Homo inter homines sum – The impor- tance of age for freedmen’s construction of masculinity in Late Republican and Early Imperial Rome1

1 — Acknowledgement is due to The Swedish Foundation for Humanities and Social Sciences (Stiftelsen Riksbankens Jubileumsfond), which financed my attendance to the conference where this paper was presented: Arachne viii: Ages, ageing, old age in the Greco-Roman world in 25-27 October 2017, and the research. I am grateful to the two anonymous readers for their valuable critique. Any errors that remain are my own. 2 — Petron. Sat. 39.4. Latin texts from the Brepolis LLT-A (Library of Latin Texts). All translations are my own, unless otherwise noted.

These are the words of the famous freedman Trimalchio, expressing that he is now able to enjoy the good life and that he is part of society, he is "a man among men". But what does this "a man among men" mean and why was it important for a freedman to be seen as a man?
In the Greco-Roman world the slave was seen as a boy and manumission was perceived as a maturity process, where the manumitted slave could be seen as coming of age. This article will investigate how age can be used as a means to study how Roman male freedmen were positioned in relation to the ideal Roman masculinity and how this positioning can be seen as an expression of (gender) identity and status. The article addresses these questions by doing a critical close reading of a sample of Late Republican and Early Imperial Roman material, applying masculinity theory and theories of age and gender, drawing on modern conceptions and constructions of masculinity. A selection of iconographic, epigraphic and literary sources has been made, where I have chosen to present material that illustrates discourses on masculinity connected to age.
As a key concept to discuss the material, Connell's concept "hegemonic masculinity", as an ideal and normative masculinity, is used. To use masculinity theory is a rather new method within the study of ancient social history. With the exception of my research on Roman freedmen and virtus 3 , it has not been used to study Roman freedmen and neither has it been connected to age when studying Roman masculinity. In fact, theories of multiple masculinities have rarely been applied in research on Roman masculinity as earlier studies on the Republican and Imperial period have focused almost exclusively on elite (hegemonic) masculinity 4 .
To begin with, I will briefly explain theories of hegemonic masculinity and age connected to gender as these concepts will be used to understand and discuss the importance of age for Roman freedmen's construction of masculinity.

Gender and age
The concept "hegemonic masculinity" was first used by R. W. Connell, more than 30 years ago in her Gender and Power, and since then it has been further refined by her and James W. Messerschmidt 5 . Connell and Messerschmidt argue that a variety of masculinities coexists in a society: a dominant, "hegemonic masculinity", performed by an elite, and "subordinate masculinities" 6 . The hegemonic masculinity is normative, 3 -Hagelin 2019. 4 -This problem is addressed by McDonnell 2006, 166, but is not further discussed by him. The studies of Greek, Late Roman and early Christian masculinities are more diversified, see e.g. Berg 2010;Masterson 2014;Stewart 2016;Conway 2008. 5 -E.g. Connell 1987Connell , 1995Connell and Messerschmidt 2005;Messerschmidt 2012. The concept has been criticized, see e.g. Demetriou 2001;Whitehead 2002, 88-96. 6 -Connell 1995. The idea of a plurality of masculinities is further developed into a even though only a minority of men really enacts it, as it is seen as "the most honoured way of being a man" and "it requires all other men to position themselves in relation to it" 7 . The hegemonic masculinity must be understood neither as a fixed character type nor as a fixed, transhistorical model. Rather, the hegemonic masculinity is the masculinity that upholds the hegemonic position in a given pattern of gender relations, a position that can always be questioned. The hegemonic masculinity is always subject to change and older forms of masculinities can be displaced by new ones, and also non-hegemonic patterns of masculinities can be incorporated into the hegemonic. In this way, it is a cultural ideal that can change according to time and space 8 . The Roman hegemonic masculinity belonged to the elite and the public sphere, as masculinity was defined by public performances. Roman masculinity involved issues of body, dress, sexuality, social performance and competition and it was continuously judged and scrutinized by others 9 . It very much acted in accordance with the way masculinity is constructed and enacted in a modern society according to the theory of hegemonic masculinity, and as is seen in other studies on contemporary masculinities. In modern masculinity theory, age plays an important part, as it connects masculinities with physical appearance and bodily experience 10 . Indeed, the body in itself is essential for the construction of masculinity 11 . As stated by Whitehead: "The male body can be understood [...] as the place from which masculinities appear both as illusion and as materiality" 12 . Also in ancient Rome age was an important factor in how an individual was perceived and in how gender was constructed, and the interrelationship of gender and age in antiquity in terms of agency or individual empowerment must be recognised, as pointed out by Harlow and Laurence 13 . In addition, there are similarities between gender and age system of three different levels of hegemonic masculinities, i.e. global, regional and local in Connell and Messerschmidt 2005. 7 -Connell and Messerschmidt 2005, 832. 8 -Connell 1995Connell andMesserschmidt 2005. 9 -McDonnell 2006, 165-166; this is illustrated in the studies of Gleason 1995 andGunderson 2000. 10 -See further Drummond 2007, 10-13. 11 -See e.g. Connell 1987Connell 1995, 45-66;Connell 2000, 57-99. 