III SUCCESSION
I. Wills
It was open to any Roman citizen who was of age to make a will; a woman would require the authority of her tutor.
The law of succession is one of the most complex areas of Roman law, and there was a large number of formal requirements which had to be met for the will to be valid. The full details need not be considered here, but included such things as formal words for the appointment of heirs and legatees; the need for the appointment of the heir to be made before any other dispositions were made; the need for children and certain others, if they were not being appointed as heirs, to be disinherited either by name or in a general clause; and attestation by a number of witnesses. What is important is that all of these requirements were requirements of form and not content. Provided the testator succeeded in complying with the form, which with legal advice need not have been difficult, the will could leave property to anyone the testator wanted.Non-compliance with the formal requirements meant invalidity and intestacy. An interesting illustration is given by the following text:
‘I, Lucius Titius, have made this will without any legal expert, observing the reason of my own mind rather than excessive and miserable pedantry; if I have done anything without due legality or skill, let the wishes of a sane man be treated as valid in law.’ He then appointed his heirs. A question arose when the property was claimed on intestacy... (Scaevola, D. 31.88.17)
The jurist does not trouble to say what was wrong with the will. But the testator’s plea ad misericordiam was evidently rejected; the will was void and therefore intestacy supervened.
Given the number of pitfalls, it may be that legal help in making a will was usual.Content of wills
Wills had to appoint an heir or heirs. The heir was responsible for continuing the deceased’s family sacra or religious observances. The heir also succeeded not only to all the deceased’s assets and rights but also to his or her obligations (so far as they were capable of surviving his or her death). This applied in particular to debts. The position of heir was therefore not merely a responsible but might also be an unprofitable one, since the debts of the inheritance might well exceed its assets and the heir would none the less remain liable to pay them in full. There is good evidence that, to spare their children these responsibilities and burdens, some testators preferred to disinherit them, and appoint someone else as heir charged with the duty of transferring property to the children, usually by way of fideicommissum (see below; Ulpian, D. 28.2.18 and D. 38.2.12.2). By this means the children could be spared the burdens which heirship imposed. Various technical devices were in due course devised to attempt to improve the position of the heir (Buckland 1963: 304—6, 316—19).
Apart from the appointment of an heir, the contents of a will were optional. Typically, a paterfamilias might appoint tutors for his children in minority; free some well-deserving slaves; and make legacies.
Legacies are of particular interest, because all the evidence is to the effect that these were so far from being few and modest tokens of esteem that they had actually to be curtailed by statute (Duncan-Jones 1982: 21). Much property was dispersed among many recipients; in part these, it seems, were payment for what during the testator's lifetime had been obtained through friendship and patronage. There was a series of statutes restricting the testator's power to dissipate his estate by means of legacies.
Here it is enough to note the last of those statutes, the lex Falcidia of 40 bc, which provided that after paying legacies the heir (or heirs) must be left with a quarter of the net estate; if the total legacies exceeded that quarter, the legacies were abated in order to preserve it. This rule provided the jurists with a number of thorny technical problems, and the remnants of their discussions can be read in D. 35.2. What is most striking is that anybody should think of leaving more than three-quarters of his or her estate to persons other than the heirs appointed under the will.Legacies Several books of the Digest are entirely made up of discussion about types of legacies. They provide an invaluable picture of the sorts of property a Roman testator might have and might wish to single out and leave in a bequest. All that is possible here is to note some of the categories of legacy to which the jurists devoted particular attention: dowry; wine, corn or oil; farm equipment; peculium; food (perns); furniture; jewellery; freedom. The detailed discussions of the jurists can be found in books 33 and 34 of the Digest.
It is worth commenting in slightly more detail on four types of legacy.
(1) Usufruct is discussed in more detail in the next chapter, as an institution of the law of property. For the purposes of this chapter, what is important is that this enabled a testator to leave the ownership of his estate to one person but a legacy of the usufruct in it to another: the legatee was entitled to use the property and take its income or fruits, either for a fixed term or for life. By far the commonest arrangement was that a paterfamilias left a usufruct in his estate to his widow for her lifetime and ownership of the property to the children (D. 33.2).
