1

Recent Work on Gratian’s Decretum

Recent Work On The Making Of
Gratian’s Decretum

Anders Winroth

The last twenty-five years have been twenty-five very good years for the study of Gratian and his Concord of Discordant Canons.[1] Today we know a great deal more about the Decretum then twenty-five years ago, although Gratian himself has become even more enigmatic than ever.

Twenty-five years ago, Gratian scholars thought that they knew their man. He was a Camaldolese monk with his home in a known monastery in Bologna. But in 1979, John Noonan published an article demonstrating that this man was a myth built up over the centuries.[2] About Gratian, we can only know with any certainty that he composed the Decretum in Bologna in the 1130s and the 1140s, and that he was a teacher with theological knowledge and a lawyer’s point of view. Nothing much has been added to Noonan's portrait, except for cautious suggestions that Gratian might have been a monk or a bishop, but real evidence is lacking for either view.[3]

Twenty years ago, at the Congress in Cambridge in 1984, Stephan Kuttner gave a magisterial lecture about the accomp-lishments of Gratian research during the previous half-century and about what still remained to be done.[4] He drew up a five-point program for future research on the Decretum. I want to take this opportunity to take stock of what has been done since then and what still needs more research. I will deal with one point at a time, and I will take the liberty of going through them backwards.

Kuttner’s fifth and last point asked for Gratian’s immedi-ate sources: where did Gratian find the texts that he copied into the Decretum? Earlier scholars tended to think that Gratian used a large library of canonical collections. In 1984, Peter Landau suggested that it makes better sense to presume that Gratian used only a small number of sources. He tested this assumption by applying a set of well-defined criteria for identifying the source of a canon.[5] Using this method, he was able to demonstrate that most of the canons in the Decretum derive from Anselm of Lucca’s collection, Ivo of Chartres’s Panormia, the Tripartita, Gregory of St. Grisogono's Polycarpus, and the Collection in Three Books. In specific sections of his work, Gratian also used Isidore’s Etymologies, the work of Alger of Liège, and the Sententiae magistri A. Further research has confirmed the validity of Landau’s elegant demonstration. There remain, however, several dozen canons for which no source can be found. This state of affairs is annoying to many, but – I suspect – to no one more than to Peter Landau himself. He has published a series of articles discussing possible sources for one group of such canons after another, for example Burchard’s Decretum and the writings of Ambrose of Milan.[6]

Study of Gratian’s immediate sources has greatly benefited from the edition of several of them. I am thinking of Robert Kretzschmar’s 1986 edition of the De misericordia et iustitia by Alger of Liège, and the partial edition of the Sententiae magistri A. by Paule Maas in 1995. Even more important than these completed and printed editions are the editions under work by Martin Brett and Bruce Brasington of the Collectio Tripartita and the Panormia of Ivo of Chartres, which with characteristic generosity they have made available to the scholarly world. Another similarly important tool is Linda Fowler-Magerl’s CD-ROM, Clavis canonum, which contains entries recording the inscription, incipit, and explicit of more than 100,000 canons found in many dozens of collections from the period 1000-1140. The study of Gratian’s sources rests on solid ground.

As his fourth point, Kuttner asked for some way of managing the overwhelming number of extant manuscripts of the Decretum, in order to establish the text of a much desired new edition. We now know the manuscript tradition much better than in 1984, mostly thanks to the research of the much missed Rudolf Weigand, who studied in person or on microfilm more than 200 of the oldest Gratian manuscripts. His purpose was to study the glosses to the Decretum that were produced before the Ordinary Gloss of Johannes Teutonicus (ca. 1216). This project resulted in numerous articles and, most importantly, in his two-volume work Die Glossen zum Dekret Gratians published in 1991. Weigand’s manuscript descriptions make up a sound foundation for further research, in effect replacing parts of Kuttner’s 1937 Repertorium. Manuscript work still continues, especially in the project led by Carlos Larrainzar at the Instituto de Derecho Europeo Clásico, Tenerife.

In contrast, much more work is needed to address Kuttner’s third point, which asks for Gratian’s purpose and outlook. The text of the first recension contains many hints that Gratian 1 was familiar with ideas of the theological schools of Northern France, as has been pointed out by Richard W. Southern, Enrique De Leon, Titus Lenherr, myself, and others.[7] He clearly knew and quoted many of the biblical glosses produced in the circle around Anselm of Laon, and he used several theological works as sources for his canons, including Alger of Liège and the Sententiae magistri A. Many of the questions Gratian posed seem to be inspired by biblical exegesis. C.32 q.4 asks, for example, whether one is allowed to conceive children with a maid if one’s wife is infertile, as Abraham did with Sara’s maid Hagar, and Jacob with Rachel’s maid Bilhah. These are some of the same questions as French theological writers of the early twelfth century posed in their marriage treatises. Were such questions primarily interesting for biblical exegetes or also for practicing lawyers?

Alongside such passages are others which appear to have been more inspired by a lawyer’s practice, such as ‘causa’13. As Fred Paxton has pointed out, the argument in this ‘causa’ appears as a back-and-forth between two parties in a lawsuit, frequently even slipping into the first person.[8] These and other hints about Gratian's background and interests should be further explored.

