγόνοι (ancestors) of Pausanias, who was only a Spartan regent, with the earlier list belonging to Leonidas, thus indicating that Herodotus was concerned with kinship and a figure who was not a king. 195 Cartledge, however, in arguing against the position that the lists are genealogical, gives reasons why certain rulers are missing from the lists and gives evidence from fragments of Early Greek poetry on papyri that some of the names from Herodotus’ list, who were previously unknown as rulers, were indeed called kings in Sparta.196 Cartledge concludes, “on balance” that 194 Henige 1974, 208. 195 Cartledge 2002, 294. 196 Cartledge 2002, 294-95. 100 “Herodotus did indeed mean the lists for king-lists” and eventually concludes that in Herodotus’ lists “we have access to king-lists.”197 That we may have access to information on kings of Sparta, however, does not mean that Herodotus intended his lists to be king-lists. There is an important distinction to be made here between king-lists (lists of kings in succession) and lists with kings. I agree with Henige’s assessment, that Herodotus was first and foremost creating genealogies. There is simply no getting around the fact that Herodotus connects the names through a formula of descent relationships: x, son of y, son of z, etc.... The lists, whether they contain the kings of Sparta or not, are expressed as a genealogy recounting ancestry. A further indicator that Herodotus’ lists were genealogies is his statement after the list associated with Leotychidas: οὗτοι πάντες, πλὴν τῶν ἑπτὰ τῶν μετὰ Λευτυχίδεα πρώτων καταλεχθέντων, οἱ ἄλλοι βασιλέες ἐγένοντο Σπάρτης (All of the others were kings of Sparta, except the first seven recorded after Leotychidas) (Hdt. 8.131.3). Prakken writes that this statement would “suggest that Herodotus considered his table a kinglist, with certain exceptions, as well as a genealogical chart.”198 Cartledge also interprets the statement as implying that Herodotus thought he was making king-lists.199 I think that it suggests something quite different, that Herodotus produced a genealogy that involved kings and the sons of kings and that precisely because what he produced was not meant to be a king-list, he had to include the connection between kinship and kingship after the fact. This, added to the cross-referencing between the προγόνοι (ancestors) of Leonidas and Pausanias, 197 Cartledge 2002, 294-95. 198 Prakken 1940, 462. 199 Cartledge 2002, 294. 101 who moreover was not a king, and the overall form of the lists which clearly recount ancestry relationships, suggests that we are very much in the realm of kinship and not kingship. 200 Kingship does, however, come into the picture, but only by nature of the men whose ancestries are being recounted, men who are important, have illustrious ancestries able to be traced back to a demi-god (Herakles) and available to Herodotus. A genealogy of kings, even a mythical one, is not necessarily a king-list. They may overlap, especially where there is hereditary succession or the desire to create the impression of hereditary succession. However, instances of broken or collateral succession (brothers succeeding brothers) can throw a king-list out of alignment with genealogy, even when trying to maintain the impression of unbroken succession. Because the genealogies in Herodotus, including that of Alexander of Macedon, involve kings and the sons of kings does not make them king-lists, even besides the fact they do not exclusively involve kings. Moreover, many Early Greek genealogies involve members of royal families who were never kings, e.g., the genealogy of Aineias at Il. 20.200-41. Should we call the genealogy of Aineias a king-list because many of his ancestors were kings of Troy? Probably not. The important question seems not to be about Herodotus’ intentions as much as it is about the nature of his sources. Were they king-lists, genealogies, or something else? Did he use multiple sources? Were they as list-like as Herodotus’ product? The sources of Herodotus are difficult to assess and the trick here is to advance far enough to help make sense of the scope, style, and structure of the genealogies in Herodotus without falling into the bottomless vortex that Quellenforschung, if taken too far, could create. Prakken’s assertion, following Meyer, and supported by Jacoby and Wade-Gery, that the king-lists 200 Similarly, Möller 2001, 252-53. 102 compiled by Hekataios were Herodotus’ source for the Spartan lists, is tenuous at best and searching for the closest figure in the dark at worst.201 The connection with Hekataios seems only to be based on two tenuous and circuitous sets of calculations. 202 The first set concerns the date of Herakles in Herodotus (placed at 1330 BCE), the number of generations recorded in the Spartan lists (21), and the supposedly resulting use of forty-year generations, which are supposedly used by Hekataios but not usually by Herodotus (who says that there are three generations every hundred years, but only says so once [Hdt. 2.142]). The first calculation rests on the assumption that Herodotus was concerned about keeping the number of generations in the genealogy consistent with a date for Herakles and that he had himself done the math. That he does the math for the Egyptian chronology, where he makes his statement that there are three generations in every hundred years, is not as relevant here as it might seem. There Herodotus is dealing with a list clearly associated with time-keeping and is working out a particular puzzle. The Spartan lists in Herodotus, however, are genealogies not chronological lists and there is no indication that Herodotus treated them as time-keepers or felt that they had to match up with any particular date for Herakles or follow any particular generational scheme. The second set of calculations, relying on the argument that Herodotus adapted kinglists to make his genealogies by replacing certain contemporary names, places the compilation of the ‘original’ king-lists during the reign of Demaratos, ca. 510-491 BCE, and 201 Meyer 1892, 170-71; Prakken 1940, passim; Wade-Gery 1952, 76n28, 90-91. For scepticism on this theory see Pearson 1939, 105-6; Mitchel 1956, 64-66; Möller 1996, 26. 