(2) Legacies of annuities.
A title of the Digest is devoted to legacies of annuities, that is payment of annual sums for the remainder of the lifetime of the payee (D. 33.1). The jurists construed these as a series of annual payments, of which the first was unconditional and the remainder were subject to the condition that the legatee was alive at the time due for payment (Paul, D. 33.1.4). There are examples of annuities in favour of wives, children, and freedmen. It might be necessary to value the annuity in order to ensure that the lex Falcidia was complied with. An intriguing text of Aemilius Macer sets out how this was to be done, by making assumptions about the legatee's life expectancy at a given age: the range is from five years for those of age sixty or more to thirty years for those between birth and age twenty (D. 35.2.68 pr.). These assumptions, although not generous, appear not to be too unrealistic (Frier 1982; Duncan-Jones 1990: 93-104).(3) A similar sort of arrangement was the legacy of an allowance for food or clothing (alimenta). The Digest suggests that this would most commonly be left to freedmen and might be for payment in monthly or annual instalments for a fixed period of years, or until the age of majority, or for life (D. 34.1).
(4) Legacies for purposes. Philanthropists seem to have favoured what the jurists called legacies sub modo, legacies for purposes to be carried out. Inscriptions as well as texts in the Digest attest a large number of these, for such things as providing games, heating public baths, paving roads, and constructing buildings, or for the commemoration of the deceased. Here is one example: ‘Lucius Titius left a legacy in his will of 100 to his hometown, Sebaste, so that from the interest on it games should be celebrated in his name every other year...' (Scaevola, D. 33.1.21.3). In order to ensure the durability of such arrangements, the legatee chosen would have to be a non-natural person, such as a municipality: the legatee was then liable, so long as it existed, to perform the purpose for which the legacy had been left.
In classical law, however, such arrangements suffered from the weakness that the modus, the purpose, could not be enforced as such: the legatee could only be forced to pay damages. (This was a general rule of the classical procedural system: see chapter 6.) There was no guarantee therefore that arrangements of this sort would be observed. In later classical law, however, it looks as if the purpose might have been enforced as such if there was a public interest.2. Fideicommissa or trusts
The most versatile institution of the law of succession was the fideicommissum or ‘trust'. This was a device which - like a legacy or succession under a will - operated only on the death of the person establishing it. But whereas appointment as heir was made directly by the testator, and a legacy was paid directly by the heir to the legatee, the trust worked indirectly, through an intermediary. The person making the fideicommissum would entrust (committere) property to the faith (fides) of an intermediary, for it to be conveyed ultimately to the beneficiary. The intermediary was usually, but did not have to be, the testator's heir.
Initially trusts were not recognized by the law, and the only entitlement the beneficiary had was a moral one which was not enforceable in the courts. That changed when the emperor Augustus introduced a jurisdiction charged with the enforcement of such trusts; at first the consuls were responsible for this jurisdiction. Later on, a special praetor performed the same role. An institution which had depended purely on faith and friendship now came to depend on legal obligation.
The trust was very versatile, but we can review at least its principal uses.
(1) It was used to transfer property to beneficiaries who as a matter of ius civile were unable to receive legacies or inheritances: for example, foreigners, proscribed criminals, those debarred under statutes such as the lex Voconia or the Augustan marriage legislation.
