CHAPTER XIX. RELEASE FROM SLAVERY. GENERALIA. OUTLINE OF LAW OF MANUMISSION DURING THE REPUBLIC.

It is not necessary to attempt the hopeless task of defining liberty. Justinian adopts from Florentinus^[1643] the definition: Liberty is the natural capacity {facultas) of doing what we like, except what, by force or law, we are prevented from doing.

The conception of manumission needs some examination. It is not in strictness transfer of dominium. A man has no dominium in himself or his members². Nor is it an alienation of liberty. The right received is not that of the master, and the rule that a man cannot give a better liberty than he has is intelligible without reference to such an idea. Nor is it a mere release from the owner’s dominium: that is derelictio, from which manumission differs in several ways. Dereliction does not make the man free, it merely makes him a res nullius³. Moreover manumission leaves many rights in the master, and there is no such thing as partial dereliction⁴. If it had contained a dereliction, then, since derelictio is purely informal, a manumission which failed for lack of form would have been a dereliction. But this was not the case. At civil law such a defective manumission produced no effect at all, and even under the Praetorian law and the lea; lunia it left large rights in the master, and entitled no third person to seize⁶. We have seen⁶ that the Roman conception of slavery was subjection to ownership, actual or potential: a slave was a human res.

There are some general rules which may be shortly stated here, though some of them will need more detailed treatment later.

An ingenuus is a freeborn person who has never been in lawful slavery². One who has been a slave is, on release from that position, a libertinus³. The law favours freedom on the one hand but guards the purity of ingenuitas on the other. An ingenuus does not cease to be one by being sold by his father⁴, or by manumission from apparent slavery®, or by being treated in any way as a slave, wrongly®. Even a declaration by the man himself under pressure that he is not an ingenuus does not deprive him of that position⁷. Adverse decision does not prevent repeated assertion of liberty, though a decision in favour of liberty may prevent its being again disputed®. In the same way a man cannot become a slave by lapse of time spent in apparent slavery’, though he may be free by prescription¹⁰. All these are evidences of the favour shewn to liberty: infinita est aestimatio libertatis¹¹. On the other hand, though a libertinus may be adopted, at any rate by his patron, he does not thereby become an ingenuus¹³, so far, at least, as rights in relation to third persons are concerned¹³.

To these general rules there are some exceptions, little more than apparent, which need only mention. They can be grouped under three heads.

(i) A libertinus may by special grace acquire the rights of an ingenuus. With this case we shall not deal¹⁴.

(ii) It is possible in certain cases, already discussed, for a person to be a libertus without having been a slave¹Â®.

(iii) It is possible for one who has been validly enslaved to become an ingenuus on again becoming free¹Â®.

There is a general tendency, doubtless accentuated in later law, to interpret rules and facts as far as possible in favour of liberty. It is a

'Post, p. 439. ^SG. 1.11. > 1.6. 6; In. 1.6. pr.; G. 1.11.

� P. 5. 1. 1; C. 7. 16. 1, ante, p. 420. « P. 5.1. 2; In. 1. 4. 1.

» P. 5. 1. 3. ’ h. t. 4. ³ C. 7. 16. 2, 4.

• C. 7. 22. 3. 10 C. 7. 21. 7; 7. 22. 1, 2; post, Ch. xxviii.

ii50. 17. 176. 1. і» 1. 5. 27 ; 1. 7. 46; Aul. Gell. 6. 19. 12.

¹⁸ 1. 7. 15. 3, 46; 2. 4.10. 2; 37.12. 1. 2; 38. 2. 49. An ingenuus adopted by a libertinus was still of course ingenuus, 1. 7. 35. From the texts cited in this and the last note, with others, it appears that adrogation of libertini, by any but the patronus, was one of the things which non debent fieri, sed facta valent.

¹⁴ See, e.g., Moyle, ad In. 1. 5. 3. ¹⁵ Ante, p. 412.

¹⁶ Thus captivus is a slave but reverts to ingenuitas by postliminium (ante, p. 304). See also the case of children sold, ante, p. 420, and see p. 410. general principle that in doubtful or ambiguous cases it is best to follow the more liberal view¹. Liberty being of immeasurable value and omnibus rebus favorabilior^{[1644] [1645]}, the principle is naturally laid down that in doubtful questions affecting liberty, secundum, libertatem respondendum erit^[1646]. Countless illustrations of this tendency will be found in the following chapters.