202 For an outline and critique of Meyer’s long-standing argument that Hekataios developed a chronological scheme based on forty-year generations, see Mitchel 1956, 64-66. See also Möller 2001, 251-53, for a critique of the place of Herodotus’ Spartan genealogies in the beginnings of chronology. 103 this is then compared to the dates of Hekataios’ career and his Genealogies, which are only tentatively placed within that time. 203 Linking up these two sets of calculations, the argument goes that Hekataios’ work informed Herodotus’ because of forty-year generations and because Hekataios may have worked during the period in which they may have been compiled. Even besides that fact that we have no evidence of such a list or generational chronology in the fragments of Hekataios, the argument is very shaky. The calculations and argument on the whole seem to be ripe with circularities and hopeful speculation, especially the attribution to Hekataios on the basis of the time in which he lived and worked. This approach does not seem particularly helpful in determining the nature of Herodotus’ sources. Better results may be achieved by looking at the genealogies in Herodotus and any clues he may give us about his method and by considering other genealogical material as comparanda. The other part of Prakken’s argument about the sources of Herodotus is that they were king-lists.204 This argument is tenable. That each man who gets his genealogy told in Herodotus is a member of a royal family (our three Spartans and the Macedonian Alexander), suggests that the sources may have indeed been king-lists. Herodotus’ statement at 8.131.3, that all but two of the ancestors of Leotychidas were kings, suggests that his sources contained more information than he relates in his genealogy. Herodotus shares after the genealogy some further information that may have been present in his sources. Such information about kings could have been derived from a source that presents information on kings, i.e., from a king-list, but not necessarily. It is also possible that Herodotus, like the later authors who quote and paraphrase the mythographers, derived his descent and ancestry 203 See Prakken 1940 for specific details of the argument. 204 Prakken 1940, 466. 104 information as well as information about kinships from genealogical material. 205 Take the example of Aineias’ genealogy again (Il. 20.200-41). We learn from this genealogy that his ancestors were kings of Dardania and Troy and this type of information is important to the genealogical narrative. Ultimately, however, as is often the case, the author’s sources remain elusive and the matter ambiguous. Thus Prakken’s statement that “there can be no doubt that in the form in which he found them they were king-lists” is far from reflecting the reality of the situation.206 It certainly seems likely, but not free from doubt. So what did Herodotus do with those sources, whatever they were? While I am sceptical about Prakken’s identification of Hekataios as Herodotus’ source, I can agree with Prakken that Herodotus applied techniques of selection and compilation in creating his genealogical lists. If Herodotus’ sources were indeed king-lists, then the process of selection and compilation is certainly at work here. He would have had to select and adapt the information to create a genealogy from a king-list. If his sources were genealogies with information on kings, he would have had to distil the information to create his terse genealogies. We can also see something of Herodotus’ process of genealogy-making, in the cross-referencing of the genealogy of Leonidas with the προγόνοι (ancestors) οf Pausanias at 9.64. Herodotus does not give the genealogy of Pausanias, because he writes that he has already recounted the names in the genealogy of Leonidas, τῶν δὲ κατύπερθέ οἱ προγόνων τὰ οὐνόματα εἴρηται ἐς Λεωνίδην· ὡυτοὶ γάρ σφι τυγχάνουσι ἐόντες (The names of his ancestors have been mentioned above in relation to Leonidas, for they 205 Furthermore, if Herodotus’ source was indeed Hekataios and the Spartan information would probably belong to the Genealogies, as Prakken himself argues (1940, 467-68) and therefore be a genealogy. But I still think the connection is tenuous at best and this argument therefore rather irrelevant. 206 Prakken 1940, 466. 105 happen to be the same) (Hdt. 9.64). Herodotus recognizes that the genealogies of the men overlap. His means of coming to this conclusion is unknowable to us, but we can speculate on a few options. He may have compared separate sources containing genealogical information for the two men, which would mean that both Pausanias and Leonidas had their own separate source. It seems improbable that there were two distinct sources out there for each man, when they share the same lineage. Moreover, it seems unlikely that Pausanias as a regent and an unexpected leader would have had his own genealogical source. Another option is that knowing Pausanias’ immediate ancestry, his father and grandfather but not having a separate genealogical source for Pausanias, he linked his descent with that of Leonidas. This seems more probable, given Pausanias’ status as regent. While we cannot know from which sources exactly Herodotus compiled the genealogies of Pausanias and Leonidas and by what methods he connected them, it is apparent that he synthesized the genealogical information from his sources. Herodotus correlated the two genealogies, recording only one from beginning to end, he then distilled the information into a simple statement to stand in for the genealogy of Pausanias. In this cross-referencing, we see Herodotus’ method of compiling, synthesizing, and distilling information at work in the creation of genealogies. Herodotus thus appears to have used literary techniques to construct his list-like genealogies, working from more detailed sources, king-lists or genealogies. His genealogies are the product of distillation and compilation. Therefore, list-like genealogies, such as may be useful in inspiring chronological organization and thinking, actually do not seem to appear before the influence of literary techniques on genealogy making. 