Full text of DOI:10.1177/036319902236623

1JPO0h.aU1n1Rg7N/7A/F0AL3M6O3IF1L9FI9EA0SM22OI3LF6Y6R2HO3IMSTAONRSYO/LODcItEoRbeSr 2002
THE FAMILIES OF ROMAN SOLDIERS
(FIRST AND SECOND CENTURIES A.D.):
CULTURE, LAW, AND PRACTICE
Sara Elise Phang
During the first two centuries A.D., Roman soldiers were prohibited from contract-
ing legal marriage; the masculine nature of Roman military discipline was the
likely motivation for the ban. Nevertheless, many Roman soldiers formed de facto
unions with women and fathered children. This marriage ban is directly attested to
in papyri from Roman Egypt that show that soldiers’marriages were illegal and
that their children born during service were illegitimate. Military diplomas
granted to discharged soldiers show that veterans were given the right to marry.
This article explores the nature of the soldiers’de facto unions in Roman law and
the impact of the ban on the soldiers’illegitimate children. This article also dis-
cusses the soldiers’union in social practice, based on a survey of Latin epitaphs.
Soldiers more frequently married in the second century; other features, such as
relatively late marriage or celibacy, seem characteristic of the homosocial mili-
tary subculture.
Soldiers’ conjugal families have usually been problematic in institutional terms. The
soldier’s first obligation is to the military institution to which he belongs; responsibil-
ity for a family may be thought to conflict with obligation, and should the military be
responsible for the family should the soldier die? Various professional armies have
adopted the drastic solution of prohibiting soldiers, usually below a certain rank, from
entering into a legitimate marriage during service, thus regarding family formation
as incompatible with military service. These restrictions of marriage characterized
eighteenth- and nineteenth-century European armies.1 In the United States in the nine-
teenth century, the military did not prohibit soldiers from marrying, but neither did it
encourage the recruitment of married men.2 This attitude, which persisted into the pre–
AUTHOR’S NOTE: In this article, studies in languages other than English have been kept to a minimum.
For a more extensive bibliography, see S. E. Phang, The Marriage of Roman Soldiers, 13 BC-AD 337: Law
and Family in the Imperial Army (Leiden, the Netherlands, 2001).
Sara Elise Phang received her Ph.D. in Roman history at Columbia University in 2000 and is presently an
Andrew W. Mellon Postdoctoral Fellow and Lecturer in the Department of Classics at the University of
Southern California (2001-2). Her first book is titled The Marriage of Roman Soldiers (13 B.C.-A.D. 235):
LawandFamilyintheImperial Army (2001). She is currently working on a study of Romanmilitary discipline.
Journal of Family History, Vol. 27 No. 4, October 2002 352-373
DOI: 10.1177/036319902236623
 2002 Sage Publications
352
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

Phang / FAMILIES OF ROMAN SOLDIERS
353
World War II period,3 is reflected in the saying, “If the Army had wanted you to have a
wife, it would have issued you one!”
The contemporary American military’s unrestricted permission to marry and its
welfare provisions for soldiers’ families are a result of mass service in the Second
World War and of the voluntary recruitment policy from 1973 onward.4 Marriage
restrictions continue in the French Foreign Legion, in which enlisted men are prohib-
ited from legal marriage; the legionnaires’ concubines, however, are tolerated.5
This second approach was the policy of the imperial Roman army, often considered
the first “professional” army. During the first and second centuries A.D., common sol-
diers were prohibited from marriages that were valid in Roman law. This marriage ban
is usually attributed to the first emperor, Augustus (27 B.C.-A.D. 14). In practice, many
Roman soldiers formed de facto marriages and had illegitimate children. The families
suffered from the legal disabilities of invalid marriage but developed coping mecha-
nisms. Many soldiers remained celibate as the Roman military culture regarded sol-
diers’marriages as incompatible with martial virtues. This attitude may have changed
in the second century A.D., as family formation became frequent. The marriage ban
was revoked by the emperor Septimius Severus in A.D. 197.6
The soldiers’marriage ban affected many men for a large part of their lives since the
Roman army in this period—estimated at 300,000 to 450,000 men—was a long-service
army in which men served twenty-five to twenty-eight years, depending on the branch
of the service.7 Legionaries (about 180,000 men) served twenty-five years and were
recruited from Roman citizens. Auxiliary troops (cavalry and infantry regiments) were
recruited from noncitizens and served twenty-five years. Sailors in the navy were also
noncitizens and served twenty-six to twenty-eight years. During the first and second
centuries A.D., periods of warfare were separated by long periods of peace, and the
legions became almost stationary, situated chiefly in frontier provinces along the
Rhine and Danube rivers and in the Near East. As in later periods of European history,
common soldiers were recruited from the laws (free) social stratum; the legionary was
paid a salary similar to that of the wage laborer. Noncitizen troops were granted
Roman citizenship on discharge; all veterans were granted the right to marry.
THE MARRIAGE BAN
The existence of the soldiers’marriage ban has been disputed by scholars of Roman
law because the ban is not extant in its original form and the evidence attesting to it is
very diverse. Greek and Roman literary authors, writing in the period of the ban or
soon afterward, mention that soldiers were not permitted to marry. Cassius Dio’s Ro-
man History states that the emperor Claudius in A.D. 44 “granted the soldiers the rights
of married men, since they could not have wives according to the laws.”8 These “rights
of married men” signify privileges granted to men who married and had three children,
in accordance with the emperor Augustus’s legislation promoting the reproduction of
Roman citizens. The soldiers’ marriage ban was ended by the emperor Septimius
Severus (193-211) in A.D. 197; the early third-century historian Herodian says that he
“permitted soldiers to live with their wives” (3.8.5). The term “live with,” the Greek
synoikein, usually connotes legal marriage. Writing in the late fourth century, the
AntiocheneGreekorator Libanius (Oration 2.39-40) says of present-daysoldiers that
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

354
JOURNAL OF FAMILY HISTORY / October 2002
nothing stops them from getting married. . . . This was not the case in those good old
days which I comment. . . . The men themselves were sturdy and brave, specialists in
warfare, and they remained unmarried; it was ensured that they would even have no
need of marriage.
The most important evidence of the ban is the Cattaoui papyrus, discovered at the
end of the nineteenth century. The Cattaoui papyrus is a collection of excerpts or sum-
maries of legal proceedings before the prefect (governor) of Roman Egypt or his depu-
ties in the early second century A.D.9 Three concern the status of soldiers’children born
during their fathers’service. The children are stated in these excerpts to be illegitimate
and, moreover, that soldiers cannot marry:
Longinus Hy . . . stated that he himself was a Roman citizen and had served in the first
cohort of Thebans under Severus, and that during his military service he had cohab-
ited with a Roman citizen woman by whom he had begotten the sons Longinus
Apolinarius and Longinus Pomponius, and he asked that these be certified [as Roman
citizens]. Lupus after consulting with the jurists, said, “The boys . . . since they were
born from a Roman woman. You also desire to establish them as legitimate, but I can-
not make you their legal father.” (3.11-22.24 October 114)
Roman citizens in Egypt were certified as such through an official examination of sta-
tus. The prefect of Egypt, Rutilius Lupus, probably stated in the lacuna that the sons of
Longinus could be certified as Roman citizens. When a marriage was not valid in
Roman law, the children took the status of the parent of inferior status; in this case,
since both parents were Roman citizens, the children were Roman citizens.
The case of Chrotis and Isidoros illustrates the change in legal status through mili-
tary service:
Chrotis stated through Philoxenos her advocate, that she, herself a citizen of Alexan-
dria, had married Isidoros, an Alexandrian citizen, and thereafter while [Isidoros] was
serving in his cohort, a son Theodoros was born to them. Concerning [Theodoros] she
now petitions, arguing that [although she had omitted to make a declaration of his
birth], it was nevertheless clear in the will that [Isidoros] had written that he was his
son, for [Isidoros] had made him the inheritor of his property. On examining the will
of Iulius Martialis, soldier of the 1st Cohort of Thebans, and after consulting with his
advisers, Lupus replied: “Martialis could not have a legitimate son while he was serv-
ing in the army, but he named him as heir legally.” (4.1-15.4 June 115)
Isidoros entered the auxiliary forces as an Alexandrian but was probably granted Ro-
man citizenship on discharge and, although veterans were not required to change their
names, took a Roman name: Iulius Maritalis. A tax was levied on the estates of de-
ceased Roman citizens, from which only legitimate children were exempted;
Martialis’s son did not meet this criterion as while Martialis was free to name him as
his heir, illegitimate children were not excluded from testate inheritance in Roman law.
The case of Octavius Valens further illustrates not only the marriage ban but also the
indignation of soldiers who truly believed that they were married:
When Octavius Valens and Cassia Secunda, whose case had been adjourned from yes-
terday, cameintocourt, [theprefect]Eudaimon, having consulted with his legaladvis-
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

