Introduction

The evidence on manumission and the social and legal position of manumitted persons in the Greek world differs greatly from that pertaining to slavery alone. Whilst the evidence for the latter derives predominantly from Athens, evidence for the former is more broadly distributed. Many inscribed manumission documents survive on stone, the bulk of them from northern Greece (though such documents are scattered across the ancient world). Close study of these documents and comparison between corpora from different Greek settlements allow scholars to grasp the legal position of manumitted persons in the Greek world, but the nature of the evidence raises a number of problems that continue to stir debate.

The Evidence

The earliest evidence mentioning apeleutheroi dates to the 5th century bce,¹ and for this reason it has been commonly held that the practice of manumitting slaves in Greece did not develop before the classical period. This is perhaps also the reason why very few works have analysed manumission in the Homeric poems.² A close look at the two poems, however, shows that specific forms of manumission (in its legal meaning of a transition from slavery to freedom) already existed in the archaic period. In the Iliad, manumission results from the payment of a ransom (apoina) for the purpose of releasing (lyein/apolyein) war captives who had been reduced into slavery by the enemy (for instance, Hom., Il., 6.46–50). In the Odyssey, the only reference to manumission is when Odysseus promises his slaves Eumaeus and Philoitius that if they will help him to fight against the suitors, he will make them hetairoi (companions) and kasignetoi (cousins) of Telemachus (Hom., Od., 22.212–216). The relationship between these individuals and Odysseus is no longer to be one based on ownership, since the term hetairos refers to the ideal of equality and the rule of reciprocity between individuals, whilst kasignetos implies the creation ex novo of a kinship tie: in this case the liberation of the two slaves implies their entrance in their master’s oikos no longer as slaves (i.e., property) but as equals and free persons.

Moving beyond the Homeric poems to later evidence, it is worth noting that the nature of the sources has long prevented a proper understanding of the condition of apeleutheroi. On the one hand, evidence for manumission and manumitted slaves is mainly provided by literary and epigraphic sources which often refer very briefly to manumission and freedmen; on the other hand, the geographical and chronological distribution of the evidence is far from uniform. In Athens, for example, the evidence for apeleutheroi is limited to the information provided by few forensic speeches from the classical age (Dem., 59.30–32, 40, 45–46; Hyp., 3.4–7) and brief remarks by later authors (Harp., s.v. apostasiou; Athen., Deipn., 6.267b–c; D. L., 5.54–55, 69–74, 15), whereas several poleis from central Greece (mainly Delphi) provide a large number of inscriptions dating to the Hellenistic period. Evidence from inscriptions is widely scattered geographically and can be found in places as diverse as Boeotia, Phocis, Locris, Thessaly, Epirus, Taenarum in Laconia, and Lemnos. Another group of inscriptions can be found in the site of Seleuceia-by-the-Eulaios, which provides interesting information on manumissions carried out in the Greek cities of the Seleucid Empire.³ One puzzle is why so few of these documents derive from the more famous poleis, places such as Athens, Corinth, or the Ionian cities. One answer may lie in the issue of differing regional epigraphic habits; another may lie in the use of papyrus documents in some of the more economically advanced cities. Papyrus documents have generally not survived, apart from some found in Egypt and dating to the Ptolemaic period (C.Ptol.Sklav., 28–36), as well as the wills of the philosophers quoted verbatim by Diogenes Laertius. These provide much valuable detail on testamentary manumission in late classical and early Hellenistic Athens.⁴ At any rate, one should not draw any simplistic relationship between the distribution pattern of evidence and the frequency of manumission from one region to the next.

The Legal Aspects of Manumission

The language of freedom in ancient Greek can be employed in a flexible fashion, and applied to both legal and extralegal contexts. Slavery (douleia) and freedom (eleutheria) can describe in a technical sense the condition of persons owned or not owned by others, but the same language can also be used to describe various extralegal relations of dependence and domination. Law court speeches (e.g., Dem., 57.34) show that Greeks could make a pragmatic distinction between the legal register of slavery terminology and the looser shades of meaning that could be employed in extralegal contexts.

