Introduction

The purpose of this paper is to discuss the role of witnesses in the establishment of a will as it may be found in sources of written law and documents of legal practice, preserved mostly in Egyptian papyri. Consequently, the evidence presented hee deals largely with wills preserved in writing².

Civil Law

A discussion of the role of witnesses in Roman wills must start with the mancipatory will, as it was the earliest type of Roman testament witnessed by private witnesses³. The function of witnesses in mancipatory wills was rather secondary; they witnessed a ficti-tious mancipatio which was a required element of the will⁴; Gaius Inst. 2, 104: eaque res ita agitur: qui facit testamentum, adhibitis, sicut in ceteris mancipationibus, V testibus ciuibus Romanis puberibus et libripende, postquam tabulas testamenti scripserit, mancipat alicui dicis gratia familiam suam.

Witnesses are traceable in documentary sources from the moment when the first written Roman wills appear in papyri in the first century AD⁵. In wills on papyri from the period prior to the issue of the Constitutio Antoniniana, there are two elements related to the func tion of witnesses in written wills – witnesses’ clauses and mancipatory clauses.

The witnesses’ clauses are attested in both the Latin originals, and in the Latin and Greek copies of those originals – composed as either an αὐθεντικόν required by law at the moment of the official opening of the will, or as an ἀντίγραφον written for other reasons at any time after the opening⁶. They were based on a simple repetitive model which contained solely information about who the witnesses were, as in a will of the Roman knight Antonius Silvanus⁷. FIRA III 47 (CPL 221; Alexandria, AD 142): Nemonius – dupli carius turmae Mari signavi. There are six more, probably holographic signatures. As the entire text of the will is preserved, one can compare the signatures with the mancipatory clause: familiam pecuniamque testamenti faciendi causa emit Nemonius duplicarius turmae Mari, libripende M. Iulio Tiberino sesquiplicario turmae Valeri, antestatus est Turbinium signiferum turmae Proculi.

Pasquale Voci observed that these signatures belong to the persons involved in mancipatio, that is to the five witnesses (starting from the honorary one called antetestatus), the familiae emptor and the libripens⁸. In other documents, mostly copies of original wills, the same pattern is visible⁹. After comparing this practice with the Gaian text we may conclude that the original function of witnesses was their physical presence at the very act of mancipatio and witnessing thereto, but not to the will as such, and that the witnesses’ role was to witness the act, but not the content of the will¹⁰.

This function is also made evident by the catalogue of persons excluded from the function of witnesses to mancipatory wills (Gaius Inst. 2, 105-107, Tit. Ulp. 20, 1). Those who were not capable of acting as witnesses at the act of making a will were men connected to either the testator or the familiae emptor through patria potestas. However, an heir himself could easily play the role of witness to a testament which appointed him an heir¹¹.

Nevertheless, the papyri do not allow us to conclude that this was the only role of Roman citizens who took part in the composition of wills. The examination of the aforementioned documents, composed during will-opening ceremonies, indicates that the function of witnesses was also related to the very act of sealing (but not subscribing)¹². In almost all Roman wills in Egypt written before the Constitutio Antoniniana testamentary witnesses are called cφραυιcταί or signatores.

Moreover, in the documents of the opening, their role is to recognise seals. A record would first make known that a will had been opened and read out in front of sealers who had recognised their seals, and would then be followed by a list of those who identified their seals.

– P. Diog. 10 (= ChLA XLVII 1403 = P. Coll. Youtie I 64; Ptolemais Euergetis, AD 211): apert (um) et rec(itatum) (…) praes (ente) pl(urima) part(e) signat (orum) f(igentium) sig(na), L(ucius) Valerius Lucretianus adg(noui). M(arcus) L[……]nus adg(noui). Fl(avius) Diogenes adg(noui). Arrius Nigerus adg(noui) M(arcus) Aurel(ius) Anubion. L(ucius) A[……] Cottarus.

– BGU I 326 (= FIRA III 50 = M. Chr. 316 = Sel. Pap. I 85 = Jur. Pap. 25; Arsinoite nome, AD 194): ἠνύγη (l. ἠνοίγη) καὶ ἀνεγνώσθη (…). oἱ λοιποὶ cφρα(γιcταὶ) Гαῖoc Λoγγῖνoc Ἀκύλαc, ἐπέγνοι (l. ἐπέγνω), Ἰούλιοc Boλύccιoc, Μᾶρκοc Ἀντίcτιoc Πετρωνιανὸ<c Ἰούλιοc> <Γεμέλλοc οὐετρ[α]νόc>¹³.