12 -Whitehead 2002. See further discussion in his chapter on "materializing male bodies" 181-203, pages 199-202 focusing on age, discussing men's ageing bodies. See also Petersen 1998, 41-71, concluding: "To argue that the body is socially constructed is not to deny its materiality [...], but simply to recognise the fact that its specific materiality is a product of shifting relations of power/knowledge", Petersen 1998, 70. Petersen 116-117 criticises Connell's work Masculinities, where, according to Petersen, "the specific materiality of the body is taken as a given", Petersen 1998, 116. Following Butler (1993 Petersen argues that "the materiality of the body is an effect of power, or a discursive production", Petersen 1998, 122. 13 -Harlow & Laurence 2002 as social categories 14 . Both correlate with differences in social power and status, as expressed by Judith Kegan Gardiner: Age categories, like gender ones, correlate with differences in social power, status, and access to resources. Like gender, age categories form part of systems of power relations that shape and are shaped by all other social hierarchies [...]. Similarly, like gender, age appears to have an obvious biological basis in a way that many other social categories, like social class, do not, although biological sex manifests itself as more dimorphic than either such markers of gender as aggression or activity or the gradations of biological aging 15 . As Gardiner states, age also gives the impression to have a biological basis in the same way as gender does, although, like gender, it is in reality a social, not merely biological, category 16 , as will be discussed in the following. Modern scholars often divide age into different categories or meanings as biological or physiological, chronological/calendar and social, dividing the social and biological aspects. Sara Arber and Jay Ginn suggest that there are at least three meanings of age: Chronological (or calendar) age, physiological age and social age. Chronological age is essentially biological, but it has to be distinguished from physiological age since the latter is a medical construct, that refers to the physical ageing of the body. According to Arber and Ginn, "Social age refers to the social attitudes and behaviour seen as appropriate for a particular chronological age, which itself is cross-cut by gender" 17 . This means that there is an expected behaviour according to a person's age, as well as according to her or his gender, and that these two factors intersect in various ways. Arber & Ginn argue that the meaning of social age can be comparable to the notion of gender in some ways since social age is "socially constructed and refers to age norms as to appropriate attitudes and behaviour, subjective perceptions (how old one feels) and ascribed age (one's age as attributed by others). Age-based norms, like gender norms, are maintained by ideologies which are resistant to change" 18 . Thus, age as well as gender are important factors for a person's social status and identity. 14 -Gardiner 2001a proposes the advantages to feminist theory of "thinking gender through analogies with the multiple categories of age rather than in exclusively binary terms" and argues that such a theory could oppose "being a man" not to being a woman or a male homosexual but to being a boy, Gardiner 2001a, 16-17. See also e.g. Kramer 2017 on connecting the categories age and gender.

Slaves as pueri -boys, not men
Drawing on the assumption that age is an important factor in how an individual is perceived and in how gender is constructed, let us now go on to examine how age can be used as a means to study Roman male freedmen and masculine positions, starting by discussing the gender and social age of the slave.
Roman slaves were often referred to as pueri, boys, and the slave was seen as a perpetual child in many ways 19 . In fact, in the Digest the jurist Paul's first definition of the Latin word "puer" is slave 20 . The infantilization of slaves can also be seen in art, where adult slaves were often shown as about half the size of free people 21 . In his work Adults and children in the Roman Empire Wiedemann defines slaves as "in a sense children who had not been allowed to grow up" 22 . Golden, in his discussion on the dual sense of the Greek word "pais", meaning child as well as slave, thinks that "the Athenians saw slaves and children as occupying similar statuses within the structure of their society" and that this analogy is the reason for the extended use of the word "pais" in Greek. According to him, "[t]he relationship of children and slaves is therefore a special case of a more general phenomenon, the Athenian tendency to emphasize the similarities between subordinate social groupings rather than their differences" 23 .
Although there are many differences between the Athenian and Roman societies as regards slavery (as for example manumission practices and the status of freed slaves), I maintain that the argument of subordination can be applied to Roman society as well. Romans slaves were subject to the will of the master and they did not have the right over their own body. Slaves could not engage in a marriage that was legally valid, and any children they might have were the property of the slave's owner 24 . As expressed in Roman law, the slave was an instrument, owned and used by the master 25 .  Bradley 1984, Garnsey 1996and Watson 1987. See Blake 2012 on the notion of slaves as the "prosthetic limbs" of the master.

Puer -boy/slave life course
When a male slave gained his freedom he was given the right to his body and self. The act of emancipation was perceived as a maturity process where the slave boy, puer, grew up and earned his freedom, thus becoming an adult. Indeed, the status of the puer, meaning "boy", and the puer "slave", was similar in many ways in Roman society 26 . They were dependent and subject to another man's will, and the patria potestas and the potestas of the dominus, dominica potestas, were in many ways similar. The boy as well as the slave lacked the rights to citizen women, in that they could not marry and play the active part in a sexual relationship, and they could instead be sexually abused 27 . And as a Roman proverb expresses, nec puero gladium 28 , they had not the right to carry arms. In this way, the negative connotation of the word puer was the same for boys as for slaves 29 . Both groups were deprived of essential components for constructing masculinity, i.e. independence and freedom, the right to one's sexuality and to be the head of household, and the right to actively defend oneself.
