The Two Recensions of Gratian's Decretum
by Anders Winroth
--------------------------------------------------------------------------------
© 1996 Anders Winroth. All rights reserved.
This paper was published with footnotes in Zeitschrift der Savigny-Stiftung für Rechtsgeschichte: Kanonistische Abteilung 83 (1997) 22-31. The text (but not the footnotes) is provided here for reference. Quotations should refer to the printed version.
--------------------------------------------------------------------------------
Fifty years ago, the Polish canonist Adam Vetulani published his ground-breaking study of Roman law in Gratian's Decretum. He showed, convincingly, that most of the excerpts from Roman law sources were added after the completion of Gratian's work, probably by another author. The success of Vetulani's admirable thesis prompted him as well as others, and particularly Jacqueline Rambaud, to explore the possibility that other components of the text of the Decretum also were late ad ditions. They suggested that Gratian's rubrics, the tractatus de consecratione, the tractatus de penitentia (entirely or in part), and the canons from the first two Lateran Councils were late additions to an already completed text. These theories rested on precarious evidence and have, on the whole, not won acceptance. In this paper, I will argue that Vetulani's and Rambaud's theories were not too bold. Rather, they were not far-reaching enough. Newly discovered evidence shows that Gratian originally circulated his Concordia discordantium canonum in a form considerably shorter than the text we know.
This early recension of the Decretum is preserved in at least three manuscripts: Admont, Stiftsbibliothek 23 and 43, Barcelona, Arxiu de la Corona d'Aragó, Ripoll 78, and Florence, Biblioteca Nazionale Centrale, Conventi Soppressi A 1. 402. These manuscripts have been known for some time, but their significance has not been recognized, although Stephan Kuttner once remarked that the Florence manuscript "may turn out to be of especial value for tracing the stages of revision of the Decretum." Other scholars who noticed that the text in these manuscripts is shorter than usual concluded that this was yet another of the many abbreviations of the Decretum. A close study of their contents shows, however, that this is not the case, and that their text cannot be the result of an abbreviator's selections from a complete text. The only possible interpretation is that this is the first recension of Gratian's Decretum, which gained a limited circulation before being supplanted by the much longer text with which we are familiar.
The first recension of the Decretum was already organized in largely the same way as the final version. But each section was shorter, so that the entire work was slightly more than half the size of the second recension. Part one was probably not yet divided into distinctions, while part two contained 36 causae, which were further subdivided into questions. The de penitentia was present, but the third part, the de consecratione, was added only in the second recension. The manuscripts of the first recension also confirm that, as Vetulani suggested, the Roman law material was not originally a part of Gratian's compilation.
How can I prove that what I here call the first recension really is an early form of the Decretum and not an abbreviation? My approach, which I can only outline here, follows two parallel lines, one of which analyzes textual details in the three manuscripts; the other studies the order in which Gratian used his sources and the sequence in which he incorporated the individual canons into the Decretum.
The textual evidence shows that the three manuscripts I mentioned contain numerous readings which are not, to the best of my knowledge, found in other manuscripts, but which must be what Gratian originally wrote. For example: The inscription of C. 24, q. 1, c. 4 reads in most manuscripts and in Friedberg's edition "Unde Alexander secundus scribit Valeriano Episcopo Martiri," that is, Gratian gives the addressee, Valerian, the titles of bishop and martyr. However, Gratian's source, the Panormia compiled by Ivo of Chartres, gives his title as "marquise," in Latin "marcioni" (as does the entire canonical tradition of this text before Gratian). Now, the Admont manuscript does indeed give Valerian the title "marquise" and the Florence manuscript contains the ambiguous abbreviation "marti." which could be expanded "martioni" as well as "martiri." (The Barcelona manuscript does not contain causa 24). Gratian most likely wrote "marcioni," like his source, and the fact that this reading is only preserved (as far as I know) in the Admont manuscript, and probably in the Florence manuscript, supports my thesis that these manuscripts contain an early recension of the Decretum rather than an abbreviation.
