On-line version ISSN 2411-7870
Fundamina (Pretoria) vol.20 n.2 Pretoria 2014
Dr. Dr. h.c., Professor Emeritus, University of Exeter
Justinian's Digest was compiled from opera jam paene confusa et dissoluta (Tanta 1). That cannot mean an organised library. It was a pile of manuscripts discarded after being used to compile a composite, inscription-less work to implement the Law of Citations. That work was the multitudo auctorum of Deo Auctore 6, and the multitudo antiqua of Tanta 17. The compilers started reading entire works in three Masses, and ended with fragmented or incomplete works or mere scraps of manuscript in the Appendix.
The materials used for the compilation of Justinian's Digest were jam paene confusa et dissoluta: they were "mixed up together and falling to bits".1
There is no parallel passage in the Greek version, Δέδωκεν.2
They are, however, very important, for two reasons.
1. The works read by the compilers were confusa et dissoluta. That is a clue as to where they were found or where they came from.
I think it was in April 1994 that I attended an advanced seminar5 run by Aldo Cenderelli for his students in Milan. He started with the words confusa et dissoluta and asked us: Does that sound like Tribonian's private library?6 Of course, the answer was no. He went on, like an advocate enjoying his cross-examination: Does it sound like the Constantinople law school library?7 No. Does it sound like Justinian's imperial palace library? No. Does it sound like a collection of works from all over the eastern empire? No. What does it sound like? A heap of discarded manuscripts, which have been heavily used but are no longer wanted.
Cenderelli's original view was that in the 430s the compilers of Theodosius' planned but never completed code of jurisprudence read all the relevant works, excerpted them and built up a schedatura sistemática dei vari libri,8 a systematic card index; and it was that schedatura that Tribonian and perhaps the other compilers found in 528. But confusa et dissoluta does not sound like a schedatura sistematica, and in his later article all reference to schede and schedatura has disappeared.9 What Justinian's compilers found were not schede compiled from the original works, but the original works themselves, ipsa vetustatis studiosissima opera (Tanta 1).
There is another possibility. After the Law of Citations, it was no longer sufficient for an advocate to cite one jurist, Ulpian for example, in support of his argument. The judge would need to know what the other four said.10 In fact, what was really needed was a composite work setting out the majority view on as many legal points as possible. Individual inscriptions could be omitted. There was no point in having five or more inscriptions for a single rule. And the original works could then be discarded and forgotten. That explains Justinian's statements in Tanta 17: quorum et nomina antiquiores homines non dicimus nesciebant, sed nec umquam audiebant, the older men not only did not know their names but had never even heard them; and in quibus multi fuerant et ipsis eruditissimis incogniti, many of them were unknown even to the very learned.
The composite work was obviously displaced by Justinian's Digest and has not survived. Is there any trace of it? I think there is, in the multitudo auctorum in Deo Auctore 6, and the multitudo antiqua in Tanta 17. Let us take the second text first.
Tanta 17: Mirabile autem aliquid ex his libris emersit, quod multitudo antiqua praesente brevitatepaucior invenitur... ut egena quidem antiqua multitudo inveniatur, opulentissima autem brevitas nostra efficiatur. ("Now there is something remarkable that appears from these books: it turns out that the multitudo antiqua is smaller than the present abridgement (the Digest) ... the multitudo antiqua is worthless but our abridgement is enormously valuable.")
Now what was the multitudo antiqua?11 At first sight one would expect it to refer to the whole of classical law,12 but in that case the text says that the fifty books of the Digest are bigger than the two thousand books (Justinian's figure) of classical law from which it is derived: which is patent nonsense. And it says that classical law is worthless: which is an extraordinary statement from someone whose compilers have just spent three years working on it. That cannot be right.13 If, however, the multitudo antiqua was a composite work of inscription-less fragments, both statements are reasonable and plausible.
We can now return to Deo Auctore 6:
Sed neque ex multitudine auctorum quod melius et aequius est judicatote, cum possit unius forsitan et deterioris sententia et multos et maiores in aliqua parte superare. ("Do not judge from the multitudo auctorum what is better and fairer. The opinion of an individual jurist of lower rank may sometimes be better than (the opinion of) the many of higher rank.")14
The multitudo auctorum cannot be all the classical lawyers. There would be no one left. But it could be a composite work of inscription-less fragments drawn from the works of the jurists singled out by the Law of Citations.
2. The works read by the compilers were confusa et dissoluta. That may explain the order or apparent disorder of the fragments in some Digest titles.
