Microsoft Word - Kazakevich.dot Greek, Roman, and Byzantine Studies 48 (2008) 343–380 © 2008 GRBS Were the χωρὶς οἰκοῦντες Slaves?* Emily Grace Kazakévich Edited by Deborah Kamen Editor’s Introduction Born in 1911, Emily Randolph Grace was one of the six children (including the archaeologist Virginia Grace) of Lee Ashley and Vir- ginia Fitz-Randolph Grace. She received her B.A. in Greek cum laude from Bryn Mawr in 1933, staying on at Bryn Mawr as a Fellow dur- ing the academic year 1933–1934 and receiving her M.A. in Greek in 1934. Entering graduate school in Classics at Yale in 1935, she co- wrote (with Margaret Crosby) an essay on Yale’s Achaean League coin collection, An Achaean League Hoard, which was published by the American Numismatic Society in 1936. Some time thereafter, Emily married Vladimir Kazakévich, an Economics instructor in the Army Specialized Training Program at Cornell University. Both Emily and Vladimir were committed Marxists: in fact, in the 1940s they trans- lated at least two Communist works from the Russian.1 Emily earned her Ph.D. from Yale in 1949, with a dissertation entitled The Sparta of Agis and Cleomenes: A Study of the Ancient Literary Sources. In the same year, Vladimir was exposed as an alleged Soviet spy,2 and the couple promptly moved to Moscow.3 Beginning in 1953, and continuing for *Originally published in VDI 73.3 (1960) 23–42 1 Political Economy in the Soviet Union: Some Problems of Teaching the Subject, transl. E. G. and V. D. Kazakévich (New York 1944), a translation of an anonymous article originally published in the Soviet journal Pod Znamemem Marksizma (Under the Banner of Marxism) 7–8 (July–August 1943); and A. L. Leontiev, Marx’s Capital, transl. E. G. Kazakévich (New York 1946). 2 The former Communist spy Elizabeth Bentley testified against Vladimir on June 5, 1949 (see “Miss Bentley Lists Ten More as Spies,” New York Times, 6 June 1949); cf. the August 27, 1948, issue of the newsletter Counterattack: Facts to Combat Com- munism, which reported that she would testify against Vladimir in September 1948. 3 See “Accused Russian Returns to Soviet,” New York Times, 18 January 1950. The Kazakéviches’ life in Moscow is vividly described by the Canadian diplomat John Benjamin Clark Watkins in his dispatches and letters home to his colleagues. Watkins and Vladimir, who had first met at Columbia University in 1930, recon- 344 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? three decades, Emily was employed at the Institute of World History of the Academy of Sciences of the USSR, where she published num- erous articles on social status and Greek law.4 She died in 1986. Although Emily’s article “Were the χωρὶς οἰκοῦντες Slaves?” is frequently cited as the definitive study on this particular status group in classical Athens,5 most scholars have been unable to read it— hence my desire to bring to light, at long last, an English translation. The following translation derives primarily from a typescript Emily mailed to Professor Ronald Stroud piecemeal in 1977 and 1978 (with addenda submitted in May 1979). The typescript does not represent an exact translation of the original article: in a letter to Stroud dated May 3–8, 1978, Emily wrote, “Quite a bit of revising, expanding (partly from marginal notes on my working copy) has been incor- porated, as you will see and have seen.” Because she neglected to ___ nected in Moscow in 1951. See J. Watkins, Moscow Despatches: Inside Cold War Russia (Toronto 1987) 27–33, 74–76. 4 As E. G. Kazakévich: “The Term δοῦλος and the Meaning of ‘Slave’ in Athens in the 4th Century B.C.,” VDI 57.3 (1956) 119–136; “On the Role of the First Forms of Capital in Athens in the 4th Century B.C. (In Light of the Economic Theory of K. Marx),” VDI 64.2 (1958) 29–39; “Slaves as a Form of Wealth in Athens in the 4th Century B.C.,” VDI 64.2 (1958) 90–113; “Were the χωρὶς οἰκοῦντες Slaves?,” VDI 73.3 (1960) 23–42; “On Slave Agents in Athens,” VDI 77.3 (1961) 3–21; “The Revolt of Saumakos: Contribution to a Polemic,” VDI 83.1 (1963) 57–70. As E. Grace: “Concubines in Classical Athens,” VDI 103.1 (1968) 28–52; “Athenian Views on What is a Slave and How to Manage ‘People’,” VDI 111.1 (1970) 49–66; “Status Distinctions in the Draconian Law,” Eirene 11 (1973) 5–30; “Aristotle on the Enfranchisement of Aliens by Cleisthenes,” Klio 56 (1974) 353–368; “The Legal Position of Slaves in Homicide Cases,” VDI 128.2 (1974) 34–56; “A Note on Dem. XLVII 72: τούτων τὰς ἐπισκήψεις εἶναι,” Eirene 13 (1975) 5–18; “Status Distinctions in Plato’s Homicide Law,” VDI 139.1 (1977) 71–81. Posthumously, as E. Grace: “The ‘Altogether Xenos’ as Killer in Attic Law,” VDI 203.4 (1992) 28–46 and VDI 204.1 (1993) 25–39, both parts found in Grace’s personal archives and translated into Russian by L.V. Semechenko. 5 E.g., by R. S. Stroud, “An Athenian Law on Silver Coinage,” Hesperia 43 (1974) 182 n.97; E. Perotti, “Esclaves χωρὶς οἰκοῦντες,” in Actes du Colloque 1972 sur l’esclavage (Paris 1974) 47 n.1; E. Erxleben, “Die Rolle der Bevölkerungsklassen im Aussenhandel Athens im 4. Jahrhundert v.u.Z.,” in E. C. Welskopf (ed.), Hellenische Poleis I (Berlin 1974) 477 n.132; H. Klees, Herren und Sklaven (Wiesbaden 1975) 5 n.28; G. E. M. de Ste. Croix, The Class Struggle in the Ancient Greek World (Ithaca 1981) 563 n.9, who says that I. Bie-żuńska-Małowist in “Quelque formes non typiques de l’esclavage dans le monde ancien” in Antichnoe Obshchestvo (Moscow 1967) 92 n.1 cites “an evidently useful article in Russian (which I do not read) by Emily Grace Kazakevitch [sic]”; H. Klees, Sklavenleben im klassischen Griechenland (Stuttgart 1998) 144 nn.92, 93, 94, 97, 146 n.105; E. E. Cohen, The Athenian Nation (Princeton 2003) 149 n.99. EMILY GRACE KAZAKÉVICH 345 translate the last three pages and the Appendix,6 a translation of these pages has been provided by Olga Levaniouk. I have added some bibliography and editorial comments, enclosed in brackets, to the notes.7 June, 2008 Department of Classics University of Washington Seattle, WA 98195 dkamen@u.washington.edu Author’s Introduction8 My discussion of the χωρὶς οἰκοῦντες problem is the fourth of six research articles on various aspects of slavery in Athens.9 All six appeared in Vestnik drevnei istorii before the journal began (in 1967) to provide English summaries. Since among classical scholars and ancient historians in other parts of the world there are still very few who read Russian, these early studies went largely unremarked. The notable exception is the comment by F. Gschnitzer on two of my slave-term articles10 in his study of the same subject, Studien zur grie- chischen Terminologie der Sklaverei I: Grundzüge des vorhellenistischen Sprach- gebrauchs (Wiesbaden 1963). Recently there has been a sprinkling of references to my chôris oikountes study, but the way it has been cited has made it obvious that those citing it could not have read it. Indeed how could they? Here, too, an exception must be registered. E. Erxleben, in his article “Die Rolle der Bevölkerungsklassen im Aus- senhandel Athens im 4. Jahrhundert v.u.Z.,” in E. C. Welskopf (ed.), Hellenische Poleis I (Berlin 1974) 477 n.132, comments favorably on my conclusion toward the end of the article that by the “dwellers 6 In another letter to Stroud, dated September 1977, she says she stops three pages before the end of the article, “first because there is no time, but also because I regard what follows as not very helpful guesswork, which might be found adequately summed up by E. Erxleben in his article in Hellenische Poleis, Berlin (East), 1973 [sic], p. 513, n.132. Erxleben thought this the main point of the article; for me it is the least important thing in it. Who these people were is a problem that really does need much digging and thought (my present notion hovers over mercenaries or the kind of refugees we have evidence of a little later in the century—but, in short, I haven’t done enough about this to make a sensible proposal, really).” 7 I am very grateful to Ron Stroud for providing me with his correspondence with Emily, and to Olga Levaniouk for translating the final pages of the article. 8 [This Introduction was not part of the original article, but was written by the author in May 1978 and mailed to Stroud.] 9 [The other five are “The Term δοῦλος,” “The First Forms of Capital,” “Slaves as a Form of Wealth,” “Slave Agents in Athens,” and “The Revolt of Saumakos.”] 10 Namely, “The Term δοῦλος” and “Slaves as a Form of Wealth.” 346 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? apart” Demosthenes must have had in mind free foreigners. The only thing new about this conclusion is the “must,” and that is founded on the necessity to accept what remains when other pos- sibilities have been eliminated. For me the question as to who these people were was of secondary importance as compared with the prob- lem of establishing who they were not. For the prevailing identifica- tion of them with a category of privileged slaves had led to distortion of valuable evidence on different ways in which slaves were exploited economically (a subject I was investigating when this article was written). Rented-out slaves are lumped together in one category with slaves allowed to conduct enterprises on their own account simply because both often “lived apart” from their owners. The category hoi chôris oikountes is even treated as a synonym of ἀνδράποδα μισθο- φοροῦντα (rent-bearing slaves). Thus an important term distinction made in the contemporary sources—the distinction between μισθός- and ἀποφορά-yielding slaves—is disregarded in favor of plainly speculative definitions found in late sources. As for who the “dwellers apart” were in the time of Demosthenes, I am afraid that I can make no better answer now than I did twenty years ago: since (if my argument then was sound) they could not have been slaves or freedmen, they must have been free foreigners. Erx- leben does not comment on the validity of my negative points, unless his finding my “positive” conclusion worth considering implies ap- proval of what went before. Instead he cites A. R. W. Harrison, The Law of Athens I (Oxford 1968) 167 nn.5 and 6, where there is a con- venient summation of modern views on the status (slaves or freed- men?) and legal position of the “dwellers apart.” In Harrison’s own opinion they were “a clearly distinguishable category of slaves” and in n.6 he nicely exemplifies the muddle this question has got into when he says that in Men. Epitr. 378–380 (my A 3 infra) Syriskos “is shown … to be a δοῦλος μισθοφορῶν.” Perhaps the foregoing remarks are sufficient to explain the author’s concern that this article be made accessible to interested colleagues everywhere, and to justify its re-issue, in revised and ex- panded form, as a contribution to the general exchange—and clash —of opinion that is so necessary to life and health in any field of enquiry. * * * EMILY GRACE KAZAKÉVICH 347 N THE FOURTH CENTURY B.C. the expression χωρὶς οἰκοῦντες, “those who live separately” or “apart,” was used of persons who (like metics and citizens) might be called up for service in the Athenian fleet. This bit of information is given in a passage from a speech by Demosthenes; nothing else is said about these people in sources of the classical period. It would seem, in the circumstances, no easy task to make out just who are meant by “the dwellers apart,” what place they filled in Athenian society of that time. Yet modern scholars have found no great difficulty in placing them. Taking a cue from one lexicographer of the Roman era and another of the Byzan- tine, whose interpretations of the phrase are fairly close to one another (the later definition is probably a modification of the earlier one), modern opinion has come to regard the question as settled well enough within the limits of two variants. The solution, especially in the variant that will chiefly concern us here, is noticeably in harmony with prevailing ideas about the condition of slaves in classical Athens. It was arrived at less by comparative analysis of the pertinent sources than by inflating the significance of what may be called the terminological aspect of the concept “dwellers apart,” an approach to the problem which at any rate gives the appearance of being based on source analysis. The result is that, like a ship encrusted with barnacles, the essential question—what sort of group did Demosthenes have in mind?