A document published as BGU I 361 (= FIRA III 57 = M. Chr. 92; Philadelphia, AD 184) reports a family dispute regarding an opening of a will which took place in the presence of Apollonios, the strategos of the district. As the son of a testator, supported by his half-brother, tried to stop the opening of his father’s will, he questioned the authenticity of the seals, as well as their recognition and, consequently, the authenticity of the document itself¹⁴.

– BGU I 361 (= FIRA III 57 = M. Chr. 92; Ptolemais Euergetis, AD 184): καὶ περὶ τῆc διαθήκηc δὲ ἀποκρείνομαι (l. ἀποκρίνομαι), ὅτι ἐν πάcαιc τ[αῖc δια]θηκαιc ἑπτά εἰcιν cφραγιcταί. εἰ οὖν ἐκῖ (l. ἐκεῖ) ἐν ταύτῃ ἑπτὰ ἐc̣[φράγι]cαν, ἡκέcτωcαν καὶ τὰc cφραγῖδαc αὑτῶν πρότερον ἐπιγν[ώτω]cαν. εἰ δὲ ἀπὸ τῶν ἑπτὰ τέccαρεc ἐνθάδε εἰcὶ καὶ νομικὸ[c (…) βή]μ̣α̣τοc ἔρχεται ἀξιῶν λυθῆναι τὴν διαθήκην, οὐκ ἐν ὀλίγῃ ὑπ̣[οψίᾳ] γε[ί]νομαι¹⁵.

These passages illustrate how the fonction of witnesses evolved from the secondary role of testifying to the act of mancipatio to the primary role of guarantors of the authenticity of documents, composed in secret in order to preserve the content of a testator’s disposition (expressed on tablets accompanying the mancipatory will). As no other proof of the authenticity was used, this function seems to be particularly important.

On the other hand, there is no way of ascertaining whether the act of mancipatio nummo uno was ever performed in Egypt. It seems improbable, as there is not much evi dence of the application of any kind of mancipatio in papyri, except for mancipatory clauses in wills¹⁶. Also outside of Egypt the practice of both a symbolic and a real mancipatio is unsatisfactorily attested¹⁷. As the mancipatio was a very Roman act of archaic origin, it should have been incomprehensible to provincials of a much later period, who were rooted in a different legal culture. It is therefore hardly plausible that this act was effectively per formed during the composition of wills. This makes more credible the supposition that, at least in Egypt, the sole role performed by the witnesses was effectively to guarantee the authenticity of documents composed, in order to preserve the content of wills.

This conclusion is further supported by the fact that no local will, composed orally, existed in Roman Egypt¹⁸. Moreover, a valid will had to be sealed by witnesses and no other requirements were applied. In a papyrus recording a family dispute there is a state ment that a valid will was one witnessed by a proper number of witnesses¹⁹. Documents concerning opening of wills prove the same: a legally valid will was one sealed by wit nesses²⁰. Therefore, one can conclude that Roman wills were performed in the local way, even though they were framed on the Roman formulary²¹.

Praetorian Law

In the doctrinal sources of Roman law, one can find information about how praetorian law maintained the validity of wills which were not composed according to one or more requirements concerning the external form of a will²². A brief summary of the rules can be read in the classical source Tituli ex corpore Ulpiani.

Reg. Ulp. 33: si septem signis testium signatum sit testamentum, licet iure civili ruptum vel irritum sit, praetor scriptis heredibus iuxta tabulas bonorum possessionem dat, si testator et civis Romanus et suae potestatis, cum moreretur, fuit; quam bonorum possessionem cum re, id est cum effectu, habent, si nemo alius iure heres sit.

The described practice might be applied to the tablets composed according to civil law. The seals of five witnesses to mancipatio, plus those of familiae emptor and libripens seem to have been the seven seals required by praetorian law in order to make valid a defective will²³. This is the case of Antonius Silvanus’ will, since among the sealers there are persons who were mentioned, in a mancipatory clause, as those who played the role of the ficti tious acquirer of the inheritance and the scale-holder.

The statement is also supported by documents of opening²⁴. Thus the witnesses known from Egyptian wills played a double role of witness to both a mancipatio and to the document itself²⁵. Moreover, praetorian law conferred authority only upon wills accompanied by tablets and, as the validity of a document depended on the presence of seals affixed to the document, witnesses to wills became an element guaranteeing the entire existence of a will, at least at the level of ius honorarium.

The conclusion to be drawn is that there is a dualism in the role of witnesses in the process of making wills in the classical period. It is a dualism of acting as witness to the act and, at the same time, to the document accompanying this act²⁶. It is a dualism of – to use modern terminology – the probationary and the constitutive character of witnessing. Though witnesses gave an expectation of the validity of an act under praetorian law, it was not an absolute expectation. In the above quoted passage of the Tituli we read: quam bonorum possessionem cum re, id est cum effectu, habent, si nemo alius iure heres sit.