One example is Quintus Mucius Scaevola's liber singularis όρων, traditionally assigned to the Appendix. There are four fragments in the Digest. Two of them come from the last two long titles in the Digest, where the displacement of texts is exceptional: D. 50.16.241 and 50.17.73. In the first, Mucius precedes Javolenus 2 ex post Lab. In the second, it follows Javolenus 3 ex post Lab. The other two fragments are consistent with either position. In D. 43.20.8 it stands alone and could therefore fit in anywhere in the Appendix. In D. 41.1.64, it stands ahead of Labeo's pithana, which come later than Javolenus ex post Lab. , so that it could still come on either side of Javolenus. So the question boils down to this: where there is a conflict between D. 50.16 and D. 50.17, what conclusion should we draw?
Bluhme did not have a problem. He attributed D. 50.17.72 to the S Mass, so that D. 50.17.73 stands on its own and is consistent with any order.15 It is now widely accepted that D. 50.17.72 also belongs to the Appendix,16 so that the conflict is unavoidable. Mantovani concludes that it is not possible to decide where the fragments of Mucius belong, nor even to which Mass they belong.17 Osler maintains that D. 50.17 contains the primary evidence and that the order must have been altered in D. 50.16.18 Honoré,19 "with some hesitation," but no further reason, follows Osler's suggestion and places Mucius after Javolenus in his table on page 113, but in its Bluhme position before Javolenus on page 160 in his Addendum A, the BK Ordo Updated in the Light of Modern Scholarship.20
It is an unspoken assumption on the part of everyone from Bluhme onwards that the compilers worked on entire manuscripts. That is why he found it so remarkable that Proculus epistolae was split into two parts, separated by Pomponius variae lectiones.21 If they had kept in mind the words confusa et dissoluta and in particular, the word dissoluta, they might have reached a different conclusion. The simplest explanation for the conflict between D. 50.16 and D. 50.17 is that the Mucius manuscript had fallen into at least two bits, which had been picked up and read separately, one before Javolenus and the other afterwards. Where the other two fragments belong, it is impossible to say.22
A second example is Scaevola's lib. sing. quaestionum publice tractatarum, also from the Appendix. Bluhme placed it after Pomponius epistolae and senatusconsulta because of D. 46.3.93. Mantovani agreed.23 But there is also D. 24.3.65 (apparently overlooked by Mantovani) in which Scaevola precedes Pomponius epistolae. It also precedes Javolenus ex post Lab., which puts it next to Mucius (D. 50.16 fragment), though in what order we cannot tell. Krüger does not give it an asterisk. Honoré classifies it as a displacement designed to provide a mass introduction,24 and includes it in his list of displacements (inversions of the normal order) in the Appendix under the heading "Did the Editors insert the Appendix Works strictly in Order?"25 He there raises the question whether the number of such displacements or inversions is unusually high and concludes that it is difficult to say. Where that question arises, the first step is to check that the standard order is correct.
There are eleven Digest titles in Honoré's list.26 In four of them, the books of Scaevola's digesta appear in the wrong numerical order. So such displacements do exist. The question is whether the other seven titles are also cases of displacement. We have already looked at Mucius lib. sing. in which the conflict between D. 50.16 and D. 50.17 could more easily be explained by the fragmentation of the Mucius manuscript and its reading and excerpting at two different points in Bluhme's table. In the case of Scaevola's lib. sing. the conflict between D. 46.3 and D. 24.3 may be explained in the same way.
Our third example is Paul's work imperialium sententiarum. Bluhme placed it at the head of what he called Pp27 or pP,28 which since 1837 we have all called the Appendix. There has been much inconclusive discussion recently on whether it came at the end of the P Mass or the beginning of the Appendix.29 That need not concern us here, where we are interested in the conflict between D. 35.1.113, which follows Scaevola's digesta and Pomponius epistolae, and D. 50.16.240, which precedes them (and Javolenus ex post Lab. and Labeo's pithana). Osler concludes, that "since the evidence is conflicting, we must concede that in at least one case the fragment has been displaced." There is an alternative solution: that Paul's manuscript had fallen apart into at least two sections (opus dissolutum) which had been read and excerpted at different times. D. 50.16.240 comes from book 1, and D. 35.1.113 comes from book 2, so that it is possible that the break came between the two books. The only other fragment from book 2 is D. 40.1.10. Osler says that "the text has clearly been displaced." However, it stands between the S and E Masses; it is a fairly long text reporting an actual case between named parties; and it has no connection with the texts on either side: I should say that it has clearly not been displaced. But the important point is that it stands on its own and is therefore no evidence of its relationship with other works in the Appendix.30 The only other possibly relevant text is D. 36.1.83. It comes from book 1. It comes after two texts from Scaevola's digesta out of order, and its inscription contains two Roman numerals, which was forbidden. It should probably be ignored; but if it is taken into account, the break in Paul's manuscript came part of the way through book 1.