—is now overgrown by conceptual material of a much later epoch, which, if further progress is to be made, must be dealt with first of all. Harpocration (s.v. τοὺς χωρὶς οἰκοῦντας) states his opinion that the “dwellers apart” were obviously freedmen; a Byzantine lexicographer repeats this definition and also suggests an al- ternative interpretation: that the expression refers to slaves who live separately from their owners. Both variants have found impressive backing in the modern literature. The first was adopted by G. Busolt in the third edition of his Griechische Staatskunde;11 the second is preferred by W. L. Westermann 11 See G. Busolt, Grieschische Staatskunde I (Munich 1920) 274 and II (Munich 1926) 985; O. Schulthess, RE s.v. μισθοφοροῦντες; [L. Gernet, Droit et société dans la Grèce ancienne (New York 1955) 169 with n.4; J. H. I 348 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? and, more recently, by A. R. W. Harrison.12 Adherents of the second variant identify the “dwellers apart” with slaves permitted to have their own enterprise and retain part of the proceeds, enjoying a considerable measure of economic inde- pendence and also other privileges, the right to live separately from their owners being regarded as the most characteristic of them.13 In connection with this interpretation of chôris oikountes, ___ Lipsius, Das attische Recht und Rechtsverfahren (Leipzig 1905–15) 622 n.6 (also 798 n.29); see also the Loeb (“freedmen”) and Budé (“les affranchis”) trans- lations of Dem. 4.36]. 12 See [M. Clerc, Les métèques athéniens (Paris 1893) 281, 291;] G. Gilbert, Handbuch der griechischen Staatsalterthümer (Leipzig 1893) 191; L. Beauchet, Histoire du droit privé de la république athénienne (Paris 1897) II 445ff.; J. Partsch, Griechischer Bürgschaftsrecht (Leipzig/Berlin 1909) 136; A. E. Zimmern, The Greek Commonwealth (Oxford 1922) 264 n.2; [A. Diller, Race Mixture among the Greeks before Alexander (Urbana 1937) 145 n.47;] G. R. Morrow, Plato’s Law of Slavery (Urbana 1939) 18, 73; W. L. Westermann, The Slave Systems of Greek and Roman Antiquity (Philadelphia 1955) 12, 16–17, 23, 38, 122; V. Ehren- berg, The People of Aristophanes (New York 1961) 188 n.6; Harrison, Law of Athens I 167; [Perotti, “Esclaves”; E. Perotti, “Contribution à l’étude d’une autre catégorie d’esclaves attiques: les ἀνδράποδα μισθοφοροῦντα,” in Actes du Colloque 1973 sur l’esclavage (Paris 1976) 181–194; de Ste. Croix, Class Struggle 142 with 563 n.9; Y. Garlan, Slavery in Ancient Greece (Ithaca 1988) 71; M. H. Hansen, The Athenian Democracy in the Age of Demosthenes (Oxford 1991) 121; E. E. Cohen, Athenian Economy and Society: A Banking Perspective (Princeton 1992) 97; J. Trevett, Apollodoros, the Son of Pasion (Oxford 1992) 155; S. C. Todd, The Shape of Athenian Law (Oxford 1995), esp. 188; Klees, Sklavenleben, esp. ch. 4; Cohen, Athenian Nation 130–154; E. E. Cohen, “Athenian Prosti- tution as a Liberal Profession,” in G. W. Bakewell and J. P. Sickinger (eds.), Gestures: Essays in Ancient History, Literature, and Philosophy presented to A. L. Boegehold (Oxford 2003) 218; R. Zelnick-Abramovitz, Not Wholly Free: The Concept of Manumission and the Status of Manumitted Slaves in the Ancient Greek World (Mnemosyne Suppl. 266 [2005]) says (215–216) that the term could refer to either slaves or freedmen, but elsewhere (e.g. 269, 289, 293) she as- sumes that they are slaves]. [Uncertain about their status: A. Calderini, La manomissione e la condizione dei liberti in Grecia (Milan 1908) 374–375; D. Whitehead, The Ideology of the Athenian Metic (Cambridge 1977) 25 n.87.] 13 The persistence of this interpretation is perhaps best shown by quoting some studies devoted mainly to other wider questions: Partsch, Griechischer Bürgschaftsrecht 136, citing L. Beauchet, Histoire II, calls the chôris oikôn a “half- free slave.” Zimmern, Greek Commonwealth 264, remarks simply that slaves who worked “on their own” were called chôris oikountes, citing Dem. 4.36. EMILY GRACE KAZAKÉVICH 349 reference is often made to a group of passages in the classical sources where there is mention of slaves working under the supervision of persons other than their own masters. The same group of texts provides a word combination, ἀνδράποδα μισ- θοφοροῦντα,14 often taken as another name for the “dwellers apart” category of privileged slaves, and also the term ἀποφο- ρά, which denotes a payment made by the slave to his owner. My purpose here is not to question the fact that there were slaves in Attica whose masters might grant them this or that privilege of an economic or personal nature. Neither do I doubt that a substantial number of slaves lived separately from their masters. There is, however, an a priori unlikelihood that the two groups—privileged slaves and slaves living apart from their masters—comprised a single category. Yet the identifica- tion is often made in the modern literature. It is brought to its logical extreme by Westermann, who includes in his category of privileged slaves those who live apart from their masters because they have been rented out for exploitation by owners of workshops, mining enterprises, and the like. This is surely to enroll in the ranks of the privileged too large a part of the Attic slave population. So wide-ranging a concept of the category of privileged slaves is clearly the result of undiscriminating treatment of the texts cited as evidence that such a category existed, which treat- ment must, I suppose, also be held responsible for the identify- ing of this category with slaves called andrapoda misthophorounta, a designation found twice in the classical sources.15 In the litera- ___ Morrow, Plato’s Law of Slavery 73, writes that many slaves in Athens lived apart from their masters and had an enterprise of their own; in a footnote he terms such slaves chôris oikountes. For Harrison the chôris oikountes were “a clearly distinguishable category of slaves … who were not part of their master’s household at least in the narrow physical sense” (Law of Athens I 167). 14 Or δοῦλοι μισθοφοροῦντες—essentially, of course, the same idea, but expressed in a form which is attested only for the Byzantine period: see Bekker, Anec. 1 212.12, where it is said of μισθοφοροῦντες δοῦλοι that “nowadays” (νῦν) they wait about at the Anakeion, i.e. the shrine of the Dioskouroi. 15 Texts M 6 and 7 in the series appended to this article. See W. Wyse, 350 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? ture this designation has been associated—both by those who regard chôris oikountes as freedmen and by those who think they were privileged slaves—with slaves who were allowed to exercise some initiative in the organization of their work. Moreover the “μισθός-bearing” slave has been assigned to the “privileged category” even by an author who elsewhere, in the same publication, defines andrapoda misthophorounta as slaves subjected to the double exploitation of being rented out to another master.16 Similarly, in respect to “those who live apart” we find it said in Gilbert’s Handbuch der griechischen Staats- alterthümer, first that they were slaves, next that, in the only text from the classical period in which they are named (Dem. 4.36), the reference is to freedmen.17 Term usage in texts referring to forms of payment to a slave’s owner is subjected to the same arbitrary treatment: the word apophora is associated with the payment made to his owner by the slave allowed to run his ___ The Speeches of Isaeus (Cambridge 1904) 115f., commentary on Is. 8.35 (my M 7). [See also Cohen, Athenian Economy 97–98: “certain slaves, the so-called douloi ‘living on their own’ (khôris oikountes) or ‘self-supporting’ (mistho- phorountes)”; Klees, Sklavenleben 103: “die sog. χωρὶς οἰκοῦντες, für die auch die Benennung μισθοφοροῦντες angenommen wird” (although he does ac- knowledge on 143–144 that there is debate about whether these terms are synonymous, citing this very article); Zelnick-Abramovitz, Not Wholly Free 216 n.67: “The choris oikountes might have resembled slaves hired out by their masters (ἀνδράποδα μισθοφοροῦντα), if the latter were allowed to keep part of their wages for themselves.” Cf. Perotti, “Contribution,” who argues that the andropoda misthophorounta were similar to, but less privileged than, the chôris oikountes; and de Ste. Croix, Class Struggle 563 n.9: “From the χωρὶς οἰκοῦντες we must in principle distinguish slaves hired out to others (and re- ferred to by some such expression as ἀνδράποδα μισθοφοροῦντα).”] 16 Compare O. Schulthess, RE s.v. μίσθωσις, where it is said that (in my M 6 and 7) the andrapodon misthophoroun is a rented-out slave, with the same author’s earlier article in the same volume [Schulthess (s.v. μισθοφοροῦν- τες)], where he uses this term (without citing any source) as designating “mit wirtschaftlicher Selbständigkeit arbeitende Sklaven.” 17 Gilbert, Handbuch 191, 193. [See similarly de Ste. Croix, Class Struggle 563 n.9, who says that chôris oikountes in Dem. 4.36 “must refer mainly if not entirely to freedmen”; also Klees, Sklavenleben, who says that these chôris oikountes may not have been slaves (130) and that they were “vielleicht Frei- gelassene” (307 n.62).] EMILY GRACE KAZAKÉVICH 351 own enterprise18 and also with the income accruing to the owner from a rented-out slave. The same disregard for ancient term distinctions is manifested when an author (in this case Schulthess) who takes pains to stress the difference between these two forms of exploitation, citing for the rented-out slave my texts M 7, 6, 5, 8, 1, 4 (see below), where the payment is designated only by the term misthos, can then cite A 2, 3, 5— where nothing is said about renting out slaves—as showing that the owner’s income from a rented-out slave was called apo- phora.19 When the language of the sources is handled as freely as this, when misthos, apophora, and hoi chôris oikountes are jammed to- gether under one umbrella, there is danger of distortion in two directions. In the first place an exaggerated impression is created of the quantity of evidence on privileged slaves.20 In the second place failure to respect the term distinctions made in the 18 See Beauchet, Histoire II 445ff.; Busolt, Grieschische Staatskunde I 274; Zimmern, Greek Commonwealth 264 (“slave-rent”); Schulthess (s.v. μισθο- φοροῦντες); Lauffer, Die Bergwerkssklaven von Laureion I (Wiesbaden 1956) 69f., 84, and II (Wiesbaden 1957) 175; R. J. Hopper, “The Attic Silver Mines in the Fourth Century B.C.,” BSA 48 (1953) 247 and n.342; [Perotti, “Esclaves” 50, 55; Perotti, “Contribution” 185–186; Cohen, Athenian Econ- omy 93–94; Todd, Shape of Athenian Law 188; Klees, Sklavenleben 144; Zelnick- Abramovitz, Not Wholly Free 215;] and finally, the definition in LSJ9 s.v. ἀποφορά, “esp. money which slaves let out to hire paid to their masters,” citing my A 5, 4, 1, 3, and the Menander fragment cited in my n.63. 19 Schulthess, s.v. μίσθωσις. On the meaning of μισθοῖντο in M 8 see my n.62. On apophora as payment for a rented-out slave, see also A. Boeckh, Die Staatshaushaltung der Athener 3 I (Berlin 1886) 90f. (where my texts M 3 and 7 are cited) and the criticism of Boeckh’s position on this in M. S. Kutorga, The Social Position of Slaves and Freedmen in the Athenian Republic (in Russian), published posthumously in Sobraniye Sochinenii (St. Petersburg 1894) 153– 560, esp. iv, 230f., 179f., 213 (citing Boeckh, Die Staatshaushaltung der Athener2 I [Berlin 1851] 101). 20 Some historians—e.g. Busolt, Grieschische Staatskunde I 283; cf. Schul- thess (s.v. μισθοφοροῦντες)—underline the relatively negligible size of the category of privileged slaves [cf. Cohen, Athenian Nation 130–154]. However, the very fact that slaves mentioned in my M-texts are too often assigned to this category (e.g., by Busolt 274 n.3, with reference to M 6 and 7) en- courages the inclusion in this category of all slaves let out for hire. Much de- pends, of course, on how one understands μισθοφορέω when used of slaves. 