SISTERHOOD, AFFECTION AND ENSLAVEMENT IN HYPERIDES’ AGAINST TIMANDRUS

Abstract

A recently published fragment of the fourth-century speechwriter Hyperides contains a speech for the prosecution of Timandrus, accused of mistreating four orphans in his care. This article draws out from the fragment three important contributions to our understanding of Athenian conceptions of family relationships, particularly the relationships of marginalized groups: girls and enslaved people. First, the fragment constitutes a rare portrayal of a relationship between two sisters. Second, the fragment clearly articulates the idea that affective family relationships are not a biological inevitability but arise from socialization, a departure from other fourth-century thinking. Third, the speaker applies this statement to enslaved people, claiming that the separation of children from close family members is so cruel that even slave-traders avoid it in their sale of human beings. Though this claim seems to have been untrue except in a very limited sense, its place in the argumentation of the speech assumes broad recognition of the existence and value of family relationships between enslaved people, vivid evidence of the paradox that slave societies recognized the humanity of people they simultaneously insisted were subhuman.

The publication in 2005 of a fragment of Hyperides’ Against Timandrus ¹ produced a flurry of scholarship: on the text and its transmission (Horváth, Easterling, Ucciardello); on how the speech fitted into what sort of trial (Thür, Whitehead); on the alleged practice of keeping enslaved families together at sale (Jones, Schmitz); on the law concerning the care of orphans (Rubinstein).² Part of the ‘Archimedes Palimpsest’, the fragment consists of two pages (64 lines) from Hyperides’ speech for the prosecution of one Timandrus, guardian (ἐπίτροπος) of four orphans and their property; adjacent pages contain remnants of Hyperides’ Against Diondas. The passage of Against Timandrus, which is all that remains of the speech, begins and ends mid-sentence; it is not clear how much of the text originally appeared on the parchment, or in what context. The lines come from the πίστις (‘proof’) section of the speech.³ The narrative (διήγησις) had probably been given by a previous speaker; the speaker of the fragment makes brief reference to some events which this other speaker may have mentioned (lines 20–7, 44–9, 63–4), but this is only guesswork. The central complaint in the fragment is that Timandrus separated a child from her siblings.

The fragment makes valuable contributions to our understanding of Athenian conceptions of family relationships, particularly the relationships of free girls and of enslaved people. This article elucidates three of these contributions. First, the fragment is a rare portrayal of a relationship between sisters—ironically, through the severance of that relationship. It attests to an emotive interest among Classical Athenians in sororal relationships, abundantly clear in tragedy but otherwise unusual in literature, even in forensic speeches concerned with family dynamics. Second, the fragment articulates the idea that affective family relationships are not biologically inevitable but arise from socialization (lines 35–42)—a departure from other fourth-century thinking. Third, the speaker applies this insight to enslaved people, claiming that the separation of children from close family is so cruel that even slave-traders avoid it (29–35). Whether or not this was true, in drawing such an analogy Hyperides recognizes—and expects the jurors to recognize—that enslaved people formed affective relationships comparable to those of free people and worth preserving. This article uses the fragment to develop our understanding of Athenian thinking about family and family relationships, particularly between those whose emotions and relationships are less visible in the historical record.

THE SIBLINGS TRAGEDIZED

In the fragment, Hyperides stages a family tragedy. Griffin has shown that the ‘acute suffering, extreme situations, and agonizing decisions’ which appear on the Athenian tragic stage ‘are intimately linked to events in real, recent, and contemporary life’; these situations resonated because they could and did happen in the real world.⁴ He draws parallels between events of the tragic stage and stories told in the historians; there are similar parallels to be drawn with stories told in the orators.⁵ Andocides tells of a woman who tried to hang herself out of shame at the quasi-incestuous sexual relations within her family (1.124–5; compare Jocasta in Sophocles’ Oedipus Tyrannus); Andocides suggests that the child of one of the women involved should be called Oedipus (128–9).⁶ Isaeus presents a situation in which someone tries to prevent a dead man's relatives from burying him, though the relatives manage to bury him at night (6.39–41; compare Sophocles’ Antigone). Similarly, elements in Hyperides’ narrative are the stuff of tragedy: family members separated in childhood (cf. Creusa and Ion in Euripides’ Ion; Electra and Orestes in Sophocles’ and Euripides’ Electras and Aeschylus’ Choephori); mistreatment of vulnerable children by guardians (cf. Creon, Ismene and Antigone in Sophocles’ Oedipus at Colonus). We know of fourth-century reperformances of the four Sophoclean dramas just mentioned, for example:⁷ Hyperides is evoking patterns on which the tragedians also drew, familiar to his audience from the stage and from life.

Making a quick, compelling sketch of the children and their relationships to elicit maximum pity, Hyperides presents them as two sets of siblings: δυοῖν ἀδελφοῖν καὶ ἀδελφαῖν δυοῖν, ‘two brothers and two sisters’, or, to push the dual further, ‘a pair of brothers and a pair of sisters’ (21); contrast Isae. 2.3, without pathos, ἦμεν δὲ αὐτῷ παῖδες τέτταρες ἡμεῖς, δύο μὲν ὑεῖς, δύο δὲ θυγατέρες, ‘we were his four children, two sons and two daughters’. By presenting the girls as ‘a pair of sisters’, Hyperides establishes the relationship between them which will be broken later (44). The chiasmus marks the phrase; contrast the Suda's δυοῖν ἀδελφοῖν καὶ δυαῖν ἀδελφαῖν (π 847 Adler), referring to this speech. Hyperides contrasts the implicit closeness of the children with the emphasized loss of both mother and father (ὀρφ̣α̣ναῖν καὶ μητρὸς καὶ π(ατ)ρ(ό)ς, 21–2).⁸ ὀρφανός/-ή denoted a child who had lost his or her father; one could be ὀρφανός with a living mother. The phrasing presents the loss of two parents as a triple blow: the children are ‘orphans/fatherless [having lost] both their mother and their father’. Compare the opening of Sophocles’ Antigone (1–60, especially 49–60) where the closeness and the interdependence of the two sisters are contrasted with the loss of both their father and their mother (and then their brothers). Both sisters repeatedly use duals to describe themselves (Antigone: 3, 21; Ismene: 50, 58, 61–2) and their brothers (Antigone: 21; Ismene: 55–7), although, after Ismene states that she will not help Antigone bury their brother, there are no more duals in the scene, and Antigone distinguishes emphatically between ‘I’ and ‘you’ (69–72, 76–7).