There are four Digest titles in which Labeo's pithana precede Scaevola's digesta, as against ten in which Scaevola precedes Labeo.31 Bluhme and Krüger follow the ten in reconstructing their list. Are the four displacements, or is there some other explanation for them? Krüger does not give any of them an asterisk, nor does Honoré in his "Analysis of Displaced Texts".32 Three of the texts come from Labeo book 1, which has eleven fragments altogether. Those three all concern the law of legacies, and Lenel prints them together as a little group under the heading De legatis.33 So this may be another opus dissolutum in which one of the early pages has come loose and been read separately before the rest of the work, creating a break part of the way through book 1.34 The fourth text is more difficult: D. 23.3.84. The inscription attributes it to book 6. That book provides six fragments, of which four concern the law of property, and Lenel prints them together under the heading De acquirendo rerum dominio et rei vindicatione. Our text and one other are left under a heading of two question marks. It is possible therefore that there is an error in the inscription: VI instead of I. Otherwise this text remains unexplained.
Honoré ends his discussion of these texts with the words: "In the upshot the explanation of the irregular sequence of these works of Scaevola and Labeo is obscure, but it may be relevant, as I have argued elsewhere, that both belong to a group of 48 books excerpted as a whole but divided into two lots of 24 books each." I do not understand that explanation. But there is no problem if Labeo's manuscript was an opus dissolutum and one of the early pages had come loose and been read separately at an earlier stage than Scaevola and the rest of Labeo.
The final Digest title presenting a problem is D. 46.1, where Javolenus ex post Lab. follows Labeo's digesta instead of preceding it in accordance with Bluhme's Table (h. t. 45, 46); but the Labeo fragment comes from his tenth and last book. It may be that there was a loose page here as well, which was not read until after Scaevola.
The explanation for the apparently wrong order in all these cases is either that the compilers were reading opera dissoluta or that they made all the displacements. When we bear in mind how much work the compilers had to do in how short a time, the first explanation is preferable.
So far, we have been looking at details of individual works. We can now turn to the general character of the Appendix.35 The overall picture is one of opera confusa et dissoluta. Unlike the three original Masses, it does not begin with a major work; there is no apparent organising system or common factor. It is just a miscellaneous collection of odds and ends.36 And nearly all the works show signs of being incomplete in one way or another. Some are incomplete at the beginning: Pomponius epistolae begin with, or in, book 5 (out of 20); Venuleius actiones begin with, or in, book 4 (out of 10); the Labeo/ Javolenus work in the Appendix begins in the middle of book 2; Scaevola's digesta in the Appendix begins with, or in, book 3;37 Labeo's pithana had a loose page at the beginning, which may have been read separately and earlier. Some are incomplete at the end, or must have seemed to the compilers to be incomplete in the light of the list that became the Florentine Index after the completion of the Digest:38 Paul's imp. sent. stops at book 2 (out of 6); Scaevola's digesta stop at book 34 (out of 40); Furius Anthianus on the edict stops at book 1 (out of 5); Javolenus ex post Lab. may have had a loose page at the end, which was read separately and later. The monographs by Mucius and Scaevola seem to have been fragmented. The only clearly complete works in the list seem to be Pomponius senatusconsulta (5 books) and Venuleius interdicta (6 books). There is at least one fragment from all the books in their works. That may be because they were attached in the same manuscripts to the larger works by the same authors.
When Tanta 1 spoke of opera confusa et dissoluta, it was not confined to the Appendix or in any other way. It applied to all the works read by the compilers, including the three original Masses. After the initial works in the S and E Masses, down to the digesta of Julian and of Celsus and Marcellus, there is no appearance of any overall plan: some works are grouped by author; some works are grouped by subject matter;39 some works are not grouped at all.40 The contents of the Masses, and the order of the groups and individual works in them, seem to be largely haphazard.41 The compilers working on each Mass seem to have selected the works to read independently of each other and as they came to hand in the pile of opera confusa et dissoluta. If the works were complete, they were read straight away. If they were not complete, they were set aside until the missing parts turned up. As they neared the end of the reading stage, they had to read the incomplete works even though the missing parts had not turned up. In Volusius Maecianus 14 libri de publicis judiciis in the S Mass (BK 56) they had books 5, 10 and 11. In Venuleius 4 libri de officio proconsulis (BK 91) they had two books. In Tarruntenus Paternus 4 libri de re militari in the E Mass (BK 172) they had two books. In Callistratus 6 libri ad edictum monitorium (BK 176), they had four books.42 And in Papirius Justus 20 libri constitutionum (BK 177) they had books 1 and 2 and one fragment of book 8.