When discussing the dual sense of the Greek word "pais", Golden states: "the word pais was applied to a social as well as a biological stage -that is, it indicated a person's relation to his society. To move from the status of pais, an Athenian boy had both to come of age and to be admitted to a deme" 30 . Also in Roman society, the life courses of the puer-boy and the puer-slave were similar, as they were both subject to a maturity process, where the final goal was to achieve citizenship. The boy as well as the slave were striving for the toga virilis, which was associated with a difference in behaviour 31 . Thus, in striving for the toga virilis, the boy and the slave also shared the positive connotation of the word puer 32 .
For a boy/young man, the toga symbolized the acquisition of full citizenship and freedom from paternal control and therefore it was also called toga pura or toga libera 33 . The toga virilis could be seen as a "visual 26 -On legal similarities and differences, see Pelloso 2018. 27 -Freeborn boys and young men were seen as sexual objects who had to be guarded against abuse (e.g. Plin. Ep. 3.3.4) and wore a bulla that distinguished them from slave boys, see further e.g. Richlin 2014. Nonetheless, it was against the law to sexually abuse freeborn children as it was considered stuprum, see e.g. Williams 2010, 103-136;Gardner 1986, 121-25;Richlin 1983, 224-225;Richlin 1993, 561-566;Dig. 47.11.1.2. 28 -Aug. Ep. 104.7. 29 -Maurin 1975, 223-225. 30 -Golden 1985, 94. 31 -Harlow & Laurence 2002, 69. 32 -Maurin 1975, 228. 33 -See Olson 2017, 48 and Dolansky 2008, 54-55 for references. According to Richlin 1993, 546, the fact that "toga virilis and toga pura were used interchangeably reminds us of the sexual overtones of impurus -definitely un-manly". Also Dolansky argues that pura may have ethic symbol of the young man's process of socialization" 34 . According to Olson, in her excellent study on masculinity and dress in Roman society, in ancient sources the toga is never characterized as manly clothing or masculine, but "reveals itself as a masculine garment when its absence is spoken of: on foreigners, slaves and women" 35 . Olson sees the toga as "quintessential Roman male attire", investing the sexually dominant male, whereas the toga-less person was penetrated, powerless and subjugated. "Persons who did not wear the toga were located outside traditional power structures", according to Olson 36 . However, it was not always easy to discern slaves from free men by their clothing, as seen in e.g. Petronius' Satyricon 37 .
The boy achieved his toga, the token of his citizenship, in two ceremonies, a private one at home and a public one that often took place during the liberalia 38 . The liberalia was a feast day devoted to Bacchus, where the name liberalia in itself indicates its close connection with the Latin word liber, free 39 . The ceremony was witnessed by family members and friends, and it is possible that also slaves were present, linking the male body with "place, dress, and male bonding" 40 . This public ceremony illustrates that, although the boy attained citizenship simply by growing up and being biologically mature, his maturity had to be recognized by society 41 . That is, his chronological age had to be sanctioned as a social age.
Modern masculinity studies have shown that such rites of passage are often seen as crucial for achieving manhood, that is, for boys to grow into men, since they function as a social confirmation or "proof " of a masculine identity 42 . As expressed by Kimmel, masculinity is always subject to doubt and needs "constant validation" 43 . An essential part of such rites is that they are affirmed by the peer group and by adult men. Rites making men out of boys must involve separation from the weakness and connotations, where the freeborn boy is seen as "unspoiled", and "those permitted to wear the toga pura were in some way 'pure' because of their freeborn status", Dolansky 2008, 54. 34 -Olson 2017, 48. 35 -Olson 2017. On the romanitas and masculinity of the toga as opposed to barbarian clothing see e.g. Harlow 2004, 44-45. 36 -Olson 2017, 54. 37 -Hemeros, one of the characters of the Satyricon, states that when he was a slave "no one could tell whether I was slave or free", nemo tamen sciit, utrum servus essem an liber, Petron. Sat. 57.9.
38 -See Dolansky 2008 for a thorough description and discussion on assuming the toga virilis. See also Olson 2017, 48 for further discussion.
43 -Kimmel 1990, 100. dependence of childhood and from the dependence on, and association with, the maternal, the feminine. This separation provides "a new sense of belonging to a distinctive world of adult males" 44 . It is important that this new male identity is adequately affirmed for men to acquire a confidence in their masculinity. As stated by Whitehead, "manhood must not only be attained, it must be seen to have been attained by others -men, for it is men who are its ultimate judges and arbiters, not women" 45 . Thus, manhood can be seen as a culturally sanctioned stage in a man's life, "which purports to connect to a deeper [...] male essence" and in so doing strengthening men's status 46 .
Although the puer-slave and the puer-boy were both striving for the toga, there was a huge difference between the boy and the slave in Late Republican and Early Imperial Rome. The boy became a man just by growing up biologically (although this growth had to be sanctioned and affirmed by other men), whereas the slave was always dependent on his master's will to be manumitted and thus become an adult. That the manumission was always at the mercy of the slave's master is very well illustrated in the words of Trimalchio, cited in the beginning (Petron. Sat. 39.4), which clearly state, "he (the master) wanted me to be a man", i.e. free: Patrono meo ossa bene quiescant, qui me hominem inter homines voluit esse (may the bones of my patron rest well; he wanted me to be a man among men). In short, the boy would surely grow up and be independent, but this was not the reality for every slave. Many slaves were never freed, and this made them remain boys. Remaining slaves, that is pueri, meant that they had a social age that was not corresponding to their physiological or chronological age.