Further, the children are very little (παιδαρίων, 22). The speaker describes the girl as ἴσως … ἑπτὰ ἐτῶν (‘perhaps seven years old’, 27); her elder brother was ἴσ̣ω̣ς̣ … δέκα ἐτῶν (‘perhaps ten years old’, 23–4). It seems unlikely that he did not know their ages: explicit references to children's ages elsewhere in Classical literature show that the numerical ages of both boys and girls were known and important.⁹ The speaker's vagueness may be because the children were in fact older and the speaker is manipulating the facts for sympathy; or they were ten and seven, but ἴσως allows the audience to think of them as being younger.¹⁰

Hyperides portrays the severance of the children's closeness with similarly brief, emotive detail (25–31):

τὴν νεωτέραν αὐτῶν ἀδελφὴν ἀποσπάσας οὑ̣τοσὶ Τίμανδρος ἔτρεφε παρ᾽ αὐτῷ ἀποκ̣ο̣μ̣ί̣σ̣(α̣ς̣) εἰς Λῆμνον ἴσως οὖσαν ἑπτὰ ἐτῶν. κα̣ίτ̣ο̣ι̣ τοῦτο μὴ ὅτι ἐπίτροπ(ος) ἢ εὔνους <ἂν> ἄν(θρωπ)ος ποιήσαι, ἀλλ’ οὐδ’ οἱ κατὰ πόλεμον ἐγκρατεῖς γιγνόμενοι τ(ῶν) σωμάτων, ἀλλὰ καὶ κατ᾽ οἰκίαν πωλοῦσιν ὅτι μάλιστα.

This man Timandrus dragged away their younger sister, took her away to Lemnos and brought her up in his house; she was perhaps seven years old. Now, this is a thing no guardian would do, nor any human being with his heart in the right place, not even those who get control of slaves in war, but even they sell them as families as far as possible.

The speaker refers to the younger girl (not named, in accordance with Athenian custom)¹¹ not as ‘the younger girl’ but as ‘their younger sister’ (τὴν νεωτέραν αὐτῶν ἀδελφήν), thereby focalizing through the other children.

The image conjured up at 28–31, where the speaker compares Timandrus’ behaviour to that of conquering armies, is ‘the truly frightening one of andrapodismos’.¹² ἀνδραποδισμός involved soldiers who captured a settlement extracting the older children and young women to exploit themselves or sell to traders for exploitation by others; to achieve this end, they typically killed any surviving men of fighting age, then slaughtered a number of unwanted women and children (the old or very young) to terrorize the rest into submission, then, by cudgelling and other tactics, separated out the desired women and children.¹³ ἀποσπάω (25) is part of this image: it evokes the violent manhandling associated in Classical Athens with slavehood, and specifically the sexual violation of enslaved girls, including war captives;¹⁴ aggravated sexual violence against women and girls was central to ἀνδραποδισμός.¹⁵ Paired with ἀποσπάω, the shared prefix of the more neutral verb ἀποκομίζω (‘carry away’, 26) gives it a sinister tone, developing the image (it is once used of war captives, Thuc. 7.82, and more often for people in a state of helplessness: Isoc. 19.39, Andoc. 1.61).

Again, the speaker's story has a Sophoclean counterpart. In Oedipus at Colonus, Creon, the uncle and soon-to-be-guardian of Ismene and Antigone, kidnaps Ismene and takes her away from her sister and father in Attica to his home in Thebes (818–47). Timandrus, being the children's ἐπίτροπος, could well have been an uncle and was almost certainly a male relative.¹⁶ Creon later kidnaps Antigone too, dragging her away onstage. ἀποσπάω is not used but equivalents are: ξυναρπάσας (819), ἐξάγειν | ἄκουσαν (826–7), ἀφήσεις (835), ἀφέλκομαι (844), πρὸς βίαν πορεύομαι (845).

Hyperides portrays the separation of the siblings as a great cruelty, which suggests a shared understanding among those he addresses of the value to children's well-being of maintaining sibling relationships and the suffering caused by their severance. Given high infant mortality rates, it is likely that many jurors would have lost a sibling in childhood;¹⁷ Hyperides may thus be drawing on the jurors’ memories to arouse their grief and indignation over siblings lost to each other while still alive. If so, it is notable that he focusses on the thwarted relationship between the sisters, not between the sister and her brother(s); the jurors are expected to sympathize across the gender division, a point to which I return. The immediacy of the separation is also a grievance: it happened ‘right in the first year when their father died’ (εὐθὺς τῷ πρώτῳ ἐνι̣αυτῷ ὡ̣(ς) <ὁ> π̣(ατὴ)ρ̣ αὐτῶ̣(ν) ἐ̣τ̣ε̣λ̣ε̣ύ̣τησεν, 63–4). Though it seems that the Athenian mourning period lasted until the rites performed thirty days after the death,¹⁸ it may be that the first year was felt to be a particularly sensitive time. This fits with the possibility that the annual rites performed for the individual dead (ἐνιαυσία) were celebrated on the anniversary of the death (as opposed to the anniversary of the person's birthday or funeral).¹⁹ In another prosecution of a guardian, after the dead man's wife and children were told of his death they continued to live in his house ‘for the first year’ (τὸν … πρῶτον ἐνιαυτόν, Lys. 32.8). Only after the first year did their guardian Diogeiton split them up, perhaps making him a more sensitive guardian, at least initially.²⁰