Manumission inscriptions reveal further the sophisticated and pragmatic understanding among the Greeks of the technical distinction between slaves and free persons. Those who have passed from the former to the latter condition via manumission are called apeleutheroi in Greek. Legally, slavery is a relationship based on ownership: this means that slaves were included among their masters’ property and that the latter could exercise over them all the powers and rights pertaining to ownership (such as, typically, the rights to possession, use, income, capital, security, and transmissibility, as well as the absence of term and liability to execution).⁵ As an effect of manumission, freedpersons were no longer the property of their (former) masters, but were rather considered to be—according to a formula often used in the Greek sources—“masters of themselves.” The rights and powers pertaining to legal ownership could no longer be exercised over them. As with the Greeks, the Romans had a category of freedpersons, and designated freedmen/freedwomen with the term liberti/libertae. Although both apeleutheroi and liberti were legally free, there was one fundamental difference between the two: whereas liberti, as an effect of formal manumission, usually acquired citizenship with their freedom (becoming cives Romani), apeleutheroi acquired citizenship (politeia) only exceptionally as an effect of the enactment of a public decree (psephisma) by the polis. The most famous example is the case of Pasion (see Dem., 59.2 and the discussion on Pasion below).

Traditional scholarship has grouped the main ways of manumitting slaves into specific categories.⁶ The first basic distinction is between “public” and “private” manumissions: public manumissions were those in which the polis liberated either public slaves (i.e., slaves owned by the polis) or privately owned slaves as a reward for a specific service they rendered to the state. A classic illustration of the latter type lies in the events surrounding the Battle of Arginusae in 406 bce, when the Athenian state, in need of manpower, promised freedom to any privately owned slaves who enlisted as rowers in the fleet (Xen., Hell., 1.6.24; Ar., Ran., 693–694; Hellanicus, FGrHist, 323a F 25; Diodorus, 13.97.1). In Sparta, a particularly unusual situation prevailed: helots, although privately owned, could not be privately manumitted, a right that was vested in the state alone (Ephorus, FGrHist, 70 F 117). Private manumissions were those by which masters liberated their own slaves, either for free or, more often, in exchange for money. Among private manumissions, traditional scholarship distinguishes between “sacral” and “secular” manumissions: the first group implies the intervention of the god in the manumission procedure; the second group is defined as such because it does not involve the religious sphere, but merely the civic.

Manumission at Delphi

The bulk of evidence for the so-called sacral manumissions comes from Hellenistic central Greece, particularly from inscriptions found at Delphi, which are believed to represent “manumission through sale” of a slave to the god (e.g., SGDI II 1688).⁷ This epigraphic material constitutes the most important source of information we possess about manumission and the condition of apeleutheroi, since the wall of Apollo’s temple in Delphi has preserved over one thousand inscriptions recording the liberation of more than 1,200 slaves. The content of these inscriptions is highly formulaic and has traditionally been interpreted as describing manumissions as sales of slaves to the god (generally Apollo) by their masters: as an effect of sale, the god would become the slave’s new master; the god would then manumit the slave at a later point in time. This view, however, is undermined by a fundamental problem: legally speaking, sale involves a conveyance, that is, a transfer of ownership (over a slave from one master to another), whereas the ultimate effect of manumission is the cessation of any rights of ownership over a slave, who therefore becomes a free individual. Identifying manumission with a sale is therefore highly problematic, since the two acts lead to opposite legal effects (cessation of ownership versus transfer of ownership). Traditional scholarship has circumvented the problem by describing these manumissions as “fictive sales”; however, their vocabulary and clauses indicate that no actual or fictitious sale took place. The manumission inscriptions from Delphi show that in all these cases slaves were made free in exchange for money, and that the money used for “buying” freedom was provided by the slaves themselves. Slaves could give money to their masters, but the transaction had no legal meaning because the slave had no rights and could not stricto sensu be a party to a legal transaction with his or her master. For this reason, they entrusted (pisteuein) their money to the god with the assurance that the god would then transfer the slaves’ money to their masters, thus completing the transaction which would ultimately result in the slaves “buying” their own freedom.⁸ The god, in other words, would give the money to the masters “on behalf” of the slaves, without acquiring any right or claim over the slaves thus manumitted. The Delphic inscriptions show that manumission in exchange for money has the nature of a bilateral legal transaction between the slaves’ masters and a third party, whose role is simply to hand over the money, in order for the slaves to be validly manumitted. SGDI II 1685.1–5 illustrates a typical case:

ἄρχοντος Θρασυκλέος μηνὸς Ἡρακλήου, ἀπέδοτο Ἄθαμβος Ἀθανίωνος τῶι Ἀπόλλωνι τῶι Πυθίωι σῶμα γυναικεῖον ἇι ὄνομα Ἁρμοδίκα τὸ γένος ἐξ Ἐλατείας, τιμᾶς ἀργυρίου μνᾶν ἕξ, καὶ τὰν τιμὰν ἔχει πᾶσαν, καθὼς ἐπίστευσε Ἁρμοδίκα τῶι θεῶι τὰν ὠνάν, ἐφ’ ὧιτε ἐλευθέρα εἶμεν καὶ ἀνέφαπτος ἀπὸ πάντων τὸμ πάντα χρόνον ποιέουσα ὅ κα θέλη καὶ ἀποτρέχουσα οἷς κα θέλη. βεβαιωτῆρες κατὰ τὸν νόμον τᾶς πόλιος·‎

(Under the archonship of Thrasykleos, in the month of Herakleios, Athambos son of Athanion sold to Pythian Apollo a female slave named Harmodika, Elateian by race, at the price of six silver minae, and he (Athambos) received the whole payment, since Harmodika entrusted the purchase money to the god, under the condition that she is free and not claimable by anyone forever, being (and she is) free to do whatever she wants and to go wherever she wants. Guarantors according to the law of the city.)

Earlier scholars viewed this procedure as a kind of fossilized survival from an earlier, pre-droit period when the institutional features associated with the classical polis had not yet developed.⁹ Such a view is implausible. Not only is there no evidence from the archaic period for any such practice but secular manumission does appear to be present in Homeric epic. The sacral nature of these documents is, rather, a pragmatic way of (1) circumventing the legal obstacle posed by the slave’s lack of legal personality, (2) involving a trustworthy institution as an intermediary in the transaction, (3) maximizing publicity by inscribing a record of the transaction in the sanctuary, and (4) creating warranty of title—as the seller/manumittor was obliged to warrant the freedom of the freedperson if it were challenged by a third party.

Consecration

One further group of inscriptions which is believed to represent a specific mode of sacral manumission comes from Hellenistic Boeotia and is customarily labelled as “manumission through consecration” of slaves to the god¹⁰: these sources describe masters who dedicate (anatithemi) their own slaves as sacred (hieroi) to a god, usually Serapis. A typical case is IG VII 3312, from Chaironeia in the 2nd century bce:

ἄρχοντος Ἀρχεδάμου, μηνὸς Ὁμολωΐου πεντεκαιδεκάτ[ῃ], Θέων Σωμήλου Λεβαδεύς, παρόντος αὐτῷ τοῦ υἱοῦ Σάμωνος, ἀντίθησι τὸν ἴδιον θρεπτόν, ᾧ ὄνομα Σωσίδαμος, ἱερὸν τῷ Σεράπει, πο[ιο]<ύ>μενος τὴν ἀνάθεσιν διὰ τοῦ συνεδρίου κατὰ τὸν νόμον‎.

(Under the archonship of Archedamos, in the fifteenth day of the month of Homoloios, Theos Lebadeus son of Somelos, being his son Samonos with him, dedicates his own houseborn slave, named Sosimados, as sacred to Serapis, having performed the dedication before the council according to the law.)

The conventional view holds that these dedications were fictitious since, as an effect of consecration, the slaves became free. However, interpreting this evidence as referring to manumission is problematic. On the one hand, both the verb anatithemi and the adjective hieros technically refer to the act of consecration and the condition of consecrated items, which therefore belong to the gods. On the other hand, the vocabulary of these inscriptions is highly formulaic, and the same formulae are found elsewhere in the Greek world for regular dedications to the gods. This suggests that these inscriptions do not record fictitious but real consecrations of slaves who would then become property of the god (hieroi). That consecration ultimately results in a transfer of ownership from humans to gods is also confirmed by other Greek sources, which always refer to consecrated slaves as “slaves of the god” (see, for instance, Eur. Ion; Strab. 8.6.20). This shows that, from a legal perspective, hieroi were slaves of the god, even if the absence of a human owner who could concretely exercise the rights and powers pertaining to ownership made their de facto condition akin to that of free individuals. Legally speaking, if hieroi were property of the god, they were not free, and therefore the consecration of slaves to the god can hardly be described as manumission. In fact, on some occasions apeleutheroi (individuals already manumitted before consecration) were consecrated as hieroi by their former masters (IG VII 3318). This illustrates one of the problems of the traditional taxonomy used by scholars of Greek manumission, namely the collapsing into one category (“sacral”) of two very different kinds of operation.