106 The Genealogy of Miltiades the Oikist The genealogy of Miltiades the Oikist or the Philaid genealogy, as it is more commonly called in the scholarship, is probably the most studied genealogy from ancient Greece. It appears in discussions on the development of historiography, on elite selfpromotion, on political propaganda and positioning in fifth century Athens, on aetiological genealogies, and on aristocratic families and family tradition.207 A problem with all of this attention being paid to this genealogy, however, is that it is highly unusual and does not accurately represent the evidence of Early Greek genealogies and genealogical thinking. It is not the norm, but the exception. Compared with other fragments of the prose and poetic genealogists, Pherekydes included, the genealogy stands out for its length, its treatment of figures we know to be historical, its ‘completeness’ in full generations from an early legendary ancestor to an historical figure, its very limited element of story-telling, and its preservation of a list-like style through several generations and through what small segments there are of story-telling. Unlike most other genealogical material, it would be well-suited to inspire chronological thinking. It has the appearance of a highly ordered genealogy. But this is not the norm for Early Greek genealogies or even those of the fifth century. This has implications for arguments made with the genealogy as evidence. For example, the idea that there was a tradition of genealogies stretching from the present into the distant legendary past, inherited and preserved by aristocratic families, cannot be supported by this one unusual fragment.208 It is unlike most other genealogical material and therefore cannot alone 207 E.g., Meyer 1892; Jacoby 1947; Wade-Gery 1951, 1952; Huxley 1973; Viviers 1987; R. Thomas 1989; Jacob 1994; Möller 1996; Higbie 1997; Duplouy 2006. 208 Cf. Momigliano 1971, 24. 107 represent standard genealogical practice in the mid-fifth century, neither in arguments about the structure, style, and scope of Early Greek genealogies in general, nor in arguments about their context, purpose, and use. So what, then, can we establish about the structure, style, and scope of this particular genealogy so that we may situate it among the other evidence and be able to use it while discussing the context, purpose, and use of genealogies in the Early Greek world in the following chapter? In the following paragraphs, I will outline a few possible explanations for the unusual character of the genealogy of Miltiades the Oikist, comparing and contrasting the genealogy with other fragments and genealogical materials in order to situate it among what seems to be more standard Greek genealogical practice in the fifth century. One possible explanation for the unusual ‘completeness’ and length of the genealogy of Miltiades the Oikist lies in its construction from contemporary information and the use of literary methods to extend that information into a ‘complete’ genealogy. R. Thomas argues that the genealogy is the product of literary methods applied by Pherekydes to oral family tradition. The argument goes that Pherekydes, working from a family tradition which was subject to telescoping (the common phenomenon of connecting only ancient ancestors with contemporary figures and leaving out the middle links), used literary methods to stretch out names known from family tradition in a linear fashion in order to create a genealogy that was unbroken from ancient legendary ancestor to contemporary subject.209 Thus the source of Pherekydes’ genealogy would be family tradition and the literary method used on that source would be largely the chronological manipulation of information. The unusually list-like structure of this one genealogy in Pherekydes, then, extrapolating from R. Thomas’ 209 For the full argument, see R. Thomas 1989, 161-73. 108 argument, would result from the simple adoption of names from family tradition. This is a likely possibility for Pherekydes’ method of compiling and composing. However, while R. Thomas’ arguments suggest what may have been Pherekydes’ sources and method, we still cannot explain why the end product is so exceptional in its style. A challenge in dealing with this genealogy is that because it is unusual, it is tempting to distance this one fragment from the rest of the fragments of the prose genealogists and discuss it alone. We should consider, however, that its exceptionality may be better understood through its commonalities with other material. Although the genealogy of Miltiades the Oikist is unusual and therefore cannot stand alone as evidence of standard Greek genealogical practice, it is not entirely dissimilar to other Greek genealogical material and, therefore, is probably best understood within that context. Two points of comparison are particularly enlightening toward understanding the listlike style of the fragment. First, the genealogy is not absolutely list-like, in