Phang / FAMILIES OF ROMAN SOLDIERS
355
ers, said: “Yesterday when the memoranda of the distinguished Heliodorus were read
out, the reason for an adjournment became clear, in order that judgment should be
reached in the case of those who are forbidden (to marry), in the presence of the child’s
mother. Today, having considered the relevant circumstances in this matter, I confirm
[the judgment] which I gave yesterday. Whether this man enlisted in a legion or in a
cohort or an ala, he is not able to have a legitimate son. Moreover, since he is not the le-
gitimate son of his father who is an Alexandrian citizen, he cannot be an Alexandrian
citizen. Therefore this boy, born when Valens was serving in a cohort, is a bastard; he
cannot be admitted to the city-state of the Alexandrians.” And he added, “Yesterday
you said that you had other sons; what ages are they, when were they born?” Octavius
Valens replied, “One just recently, the other is older.” Eudaimon said, “Where were
you serving when the older was born?” Valens replied, “In the cohort, and the younger
one too.” Eudaimon said, “Know that they are in the same class as this one. There are
things which cannot be changed.” Valens said, “Now, if I manage to go abroad, you
yourself could sign my petition, so that I may obtain my rights through a legal repre-
sentative. What wrong have the children committed?” Eudaimon said, “I have been
generous [or foolish] in explaining in many words what I could have said in a few
words. You are reaching after impossible things; neither this son nor your other sons
are Alexandrian citizens.” (4.16-5.26.26 August 142)
Clearly, themarriageban applied to both legionariesand auxiliaries (cohorts and alae).
In Papyrus Cattaoui, three other cases also concern the validity of soldiers’ mar-
riage. In a legitimate Roman marriage, the wife gave a dowry to her husband. If they di-
vorced, or if he died, the wife was entitled by law to the repayment of her dowry from
his estate through a special legal remedy, the actio rei uxoriae.10 Soldiers’women were
refused this legal recourse since dowry in an invalid marriage was not recognized.
Lucia Macrina states through her advocate Fannius that she demanded her deposit
from the property of the deceased soldier Antoinus Germanus. (The prefect) Lupus
said, “We know that deposits are dowries. For such reasons as these I cannot grant a
judgment. For a soldier cannot marry. If you had demanded the dowry and I give a
judgment, I will seem to have been persuaded that the marriage is legal.” (1.5-13.5
January 117)
This passage, the Greek text of which is much disputed, shows that soldiers and their
“wives” attempted to get around the marriage ban by exchanging not dowry, which
was legally invalid, but a deposit or type of loan; they hoped that the “wife” could
recover her deposit from the soldier’s estate through the legal suit known as actio
depositi. The usual procedure was for the magistrate, here the prefect of Egypt, to
nominate a judge who would try the case himself. Here the prefect Lupus states that
“we know that deposits are dowries” when women transact them with soldiers. He
refuses the action on the deposit on the grounds of fraus legis, fraud attempting to cir-
cumvent the law.11
Another fragmentary case shows that a soldier’s woman could have difficulty in re-
covering her deposit or dowry should the soldier refuse to cooperate, as in a hostile
divorce:
Chthinbois (lodged a suit) against Cassius Gemellus, a cavalryman of the Vocontians,
in the presence of her father Orestouphi. Apollinarius the advocate said that since
Gemellus owed 700 drachmas to Chthinbois according to a bank draft, he must hand
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

356
JOURNAL OF FAMILY HISTORY / October 2002
them over according to it. But Alexander and Herakleides the advocates of the defense
accused Chthinbois of wanting to demand back a dowry from him, as if she had mar-
ried [Gemellus], which [dowry] she had given to him. They said that it was not possi-
ble for soldiers to take wives. [What follows is fragmentary.] (2.14-3.10.25 February
134)
The final case in Papyrus Cattaoui reiterated that soldiers cannot marry and refuses the
action for the deposit/dowry. This case also shows that gifts between a soldier and his
woman during service were valid, proving that the union was not a legal marriage as in
legal marriage, gifts between husband and wife were invalid. The institution refers to
substantial property transfers, not trinkets.12
[Record] of the idios logos Iulianus.
Sarapion son of Apollonius and Amoisosas son of Heliodorus, represented by the
advocate Apollonius the elder, brought a charge against Cornelia that she had taken
possession of seven slaves conveyed to here in the time of their cohabitation by (the
soldier) Iulius Acutianus, who died intestate. Theon, the advocate defending
Cornelia, said that the marriage was not legitimate, for they had united when
Acutianus was in military service, and that she had bought the slaves. He recognized
the purchase of Mousa with her nursling in the twelfth year [of Hadrian], and that of
Daphne in the nineteenth year, but that Threputs and Syntrophus were house-born.
Horeion, the advocate defending Cornelia, said that Severus the nursling was already
born from Mousa, but that Elpidephoros could not be house-born, although this is the
case among many, (to raise) children who are house-born so as not to pay for them.
Iulianus said, “You may not claim as house-born the slaves purchased after his mili-
tary service . . . they go to the imperial account; the others I release to you.” Cornelia
petitions that she be given a talent (6000 drachmae) which Acutianus had kept for her
as a deposit, and she recognized the document, but the accusers said that this was a
marriage contract, for soldiers thus form marriages. But Iulianus said, “I refuse to rec-
ognize a loan arising from a marriage contrary to law.” (6.1-23.23 November 136)
“House-born” slaves, those born to slave mothers, were highly valued. Cornelia was
permitted to keep the slaves whom she acquired during Acutianus’s military service,
but she was not permitted to keep the slaves acquired after his discharge, when the
union became a legitimate marriage.
Another papyrus document, Berliner Griechische Urkunden (BGU) 140, a procla-
mation in letter form by the emperor Hadrian (117-138) addressed to the prefect of Al-
exandriain A.D. 119, grantscertain legal privileges to soldiers’illegitimate children:13
Copy of a letter to the emperor, translated . . . which was publicly displayed in the third
year of Trajan Hadrian Augustus, in the consulship of Publius Aelius for the 3rd time
and of Rusticus . . . in the winter camp of the III Cyrenaica legion and the XXII
Deiotariana on the 4th of August which is the 11th of Mesore, at headquarters.
I know, my dear Rammius, that those whom their parents in their period of military
service acknowledged as their children have been debarred from succeeding to their
fathers’property, and this measure did not seem harsh since their action was contrary
to military discipline. But I find it most pleasant to put forward precedents through
which I may interpret more liberally the rather harsh rule established by the emperors
before me. For although those acknowledged in the period of military service are not
legitimate heirs of their fathers, nevertheless I decide that they also can claim posses-
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

Phang / FAMILIES OF ROMAN SOLDIERS
357
sion of property from that clause of the edict which gives this right to cognate rela-
tives. It will be your duty to make this grant of mine well known both to my soldiers
and to the veterans, not to enable me to be exalted in their eyes, but so that they may
profit by this, if they are unaware of it.
Hadrian’s letter shows clearly that soldiers in service could not have legitimate chil-
dren and that their procreation was considered contrary to military discipline. How-
ever, he grants the illegitimate children inheritance rights as a humane policy.14
These papyri attesting to the marriage ban were found in Roman Egypt, but nothing
about them suggests that the ban applied only to Egypt, as some revisionists have sug-
gested.15 The literary authors, such as Cassius Dio, Herodian, and Libanius, mention
the marriage ban as if it applied to all soldiers everywhere; Egypt was a distant and for-
eign land to them. The Roman army’s administration was fairly uniform throughout
the Empire.
The military diplomas, granting discharge privileges to veterans, suggest the exis-
tence of a ban during service. These diplomas are small bronze tablets engraved with
copies of the imperial legislation granting privileges to veterans of the Roman armed
forces. Some four hundred diplomas have survived and have been published.16 In the
diplomas, auxiliaries, sailors, and other noncitizen troops were granted Roman citi-
zenship and the right to marry (conubium) at discharge; their children were also
granted Roman citizenship. The formula runs as follows:
The emperor grants to them and to their children and their posterity, the citizenship
and the right of marriage (conubium) with their “wives,” the women whom they had
when the citizenship was granted to them, or if they were unmarried, with those whom
they married afterwards, limited to one woman for each man.17
Praetorian guardsmen, who were already Roman citizens, were granted only
conubium. Conubium existed automatically between two Roman citizens, but had to
be conferred upon unions of Roman citizens with non-Romans. In the case of the aux-
iliaries and sailors, the veterans and their children were granted Roman citizenship but
the wives were not, requiring a grant of conubium. The conferral of conubium suggests
that auxiliaries, sailors, and Praetorians’wives were non-Roman, at least at the incep-
tion of the grant of diplomas, begun by Claudius in A.D. 52. Diplomas continued to be
granted to the auxiliaries down to A.D. 203, to sailors and to the Praetorians through the
third century, but it should not be assumed that all wives of these troops were always
non-Romans. Epitaphs attest that most of the wives of Roman soldiers had Roman-
type names by the second century. Legionaries did not receive diplomas; it is probable
that most of their wives were Roman citizens and hence they did not need conubium or
citizenship for their children. It is notable that Roman citizenship was not conferred on
the noncitizen wives of soldiers.18
The grant of permission to marry does not seem to have been dependent on the sol-
diers’ status or conduct. The military diplomas granted conubium to all veterans, and
only a few diplomas suggest that a probatio or “examination” was required, as in the
naval diploma formula of the late second century (after 158):
[The emperor] granted to the [veterans] and to their children, whom they had begotten
from women with whom they proved that they had lived in a permitted relationship
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