Manumission and the Law at Athens

Given the Athenocentric nature of evidence for many aspects of ancient Greek life, it may come as a surprise that Athens provides little evidence for so-called secular manumission. The bulk of scholarly discussion on manumission at Athens has traditionally focused on a group of thirty-three inscriptions recording the dedication of phialai (bowls) each weighing one hundred drachmas.¹¹ The traditional view held that these sources constituted the only Athenian epigraphic evidence for manumission and that they represented fictitious dikai apostasiou (“suits for desertion”) brought by masters against their slaves for the purpose of manumission. However, recent studies have shown that these sources do not deal with manumission and apeleutheroi but with acquitted defendants who had dedicated these phialai as a mark of thanks for victory in the trial.¹² The evidence from these inscriptions cannot, therefore, stand as documentation pertaining to manumission at Athens.

The available information on manumission and the condition of freedmen in Athens derives in the main from forensic oratory (e.g., Dem., 59; Hyp., 3). This evidence shows, on the one hand, that whenever slaves were manumitted in exchange for money, the intervention of a third party (other from the slave) was necessary for conveying the money to the slave’s master and, on the other hand, that manumission in exchange for money had the nature of a bilateral legal transaction between the slave’s master and a third party. In other words, this “secular” transaction was substantively similar to the “sacral” transaction noted at Delphi, the difference lying in the identity of the third party: at Athens, a human agent; at Delphi, the god Apollo, represented by his priests. Payment for manumission was common throughout the Greek world and this practice has been interpreted by traditional scholarship either as a specific mode of manumission known as “sale for the purpose of freedom” (prasis epi eleutheria) or as a fictitious sale.¹³ The first definition implies that the third party first purchased the slave, thus becoming the new owner, and then manumitted the slave. However, the vocabulary of the sources never implies this two-step process but always suggests that the slave’s liberation ensued as an immediate and automatic result of payment by the third party. The second idea (manumission as a fictitious sale) is also misleading, since the vocabulary used in the sources for describing manumission is different from that of sale and simply suggests that the third party pays the money “on behalf” of the slaves, who could not conduct legally valid transactions with their masters.

Specific duties were imposed on those freedmen who decided to reside in Athens.¹⁴ First, they had to pay the metoikion, a residence tax that was paid also by metics and set at twelve drachmas a year for men and six drachmas a year for women (Harp., s.v. metoikion). Another tax, the triobolon, mentioned only by Harpocration (Harp., s.v. metoikion), was perhaps imposed on freedmen. Second, freedmen in Athens had to have their former owners as prostatai (a similar duty was imposed on metics, who had to register an Athenian citizen as their prostates), whose function was to represent apeleutheroi especially in courts. This implies that whenever freedmen brought an action or were brought to trial, they had to be represented in court by their prostatai. Third, they were liable to be called up to fight for the polis in times of war. The jurisdiction on those cases which involved the apeleutheroi both as plaintiffs or as defendants was held by the Polemarch (Arist., Ath. Pol., 58.3). There was however a specific private action, known as dike apostasiou, which directly involved apeleutheroi. The dike apostasiou could be brought before the Polemarch by manumittors against manumitted slaves in three cases: (1) if apeleutheroi went away from their manumittors; (2) if they registered a different person, other from their manumittor, as their prostates; and (3) if they failed to do “what the laws prescribed” (Harp., s.v. apostasiou). Harpocration further specifies that those apeleutheroi who lost the trial reverted into slavery, whereas those who were acquitted became “completely” free.