A slave may become free either as the result of manumission by his dominus, or without the latter’s consent. It is convenient to begin with manumission, and, as the topic is somewhat complicated, to deal first with the simplest case.

With the very early law we are not concerned, and indeed little but guesswork is possible in relation to it. The origin of manumissions is unknown. Dionysius of Halicarnassus credits the foundation of the law on the point to Servius Tullius^{[1647] [1648], but as he refers nearly everything else to that king no particular weight attaches to his testimony. The XII Tables shew that at their time it was an established institution[1649]. All manumission is regarded as an institution of the ius gentium*. It is a datio libertatis: liberatur (servos') potestate7. But it is more than that: it is, at any rate during the Republic, the making of a civis. Ulpian tells us that legitime manumissi, nullo iure impediente*, become cives. In the Digest he speaks of the patron’s rights as a return for having made cives of the slaves9. Thus citizenship is always the ordinary result of a typical manumission. From this characteristic of manumission it follows that all the modes of manumission are public, i.e. are in some way under public control. The State is interested in seeing that civitas is not bestowed on unworthy persons10.}

Of these modes of manumission there are three.

I. Censu. Although the Census survived into the Empire, it is so essentially a republican institution that it seems best to say here the little that is to be said about it.

It is not necessary to discuss the Census in general: we are concerned with it only as a mode of manumission. It is probable that this form of manumission is extremely old, but it hardly survives into the classical law with which we are really concerned. There was a census in A.D. 74, and there was at least the name of one in 243. But this form of manu­mission was really extinct.

The Census, taken every five years’, is in essence a list of cives made for fiscal purposes and for the regulation of military service. The form of manumission is the inscription of the name of the man on the list of citizens. It involves three steps. The slave presents himself and claims to be a citizen: censu profitebantur¹â€™. The assent of the owner is shewn: iussu or consensu domini*. The Censor inscribes the name on the list of cives. On each of these three requirements there is some­thing to be said⁹. The professio mentioned is the formal presentation of himself which each civis was bound to make to avoid the penalties falling on an incensus¹⁰. The iussum of the master does not seem to have been a formal part of the ceremony, though of course the Censor would not enrol the name without it^[1651]. It is an authorisation to the slave, not to the Censor, and its informal nature is expressed in Cicero’s description of it as consensus¹*. We are not expressly told that the master had to be present, but this was probably the case, especially in view of the fact that in 176 B.C. a temporary rule was laid down that in certain cases of manumission before the Censor or other magistrate the owner was required to take a certain oath¹³.

The process could take place only at Rome, for it was only at Rome that the true Roman Census was held^[1652]. The slave must be the property ex iure quiritium of the manumitter², and it is plain that no modality of any kind could be attached to manumission in this form. But it must be borne in mind that the freedman’s oath was not a condition, and, no doubt, by means of it many conditions could practically be imposed, breach of the undertaking being a punishable case of in­gratitude.

The procedure of the Census is a long business: the new lists cannot be prepared in a day. Apparently it was not usually till towards the end of their eighteen months of office that the lists were completed and the Censors proceeded to the formal act, lustrum condere, which brought the new lists into operation. It was not clear whether the slave was free from the moment of enrolment, or only when the new lists came into, operation. The doubt is referred to by Cicero, and again in the much later Fragmentum Dositheum³. It was not confined to this question, but extended to all acts of the Censor taking effect in the Census Roll, e.g. notae censoriae and the like. It would seem that the question must have been of great practical importance, and yet that it was never determined. It may be that these various acts, which were more than mere records of fact, were postponed till the last moment. Logic seems to require that, at least in our case, the later date should apply. It seems that it must have been so, so far as concerns public law. But it may well be that for private law the practice was otherwise. The entry does not purport to make him a civis : it is a fictitious renewal of an entry, and the Censor is recording the fact that the man is a civis, not making him one. Strictly indeed he is only recording the fact that the man has claimed to be a civis, and if such an entry is made in error it is null¹, and cannot operate by lapse of time, for it is not till much later that we find rules as to liberty by prescription’.