Hyperides’ emphasis on the relationship between the sisters is remarkable. The emotional climax of the fragment is the sisters’ inability to recognize each other after years of separation (42–5):

Τίμανδρος τοίνυν τούτου α̣ὐτ̣ο̣ῦ̣ γε̣ αἴτιος γέγον(εν) ὥστε τ̣ὰς̣ μὲν ἀδελφὰς̣ ἀλλήλας μὴ ἀναγνῶν(α̣ι̣) μήτε ἐν ὁ̣δ̣ῶ̣ι̣ μήτε ἐν ἱ̣ε̣ρ̣ῶ̣ι̣ ἰδούσας

And so Timandrus was guilty of this: that the sisters did not recognize each other when they saw each other in the street nor at the temple

Hyperides attests to the potential closeness and importance of sororal relationships by showing us one thwarted. Though the (failed) interaction which he describes happened in adulthood, it is used to argue for the cruelty of their treatment as children and relies on the pathos of a childhood spent apart; his interest is in the emotions and affective relationships of children, girls at that.

Why does Hyperides specify the street and the temple? The speaker may be revisiting details from a fuller, earlier account—perhaps the διήγησις probably given by another speaker—which described how the younger sister visited Athens, or the older sister visited Lemnos (as her brother did, and perhaps with him) and the two women, now in their twenties and probably married with children of their own, walked around the same city without knowing each other. This lost narrative would have been a rare account of female sociality and interaction in public spaces, both sacred and not. Another possibility is that this narrative of non-interaction is introduced here for the first time; the specific details encourage the jurors to imagine precise situations and bring the sisters’ thwarted reunion to life. Streets and temples would have been among the most public places accessible to women. By specifying these places, emphasized by their coordinating negatives (μήτε ἐν ὁ̣δ̣ῶ̣ι̣ μήτε ἐν ἱ̣ε̣ρ̣ῶ̣ι̣), the speaker stresses that Timandrus’ separation of the girls deprived them not only of the domestic intimacy which female relations could enjoy but even of the pleasure of knowingly encountering each other in public (temples and sanctuaries were favourite sites for women's social interaction).²¹ Further, the mention of the temple brings the gods into the narrative, which allows the jurors to imagine divine disapproval.²² The extant narrative contributes to our increasingly complex picture of female ‘public’ visibility and invisibility in Classical Athens:²³ it suggests that it was unremarkable that free élite Athenian women would encounter unfamiliar women in streets and temples, and that streets and temples could be sites of female visibility.

The abducted woman's brother did not recognize her on sight either, but no emotional point is made of this. Her sister's non-recognition is contrasted with her brother's recognition; that he too did not recognize her on sight is added as an explanatory detail (47–9):

τὸν δὲ ἀδελφὸν τουτονῒ Ἀκάδημον ἀναγνωρίσαι τὴν ἑαυτοῦ ἀδελφήν, ἐλθόν(τα) δὲ̣ ε̣ἰ̣ς̣ Λ̣ῆ̣μ̣ν̣ο̣ν̣ μ̣ὴ̣ γ̣ν̣ῶ̣ναι̣ ἰδόν(τα).

but it took her brother here, Academus, to recognize his own sister, though he had come to Lemnos and did not know her when he saw her.

The girl's relationship with all three siblings was severed, but Hyperides foregrounds the thwarted sororal relationship even over her relationship with Academus, present in court (τουτονί, 47). We may infer that if a girl had both brother and sister, her childhood relationship with her sister was expected to be the stronger one. Certainly Hyperides’ use of the thwarted sororal relationship as an illustration of Timandrus’ cruelty—the central illustration, at least in the text that we have—suggests that the jurors would generally have accepted that sororal relationships were special and worth preserving.

Golden laments the lack of evidence for sororal relationships in Classical Athens, which he attributes in part to selective exposure of baby girls (to which we might add gendered undernourishment)²⁴ but mainly to ‘our sources’ systematic scanting of women’.²⁵ Coo refines this judgement, noting that ‘while we have little evidence for the real-life relationship between sisters in Classical Greece, we find an abundance of sisters in myth and literature’²⁶ and compiling a list of 17 extant and 27 fragmentary or lost tragedies which certainly or probably featured pairs or groups of sisters. Sophocles’ portrayals of sororal relationships in Oedipus at Colonus, Antigone and Electra are well known. Coo argues further for the centrality of sisterhood in Sophocles’ Tereus and Euripides’ Erechtheus. Her list and analysis attest to an emotional interest among Athenians in sororal relationships, upon which Hyperides—sometimes using tragic motifs—draws.

Among the evidence for ‘real-life sisters’, Golden notes a monument commemorating two sisters, Melino and —ostrate (IG II² 5673, c.350, Piraeus):²⁷

καὶ ζῶσαι̣ πλούτου πατρικοῦ μέρος

εἶχον ὁμοίως, | τὴν αὑτῶν φιλίαν κ̣α̣ὶ

χρήματα ταὔτ’ ἐνόμιζον.

οὐδ]ένα λυπήσασα τέκνων δ’ ἐπιδοῦσ-

α ἔτι] παῖδας | τ̣ῆς κοινῆς μοίρας πᾶσ-

ιν ἔχει] τὸ μέρος.

——οστράτη, Μελινώ

——νος Ἀναφλυστ[ίου]

While they were alive, they had an equal portion of their father's wealth;²⁸ they considered their affection and wealth the same.