The slave-puer as a sex object
As stated above, the slave did not have any rights over her/his own body, s/he was her/his master's or mistress' tool, to be used in whatever way it pleased him/her, also sexually 47 . This was true for male as well as female slaves. The fact that male slaves could be referred to as boys, pueri, in itself expresses that they were seen as sexually available, as does also its female counterpart puella. In this context, a puer or puella is a sexual 44 -Segal 1990, 131. 45 -Whitehead 2007, 380. 46 -Whitehead 2007. See further Kimmel 1995. 47 -Watson 1987, 119 comments on the absence of legal protection for Roman slaves of any age against sexual abuse by their masters. Green 2015, 144: "Roman men also envisioned slaves as instruments meant to fulfill their own sexual needs and desires". Green 2015 also discusses how Roman men used male and female slaves' sexual lives as instruments for their status performances and points out that Latin writers portrayed slaves as willing and consensual partners when performing domestic work. partner, a sexual object 48 . In effect, an anal penetration of a woman could be expressed by saying that she was "taking the part of a puer", as seen in e.g. Mart. 9.67.3 (illud puerile) 49 . This can be compared with the expression muliebra pati, used for anal intercourse with a man, discussed in the following section. The point is that the penetrated partner is not a real man -this position is reserved for unpenetrable men 50 -the partner is a woman or a boy 51 . The role of the penetrated partner was appropriate for a puer, "a term that covers both male children and male slaves of any age" 52 . Richlin, discussing boy-love and child-love, in fact raises the question of "whether the lexical blurring in the words for 'child/slave/beloved boy' in Greek and Latin represents a blurring, in practice, of categories [...] -'child' and 'sex object'" 53 .
A slave boy kept for sexual or visual pleasure could be called delicatus, delicium or deliciae 54 . In Sat. 75.10-11 Trimalchio tells that he served as a deliciae for both his master and mistress 55 . A slave who was aimed to please could be recognized by various visual features, such as make up and long, and often curly, hair. Make up was associated with women and effeminate men. The long or curly hair was a sign of desirability and sexual availability, and these slave boys were referred to as capillati in ancient sources. It is possible that the word capillatus was used as a metonym for a sexually attractive/available boy whether he actually had long 48 -E.g. Richlin 1983, 35-56;Richlin 1993;Parker 1997, 49-50;Walters 1993Walters , 29, 1997Williams 2010, 19, 83. Puella sexual OLD s.v. 3a; puer sexual OLD s.v. 3a.
49 -See Williams 2010, 83 and Walters 1997, 31 for discussion. 50 -See Fredrick 2002and Walters 1998 for the notion of penetration and "penetrability", also in a broader sense. For the notion "power is penetration" see e.g. Vout 2007, 19-20, passim. 51 -See e.g. Halperin 1990, 33-35 on boys and women as interchangeable. Richlin 1983, 32-56 discusses resemblance and difference between the erotic ideal of pueri and women as seen in ancient erotic literature. 52 -Walters 1997, 31. It has even been suggested that puer could be seen as a third gender in antiquity, based on the notion that the puer is penetrable despite his anatomical gender, Richlin 1997b, 30-31. In analogy with this idea, Fredrick 2002, suggests that a third or fourth gender should probably be proposed for successful freedmen, Fredrick 2002, 242. However, Walters 1993 to the view of puer as a third gender and argues that gender is a culture-bound social construct. Instead, he sees a group of "unmen" or "not fully men" though male in sex, comprising of youths, slaves, eunuchs and sexually passive males. 53 -Richlin 2014, 352. 54 -The literary and epigraphic evidence for delicatus, delicium/a and deliciae is vast, see Laes 2010, Pomeroy 1992, Richlin 2014 and W. J. Slater 1974 for discussion and further references. George 2013 discusses deliciae and its connection to the "cupid punished motive" in Roman art. According to Laes, Pomeroy and Slater probably not all deliciae were sexually abused. They could be small girls and boys who were kept in the house to provide company and amusement. They could be slaves but also natural children or foster children, and delicium/delicata could even be used as an epithet to denote one's spouse, Laes 2003Laes , 2010. However, as has been pointed out by Rawson 2003, 261: "The line between indulgent affection and sexual exploitation must have been blurred". This is also the conclusion of Richlin 2014 and of George 2013, 168-170, citing Rawson. 55 -See e.g. Bodel 1989;Pomeroy 1992 and Richlin 2014 for discussion.
or curly hair or not 56 . In the Satyricon, one of the freedmen, Hermeros, refers to himself as once a puer capillatus, a young longhaired slave, and proudly states that he is now "a man among men" and "walks around with his head bare": Homo inter homines sum, capite aperto ambulo 57 . Also Trimalchio had been a capillatus 58 . A slave boy kept for visual or sexual pleasure could also be depilated and wear a high-belted tunic. For elite men a tunic girded too short was cause for ridicule and scorn, as it was associated with lower class or even slaves and their sexual availability 59 . The elite's disgust for this kind of effeminate appearance and how degrading it was for a slave to be a sexually available puer is clearly articulated by Seneca in Ep. 47.7 cited below. To be adorned like a woman, muliebrem modum ornatus, was not suitable for a man's dignity, as expressed also by Cicero in Off. 1.130: viro non dignus ornatus. Men and women ought to be clearly distinguished by clothing, and crossing gender boundaries was often censured and ridiculed 60 : Alius vini minister in muliebrem modum ornatus cum aetate luctatur: non potest effugere pueritiam, retrahitur, iam que militari habitu glaber retritis pilis aut penitus evulsis tota nocte pervigilat, quam inter ebrietatem domini ac libidinem dividit et in cubiculo vir, in convivio puer est.