Some of the early Emperors were Censors, and Domitian was Censor for life. It does not seem that he proceeded to any census, or lustrum condidit in the old sense. There are no signs of manumission before him as Censor: the whole institution is at an end.

II. Vindicta. This is a “ fictitious ” application of the procedure in a causa liberalis. If a claim of liberty was made on behalf of a man alleged to be wrongully detained as a slave, the claim took the form of an action brought by an adsertor libertatis, claiming him as a free man, the form being, at this time, that of sacramentum¹. Used as a mode of manumission it was essentially a case of cessio in iure^{[1653] [1654] [1655] [1656] [1657]. The adsertor libertatis who, at least in later times, was often a lictor, claimed him before a magistrate as free, touching him with the wand which appears in sacr amentum, and which gave its name to this mode of manumission. The dominus made no defence and the magistrate declared the man free3. As it was an actus legitimus, no condition or suspension was possible: by addition of dies, or condition, actus legitimi in totum vitiantur*.}

From the form and nature of the process it is clear that the presence and assent of the magistrate were necessary. From the text of Livy already cited³ it may be assumed that the actual presence of the dominus was needed, though the oath there referred to was a temporary matter^[1658]. As it was in form a vindicatio, the slave must be on the spot.

We are not concerned with the position of a libertinus, but it may be as well to observe that it was not unusual to exact an oath, before manumission, that the man would render certain services. The oath was not in itself binding, but was regarded as putting the slave under the duty of swearing again, or promising, immediately after the manu­mission⁷. Breach of the undertaking would expose the freedman to the ordinary liabilities for ingratitude⁸. Even though the libertus were impubes, if he were old enough to take an oath, an actio utilis would lie to enforce the duties after puberty, and there were some duties which he could render even impubes; for instance, he could act as histrio or nom emulator¹¹.

III. Testamento. Gratuitous benefits are, naturally, given most readily at death. This mode of manumission was therefore by far the most important in the law. It will be necessary to deal with it at some length when we are discussing the classical law¹⁰: here it will suffice to describe its general nature, and to lay down a few main rules.

The origin of the institution is not certainly known¹¹. It is clearly as old as the XII Tables. Pomponius tells us that they gave a very wide power of, inter alia, manumission, by the uti legassit clause, a power afterwards restricted in divers ways¹³. So Ulpian, in an imperfect text, seems to base the law on the XII Tables^[1659]. But there is reason to think it older. The incomplete text of Ulpian ends with the word confirmat, and may mean that the lex confirmed an existing practice, or that it confirmed the testator’s declarations. But the most important point is that the lex contained detailed rules as to succession to freedmen, as to conditional gifts of liberty, as to the person to whom a man might make a payment on which his liberty was conditioned, and so forth⁸, a state of things which is most unlikely if the whole institution was new. We will not enter on the still more speculative question as to the relative antiquity of this and the other two modes, but will merely remark in passing that it is the only one of the three which is direct, i.e. is not based on a fiction. From this it has been inferred³ that it was the oldest: the contrary conclusion seems more reasonable.

The will of this age was, of course, the testamentum in comitiis calatis. The disappearance of public control is shewn by the extension of the rules to the mancipatory will of the later Republic: in that will the public aspect of the transaction has become a mere tradition.

Liberty could be given directly only to slaves of the testator, the model followed being plainly that of a legatum per vindicationem*. In conformity with the same principle Servius established the rule that the slave must have been the property of the testator both at the time of the will and at the time of death⁵. This is the rule in such legacies⁶, apart from the lex Papia which postponed dies cedens to the opening of the will. We are not indeed told that our rule has anything to do with dies cedens, but it seems probable that it had, and this may serve to explain an apparent conflict. As in the case of legacy, media tempera non nocent. Sale and reacquisition after the will was made left the gift valid⁷. Maecianus tells us that if liberty was given to a slave, and he was sold, but became again the property of the hereditas, before aditio, the gift was valid⁸. Gifts to slaves ceded only on aditio*, a fact which brings the present rule into connexion with the theory of dies cedens¹â€ś.