358
JOURNAL OF FAMILY HISTORY / October 2002
(quas secum concessa consuetudine vixisse probaverint), the Roman citizenship and
the right of marriage with the same women.
The “permitted relationship” (concesssa consuetudo) was not legitimate marriage, as
some have thought. It is possible that this proof monitored the status of the women and
required durable or monogamous relationships.19 But very little is known about the
social or political context of this concession; probably, the sailors were required to
document only that they had been “married” for some time, preventing the substitution
of other women. The military diploma formulas in general suggest that any social con-
trol was limited to restricting the grant of the citizenship and of conubium to one
woman per soldier and to her children to prevent fraudulent acquisition of the citizen-
ship, not as an instrument of moral control.20
MOTIVES FOR THE MARRIAGE BAN
Modern scholars have suggested practical motives for the policy of the marriage
ban. For example, Parker suggested that women and families hindered the army on the
march; Watson suggested that soldiers who formed family ties with local civilians
were reluctant to be transferred and might even desert as a result.21 Modern armies
often restricted soldiers’marriages ostensibly to keep down costs.22 Similarly, Camp-
bell and Lindsay Allason-Jones suggested that through the marriage ban, the army
evaded pay raises or the provision of benefits for the soldiers’ families.23 Colin Wells
has suggested that the marriage ban was meant to prevent the depopulation of Italy by
discouraging women from joining the Italian legionaries at the frontier.24 These
motives depend on the assumption that all soldiers marry, which was not the case.
Some of these motives are based on gendered assumptions that women and children
are helpless and parasitical, emotional, and financial burdens; Libanius depicts fourth-
century soldiers as unable to support their wives and children (Oration 2.39-40). In
practice, the Roman soldiers’ women brought dowries that could be used for their
support.
It is more likely that the reasons for the marriage ban were political or cultural.
From prehistoric times until very recently, warfare has been the prerogative of males;
this was so in Roman society where women were excluded from political life.25 Roman
military virtues such as virtus and disciplina militaris are associated with the mascu-
line gender: the Latin virtus or “courage” derives from vir (man) and retained this asso-
ciation even when it came also to mean “virtue.”26 Disciplina militaris connoted obedi-
ence and martial competence through training but also required a masculine austerity
of the soldier who was required to endure hardships and abstain from luxuries.27 These
luxuries often included women. A famous anecdote (Livy 23.18.10-16) concerned the
deterioration of Hannibal’s army stationed in Capua, a South Italian city notorious for
luxury: the soldiers became “entangled with prostitutes” and effeminized by the civil-
ian luxuries.28 Many similar anecdotes exist for the Roman army; the Roman historian
Sallust depicts the late Republican general Sulla’s army as degenerated by service in
Asia:
Besides all this, Lucius Sulla, in order to secure the loyalty of the army which he led
into Asia, had allowed it a luxury and license foreign to the manners of our forefa-
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

Phang / FAMILIES OF ROMAN SOLDIERS
359
thers; and in the intervals of leisure those charming and voluptuous lands had easily
demoralized the warlike spirit of his soldiers. There it was that an army of the Roman
people first learned to indulge in women and drink; to admire statues, paintings, and
chased vases, to steal them from private houses and public places, to pillage shrines,
and to desecrate everything, both sacred and profane.29
Soldiers’indulgence inwomen iscoupled withgeneralindiscipline and moraldeterioration.
In contrast, good discipline was maintained by a celibate army. Roman tradition
had it that, in the old Republic, women were excluded from castra (military camps).
The etymologizer Servius (fourth century A.D.) defines the term castra thus: castra,
quasi casta, vel quod illic castraretur libido, nam numquam his intererat mulier
(Servius ad Aeneid 3.519). “Camps are called chaste, because there lust is castrated,
for never was a woman present in these.” Commanders restoring discipline might
throw women out of camps: in 134 B.C., Scipio Aemilianus expelled female prostitutes
from the demoralized Roman camp at Numantia in Spain.30 The exclusion of women
from military affairs was a recurrent theme in the civil war propaganda of the late
Republic. Servius, commenting on Mark Antony’s union with Cleopatra, states that “it
was not only monstrous that a Roman man married such a woman, an Egyptian, but
also because female camp-followers were regarded as shameful by our ancestors; thus
when he was about to fight, Pompey left his wife in Lesbos.”31
Whether this taboo on the presence of women in military camps existed during the
Principate is dubious. In A.D. 21, the Roman senate debated whether commanders’
wives should accompany them to their stations; despite the misogynistic language in
Tacitus, Annals 3.33, they were allowed to go. Aristocratic commanders’and officers’
wives resided in the commander’s quarters in the forts.32 Archaeologists have found
women’s and children’s shoes, brooches, rings, combs, hairpins, and beads in even the
common soldiers’barracks.33 Carol Van Driel-Murray suggests that some soldiers were
allowed to cohabit with their wives in the barracks, as in the Dutch army in nineteenth-
century Indonesia.34 However, though the archaeological evidence suggests that some
women were occasionally present in the barracks, since artifacts are mute, nothing is
known about the origin, status, or occupation of the women.
Several passages suggest that soldiers’marriage might be associated with indiscip-
line. Caesar’s Civil War 3.110 describes Roman soldiers who had “gone native”:
These consisted of soldiers of Gabinius who had habituated themselves to Alexan-
drian life and licence and had unlearnt the name and discipline of the Roman people
and had married wives by whom very many of them had children. To them were added
men collected from among the freebooters and brigands of Syria and the province of
Cilicia and the neighboring regions; also many condemned criminals and exiles had
joined them. All our own fugitive slaves had a sure place of refuge at Alexandria.
The soldiers had lost the qualities of disciplined Roman soldiers and had merged with
the dissolute Alexandrians; they are associated with brigands, condemned criminals,
and fugitive slaves. This suggests that the marriage ban maintained a distance between
the Roman army and the provincials that was necessary to legitimize Roman rule.35
Herodian 3.8.5 depicts Severus as corrupting the discipline of the Roman army by al-
lowing soldiers to marry:
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

360
JOURNAL OF FAMILY HISTORY / October 2002
To the soldiers he gave a very large sum of money and many other privileges that they
had not had before; for he was the first to increase their pay and he also gave them per-
mission to wear gold rings and to live [in marriage?] with their wives. All these things
are normally considered alien to military discipline and efficient readiness for war.
And he was the first to undermine their traditional rigor, the austerity of their mode of
life, their acquiescence in hard work and their well-disciplined respect for their offi-
cers, teaching them to covet money and turning them aside to soft living.
Herodian also associates marriage with general indiscipline: with luxury, mercenary
behavior, and a lack of respect for officers. The Greek orator Libanius denounces the
corruptionofthelaterRoman army’s officers and the poverty of the common soldiers:
They are obliged, of course, to spend their money on the wife and children—for every
one of them has both. Nothing stops them from getting married and they don’t concern
themselves with what the mothers and the children will have to live on. So when sol-
diers’ rations are so sub-divided, where can the man get his fill? The harm resulting
from this is a loss of military efficiency. This was not the case in those good old days
which I commend. Then the officers hankered after glory, not cash, and no one would
rob the soldiers of what was theirs. And the men themselves were sturdy and brave,
specialists in warfare, and they remained unmarried: it was ensured that they would
even have no need of marriage. (Oration 2.39-40)
It seems that marriage was thought to feminize soldiers, weakening their effectiveness.
Augustus, in prohibiting soldiers’marriage, presumably wanted to restore good disci-
pline after the civil wars of the late Republic; maintaining disciplina also legitimized
his rule as emperor.
The Roman military marriage ban may have influenced similar institutions in later
European history. According to David F. Noble, the gendering of military service as a
“world without women” influenced the development of early Christian monastic and
clerical celibacy; monastic rules and the hierarchy of the Church were influenced by
Roman military organization and discipline.36 The humanist Justus Lipsius, in De mili-
tia romana libri quinque (1596), praised the chastity of Roman soldiers. And, perhaps
influenced by the Renaissance humanists’ advocacy of Roman military discipline,
including celibacy, seventeenth- and eighteenth-century European military reformers
forbadesoldiers tomarry or permitted marriage only to a fraction of each regiment.37
THE SOLDIERS’ UNIONS IN ROMAN LAW
Law was the Romans’main intellectual achievement, and the Roman legal system
had great influence on subsequent canon and civil European law.38 Roman lawyers, the
jurists, wrote systematic treatises on legal topics, and Roman jurists, magistrates, and
emperors also often gave advice in or judged real cases. Many legal documents survive
in the papyri, including, for example, the Cattaoui papyrus. Excerpts from the treatises
and responses of Roman jurists and the judgments of emperors are preserved in a col-
lection known as the Corpus Iuris Civilis (Digest and Codex Iustinianus) authorized
by the sixth-century emperor Justinian.39
Law is one of our main sources for the social history of ordinary persons in the
Roman worlds, who are so little attested in the aristocratic classical authors. From
legal sources, scholars have reconstructed the law and practice of marriage, concubi-
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