that, like other genealogical material, it has an element of story-telling, providing information beyond that of descent or ancestry. Alongside descent relationships, the genealogy records: that Philaios settled in Athens; that during Teisander’s archonship something happened (exactly what we do not know because of a lacuna in the text); that during Hippokleides’ archonship the Panathenaic festival was established; and that the last Miltiades in the list settled the Chersonese. These pieces of information consist of only a few words and do not quite seem to measure up to the elaborate stories we see in most Early Greek genealogies, but they do look like remarks made in other Early Greek genealogical material, in which only an epithet or a small amount of information is provided. The common formula ἐφ’ οὗ ἄρχοντος (in the archonship of) is nicely suited to such a remark, being easy to insert and follow with the rest of the genealogy. In Early Greek genealogical material short stories and remarks are 109 often contained in relative clauses (e.g., Il. 20.233-35; FGrHist 1 F15; FGrHist 3 F39, F101). Such insertions, at the very least, indicate that the genealogy of Miltiades the Oikist by Pherekydes did not just contain descent information, despite its highly list-like character. The second point of comparison is that a list-like style similar to that in the genealogy of Miltiades the Oikist appears, albeit in very short stretches, in other fragments of the prose genealogists held to be direct quotations (FGrHist 2 F3, F44; FGrHist 3 F20, F66; FGrHist 4 F4).210 All of these examples have already been introduced above in connection with storytelling and the combination of descent information and stories in Early Greek genealogies, because they depict the movement between descent information in a list-like style and the telling of stories. The fragment of Hellanikos (FGrHist 4 F4) illustrates this movement particularly well, proceeding from list into story. We also see movement from story into list (FGrHist 3 F20, F66). What we see illustrated here, captured by these fragments, are genealogies as they progress back and forth between list-like sections involving descent information and sections of story-telling or elaboration on an individual figure in the genealogy. It could be then, that what we have in this fragment of Pherekydes is a piece of a list-like section between stories like we see in other Early Greek genealogical material, the fragments of Pherekydes among them (FGrHist 3 F20, F66). It is important to recognize here that what we have in FGrHist 3 F2 is only a fragment of the original, a portion selected by later authors. Whether or not Miltiades the Oikist was actually the final figure in the genealogy depends on whether or not we have the full genealogy preserved in our source, quoted in its entirety by both Didymus and Marcellinus. 210 Thanks are owed here to Robert Fowler for drawing my attention to some of the fragments of the mythographers which also employ list-like styles. 110 Although Marcellinus’ intentions were supposedly to connect the genealogy to Thucydides and therefore likely would have continued the genealogy down as far as he possibly could, we cannot assume that we have the full genealogy as far as it went. We do not have Didymus’ intentions for citing Pherekydes on this matter and therefore cannot surmise the length of his quotation. Moreover, given that we do not have the context out of which the quotation was selected and it seems doubtful that this small piece of prose (despite its length in generations, it is not a wordy genealogy taking up only a few lines of text) was published or produced on its own, it is safe to say that we probably do not have the complete genealogy. There could have been further generations in the genealogy. But more likely, judging from the pattern of other genealogical materials, including that of Pherekydes, is that there were other branches, leading down from Aias or the other sons of Aias. We know that Pherekydes wrote great genealogies collected in a large single work along the lines of the Catalogue of Women and the works of Hekataios and Akousilaos. The genealogy of Miltiades the Oikist belonged to one of these great genealogies of Pherekydes. It is one branch among many in a great genealogy, likely of the Aiakidai, the descendents of Aias, or the Asopidai, the descendents of Asopos, including Aias and Achilleus.211 Marcellinus relates that the quotation comes from the first book of Pherekydes’ Histories (FGrHist 3 F2.3) and that Hellanikos covers the information among the Asopidai (FGrHist 4 F22). The rather list-like genealogy of Miltiades that Oikist that has come down to us is a portion of a larger genealogy, perhaps a list-like section in between sections of story-telling, belonging to 211 Following this argument, the genealogy should probably be called a branch of the Aiakidai or Asopidai genealogies, but since it survives for us as a fragment and we thus need a separate name as shorthand, I think the ‘genealogy of Miltiades the Oikist’ is the most descriptive of the information in the genealogy. 