Paramone

The fact that manumittors could prosecute apeleutheroi if they “went away” from them suggests that after manumission apeleutheroi could be subject to a specific duty of “remaining by” their former masters in order to perform some services for them. This duty is usually labelled by scholars with the comprehensive term paramone, and clearly recalls the operae libertorum that liberti had to perform for their patroni in Rome. The categorization of apeleutheroi with paramone duties has often puzzled scholars: some of them describe their legal condition as one of slavery,¹⁵ whereas others considered it an unspecified form of semi-freedom or half-slavery.¹⁶

Paramone is most thoroughly attested in the Delphic manumission inscriptions: about one quarter of them record the imposition of post-manumission obligations on freedmen.¹⁷ Some of these inscriptions, after describing the liberation of slaves in the form of a “sale” to the god, prescribe that apeleutheroi must “remain with” their former masters (or, in few cases, with different persons designated by the manumittor, but usually belonging to the latter’s family) for a specific period of time. The duty of “remaining with” is usually expressed with the aorist imperative of the verb paramenein (parameinato, parameinanton), which has an aspectual value and thus indicates that this obligation comes with and as an effect of manumission. In most cases the inscriptions do not mention the specific duties imposed on apeleutheroi under paramone, but generically state that freedmen must paramenein with their manumittors and do whatever they are ordered to do. The duration of paramone is always specified in the inscriptions: its length could be several years (e.g., FD III 2:217) or, more often, the rest of the manumittor’s life (e.g., SGDI II 1703).

The existence of post-manumission obligations upon freedmen is also attested by some Athenian sources, even though the evidence for paramone in Athens is scanty: the most relevant sources are Harpocration (s.v. apostasiou) and the wills of the philosophers Aristotle, Theophrastus, and Lyco preserved by Diogenes Laertius (5.54–55, 69–74, 15); implicit references to paramone in Athens are also made by Plato (Laws, 914e–915c) and Athenaeus (Deipn., 6.267b–c). These sources, however, do not always refer to post-manumission duties with the verb paramenein since therapeia is also mentioned as a general duty binding apeleutheroi to their former masters. However, in compelling manumitted slaves to perform further services towards their manumittors, they clearly refer to the obligation of the former to remain in the household of the latter no longer as slaves but as free dependants.¹⁸

Paramone did not always characterize the condition of apeleutheroi after manumission. Those freedmen who, after manumission, were not legally obliged to perform paramone obligations towards their former masters were free not only from a legal point of view (i.e., they no longer belonged to their former masters) but also de facto, in the sense that they could live separately from their former masters and constitute a separate household. Freedom of movement and action as characteristic of the condition of freedpersons released from paramone obligations is often stressed in the Delphic inscriptions, which state that the apeleutheros, after manumission, “is free to do whatever he/she wants and to go wherever he/she wants” (e.g., SGDI II 1697). Conversely, those apeleutheroi who, after manumission, remained under paramone obligations towards their manumittors were still attached to their former masters’ households. They did not enjoy complete freedom of movement and action but were nonetheless free individuals from a legal point of view. The fact that some apeleutheroi were held under paramone obligations and others were not suggests the existence of two categories of manumitted slaves. The distinction between these two categories was their de facto condition, viz. their capacity to live in a different household (in a legal, not a physical sense) from their manumittors. In both cases they were labelled as apeleutheroi and were legally free individuals. At least in Athens, freedmen who were not under paramone obligation could also be referred to as khôris oikountes (literally, “dwellers apart”: Dem., 4.36–37).¹⁹

The nature of paramone was that of a post-manumission obligation imposed on freedmen in favour of specific individuals, usually their former masters. The only example of post-manumission obligations automatically following manumission comes from 1st-century BCE Calymna, which provides the only evidence for what has been called by scholars paramone ex lege. In Calymna, paramone was not the object of an agreement between manumittors and freedmen but was a compulsory feature imposed on manumitted slaves by the apeleutherotikoi nomoi, that is, by those laws specifically aimed at regulating the condition of freedmen. In other words, this meant that in Calymna, paramone was not merely a possible option that owners could impose on their slaves upon manumission but a duty imposed on all freedmen by law.²⁰