She caused grief to no one, but having even seen children's children, she has a portion of the fate that is shared by everyone.²⁹

——ostrate, Melino

[daughter(s)] of ——on of Anaphlystus

The couplets were inscribed at the same time. Kaibel suggests that the sister praised in the second couplet died first (—ostrate, named first?) and that her sister (Melino?) set up the monument for her before being buried in it herself; in his view, Melino had the couplet referring to both of them inscribed in preparation for her own death at the time she commemorated —ostrate.³⁰ The sisters were almost certainly ἐπίκληροι, which likely motivated the emphasis on their inheritance. This status may have given them a sense of their own significance, along with some level of access to or influence over financial resources, which enabled the erection of such a monument, into which Melino must have had input: this is a sister's description of her sororal relationship. The phrase τὴν αὑτῶν φιλίαν κ̣α̣ὶ χρήματα ταὔτ’ ἐνόμιζον (‘they considered their love/affection and wealth the same’) seems to mean ‘they valued their relationship as much as their money’.³¹ Though —ostrate, a grandmother, was at some point married, the sisters apparently maintained a strong relationship throughout their long lives.

It is this kind of sororal relationship, close and persisting into adulthood, that the speaker of Against Timandrus accuses Timandrus of thwarting. The speech's emotional engagement with real-life sisters makes it extremely unusual in Attic forensic oratory. Elsewhere in the orators there are a few pairs or groups of sisters, but we are given almost no sense of their relationships with each other: mostly speakers just note their marriages and dowries.³² In a handful of cases, sisters are presented in a way which implies an affective or cooperative relationship. In Isaeus 6, the two daughters of Euctemon, who have three brothers (§6), are situated together socially within their family and their father's phratry and deme (§10). More importantly, they are said to have come together with their mother after their marriages to bury their estranged father when he died (§§40–1).

In Isaeus 12.5, the speaker portrays two men's willingness to testify to the citizen status of their brother-in-law Euphiletus as facilitated by the shared attitudes of the sisters married to them. The speaker, Euphiletus’ paternal half-brother, claims that these men would not have given false evidence in favour of Euphiletus because Euphiletus’ mother was their wives’ stepmother:

εἰώθασι δέ πως ὡς ἐπὶ τὸ πολὺ διαφέρεσθαι ἀλλήλαις αἵ τε μητρυιαὶ καὶ αἱ πρόγονοι· ὥστε εἰ οὗτος ἐξ ἄλλου τινὸς ἀνδρὸς ἦν τῇ μητρυιᾷ καὶ οὐκ ἐκ τοῦ ἡμετέρου πατρός, οὐκ ἄν ποτε, ὦ ἄνδρες δικασταί, τοὺς ἑαυτῶν ἄνδρας αἱ ἀδελφαὶ μαρτυρεῖν [εἴασαν καὶ] ἐπέτρεψαν.

and stepmothers and daughters from a former marriage mostly tend to be at odds with each other, so if this man had been [born] to the stepmother from some other man and not from our father, our sisters would never, jurymen, have allowed their husbands to be witnesses.

The speaker implies a unity of feeling between the sisters and a united response to the situation.

Demosthenes 27 provides indirect evidence for an affective relationship between sisters, Demosthenes’ mother Cleobule and her sister Philia. Demosthenes says that Philia's husband helped Cleobule by acting as her advocate (14–15), and Philia's son married Cleobule's daughter ([Plut.] Mor. 847C), though her dowry had been appropriated by her guardians (Dem. 27.5–6, 65, passim).³³ Philia's husband's and son's support for Cleobule and her daughter suggests a close, supportive relationship between Cleobule and Philia herself.³⁴ Lys. 3.6–7 offers a rare glimpse of sisters in childhood: the speaker mentions his fatherless nieces, who were allegedly intruded upon in the γυναικωνῖτις, the women's space of the house. He describes them as παῖδας κόρας καὶ ὀρφανάς (‘young girls, orphans’) and says that ‘they have lived such orderly lives that they are ashamed to be seen even by their [male] relatives’ (οὕτω κοσμίως βεβιώκασιν ὥστε καὶ ὑπὸ τῶν οἰκείων ὁρώμεναι αἰσχύνεσθαι). We infer that the sisters shared a living space and lifestyle, the kind of συνήθεια (‘shared life’, lines 36–7 of our fragment; cf. τὸ̣ συντρόφους … εἶναι, ‘being brought up together’, 37) that later in the speech Hyperides says leads to εὔ̣νοιαι (‘affectionate feelings’, 36) among family members; the second part of this article discusses this claim in detail.³⁵

Tchernetska and her co-editors suggest that ‘Akademos, now head of the family, was in a position to establish the identity as well as the whereabouts of his younger sister, but when he first saw her in Lemnos, he failed to recognize the young woman he had last seen over thirteen years ago when they had both been orphaned in early childhood.’³⁶ Though Lemnos might have a particular ethical resonance given that a ‘Lemnian’ deed was proverbial for a cruel deed,³⁷ it is more relevant that in fourth-century Athenian courts Lemnos was associated with estranged relatives and difficult-to-establish identities of family members. Lemnos was home to many Athenian κληροῦχοι, emigrants who lived some or all of the time in Lemnos and held land there, but retained Athenian citizenship and membership of the deme and tribe to which they had belonged; Timandrus, whose home was in Lemnos (26–7), was probably one.³⁸ ‘Athenian litigiousness did not fail to seize the opportunities of fraud presented by the existence of a body of citizens practically unknown in Attica.’³⁹ In Isaeus 6, the speaker's opponents claim that the testator had sons with a second wife from Lemnos; the speaker contends that no such woman existed and that the claim her father came from Lemnos was a convenient excuse for his untraceability. In Terence's Phormio, one character, like Academus in Against Timandrus, travels to Lemnos to search for a female relative from whom he has been separated and who has since grown to adulthood (Phorm. 568–72). Donatus’ commentary on the play strongly suggests that its plot is closely based on Apollodorus’ Epidikazomenos.⁴⁰ In this case the lost relative is the man's illegitimate daughter: during a previous stay in Lemnos, he had contracted a secret second marriage under a false name, though already married in Athens, and had a child.