Of the form of the gift little need be said here: it will be considered later. There must be a clear expression of intent that the man should be free. Thus it might be liber sit, liber esto, liberum esse vubeo, and the like^[1660]. Implied gifts are not readily admitted even in later times¹⁸. It will be seen that the above forms follow closely that of legatum per vindicationem. The question may be asked: what would be the effect of such a gift as heres meus damnas esto Stichum liberum esse sinere I Such a gift must, it seems, have been null. A legacy in that form gave only a ius in personam and this could not have given more. But it could not have given a right of action to the man, and fideicommissary gifts were not yet invented. The same would no doubt be true of a gift heres meus damnas esto Stichum manumittere¹. The same reasoning applies. No doubt the institution might be made conditional on the manumission by the heres of his or a third person’s slave. But this is a wholly different matter: it is in no way enforceable by the slave. It is another question how far a person who disregarded such an injunction to free one of the slaves might incur the disapproval of the Censor.

The gift takes effect only upon the actual operation of a valid will, but if the heres has accepted, the gift remains effective even though he afterwards gets restitutio in integrum*. If, however, the will is r upturn, it was never valid at all, and, apart from collusion, a liberty which may have apparently taken effect by entry is void’. So also it fails in an ordinary case of testamentum destitutum, where, apart from collusion, no heir enters⁴. There are, however, exceptional cases in which the gift will be effective, though the will does not operate, for instance, where a heres, entitled both by the will and on intestacy, takes on intestacy— omissa causa testamenti. Other cases of this type will be considered later⁶.

Such gifts might, like legacies, be adeemed, and though they were not subject to the lea; Falcidia, passed just at the close of the Republic, their existence in a will gave rise to some difficult questions when that lea; operated. Both these topics will be more conveniently considered in connexion with the classical law⁶.

These three were the only forms of manumission which were recognised during the Republic. They all, whenever they were valid at all, made the freedman a civis, if we leave out of account for the present the slave freed by a Latin owner. But it is obvious that occasions must have arisen under which the intention to free a man, there and then, was expressed in less formal ways. Two such are in fact recorded. They are the declaration, inter amicos, that the man is free, and writing him a letter of enfranchisement’. Such declarations were void in early law. But, towards the close of the Republic, the

¹ Accarias, Precis, § 56, thinks such & gift valid. He does not advert to the question of the remedy.

⁸ C. 7. 2. 3. As to case of heres necessarius, post, pp. 505 sag.

⁸ h. t. 12. 2; D. 40. 5. 24.11. 4 C. 7. 2.12.

⁸ 40. 4. 23. pr. Post, Ch. xxvn. « Post, pp. 473 sqq.

⁷ See e.g. G. 1. 44. Amici are testes· See G. 2.25, and Bruns, Syro-Eoman Law-book, 195. See also Suetonius, de Rhet. 1. As to manumission in convivio, post, p. 446.

ch. xix] Persons in libertate tuitione Praetoris 445

Praetor interfered to protect persons who had been so declared free and gave them de facto enjoyment of liberty. Hence they were said in libertate servari auanlio Praetoris^[MDCLXI], in libertate tuitione Praetoris esse², in libertate domini voluntate morari³ (or esse). The texts are explicit that, notwithstanding the declaration, they were still slaves (manebant servi, non esse liberos, ex iure quiritium servi), and in accordance with this we learn that their peculia and all that they acquired belonged to their (former) master. It is clear that it was not revocable, and there is no reason to doubt that it was binding on successors in title of all kinds. It is also fairly clear on our few authorities that the status was not heritable: such persons were slaves and the child of a woman in such a position would be an ordinary slave. The main, indeed, as far as can be seen, the only, effect was to free them from any duty of working, so that if the owner tried, by force, to make them work for him, the Praetor intervened to prevent it⁴. They were evidently not derelicti: the informal declaration that they were to be free was very far from an abandonment of all rights.

Doubt may be thrown on some of these conclusions by other language of these texts. Thus Gaius speaks of them as ex iure quiritium servi, and goes on to speak of the master as patronus³. And the Frag. Dositheum speaks of him as manumissor and patronus, and says that the person so dealt with, omnia quasi servus acquirebat manumissori“. But it must be remembered that these texts were written a century and a half after the lex lunia had turned these processes into real manumissions, and this part of the language is coloured by that fact. More weight must be given to the words which express what was certainly not the law of the age in which they were written.