The nature of paramone as an obligation between freedmen and manumittors implies that freedmen under paramone were legally free. This is further shown by some clauses mainly mentioned in the Delphic inscriptions and aimed at enforcing the performance of paramone by freedmen. Some inscriptions, for instance, state that if freedmen did not paramenein with their manumittors, the latter could punish them in any way they liked but could not sell them (e.g., SGDI II 2171). Prohibition to sell clearly indicates that the relationship between manumittors and manumitted slaves was no longer based on the right of ownership, since the right to alienate through sale (or gift) is one of the fundamental rights pertaining to ownership (cf. Arist., Rhet., 1.5.1361a21). The fact that the beneficiaries of paramone could sometimes be entitled to punish disobedient apeleutheroi is not inconsistent with the latter’s legal condition as a free person. Debtors, as free individuals, could be subject to physical coercion by creditors, as is attested in early Roman law (see, for instance, the nexi). Another clear indication of the legal condition of freedom enjoyed by apeleutheroi is the necessity, mentioned in some inscriptions, to abide by the decision of three arbitrators in the event of any disagreement between manumittors and manumitted persons on the fulfillment of paramone duties and the express statement that the arbitrators’ decision had to be considered valid and binding both by manumittors and freedpersons (e.g., SGDI II 1689). This clause shows that apeleutheroi held under paramone were not slaves, since nobody could interfere with the masters’ ownership rights over their slaves. The fact that some inscriptions refer to apeleutheroi under paramone as douloi (e.g., FD III 2:129) or somata (e.g., FD III 1:304) does not necessarily mean that they were slaves in a technical sense: it may reflect a metaphorical use of the vocabulary of slavery, which was common in Greece and refers to the de facto condition of this specific category of apeleutheroi.

The performance of paramone by apeleutheroi was enforced in different ways. The Delphic inscriptions illustrate two possible scenarios. Some of them state that if the apeleutheros did not paramenein, the only possible remedy was physical punishment (e.g., SGDI II 1703). Freedmen, in this case, remained legally free. Some inscriptions, on the other hand, mention that if apeleutheroi did not paramenein, their manumittors could sell them as slaves (e.g., FD III 3:329) or their manumission was to be considered invalid (akyros: e.g., SGDI II 1702). In these latter two cases, the result was effectively the same: the apeleutheros was no longer considered to be a free person and reverted to the legal condition of slavery. Reversion into slavery as a penalty for the nonfulfillment of post-manumission obligations was common also in Athens, since Harpocration (s.v. apostasiou) specifies that if apeleutheroi were acquitted in a dike apostasiou, they would be “completely” free, whereas if they lost the trial, they would revert to being slaves. Harpocration’s definition of dike apostasiou, together with the evidence from the Delphic inscriptions, illustrate that the condition of apeleutheroi under paramone was simply a temporary one. At the end of the paramone term, the slaves would become “completely” free: together with legal freedom (which they enjoyed throughout the period of paramone), they also acquired de facto freedom, which would allow them to live separately from their manumittor’s household and to enjoy freedom of movement and action.

Apolysis

One of the most telling indications that apeleutheroi who were held under paramone obligations were legally free is the process called apolysis in the inscriptions.²¹ Apolysis is attested in a number of inscriptions from Delphi and effectively refers to the “early release” of freedpersons from their paramone obligations in return for paying a specific sum of money to the beneficiaries of their services (e.g., FD III 3:354). The fact that apeleutheroi under paramone could themselves pay money to their former masters in order to obtain early release from paramone marks a fundamental distinction between their legal condition prior to and after manumission, and further proves that this category of manumitted slaves was legally free. On the one hand, it shows that freedpersons under paramone legally owned their money (whereas slaves could only “possess” money, the legal title to which always lay with their owners, similar to the institution of peculium in Roman law). On the other hand, it demonstrates how freedmen under paramone could validly pay money to their former masters, which implies that they could negotiate and enter into legally binding agreements as free persons (whereas slaves could not pay money to their masters for their manumission and therefore needed the intervention of a third party in the manumission procedure, given their lack of legal personality).

The Economic and Social Dynamics of Manumission

The Greeks understood the legal distinction between slavery and freedom and could also find ways to maintain the domination of the master even after he surrendered many of his legal over his slave, namely by the inclusion of special paramone obligations in the manumission act. It is also worth considering the social and economic dynamics that shaped these legal relations. How common was manumission in the Greek world, and how could slaves amass enough money to pay for their manumission if they had no legal right to it? Athens provides a broader range of evidence than the inscribed manumission documents from northern Greece and elsewhere.

