CONCLUSION 233which won formal recognition in imperial rescripts from the early second
century or earlier. Slaves were chattel; their humanity was given some limited
recognition in the law, again through the decisions of emperors.
Turnover in senatorial and equestrian families was extremely high by any
historical standards; these orders and to a lesser extent the urban elites were
in constant need of replenishment from below. Ex- soldiers and ex- slaves
formed two upwardly mobile groups. The promotion of veterans was an
outcome of the professionalization of the army. Pay and donatives were
suffi cient to enable veterans to retire with modest, and in the case of offi cers,
substantial, wealth and assume positions of responsibility in local government.
In contrast, the emancipation of slaves was a private affair; Augustus
regulated but did not block the practice. Thus ex- slaves, selected as suitable
recipients of property by wealthy men who lacked natural heirs or adopted
sons, contributed a steady trickle of sons to the local aristocracy. In this way
manumission played a part in the wealth- transferring process. In the East,
where for technical reasons men of servile origin are less easy to pick out in
the relevant Greek- language documents, it is a safe inference that the local
elite replaced itself with select clients, freed or freeborn. In a society where
wealth was in land and transmitted through the family, a propertied class
that could not reproduce itself was replenished through controlled cooptation.
Augustus was aware of the importance of the family in society; he tried
to reduce social mobility at the top of the social hierarchy by encouraging
senators to marry, bear children and keep their property within the family.
He was attempting the impossible, essentially because senators had devised
what were to them satisfactory alternatives to constant childbearing, in
particular, recourse to natural daughters as successors and the adoption of
adult sons. In general, emperors were unwilling to bring Roman law as it
related to the family into line with social behaviour. The contrast between
legal principles and social realities is nowhere clearer than in the matter of
parental authority, though the scale of the contrast has escaped modern
commentators. The standard image of the Roman family as a patriarchal
household ruled by an authoritarian, elderly paterfamilias and including his
wife, sons and unmarried daughters, plus his sons’ children is untenable. In
particular, low life expectancy at birth (about twenty- fi ve), the late age of
marriage of men (the late twenties), and therefore the generational age gap
(about forty) substantially reduced the effects of paternal authority over
sons. Few fathers, around 20 per cent (25 per cent in the case of aristocratic
men), were alive at the time of their son’s marriage. Women married younger,
at 13 or 14 if aristocrats, in their late teens or early twenties otherwise.
Many lacked fathers to witness their marriages (in the case of non- aristocratic
women, this was true of more than half). As to husband–wife relationships,
the effects of the ideology of inferiority and the age differential between
wives and husbands have to be weighed against the wife’s independent
control of her own property after her father’s death, her right to divorce and
to take much of the (typically modest) dowry.