352 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? sources leads to the obscuring of valuable data on the different ways a slave might be exploited when not working under the immediate supervision of his owner. The data lie in the terms (misthos and apophora) used of pay- ments due the slaveowner for the use of his slave’s income- producing capacities and in the context in which the term appears in each series. The connection between these data and the chôris oikountes problem is a matter more of conjecture (or assumption) than of observation and is tangential to the subject of this article. As the question stands now, however, the desig- nation chôris oikountes is caught up in a tangle of other terms (including the two just mentioned), which have been, in various combinations, tacked onto the concept “privileged slaves.” It will therefore be necessary to present a brief account of the results produced by a separate investigation of the M- and A- texts. So much is needed for the light it sheds on the true relationship between these income-terms and the category of privileged slaves, and also as a preliminary to a critique of Wes- termann’s treatment of that category. Largely on the basis of the evidence provided in the texts M 1–4 and A 1–3,21 I came to the overall conclusion that the difference in term usage between the two series is not a chance or “style” variation but reflects a real economic distinction, a difference in the form of exploitation. The texts in the two series have in common the fact that they deal with slaves who work not under the supervision, or for the sole profit, of the slave’s owner. From the slave’s standpoint, on the other hand, the relations indicated by the M-texts would be radically differ- ent from those indicated by the A-texts, and this difference does coincide with a difference in the term used to denote the form of payment in each case. 21 Two texts from the post-classical era are omitted from the series appended to this article: Teles ap. Stob. Flor. 95.21 and 5.67, which are sometimes cited in connection with privileged slaves. Both texts appear in a compilation of the fifth century CE, where they are attributed to a philos- opher whose floruit is conjecturally put in the third century B.C. (Christ- Schmid II 88). Both the nature of the source and the uncertainty as to the date of composition make doubtful the usefulness of these texts as evidence, particularly as regards terms. EMILY GRACE KAZAKÉVICH 353 Texts M 1–4 show quite clearly that when the term misthos (or its cognates) is used in a context involving the master-slave relationship, the slave is the object of transaction between the man who owns him and the man who has hired the use of his skills. In other words the term misthos in such a context denotes not pay for work, not “wages” (as in the case of misthoi received by the free hired worker, μισθωτός),22 but “rent” paid for hired property, that is, for the slave himself. The evidence of M 1–4 further shows, a fortiori, that the transaction implied by the words andrapodon misthophoroun in M 6 and 7 (a word combina- tion found only here) was of the same sort. Indeed, as already noticed, M 6 and 7 have sometimes been understood in this sense even by those who on another occasion may use the ex- pression as a general designation for the privileged slave. These observations regarding M 6 and 7 are not invalidated by the fact that in application to people (and not to houses or animals) the verb μισθοφορέω often has the meaning “to be paid wages.”23 My point is that in his business calculations the slaveowner counts his slave as an object, at most as an income- yielding object. This is his normal starting point when he con- templates the economic uses of his slave. It can be modified in certain circumstances, such as when the slave appears as a “subject,” an independent operator. But in M 6 and 7, both the evidence of other texts, on which my overall conclusion was based, and the immediate context in which the words andrapoda misthophorounta appear show that the participle means “yielding rent.” In M 7 these words are used alongside and, so to speak, 22 Slaves allowed to make money on their own account might, of course, work for hire, in which case their earnings would presumably be termed misthoi. Pl. Leg. 742A (a passage to which Y. A. Lentsman called my at- tention) is perhaps an example of this; but the sum they paid to their owners for the right to earn money independently would still be called apophora. On the misthôtos see Pl. Resp. 371E. 23 In the Demosthenic corpus, for instance, it always has that meaning. Cf. Xen. Oec. 1.4. The term misthos can also indicate payment for work done by free craftsmen (Pl. Leg. 921C), and μισθαρνία is used by Aristotle in reference to the same category of workmen; it refers also, of course, to the wages of free hired labor (e.g. Dem. 49.51–52; see my n.31) or the pay of military personnel (Dem. 50 passim). 354 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? on the same parallel with οἰκία μισθοφοροῦσα, “rent-bearing house.” In M 6 the parallel is drawn, though perhaps some- what ambiguously (see n.61), between ζεῦγος (team) and andra- podon misthophoroun. There is in fact no ground for supposing that “rent-yielding” slaves were privileged as a species, still less that andrapoda misthophorounta was used in this sense. On the term apophora we have much less information. From the late fifth and fourth centuries I have so far been able to find only six instances of the use of this term for payments made by a slave to his owner.24 This is not too firm a base on which to found conclusions about the nature of the economic relations indicated by this form of payment. However, by combining the information given in the A-texts with indications provided by other types of contexts it is possible to introduce an element of firmness into the rather vacillating conception of these relations which has hitherto prevailed. This conception tends to associate with the word apophora two different forms of exploitation (see above), although both the term usage and the general sense of the passages cited in evi- dence (i.e. mainly my M- and A-texts) speak to the contrary. For one thing the terms misthos and apophora are never found in joint reference to one and the same type of income, as might be expected if, for instance, misthos were the general term and apo- phora the particular, more specific designation of “slave-rent,” as Zimmern supposed. For another, in the exploitation relation connected with the term apophora the slaveowner received this payment not from a third party for the slave but from the slave himself.25 Another characteristic of apophora relations is the evi- dent possession by the slave of means “from” (apo-) which to 24 Among them I count Men. fr.431 Kock (see my n.63) which, brief as it is, at least does nothing to upset my position as to the meaning of the term. The word apophora appearing in Marchant’s OCT text of Xen. Vect. 4.49 is an emendation (see his apparatus). In all the other numerous mentions in Vect. 4 of slaves rented out for work in the mines, the form of payment, where expressed, is designated by the term misthos, which payment, as the context makes clear, is hire-money (rental) due to the owners of the slaves from mining entrepreneurs who have hired (rented) them. 25 See A 1–3; cf. Men. fr.431 Kock. In A 4 the owner implies that he went himself to Laurium to fetch the apophora. Cf. A 3. EMILY GRACE KAZAKÉVICH 355 make the payment. This also follows from the whole context of A 5 (see n.64). I also find instructive the use of apophora and ἀποφέρω in several other texts of the fifth and fourth centuries, in which these words are used of payments exacted from a juridically free population. Tribute of this sort was, according to Herodotus, paid to the Egyptian king Sesostris by his sub- jects from the produce of land allotted to them (2.109). The Lepreans of Triphylia, half of whose land was at one time held forfeit to Elis, were allowed to remain in possession provided they made periodic payment (ἀποφέρειν) of one talent to Zeus Olympios (Thuc. 5.31.2–5). In criticizing Plato’s first ideal state Aristotle points to the folly of supposing that harmony can prevail among a citizenry split in two, the smaller part being an encampment of high-minded soldiers (φύλακες) destined to live in communal frugality, ever on watch against enemies without and within the city, the others being each in control of his own oikos and free to dispose of his income provided that a part of it go to maintain rulers he has had no say in choosing (Pol. 1264a11–b25). Aristotle calls the phulakes “garrison troops” (φρουροί), the maintenance tax apophora, and observes that such a levy imposed on citizens would surely arouse in them a spirit more rebellious than that displayed by the helots, penestai, and other subject peoples (δουλεῖαι) in the real world.26 Common to these three cases is the fact that apophora is paid from (apo-) means put or left at the disposal of the payers, to a person or collectivity recognized as master of those means or of 26 Pol. 1264a32–36. Plato’s Socrates calls this levy misthos, solider’s pay (in kind): Resp. 416DE, 464BC, cf. 419–420A. He perceives the danger of staseis among his bifurcated citizenry, one part of which will have a permanent monopoly of arms and office while the other will retain individual family and property rights on the normal oikos principle. To avoid the danger he resorts to the ideological weapon, proposing to tell his citizens “stories” which he hopes will in time come to be believed, and so help to knit his ideal community into One (414C–415D). In the Laws Plato retreats to a more familiar pattern. The citizens, who will do their soldiering jointly, are to be fed by a levy on the produce of land worked by slaves: γεωργίαι δὲ ἐκδεδομέναι δούλοις ἀπαρχὴν … ἀποτελοῦσιν (806DE). Though the term is different (ἀποτελοῦσιν), the economic relationship is essentially that indi- cated in my A-texts, but on a scale more consonant with helotry (cf. Plut. Mor. 239E). Aristotle takes no notice of this in his critique of the Laws. 356 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? the persons using them or of both. Leaving aside my A-texts, the term is semantically well suited to denoting income from slaves placed in a similar economic relationship. Yet M. S. Kutorga, who in his pioneering study of the different modes of exploiting slaves in the “Athenian Republic” discusses all three passages, would not accept this. He insists that apophora was used only of payments made by free persons, a position which led him to twist my A-texts into unlikely shapes (he did not live to see A 3). The analogous mode of exploiting slaves he associ- ates with the chôris oikountes (see n.47), and takes his term for the payment made by such slaves to their masters (ἀναφορά) from the lexicographer Ammonius.27 That for Aristotle the word apophora had “slavish” overtones is shown by the way he uses it to sharpen his criticism: directly, it denotes a payment made by free men, the citizens of the ideal state, but his main point is the analogy it suggests with servile conditions. For I do not think that his substitution of apophora for Plato’s misthos was a random variation. Consideration of the use of apophora to denote payments made by persons who were legally free tends to confirm the principle underlying my classification of the texts most often cited as pertinent to the chôris oikountes problem, namely my view that apophora was used to denote a form of slaveowner’s revenue distinct from misthos, associated with a quite different type of exploitation. Slaves who paid this levy and were other- wise left to organize their work as they pleased may indeed be classed as privileged