The incentive value of manumission and its ability to motivate slaves was well known (e.g., Arist., Oec., 1344b; Arist., Pol., 1330a32–3). Recent work suggests that certain occupations—especially those requiring high levels of motivation, trust, and skill (so-called care-intensive roles)—were likely those in which manumission was dangled by the master as a kind of carrot for the slave to work towards. The owners of slaves working in occupations that required little more than brute strength, or low-skilled jobs (so-called pain-intensive roles) are less likely to have attempted to encourage them with incentives such as manumission, and more likely to have employed lesser incentives, such as better food and clothing or the opportunity for sexual liaisons (Xen., Oec., 9.5, 13.6–10), and the threat of punishment (Arist., Oec., 1.5.3).²²

The promise of manumission was among the most potent tools in the master’s arsenal for motivating his slaves, though the high value of slaves meant that manumission entailed a significant financial loss for the owner. Both formal and informal means were used to circumvent this potential problem. On the informal side, it is likely that most slaves in the Greek world were actually allowed to amass possessions. Theophrastus (Char., 30.9), for example, regarded it as shameful and unusual for a man to force his slaves to hand over any coins they found on the street. Some slaves, though, might earn significant amounts of money beyond the trivial pocket money to which Theophrastus alludes. One informal way of motivating slaves was for masters to have them pay a fixed fee (apophora) from their earnings (e.g., Aeschin., 1.97); any money earned over and above this was retained by the slave, meaning that lucky and industrious slaves working under such an arrangement might amass enough money to buy (eventually) their freedom. It is vital to grasp both the legal and extralegal dimensions of this practice. On a formal legal level, anything slaves possessed belonged to their owners. But on an informal level, by allowing slaves to “keep” a portion of their earnings, masters could both motivate their slaves to work harder and recompense themselves towards the end of the slaves’ working lives, by having the slaves pay their savings back to the owners, giving the latter the money to purchase younger replacements. Such a system was self-perpetuating. In fact, one writer, the “Old Oligarch,” specifically points out the connection between apophora arrangements and allegedly high levels of manumission at Athens (Xen., Ath.Pol., 1.11). That, at least, is the cynical calculus that must have lain behind the promise of manumission many masters offered their slaves. Furthermore, it has recently been suggested that manumission could also be used by some slave owners as an economic tool to avoid having to provide for slaves in the periods when urban demand dropped.²³

However, we should not suppose that real emotions never interfered with the “rational,” profit-maximizing motivations that economists tend to focus upon. The wills of the philosophers provide us with fine-grained information showing how a single master might use a variety of approaches to reward (or not) his slaves upon his death, and expose some of the emotional connections. For instance, the will of the metic and philosopher Aristotle (Diogenes Laertius’ Lives of the Philosophers (D. L.), 5.12–16) mentions how some slaves were gifted to individuals, whilst another slave, Ambracis, was to be freed and given, on the occasion of Aristotle’s daughter’s marriage, ownership of the slave girl the daughter possessed as a maid. Other slaves were to be given their freedom on the occasion of the aforementioned marriage, and Aristotle stated in his will that no slave who attended him should be sold, and all who deserved it should be at some point manumitted. This shows a very flexible approach and hints at the relationships that Aristotle had built up with his slaves—some close, others less so. As a point of his generosity, no slave paid for freedom: any persons freed as an effect of the will were freed gratis, and some were even given gifts. But not all owners were as affluent as Aristotle, or as generous.

Two case studies demonstrate ways slaves may have used their skills or attributes to achieve manumission, and in some cases to thrive as freedpersons: the case of the bankers Pasion and Phormion and the case of hetaira Nearia.