Another example is found in C. 11, q. 3, c. 50, which quotes Augustine saying that unjust excommunication does not harm an innocent person. He originally said: "How does it harm a man that human ignorance does not want to recite his name from this table, if a guilty conscience does not erase him from the book of the living?" The "book" referred to is the book of life mentioned in Revelation, while the "table" is the diptych which listed persons whose names were recited in mass. In Friedberg's edition, the wording of this passage is different, although the general sense of the canon remains the same. Instead of "non vult eum recitare," Friedberg has "vult eum delere," which he claims to be the reading of all manuscripts that he inspected. All but three of the manuscripts which I have examined have Friedberg's reading. The exceptions are the three manuscripts of the first recension, which read "non vult eum recitare," as does the entire canonical tradition before Gratian. Clearly, Gratian first reproduced Augustine's words faithfully and later decided to change them. Only the Admont, Barcelona, and Florence manuscripts contain Gratian's first thoughts on the subject, which further supports my thesis. Other similar examples could be cited, but I am now going to turn to my other line of argumentation, which provides the most conclusive proof for my thesis.
Several scholars have pointed out that Gratian did not write the Decretum in one giant sweep, but that he repeatedly returned to subjects he had already treated in order to add new material. This process is obvious to modern scholars from the many visible "seams" in the text, that is to say: minor inconsistencies and formal peculiarities. A decade ago, the German canonist Titus Lenherr combined evidence about Gratian's reworking the text with evidence about the source he used for each of the canons. Lenherr's results were extremely interesting. Looking at C. 24, q. 1, he found that Gratian originally constructed the questio around three canons which he had found in the Panormia. Gratian then added 15 canons extracted from another canonical collection, the Polycarpus. Finally, he added 18 canons from the Collection in Three Books and from the Collectio Tripartita. Lenherr was the first to realize that different stages in the composition of the Decretum might correspond to Gratian's use of different sources. He proved that his intuition was correct in the case of C. 24, q. 1. In making similar examinations of other parts of the Decretum, I found further confirmation of Lenherr's thesis. In general, Gratian first used the Panormia and the Polycarpus. The canons deriving from the Collection in Three Books and the Collectio Tripartita were usually added at a later stage of composition. As is to be expected, there are exceptions to this rule. But some of them may be only apparent exceptions, since it remains difficult to identify securely Gratian's source for every one of his canons.
Applying these findings to the contents of the manuscripts in Barcelona, Florence and Admont, I discovered a clear pattern. Causa 24, for example, contains 91 canons in the second recension. I found that 16 of them derive from the Collectio Tripartita and 29 from the Collection in Three Books. Not a single one of these appear in the Florence or Admont manuscripts. (Again, the Barcelona manuscript does not contain causa 24). Of the 14 texts which derive from the Panormia, 12 were originally present in the Florence manuscript and 13 in Admont. 23 canons come from the Polycarpus; 22 among them were originally present in these two manuscripts. Anselm of Lucca is represented in causa 24 by a sole text, which was not present in these manuscripts. For the sake of completeness, I should also mention that the one fragment deriving from the Glossa ordinaria to the Bible is present in both manuscripts, while the two canons from the First Lateran Council are absent from the Florence manuscript but present in Admont. I have not been able to identify the source of the remaining 5 texts, which are present in both manuscripts.
Clearly, the text of causa 24 in the Admont and the Florence manuscripts shows what the Decretum looked like before Gratian had used the Collection in Three Books and the Collectio Tripartita.
Let me illustrate what lies behind these abstract numbers. C. 24, q. 2, c. 2 is an excerpt from a letter sent by Pope Gelasius to Faustus, who at the time was the pope's envoy at the imperial court in Constantinople. This text appears in various forms in several pre-Gratian canonical collections. In order to determine which one was Gratian's source, I compared variant readings in the different collections. The result was curious although far from exceptional: My collation showed unambiguously that Gratian must have used two sources. The beginning of the text derives from the Collectio Tripartita, while the last few lines (from the word "mortuos") were taken from the Panormia, which contains only these lines. Since the Tripartita contains the entire text, it appears that Gratian first extracted the text from the Panormia and then expanded it when he found the longer excerpt in the Tripartita. The Panormia wrongly attributes the text to Pope Gregory, while the Tripartita correctly ascribes it to Pope Gelasius. Now, if one looks at the text in the Admont manuscript, one finds only the second part of the text, from the word "mortuos," and the inscription is "Gregorius papa." This is exactly what the text looks like in the Panormia. Similarly, the Florence manuscript originally contained the same shorter text and probably the same false inscription. The inscription was, however, at some point erased and replaced by the correct one, and the missing first part of the text was added in the margin by a later hand. The only possible interpretation seems to be that the two manuscripts preserve the text in the form it had before Gratian was aware of the longer text and the correct inscription in the Tripartita.