However, together with a constitution issued by Antoninus Pius, which made such bonorum possessio cum re, the need for the dualism disappeared.

Gaius Inst. 2, 120: sed videamus, an etiam si frater aut patruus extent, potiores scriptis heredibus habeantur; rescripto enim imperatoris Antonini significatur eos, qui secundum tabulas testamenti non iure factas bonorum possessionem petierint, posse aduersus eos, qui ab intestato uindicant hereditatem, defendere se per exceptionem doli mali.

By granting an exception against the hereditatis petitio to heirs instituted in tabulae, their position was safeguarded; otherwise an intestate heir would have been able to take prece dence over an heir instituted in a civil will valid solely iure praetorio. The Emperor thus abolished the need for the performance of the act per aes et libram. From the moment this constitution was issued the position of heirs only written down in tablets was as solid as the position of heirs instituted through the proper civil act.

Late Antiquity

What consequences did this constitution have for the function of testamentary witnesses? First of all, it meant the disappearance of a mancipatory will. Since bonorum possessio secundum tabulas became cum re, at least at the level of legal practice mancipatio nummo uno was no longer necessary. Nonetheless, the wills including mancipatory clauses remai ned standard in the papyri until the second quarter of the third century. The factual disap pearance of a mancipatory will takes place when Alexander Severus issued the constitution on the language of wills²⁷. The decay of mancipatory wills meant also the disappearance of mancipatio witnesses in testamentary practice²⁸. It is worth stressing that a perfectly effective will had to be sealed by seven persons and composed according to the require ments concerning their inner form – the ordo scripturae.

Still, together with both the progressive decay of testamentum per aes et libram and the disappearance of the distinction between civil and praetorian testamentary succession, the function of witnesses started to become more and more significant and uniform²⁹. Justi nian’s Institutions 1, 2, 10, 10 include a list of persons excluded from being witness to a will³⁰. The persons who were not allowed to play the role of witnesses were heirs to a will in which they were instituted and men connected with them through parental power, whereas before, this role could not be played by people related to the testator and the familiae emptor. This is to be interpreted as a change in the function of witnesses. They now acted as witnesses to the will itself, not to the act on which the will was framed.

The function of witnesses became more significant. The more constitutions abolished the inner formalities, the more they emphasised the importance of witnesses in wills. This tendency is visible in the dogmatic sources as well as in the papyri. Since the earlier requi rements, as solemnity of heredis institutio and exhereditatio or the presence of mancipatio and nuncupatio disappeared either through formal abrogation, as in the case of heredis institutio (C. 6, 23, 15, possibly AD 320), or through disuse, as in the case of mancipatio and nuncupatio, witnesses became the sole element required for the validity of wills³¹. The constitutive function of witnesses is literally and repetitively expressed in the constitutions starting from the beginning of the fourth century³².

This trend is underlined in the documents of practice as well:

– SPP I, p. 6-7 (= FIRA III 52; Antinoopolis, V AD): κυρίαν δὲ οὖcαν καὶ βεβαίαν αὐτὴν ἐθέμην ἐφ’ ὑπογραφῆc ἐμῆc καὶ τῶν αὐτῶν ἑπτὰ νομίνων μαρτύρων ὁμο`ῦ´ cυνηγμένων καὶ cφραγιζόντων κατὰ τὴν θείαν διάταξιν.

– P.Oxy. XVI 1901 (Oxyrhynchus, V/VI AD): ταύτηc μου τῆc διαθήκηc κ̣α̣λ̣ῶ̣c̣ [ἐχούcηc, ποιηθείcηc] ὑπάτοιc τοῖc προκειμένοιc, καὶ ἠξίωcα τοὺc ἐξῆc ἀξιοπίc[τουc μάρτυραc] ἐνθεῖναι τὴν αὐτον μαρτυρίαν καὶ cφραγῖδα μετὰ τὴν ἐμήν̣ [ὑπογραφὴν πρὸc] ἀcφάλειαν καὶ βεβαίωcιν τῶν ἐμοὶ παραcτάντων.