Recognition scenes between long-lost family members were frequent in tragedy and, under its influence (particularly that of Euripides), New Comedy. Particularly relevant is the Electra/Orestes recognition scene between a brother and a sister separated in early childhood. Euripides’ and Sophocles’ Electras and Aeschylus’ Choephori all feature such a scene; in none of them do the pair recognize each other on sight. Euripides makes explicit what is implicit in Hyperides (Eur. El. 282–4); the language used in the two passages is similar:

Ὀρέστης: εἴθ᾽ ἦν Ὀρέστης πλησίον κλύων τάδε.

Ἠλέκτρα: ἀλλ᾽, ὦ ξέν᾽, οὐ γνοίην ἂν εἰσιδοῦσά νιν.

Ὀρέστης: νέα γάρ, οὐδὲν θαῦμ᾽, ἀπεζεύχθης νέου.

Orestes: If only Orestes were nearby to hear this.

Electra: But stranger, I would not know him if I saw him.

Orestes: No wonder, for you were both young when he left you.

Euripides’ reworking of Aeschylus’ rendition (Eur. El. 215–431, 508–84; Aesch. Cho. 212–45) suggests the popularity of the theme, but where in Aeschylus’ and Euripides’ versions the reunion scenes come early and are fairly brief, in Sophocles’ Electra the recognition between long-lost brother and sister is much delayed and forms the emotional high point of the play.

THE SOCIAL CONSTRUCTION OF THE FAMILY RELATIONSHIP

The speaker claims that ‘affectionate feelings between people exist because of shared lives and their having a shared upbringing rather than because of shared blood’ (αἱ … εὔ̣νοιαι τ̣ο̣ῖ̣ς̣ ἀνθρώπ̣ο̣ι̣ς̣ εἰσὶ διὰ τὴν συνήθειαν καὶ τὸ̣ συντρόφους αὐτοὺς εἶναι μᾶλλον ἢ διὰ τὰς συγγενείας, 35–8).⁴¹ The ‘evidence’ (τ̣εκμήριον) for this claim, he continues, is that ‘fathers would not be fond of their children if they [sc. the children] had not been brought up by them from infancy, if someone had immediately dragged them off [text missing: ‘and kept hold of them’?], nor would children be fond of their parents if they had not been brought up by them’ (38–42). Hyperides’ absolutist examples here recommend the translation of μᾶλλον ἤ as ‘rather than’ instead of ‘more than’. How typical was such a sentiment in Classical Athens? Let us compare the analyses of two contemporary authors who engage with the question of the nature, origin and significance of affection between kin: Aristotle and Isaeus.

In the Nicomachean Ethics, Aristotle, without a prosecution case to make, offers a different account (1161b3–4).⁴² As elsewhere in this work, he begins his analyses from, and tests them against, ἔνδοξα (‘reputable views’; the views of the majority or, in certain cases, experts); he often begins from definitions he claims are generally accepted or cites proverbs to support his arguments.⁴³ Aristotle's concern is with φιλία, friendship or affection, rather than εὔνοια, the term Hyperides uses (in the plural). Aristotle asserts that εὔνοια (singular) is commonly used to mean ‘wishing another's good’; as such, φιλία is commonly understood to consist in reciprocal εὔνοια (1155b).⁴⁴ Outside philosophy, εὔνοια describes an established positive interpersonal emotional attitude ranging from the weaker ‘goodwill’ (for example in diplomatic contexts) to a more earnestly felt, devoted affection as between close friends and family members.⁴⁵ A comparison between Hyperides’ account of εὔνοιαι and Aristotle's of φιλία is therefore still productive.

Aristotle argues that all affection (φιλία) inheres in κοινωνία (community, having something in common), but categorizes affection between family and friends as different from friendship between fellow citizens, fellow tribe members and colleagues, because it is less contractual. He argues that all family affection (ἡ συγγενικὴ [φιλία]) depends upon paternal love,⁴⁶ ‘for parents love their children because the children are something of themselves, but children love their parents because they are something from them’ (οἱ γονεῖς μὲν γὰρ στέργουσι τὰ τέκνα ὡς ἑαυτῶν τι ὄντα, τὰ δὲ τέκνα τοὺς γονεῖς ὡς ἀπ᾽ ἐκείνων τι ὄντα). Parents are ‘close’ to their children, he continues, because the children have come from and belong to their bodies; therefore, parents love their children as soon as they come into existence, while children require knowledge, or perception, to love their parents. Parents love their children as being part of themselves; children love their parents as being their source; brothers love each other as sharing a source. This biological or essentialist explanation differs from Hyperides’ claim that fathers separated from their children in infancy would not be fond of them, and children brought up by people other than their parents would not be fond of their parents (38–42): Aristotle asserts that mothers who give their children to be raised by others (διδόασι τὰ ἑαυτῶν τρέφεσθαι) still love them (καὶ φιλοῦσι), though they do not expect to be loved by them in return (ἀντιφιλεῖσθαι δ᾽ οὐ ζητοῦσιν, 1159a). For Hyperides in this fragment, unlike for Aristotle, ‘sharedness’ between siblings (Aristotle's κοινωνία, Hyperides’ συνήθεια and τὸ̣ συντρόφους … εἶναι) is not dependent on the children's relationship to their parents.