111 a branch of a non-linear, ancestor-focused genealogy, such as we see in the Catalogue of Women and the works of prose and poetic genealogists, Pherekydes among them. The genealogy in Pherekydes fragment 2 is not extraordinary for its employment of a list-list style, but it is extraordinary for the length and preservation of its list-like style through so many generations and during segments of non-descent information. A comparison with the other two fragments of Pherekydes that employ a list-like style illustrates the similarities and this important difference nicely. All three of the fragments begin with details about an ancestor. In both fragments 2 and 20, the information offered is about where that ancestor lived or settled. Toward the end of the list-like section, fragments 2 and 66 begin to add information beyond descent.212 Once fragment 66 expands into a story, it abandons its list-like syntax and shortly thereafter the quotation ends. Fragment 2, however, begins to add short remarks several generations before the end of the fragment, incorporating them so formulaically that it never loses its list-like style, adding remarks only at certain generations. This strict adherence to a list-like style is unusual. It is also unusual for its length. The genealogy of Miltiades the Oikist lasts for eight generations of purely descent information straight and for four more after other information begins to be added, maintaining its list-like character for twelve generations. The other fragments have list-like sections that are either much shorter or truncated. The genealogy is, therefore, similar to other fragments with list-like sections, but different in two key ways. So if the fragment represents a list-like section between stories, how can we explain the lengthy and pervasive list-like style? 212 This does not happen in fragment 20, as the quotation appears to be truncated after Echepolos as the scholiast who quotes it appears only to be interested in his descent (Schol. (T) Il. 23.296c). 112 A consideration of the fragment’s transmission may be helpful here. The genealogy of Miltiades the Oikist is presented as a direct quotation in critical editions of Pherekydes’ fragments and of Marcellinus, in whose work it is quoted by Didymus. 213 The genealogy is a quotation of a quotation and it is possible that it is not a direct quotation. The question is: was it Pherekydes or a later writer who gave the information its list-like structure and style? It is possible that the information in the genealogy of Miltiades the Oikist could have been distilled by a later quoting author, either Didymus or Marcellinus and that the style and structure of the genealogical material of Miltiades the Oikist in Pherekydes’ work may have been much more detailed, along the lines of his other fragments. Marcellinus would have had ample reason to do so, being that his purpose in using Pherekydes was ostensibly only to show that Thucydides was descended from Aias. Marcellinus’ syntax and word choice, however, suggest that he believed it to be a quotation: καὶ τούτοις Δίδυμος μαρτυρεῖ, Φερεκύδην ἐν τῆι πρωτηι τῶν Ἱστοριῶν φάσκων οὕτω λέγειν· ... (And Didymus gives evidence of these things, saying that Pherekydes wrote thus in the first book of his Histories: …) (Marcellinus, Vita Thuc. 3; FGrHist 3 F2). It is also possible that Didymus could have paraphrased and not quoted Pherekydes on the genealogy and the result could have been a list-like paraphrase such as we see done by other later writers. This paraphrase then could have been taken up as a quotation and mistakenly cited as such by Marcellinus. This is not a suggestion of sloppiness in Didymus’ scholarship, in the vein of ancient and modern criticism of his scholarly abilities.214 There 213 See critical editions by Jacoby (FGrHist 3 F2), Fowler (EGM Pherekydes F2), and Piccirilli 1985 (Marcellinus, Vita Thuc. 2-5). 214 On Didymus’ ancient and modern reputation as a scholar and in defence of Didymus, see Gibson 2002, 5462; Harding 2006, 31-39. 113 was lack of accepted ancient principles for quoting or citing the ideas of other writers, and so the interests, agenda, and purpose of the quoting author largely determine how ideas were excerpted and represented.215 Didymus’ style and technique of excerption from other writers, however, as far as can be seen from P. Berol. 9780 (Didymus on Demosthenes), lean heavily toward detailed citations and large verbatim quotations. 216 And so quotations seem to have fared quite well in Didymus’ hands, especially if, as Gibson suggests, the fragments of the papyrus were excerpted by another scholar from a larger commentary.217 His interest seems to be in recording verbatim the opinions of other scholars, rather than in selecting and condensing particular pieces of information to present in paraphrase. It seems less likely then that Didymus himself would have paraphrased the genealogy quoted by Marcellinus from a more detailed source. I have presented three possible explanations for the unusual list-like character of the genealogy of Miltiades the Oikist preserved in Pherekydes fragment 2. The first, argued by R. Thomas, is that Pherekydes used contemporary information from family tradition to ‘complete’ the genealogy generation by generation from legendary ancestor to recent historical figure. While this may have been Pherekydes method, it does not fully explain why the end product appears to be so different from other genealogical material. Comparison with other Early Greek genealogical material gives us a better idea of the genealogy’s context and suggests some other explanations for its unusual-looking character. The second explanation is that the genealogy of Miltiades the Oikist may be a paraphrase of a more 215 On ancient methods of excerption from sources, see Gibson 2002, 6-7. 216 Harding 2006, 20, 31, 34. 217 Gibson 2002, 66-69. Cf. Harding 2006, 13-20, on the nature of the work transmitted by P. Berol. 9780. 