– P.Cair. Masp. III 67324 (Aphrodito, ca. AD 525): ἥνπερ δ̣ι̣[αθ]ήκην ὑπεγόρευcα (l. ὑπηγ-) καὶ παρεcκεύ[α]cα α(ὐ)τὴν γραφῆναι κ(αὶ) ὑπογραφ[ῆναι δ]ι̣[ὰ τ]ῶν ἑξῆc ὑπογραφόντων μαρτύρων ἑπτὰ τὸν ἀριθμόν, οὕcπερ μάρτυραc προcεκαλεcάμην, κ(αὶ) παρανέγνων αὐτοῖc τὴν δύ̣ν̣αμιν τῆc δ̣ι̣α̣θ̣ή̣κηc κ̣α̣ὶ̣ π̣έ̣π̣ε̣ι̣κ̣[α] κ̣α̣τ̣ὰ̣ τ̣ο̣ὺ̣c̣ ν̣[όμουc] ὑπογράψαι³³.

– P.Cair. Masp. III 67312 (Aphrodito, AD 567): ταύτην τὴν διαθήκην τίθημι, ἐπὶ παρουcίᾳ τῶν κατὰ παράκληcιν ἐμὴν προcκληθέντων καὶ ἐπὶ τὸ αὐτὸ cυνηγμένων ῥογάτων νομίμων ἑπτὰ μαρτύρω(ν), πολιτῶν ὄντων Ῥωμαίων, ἐφήβων, καὶ ὑπολήμψεωc ἀντιποιο(υ)μένων, τῶ`ν´ κ(αὶ) ἑξῆc ὑπογραφόντων ταύτῃ μο(υ) τῇ δ[ιαθ]ήκῃ καὶ cφραγιζόντων αὐτὴν ἐν μιᾷ cυνόδῳ καὶ ῥοπῇ καὶ ὥρᾳ, μηδεμιᾶc ἑτέραcπράξεωc μεcολαβούcηc, κατὰ τὴν τῶν νόμων δύναμιν.

– Will of Gregorius in P.Ital. I 4-5 (= ChLA XVII 653; Ravenna, AD 552): testium quoque rogatorum numero competenti ad hanc tantum causa, scientium quur venirent, uno tempore eundeque in loco sub meorum visione conspectuum suscribtionibus signaculisque firmavi, quem claudi signarique praecipi, et valere iussi.

The important question in this context is the scope of the witnesses’ act, namely, whether it was either the act of composition of the document, or the contents of the will. In the documents, both the presence of witnesses and physical signs of this presence – seals and signatures – are stressed. This is what one observes in the already quoted examples; howe ver, in the doctrinal sources the differentiation is clear.

C. 6, 23, 21 (= Nov. Th. 16, 1): hac consultissima lege sancimus licere per scripturam conficientibus testamentum, si nullum scire volunt quae in eo scripta sunt, signatam vel ligatam vel tantum clausulam involutamque proferre scripturam vel ipsius testatoris vel cuiuslibet alterius manu conscriptam, eamque rogatis testibus septem numero civibus Romanis puberibus omnibus simul offerre signandam et subscribendam, dum tamen testibus praesentibus testator suum esse testamentum dixerit quod offertur eique ipse coram testibus sua manu in reliqua parte testamenti subscripserit: quo facto et testibus uno eodemque die ac tempore subscribentibus et consignantibus valere testamentum nec ideo infirmari, quod testes nesciant quae in eo scripta sunt testamento.

The moment a will was completed was the moment it was sealed and subscribed and, therefore, became a valid will. Thus, witnesses played a role which, according to modern terminology, might be named the constitutive function, as their seals and signatures together with testators’ subscription constituted the will³⁴. As the authors of the Institutions 1, 2, 10, 3 rightly observed, this concept had been adopted from the praetorian law by legal practice (approved by imperial constitutions).

The above statement is also supported by the documents recording ceremonies of ope ning. The first one is a dialogue between an official responsible for the opening of wills, logistes, and a party claiming the opening.

P.Oxy. LIV 3758 (Oxyrhynchus, AD 325): ὁ λογι(cτὴc) εἶ(πεν) · πόcοι εἰc[ὶ] cφραγιcταί; Διογένηc εἶ(πεν) ἑ̣π̣τ̣ά̣, τέccαρεc δὲ πάρειcι. ὁ λογι(cτὴc) εἶ(πεν) ὑπογ̣[ρ]α̣[ψάτωcαν] [οἱ] τ̣έ̣c̣[cα]ρ̣[ε]c̣ ἐ̣[πεγνω]κ̣έ̣ν̣αι ἑ̣[α]υ̣τ̣ῶ̣ν̣ τ̣ὰ̣c̣ c̣φραγῖδα̣c. καὶ τῶν̣ π̣α̣ρ̣ό̣ν̣τ̣[ων] cφραγιc̣τ̣ῶ̣ν ̣ὑποcημιωc̣αμένων (l. -cημειω-) ἐπεγνωκέναι ἑ̣α̣υ̣τῶν τὰc cφραγῖδαc, ὁ λογι(cτὴc) εἶ(πεν) · λυθήτω τὸ γραμμάτιον κ̣[αὶ] ἀναγνωcθήτω.