We can now give a simple explanation of the Appendix. When the compilers started working on the opera confusa et dissoluta they could not conveniently do so in the room in which they found them. They therefore took them into three neighbouring rooms, where they worked on them separately as they progressed: hence the first three Masses. At the end there were only a few works and scraps of manuscript left in the original room. They therefore moved back there and finished them off. They were not read late because they arrived late; they were read late because in one way or another they were incomplete. Scaevola's digesta did not turn up late.43 That would suggest that they were not readily available at the beginning, either in Constantinople or Beirut, and were not brought from Beirut by the two antecessores, Dorotheus and Anatolius, but that they turned up later in some lesser university town and came from there. That seems most unlikely. They were part of the opera confusa et dissoluta left aside because the compilers only had thirty-four books and were waiting to find the other six. And that is the origin of the separate and miscellaneous group of works in the Appendix.
1 "Thoroughly intermixed and broken up as they may almost be called." (Monro, 1909); "Already to some extent confused and fragmented as they were." (Watson, 1975); "Gia frammentate e pressoche mescolate fra di loro." (Aldo Cenderelli, BIDR, 1993/94, 540); "Die schon fast ganz in Wirrnis und Auflosung geraten waren." (O. Behrends, R. Knütel, B. Kupisch, H.H. Seiler, Corpus Juris Civilis, Text und Ubersetzung, 2, Digesten 1-10 (Heidelberg, 1995)). [ Links ]
2 There is no explanation in Friedrich Ebrard, ZSS, vol. 40 (1919) 113-135, or in T. Wallinga, Tanta/ Dedoken Two Introductory Constitutions to Justinian's Digest (Groningen, 1989). [ Links ]
3 F. Bluhme, Die Ordnung der Fragmente in den Pandectentiteln, Zeitschrift fur geschichtliche Rechtswissenschaft 4 (1820 repr. Keip Verlag, Goldbach, 1997) 257-472. [ Links ] He refers to Tanta 1, in footnotes on pages 355, 356, 357 and 360, but never to these words.
4 T. Honoré, Tribonian (London, 1978) and Justinian's Digest Character and Compilation (Oxford, 2010). [ Links ] These words appear on page 28, note 151, of Justinian's Digest, but the text is concerned with what the compilers did to the material and not to the state in which they found it.
5 Un corso universitario seminariale progredito, Cenderelli, (1993/94) 534.
6 Franz Hofmann, Die Compilation der Digesten Justinians (Vienna, 1900) 75, [ Links ] followed by Honoré (1978) 68.
7 Hugo Krüger, Die Herstellung der Digesten Justinians und der Gang der Excerption (Münster, 1922) 3. [ Links ]
8 Aldo Cenderelli, Digesto e Predigesti Riflessioni e ipotesi di ricerca (Milan, 1983) 43 and chapter III passim. [ Links ]
9 Honoré, (2010) in an extensive bibliography (210-214) includes Cenderelli's book but omits his later article.
10 There would be a similar problem in the law schools. After the Law of Citations, if the professor cited Ulpian, one can imagine a clever student putting up his hand and asking what the other four said. All the lectures would have to be re-written. For legal education, see Omnem.
11 The translation by Monro is inconsistent: "the old books, plentiful as they were," followed by "the ancient plenty." So is the translation by Watson: "the mass of old work," followed by "the ancient plenitude."
12 As it does in Omnem 1: legum multitudo.
13 There is no parallel passage in Δέδωκεν. Cp. n. 2 above. It appears that, although Tanta and Δέδωκεν were formally addressed to the same universal audience, Tanta was intended more for lawyers who might be interested in history and legal technicalities, and Δέδωκεν was intended more for the general public who were not.
14 "You must however, when comparing a number of authors, not pronounce on the work of one as better and juster, as it is possible for the opinion of one writer, and that one of inferior merit, to be preferable in some points to many and even better authors." (Monro); "Out of a large number of authors, you must not make a judgment that the work of one is better and more equitable, since it may happen that the opinion of one writer, perhaps of inferior merit, is better at some point than those of many other authors, even superior ones." (Watson). The translations of multitudo are not consistent with note 11, above.