slaves. But, if the ancients gave it a name at all, this category could surely not have been called andrapoda misthophorounta. Before discussing the other supposed name of the privileged- slave category, “the dwellers apart,” it will be useful to examine Westermann’s treatment of it. The bearing of the designation “dwellers apart” evidently presented no problem to Wester- mann, since he offers no argument to support his understand- ing of it. Consequently much of what he writes about this category remains undocumented or weakly documented by 27 Ammon. [Adv.voc. p.17 Nickau]; for Ammonius see also Dindorf’s edi- tion of Harpocration (II 95). EMILY GRACE KAZAKÉVICH 357 source references. Even in formulating his own views on the subject he often uses language so broadly general, and at the same time often cluttered with unanalyzed terminological ma- terial, that it is hard to be sure of his meaning. But one thing is clear: the “dwellers apart” are central to his conception of the condition of slaves in Athens. His allusions to them are scattered through different sections of his study on slavery in antiquity28 but, taken as a whole, do little to clarify his conception that the words chôris oikountes and andrapoda misthophorounta refer to the same category of slaves. The position is stated most fully on p.12 of his book Slave Sys- tems: According to the method of their employment slaves were distinguished as douloi or oiketai, who worked directly for their owner, and the andrapoda misthophorounta [his n.1 reads “Isaeus, Or. 8, 35”] or douloi misthophorountes who are to be identified with the chôris oikountes, those who lived apart from the owner’s res- idence, working at occupations of any sort and paying to their owners all or some percentage of their earnings. It should be stressed that, aside from the reference to Isaeus (my text M 7), no documentation whatever is offered for this ostensibly well-grounded definition. In the statement just quoted two categories are discernible, distinguished from one another by a purely formal criterion— the slave’s place of residence: 1. Slaves called douloi or oiketai,29 who worked directly for and apparently (such is the implication) lived with their masters. 2. Slaves called andrapododa misthophorounta (or douloi misthopho- 28 Westermann, Slave Systems 12, 16–17, 23, 38, 122. In the index (under chôris oikountes) only p.12 is cited and the term is given the enigmatic definition “general utility, non-tenant slaves.” Since the work was published after the author’s death oddities in the index should perhaps not be ascribed to him. The section on classical Greece is apparently an English version of Westermann’s “Sklaverei” article in RE; see Slave Systems ix. 29 There is no true terminological distinction here, of course. Slaves who lived apart from their owner were as likely to be called either douloi or oiketai —or andrapoda for that matter—as were slaves who lived under his roof, just as a house was an oikia whether or not its owner had let it. On slave-term distinctions see further n.39. 358 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? rountes) = chôris oikountes, who lived apart from their owners, turning over to them all or some part of their earnings. A closer look at the second category reveals that in Wester- mann’s view the “dwellers apart” embraced what are actually two sub-groups: (1) slaves let out for hire like any other type of property, and (2) slaves who worked for themselves, turning over to the slaveowner a fixed part of their income. Wester- mann himself makes nothing of this distinction, which I have “restored” from fragments in his rather jumbled second cat- egory. In the slaves who worked for themselves Westermann does not recognize a separate category—or rather, for him, this category tends to absorb all slaves who do not live with their masters. Here Westermann goes further than Schulthess, who also gives the same name (andrapoda misthophorounta) to this com- paratively privileged group of slaves and to rented-out slaves but draws a sharp distinction between the two groups in respect to “method of employment.” On pp.16–17, speaking of the legal position of slave emporoi and those “whose services were leased by their masters” [his n.104 reads “Greek: ἀνδράποδα μισθοφοροῦντα”], Wester- mann, in apparent reference to both groups, speaks of them (p. 17) as “slaves who lived apart.”30 The circumstance that slave lived separately from master is again elevated to a category criterion on p.23 of Slave Systems, where it is said of “the sep- arate-living slaves (chôris oikountes)” that they could marry and “found households,” although “nothing is definitely known” about their property rights. On p.38, speaking of the Greek law of slavery in the Hellenistic period, the author simply asserts, 30 If the economic rather than place-of-residence factor is acknowledged as the true criterion, the slave emporoi, who are here lumped together by Westermann with rented-out slaves, belong in fact to a quite different cat- egory. Though the nature of his activities required that the slave engaged in overseas trade live apart from his owner, he was working directly for his owner inasmuch as he represented his owner’s interests and was account- able to him. It is entirely possible, indeed probable, that the slave agent en- joyed various personal privileges as well as juridical competence, but his position differed, as the day from night, from the slave who, like any ox or horse, was rented out to a third party. See Kazakévich, “Slave Agents in Athens.” EMILY GRACE KAZAKÉVICH 359 without any qualifications, that “in the older Attic law” the property rights of slaves who “‘lived apart’” had been “fully recognized.” This is restated on p.122, where the property rights of the chôris oikountes in classical Athens are contrasted with the conditional rights of slaves over their peculium under Roman law. As we have seen, in the original definition of his second slave category on p.12, and again on pp.16–17, Westermann com- bines into one homogeneous group (“according to the method of their employment”), and endows with various rights and privileges, slaves who bring in misthoi and, generally, all slaves who “lived apart.” On pp.12 and 22 he underlines the eco- nomic significance of the “lease system” (i.e. the renting out of slaves) in “handicraft-industrial city-states” such as Athens. The source material cited on p.12 (nn.6–14) as showing the great variety of types of labor associated with the “lease system” in- cludes a group of cases which, with the exception of the first (the slaves rented out by Nikias and others to work in the silver mines), can hardly be assigned to the category of slaves working away from their masters on the lease system.31 Thus (if I have rightly understood him here) Westermann’s category andrapoda misthophorounta32 = hoi chôris oikointes, with all the rights and priv- 31 For example, Dem. 49.51–52: the μισθωτοὶ ἢ οἰκέται who would have brought the copper to Pasion’s bank, and also Pasion’s own oiketai, who would have received the copper if the others had brought it; the misthôtoi in this passage would not be slaves since they could not be “handed over” (for basanos) and there is no reason to suppose that the oiketai in the first instance would be rented-out slaves, while those of the second instance are obviously not. Cf. Dem. 27.9 and 36.11 (Westermann, Slave Systems 12 n.12, has a wrong reference), referring to workshops owned by Demosthenes Senior and Pasion. This series of source references immediately follows the author’s statement that in cities like Athens, Corinth, and Megara, the “rapid de- velopment of the custom of capital investment in slaves as instruments of production, earning money for their owners under the lease system, and the wide diversification of such labor, are clearly apparent” (12). The cases cited immediately after this statement are, taken together, evidently intended to exemplify both the diversification of slave labor (which they do) and the “lease system” (“such” labor), which they are very far from doing. 32 On p.22, perhaps only for variety’s sake (he does not say), Westermann calls these “pay-earning slaves,” misthophorounta sômata. 360 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? ileges assigned to it, embraces virtually all the slaves engaged in trade, transportation, and industry. Leaving aside the somewhat padded documentation of the “lease system,” we may reconstruct the course of Westermann’s reasoning as follows: since in the economically most-developed city-states the practice of renting out slaves was widespread (a premise it would be hard to quarrel with) and since all slaves living apart from their masters, including rented-out slaves, were granted a number of privileges (no proofs here), the con- clusion must be that in cities with the most highly developed slave economies the condition of the slaves was the pleas- antest.33 If the conclusion of this syllogism is not exactly con- vincing, its internal coherence must be granted. The trouble lies not in the argument but behind it, in the prior assumption that andrapoda misthophorounta and hoi chôris oikountes refer to one and the same category of privileged slaves, while for the first expression the meaning “rent-bearing slaves” is retained (see Slave Systems 16 with n.104). Westermann seems to have set his sights very high. He at- tempts nothing more or less than to soften the often harsh impression left by ancient comments on slavery in the Greek cities. For our immediate purposes, however, the most im- portant aspect of his conception is the weight he puts on the criterion of living separately. It was this which led him to make 33 See, e.g., p.22, where Westermann explains the “marked leniency” in the attitude towards slaves in Attica by the “relatively high development … of slaves as objects of capital investment, living in the semi-independent condition of the misthophorounta sômata.” On p.16, without documentation, he says that “the master in return for efficient service was accustomed to set aside a part of the earnings of the slave which might ultimately serve for the purchase of his liberty.” Cf. the “impression” gained by the same author from a combined study of the Attic “silver-bowl” inscriptions (see my n.46), the Delphic manumissions, and the wills of the philosophers, “that the ways of gliding over from slave status into free status in the busy handicraft center of Athens in the fourth century … were numerous and fairly easy”: “Two Studies in Athenian Manumission,” JNES 5 (1946) 103–104. In particular Westermann finds that the paramonê obligations laid on manumitted slaves facilitated this gliding progress toward freedom. No doubt it did for the manumittor, who was thereby freed from the need to maintain the slave while still exploiting him. EMILY GRACE KAZAKÉVICH 361 no distinction between wretches who were bought and then leased out wholesale for exploitation in the mines, on the one hand (see, besides Xen. Vect. 4.14, Hyp. Lyc. 1f., Eux. 35), and on the other the perfume-seller Midas with his pretty son and load of debts (Hyp. Ath., see Slave Systems 12 n.10). Any criticism of Westermann’s belief (which still has considerable influence) in the mildness shown their slaves by Athenian slaveowners34 must start from a reexamination of the assumption underlying this conception: that the expression “the dwellers apart” re- ferred to slaves. In the classical sources (and this also goes for the Hellenistic period), the expression οἱ χωρὶς οἰκοῦντες appears once only, in the First Philippic of Demosthenes, which is dated to 351 B.C. Demosthenes reproaches the Athenians for their apathy and unreadiness for waging war: In what pertains to war and its equipment, everything is ill- arranged, ill-managed, ill-defined. Consequently we wait till we have heard some piece of news, and then we appoint our ship- masters, and arrange suits for exchange of property, and go into committee of ways and means, and next we resolve that the fleet shall be manned by resident aliens and the chôris oikountes, then again by citizens, then by substitutes, then, while we thus delay, the object of our cruise is already lost.35 (Dem. 4.36–37) What can be deduced from this about the chôris oikountes? With certainty, only this: there was a group, in some way analogous to the metics, which was familiar to Demosthenes’ audience under the name hoi chôris oikountes. The same combination of words, here shown by the context to refer to a definite group, is found several times in other parts of the Demosthenic corpus, 34 The sentimental or moralistic approach to the history of slavery was alien to Westermann. See the comments in his preface to Slave Systems x; cf. 26 and criticism of Wallon [= H. Wallon, Histoire de l’esclavage dans l’antiquité I–III (Paris 1847)] on 152ff. He prefers to explain the “mildness” shown to slaves in Athens by economic factors rather than by the “greater humanity” of the slaveowners (22). In other words the “leniency” in the treatment of slaves came not from humaneness but from the very nature of a developed slave society. To my mind the one finding is as little credible as the other. 