Modern kinship studies tend to de-emphasize the role of biology that was so important to Aristotle. Most insistent is Sahlins, who argues that kinship is entirely cultural, and counter-intuitively uses this passage as an illustration: ‘Anchored as it may be in concepts of birth and descent, Aristotle's discussion of kinship at once goes beyond and encompasses relations of procreation in larger meanings of mutual belonging that could just as well accommodate the various performative modes of relatedness. Or so I read the possibilities of his sense of kinship as “the same entity in discrete subjects.”’⁴⁷ This overstates the flexibility of Aristotle's conception of kinship: Aristotle's ‘same entity in discrete subjects’, his ‘being of [the other, where in fact, “the other” is always and only the parent]’, is biological: parents are ‘close’ (οἰκεῖος) to their children because the children have come from and belong to their bodies, like a tooth or hair (1161b2). However, Aristotle adds a socio-cultural element, claiming that φιλία between brothers is much increased (though not generated) by their shared upbringing and similar age (μέγα δὲ πρὸς φιλίαν καὶ τὸ σύντροφον καὶ τὸ καθ᾽ ἡλικίαν); he cites the saying οἱ συνήθεις ἑταῖροι (literally, ‘those who spend time together [become] friends’). For this reason, affection between brothers is similar to affection between friends (διὸ καὶ ἡ ἀδελφικὴ [φιλία] τῇ ἑταιρικῇ ὁμοιοῦται).

When Aristotle is speaking in apparently general terms, he predominantly uses language specifically relating to men. Hyperides does the same in this speech: though he says that slave-traffickers and slave-dealers try not to separate a mother from her little children (31–5), the first ‘proof’ he gives of his explanation that separation prevents family affection is that fathers (π̣(ατέ)ρ̣(ε)ς) would not love their children if they had been brought up by someone else (38–40). In seeking the jurors’ emotional engagement, the speaker appeals to them in terms of men's emotional experiences and attachments. He continues, however, ‘nor would children love their parents’ (τοὺς γονέας, 41–2);⁴⁸ the elision of gender in Hyperides’ argumentation is discussed below. In Aristotle, ‘the word for brothers, “adelphoi”, can be translated “siblings”, but [Aristotle] probably takes it for granted that our attention will be on male adelphoi.’⁴⁹ It is hard to know how far Aristotle would have considered his analysis applicable to girls or women. Hyperides’ attention to the emotional lives and affective relations of young girls in Against Timandrus is unusual in Attic literature outside tragedy.

The question of what constitutes family closeness is thematic in Isaeus, who deals extensively with family relationships in the context of disputed inheritances, often arguing that his client was a closer relative to the dead man than to his opponent. Isaeus uses a framework in which being related to someone is characterized and constituted by blood kinship (συγγένεια) and certain shared life experiences, like cohabitation and shared religious practice (analogous to Hyperides’ συνήθεια), and shared upbringing (cf. Hyperides’ and Aristotle's τὸ σύντροφον). Unlike Hyperides and Aristotle, Isaeus portrays emotional connection not as arising from biology and/or shared life experiences but as complementary to them in constituting family closeness, and demonstrated by behaviours such as raising a child in one's home and including her or him in religious rites.⁵⁰ This is because his arguments sometimes rely on reconstructing the supposed intentions of the deceased, where their alleged behaviour is taken to indicate their feelings. For Isaeus (and, he must assume, the jurors), feelings of affection between family members, combined with and demonstrated by certain behaviours, attest to the existence of a family relationship. Family relationships themselves are never understood as solely biological, not least because of the prevalence of adoption. Where necessary, Isaeus prioritizes shared life experience over biology.

Most relevant is Isaeus’ frequent foregrounding of the familial context of a child's upbringing. In Isaeus 8, part of the proof of the legitimacy of Ciron's daughter by his first wife (that is, of the family relationship between Ciron and his daughter) is that Ciron brought her up with his second family: καὶ ἐκείνην τε ἔτρεφε παρὰ τῇ γυναικὶ καὶ μετὰ τῶν ἐξ ἐκείνης παίδων (‘and he brought her up with his [second] wife and alongside his children by that wife’, 7); she was τρεφομένην ἔνδον (‘brought up in his home’, 14). Ciron was willing to spend money on bringing her up and publicly to acknowledge her as a member of his household and family. Isaeus 9 describes how a remarried woman brought her son Astyphilus from her first marriage into her second marriage; when she had a son, Theophrastus, with her new husband, the boys were educated together (28). Theophrastus uses this to argue for his close family relationship to Astyphilus and therefore his claim to inherit. In Isaeus 7, Thrasyllus claims the estate of Apollodorus, who he says adopted him. The adoption was not finalized before Apollodorus’ death, so Thrasyllus must argue that Apollodorus intended to adopt him and that his own claim trumps that of Apollodorus’ first cousin. The thrust of Thrasyllus’ argument is that Apollodorus was full of goodwill to him and ill-will to his cousin. The context of Apollodorus’ goodwill towards Thrasyllus was that Thrasyllus’ grandfather Archedamus brought up Apollodorus. As Thrasyllus tells it, after Apollodorus’ father died, Apollodorus went to live with his uncle and ἐπίτροπος, while his mother was remarried to Archedamus. Archedamus saw that the ἐπίτροπος was financially abusing his wife's son and ‘brought him to live with him and the boy's mother and brought him up when he was a child’ (ἔτρεφέ τε αὐτὸν παῖδα ὄνθ᾽, ὡς ἑαυτὸν καὶ τὴν μητέρα κομισάμενος, 7; ἔτρεφέ is promoted by a kind of hysteron proteron). Apollodorus’ goodwill towards Archedamus is said to have come from being brought up by him, not from blood relationship—since there was none.

In the context of the speech, Hyperides’ claim that ‘affectionate feelings between people exist because of shared lives and having a shared upbringing, rather than because of shared blood’ casts Timandrus’ alleged separation of the siblings as the wholesale destruction of a family relationship, which could never have been in the children's interests. This is part of the rhetorical strategy which probably set out to disable a claim by Timandrus that it was in the separated sister's interest to be raised in Lemnos and in the other siblings’ interest to be raised in Athens.⁵¹ In the context of fourth-century Attic thought on kinship, Hyperides’ argument is not entirely novel, but its formulation is extreme. It is striking that he expected the jurors to accept it.