Phang / FAMILIES OF ROMAN SOLDIERS
361
nage, and inheritance in the Roman world, to name only a few aspects of family his-
tory. Yet this approach poses the problem of whether average Roman citizens knew or
heeded legal institutions, using them as strategies to guide their actions in such areas as
marriage and succession. The relation of law to social practice cannot be assumed to be
one-to-one.40
The legal sources on soldiers’families pose a particular challenge in that most of the
excerpts in the Digest and Code that concern soldiers’families date from after the end
of the marriage ban. Thus, these excerpts depict soldiers’ legitimate marriages. The
juristic writings articulating the soldiers’ marriage ban before 197 were presumably
discarded as obsolete by the Byzantine compilers of the Digest. Hence, the nature of
the soldiers’unions in law before 197 is not attested. The common modern view is that
before 197, the soldiers’women were “concubines.” Concubinage in the Roman world
was a respectable alternative to marriage; it was monogamous, and the concubine was
regarded as honorable.41 But it is defined with respect to upper-class men’s relation-
ships with women of much lower status; these relationships were intended not to pro-
duce children as concubinatus was practiced chiefly by young men before legitimate
marriage or by aged widowers.42 The soldiers, of nonelite status and of middle years,
do not fit this pattern as many of their unions produced children. It is most likely that
the soldiers’ unions were de facto marriages, or matrimonia iniusta, “invalid mar-
riage.” In Latin epitaphs, the soldiers and their women represent themselves as hus-
band and wife.43 The excerpts from the Papyrus Cattaous show that the soldier’s
woman brought a dowry to the union, even though it was not legally recognized: dowry
was a characteristic of marriage.
Many such de facto marriages existed in the Roman empire, often between slaves,
who were legally unable to marry, or freedpersons and slaves.44 Casually manumitted
freedpersons (Junian Latins) also lacked the right of legal marriage. Roman citizens
could not legally marry non-Romans without a grant of conubium. These unions were
not granted legal recognition, but they were acknowledged by society and are depicted
as marriages in epitaphs. The children of these unions and of soldiers’ unions were
technically illegitimate when the mother was free; if she was a slave, the children were
born slaves. Yet these children were distinct from those actually conceived in promis-
cuity or adultery. As John Crook, an expert in Roman law says, “The opposite of mar-
riage was not ‘living in sin.’ ”45 The de facto unions should not be analogized to Peter
Laslett’s “bastardy-prone sub-society” as they were not a “social problem” in the mod-
ern sense.46
The ban’s seemingly delayed impact—its effect felt long after the formation of the
union—is based on the nature of Roman marriage as a formless union of fact.47 Legal
Roman marriage required compatible statuses (the partners must possess conubium
and be above puberty and nonincestuous), the consent of the partners’fathers or pater-
familias, and the intention to be married (affectio maritalis). No civic or religious cere-
mony was legally required, and married couples were not registered; hence, the couple
did not encounter the authorities at the formation of the marriage.48 A “ban” could not
be an actual impediment to marriage; there was no way that authorities could prevent
the formation of the unions, short of confining soldiers to forts. A contrast is provided
by eighteenth-century Göttingen, in which the soldiers of two regiments were required
to obtain official permission to marry from their commanders and to present this certif-
icate to the parish priest who performed the marriage ceremony.49
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

362
JOURNAL OF FAMILY HISTORY / October 2002
Instead, the Roman ban took the form of the denial of the legal effects of marriage.
One effect of modern European marriage, the creation of a communal property regime,
did not occur in Roman marriage of the imperial period. Roman partners’ property
remained separate and independent, except for the dowry that belonged to the husband
while the marriage lasted.50 As seen in Papyri Cattaoui, the effects of marriage became
apparent at the termination of the union by divorce or death.51 In a legitimate marriage,
the wife had the right to sue for the return of her dowry from her husband’s property
when he divorced her or died. Legal marriage resulted in legitimate children, but this
effect would have an impact chiefly if the father died and the children’s right to inherit
became an issue. The injustice is that the impact of the marriage ban might be felt many
years after the creation of the marriage, during which time the soldiers regarded them-
selves as married.
Since soldiers’ children born during service were illegitimate, they were in their
fathers’ paternal authority (patria potestas). Patria potestas was the main principle of
Roman family law; the paterfamilias, or head of the family, theoretically had absolute
power over his children, whom he could sell or kill. Such absolute power was referred
to in archaic legends; in the imperial period, patria potestas was less oppressive and
chiefly determined the children’s property rights and inheritance rights.52 The legiti-
mate children or descendants in the male line of a paterfamilias were unable to own
property during their lifetime; all family property was the father’s. The children, how-
ever, were automatic heirs to the father’s estate upon his death, should he die without a
will.
Patria potestas over children was chiefly generated through the procreation of chil-
dren in legal marriage.53 If soldier fathers died intestate, their illegitimate children had
no claim whatsoever to their property. This was ameliorated by the emperor Hadrian in
his letter of A.D. 119: he granted to soldiers’ illegitimate children the right to inherit
upon their fathers’ intestacy if no legitimate children or descendants existed. Illegiti-
mate children were not excluded from testate inheritance before the later Empire since
Roman citizens had great freedom of testation.
It is often assumed that the soldiers’children were legitimated by the military diplo-
mas’ grant of the citizenship and legitimation of the marriage. This assumption is
based on a misunderstanding of illegitimacy in Roman law and society. Illegitimacy, as
such, was not stigmatized in Roman society, in contrast with Greek city-states.54 The
dismay expressed by the Alexandrian Octavius Valens in Papyrus Cattaoui is because
Alexandria was a Greek city, in which citizenship depended on legitimate birth from
citizen parents. In Roman society, citizenship was not dependent on legitimacy; the
legitimate child of a Roman citizen woman was a Roman citizen, and Roman citizens’
slaves at manumission became citizens themselves, even though they had no legal par-
ents. The soldiers’ children born during service were illegitimate but were given
Roman citizenship (before 140 in the case of the auxilia). Legitimation of soldiers’
existing children was not conferred explicitly at the veterans’discharge, nor was it the
effect of the legitimation of the marriages, though any children born after discharge
were legitimate.55 Legitimation of existing children by subsequent marriage is a devel-
opment only of the later Roman empire (fourth through sixth centuries). It is doubtful
whether after Hadrian’s edict of 199 granted them inheritance privileges that the sol-
diers’ illegitimate children would even want to be legitimated since that would bring
them under the patria potestas of their fathers and deprive them of the right to own
property.
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

Phang / FAMILIES OF ROMAN SOLDIERS
363
It is clear that the soldiers’ children, such as Longinus’s in Papyri Cattaoui, were
socially legitimate: that is, their fathers acknowledged them in practice.56 In traditional
Roman nomenclature, illegitimate children or spurii were given the mother’s nomen or
family name, for example, the illegitimate son of Sempronia Gemella is Marcus
Sempronius Sarapio and the filliation Sp(urii) f(ilius/a), but this “correct” nomencla-
ture appears very rarely in soldiers’ children and then only in the most official con-
texts.57 Usually, in the epitaphs, a soldier’s children have the father’s nomen because he
was known to be the father. This appears in Papyrus Cattaoui 3.11-22 where the sons of
Longinus are Longinus Apolinarius and Longinus Pomponius.
The legal interpretation of the ban put the soldiers’ “wives” at a greater disadvan-
tage than the children. The soldier’s “wife” was dependent on the soldier’s good will.
As the Papyrus Cattaoui cases show, she had no legal recourse to recover her dowry
upon divorce or the soldier’s death. The soldier could name his “wife” his heir or leave
her a legacy in his will, but he was not required by law to do so, and if he left nothing to
his “wife,” she had no legal recourse. If he died intestate, she had no claim on his estate.
Due to the scarcity of wage labor for women in Roman society, unless she inherited
property from her relatives, the loss of the dowry rendered her unable to support her-
self as a widow or to remarry.58
An additional possible trap for soldiers’ women existed in a provision of the
emperor Hadrian that “a woman on whom a suspicion of disgrace may fall cannot
receive anything from a soldier’s will” (Digest 29.1.41.4) and a similar excerpt of the
jurist Papinian (Digest 34.9.14): “A woman who was convicted of fornication
(stuprum) and who cohabited with a soldier should not be permitted to benefit from the
soldier’s will.” This “disgrace” was probably sexual: since a Roman woman’s norma-
tive social role was that of a chaste married woman, a matrona or materfamilias, a
woman’s sexual activity before or outside of marriage was viewed as shameful and
was denoted as adulterium (adultery) or stuprum (illicit sexual relations in general).59
The problem is whether all women who were not married in Roman law were viewed
as committing stuprum in their sexual relationships. The jurists debated as to whether a
man should have as concubine a respectable woman or only one “with whom stuprum
cannot be committed”: women of innate dishonorable status such as slave women or
prostitutes.60
It is most probably that the women penalized in Digest 29.1.41.1 and 34.9.14 were
suspected or convicted adulteresses.61 The lex Iulia de adulteriis, an Augustan law,
established criminal prosecution for adultery and sexual relations with unmarried
women. Before this, the punishment of adultery had usually been left to the family.62 A
husband who discovered his wife’s adultery was required to divorce her and to prose-
cute; if convicted, she and her lover were exiled to different islands. Adulteresses were
deprived of inheritance rights.
It is not certain whether Digest 29.1.41.1 and 34.9.14 were meant to trap all sol-
diers’ women, who were potentially committing stuprum since they were not legally
married. However, Roman soldiers’ women, as a group, are not depicted as sexually
immoral, in contrast with the “camp followers” of early modern Europe. The Roman
soldiers’ unions appear to be long-term monogamous quasi-marriages.63 Women in
quasi-marital but legally invalid unions were not stereotyped as “whores,” and they
probably were not prosecuted for stuprum, especially if they belonged to the lower so-
cial strata. These unions were socially legitimate though not legally valid; what mat-
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