35 [Translations are from the Loeb Classical Library, with some modifica- tions.] 362 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? where, however, its reference is different. For example, in 43.19, in the context of an inheritance dis- pute, the speaker says: Bouselos … was a member of the deme Oeon, and to him were born five sons … And all these sons of Bouselos grew up to manhood, and their father Bouselos divided his property among them all fairly and equitably, as was fitting. And when they had divided the property among themselves (νειμάμενοι), each of them married a wife according to your laws, and sons and grandsons were born to them all, and there sprang up five households (οἶκοι) from the single one of Bouselos; and they dwelt apart, each having his own home36 and begetting descen- dants (καὶ χωρὶς ἕκαστος ᾤκει τὸν ἑαυτοῦ ἔχων καὶ ἐγγόνους ἑαυτοῦ ποιούμενος). In 47.34–35 the speaker tells how he was looking for Theo- phemos in order to get back from him certain public property; on meeting Theophemos’ brother Euergos, he asked him whether he had divided the estate with his brother, or whether their property was held in common. On Euergos’ an- swering me that it had been divided and that Thophemos lived in a house by himself, but that he (Euergos) lived with his father, I thus learned where Theophemos lived, and taking with me a servant from the magistrates, I went to Theophemos’ house (ἀποκρινομένου δέ μοι Εὐέργου, ὅτι νενεμημένος εἴη καὶ χωρὶς οἰκοίη ὁ Θεόφημος, αὐτὸς δὲ παρὰ τῷ πατρί, πυθόμενος οὗ ὤκει ὁ Θεόφημος, λαβὼν παρὰ τῆς ἀρχῆς ὑπηρέτην ἦλθον ἐπὶ τὴν οἰκίαν τοῦ Θεοφήμου). In both passages the condition χωρὶς οἰκεῖν involves a division of property. Although separate residence is plainly indicated, especially for Theomphemos in the second passage, the ex- pression may have in this context another meaning as well. In the fourth-century literary sources the word oikos was used in at least two senses. It might mean simply house, though perhaps oikia was more often used in this sense, as it is in the passage just quoted from Dem. 47, or it might mean family in the sense of the social unit whose function was to transmit property (also 36 One manuscript reading (A) is τὸν ἑαυτοῦ οἶκον ἔχων; the other mss. omit οἶκον (see Rennie’s apparatus in the OCT of 1931). EMILY GRACE KAZAKÉVICH 363 family cult obligations) from one generation to the next. In some cases (e.g. in misthôsis oikou) this second meaning narrows down to the sense “totality of heritable property,” or the whole property that a man owned (Xen. Oec. 1.5). At one time, prob- ably, the physical and social meanings had not been differ- entiated: in the so-called “patriarchal” type of household, be- longing to a family, which might embrace three or even four generations, went with one common place of residence. Con- versely, living separately amounted to breaking away from the family group. It may be that in the smaller family of later times the second meaning of oikos (transmitter of family property) is the one reflected in the expression χωρὶς οἰκεῖν, that this ex- pression represents a relic of customary law from an earlier epoch, when to separate oneself from the “family compound” meant not so much to found a new oikos, equal in status to the paternal one, as to incur loss of status in a more self-contained, monarchically ruled family group, which was then the basic social unit. The expression chôris oikein is also used in the Demosthenic corpus in connection with the separation of a freedwoman from the household of her former owner. In 47.68ff. the speak- er tells how he enquired of the exegetai (interpreters of ritual) what he should do about the killing of his old wet-nurse. They referred him to the law and from the stele bearing the law of Draco he confirmed what they had told him, that the killer (like the killer of a relative) could be prosecuted only by the close kin of the victim’s owner.37 In this case, however, “the woman was in no way related to me by blood, she had only been my nurse; nor again was she a slave; for she had been set free by my father, and she lived apart and had taken a husband” (καὶ χωρὶς ᾤκει καὶ ἄνδρα ἔσχεν, 47.72, cf. 70). The right (or duty) to pursue a killer belonged to any relative of the victim who came within the ἀγχιστεία, a circle of kin within a prescribed degree which could, and usually did, reach wider than the oikos- 37 [On Draco’s law, see, e.g., R. S. Stroud, Drakon’s Law on Homicide (Berkeley 1968); Grace, “Status Distinctions in the Draconian Law”; M. Gagarin, Drakon and Early Athenian Homicide Law (New Haven 1981).] 364 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? family.38 In this latter, smaller unit the slave was regarded as an adjunct of his owner and came under the homicide law through him. Once the speaker’s nurse had been set free her killer could no longer be prosecuted by her former master’s kin. Here, too, living apart involves separation from an oikos; but in this case, unlike those of Theophemos and the sons of Bouselos, who still had anchisteis, separation also meant removal from that protection of the homicide law which attachment to a citizen oikos had given. The social implications of the phrase chôris oikein, said of the freedwoman in Dem. 47, are also brought out in another pas- sage in the same speech: “After she had been given her free- dom she lived with her husband, but after his death, when she herself was an old woman and there was no one to care for her, she came back to me” (55). The statement in 47.72 that she had “lived apart” from the family of her former owner thus acquires a special significance: it is emphasized by the speaker as an essential element of proof that as a freedwoman she no longer belonged to her master’s oikos and had no legally recog- nized connection with his anchisteis. Her later residence in the speaker’s house did not change this situation. This story about the freedwoman has some relevance to the distinction drawn between the terms doulos and oiketês in a definition ascribed by Athenaeus to the third-century Stoic philosopher Chrysippus (Ath. 6.267B): “Chrysippus says … that a doulos differs from an oiketês in that apeleutheroi are still douloi, whereas those who have not been released from ownership are 38 On the anchisteia see, e.g., H. E. Seebohm, On the Structure of Greek Tribal Society (London 1895) 55, 75ff.; cf. G. Thomson, Studies in Ancient Greek Society: The Prehistoric Aegean (London 1954) 109ff.; [Harrison, Law of Athens I; L. Rubinstein, Adoption in IV. Century Athens (Copenhagen 1993); Todd, Shape of Athenian Law; C. B. Patterson, The Family in Greek History (Cambridge [Mass.] /London 1998)]; see also Dem. 43.57; IG I2 115 [= IG I3 104]. As a social unit, the oikos seems to be connected with the genesis of private property, while the anchisteia looks back to an earlier, pre-polis period. With the appearance of the classical oikos, the kinship principle may be said to have been “negated,” while continuing to manifest itself in new ways (e.g., the wife enters her husband’s oikos while remaining attached to her paternal genos), as family property becomes more exclusive. EMILY GRACE KAZAKÉVICH 365 oiketai (οἰκέτας δὲ τοὺς μὴ τῆς κτήσεως ἀφειμένους). ‘For’, says he, ‘the oiketês is a doulos appointed thereto by ownership (κτήσει κατατεταγμένος)’.” In my view this statement has more validity as a description of the de facto difference between the positions of slave and freedman than it has as an accurate definition of the slave terms in question, at least as they are used in courtroom language of the fourth century in Athens.39 It gen- eralizes from a distinction which in Dem. 47.72 is revealed indirectly in reference to a particular question of law. The definition ascribed to Chrysippus clearly qualifies the freedman as a person who has been separated from his former owner’s property. At the same time it makes allusion to the freedman’s continuing dependence on his former master, since, it would seem—at least outside the context of legal proof—the term which normally defines slave status might still be applied to a freedman.40 39 While in fourth-century Athens oiketês (female: therapaina) and doulos were both general terms for “slave,” and it would usually be misleading to translate them otherwise than by that word, doulos was first and foremost the status-defining term: see e.g. Dem. 49.55. What Chrysippus says about oiketês alone does agree with fourth-century Athenian usage, in which this term appears chiefly in reference to the slave as a person bound to some- one’s oikos (i.e., κτῆσις). The third general term, andrapodon, typically refers to the slave regarded as embodiment of exchange value, whether the context is one of war booty or, say, the assets of an estate (e.g. M 7); sometimes this term refers more specifically to the slave as live agricultural inventory, along with cattle. See my slave-term articles: Kazakévich, “The Term δοῦλος” and “Slaves as a Form of Wealth.” F. Gschnitzer came to similar conclu- sions, especially in regard to doulos and andrapodon, though he did not know of my work till his was in press: Studien zur griechischen Terminologie I (Wies- baden 1963) 25f., 26 n.2. [On the Greek terminology for slavery, see also M. I. Finley, “The Servile Statuses of Ancient Greece,” RIDA 7 (1960) 165– 189; F. Gschnitzer, Studien zur griechischen Terminologie II (Wiesbaden 1976); Garlan, Slavery 19–22; Klees, Herren ch. 1; Zelnick-Abramovitz, Not Wholly Free 27–39.] The relevance of Chrysippus’s definition to the actual condition of freedmen was noted by C. B. Gulick in the Loeb edition of Athenaeus III 200 n.a; cf. L. Gernet, Droit 169 n.1. 