The Fragment gives some further details. The protection of the Praetor did not proceed as a matter of course, but only if the Praetor thought the slave a fit person to have this de facto liberty⁷. Moreover voluntas domini spectator, and thus his consent must be real. The Praetor would not intervene if the master’s declaration that he wished the slave to be free was made under pressure⁸. He would not intervene if the owner was a woman who had not her tutor’s auctoritas, or, pre­sumably if it was an impubes in the same case. The text remarks that the Praetor would not intervene if the owner were under twenty. As it stands this may be a result of the lex Aelia Sentia, but it is equally probable that that enactment only followed in this respect what had

446 Persons in libertate tuitione Praetoris [pt. n

been the practice of the Praetor. The age of the slave was immaterial¹. The master must be one who held the slave in bonis, but he need not be the quiritary owner’. If the slave were common, the declaration by one of the owners that the slave was to be free produced no effect at all’. These texts are mainly concerned with latinity under the lex lunia, but are made relevant to our case by some words which indicate that that latinity was granted on informal manumission under such circumstances as would have led the Praetor to protect de facto liberty⁴. This entitles us to say that this partial relief might be given when a master, incapacitated from formal acts by physical defect, yet wished to free his slave. Thus Paul tells us that a mutus surdus, though he could not manumit vindicta, could do so inter amicos^e. Wlassak, in the course of an exhaustive article⁸, in which he shews that Praetorian rights were not exempt from rules of form, establishes certain con­clusions in relation to these manumissions. He shews that in such manumission there was needed express declaration of intention to free, not merely to allow to be in libertate. He shews further that the evidence on which it has been generally held⁷ that there were many of these modes of manumission shews only that there were many ways of obtaining latinity, and that of the well-known three, that in convivio is not mentioned till the later Empire⁸. He objects to the expression “ informal,” since in fact each has its form. It seems, however, justifi­able to call them informal, since the presence of witnesses is rather a substantial guarantee than a formal one’. He discusses¹⁰ certain texts which suggest that it was enough that the master had expressed a willingness for the slaves to be in libertate, i.e. that animus manumittendi was not needed. The non-juristic texts¹¹ he holds to be mere inaccuracies of expression. This is probably correct, but such allusions shew that manumissions which required only declaration before an unspecified number of witnesses must have taken place under such varying conditions as to have given an impression of formlessness. Two juristic texts¹² which raise the same suggestion are referred by Wlassak to another matter¹⁸.

¹ Ft. Dob. 13—15. 2 Ft. Dob. 9.

⁸ Fr. Dos. 10. * Fr. Dos. 5—7; Gt. 3. 56.

« P. 4. 12. 2. « Z. S. S. 26. 367 sqq.

1 In. 1. 5.1; 1. 5.3; 3.7.4; C. 7. 6.1.

⁶ He cites (op. cit. p. 404) as mentioning it Pseudo*Dion., ad Paul. Sam.; G. Ep. 1.1.2; Lex Rom. Burg. c. 44; Theoph. ad Inst. 1. 5. 4. It is not in the Digest or in the comprehensive C. 7 · 6.

⁰ It may be noted that in the record of manumission inter amicoa which we possess (Girard, Textes (3), Appendice) the witnesses are not named and do not sign as in the recorded mancipations (Girard, op. cit. 785, 806 sqq.).

¹⁰ p. 391. ¹¹ Pseudo-Quintii. Declam. 340, 842. ¹³ 40. 12.24. 3, 28.

¹⁸ Post, Ch. xxvm. That such transactions as those recorded in Suetonius and the De* clamationes, locc. citt. should have been thought valid by anyone shews how little any notion of form enters into the matter. In dealing with Latini luniani we shall see that either the conception, inter amicoa, was very widely construed or before Justinian a number of informal modes had come to be recognised.

ch. xix] Gifts of liberty in a Praetorian Will 447

The texts do not touch the question whether a manumission in a Praetorian will could be enforced in this way. It seems very unlikely for many reasons. It seems almost certain that if it had been so we should have heard of it, for we hear in various places a good deal about these wills, and about the ways in which Junian latinity could be obtained, and it is nowhere mentioned in either of these connexions. All informal manumission seems to be contemplated as inter vivos. Moreover the expression Praetorian will is a little misleading. It is far less than a will. It operates under certain edicts, to the effect that if the document is in a certain form claimants under it will be given possession of the bona. All that can be got under it is bonorum possessio, and this de facto liberty cannot be brought under that conception. The enactment of Justinian abolishing latinity deals with two cases closely akin to this. Liberty given by codicils is mentioned allusively¹, and the case is discussed of a direction by the testator that certain slaves should share in his funeral, wearing the pileus which was the sign of liberation. This last he adds to his list of informal manumissions, remarking that as such a gift it had been of no effect before his time². It seems that so verbose a draftsman must have adverted to the case we are concerned with, if it had existed.