The second one also records the procedure of opening. The text of the opening of five wills from Ravenna was based on one scheme³⁵. After a depositary’s petition for the opening of a will, the offical said: suscipiatur carta testamenti, quae offertur, et testibus praesentibus ostendatur. Then follows: cumque carta testamenti suscepta fuisset et testibus praesentibus ostensa. (A witness) responded: constat me in hoc testamento interfuisse, in quo agnosco signaculum anuli mei, superscribtionem meam, et infra subscribi. Then, other witnesses present at the ceremony repeated the same formula. Thereafter, the official said: quid de alio teste, cuius signaculum superscribtionem inprexam vidimus? In response, the witnesses present said: constat (name of a witness/witnesses) una nobiscum in hoc interfuisse testamento, in quo agnoscimus anuli eius signacula, superscribtionem, sed nunc absens est. The narration: quoniam de agnitis signaculis vel supersubscriptionibus testium responsio patefecit, nunc carta testamenti resignetur, linum incidatur, aperiatur et per ordinem recitetur.

This text clearly illustrates that a valid will was a document with authentic seals, which is indicated by the fact that the recognition of the seals resulted in the opening of a will, in other words, with the recognition of the authenticity. Thus, witnesses in late Roman law played two roles: they made a will valid (through a seal and a signature) and guaranteed the correctness of the very act of completion which was secured by all of the conditions imposed on it in the imperial constitutions, like the obligation of unitas actus³⁶. The need for securing the act of completion of a will is also visible in the deeds of practice, which constantly mention the fact that the presence of some witnesses fulfilled the legal require ments, or at least that some witnesses were present during the act of completion.

– P.Col. VII 188 (= SB XX 14379; Karanis, AD 320): Οὐ]αλέριοc Ἰcίδωρο[c ἑκατόνταρχοc cυγκολλήγαc cυνέcτηκα καὶ μαρτυρῶ.

– A will of Gregory of Nazianzus: Ἀμφιλόχιοc ἐπίcκοποc τῆc καθολικῆc ἐκκληcίαc τοῖc ἐν Ἰκονίῳ παρὼν τῇ διαθήκῃ τοῦ αἰδεcιμωτάτου Γρηγορίου καί παρακληθεὶc παρ’ αὐτοῦ ἐπέγραψα χειρὶ ἐμῇ.

– P.Oxy. XVI 1901: Φο̣ι̣β̣[άμμων υἱὸc] Θεοτίμου μαρτυρῶ τῇδε τῇ διαθήκῃ ἀκούcαc παρὰ Ποῦcι κούρc[οροc τοῦ διαθεμένου] ὡc πρόκ(ειται).

– P.Cair. Masp. III 67324: ☧ Ψενθαῆcιc Ὀλλοτοc πρεcβ(ύτεροc) μαρτυρῶ̣ τῇ διαθήκῃ ἀκούcαc παρὰ τοῦ θεμένου³⁷.

– P.Ital. I 6 (= ChLA XXI 714; Ravenna, AD 575): huic testamentum rogatus a Mannae viro devoto, filio quondam Naderit, ipso praesente et subscribente, atque ei testamentum relictum, per quo constituit heredem sanctam catholicam ecclesiam Ravennate, testis suscribi.

The quoted subscriptions have a double meaning. First of all, they attest the formal cor rectness of the act of completion. Secondly, they inform us – especially in the later documents – that witnesses heard the will when it was dictated: μαρτυρῶ̣ τῇ διαθήκῃ ἀκούcαc παρὰ τοῦ θεμένου. Thus, they might confirm that the written content was in accordance with what a testator dictated. They could probably testify if any legal dispute arose or if the document was lost. Despite the constitution (C. 6, 23, 21 = Nov. Th. 16, 1, AD 439) reminding us that there was no legal necessity to disclose the content of a testament in front of the witnesses, the practice of making the content of wills known to witnesses was wide-spread in both East and West³⁸.

To conclude, the role of witnesses as recorded in papyri and sources of written law underwent a visible process of evolution. Initially, their function was to witness the formal act of mancipatio; however, as the last institution was never perfectly performed (at least in Egypt) the character of witnessing evolved into the constitutive one. The witnesses de facto guaranteed the correctness of the document, since the document replaced the act per aes et libram. The final stage of the evolution was the acceptance of this function by imperial law.