40 [For this same idea, see Todd, Shape of Athenian Law 193.] After manu- mission a slave was often still bound to his owner by so-called paramonê obligations, which might continue in force till death. See Calderini 1908, 277–288; [A. E. Samuel, “The Role of Paramone Clauses in Ancient Documents,” JJP 15 (1965) 221–311; K. Hopkins, Conquerers and Slaves 366 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? Though chôris oikein seems to have had in certain contexts a more “technical” meaning than simply “take up separate resi- dence,” there is no ground for seeing in these words references to the name of a social category. In this semi-technical meaning they might be used, as we have seen, either of free men, even citizens, or of freedmen. Their use of particular persons stand- ing in certain defined relationships to private-law institutions (the family or its property) gives them a random, transitory quality from the standpoint of the community as a whole. Therefore it seems a priori most unlikely that such persons were regarded as composing a stable, homogenous group, clas- sifiable under the name “the dwellers apart,” as Beauchet seems to imply: “Les frères qui avaient procédé au partage de la suc- cession paternelle et qui, par suite, vivaient séparément, s’appe- laient χωρὶς οἰκοῦντες.”41 If it was said of such brothers that they chôris ôikoun, this hardly means that they were known as hoi chôris oikountes. Yet the special sense in which, as Beauchet ob- served, the expression chôris oikein was used in Athens of certain individuals can help us understand what sort of group was called by this name in category form. We now return to the discussion of that question. From the second century A.D. we have Harpocration’s com- mentary on Dem. 4.36—which, it should be remembered, is the only surviving contemporary testimony to the existence in the classical period of a group known as the “dwellers apart”— and also, from a still later epoch, two mutually exclusive defi- nitions of the term, the first of which clearly stems from Harpocration, by the compiler of the Lexeis rhêtorikai. The exe- getic manner in which Harpocration quotes from Demosthenes ___ (Cambridge 1978) ch. 3; Klees, Sklavenleben 325–330; Zelnick-Abramovitz, Not Wholly Free 222–248]. See also Harp. s.v. ἀποστασίου; cf. Pl. Leg. 915A. According to Westermann, “Two Studies,” the imposition of paramonê obligations eased the way of a slave towards freedom; but Morrow’s com- ment seems likely to be closer to the truth: “according to the common Greek notion … the difference between slave and free status was so sharply marked that only in rare cases could emancipation obliterate it” (Plato’s Law of Slavery 105). 41 Beauchet, Histoire III 642 n.2. Elsewhere (see my n.12) Beauchet ap- plies the same group designation to a category of slaves. EMILY GRACE KAZAKÉVICH 367 at once suggests that he is guessing:42 “Demosthenes in the Philippics: ‘and after that the decision to call up to the ships the metics and those who live apart from their masters’. However, the meaning would be plain without this addition, because the freedmen lived on their own [or “by themselves”] apart from those who set them free, whereas while they were slaves still they lived together (with their masters)”—καὶ τοὺς χωρὶς οἰκοῦντας τῶν δεσποτῶν. οὐ μὴν ἀλλὰ καὶ χωρὶς τοῦ προσκεῖσθαι φανερὸν ἂν εἴη τὸ δηλούμενον, ὅτι οἱ ἀπελεύθε- ροι καθ’ αὑτοὺς ᾤκουν χωρὶς τῶν ἀπελευθερωσάντων, ἐν δὲ τῷ τέως δουλεύοντες ἔτι συνῴκουν.43 Perhaps Harpocration’s definition was influenced by ac- quaintance with the story about the freedwoman in Dem. 47 who “lived apart” from her former master. But then the fact that this woman later returned to live in her former master’s house, without incurring any change of her “separated” con- dition, would conflict with Harpocration’s evidently “geo- graphical” interpretation of the concept “live apart” in the 42 Cf. Kutorga, Social Position 220, 232. 43 Harp. s.v. τοὺς χωρὶς οἰκοῦντας (p.291.12–15 Dind.); from our texts of the First Philippic τῶν δεσποτῶν is missing. It is worth noting that Har- pocration uses χωρὶς οἰκεῖν and οἰκεῖν καθ’ αὑτόν as synonyms, “defining” the latter by the word χωρίς in the next phrase (line 14). Cf. comm. by Valesius in vol. II of Dindorf’s ed. of Harp. (p.291), where both expressions are translated by seorsum (habitare). In Dem. 49.10 ᾤκει καθ’ αὑτόν is said of someone who, preferring not to marry, agreed to leave the family property undivided and went to live “by himself” in Salamis. This purely physical separation left the paternal oikos undisturbed. In Dem. 36.4 ἤδη καθ’ ἑαυτὸν ὄντι, referring to Phormion, to whom Pasion had leased the business (6) of his bank and shield workshop (ca. 371 B.C.), means in my opinion no more than that at the time of the lease Phormion was in business “on his own,” no longer a clerk in Pasion’s bank, and not that he had been manumitted: see F. A. Paley and J. E. Sandys, Select Private Orations of Demosthenes II (Cam- bridge 1886) 7; L. Gernet, Démosthène: Plaidoyers civils I (Paris 1954) 207, “alors qu’il était déjà libre,” cf. 200, “déjà affranchi”; [Cohen, The Athenian Nation 134 n.25, “noting explicitly that Phormiôn had already obtained his freedom”]. Sandys straddles a bit: “‘Doing business on his own account, as his own master’, no longer subject, as a slave, to the control of another”; but this position would be equally compatible with his being a slave ἐπ’ ἀπο- φορᾷ. My point is not that Phormion was a slave at that time, only that καθ’ ἑαυτὸν ὄντι does not define his status. 368 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? category term hoi chôris oikountes, which would cast doubt on the correctness of his definition. If in fact Harpocration was taking into account not only the passage in Dem. 4.36 but also how the words chôris oikein are used elsewhere in the Demosthenic corpus, it must be supposed that he ignored the “group” of sons who took their shares and “lived apart” because of the extreme unlikelihood that these people formed a group at all, much less a group appropriate to the context of the passage in the First Philippic. Indeed persons so described could hardly have formed a special class, distinct from citizens generally, of citizens liable to call-up for the navy. But of course it must be borne in mind that many more literary sources were available to Harpocration than are to us. It would be arbitrary to exclude the possibility that his interpretation of the phrase in this passage was based on other, more apt exam- ples in texts now lost. The fact that he cites no parallels, which he often does in defining other terms, is noteworthy; but it can- not be taken as proof that he knew of none. There are, however, other reasons for finding Harpocration’s identification of the “dwellers apart” with freedmen uncon- vincing. First of all, if Demosthenes had freedmen in mind, why should he allude to them in so roundabout a way when he had at hand the far more precise term ἀπελεύθερος? Even supposing what was very likely the normal case, that freedmen actually did live apart from their former owners, would this be the aspect of their condition which distinguished them from the rest of the population in the eyes of the citizens addressed by Demosthenes? Was not their distinguishing feature rather the fact that they had once been slaves and were now freed from that condition—which is just what is conveyed by the term apeleutheros? If we are to suppose that freedmen were also termed “dwellers apart,” should we not perhaps expect to find slaves referred to as hoi sunoikountes? Then, it will have been noticed that Harpocration, in quoting the passage from the First Philippic, “emended” the text by inserting δεσποτῶν. That this was added by Harpocra- tion, that he did not find it in his text (as supposed, e.g., by J. E. EMILY GRACE KAZAKÉVICH 369 Sandys),44 is made sufficiently clear by the word προσκεῖσθαι and also by consideration of Attic usage in the classical period. The word despotês would, properly speaking, be out of place in this context. In a context relating to slavery it is regularly used of a slave’s master (owner) rather than of a freedman’s manu- mittor. It is not likely that Demosthenes himself would have accompanied a reference to the category of freedmen by so “free” a use of the term despotês45 (see his strictures on accuracy in the use of terms in 20.29). The fact that Harpocration felt the “addition” to be necessary, despite his own denial that it was really needed, suggests that he was aware of some in- adequacy in hoi chôris oikountes as a name for freedmen. In gen- eral, his interpretation of this name has an air of justification— as though he were defending a dubious conjecture—quite unlike his assured and circumstantial treatment of, for instance, the juridical term δίκη ἀποστασίου.46 It is my belief that Harpocration was hazarding a guess as to the meaning of hoi chôris oikountes. Moreover it seems quite possible that it was this rather awkward interpolation of τῶν δεσποτῶν which inspired the hesitant double definition of the same expression in Lexeis rhêtorikai: οἱ ἀπελεύθεροι, ἐπεὶ χωρὶς οἰκοῦσι τῶν ἀπελευθερωσάντων ἢ δοῦλοι χωρὶς οἰκοῦντες τῶν δεσποτῶν (“freedmen, since they live apart from those who set them free, or slaves living apart from their masters”: Bekker, Anec. I 316.11). In the first part of this definition Harpocration 44 J. E. Sandys, First Philippic and the Olynthiacs of Demosthenes (London 1910) 113. 45 One cannot of course entirely exclude the possibility that Harpocration had an emended copy or glossed copy of the speech. A “free” use of despotês might naturally occur in an emotionally charged recollection of one’s own slave condition (see, e.g., Dem. 24.124, cf. Isoc. 6.96) [see also Klees, Sklavenleben 338, on this use of despotês]. The strict use appears in Pl. Leg. 915A: τῷ γενομένῳ δεσπότῃ. 46 In Slave Systems Westermann makes no mention of Harpocration’s definition of this procedure (s.v. ἀποστασίου), by which a manumitted slave might be returned to slave status for not fulfilling “legal” obligations. Westermann translates the words δίκαι ἀποστασίου as “judgments for re- lease,” in accord with his interpretation of the silver-bowl inscriptions (IG II2 1553–1578) as recording the results of “simulated trials” (Slave Systems 25f.). 370 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? is plainly recognizable, even to the “justificatory” air. M. S. Kutorga stressed the contradictory nature of the defini- tion: it was unthinkable that the Athenians would apply the same term to two different social categories, each of which had its own proper designation. He takes Boeckh to task for appar- ently accepting it whole as a true definition of hoi chôris oikountes, and it would be hard not to take his point. However, Kutorga’s own recommendation—to scrap the first meaning in favor of the second, thus in effect anticipating the modern consensus— seems equally unsatisfactory.47 But then it is no easier to accept the position taken by Busolt in the third edition of his hand- book, namely that the second definition should be rejected in favor of the first; or Gilbert’s acceptance of the first meaning for Dem. 4.36 and the second as good for other (hypothetical) applications of the same expression to slaves (see my nn.12 and 17). Busolt’s suggestion that the name “dwellers apart” denoted a subgroup among freedmen, namely “Freigelassenen, die ihren eigenen, vom Freilasser getrennten Wohnsitz hatten,”48 looks like an attempt to overcome the terminological difficulties posed by Harpocration’s definition in light of the story about the freedwoman in Dem. 47. There are several objections to 47 Kutorga, Social Position 220f., 233f. (citing Boeckh, Staatshaushaltung2 I 365). The grounds for his choice are the uncertain tone of Harpocration’s definition, the practice of admitting freedmen to metic status, and a hypothetical superior source for the second definition. In respect to source criticism, Kutorga was rather old-fashioned for his time (his professional life extended from 1836 to his death in 1886), but he was well ahead of his time in undertaking a full-length enquiry into the socio-economic stratification of the slave population in classical Athens. His thinking on this subject was ob- viously influenced by his familiarity with the institutions of serfdom in pre- Reform Russia. But this familiarity also enriched, without sentimentalizing, the sympathetic imagination he brought to the study of ancient forms of servitude and did not in the least disturb his admiration for the political achievements of the citizen community, among which he particularly stresses (in a letter to a friend, see Social Position vii–viii) the working out of two great ideas, “the idea of the freedom of the citizen and the idea of the freedom of thought.” 