Another (slave), who serves the wine, dressed like a woman, struggles with his age: he cannot escape from his boyhood, he is dragged back to it; and although he already has a soldier's body, he is kept smooth by having his hair smoothed away or plucked out by the roots, he must stay awake throughout the night, dividing his time between his master's drunkenness and his lust, and in the bedroom he is a man, at the banquet he is a boy.
I pass over the flocks of unfortunate boys, whom other shameful treatment in the bedroom awaits, after the banquet is over. I pass over the troops of exoleti, ranked according to nation and colour, who must all have the same smooth skin, the same amount of first down on their cheeks and 56 -Olson 2017, 136-140 the same hair style, so that no boy with rather straight hair get mixed in with the little curly haired boys.
The passages cited above illustrate how slave owners sometimes tried to prolong their slaves' boyhood physically by removing body hair, thereby making them remain sexual pueri. The growth of facial and body hair was a marker and symbol of coming of age, and the first shaving was of crucial importance for a young Roman man, as it showed that he had crossed the boundary of manhood 61 . The first shavings of the emperors Augustus and Nero were followed by public celebrations and rituals, and Suetonius writes that Nero put his shavings in a golden box adorned with pearls of great price dedicating it to the Capitoline gods 62 . Also Trimalchio had his first beard saved in a golden box 63 . What is more, when Trimalchio tells his guests at the banquet that he was the deliciae of his master and mistress, he also says that he used to grease his lips with the oil of a lamp to make his beard grow faster 64 .
The beard clearly played a key role in the expression of Roman masculinity, as a symbol of the adult man. Thus, it was also a symbol of power and authority. From the work of the dream interpreter Artemidorus of Daldis, it is evident that dreaming of having a beard had different meanings shifting according to the sex, status and age of the dreamers. "When non-male dreamers are made 'male' by dreaming of a beard, they sometimes become empowered" to use Montserrat's words 65 . This is the case for the young boy or slave: "But for someone who is an adolescent, and who will soon grow a beard of his own, whether he is now a slave or a free man, it signifies that he will be his own master, since the beard shows that he is full-grown and responsible for himself " 66 . Slaves dreaming of growing a beard could thus expect to be freed; they would grow up and in consequence gain autonomy and the beard was a sign of this autonomy.
The fact that boyhood could be prolonged by removing facial and body hair shows that the word puer expresses the physical age as well as the social age of the puer as a sexual object 67 . Seneca's words cited above can be compared with the use of puer in Statius' and Martial's poems to the eunuch Earinus, freedman of Domitian 68 . The use of the word puer emphasizes Earinus' status as a sexual puer and the fact that he has not been able to grow up and become a man physically, although he is a freedman, due to his castration 69 .

Roman masculinities and the body
To been seen as a sexual object and not to have the right over one's own body had consequences for the slave's embodiment of masculinity and sense of a masculine identity, since the body was of major importance for the construction of Roman masculine identity. To embody the hegemonic masculinity it was essential to be able to guard one's bodily integrity and to be in control of one's body 70 . In fact, some Roman elite men, such as Juvenal and Seneca, even regarded gladiators as unmanly, as they were seen to have given up control of their own bodies, and -even worse -given it to other men to be used for pleasure 71 .
The ideal body of a Roman man was healthy and strong, linked to the notion that the characteristic of a man is that he is hard, hot and dry, as opposed to the soft, cold and damp woman 72 . The notion of the healthy and strong manly body was also connected to the idea that a man's body mirrors his virtues. As expressed by Gunderson, the body of a man "must represent the virtue of the character who bears it" as it was seen as a public object 73 . An ideal Roman man was a vir bonus, a real and good man. Bonus signified good, but not only in a moral sense. The adjective bonus also expressed social standing and the wealth that came with a certain position in society 74 . A vir must never act in a "slavish" or "womanish" way, serviliter or muliebriter, as expressed by Cicero in Tusc. 2.55. In his opinion, a man must "beware of anything immoral, loose, unmanly", cave turpe quicquam, languidum, non virile (Tusc. 2.51).