15 These texts in D. 50.17 may have helped Bluhme to his conclusion that texts inscribed Javolenus belonged to the S Mass and texts inscribed Labeo belonged to the Appendix: loc. cit. 320, where however there is no systematic examination of the texts.
16 T. Honoré, "Labeo's posteriora and the Digest Commission" in Daube Noster (Edinburgh/London, 1974) 161-181.
17 D. Mantovani, Digesto e Masse Bluhmiane (Milan, 1987) 110-111.
18 D. J. Osler, IURA vol. 39 (1988) 137-158, at 149-150. "Yet not so much a displacement as an interchange of position within the Appendix, which is not at all uncommon." But see below, on the order of Labeo's pithana and Scaevola's digesta.
19 (2010) 115.
20 If Mucius came after Javolenus there is no evidence as to whether it came before or after Scaevola's digesta.
21 Loc. cit. 449 and note 4.
22 See Cenderelli, (1993/94) 540: "Io ritengo che non si possa prescindere dal considerare e valutare anche il ruolo che puo avere giocato lo stato e l'organizzazione del materiale, senza dare per scontato (come, invece, di fatto si tende di fare) che essi abbiano lavorato sulle opere piu o meno integre dei giuristi classici."
23 Loc. cit. 102.
24 (2010) 183.
25 Idem 135-6.
26 The list overlooks D. 35.1.108-113, which Honoré actually discusses at page 114.
27 Loc. cit. 452.
28 Gustav Hugo, Lehrbuch der Digesten (2 ed. Berlin, 1828) 25. [ Links ]
29 Mantovani, loc. cit. 109-110; Osler, loc. cit. 147: "We may assign the work to either of these locations;" Honoré, (2010) 114-115: "The case for moving this work to the Papinian mass is therefore weak." In Bluhme's original First Table, loc. cit., between pages 266 and 267, the Appendix follows the P Mass without a break.
30 Honoré, (2010) 95, note 143, lists twelve texts possibly displaced to a position between the other two Masses.
31 Details in the last note to Krüger's Ordo librorum, from which D. 33.5 should be deleted because it concerns Labeo's posteriora, not his pithana.
32 (2010) 162-209.
33 Palingenesia, Labeo, 199-201.
34 D. 33.7 is a double title, that is, Bluhme's Masses appear twice. In the first section Labeo book 1 precedes Scaevola; in the second section it follows Scaevola. Honoré concludes that the two sections were edited separately by different compilers before they were amalgamated. I think that there was no change of editor but that part of book 1 was separated before the reading stage.
35 Cp. Honoré, (2010) 117-118.
36 "The Appendix in some ways remains more like a chance collection of texts than a mass." (Honoré, (2010) 94).
37 D. Pugsley, "The Victory Riots and the Appendix Mass" in Oxford Journal of Legal Studies, vol.11 (1991) 325-339, at 333-334, [ Links ] reprinted in D. Pugsley, Justinian's Digest and the Compilers (Exeter, 1995) at 9-10. [ Links ] See also Osler loc. cit. 137-158, at 154-158. That volume of IURA, though dated 1988, was not published until 28 December 1991, so that we reached the same conclusion independently.
38 The original list, which begins with Julian and Papinian, followed by the other jurists in rough chronological order, must pre-date the Law of Citations, which altered the order of precedence, and Justinian, who abolished it: Deo Auctore 5, in fine omnibus auctoribus juris aequa dignitate pollentibus.
39 This creates problems: should Modestinus regulae be filed with Modestinus in the E Mass or with regulae in the S Mass? In fact it was filed with Modestinus. The E compilers got there first. But what about Licinius Rufinus regulae near the end of the E Mass, or Gaius regulae fairly near the end of the P Mass, which may have been read more or less at the same time? There are many other examples.
40 Ulpian's 3 libri de officio consulis in the E Mass is a striking example. It is not grouped by author, or by subject matter (public offices), and it is too small to stand on its own. Why is it not next to Ulpian's 10 libri de officio proconsulis in the S Mass, which is also ungrouped?
41 Honoré, (2010) 78 says there was a "coherent and sophisticated scheme." I cannot see it.
42 Bluhme placed all four books in this position. Krüger moved the first two up to BK 100. Mantovani moved them back again.
43 Mantovani, loc. cit. 115, 116.