48 Busolt, Grieschische Staatskunde I 274 n.3, 985; cf. Lipsius, Attische Recht 798; Schulthess (s.v. μισθοφοροῦντες). EMILY GRACE KAZAKÉVICH 371 this solution. First as to the woman, it was not her going to live apart from her master but the act of manumission which broke her tie with his household and kin. Then, the freedman who did not stay in his former master’s house was as likely to be bound to him by service obligations as the freedman who did. Where a freedman lived would not, I should think, affect his availability for service in the fleet one way or the other. What would affect it, and directly, would be his registration as a metic—and metics have their own place in the passage from the First Philippic. Finally, Busolt takes no notice in this con- nection of the parallel between chôris ôikei as used of this freed- woman after her manumission and its use of citizens who “separate” themselves, by withdrawing their share, from the paternal oikos. In both cases the physical separation is a natural but not necessary consequence of a more significant change. In neither case can it be sensibly connected with the formation of a category of reserves for call-up to the fleet. In my opinion the simplest approach to the double definition problem posed by the Byzantine lexicographer is, in the cir- cumstances, likely to be the soundest: since everybody knew that freedmen usually and slaves sometimes lived “apart” it seemed to this lexicographer that the term might have referred to either of these categories. But since he did not know for sure which, he presented alternative explanations: “freedmen—or slaves,” transferring Harpocration’s τῶν δεσποτῶν to his own second definition. More significant for us than the question as to why these lexicographers interpreted hoi chôris oikountes as they did is the fairly plain fact that neither of them was sure what it meant. And so we are again left with the passage in Demosthenes’ First Philippic as the only firm ancient testimony regarding the group as “the dwellers apart.” As I have already argued above in connection with the lexi- cographers’ explanations of the term in Dem. 4.36, the words hoi chôris oikountes could hardly have been used to designate the freedmen, or even some part of this category (as Busolt pro- posed). Could such an expression have been used of any group among the slave population of Attica in the fourth century? Undoubtedly there were among this population slaves who 372 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? lived apart from their owners. Slaves rented out for work in the silver mines are an example.49 Even slaves who worked under the supervision of their owners sometimes lived apart from them. Arizelos (father of the Timarchos who figures in A 1), for example, once owned two workshops in the mines.50 The slaves who processed silver ore in these ergasteria, like the thirty slaves of Pantainetos in Dem. 37, were presumably lodged in some kind of barracks near their place of work.51 Of these slaves it could also be said that they “lived apart.” As regards the slaves who worked independently, turning over to their owners payment in the form of apophora, the meagerness of the evidence we have about them does not show conclusively that they always lived apart from their owners. Diokleides’ slave (A 4) probably did so, since his owner sets out to Laurium to fetch his apophora. On the other hand there is nothing to show that the nine or ten leather workers of Ti- marchos (A 1) did not exercise their craft in the immediate vicinity of their master’s house if not within its walls. They were not so very many. Demosthenes Senior’s house had room for two workshops manned by a total of more than 50 slaves; Lysias’ shield factory in Peiraeus, employing some 100 slave craftsmen, was in the same compound where the owner lived.52 Menander’s language in A 3 suggests that Syriskos and his wife often spent nights in their master’s house, if they did not “live” there. There are no grounds for concluding that the right to run an independent enterprise necessarily implied the residen- 49 P. Guiraud, La main-d’oeuvre industrielle dans l’ancienne Grèce (Paris 1900) 133; see also Lauffer, Bergwerkssklaven Ι 59 n.4. 50 Aeschin. 1.101. That these workshops were manned by slaves is shown in 105. 51 See Lauffer, Bergswerkssklaven Ι 61f. 52 Lys. 12.8, 12, 19. In 19, 120 andrapoda are mentioned as part of the property of Lysias and Polemarchos which was seized by the Thirty. But we cannot be sure that all these slaves were craftsmen. Such a rich estab- lishment must have maintained a large staff of domestics. This was noticed by G. E. M. de Ste. Croix in “Slavery: Review of W. L. Westermann’s The Slave Systems of Greek and Roman Antiquity,” CR 7 (1957) 55f., though without reference to the term andrapodon, on which general term for “slave” see Kazakévich, “Slaves as a Form of Wealth” [and my n.39]. EMILY GRACE KAZAKÉVICH 373 tial separation of slave from master. Thus at least some of the slaves who, according to the mode of their exploitation, belonged to any one of three categories— slaves working under the direct supervision of their owners, slaves rented out for exploitation by other people, and slaves permitted by their owners to go into business on their own account—could be classed as chôris oikountes in the purely formal sense in which both Harpocration and Westermann take these words. On the other hand, some slaves belonging to the same three categories lived in their owner’s house. Evidently the geographical location (place of residence) and the “method of employment” (Westermann’s phrase) by no means always co- incided. The place where the slave lived probably depended on the nature of his work; it certainly did not depend on the economic relations obtaining between slave and master. Setting aside the economic relationship, could Demosthenes have had in mind slaves regarded as a group simply because they did not reside with their masters? The answer is clearly no. Such slaves were neither more nor less available for call-up than slaves who lived in their master’s house. If the demos could (or wanted to) recruit for service in the fleet slaves who lived apart from their owners, nothing could stop the demos from also calling up slaves who lived with their owners. Either measure would encroach on the private interests of the slave owners, who comprised a large section of the demos itself. The same factor works against the thesis that hoi chôroi oikountes designated slaves who were allowed to conduct their own enterprises and so might be regarded as being in some degree separated from the master’s household economy. These slaves were as much articles of property as were slaves directly exploited by the slaveowner or rented out by him to someone else. The truth of this statement is confirmed by one indisput- able piece of evidence: Timarchos sold the leather workers who had been paying him apophora.53 In other words these relatively 53 Aeschin. 1.99. We have only Aeschines’ interested assertion for this, but for present purposes that is enough, for it means that such a sale was possible. Schulthess (s.v. μισθοφοροῦντες) stresses that the right to conduct his own enterprise was granted to a slave on the private initiative of his 374 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? privileged slaves were also κτήσει κατατεταγμένοι, that is, they were component parts of their master’s oikos. It seems, then, out of the question, or at least in the highest degree unlikely, that in 4.36 Demosthenes was talking either about freedmen (who would probably be classed as metics anyway,54 if not as apeleutheroi), or about any group of slaves. Any possibility that by chôris oikountes he meant some group of citizens is, besides the considerations already discussed, ex- cluded by his clear reference to citizens, as such, immediately after the mention of chôris oikountes: “then to go aboard our- selves (αὐτούς).” It can only be supposed that the “dwellers apart” were non-citizens. As noted at the start, the context in which the phrase hoi chôris oikountes occurs in this passage, in- expressive though it may appear in some respects, still suggests some group analogous to the metics, who are mentioned in the same breath as also liable to service in the fleet. The logical conclusion is that the “dwellers apart,” since they cannot have been freedmen or slaves, were some category of non-Athenians who were not made to become metics. I cannot undertake here a proper investigation of sources pertaining to the status of persons residing in Attica who were not registered as metics and were not slaves.55 I will confine myself to pointing out those considerations that incline me to think that Demosthenes could have been referring to any or all of the various free populations groups in Attica and nearby areas, namely those groups who were neither Athenians nor ___ owner and that in all other respects the position of a slave who enjoyed this right was the same as that of any other slave. 54 Pollux (3.55) defines as a metic anyone who pays the metoikion; cf. Dem. 25.57; Harpocration (s.v. μετοίκιον) says freedmen also paid this tax. See [Clerc, Métèques 282; Beauchet, Histoire II 481 and passim; G. Foucart, De libertorum conditione apud Athenienses (Paris 1896) 50]; A. M. Andreades, History of Greek Public Finance I (Cambridge [Mass.] 1933) 277f.; [Whitehead, Ideology 115 and 116; D. M. MacDowell, The Law in Classical Athens (London 1978) 82; Garlan, Slavery 80; Cohen, Athenian Economy 109–110; Cohen, Athenian Nation 150; cf. Zelnick-Abramovitz, Not Wholly Free 308–319, who does not equate the two groups.] 55 [Kazakévich’s own translation stops here. What follows is Levaniouk’s translation.] EMILY GRACE KAZAKÉVICH 375 metoikoi—either because they had not yet acquired the status of metoikoi or because their relationship with the Athenians was de- termined by some other, more or less permanent, ties.56 From the fact that in 406 B.C. non-citizens were enlisted into the navy in large numbers (see especially Diod. 13.97.1)57 it is possible to conclude that such populations were of significant size. The majority of them could have been professional sailors, a per- manent reserve for the navy of what was still one of the greatest sea powers in the Aegean. How can we understand the expression hoi chôris oikountes as a designation of such groups? The answer is suggested by the name of a group which is in some sense analogous yet at the same time opposed to them, namely the group of metoikoi. Ap- plied to members of the latter group, the words μέτοικος and μετοικέω are commonly understood as designating people who changed where they lived in order to come and live in Athens. And indeed the prefix meta- most often carries the idea of change, of departure from a place, of moving from something to something. But this prefix can also convey the meaning of “together with” (someone or something) with the connotation of “participation in” (something) (μετέχω, μεταδίδωμι). The Athenians celebrated a festival called Συνοίκια in honor of the political unification of Attica, the synoecism ascribed to Theseus (Thuc. 2.15.1, schol. Ar. Pax 1019); Plutarch calls the same festival Μετοίκια (Thes. 24). Is it not possible that this was its older name? In any case, the ideological content of both words is evidently the notion of a unity deeper and more or- ganic than a simple coexistence in the same place. And indeed such coexistence was actually absent in Attica until that time when the invasion of Archidamos forced those living in the χώρα to seek refuge inside the Long Walls. It seems possible to 56 For example, such semi-subjugated neighbors of the Athenians as the inhabitants of Oropos, who were in that position, it seems, for much of the fifth century. See Thuc. 2.23.3 and comments in A. W. Gomme, A Historical Commentary on Thucydides II (Oxford 1954) 80; cf. Diod. 15.76.1. But this question, of course, needs separate study. 57 [On slaves in Greek warfare, including their role in the navy, see P. Hunt, Slaves, Warfare and Ideology in the Greek Historians (Cambridge 1998).] 