As Cicero explains in the same section, being a man "is to be master of yourself " (Tusc. 2.53). Cicero's words can be understood as descriptions of a masculine position, where Cicero is distancing himself from slavish and womanish men and behavior. Moderation, courage, reason, activity 69 -Vout 2007, 167-212. 70 -Cf. Stewart 2016Walters 1998;Williams 2010, passim. 71 -Juv. 2.143-5;Sen. Q Nat. 7.31.3. Stewart 2016, 49;Walters 1998, 364;Williams 2010, 154-155. Walters 1998 argues that the (male) spectator could be seen as a penetrator. Cf. Bartsch 2006, 152-164, who discusses the "penetrative viewing" and the cultural effeminization of the actor and of the body on display. See Edwards 1997 on the unmanliness of actors for the same reason. Cf. Juv. 6.110 where gladiators are portrayed as pretty boys loved by women, Hyacintii, i.e. pueri, Vout 2007, 94. 72 -E.g. Quint. Inst. 11.3.28;Sen. Dial. 7.7.3;Varro Rust. 1.41.4. See e.g. Corbeill 1997, 107-109 for further references and discussion on how this contributed to the construction of Roman masculinity. Cf. Edwards 1993, 174. 73 -Gunderson 2000, 61, 70. 74 -Gunderson 2000 and control were intricately linked to the ideology of masculinity in the Roman Late Republic and Early Empire society, and this was closely connected to the male body. Therefore, it was essential for a Roman man to be in control of his body in every possible way. A man who did not train and take care of his body, ran the risk of his body becoming weak and soft. Such a weakness made a man physically and mentally effeminised, as the male body was considered strong and healthy as opposed to the weak and less valued female body. This notion is expressed in Quintilian (Inst. 11.3.19) when he advices the orator to uphold physical strength, firmitas corporis, for fear that "the voice be thinned out to the frailty of a eunuch, woman or sick person" (ne ad spadonum et mulierum et aegrorum exilitatem vox nostra tenuetur) 75 . Control and training of the body hence became an essential part of the hegemonic masculinity and a way to maintain that status. As seen above, clothing and looks were also important for a man to conform to the hegemonic ideal and a man who did not embody the traditional masculine looks could be called effeminatus or mollis 76 .
Being fit and being in control of one's body is of importance when constructing masculinity in modern societies 77 . Being in control is associated with men as opposed to women's supposed inability to control their emotions and feelings 78 . A desire to control both self and others is a fundamental principle in many men's sense of their masculinity, and this desire is often manifested in sexual practices 79 . The notion of the man's impenetrability is also central to this discourse 80 . In Roman society, power and control of self and others were essential concepts to the masculine ideology and sexual penetration played a key role in the semantics of gender, as the feminine was associated with the sexual role of being penetrated 81 . This discourse was clearly expressed in the ideals that defined sexual practices and relations: In an ideal heterosexual relationship, the man, the vir, was always the active penetrator, playing the part that was considered superior and dominant.
The woman, on the other hand, played the passive and penetrated part. The mere fact that the woman was penetrated was perceived as an expression of submission and subordination, drawing on the notion that 75 -Corbeill's translation, Corbeill 1997, 125 n. 36. 76 -Olson 2017. On softness, mollitia, as an "antithesis of masculinity", Williams 2010, 139-140. Cf. e.g. Gunderson 2000Fredrick 2002. For the concept mollitia see also Olson 2017, 156 n. 22 (references to modern studies), 166 n. 140; Edwards 1993, 63-97, 174. 77 -E.g. Petersen 1998, 51. 78 -Lloyd 1984, 2. 79 -Whitehead 2002 80 -E.g. Thomas 2001. 81 -Williams 2010Williams 2014. she had to suffer something, pati, beyond her control. The passive verb, patior, means "suffer", "undergo" or "experience" and so is used of "being penetrated". Women were, in fact, "born to be penetrated", pati natae, as expressed by Seneca in Ep. 95.21. Accordingly, a man who was penetrated by another man could be referred to as "having a woman's experience" or "suffer like a woman", muliebria pati, as mentioned above 82 . A grown man who was anally penetrated could be called a pathicus, and the term cinaedus referred to a man who was apt for and enjoyed anal penetration and in his dress and gait did not conform to the normative masculine ideal, although he could also enjoy penetrative sexual relations with women. These terms, as well as effeminatus and mollis, could be used interchangeably by Roman authors, and, although their significance is somewhat unclear, they have negative connotations and are used to describe others 83 . Thus, sexual practices were based on the binary oppositions activepassive or penetrator -penetrated 84 , expressing a power relation between the dominator and the dominated and they could, in fact, be seen as expressions of power 85 . Moreover, sexual relations and activities illustrated the valuation and positioning of male and female bodies, where the female body was seen as subjugated since it was penetrated. This applied to male slaves as well, as they were seen as sexual objects, sexual pueri, enjoying the same sexual subordinate position as women, as discussed in the previous section 86 .
For that reason, it was of major importance that the relationship or sexual act between a slave and a master followed the rules of domination and subordination, "the dominance-submission grid of Roman sexuality" 87 . As sexual acts were seen as expressions of power, the master must be the active and penetrating part 88 . In Seneca Ep. 47.7, cited in the previous section, the phrase in cubiculo vir, in convivio puer est (in the bedroom he is a man, at the banquet he is a boy), the usage of the word vir implies that the slave will play the penetrating role in the sexual act between master and slave 89 . This is degrading for the master, as he ought to be the real man, the vir, in this relationship. It was considered humiliating and morally corrupt if a man offered himself to be penetrated by a person of a lower social standing and such emasculate conduct could undermine his reputation and authority.