364
JOURNAL OF FAMILY HISTORY / October 2002
tered was the individual woman’s personal conduct. This is illustrated by the opinion
of the jurist Ulpian, in the early third century A.D.:
By materfamilias [as respectable married woman] we ought to understand a woman
who has lived not dishonorably; for a materfamilias is distinguished and separated
from other women by her character. So it will not make any difference, whether she is
married or divorced or widowed, or whether she is freeborn or a freedwoman; for it is
not marriage or birth which makes a materfamilias, but good character.
However, the jurists were divided over which women were liable for the crime of
stuprum, and the ambiguity of the law illustrates the vulnerability of the soldiers’
women.
The bias against soldiers’ women seems to have persisted even after the marriage
ban itself was rescinded. The jurists Papinian and Ulpian, writing in the early third cen-
tury A.D., show a certain coolness toward soldiers’ families. They define the soldier’s
wife as “not known to her husband through his military service” (Digest 49.17.6;
Ulpian 49.17.8) so that property belonging to her should not enter the husband’s
peculium castrense, the fund granted to a son in paternal power who is a soldier.64
Papinian holds that the wife’s dowry should not enter the soldier’s peculium castrense
(Digest 49.17.16.pr). The peculium castrense was apparently viewed as military; it
contained the soldier’s salary, donations, and any gifts or bequests from fellow sol-
diers, who were “known to him through military service.” Furthermore, soldiers’fami-
lies were denied the querela inofficiosi testamenti, the “complaint of undutiful will” in
which children could appeal for the invalidation of a father’s will that ignored them
(Code 3.28.9, A.D. 224).
The implication is that soldiers might prefer to make bequests to comrades. Male
bonding between soldiers in the Roman army is well attested in their epitaphs.65 Such
relationships were favored by the surviving legal interpretation of the marriage ban,
which, with such provisions as Code 3.28.9, protected the property interests of soldiers
and of their comrades against the potential claims of soldiers’“wives” and children.66
Before 197, there was nothing to stop soldiers from legating their women’s pseudo-
dowries to comrades or from leaving all their property to comrades by will. After 197,
the soldier’s “wife” was able to claim her dowry, but the family still could not contest a
soldier’s will that excluded them.
In practice, as papyrus documents show, the soldiers’families seemed able to cope
with these disabilities. Soldiers made declarations of the birth of illegitimate children,
perhaps to confirm their status so that they could inherit under the terms of Hadrian’s
letter.67 Soldiers’wills survive in which they make bequests to “wives” and to children
as well as to comrades.68 One document, BGU 327, a petition from the late second cen-
tury, attests to conflict between the interests of soldiers’ comrade heirs and soldiers’
families. The naval veteran Gaius Fabullius Macer left his estate to his fellow veteran
Gaius Longinus Castor. According to the petitioner, a woman named Aitete, Macer’s
will stipulated a legacy to her of two thousand drachmas, but Castor refused to give it to
her. The sum is similar to the “dowries” of soldiers’ women in other documents; one
may speculate that perhaps the legacy was said to be a dowry and hence was invali-
dated, or that Aitete was—or that Castor claimed that she was—suspected of stuprum,
or adultery.
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

Phang / FAMILIES OF ROMAN SOLDIERS
365
SOCIAL PRACTICE
A survey of some two thousand epitaphs of soldiers and their families is summa-
rized here because the social relationships of Roman soldiers also illustrate the tension
between the all-male institution and conjugal family life.69 Latin epitaphs often record
the deceased’s name and status, his or her occupation, and the person who set up the
inscription, besides expressing sentiments or, rarely in pagan epitaphs, hopes for an
afterlife. Hence, the epitaphs are of immense value for reconstructing the social rela-
tions of different groups in Roman society, such as slaves and freedmen or soldiers, all
strata below the elite.70 The patterns of commemoration attest to what relationships
were valued by the parties and were probably most frequent.71
The survey tabulates epitaphs by century, region, commemorator (parent, sibling,
wife, children, nonkin), and age at death. In the first century A.D., in all regions or
troops surveyed (North Africa, the Danubian provinces, the Praetorian Guard, and the
fleets), the rate of soldiers’marriage was quite low, from 5 percent to 15 percent of all
soldiers’ epitaphs; soldiers were most often commemorated by comrades.72 This pat-
tern is more marked in elite troops such as the equites singulares Augusti (the
Emperor’s Horse Guards) and the Praetorian Guard; their family formation only
increases in the third century A.D.73 Terms such as commilito (fellow soldier) and frater
(brother) are common in these epitaphs. Since first-century soldiers are rarely com-
memorated by “wives” and children, and the epitaphs of first-century soldiers’
“wives” and children are almost nonexistent, it is probable that in this period most sol-
diers did not have “wives” or children.
In nonelite troops in the second century, the frequency of “marriage,” indicating
here a de facto relationship, increases. Commemoration by “wives” increases to 35
percent of all soldiers’epitaphs. When the numbers of soldiers commemorated by chil-
dren (around 15 percent) are added, about half of all soldiers eventually “married” and
had children.
It may be that the ban was enforced in practice in the first century A.D., but as has
been argued, since the legal nature of Roman marriage made enforcement impractical,
there are other possible reasons for the infrequency of soldiers’“marriage” in the first
century. In the first century, the legionaries and most auxiliaries were recruited from
outsidetheprovinces in whichthey served;most came from Italy and the region of Pro-
vence and were stationed on the Rhine and Danube frontiers and in Britain, regions in
which they can have had few social contacts.74 Soldiers in the first century were also
paid less than in the second.75 It should be stressed that epitaphs are, in part, cultural
documents rather than just attestations of social practice, so it seems that the first-
century army’s culture may have been more “military.” A soldier who chose to be com-
memorated by his comrade might have had a “wife” and children but did not find them
worth recording. Such an attitude may have been due to the military culture’s discour-
agement of families, suggested by the passages Caesar B. Civ. 3.110, Herodian 3.8.5,
and Libanius 2.39-40.
A cultural shift in the second century may have caused the soldiers’ greater com-
memoration by “wives” and children and probably their actual greater frequency of
family formation. What is certain is that the second-century soldiers were more and
more locally recruited, from the province in which they served, even from the frontier
region. They would thus have many social contacts in the region and would find it eas-
ier to “marry.”
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

366
JOURNAL OF FAMILY HISTORY / October 2002
Soldiers’age at “marriage” is estimated by the shift in commemoration of soldiers
from parents or brothers to wives and children.76 This shift occurs in the soldiers’
midthirties, suggesting that soldiers “married” in this age range. An average age at
“marriage” in the midthirties might be related to promotion since many noncommis-
sioned officers are found in the epitaphs, and they were promoted on a fairly regular
track, after ten to fifteen years of service.77 However, the pattern is also found in sol-
diers who had not been promoted. It can have been economically motivated: soldiers
possibly saved to start households, as in the well-known late-marriage pattern of early
modern northwestern Europe. The soldiers’relatively late age at “marriage” may have
been a feature of Roman military culture, as modern comparisons suggest.78
Since homosocial relationships were apparently important to soldiers, the possibil-
ity of homosexual relationships should be raised. Homosexuality as such, defined by
the gender of the sexual partner, did not exist in Roman antiquity; sexual practices
were defined by the roles of the partners in intercourse, the active or penetrator role
being normative for freeborn males, the passive (vaginally or anally receptive) role
suitable only for women or slaves and deviant when ascribed to freeborn males.79 Sol-
diers’ active-role sexual relations with male slaves and prostitutes are recorded in the
imperial literary sources.80 A large number of male slaves were associated with the
army as horse grooms and baggage handlers, as well as soldiers’ personal servants.81
These individuals may have been soldiers’ sexual partners, especially if they were
younger. Homosexuality as such was not repressed by the Roman army; in fact, not all
modern Western armies have been equally repressive. Soldiers’ assumptions of the
deviant passive sexual role may have caused more concern, as this role was associated
with effeminacy, the opposite of the ideal soldier’s masculinity. Hence, homosexual
relations between soldiers would be disapproved of.82 But the issue is obscure, since
the Roman elite was most concerned with social control of the sexuality of its own
members; the sexual behavior of marked social inferiors, such as soldiers, was less of a
concern. It is probable that homosexual practices competed with soldiers’“marriage”
and reduced the frequency of the latter, but since neither are viewed as exclusive, there
is no way to quantify this. The Latin poet Martial’s friend, the centurion Aulus Pudens,
had a slave boy as catamite (Epigrams 1.31, 5.48) and later married a woman named
Claudia Peregrina (Epigram 4.13).
The social practices of Roman soldiers, as attested in their epitaphs, suggest that
homosocial relationships and family formation competed with one another and that
shifts in the balance occurred both from one century to another and through the sol-
diers’ life course.
CONCLUSION
The Roman army’s marriage ban illustrates some of the tensions between the mili-
tary and the family. Ideologically, the Roman army required a harsh, austere masculin-
ity that precluded indulgence in women; soldiers attached to families were viewed as
less effective. These representations were a discourse of the elite, and we do not know
how much they impinged on common soldiers in daily life. Epitaphs show that sol-
diers’formation of families was less frequent in the first century A.D., but the causes of
this pattern may be more demographic than disciplinary. The frontier provinces such
as Britain, Germany, or the Danubian region were still unpacified or undersettled, and
the soldiers were recruited from Italy and the region of Provence. In the second century
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