376 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? me that the notion of “participation” also entered into the semantic content of the word μέτοικος. As we know, the metoikoi took part in many aspects of Athen- ian social and economic life. They were obliged to perform regular service in the army; those of them who met certain financial qualifications had to pay the eisphora tax and perform regular public works (λειτουργίαι) (Dem. 20.18, 20); they took part in religious festivals. Thucydides had Nikias address, be- fore the last sea battle of the Sicilian expedition, the non- Athenians among the ναῦται as “participants” and “the only partners in our archê,” using the words μετείχετε and κοινωνοί. Since on the same occasion (7.63.3–4) Nikias characterizes the allies as “subjects” the scholiast was probably right in remark- ing that the sailors Nikias addresses with his flattering exhorta- tions were metoikoi.58 This was, to be sure, an appeal for help in a desperate situation, and it cannot be taken as an everyday Athenian opinion regarding the position of those whom De- mosthenes calls “the wretched metoikoi” (24.166). As Aristotle indicates, the participation of the metoikoi in the benefits of social institutions was far from complete.59 Without getting into a comparison of the misfortunes the metoikoi faced with the advantages they received from being in Athens, it is possible to say that in a limited but still real sense they were members of the Athenian community. Could not Demosthenes, when he mentioned those “living apart” along with those “living together,” i.e. the metoikoi, have had in mind those non-citizens who were left for some reason formally out- side the Athenian community, in contrast to those (the metoikoi) who were included in the common “housekeeping” (in which case, it does not matter how incomplete this inclusion was)? Then why did he not simply say τοὺς μετοίκους καὶ τοὺς 58 This was the opinion of J. Classen, Thukydides2 (Berlin 1884), but not of J. Grote, History of Greece4 VI (London 1872) 154 n.1 (ch. 60). Like Grote, Gomme, A Historical Commentary on Thucydides I (Oxford 1945) 460, it seems, did not notice that in the same passage of Thucydides the allies are called ὑπήκοοι. 59 Arist. Pol. 3.1.3, 1275a. [See, e.g., Clerc, Métèques]; E. Weiss, Griechisches Privatrecht I (Leipzig 1923) 177ff.; [Whitehead, Ideology; Todd, Shape of Athen- ian Law 194–199.] EMILY GRACE KAZAKÉVICH 377 (ἄλλους) ξένους? Possibly because he had in mind not every stranger (for example, he certainly did not mean travelers), and not only the mercenaries, but precisely all the “excluded” in- habitant foreigners, a group that could include more than just the professional sailors. The first task of this study was to test the suggestion that the term hoi chôris oikountes designated slaves, and in particular a category of privileged slaves. As we have seen, such an in- terpretation of the term, especially when it is treated as synon- ymous with andrapoda misthophorounta, seriously distorts the status of a significant part of Attica’s slaves, namely those slaves who were rented out and therefore placed under double exploita- tion. If this latter category of slaves is meant in all those texts where the form of payment is designated by the word misthos, then very little evidence remains in our sources for a group of relatively privileged slaves who were allowed to work indepen- dently and pay an apophora. Therefore, along with Busolt and Schulthess (see my n.20) we can conclude that such privileged slaves were extremely few in number in comparison with the main population of Attica’s slaves. This conclusion is opposed, in essence, only by Westermann, who based his opinion not only on equating hoi chôris oikountes with andrapoda misthophorounta, but also on the most literal “territorial” and broad interpreta- tion of the former term. The suggestion that the number of slaves who paid an apo- phora was relatively small in turn strengthens the conclusion reached earlier on the basis of different considerations: namely that the expression hoi chôris oikountes should refer to some non- slave population group. For in Demosthenes’ usage this term implies some group sizeable enough to be able to serve as a constant reserve for navy recruitment. APPENDIX Testimony of the sources regarding two forms of income received from slaves. “M” Texts (where the type of payment is μισθός) M 1. Dem. 27.20: ἔτι δὲ Θηριππίδῃ τριῶν ἀνδραπόδων, ἃ ἦν αὐτῷ ἐν τῷ ἐμῷ ἐργαστηρίῳ, μισθὸν ἀποδεδωκέναι λογίζεται. (Further- more, he charges me with money which he has paid to Therip- 378 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? pides for the hire [misthon] of three slaves of his who were in my factory.) M 2. Dem. 28.12: δύ’ ἔτη τὸ ἐργαστήριον διοικήσας Θηριππίδῃ μὲν ἀποδέδωκε τὴν μίσθωσιν. (For two years he conducted the busi- ness of the factory and paid to Therippides the hire [misthôsin] of two slaves.) M 3. Dem. 53.20: παρ’ οἷς τοίνυν ἠργάσατο πώποτε, ὡς τοὺς μισθοὺς Ἀρεθούσιος ἐκομίζετο ὑπὲρ60 [v. l. τοὺς ὑπὲρ] αὐτοῦ, καὶ δίκας ἐλάμβανε, καὶ ἐδίδου, ὁπότε κακόν τι ἐργάσαιτο, ὡς δεσπό- της ὤν, τούτων ὑμῖν μάρτυρας παρέξομαι. (I shall also bring before you witnesses to prove that Arethousios got the wages [misthous] on his account from all the persons with whom Kerdon ever worked, and that he used, as being his master, to receive compensation or give it, whenever Kerdon wrought any damage.) This is part of the proof that Arethousios is the owner of the slave in question; see Kazakévich, “The First Forms of Capital.” M 4. Thphr. Char. 30.17: συναποδημῶν δὲ μετὰ γνωρίμων χρήσα- σθαι τοῖς ἐκείνων παισί, τὸν δὲ ἑαυτοῦ ἔξω μισθῶσαι καὶ μὴ ἀνα- φέρειν εἰς τὸ κοινὸν τὸν μισθόν. ([He is the type of person who,] traveling with acquaintances, uses their slaves, and hires out [misthôsai] his own without sharing the proceeds [misthon].) M 5. Xen. Vect. 4.14: Νικίας ποτὲ ὁ Νικηράτου ἐκτήσατο ἐν τοῖς ἀργυρείοις χιλίους ἀνθρώπους οὓς ἐκεῖνος Σωσίᾳ τῷ Θρᾳκὶ ἐξεμί- σθωσεν, ἐφ’ ᾧ ὀβολὸν μὲν ἀτελῆ ἑκάστου τῆς ἡμέρας ἀποδιδόναι, τὸν δ’ ἀριθμὸν ἴσους ἀεὶ παρέχειν [MSS. παρεῖχεν]. (Nikias, son of Nikeratos, once owned a thousand men in the mines, and let them out [exemisthôsen] to Sosias the Thracian, on condition that Sosias paid him an obol a day per man net and filled all vacancies as they occurred.) M 6. Ps.-Xen. Ath.Pol. 1.17: πρὸς δὲ τούτοις ὁ δῆμος τῶν Ἀθηναίων τάδε κερδαίνει τῶν δικῶν Ἀθήνησιν οὐσῶν τοῖς συμμάχοις. πρῶ- τον μὲν γὰρ ἡ ἑκατοστὴ τῇ πόλει πλείων ἡ ἐν Πειραιεῖ· ἔπειτα εἴ τῳ συνοικία ἐστὶν ἄμεινον πράττει· ἔπειτα εἴ τῳ ζεῦγός ἐστιν ἢ ἀν- δράποδον μισθοφοροῦν.61 (In addition, the people at Athens profit in the following ways when trials involving allies are held in 60 Since the interpretation of this passage is debated, it is worth noting that the preposition ὑπέρ, apart from its regular meaning “for” in the sense of “for the sake of,” “for the benefit of” (someone or something), can also mean “for” in the sense of “in return for,” “in exchange for” (something). See also Hyp. Lyc.1f.; Dem. 13.8. 61 The word μισθοφοροῦν, even though it is in the singular, can, according to the usual Greek usage, refer both to ἀνδράποδον and to ζεῦγος. In this context such an interpretation seems to me to be the most likely. EMILY GRACE KAZAKÉVICH 379 Athens: first the one-percent tax in the Peiraeus brings in more for the city; secondly, if anyone has lodgings to rent, he does better, and so does anyone who lets out on hire [misphoroun] a team of animals or a slave.) M 7. Is. 8.35: Κίρων … ἐκέκτητο … οἰκίας δ’ ἐν ἄστει δύο, τὴν μὲν μίαν μισθοφοροῦσαν … ἔτι δὲ ἀνδράποδα μισθοφοροῦντα … (Kiron … possessed … two houses in the city, one let to a tenant [misthophorousan] … he also had slaves earning wages.) M 8. Dem. 53.21:62 ὁπότε … οἱ ἄνθρωποι οὗτοι ἢ ὀπώραν πρίαιντο ἢ θέρος μισθοῖντο ἐκθερίσαι ἢ ἄλλο τι τῶν περὶ γεωργίαν ἔργων ἀναιροῖντο, Ἀρεθούσιος ἦν ὁ ὠνούμενος καὶ μισθούμενος ὑπὲρ αὐτῶν. (Whenever they [the two slaves] bought up the produce of an orchard, or hired themselves out [or, hired others as laborers] [misthointo] to reap a harvest, or undertook any other piece of farm- ing work, it was Arethousios who made the purchase or paid the wages [misthoumenos] on their behalf.) The context is the same as in M 3. “A” Texts (where the type of payment is ἀποφορά) A 1. Aeschin. 1.97: τούτῳ γὰρ κατέλιπεν ὁ πατὴρ … οἰκίαν μέν … ἐσχατιὰν δέ … χωρὶς δὲ οἰκέτας δημιουργοὺς τῆς σκυτοτομικῆς τέχνης ἐννέα ἢ δέκα, ὧν ἔκαστος τούτῳ δύ’ ὀβολοὺς ἀποφορὰν ἔφερε [v. l. δύ’ ὀβολοὺς ἀπέφερε] τῆς ἡμέρας, ὁ δ’ ἡγεμὼν, τοῦ ἐργαστηρίου τριώβολον. (His father left him a fortune … There was a house … a suburban estate … and besides there were nine or ten slaves who were skilled shoemakers, each of whom paid him a fee [apophoran or apephere] of two obols a day, and the super- intendent of the shop three obols.) 62 This text is included in my series only because several modern scholars have cited it as an example of rented-out slaves: see Boeckh, Staatshaushaltung 3 91 n.a; Schulthess (s.v. μίσθωσις); Partsch, Griechisches Bürgschaftsrecht 135 n.8. A similar sense is given to μισθοῖντο by Sandys, Select Private Orations, and by A. T. Murray (Loeb 1939). However, the verb μισθόω in the middle voice as a rule (in the perfect middle it can have a passive meaning) is used only transitively: “hire,” “rent”—not “hire oneself,” “be hired,” or “be rented.” At least in the Demosthenic corpus there are no exceptions to his rule. As for its use in other sources, LSJ bears out my under- standing of μισθόομαι. Therefore if in M 8 persons left unmentioned (who might be free persons: see e.g. Dem. 18.51, 57.45; the first passage shows well the meaning of μισθόομαι; cf. Aristoph. Av. 1152) are understood and taken as the direct object of μισθοῖντο, then the slaves who are the subject of this verb were not rented out but, on the contrary, themselves did the renting (hiring) of other persons. If θέρος is taken as the direct object of μισθοῖντο (a possible but to my mind less attractive interpre- tation), it would have to be accepted that this text says nothing at all about hiring people. 380 WERE THE ΧΩΡΙΣ ΟΙΚΟΥΝΤΕΣ SLAVES? A 2. Thphr. Char. 30.15: καὶ παρὰ παιδὸς κομιζόμνεος ἀποφοράν, τοῦ χαλκοῦ τὴν ἐπικαταλλαγὴν προσαπαιτεῖν. ([He is the type of person who,] when he collects the tenant-rent [apophoran] from his slave, also demands to exchange the copper.) A 3: Men. Epitr. 378–380: νῦν γὰρ μενοῦμεν ἐνθάδε, | εἰς αὔριον δ’ ἐπ’ ἔργον ἐξορμήσομεν | τὴν ἀποφορὰν ἀποδόντες.63 (We’ll stay the night here. In the morning we’ll go back to work, when we have paid our dues [apophoran].) This is said by the charcoal maker (ἀνθρακεύς) Siriskos, a slave (οἰκέτης) of Chairestratos, addressing his wife. A 4: Andoc. 1.38: ἔφη γὰρ εἶναι μὲν ἀνδράποδόν οἱ ἐπὶ Λαυρίῳ, δεῖν δὲ κομίσασθαι ἀποφοράν. ([Diokleides’] tale was that he had had to fetch the earnings [apophoran] of a slave of his at Laurium.) A 5: Ps.-Xen. Ath.Pol. 1.11:64 εἰ δέ τις καὶ τοῦτο θαυμάζει, ὅτι ἐῶσι τοὺς δούλους τρυφᾶν αὐτόθι καὶ μεγαλοπρεπῶς διαιτᾶσθαι ἐνίους, καὶ τοῦτο γνώμῃ φανεῖεν ἂν ποιοῦντες. ὅπου γὰρ ναυτικὴ δύναμίς ἐστιν ἀπὸ χρημάτων ἀνάγκη τοῖς ἀνδραπόδοις δουλεύειν, ἵνα †λαμβάνων μὲν πράττῃ† τὰς ἀποφοράς, καὶ ἐλευθέρους ἀφιέναι. (If anyone is also startled by the fact that they let the slaves live luxuriously there, and some of them sumptuously, it would be clear that even this they do for a reason. For where there is a naval power, it is necessary from financial considerations to be slaves to the slaves in order to take a portion of their earnings [apophoras], and it is then necessary to let them go free.) 63 See also Men. fr.431 Kock: εἶτ’ ἀμφορεαφόρος τις ἀποφοράν φέρων. 64 This passage is one of the most difficult texts about the condition of slaves in classical Athens. It can be said that there are two things in it that are sufficiently clear and interesting for our purposes: that in Athens at that time (towards the end of the fifth century) there were slaves who had some possessions of their own (see also further in the same paragraph: τὰ χρήματα … τὰ ἑαυτοῦ), and that the form of payment known as apophora was connected to this phenomenon. But at this moment I cannot undertake to establish the precise meaning of this text. [For this reason, she uses the translation of S. I. Radtsig, taken from the appendix of his translation of Ath. Pol. (Moscow-Leningrad 1936). I use the Loeb translation here for the sake of con- sistency.]