Achieving manhood
If and when a male slave was manumitted, he became an adult, a man. However, the freed slave did not become a vir, rather, the word homo was used to define the freedman 90 . As discussed, vir was used for men of the elite and the word had positive connotations. It was closely connected to the virtue virtus, and could also mean husband or soldier. Homo had lower connotations than vir and was often used to define men of lower rank 91 . As the slave gained his freedom, he was given the right to his body and self, and so he achieved manhood. Manhood can be defined as "the period in a male's life from when he is socially recognised as a man" 92 . To be seen as a grown man was of crucial importance for the freedman due to the fact that when he was a slave, he was seen as a boy, not a man. Thus, after manumission, the former slave was finally perceived as an adult, and his social age and his physiological and chronological age were corresponding.
A way for freedmen to show this manhood to society was to make portraits where signs of age were shown, and also to make use of family portraits and inscriptions. This is visible in the funerary art of Roman freedmen from the Late Republic and Early Empire, where freed people are often portrayed as mature, or even old, men and women. This art originated in the so called "veristic" portrait style that was popular among the elite in Rome during the 1 st century BC. One characteristic of this style was that it showed features of age, such as wrinkles, loose, flabby skin and bald heads. These features must not be seen as "realistic", rather they are expressing an ideal where concepts such as sapientia, severitas and auctoritas were important and thus they represent the gravitas and dignitas of age. According to Roman belief, there were certain virtues and behaviour connected to each age, where joy was suitable for the young and 89 -Edwards 2019, 185. In her commentary to this passage Edwards also points to the fact that the master is "allowing himself to be penetrated by a person he would not deign to dine with". 90 -McDonnell 2006, 159-160. Cf. Mouritsen 2011, 61-65, 98. See Santoro l'Hoir 1992  seriousness for the old 93 . Old people were supposed to show moderation and self-control, and portraits showing old age reproduced these norms. "The exaggerated wrinkles and folds were used as a topography of virtue" to use Cokayne's words, and, as argued by her, these portraits show how these old people wanted to be seen, that is how they wanted to be perceived by society 94 .
In the late 1 st century BC to the 2 nd century AD, the "veristic" style became popular among the freed group, although it was less popular among the elite by now. With the rise of Augustus and the Principate, a new style came into fashion. This was a classical, idealised style showing serene, calm and youthful looking faces, alluding to the Augustan ideology of peace and prosperity. It has often been stated that the "veristic" style was used by freed people as a way to show off their new social status as Roman citizens, emphasizing old Roman values and mores 95 . There was clearly an ideology behind the "veristic" portrait style, possibly connected to the wearing of funeral masks of ancestors, imagines, in the funeral processions of the nobility, alluding to the mos maiorum 96 . Such associations were attractive to Roman freedmen who did not have legitimate ancestors and hence had a need to show their connection to traditional Roman values and a place in society 97 . In addition, I would like to argue that this style expresses the importance of age in itself for Roman male freedmen. By making old-age portraits, the male freedman could show himself as an adult man with conservative values, such as moderation, authority and self-control, virtues that were associated with old age, but that were also essential components when constructing hegemonic masculinity. In this way, by using an old-age portrait, the freedman could show himself as a real man, in every way different from the puer he had been as a slave.
Roman toga, the symbol of the male free citizen, and his wife as a Roman matrona wearing the traditional stola, and freeborn sons with the bulla and the toga praetexta, expressing freeborn status 100 . The birth of a son was of major importance for the family and consequently boys are much more frequent than girls are on reliefs showing freed families 101 . The tomb relief of the Servilii can be used as a representative example of freedmen funerary monuments ( fig. 1)  The boy Globulus bears his father's gentilicium and has filiation, a sign that he is freeborn and born after the manumission of his father 103 . In addition, he is identified as f(ilius), son, and wears a bulla around his neck. The freeborn status of the child is made visible in every possible way and so is Hilarus' status as his father, identified as pater, father, in the inscription. It is interesting to note that Sempronia is identified as uxor, wife, and not mother, as would be expected in analogy with Hilarus. I would like to argue that this may indicate that Sempronia functions as a means for Hilarus to show himself as a married man, a pater familias with a wife and child. Thus, the funeral monument can be interpreted as a way to emphasize Hilarus' masculine status.
Funeral portraits of couples and inscriptions commemorating spouses are very frequent among the freed group. In fact, epigraphic studies of freedman inscriptions have shown that inscriptions from the family sphere dominate the material and inscriptions between spouses are the most common type 104 .
The importance of marriage can be illustrated by the funeral monument of the freed couple P. Aiedius Amphio and Aiedia Fausta Melior (CIL 6.11284) ( fig. 2) 105 . This is a portrait of an elderly man and his younger wife. The stern and serious expression of Amphio shows that he shares the traditional values of the Roman elite and the signs of old age are clearly accentuated in his wrinkled, deeply lined face and sunken cheeks. Fausta, on the other hand, looks much younger than her husband, and her portrait appears more idealized. This is often, but not always, the case with monuments showing married couples. That the portraits of women give the impression of being more idealized may be intentional, but can also be due to the fact that women were often much younger than their husbands 106 . The conjugal bond between husband and wife is highlighted as they are joined in dextrarium iunctio (the joining of the right hands), the marker of the legal marriage, matrimonium iustum. This gesture can be seen as a symbol of the concordia (harmony) and fides (loyalty) of the married couple 107 . In addition, I maintain that the visual sign of marriage emphasizes Amphio's role as a husband, who, in a marriage cum manu, is supposed to control his wife. This control is crucial for his masculine position as it gives him the possibility to embody the hegemonic masculinity ideal of a pater familias, a head of household in control of wife and children.