114 detailed work of genealogy or a list-like section in between stories, a piece of a branch in a great genealogy. The third explanation for the genealogy’s unusual character lies in a combination of these two possibilities: it is a section of a branch of a great genealogy with a long list-like section, whose list-like style may have been furthered by paraphrasing. This would explain its not absolutely list-like character, unlike the strictly list-like genealogies in Herodotus or genealogy of Heropythos, and its relationship to other Early Greek genealogical material. These explanations are all possibilities and have been presented tentatively. I suspect that the answer may lie in the second possibility, that it is a branch of a larger great genealogy; however, its transmission is tricky and the influence of members claiming the lineage should not be entirely dismissed. Whatever the case may be, the genealogy of Miltiades the Oikist is best understood, for all its distinctiveness, alongside other fragments of Early Greek genealogy and ought not to stand alone out of context as a norm of prose genealogy and Greek genealogy in general. The Genealogy of Heropythos The genealogy of Heropythos on the tombstone from Chios is equally, if not more troublesome to interpret than the other list-like genealogies of the mid-fifth century (figs. 3.1 and 3.2). It cannot be placed within a developing tradition of historical writing, in that we know nothing of its author and it has no context within a work of history. Moreover, it is not only unusual for its list-like structure, but unique among extant Early Greek genealogical material for being an inscription and among surviving genealogical inscriptions for its early date. 115 We can, however, compare it to later inscriptions, as Chaniotis does, but this does not get us far.218 Seeing a tradition of genealogical epigraphy, Chaniotis discusses the genealogy of Heropythos in combination with three other genealogical inscriptions: two from the second century BCE from Miletus and Crete (Milet no. 422 and Inscr. Cret. III S.56 iii 8) and one from Cyrene (SGDI 4859), dated to the third century BCE by Chaniotis, but equally plausibly to the first to second centuries CE by Masson.219 These four examples, however, are too few and too disparate, coming from different areas and time periods, to represent a single tradition. In addition, as a whole, the examples of genealogical inscriptions from the ancient Greek world do not seem to have a lot in common beyond relating descent and/or ancestry and being inscriptions. The genealogy of Heropythos is similar in structure, style, and scope only to the inscription from Cyrene, in that both are list-like and deal only with descent from father to son. The other inscriptions encompass more family members, including women, and are not nearly as list-like. The genealogy of Heropythos, as an inscription, may belong to a different visual tradition of genealogy or a visual development in the same tradition. Or it may be that the genealogy’s epigraphical medium determined the structure and information presented in the genealogy. Although the medium of inscription does not necessarily impose brevity on the material being inscribed, it could have been a factor for the creator or inscriber of this genealogy. While this is a possibility, I present it warily. Any conclusion based on this unique piece of evidence can only be tentatively stated. Thus, an understanding of the listlike structure of the genealogy remains somewhat elusive. Given its date in the fifth century 218 Chaniotis 1987, 43-44. 219 Chaniotis 1987, 43; Masson 1974, 266n17. 116 and its apparent uniqueness, however, it seems unlikely that the genealogy of Heropythos represents a long or widespread tradition of list-like genealogies in the Early Greek world, but rather a unique or local development in the fifth century. The list-like genealogies that appear in the fifth century are more likely the product of the use of literary techniques on sources containing genealogical information (whether the sources are genealogical or otherwise), than they are chronological tools or indicators. Herodotus applies literary methods to compile and create his genealogies, perhaps from kinglists. Pherekydes’ genealogy of Miltiades the Oikist could owe its list-like structure and style, as R. Thomas argues, to a process of elongation inspired and accomplished by literary techniques. Since it belonged to a larger great genealogy, the quotation in Marcellinus may also represent a list-like section in between stories. The possibility also remains that it could have been paraphrased by Didymus into a distillation of the original. The structure and style of the genealogy of Heropythos is difficult to interpret because of its uniqueness. It seems to belong to a different tradition of genealogy-making that is not literary and not for recitation, but may simply just be concerned with descent information. If these list-like genealogies came about in the mid-fifth century under the influence of literary and historiographical techniques of writing and thinking about the past, it indicates that the original Early Greek genealogical impulse was not to create lists of descendents or ancestors at all, but to tell stories of ancestors or descendents largely of the distant past. The purpose and context of which genealogical stories is the subject of the following chapter. 117 Figure 3.1. The Genealogy of Heropythos, Grave Stele from Chios, ca. 450 (SGDI 5656) following Wade-Gery’s transcription (1952, 8: fig. 8) 118 Figure 3.2. The Genealogy of Heropythos, Grave Stele from Chios, ca. 450 (SGDI 5656) (Wade-Gery 1952, 8: fig. 8) The information removed is a photograph of the Genealogy of Heropythos, Grave Stele from Chios, ca. 450 (SGDI 5656) (WadeGery 1952, 8: fig. 8) 119 Chapter 4: The Types and Uses of Early Greek Genealogies There are a few ways in which scholars have attempted to categorize Greek genealogies and genealogical material. Broadbent, as discussed in Chapter 2, is concerned with the mythical versus the historical, separates genealogies into the historical, the forged or made-up, the mythic or fictitious, and the historiographic sociological. 