My conclusion that the Admont, Barcelona, and Florence manuscripts contain a first recension of Gratian's Decretum rests on a large body of similar evidence, which cannot be explained as accidental coincidences. Since time is short, I shall not give further examples. Instead, I would like to say a few words about the ways in which the discovery of the first recension changes our understanding of the Decretum.
Since three manuscripts of the first recension are known, it is clear that this text circulated but that the circulation remained limited. It is likely that it was relatively quickly supplanted by the second recension, which is found in hundreds of manuscripts. Also the three manuscripts of the first recension were affected by the success of the second recension. Already in the twelfth century, second-recension texts were added in these manuscripts by other scribes. In the Barcelona manuscript, most additions were made in the margins, but texts that could not be accommodated there were written on additional leaves which were inserted at appropriate places. The Florence manuscript contains only few marginal additions, while most second-recension texts are found on additional quires at the end of the book. The Admont manuscript is similarly organized, although the main text and the supplement are written by the same scribe. This manuscript is obviously a fair copy of a manuscript similar to the one in Florence. Several texts which in the Barcelona and Florence manucripts are additions in another hand are, however, in the Admont manuscript found in the main text. The scribe of Admont has apparently used an exemplar with marginal additions to the text. He copied these additions into the text, as he also did with at least one gloss. This is unfortunate, since the Admont manuscript is the only one of the three which is complete. The Barcelona manuscript is the surviving volume of a two-volume set and it ends with causa 12. The first few pages are missing in the Florence manuscript, which begins in distinction 28.
What can be said about date of composition of the first recension? The most recent texts are attributed to Pope Calixtus II, who reigned from 1119 to 1124. The canons from the Lateran Councils of 1123 and 1139 are entirely absent from the first recension, but there is a reference in passing to legislation of the Second Lateran Council. Unless this reference is an interpolation, the first recension must postdate 1139. On the other hand, it is clear that many of the early readers of the Decretum, such as Paucapalea and Peter Lombard, were already using the second recension in the early 1150's. Both recensions appear, thus, to have been completed within the comparatively short time-span from 1139 to approximately 1150.
It was long taken for granted that the Decretum was composed in Bologna. After John Noonan's critical analysis of Gratian's biography, there remains little evidence for placing the composition there. The Decretum mentions Bologna twice: in the standard forms for judicial appeal in causa 2 and in a decretal of Pope Paschal II, which was directed to the bishop of Bologna. It is noteworthy that neither of these passages are present in the first recension. There is, however, other evidence which point to Bologna as the place of composition.
The form letters in causa 2 in the first recension concern a dispute between Bishop Adelmus of Reggio Emilia and the archbishop of Ravenna. The same dispute is mentioned in a model letter collection which appears to have been produced at the Bolognese school of the ars dictaminis. It can hardly be a coincidence that Gratian used the same example. This may, then, be the best evidence for placing the composition of the first recension of the Decretum in Bologna. Also, the connection with the schools of rhetoric warns against interpreting the date that appears in one of the form letters as the date when Gratian worked on this part of the Decretum. He may well have taken the form letter, including the date, from some text-book in rhetoric.
The discovery of the first recension explains an aspect of the Decretum which often confuses readers: It is many times difficult to figure out Gratian's opinion on any given issue, since his text seems contradictory and inconsistent. James Brundage has pointed out that "the meaning and thrust of [Gratian's] arguments ... often seem difficult to follow the first time through." This statement holds true for the second recension but is less appropriate for the first recension, which contains most of the dicta found in the second recension but only about half of the canons. This means that most of Gratian's arguments and conclusions were present already in the first recension, while each section of the text was shorter. Therefore, Gratian's discussion appears more tightly argued and easier to follow than in the second recension Let me give an uncomplicated example. It has often been remarked that distinction 6 fits it context poorly. Gratian interrupts his discussion about the sources and theory of law in order to discuss nocturnal pollution. This distinction is missing in the first recension, in which Gratian, thus, keeps to the subject of the theory and sources of law.