As said, the slave did not have the right to form a legal marriage, matrimonium iustum, and to have legitimate children. The slave was seen as having no family connections at all, as expressed in Roman law 108 . The manumission gave the former slave the control of his body and with this came also the mastery of his own reproduction 109 . Hence, it was important for the male freedman to be able to show himself as a father and husband, i.e. a head of household. As a matter of fact, matrimony played a vital part in the constructions of masculinities in ancient societies, as it had a function in the hegemonic masculinity. To be a man in the Greco-Roman society was to be able to form a legal marriage and to have legitimate children. This masculine ideal was applied to the male slave as well because when the slave was freed, he was subject to the ideology of the life course of the freeborn male citizen 110 . In this way, the hegemonic treating Flavian freedmen.
105 -Zanker 1975, 285, abb. 16. 106 -Cokayne 2003, 25. 107 -Borg 2012, 29. See Zanker 1975  masculinity of the elite was maintained as also the freedmen strived to comply with it. In many cultures, marriage is perceived as a transition into masculine status and identity and it has been connected to the taking up of power and responsibility for the wife, children and household, where the man is seen as "head of household" 111 . Also in modern slave societies, the family was very important for freed slaves, as the slave was often deprived of paternal and familial authority and thus treated as a child and often referred to as "boy". Depriving the slave of all attributes of responsibility was a way for the white slave master to exercise his authority over the black male slave. According to Stuart Hall, this infantilization can be seen as a way of "symbolically castrating" the black man, since it is deprives him of his masculinity 112 .
In the same way as the slave in modern society, the Roman slave was forced into a prolonged childhood, not being able to take up the responsibilities of an adult man. Only when manumitted were these attributes of responsibility available to him and he was thus proud and eager to show them, since they were an essential part in constructing hegemonic masculinity. That the family was of crucial importance for freedmen can be seen in their inscriptions and funeral monuments, where family relations are often emphasized. Having a family was taken for granted among other groups in Roman society but it had a special significance for the freedmen. The awareness of its importance was something that united them as a group and constituted an important part of their collective identity 113 . This is the reason why freedmen commemorated family members in inscriptions, often stressing conjugal or parental bonds, and why other members of society did not 114 .

Concluding remarks: age as a social category
The purpose of this article was to examine how age can be used as a means to study how freedmen are positioned in relation to the ideal 111 -Heaphy 2007, 322-323. Heaphy discusses how marriage is associated with certain kinds of masculine identities and roles in relation to family and work. See also e.g. Arber & Ginn 1995 on the "head of household" model across the life course. Nevertheless, there are few studies on the experience of men as husbands, compared to the studies on women as wives, in modern society, see further Brook 2007, 381-385. 112 -Hall 2013. In addition, Connell argues for a "feminization" of colonized men, often called "boys", by the colonizers, where race is understood as a hierarchy of bodies "inextricably mixed with the hierarchy of masculinities", Connell 2000, 61. See Connell 2000, for further references: e.g. Shire 1994, 149 for the usage of "boys" regarding Shona cultures.
113 -Hagelin 2010, 166-168, 174-220;Mouritsen 2005, 40-62;Mouritsen 2011, 289. Cf. Patterson 1982 114 -Mouritsen calls this phenomenon a "socially specific response to an otherwise common experience" (that is the death of spouses or children), Mouritsen 2005, 62-63. Roman masculinity and to discuss why it was important for a Roman freedman to be seen as a man, as expressed in the Satyricon, homo inter homines, "a man among men".
As a phenomenon, Roman freedmen are interesting when studying intersections of age and gender, because they illustrate the way age can be perceived as a social category, as the slave's chronological, physiological and social ages were not corresponding. The social age of the slave was corresponding to the chronological and physiological age of a child, expressed by the use of the Latin word for boy, puer, referring to a slave. To be seen as puer was degrading for the slave in various ways, and it was of crucial importance when constructing masculinity. An observation to be made is that the notion of the slave as a child in antiquity can be compared to the infantilization and feminization of colonized and/or enslaved (black) men, who were also called "boys". This illustrates the fact that age as well as gender is a social category 115 .
Perceived as a boy, it was impossible for the male slave to comply with the Roman hegemonic masculinity. It also posited him in a social position similar to that of women, who, in many ways, were seen as children all their lives, due to their subordination and supposed inability of mental and physical control. Accordingly, when the slave was manumitted and obtained his freedom, he also achieved manhood -he became a man. As a free man, it was possible for the ex-slave to guard his bodily integrity, he was not subordinate to other men and he could marry and have children. This made him appear as a man in Roman society, since the freedman was also subject to the ideology of the life course of the freeborn male citizen. This, I would like to argue, is the reason why age, and appearing as a mature man, was of major importance for Roman freedmen. Finally, their social age corresponded to their chronological and physiological age and the freedman was perceived as an adult, not a boy. He was "a man among men", homo, inter homines, as expressed by the freedmen in the Satyricon (Petron. Sat. 39.4; 57.5).