Phang / FAMILIES OF ROMAN SOLDIERS
367
A.D., the soldiers’family formation increased, but differences from civilian family for-
mation are found: soldiers “marry” at a later age than civilian men, and many—up to
half—of the soldiers seem never to have “married” at all.
The legal interpretation of the marriage ban suggests further tensions between
Roman military society and family life. The ban’s effects excluded the property claims
of soldiers’ “wives” and illegitimate children. Where there was no legal marriage, no
dowry was recognized, and the soldiers’women were unable to recover their dowries if
widowed or divorced; illegitimate children were excluded from inheritance if the
father died intestate. The emperor Hadrian permitted soldiers’children to inherit upon
intestacy if there were no legitimate descendants. Nevertheless, the general effect of
the ban was to defend the property interests of soldiers and of their comrades to whom
they left bequests. This was the case even after the end of the ban since soldiers’fami-
lies were denied the “complaint of undutiful will” should the soldier’s will pass them
over.
However, some features of military life in Europe that appear in the seventeenth to
nineteenth centuries seem to be absent in this period. For example, Roman soldiers’
women are not stigmatized as prostitutes or “bastard-bearers,” nor are the children
stigmatized as bastards. The social context is one in which many social groups, most of
all slaves, were denied legal marriage; but Roman legal marriage was not sharply
demarcated from illegitimate unions since the formation of legal marriage was not
juridified by required ceremonies and registration. In family formation, social legiti-
macy was not necessarily linked to legal validity. Social stigma attached chiefly to
freedpersons and slaves, as well as to members of various “degraded” professions such
as prostitution. Many soldiers owned slave women; these slave women were more
unfortunate, as slaves had essentially no rights and no protection from harsh treatment
or sale.
The Roman army took an essentially laissez-faire approach to the problem of sol-
diers’ families, in contrast with some European armies of the eighteenth and nine-
teenth centuries. It should be remarked that the Roman soldiers’marriage ban does not
seem to have been used as an instrument of direct social control, in contrast with mod-
ern armies. For instance, the German army in the eighteenth century and the Victorian
British army permitted marriage to a small proportion of enlisted men (in Britain, 6
percent of the regiment) on the grounds of good character; their wives were required to
be “respectable.”83 In recent years, such scrutiny was applied to noncommissioned
officers’ wives in the French Foreign Legion.84 The permission to marry was thus
dependent on moral surveillance. There is no sign that the Roman army regulated the
soldiers’formation of marriage in this manner; given the legal nature of the formation
of Roman marriage, it was essentially impossible to do this.
Nor did the Roman army of the first and second centuries A.D. provide material ben-
efits for a minority of soldiers’families, as have modern armies.85 The Victorian Brit-
ish army employed wives of soldiers as seamstresses and laundresses, deducted sol-
diers’pay for the support of families, and provided education for soldiers’children and
pensions for soldiers’ widows. Trustram examines how these benefits all provided
opportunities for the British command’s moral surveillance and social control of the
common soldiers and their families. Soldiers’ families were a “social problem” that
exercised the opinions of the educated classes in Victorian Britain and in the Dutch
East Indies.86 This discourse seems relatively absent from the evidence for the soldiers’
marriages in Roman sources.
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

368
JOURNAL OF FAMILY HISTORY / October 2002
The Roman army excluded women more categorically. The ban appears universal,
and there is no sign that the army employed women in any way, but this exclusion was a
minimalist solution. Only the emperor Hadrian, who had a special relationship to the
army, seems to have concerned himself with the social and moral welfare of the sol-
diers. Such an approach is consistent with Roman marriage itself, the formation of
which was essentially uncontrolled by law.
NOTES
1. For Victorian Britain, see M. Trustram, Women of the Regiment: Marriage and the Victo-
rian Army (Cambridge, 1984), 29-49; F. Barwell, Mr. Kipling’s Army: All the Queen’s Men
(London, 1981), 227. For France, see R. Cagnat, L’armée romaine d’Afrique et l’occupation
militaire de l’Afrique souls les empereurs (Paris, 1892), 440. For Germany, see R. Pröve,
“Zwangszölibat, Konkubinat und Eheschliefsung: Durchsetzun und Reichsweite obrigkeitliche
Ehebeschränkungen am Beispiel der Göttinger Militärbevölkerung im 18. Jahrhundert,” in
Familie und Familienlosigkeit, ed. J. Schlumbohm (Hannover, 1993). For Dutch Indonesia, see
H. Ming, “Barracks-Concubinage in the Indies, 1887-1920,” Indonesia 35 (1983): 65-93; A. L.
Stoler, “Carnal Knowledge and Imperial Power: Gender, Race and Morality in Colonial Asia,”
in The Gender/Sexuality Reader, ed. R. N. Lancaster and M. di Leonardo (New York, 1997), 13-
36.
2. E. M. Coffman, The Old Army: A Portrait of the American Army in Peacetime, 1784-
1898 (Oxford, 1986), 25, 138, 309-10.
3. R. K. Griffith Jr., Men Wanted for the U.S. Army: America’s Experience with an All-
Volunteer Army between the World Wars (Westport, 1982), 155-57, 202-4.
4. C. M. Petillo, s.v. “Families, Military,” in The Oxford Companion to American Military
History, ed. J. W. Chambers II (Oxford, 1999), 258-60.
5. R. McKechnie, “Living with Images of a Fighting Elite: Women and the Foreign
Legion,” in Images of Women in Peace and War, ed. S. MacDonald et al. (Madison, 1987). This
article analyzes the harshly masculine image of the legionnaire.
6. J. B. Campbell, “The Marriage of Soldiers under the Empire,” Journal of Roman Studies
68 (1978): 153-66; J. H. Jung, “Das Eherecht der römischen Soldaten,” Aufstieg und Niedergang
der römischen Welt 2.14 (1982): 302-46. These discuss the legal sources for the existence of the
ban in detail. References to earlier literature are given and discussed in S. E. Phang, The Mar-
riage of Roman Soldiers, 13 BC-AD 337: Law and Family in the Imperial Army (Leiden, the
Netherlands, 2001). The bibliography cited in this article consists of the most recent and accessi-
ble studies.
7. G. R. Watson, The Roman Soldier (London, 1969), remains a standard introduction to the
Roman military as an institution. J. B. Campbell, The Roman Army, 31 BC-AD 337: A Source-
book (London/New York, 1994), contains ancient sources in translation with commentary.
8. Dio, 60.24.3.
9. The Greek text is published in B. S. Grenfell, A. S. Hunt, and P. M. Meyer, “Papyrus
Cattaoui,” Archiv für Papyrusforschung 3 (1906): 55-105; and in L. Mitteis and U. Wilcken,
Grundzüge und Chrestomathie der Papyruskunde (Leipzig, 1912), 372; the latter edition is
reprinted in Phang, Marriage of Roman Soldiers. A partial translation is provided in N. Lewis
and M. Reinhold, Roman Civilization: Selected Readings: II. The Empire (New York, 1990),
149; and Campbell, Roman Army, 257. The translations in this article are mine. The main com-
mentaries are in German: Grenfell, Hunt, and Meyer, “Papyrus Cattaoui,” 55-105; Jung,
“Eherecht,” 310-19. Campbell, “Marriage of Soldiers,” 153-55, discusses “Papyrus Cattaoui”
more briefly.
10. J. F. Gardner, Women in Roman Law and Society (London, 1986), 105-9, 112-14;
S. Treggiari, Roman Marriage (Oxford, 1991), 350-53.
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

Phang / FAMILIES OF ROMAN SOLDIERS
369
11. R. Katzoff, “Judician Reasoning in P.Catt.: Fraus legi,” Transactions of the American
Philological Association 101 (1970): 241-52. In an amicable divorce, the soldier’s woman
would experience no problem with the recovery of her deposit; the transfer and repayment of
deposit were private contracts. Trouble arose if the soldier refused to return the deposit or if he
died without legating the deposit to her in his will. In P. Catt. 1.5-13, Lucia Macrina’s soldier
“husband” may have died intestate; this is the case in P. Catt. 6.1-23.
12. Treggiari, Roman Marriage, 366-71.
13. Berliner Griechische Urkunden 140; Select Papyri I (Cambridge, 1932), 213 (Greek and
English). Translated in Naphatali Lewis and Meyer Reinhold, Roman Civilization (New York,
1966), 149; and Campbell, Roman Army, 260.
14. J. F. Gardner, “Hadrian and the Social Legacy of Augustus,” Labeo 42 (1996): 83-100, at
87-88.
15. P. Tassistro, “Il matrimonio dei soldati romani,” Studi et Documenti di Storia e Diritto 22
(1901): 3-82; C. Castello, “Sul matrimonio dei soldati,” RISG 15 (1940): 27-119.
16. Published collections of diplomas include H. Nesselhaul, ed., Corpus Inscriptionum
Latinarum XVI (1936), supplement; M. M. Roxan, ed., Roman Military Diplomas 1954-1977
(London, 1978); idem, Roman Military Diplomas 1978-1984 (London, 1985); idem, Roman
Military Diplomas 1985-1993 (London, 1994). The bibliography on the diplomas is given in
Phang, Marriage of Roman Soldiers, appendix. Watson, Roman Soldier, 133-40, is standard.
17. Campbell, Roman Army, 322-323.
18. The wives received citizenship only in exceptional grants.
19. A. A. Aly, The Roman Veterans in Egypt (Ph.D. diss., University of Michigan, 1949), 34-
35; Hartmut Wolff, “Zu den Bürgerrechtsverleihungen an Kinder von Auxiliarien und
Legionaren,”Chiron4(1974):479-510, at487;Campbell, “Marriage ofSoldiers,” 165, n. 89.
20. Some believe the formula was intended to prevent simultaneous polygamy, especially
among non-Roman soldiers and/or veterans. L. Allason-Jones, “Women and the Roman Army
in Britain,” in The Roman Army as a Community, ed. A. Goldsworthy and I. Haynes
(Portsmouth, RI, 1999), 41-51, at 49. Whether non-Roman societies were polygamous is
unclear from the notoriously unreliable ethnographic descriptions in ancient authors. It is more
likely that the formula prevented serial polygamy as a means of status usurpation (veteran mar-
ries, divorces, and remarries several times so as to provide more than one woman’s children with
Roman citizenship).
21. H. M. D. Parker, The Roman Legions (Cambridge, 1928), 245; Watson, Roman Soldier,
133.
22. Trustram, Women of the Regiment, 40, 42, 49; Stoler, “Carnal Knowledge,” 18.
23. Campbell, Roman Army, 152; Allason-Jones, “Women and the Roman Army,” 47.
24. C. M. Wells, “Celibate Soldiers: Augustus and the Army,” American Journal of Ancient
History 14, no. 2 (1998): 180-90.
25. Women could not vote or hold public office or military posts; their participation in legal
processwasrestricted. J. F. Gardner, Beinga Roman Citizen (London/NewYork, 1993), 81-89.
26. C. A. Williams, Roman Homosexuality: Ideologies of Masculinity in Classical Antiquity
(Oxford, 1999), 127, 132.
27. Masculinity is associated with austerity in Greek thought; the ideal male dominates and
restrains his physical desires, whereas the man who indulges in the sensual pleasures is
effeminized. M. Foucault, The Use of Pleasure (New York, 1985); M. Foucault, The Care of the
Self (New York, 1986); C. J. Berry, The Idea of Luxury: A Conceptual and Historical Investiga-
tion (Cambridge, 1994), 59. Military discipline is an intensification of this ethic. E. L. Wheeler,
“The Laxity of Syrian Legions,” in The Roman Army in the East, ed. D. L. Kennedy (Ann Arbor,
MI, 1996), 229-76, at 232-3.
28. On these indiscipline anecdotes, see Wheeler, “Laxity of Syrian Legions,” and on
Hannibal, 238.
29. Sallust, The War with Catiline, 11.
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