Why, then, were more texts added to the first recension of the Decretum? I cannot provide any good answers to that question, except to point out that to collect and amass texts was a central concern in medieval canon law. A medieval canonist encountering the rich storehouse of texts in the Collection in Three Books and in the Tripartita might very well make a selection of useful texts and add them to his favorite textbook. I believe that this was what happened with Gratian's Decretum. However, the additions were not always appropriate nor were they always suitably placed. More often then not, they broke up the thrust of an argument by inserting barely relevant canons in the midst of Gratian's tightly argued case. I have a suspicion, although I cannot yet prove it, that the position of the additions has more to do with the codicological makeup of a particular manuscript than with considerations concerning the flow of Gratian's discussion. Additions often come in chunks which may correspond to the amount of text that could be written on one inserted leaf. As I pointed out earlier, the Barcelona manuscript accommodates additions in such a manner, although I do not think that it is the particular manuscript in which the additions of the second recension were first made.
To me, it seems unlikely that the author of the tightly argued first recension would disturb his work in this manner. Even if he did not have any opportunity to recast his arguments in accordance with the new texts, one could at least expect him to put them at more suitable places. I would like to suggest, therefore, that the two recensions have different authors. Which of them was Gratian? Well, since the only thing we seem to know for sure about Gratian is that he wrote the Decretum, this question is somewhat disingenuous. As Charles Homer Haskins reminds us, "nothing is gained by the process which ascribes the Homeric poems to another poet of the same name." In any case, I prefer to think of the author of the first recension as the real author of the Decretum. It was he who first brought the methods of early scholasticism to bear on canon law, thereby creating the concord of dissonant canons which gave the work its original title. I like to think that his name was Gratian.
--------------------------------------------------------------------------------
Examples
C. 24, q. 1, c. 4, inscription
Gratian's source, Ivo of Chartres' Panormia:
Alexander II Willermo martioni.
Ed. Friedberg, col. 967:
Unde Alexander II. scribit Valeriano Episcopo, Martiri.
Admont, Stiftsbibliothek 43:
Unde Alexander secundus scribit Valeriano marchioni.
Florence, BNC, Conv. Soppr. A. 1.402:
Unde Alexander secundus scribit Valeriano marti.
--------------------------------------------------------------------------------
C. 11, q. 3, c. 50
Ed. Friedberg, col. 658:
C. L. Non obest homini iniusta sententia.
Idem ad Clerum Yponiensem [epist. CXXXVII.]
Quid obest homini, quod ex illa tabula uult eum delere humana ignorantia, si de libro uiuentium non eum deleat iniqua conscientia?
Augustinus ad clerum Hipponiensem, ed. A. Goldbacher (CSEL 34.337):
quid enim obest homini, quod ex illa tabula non uult eum recitari humana ignorantia, si de libro uiuorum non eum delet iniqua conscientia?
--------------------------------------------------------------------------------
C. 24, q. 2, c. 2
Ed. Friedberg, cols. 984-985:
C. II. Ecclesia uiuentes potest ligare uel soluere, non mortuos
Item Gelasius, Fausto Magistro, fungenti legationis offitio Constantinopolim.
Legatur, ex quo est religio Christiana, uel certe detur exemplum in ecclesia Dei a quibuslibet Pontificibus aut ab ipsis Apostolis, ab ipso denique Saluatore ueniam, nisi se corrigentibus, fuisse concessam. Auditum autem sub isto celo non legitur omnino, nec dicitur, quod eorum uoce depromitur: date nobis ueniam, ut tamen nos in errore duremus. Id quoque pariter ostendant qui nobis canones nituntur opponere, quibus hoc canonibus, quibus regulis, qua lectione, quoue documento, siue a maioribus nostris, siue ab Apostolis ispis (quos potiores merito fuisse non dubium est) seu ab ipso Domino Saluatore, qui iudicaturus creditur uiuos et mortuos, siue factum est umquam, uel faciendum esse mandatur? | Mortuos suscitasse Christum legimus; in errore mortuos absoluisse non legimus. Et qui hoc tantum faciendi solus habuerit potestatem, B. Petro apostolo principaliter mandauit: "Quodcumque ligaueris super terram etc." Super terram, inquit. Nam alligatione defunctum nusquam dixit esse absoluendum.
Collectio tripartita 1.46.2
Gelasius Fausto magistro fungenti legationis officio Constantinopolim. Legatur ex quo -- defunctum nusquam dixit absolui. Non solum -- debuit submoueri.
Ivo of Chartres, Panormia 5.119
Gregorius papa Fausto magistro milicie. Mortuos suscitasse Christum -- defunctum nusquam dixit absoluendum.