220 This interest in the true and the false is not terribly useful, here, in studying a culture in which the categories of true and false do not necessarily coincide with the equation of true with historical and false with mythical. Also concerned with myth and history, but in a different way, Mitchel divides Greek genealogies between poetry and prose, seeing poetic genealogies as mytho-poetic, artistic creations and prose genealogies as scholarly and non-literary works of synthesis. 221 Treating prose and poetic genealogies as separate entities in this way, however, fails to see the whole tradition of genealogy-making in the Early Greek world. Such divisions in the genealogical material seem to be partly a result of theories of historiographical development, which hold that prose genealogies were a development in historiography putting order to mythical material and partly a result of modern ideas of history and myth, which require dividing the false from the true. Either premise is faulty. A quick consideration of Early Greek genealogical material shows approaches dividing the true from the false and the poetic from the prose to be flawed. The structure of the prose material, as far as we can surmise, does not seem to be all that different from what we see in the Catalogue of Women. Indeed, the major structural difference between genealogies, as discerned above, cuts across poetic genealogy. More linear, single- 220 Broadbent 1968, 18. 221 Mitchel 1956, 49-50. 120 descendent-focused genealogies appear in Homeric poetry, whereas tendrilled ancestorfocused genealogies that branch out and down are found in the Hesiodic poetry. Moreover, poetic and prose genealogy deal largely with the same mythical material. The only difference between prose and poetry as far as genealogical material is concerned is that poetic genealogies lack figures considered by us to be historical. However, given the extremely few mentions at all of historical figures in the genealogical material and the very tiny amount of material surviving from poetic genealogists such as Kinaithon and Asios, who seem to have been very much interested in the myths of their respective poleis and may have thus included historical figures relevant to their poleis’ past, I am extremely hesitant to put too much stock in such an absence. It must also be noted that among extant Early Greek genealogical material there is only one genealogy that may be considered to be wholly made up of historical figures, the genealogy of Alexander of Macedon in Herodotus (8.139), and even that genealogy is associated with a mythical legend of how the ancestor Perdiccas claimed the kingdom of Macedon (8.137-38). Figures considered by us to be historical appear among figures considered by us to be non-historical or mythical, e.g., the genealogy of Hippokrates connecting the father of medicine to Herakles and Asklepios (FGrHist 3 F59) or the genealogies of the Spartan kings in Herodotus going back to Herakles (6.204 and 8.131). The same sort of connection between myth and history appears in the genealogical stories associated with historical figures in the praise poetry of Pindar. 222 Although the historical figures are not linked completely generation by generation through these stories to the mythical family in question, this does not represent a mythical-historical divide, but as R. 222 E.g., Ol. 2.35-48; Ol. 6.24-25 and 28-73; Ol. 7.20-38 and 92-94; Pyth. 4.247-62; Nem. 11.33-37; Isthm. 3.1317b. 121 Thomas argues, a telescoping that brings the distant past up close to the present. This association between the storied past and its figures and the contemporary world seems to be the whole point of expressing kinship here. Thus, to separate the mythical from the historical is to ignore the very connection that lies at the heart of Early Greek genealogy. Classifications of Greek genealogies, then, may more appropriately be based upon the types of connections drawn between the distant or early past and the contemporary world. How did Greek genealogies link the present and the distant past? How could genealogies solely of the legendary past relate to the present? Who and what did they link? Are the genealogies associated, directly or indirectly, with historical individuals, families, peoples, or poleis? In what context, to what end, and in whose interest were genealogies created in the Early Greek world? Many scholars see Greek genealogy as primarily associated with aristocratic families and focus on their role in establishing and securing the power and prestige of those families. 223 Momigliano, for example, writes, “Greek aristocracy shared the passion for genealogical trees which characterizes any aristocracy.”224 The problem is, however, that he is only able to cite the genealogy of Heropythos and the difficult genealogy of Miltiades the Oikist in Pherekydes, which he calls, as is appropriate to his argument, the Philaidai genealogy, as evidence that “quite a few families” produced lengthy genealogies. The argument is thus not based on an assessment of the overwhelming majority of Greek 223 Wade-Gery 1952, 92; Van Groningen 1953, 47-61 passim; Momigliano 1971, 24; Finley 1975, 27, 48; Dowden 1992, 10-11; Gras 1995; Nicolai 2007, 17. R. Thomas, although appropriately critical of the concept of lengthy inherited familial genealogies, sets genealogies squarely in the realm of family and family tradition (1989, 157). 224 Momigliano 1971, 24. 