370
JOURNAL OF FAMILY HISTORY / October 2002
30. Appian, The Spanish War, 85.
31. Servius, Commentary on the Aeneid, 8.688.
32. M.-Th. Raepsaet-Charlier, “Epouses et familles des magistrates dans les provinces
romaines aux deux premiers siècles de l’empire,” Historia 31 (1982): 56-69; L. Allason-Jones,
Women in Roman Britain (London, 1989), 53-56.
33. R. MacMullen, Soldier and Civilian in the Later Roman Empire (Cambridge, 1963), 126;
C. Daniels, “Excavations at Wallsend and the Fourth-Century Barracks on Hadrian’s Wall,”
Roman Frontier Studies 1979 (Oxford, 1980), 173-91; for a more comprehensive survey, see
L. Allason-Jones, “Sexing Small Finds,” in Theoretical Roman Archaeology: Second Confer-
ence Proceedings, ed. P. Rush (Aldershot, UK, 1995): 22-32. On shoes, also adolescent boys’
shoes, see Carol Van Driel-Murray, “Gender in Question,” in Theoretical Roman Archaeology:
Second Conference Proceedings ed. P. Rush (Aldershot, UK, 1995), 3-21.
34. Van Driel-Murray, “Gender in Question,” 12ff. She cites literature in Dutch that was not
accessible to me; in English, see Ming, “Barracks-Concubinage in the Indies, 1887-1920,” 65-
93; Stoler, “Carnal Knowledge.” Cf. Trustram, Women of the Regiment, 70-72, on the “corner”
system, in which the 6 percent of the regiment that were officially allowed to marry were permit-
ted to house their families in a curtained-off area of the barracks. The lack of privacy appalled
Victorian morality when British women were involved; the Dutch were not concerned about the
morality of soldiers’ native concubines.
35. June, “Eherecht,” 334-35; B. D. Shaw, “Soldiers and Society: the Army in Numidia,”
Opus 2 (1983): 133-53; R. Alston, Soldier and Society in Roman Egypt (London, 1995), 59-60.
Various legal privileges and restrictions, besides the marriage ban, separated Roman soldiers
from civilians.
36. D. F. Noble, A World without Women: The Christian Clerical Culture of Western Science
(New York, 1992): 53-55.
37. A. Corvisier, Armies and Societies in Europe 1494-1789, trans. A. T. Siddal
(Bloomington, IN, 1979), 174-75; Pröve, “Zwangszölibat,” 81-96.
38. D. Johnston, Roman Law in Context (Cambridge, 1999), 135.
39. On the sources of Roman law, see O. F. Robinson, The Sources of Roman Law (London/
New York, 1997), 56-66; Johnston, Roman Law in Context, 2-26.
40. R. P. Saller, Patriarchy, Property, and Death in the Roman Family (Cambridge, 1994),
155-60.
41. S. Treggiari, “Concubinae,” Papers of the British School at Rome 49 (1981): 59-81;
T. A. J. McGinn, “Concubinage and the lex Iulia on Adultery,” Transactions of the American
Philological Association 121 (1991): 335-75.
42. R. P. Saller, “Slavery and the Roman Family,” Slavery and Abolition 8 (1987): 65-87.
43. Terms for partners are tabulated by M. M. Roxan, “Women on the Frontiers,” Roman
Frontier Studies 1989, ed. V. A. Maxfield and M. J. Dobson (Exeter, UK, 1991), 462-67, at 462;
marital terms predominate, such as coniu(n)x (spouse).
44. On slaves’ and freedpersons’ families, see P. R. C. Weaver, Familia Caesaris (Cam-
bridge, 1972); P. R. C. Weaver, “The Status of Children in Mixed Marriages,” in The Family in
Ancient Rome: New Perspectives, ed. B. Rawson (Ithaca, NY, 1986), 145-70; P. R. C. Weaver,
“Where Have All the Junian Latins Gone? Nomenclature and Status in the Early Empire,”
Chiron 20 (1990): 275-304; P. R. C. Weaver, “Children of Freedmen (and Freedwomen),” in
Marriage, Divorce and Children in Ancient Rome, ed. B. Rawson (Oxford, UK, 1991), 166-90;
P. R. C. Weaver, “Children of Junian Latins,” in The Roman Family in Italy: Status, Sentiment,
Space, ed. B. Rawson and P. R. C. Weaver (Oxford, UK, 1997), 55-73; B. Rawson, “Family Life
among the Lower Classes at Rome in the First Two Centuries of the Empire,” Classical Philol-
ogy 61 (1966): 71-83; B. Rawson, “Roman Concubinage and Other De Facto Marriages,” Trans-
actions of the American Philological Association 104 (1974): 279-305; B. Rawson, “Spurii and
the Roman View of Illegitimacy,” Antichthon 23 (1989): 10-41; J. F. Gardner, “Stumbling-
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

Phang / FAMILIES OF ROMAN SOLDIERS
371
Blocks for Lower-Class Families in Rome,” in The Roman Family in Italy: Status, Sentiment,
Space, ed. B. Rawson and P. R. C. Weaver (Oxford, UK, 1997), 35-53.
45. J. A Crook, Law and Life of Rome, 90 B.C.-A.D. 212 (Ithaca, NY, 1967), 99.
46. Rawson, “Spurii,” 11, 20-21. See H. C. Youtie, “Apatores: Law vs. Custom in Roman
Egypt,” in Le Monde Grec: Pensée, Litterature, Histoire, Documents: Hommages à Claire
Préaux (Brussels, 1975), 723-40, at 738-39, on types of invalid marriages in Roman Egypt
(including soldiers’ marriages). On Bastardy-prone subsociety, see P. Laslett, Bastardy and Its
Comparative History: Studies in the History of Illegitimacy (Cambridge, 1980), 221.
47. R. P. Saller, “European Family History and Roman Law,” Continuity and Change 6
(1991): 335-46, at 336-38, stresses differences of Roman law from later European law and social
practice.
48. Crook, Law and Life of Rome, 99-103; Treggiari, Roman Marriage, 37-48, 54-57. Mar-
riage was gradually “juridified” in early modern Europe; the Council of Trent (1563) made a
church ceremony and registration a legal requirement, followed by the development of the civil
ceremony in revolutionary France. M. A Glendon, The Transformation of Family Law: State,
Law and Family in the United States and Western Europe (Chicago, 1898), 28ff.
49. Pröve, “Zwangszölibat.”
50. Treggiari, Roman Marriage, 365-66. In old-fashioned manus marriage, the wife’s entire
property became the husband’s and she owned nothing of her own, but this type of marriage was
obsolete by the first century B.C..
51. J. F. Gardner, “Gender-Role Assumptions in Roman Law,” EMC/Classical Views 39, n.s.
14 (1995): 377-400, at 392: “The only strictly legal relationships between husband and wife are
strictly important in the main only at the end of the marriage.”
52. Crook, Law and Life ofRome, 106-11;Saller, Patriarchy, Property, andDeath, 102-32.
53. It also could be created by adoption: J. F. Gardner, Family and Familia in Roman Law and
Life (Oxford, 1998).
54. For Roman society, see Rawson, “Spurii,” 17-18; for Greek and Hellenistic society, see
D. M. MacDowell, The Law in Classical Athens (Ithaca, NY, 1978), 67-68; and D. Ogden, Greek
Bastardy in the Classical and Hellenistic Periods (Oxford, UK, 1996), 203-8 (i.e., bastardy
associated with slavery, poverty, sterility, disease). However, in Greco-Roman Egypt, the atti-
tudetoward illegitimatechildren (apatores, “fatherless”)was tolerant:Youtie, “Apatores,” 737.
55. The legal argument behind this sentence is found in Phang, Marriage of Roman Soldiers,
chap. 10. Legitimation by subsequent marriage is a development of Byzantine law. A. Arjava,
Women and Law in Late Antiquity (Oxford, UK, 1996), 210-17.
56. The families of present-day long-term monogamous cohabitants serve as a comparison:
the children are biologically the father’s, and he raises and supports them. Glendon, Transfor-
mation of Family Law, 252ff. Glendon speaks of “informal family relations.” “Bastard” has
become an archaic term for such socially legitimate children.
57. For example, in birth declarations (P. Mich. III 169) and status examinations in Roman
Egypt, SB 5217; P. Oxy. 1451. Epitaphs were not official legal documents. Rawson, “Spurii,” 30,
suggests that Sp. f. was used chiefly to emphasize free birth.
58. Regarding wage labor, see S. Treggiari, “Lower Class Women in the Roman Economy,”
Florilegium 1 (1979): 65-86, at 78-79; J. K. Evans, War, Women and Children in Ancient Rome
(London/New York, 1991), 115. Regarding the dowry as support for widows, see Gardner,
Women in Roman Law and Society, 107; Saller, Patriarchy, Property, and Death, 211. To what
extent Roman women’s dowries were their share of their patrimony is disputed.
59. Treggiari, Roman Marriage, 262, 299ff.
60. McGinn, “Concubinage,” analyzes the jurists’ different positions on this issue.
61. T. A. J. McGinn, Prostitution, Sexuality, and the Law in Ancient Rome (Oxford, UK,
1998), 117.
62. The evidence for the law is collected in translation in M. R. Lefkowitz and M. B. Fant,
Women’s Life in Greece and Rome: A Sources Book in Translation (Baltimore, 1992), 104-10,
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