ENSLAVED FAMILIES AND SOCIALITY

Jones and Schmitz have attempted to assess the accuracy of Hyperides’ claim that the Classical Athenian slave trade avoided the separation of siblings and of mothers with young children.⁵² Schmitz finds one probable example of an enslaved Thracian woman being sold with her two children as a single ‘item’ in the auction lists of the property of the Hermocopidae,⁵³ and points out that the possibility of identifying an enslaved family in these lists suggests that the other enslaved people listed did not live or were not sold ‘as families’.⁵⁴ Andocides informs us that an enslaved man named Andromachus and his brother, also enslaved and living in the same household, were witnesses to the profanation of the Mysteries (1.11–12).⁵⁵ They may have been captured, sold and bought together, or (less relevant to the question at hand) born in their enslaver's home to the same enslaved mother. At any rate, they were still together in adulthood. To these examples we may add the case of Midas and his sons Pancalus and Procles/Polycles (the source gives two different names) in another Hyperides speech: a man infatuated with one of the boys intended to buy and free him, and was persuaded to buy his father and brother too (3.4–6, 24). The late fourth-century φιάλαι ‘inscriptions’, traditionally understood as records of manumissions, seem also to include some family groups.⁵⁶

However (a)typical the preservation of enslaved families was, the implications of Hyperides’ claim and of its place in his argument have not been much addressed. The argument at the centre of the fragment is as follows: it was cruel of Timandrus to separate the siblings; separating families is so cruel that even slave-traders and slave-traffickers, apparently a byword for cruelty, do not do such a thing to enslaved people; the reason they do not do this (γ(άρ), 35) is that affection between family members does not arise ‘naturally’ from blood kinship but from lives—particularly childhoods—spent together. The third stage of the argument is compressed: the intermediary stage must be something like ‘traders and traffickers do not do this because they acknowledge (actual or potential) affective relationships between enslaved family members, which would be destroyed or thwarted by separation’. The implication of this stage is significant: it humanizes enslaved people. The speaker recognizes—and implies that the jurors recognize—the capacity of enslaved people to form profound affective relationships based on shared histories, comparable to family relationships among free people. Not only does the speaker (and, by implication, the jurors) recognize the relationships, but he (and, by implication, the jurors) professes to value them and their effect on the enslaved people who form them. The argument is based on the premise that εὔνοιαι should ideally be allowed to (even facilitated in) enslaved people.

Arguing that the hallmark of enslavement is the ‘social death’ of the enslaved person, which includes ‘natal alienation’, the forcible disregarding of that person's relationships, Patterson explains that ‘this does not mean that he or she did not experience or share informal social relations … [However,] these relationships were never recognized as legitimate or binding … [P]arents were deeply attached to their children, but the parental bond had no social support … slaves had no custodial claims or powers over their children, and children inherited no claims or obligations to their parents.’⁵⁷ This aspect of slavery was recognized and discussed in Athens, for example by Callicles in Plato's Gorgias (483a–b):⁵⁸

oὐδὲ γὰρ ἀνδρὸς τοῦτό γ᾽ ἐστὶν τὸ πάθημα, τὸ ἀδικεῖσθαι, ἀλλ᾽ ἀνδραπόδου τινὸς ᾧ κρεῖττόν ἐστιν τεθνάναι ἢ ζῆν, ὅστις ἀδικούμενος καὶ προπηλακιζόμενος μὴ οἷός τέ ἐστιν αὐτὸς αὑτῷ βοηθεῖν μηδὲ ἄλλῳ οὗ ἂν κήδηται.

For this suffering, that is, being wronged, is not the part of a man, but the part of some slave, for whom it is better to be dead than alive, who, if he is wronged or trampled in the mud, is unable to help himself, or anyone he cares for.

Though the enslaved person experiences an attachment (κήδηται) to another person, he is unable to realize the relationship by carrying out its ethical obligations. Callicles makes this comment as part of a challenge to Socrates’ mode of argumentation, but despite Callicles’ cynicism both he and Socrates seem to treat it as understood that an enslaved person was incapable of decisively supporting his or her loved ones,⁵⁹ and that this set the enslaved person apart from the (free) ‘man’ (ἀνήρ).

This combination of attachment and powerlessness could be weaponized by the enslaver. The author of the pseudo-Aristotelian Oeconomica recommends that in order to ensure the obedience of enslaved people δεῖ … καὶ ἐξομηρεύειν ταῖς τεκνοποιίαις (‘it is necessary to take hostages by means of [their] begetting children’, 1134b). That is, enslaved people should be allowed to have children with each other both so that the enslaver might threaten the parents with harm to or separation from their children and so that the parents might be discouraged from escaping and incurring the cost of separation from their children.⁶⁰

Here it is worth drilling into the specific claims Hyperides makes. He asserts that those who have captured people in war (οἱ κατὰ πόλεμον ἐγκρατεῖς γιγνόμενοι τ(ῶν) σωμάτων, 29–30) try as far as possible to sell them ‘as a household/family’ (κατ’ οἰκίαν, 30). He then refers to another group of people: slave-dealers and slave-traffickers (οἱ … ἀνδραποδοκάπηλ(οι) καὶ ἔμποροι, 31–2),⁶¹ who, he claims, despite their unscrupulous profiteering, try where possible to sell young siblings or mother-and-infant groups (ἀδ̣ε̣λ̣φὰ̣ παιδάρι(α) … ἢ μ̣η̣τέρα καὶ παιδία, 33) together, even if it means incurring a loss.⁶² From later accounts of ἀνδραποδισμός, Gaca shows that, though soldiers generally did not want infants who were likely to die in transport, ‘leav[ing] infants with their young mothers can be expedient for inducing the mothers to go along without resistance’.⁶³ If Hyperides’ claim that such separations were avoided is true, this was a likelier motivation than respect for affective bonds. The claim about young siblings is harder to evaluate; was compliance a factor here too? Its rhetorical value lies in the directness of the comparison to his (free) clients, who were separated from their sibling.