122 genealogical material, but on two exceptional and difficult to interpret genealogies. It has, however, remained a profoundly influential assessment and represents the standard view of Greek genealogy among Greek historians in general. For example, Gras cites the same quotation from Momigliano above to explain the use and context of Greek genealogy in his textbook on the Archaic Mediterranean.225 Nicolai, similarly, in touching briefly on the topic of Greek genealogy in his article “The Place of History in the Ancient World” in the recently published Companion to Greek and Roman Historiography, writes that genealogies “continued and interpreted the epos and had the aim of consolidating and organizing the memories of aristocratic clans (genē).”226 As we will see in the course of this chapter, however, this understanding of Greek genealogy as primarily familial breaks down under the weight of the evidence of Early Greek genealogical material as a whole. Some scholars, however, recognizing more of the tradition of Greek genealogy, see different types of Greek genealogical material, namely, familial genealogies and civic or ethnic genealogies. Van Groningen, interpreting genealogy within a tribal model of Early Greek society, associates Greek genealogies with groups, namely clans and ‘larger’ groups, concluding that the solidarity of the group in the past was important to the solidarity of the present.227 Hall sets the genealogies he studies into the category of ethnic, and others as familial, although he appears to approach the second category more hesitantly. 228 This hesitancy is justified. The overall association of genealogies with groups, group solidarity, 225 Gras 1995, 65. 226 Nicolai 2007, 17. 227 Van Groningen 1953, 61. 228 Hall 1997, 41. 123 and group interests, requires reconsideration in light of the complex body of extant genealogical material. Möller, recognizing more dimensions in the genealogical material and that more than just group interests are involved, sees three purposes to genealogy in the ancient Greek world. 229 (1) Genealogies explained or expressed conceptual relationships. (2) They explained or expressed alliances or relationships between groups within the polis (e.g., phylai and genē) and between poleis, through mythical kinship connections. (3) Genealogies served as expressions or claims to status and prestige for families and individuals. While Möller admirably expands upon the purpose of genealogies in the Greek world, there is still a focus on groups and familial groups in her work that requires refinement. The problem still lies with the association of genealogies with kinship groups within the polis. The evidence of Early Greek genealogy as a whole does not suggest a strong connection between genealogy making and kinship groups, such as clans or tribes. As we have seen in the previous chapters, neither the structure nor the information conveyed support elaborate elite family genealogies resulting from family tradition in Dark Age and Archaic Greece. Early Greek genealogies are not concerned with drawing connections between contemporary individuals and depicting large family groups. Even when they involve large families, such as in the great genealogies in the Catalogue of Women, they are much more vertical in character and structure, tending to link generations vertically through time and not horizontally across branches drawing contemporaneous connections. Moreover, Greek genealogies are primarily patrilinear, and, although women are sometimes included (more so in the Catalogue of Women than in other examples of genealogical information), they may be left out, because it is ultimately male ancestry or 229 Möller 1996, 19-20. 124 descent that is being recounted. Such genealogies with a very limited scope of information and web of relationships do not depict kinship groups as groups, nor relationships within kinship groups nor the connections between all members of a kinship group as a group with earlier legendary ancestors. Were membership within a group, contemporary connections within the group, or mythical connections of the whole group to an ancestor of high priority, one might expect a more thorough recounting of kinship connections than what we see in Early Greek genealogical material. Early Greek genealogies also do not commonly include historical individuals, although they are often associated with historical individuals. The instances of historical figures being placed within a genealogy are very few. Instead we tend to see them in telescoping situations as in Pindar’s praise poetry, where the historical figure is said to be related to distant ancestors, but the individual is not placed within a genealogy connecting him to those ancestors. The few instances where historical individuals do feature in descent or ancestry information and are not just associated with the genealogy do not express kinship groups and their interests.230 Some of the instances come from fragments which do not contain direct quotations. 231 Others we have already discussed as unusually list-like genealogies.232 We might also include the genealogy supposedly written by Hekataios 230 Some historical events can also be found among the fragments of the prose genealogists, especially those of Hellanikos and Damastes. Such fragments are not easily identified as genealogical material, but they could represent elements of the story-telling in genealogical material. They will be discussed below in connection to the relationship between the prose genealogists and their contemporary world. 231 E.g., Hippokrates in Pherekydes (FGrHist 3 F 59); Andocides in Hellanikos (FGrHist 323a F24a, b, c = FGrHist 4 F170a, b, c); and Miltiades the Oikist in Hellanikos (FGrHist 4 F 22). 232 These are: the genealog
UBC Theses and Dissertations
Early Greek kinship Varto, Emily 2009
Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.
- 24-ubc_2010_spring_varto_emily.pdf [ 1.77MB ]
- JSON: 24-1.0068689.json
- JSON-LD: 24-1.0068689-ld.json
- RDF/XML (Pretty): 24-1.0068689-rdf.xml
- RDF/JSON: 24-1.0068689-rdf.json
- Turtle: 24-1.0068689-turtle.txt
- N-Triples: 24-1.0068689-rdf-ntriples.txt
- Original Record: 24-1.0068689-source.json
- Full Text