372
JOURNAL OF FAMILY HISTORY / October 2002
117-18; see the discussion in Treggiari, Roman Marriage, 277-90; McGinn, Prostitution, 140-
47.
63. Soldiers are sometimes attested as owning more than one female slave; the most noted
example is the papyrus Berliner Griechische Urkunden 326, the will of the veteran C. Longinus
Castor who frees his two slave women Marcella and Cleopatra. Epitaphs also attest fleet sol-
diers’ownership of two slave women. Although Roman slave owners could make sexual use of
their slaves (of both sexes), such a relationship cannot be proven in these cases. The master’s
sexual relations with several slave women were not polygamy since these relationships were not
viewed as marriage.
64. The soldier could dispose of or bequeath this peculium castrense as he pleased, even
though he did not technically own it. J. B. Campbell, The Emperor and the Roman Army: 31 BC-
AD 235 (Oxford, UK, 1984), 229-36.
65. R. MacMullen, “The Legion as a Society,” Historia 33 (1984): 440-56; R. MacMullen,
Changes in the Roman Empire: Essays in the Ordinary (Princeton, NJ, 1990), 226-27.
66. This is suggested by Wells, “Celibate Soldiers,” 189-90, from the point of view of the
soldiers.
67. Regarding birth declarations, see R. Cavenaile, ed., Corpus Papyrorum Latinarum
(Wiesbaden: O. Harrassowitz, 1958, hereafter CPL) 159; Lewis and Reinhold, Roman Civiliza-
tion II, 481; Berliner Griechische Urkunden 1690; CPL 160; Campbell, Roman Army, 256; CPL
161; P. Mich. III 169; CPL 162; Lewis and Reinhold, Roman Civilization II, 482.
68. Regarding soldiers’and veterans’wills, see Berliner Griechische Urkunden 326; Select
Papyri I 85; Lewis and Reinhold, Roman Civilization II, 191-92; Campbell, Roman Army, 175;
J. Rowlandson, Women and Society in Greek and Roman Egypt: A Sourcebook (Cambridge,
1998), 139; P. Lugd. Bat. XIII 14; CPL 221; Berliner Griechische Urkunden 1662; P. Wisconsin
I 14. Regarding the general discussion of soldiers’wills, see Campbell, Emperor and the Roman
Army, 210-29.
69. This surveyis discussed inchapter6 ofPhang, Marriage ofRoman Soldiers. Partofitwill
be published as “The Timing of Marriage in the Roman Army” in the Limes XVIII: Eighteenth
International Conference of Roman Frontier Studies collection (Oxford, UK: British Archaeo-
logical Reports series, forthcoming).
70. R. P. Saller and B. D. Shaw, “Tombstones and Roman Family Relations in the Principate:
Civilians, Soldiers, and Slaves,” Journal of Roman Studies 74 (1984): 124-55, at 126-27; B. D.
Shaw, “The Cultural Meaning of Death: Age and Gender in the Roman Family,” in The Family in
Italy from Antiquity to the Present, ed. D. Kertzer and R. P. Saller (New Haven, CT, 1991), 66-
90; G. Woolf, Becoming Roman: The Origins of Provincial Civilization in Gaul (Cambridge,
1998), 77-105.
71. Saller and Shaw, “Tombstones and Roman Family Relations”; Shaw, “Cultural Meaning
of Death,” 67-68.
72. A survey of auxiliary tombstones was made by Roxan, “Women on the Frontiers,” 462-
64, and finds the same temporal pattern.
73. S. Panciera, “Soldati e civili a Roma nei primi tre secoli dell’impero,” in Prosopographie
und Sozialgeschichte: Studien zur Methodik und Erkenntnismöglichkeit der kaiserzeitlichen
Prosopographie, ed. W. Eck (Wein, 1993), 261-76, at 268-69; M. P. Speidel, Riding for Caesar:
The Roman Emperors’Horse Guards (Cambridge, 1994), 133-35.
74. G. Forni, Il reclutamento delle legioni da Augusto a Diocleziano (Rome, 1953), 82-99;
J. C. Mann, Legionary Recruitment and Veteran Settlement during the Principate (London,
1983), 49-66.
75. R. Alston, “Roman Military Pay from Caesar to Diocletian,” Journal of Roman Studies
84 (1994): 113-23.
76. The method is used for civilians by B. D. Shaw, “The Age of Roman Girls at Marriage:
Some Reconsiderations,” Journal of Roman Studies 77 (1987): 30-46, for women; R. P. Saller,
“Men’s Age at Marriage and Its Consequences in the Roman Family,” Classical Philology 82
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015

Phang / FAMILIES OF ROMAN SOLDIERS
373
(1987): 21-34, for men, estimating civilian men’s age at marriage as the late twenties. Full bibli-
ography and graphs are found in Phang, Marriage of Roman Soldiers.
77. D. J. Breeze, “The Organisation of the Career Structure of the Immunes and Principales
of the Roman Army,” Bonner Jahrücher 174 (1974): 245-92; B. Dobson, “The Significance of
the Centurion and ‘Primipilaris’ in the Roman Army and Administration,” Aufstieg und
Niedergang der römischen Welt II.1 (1972): 392-434.
78. Trustram, Women of the Regiment, 34, displays a table (Table 3.2) showing that between
the ages of twenty-five and thirty, when more than 50 percent of civilian men were married, only
10 to 14 percent of Victorian British soldiers were married and that the percentage of married
soldiers reached 50 percent only in the forty- to forty-five-year-old age division. The reason may
lie in the military subculture. Byron Farwell, Mr. Kipling’s Army (New York, 1981), 233-35,
shows that early marriage was thought to interfere with Victorian subaltern officers’ careers.
Regarding the late-twentieth-century U.S. military, C. Enloe, Maneuvers: The International
Politics of Militarizing Women’s Lives (Berkeley, CA, 2000), 261, stated that “especially the
lower-ranking marines . . . seemed to practice marriage-avoidance.”
79. Williams, Roman Homosexuality.
80. Conquering soldiers inflicted rape on both sexes: Sallust, War with Catiline, 51.9;
Cicero, Letters to His Friends, 3.31; Cicero, Against Verres, 2.4.116; Tacitus, Histories, 3.33.
Martial wrote erotic epigrams about officers’boys: Epig. 1.31, 5.48, 8.63, and 13.69 all concern
the centurion Aulus Pudens; Epig. 9.56, 3.91, 7.80. The tone of these poems is not scurrilous,
except for 3.91.
81. M. P. Speidel, “The Soldiers’ Servants,” Ancient Society 20 (1989): 239-47; J. P. Roth,
The Logistics of the Roman Army at War (264 BC-AD 235) (Leiden, the Netherlands, 1999),
101-5, 111-12.
82. It is not possible to identify soldiers’relationships with comrades, however intense emo-
tionally, as homosexual relationships, unless there is specific evidence of an erotic relationship.
Romans would assume that in a homosexual relationship, one party must be the passive partner
and thus would be disgraced.
83. Trustram, Women of the Regiment, 190-201, explores in detail the disciplinary nature of
these welfare benefits; Pröve, “Zwangszölibat.”
84. McKechnie, “Living with Images,” 132-33.
85. In the later Roman period (fourth through sixth centuries A.D.), the state provided rations
and transport for soldiers’families. A. H. M. Jones, The Later Roman Empire 284-602 (Oxford,
UK, 1964), 630-31. These do not seem linked to conduct in any way. The policy, attested in the
imperial lawsof the period, fluctuated, as Libanius 2.39-40 attests, but verylittle is known about it.
86. Trustram, Women of the Regiment, 190-201.
Downloaded from jfh.sagepub.com at EMORY UNIV on April 19, 2015