Pasion and Phormion

The examples of Pasion and Phormion illustrate the dynamics of a slave-master relationship where the former was a talented individual working in a care-intensive role. Pasion’s life as a slave in Athens began at the end of the 5th century, when he worked as a banker for his co-owners, Antisthenes and Archestratus (Dem., 36.43). His trustworthiness and business acumen led to him being manumitted, though he remained in the bank, perhaps as a lessee, later became its owner, and accumulated a large fortune from which he made lavish benefactions to the Athenians and was eventually elevated to citizenship by a decree of the assembly. When he died he left a fortune of twenty talents, with at least a further thirty-nine talents lent out (Dem., 36.5). The career of Pasion’s own slave, Phormion, followed a comparable trajectory. Phormion began as a slave manager in Pasion’s bank, was later freed, and leased the bank from his former owner (Dem., 36.11; 36). As a condition of Pasion’s will, Phormion married Pasion’s widow and became guardian of his younger son Pasicles. Phormion too rose to become a citizen and one of Athens’ richest men (Dem., 21.157). Other non-Athenian bankers carved out comparable careers (Din., 1.43), and one ex-slave banker of the latter half of the 4th century became the tyrant of the cities of Atarneus and Assos, succeeding his ex-owner (Str., 13.1.57; Didymus, Demosthenes, cols. 4–6).²⁴

Neaira

In his speech Against Neaira, Apollodorus acts as synegoros (supporting speaker) of the plaintiff Theomnestus, and the indictment is a barely disguised attempt to attack Apollodorus’s political enemy Stephanus, who had allegedly married a foreign woman—Neaira—and attempted to pass off her children as his own legitimate Athenian progeny. Whatever the actual legal merits of the case, sections 18–49 provide a narrative of Neaira’s early life up to her residence in Athens with Stephanus. Neaira was bought as a child by a madam named Nicarete, and was raised alongside six other girls; Nicarete herself was a former slave who had belonged to a man from Elis (Dem., 59.18). She pretended that these girls were her daughters and prostituted them out from an early age for significant sums of money. The girls’ lovers were often men of distinction, such as Lysias the logographer, Xenocleides the poet, and the actor Hipparchus (Dem., 59.21, 26). Neaira allegedly gained quite a reputation in Corinth (Dem., 59.26, 28) and was bought for half a talent (around ten times the average price of a slave in Athens) by two men: one local, Timanoridas of Corinth, the other a man named Eucrates from the island of Leucas (Dem., 59.29). What occurred next is of particular interest. When her co-owners decided to marry, they wanted to be rid of Neaira, but rather than selling her to the highest bidder, they gave her the opportunity to buy her freedom, knocking a thousand drachmas off the price she had been purchased for and asking her to provide the remaining two thousand drachmas. To this end, Neaira enlisted the aid of several of her lovers, who put up the remaining money to buy her freedom (Dem., 59.30). One of them, Phrynion of Athens, took the money and paid it to Neaira’s owners to secure her freedom.

This example shows how a skilled hetaira could improve her options by drawing on the network of individuals with whom she had had intimate relations: first, her owners, who could simply have sold her for a higher price but were prepared to take a loss (perhaps) for the sake of their fondness for her, and secondly, her lovers, who were prepared to spend significant sums of money to secure her freedom. Neaira’s passage to freedom is well documented but hardly exceptional. According to Athenaeus (13.590d), the orator Hypereides bought the freedom of several expensive hetairai and set them up as his “kept women” in several locations in Attica (cf. Plu., Mor., 849d). And it was not only female slaves who might use their “sexual capital” to effect their manumission: Hypereides’ speech Against Athenogenes provides us with a male parallel, in which we hear of a man (the plaintiff, in fact) who is infatuated with a beautiful boy and is conned into buying the boy’s father and brother by their owner. As we have seen in other instances, manumittors might impose legal obligations to ensure that their sexual access to freed individuals was guaranteed.

Conclusion

Attempts to quantify the rate of manumission in this or that polis have not been convincing, and perhaps the best that can be said is that higher rates of manumission were associated with those slaves who worked in high-skilled occupations requiring considerable trust and initiative (occupations that were especially prominent in the more economically developed coastal cities), and that manumission was probably also in many cases the result of the bonds built up between owners and their slaves of long standing. The sheer variety of evidence attests to the multitude of ways in which manumission could be effected by owners across the Greek world, but at the heart of these different methods lies the same basic transition: that of being the property of another person to being the property of no one at all. But these -same documents also attest to the cynical calculus of many Greek slaveholders, who further exploited their slaves via paramone arrangements. As for the lives of ex-slaves, the evidence is particularly patchy, but from the few cases for which information is available, it is clear that legal disabilities and social stigma did not prevent the most talented individuals from occasionally achieving positions of considerable importance.