There is good evidence of this in the works of Cicero (fin. 2.55). Although it is hard to believe that this freedom remained wholly untrammelled once trusts were actionable, there is none the less evidence of their being used for avoidance of civil-law restrictions; and a series of measures through the first and second centuries ad closing loopholes confirms the — albeit diminishing — utility of the trust for getting round inconvenient rules of civil law. There are strong parallels here not just with the development of trusts in the common law but also that of equivalent fiduciary devices such as Treuhand on the continent: they too were initially associated with circumvention of the strict rules of law.(2) Under civil law the appointment of an heir was permanent. The trust, however, allowed appointment of an heir to be followed by a transfer of the whole or part of the inheritance to someone else. Gaius deals with this case ‘When we have written “Let Lucius Titius be heir”, we can add “I ask and request of you, Lucius Titius, that as soon as you are able to accept the inheritance you make it over to Gaius Seius.” ' (Inst. 2.250). In this case the heir was obliged to transfer the inheritance immediately. But often the heir was asked only to transfer it after an interval; a common case was for the transfer to be made on the trustee's death (Papinian, D. 35.1.102; D. 36.1.56 and 60.8). Here the trust provided an alternative to usufruct for granting someone a life interest in property (the difference was that, whereas under usufruct the person enjoying the life interest did not own the property, under a trust he or she did, but with the obligation to transfer it to the beneficiary of the trust). This method too seems to have been used to benefit the testator's widow during her lifetime, the property being directed ultimately to the children. There is some reason to think that this device may occasionally have been used to protect the interests of children of first marriages as against their stepmother (Humbert 1972: 207-40; but see Treggiari 1991: 392).
(3) Trusts could be imposed not only on the heir but on any person who acquired a benefit from the deceased, even if the benefit was very modest and essentially transitory, most of it being passed on to the ultimate beneficiary. That included heirs who succeeded on intestacy. The reasoning behind this was quite simple: ‘Trusts can be charged on heirs on intestacy, since the paterfamilias is regarded as intentionally leaving them his estate on intestacy' (Paul, D. 29.7.8.1). Here the trust opened up a novel path: it was possible to die without making a will, yet still direct where some or all of the estate was to go.
(4) Because a trust could be charged on a person other than the heir it was able to serve purposes going beyond the generation of the deceased's immediate successors. For example, the testator might appoint someone who received land under his will as trustee under a trust in favour of a further beneficiary, whether that was a named individual or simply a member of the family. Here is one example, provided by Scaevola: ‘A father prohibited his son and heir from alienating or mortgaging lands and entrusted to him that they would be preserved for his legitimate children and other relatives' (D. 32.38 pr.). Although in this case the son became the owner of the land, it was not his to dispose of; and its fate was already regulated by his father's will. This arrangement could extend over several generations, although classical law appears to have insisted that it was valid only so far as the beneficiaries were identifiable: the most remote beneficiaries who were still regarded as identifiable were the immediate issue of those living at the date of the settlor's death (Modestinus, D. 31.32.6). Those who have read a lot of English eighteenth- and nineteenth-century fiction are inclined to imagine that, as with the English entail, much of the land of Roman nobles would be tied up by this sort of device. But in fact there is little reason to believe that this was so.
g. Challenges to wills
So far the discussion has been concerned mainly with the content of wills, but the formal requirements for their validity have also been mentioned. From the late republic, however, there was also a substantive ground on which a will might fail: it might be challenged by the so-called ‘complaint against an undutiful will’ (querela inofficiosi testamenti). This was an action which could be brought by a descendant or ascendant of the testator if he or she had been left less than a quarter of what would have been his or her share, had the testator died intestate. For example, if the testator had three children, each had a prospective share of one third of the estate, and each could bring the querela if left less than one twelfth of it. The result of a successful claim was that the estate fell into intestacy, so the claimant received his or her full intestate share. (There might be additional complications if the claimant chose to challenge only some of the appointed heirs; in that case, the will need not be set aside in its entirety, and bequests made in the will might stand, although scaled down proportionately.)
Although it is not possible to follow through the logic of this argument (as Marcian explains in D. 5.2.2), the jurists appear to have taken over from rhetoric the notion that the invalidity of the will was in some sense founded on the testator’s insanity at the time of making the offending dispositions. In any case, to succeed in this claim, the claimant not only had to be within the necessary degree of relationship to the testator but must also have been unduly passed over. There were perfectly good reasons for disinheriting relatives, but it is rather interesting that the jurists do not appear to discuss them. It may be that they took the high- minded view that this was a matter for the rhetoricians. Imperial constitutions, however, do go into this question: people validly excluded from benefiting include those leading immoral lives and gladiators (C. 3.28.19 (ad 293); C. 3.28.ii (ad 225)).