The same question calls for the same answer, in the case of gifts made in indirect forms before the introduction of fideicommissa*.

Some peculiar cases of manumission, perhaps exceptional in form, are mentioned by other than legal writers. Festus, in two passages⁴, refers to manumissio sacrorum causa, to which we have no other reference. It is manumission by a solemn declaration that the man is to be free, the master holding him by a limb and undertaking to pay a sum of money if the man so freed departs afterwards from the sacra. Then he turns him round and releases him and he is free. This may be a case of manumissio vindicta. If this is so, either it is very incom­pletely stated (which, in view of the author’s purpose, is likely enough), or the breakdown in formality of manumission vindicta is much earlier than is commonly supposed. Mommsen* treats it as no manumission at all, but only a sale to the temple with an agreed penalty for taking him away. It is connected with a similar Greek practice⁶. In Greece it was common to sell slaves to the service of a deity, in which case they became sacred and free, at least from the secular law. Gradually the process came to be applied, as a fiction, for what was plainly manu­

i C. 7. 6. lc, 2.

ⁱ C. 7. 6. 5; post, Ch. xxiv. His language may mean that this had given latinity.

⁸ Praetorian protection in case of other defects, post, Ch. xxm.

⁴ Festus, vv. manumitti, purum.

s Mommsen, Bom. Staatsr. 3.1.421; D. P. B. 6. 2. 2. ⁶ SeeWlassak,Z.S. S. 28,pp.22sgg.

mission¹. But the process recorded by Festus never seems to have got further than the devotion of the man to the service of the deity. Festus is however very clear that it was a manumission, though his authority cannot be ranked as very high in view of the antiquity of the institution.

Aulus Gellius observes’ that many jurists had laid it down (though he treats the matter as of purely antiquarian interest) that a master could give his slave in adoption. This too seems to be ordinarily regarded⁸ as a case of manumission mndicta, with some special for­malities, and it seems to be sometimes treated as the same rule as that which Justinian attributes to Cato⁴, that is, that a master could adopt his own slave and so free him. But the two cases are not the same. The language of Aulus Gellius shews that he is contemplating an adoption by vindicta or an analogous process, and by a person who is not the dominus : serous a domino per praetorem in adoptionem dari potest. On the other hand the case in the Institutes is plainly one of adoption by the dominus himself. It is hard to see how this could be done directly by mndicta, and we have no right to suppose a transfer to a third person, followed by adoption. On the whole it seems more likely that it was a case of adrogation, in effect a shortening, by leave of the Comitia, of the form of manumission followed by adrogation of the freedman by his patron. This last we know to have been a familiar case⁵. Wlassak⁸ objects to the view that it is an adrogation on the ground that a slave could not have appeared in comitiis. If a woman could not be adrogated a slave could not. He supposes a fiduciary sale followed by adoption. There is no logical answer to this objection, but it may be doubted whether so severe a logic can be safely applied in such a case⁷. In any case the institution is not important to us for it leaves no trace in the classical law⁸.

¹ See the literature cited by Girard, Manuel, 116. See also Dareste, Recueil des Inscriptions Grecques, Sèrie 2, pp. 234 sqq. For various opinions as to the nature of the Roman institution, see Vangerow, Latini luniam, 59.

² Aul. Geli. Noct. Att. 5.19. 13—14.

⁸ Moyle, ad Inst. 1.11.12 ; Vangerow, op. cit. 62.

In. 1. 11. 12. « 1. 5. 27; 1. 7. 46e«c. « Z. S. S. 26, 387.

⁷ The effect of the transaction i§ txrmake him capax ; it is not so in the case of a woman.

⁸ The declaration apud acta that the slave is your son which seems to have given latinity before Justinian {post, Ch. nni,), is a similar institution, but it is an innovation rather than a survival.