Hyperides’ elisions and assumptions around gender and family are illuminating. Though he claims that soldiers ideally sold andrapodized people ‘as a household/family’, by this stage of ἀνδραποδισμός the household/family of the victims had been destroyed, the adult male members killed and the older female or very young members left behind. Either Hyperides elides this reality—surely familiar to the jurors—or he is using the phrase to mean ‘along with any other οἰκία members selected for sale’. Though men were not andrapodized, they were trafficked into slavery by other methods, but in speaking of family preservation by traffickers and traders Hyperides limits his claim to child siblings and mothers with infants. In neither case is the οἰκία preserved, nor are many family relationships—including paternal relationships, the value and potency of which Hyperides uses in the following sentence to explain this practice. If there is truth in his claims about family preservation in the slave system, Hyperides is extrapolating from age-, gender- and situation-specific practices to make a falsely universalizing claim about respect for all family relations among enslaved people to serve his argument about the wickedness of Timandrus’ separation of his clients from their sister.

Nevertheless, he makes the claim. In Against Timandrus, the speaker's recognition of the existence and value of social ties between enslaved people throws into sharp relief the intellectual and moral paradox of slave societies: enslavers recognized the humanity of people they simultaneously insisted were subhuman.⁶⁴ This paradox is apparent in the fragment's language: where Hyperides argues that enslaved families are kept together where possible because family feeling comes from shared experience rather than biology (27–38), he uses the word for ‘slaves’ which most reduces them to their biology: τῶν σωμάτων (‘bodies’, 29–30).⁶⁵ Among many other instances, this is the term used in Dem. 47.12 and 15, where the speaker discusses rendering an enslaved woman for torture, in which the slavehood of a person is most brutally enacted. Yet in this speech it is where Hyperides discusses ‘humane’ treatment of enslaved people and acknowledges their humanity that he uses this dehumanizing term; where he recognizes their emotions, relationships and histories he talks about them as bodies.

The passage of Aristotle discussed earlier offers an analysis of affective relationships between free and enslaved people which runs up against the same paradox (Eth. Nic. 1161a–b):

φιλία δ᾽ οὐκ ἔστι πρὸς τὰ ἄψυχα οὐδὲ δίκαιον. ἀλλ᾽ οὐδὲ πρὸς ἵππον ἢ βοῦν, οὐδὲ πρὸς δοῦλον ᾗ δοῦλος. οὐδὲν γὰρ κοινόν ἐστιν· ὁ γὰρ δοῦλος ἔμψυχον ὄργανον, τὸ δ᾽ ὄργανον ἄψυχος δοῦλος. ᾗ μὲν οὖν δοῦλος, οὐκ ἔστι φιλία πρὸς αὐτόν, ᾗ δ᾽ ἄνθρωπος· δοκεῖ γὰρ εἶναί τι δίκαιον παντὶ ἀνθρώπῳ πρὸς πάντα τὸν δυνάμενον κοινωνῆσαι νόμου καὶ συνθήκης· καὶ φιλία δή, καθ᾽ ὅσον ἄνθρωπος.

but there can be no affection/friendship/social bond⁶⁶ (φιλία) or justice towards/with inanimate things; no, not towards a horse or an ox, nor towards a slave qua slave. For they [the imagined subject and the enslaved person] share nothing/have nothing in common (οὐδὲν γὰρ κοινόν ἐστιν): for a slave is an animate tool, as a tool is an inanimate slave. So in so far as he is a slave, there can be no affection/friendship/social bond towards/with him—but not in so far as he is a human being: for it seems that for any human being there is a kind of justice [owed?] to anyone who is able to share in law and contract, and so too the possibility of affection/friendship/a social bond, in so far as the object is a human being.

Various scholars have discussed Aristotle's not entirely satisfying distinction between relating to an enslaved person qua slave and qua human being.⁶⁷ Importantly for the questions addressed in this article, Aristotle seems to be thinking of the affective relationship of an enslaver (or at least a free person) with an enslaved person:⁶⁸ he might concede that there could be something κοινόν between one enslaved person and another, even if not between an enslaved person and a free person.⁶⁹ However, if Aristotle envisions affective relationships between enslaved people, that is not explicit. Hyperides’ insistence—for the purposes of his case—on the possibility and value of relationships between enslaved people, including children, goes well beyond Aristotle. Yet the ambivalence remains: Hyperides insists on enslaved people's claim to family relationships—or at least certain family relationships—but not their claim to freedom.⁷⁰ They are human beings with the capacity for relationships beyond the solely biological; at the same time, they are merely bodies.

CONCLUSION

This new fragment is studded with precious details—on family composition, movement within and between poleis, expectations around bereavement, and more—which enrich our picture of Athenian social and family life. Fundamentally, though, its narrative and argumentation transform our understanding of Athenian thinking and feeling about families, free and enslaved. Hyperides’ brief but evocative portrayal of sibling relationships and their severance, and his condemnation of the latter, are a rare contribution to our evidence for Athenian childhood relationships and emotions, and their perception by adults. Furthermore, the development of his argument comprises both a new, more extreme articulation of ideas implicit in Isaeus and Aristotle about the nature of family relationships—namely, the importance of shared upbringing and life experience in generating affective relations between kin—and the radical application of those ideas to enslaved people, whose affective relationships the speaker claims are widely recognized. The disjuncture between the recognition of these relationships and their forcible destruction as part of the practice of slavery illuminates the psychological and moral problems faced by a slave society.