The querela is the one and only substantial restriction on freedom of testation in Roman law: a testator simply had to leave these individuals the requisite amount unless he or she had a solid reason for not doing so. But it is worth emphasizing that this tied the testator’s hands only to the extent of one quarter of the estate. For the rest he or she was free. By the standards of some modern jurisdictions, the Roman regime was extremely liberal.
It is interesting that two restrictions on absolute freedom of testation — the querela and the lex Falcidia — emerged at much the same time, in the late republic. It is tempting to connect this with a breakdown of existing conventional mores amid the turmoil of the end of the republic. What could previously be taken for granted as social practice now had to be laid down as the law (Paulus 1994).
4. Intestate succession
Intestacy might come about either because no will was made or because the will was invalid. The second of these has already been mentioned: breach of formal requirements might render a will invalid; so might other contingencies, such as the birth of a legitimate child to the testator after the will was made.
It is a vexed question how common it was to make no will (Daube 1965 and 1969: 71—5; Crook 1973; Cherry 1996). Sir Henry Maine famously spoke of the Roman ‘horror of intestacy'. There is anecdotal evidence which suggests that making a will was the norm: Cato is notoriously said to have regretted having lived a single day intestate (Plutarch, Cato maior 9.6). But in this respect as others he may not have been typical. Those who had nothing to leave are likely to have left no will. But the evidence seems to suggest that, motivated by horror or otherwise, the propertied classes at Rome typically did make wills. The material which survives suggests that men were significantly more likely to make wills than women (Champlin 1991: 46—9). Perhaps women were less prone to feeling horror.
It is of course a powerful incentive to make a will that, in its absence, one's property will be dispersed to people one would not wish to receive it. We must look therefore at the rules which applied for distribution of an estate on intestacy.
The rules were these. Children who became independent (that is, were released from paternal power) on the death of the deceased had the first claim to the estate, in equal shares. They were known as sui heredes. There might not be any (there never could be any, for example, if the deceased was a woman). In that event the next best claim was that of the nearest ‘agnate' or agnates, if more than one were equally near. They were relatives who traced their relationship to the deceased through the male line only. There might also be no agnates: in the case of freedmen there never would be, and here the rule was that if the freedman left no children the patron succeeded to his estate. For ordinary free citizens, in remote pre-classical times, if there were no sui heredes or agnates, the property went to the gens. This term is sometimes translated as ‘clan'. Perhaps ‘extended family' is less redolent of the Scottish highlands. In the developed law the gens played no part, and the praetor introduced a more complicated hierarchy, in which children, including those who had been emancipated by the deceased and so were technically no longer within the family, came first; then came agnates; while cognatic relations - that is, those whose relationship with the deceased was traced either through the female or the male line — were also given a claim according to their proximity to the deceased. The appalling details of how this question was determined are contained in a text of Paul (D. 38.10.10). Last of all came the claim of spouses, so preserving until the bitter end the separation of their property.
What is interesting about this division is the equal treatment given to the deceased's children: no preference is given to males over females, and none to the eldest male over younger children. This system of partible inheritance must have had a strong tendency to fragment the deceased's estate, and one that may seem the more surprising in a society where possession of a particular amount of property (mostly land) was what determined an individual's membership of a given class. Various factors may have mitigated the tendency to fragmentation: for instance, making a will in which preference was shown to one or more of the children; and the likelihood that only a small number of his children would survive the deceased.
What then would be the motive for making a will? A recent study of surviving Roman wills indicates that the Roman testator was most likely to appoint his children — and particularly his sons — as heirs under his will (Champlin 1991: 107—20). But it was precisely to the children, in equal shares, that the law of intestacy directed an estate. A strong reason pointing an individual towards making a will would therefore be the desire to treat the children unequally, whether or not for the purpose of preferring the sons, and whether or not with the intention of avoiding undue fragmentation of the estate. This of course would not be the only possible reason: others would be a sort of superstition (as Maine suggested) or the desire to free slaves or pay off social obligations by leaving legacies. There does not seem to be much merit in speculating about which, if any, of these reasons would have weighed with most Roman testators most of the time.