The second point on Stephan Kuttner’s research program was the date of the Decretum. In this field we know more than in 1984. Paolo Nardi recently unearthed a Siennese court decision from 1150, which unambiguously proves that the Decretum in its second recension was known by that date.[9] Since the second recension contains numerous canons from the Second Lateran Council of 1139, we may be certain that it was composed during the 1140s.

The date of the first recension is more debatable. It contains a single, very brief reference to a decision of the Second Lateran Council of 1139 (D.63 d.p.c.34). Otherwise, the text includes no material securely dated later than 1119.[10] This leaves the field quite open for differing interpretations. Jean Werckmeister and Ken Pennington believe, for example, that the first recension was composed in or around the 1120s and later interpolated with a single reference to the Lateran Council.[11] Rudolf Weigand, Carlos Larrainzar, and I consider it dangerous to hypothesize about interpolations, and we insist that the text as it is preserved in the manuscripts could not have been finished before 1139.[12] However, we are all, I think, ready to admit that Gratian might have started work on the Decretum before 1139, even long before.

Both sides in this debate have good arguments, and new evidence is probably needed to break the impasse. An interesting approach that might lead to greater agreement is to attempt to situate Gratian in relation to contemporary theological thinking. Titus Lenherr has pointed to telling parallels with the biblical Glossa ordinaria lending support to the later date. Since the chronology of the Glossa ordinaria also is problematic, this kind of approach is complicated but potentially fruitful avenue. We heard earlier this week about Gundula Grebner’s successful work to fit Gratian into his Bolognese context, which also seems to suggest a later rather than an earlier date for the first recension.[13]

The first point on Kuttner’s 1984 program asked about the making of Gratian’s Decretum. He asked whether the work was ‘drafted and completed in one grandiose thrust, or did the original version go through successive redactions’? Titus Lenherr began to answer this question in his 1987 Munich dissertation on C.24 q.1, in which he demonstrated that Gratian, rather than using all his sources at once, used them one after the other. Gratian conceived of the question he posed on the basis solely of three canons in the Panormia. He than added 15 canons from the Polycarpus, and finally 18 canons from the Tripartita and the Collection in Three Books. Lenherr was thus able to discern three stages in the composition of this specific question.

My own dissertation built on Lenherr’s results. I observed that two Decretum manuscripts contained a text of C.24 q.1 that exactly corresponded to what the Decretum would have looked like after Gratian had used the Panormia and the Polycarpus, but before he made use of the Tripartita and the Collection in Three Books. This observation inspired further research leading to the conclusion that four manuscripts in Admont, Barcelona, Flor-ence, and Paris contain an early recension of the Decretum. This first recension differs from the previously known text in some interesting ways, for example in its use of Roman law or, as Peter Landau has pointed out, in its use of some patristic sources.[14]

Since 1996, there has been a large number of articles treating the two recensions and/or drawing on first-recension manuscripts. In this field, the contributions of Carlos Larrainzar stand out. In a series of articles, he has developed interesting theses. He distinguishes among four stages in the development of Gratian’s text, three of which he identifies with particular manuscripts, and he dates them exactly:[15]

1142-1146 Excerpta (= St. Gall, Stiftsbibliothek 673)

1148Concordia Fd (= Florence, Biblioteca Nazionale Centrale, Conv. Soppr. A. 1.402 ante correctionem)

1150Decretum Fd (= FlorenceA. 1.402 post correctionem)

1155-1165Paleae

Larrainzar believes that the earliest version of the Decretum is preserved in a manuscript in the monastic library of St. Gall in Switzerland. He dates this text to 1142-1146. A second stage is identical to the original text in the Florence manuscript. This stage largely corresponds to what I call the first recension. Larrainzar dates this stage to 1148. A third stage completed in 1150 corresponds to the second recension, that is to say, largely the text of Friedberg’s edition minus the ‘paleae’. In a fourth stage (1155-1165), the ‘paleae’ were added.

This is an appealing thesis, since it introduces more precise dates and relates stages in the production of individual manuscripts to stages in the development of the text. Carlos Larrainzar is without doubt right to identify more phases than two in the textual history of the Decretum. That history continues after the publication of the second recension, with the addition of ‘paleae’ and the deletion of duplicated texts. In a posthumous article, Rudolf Weigand treated this phase (which certainly extended further than to 1165).[16]

It is easy to recognize that the first recension did not spring fully formed from Gratian’s head. Inconsistencies and apparent after-thoughts often appear. In another posthumous article, Weigand tentatively sketched three stages in the composition of the first recension of the Decretum.[17] First, Gratian collected material thematically. Then he summarized this material in larger treatises, for example ‘de coniugio’ and ‘de hereticis’. Finally, Gratian formulated the familiar ‘causae’. In a similar vein, Mary Sommar suggested that C.7 q.1, originally only treated the translation of bishops and that Gratian later, but before the publication of the first recension, added the material about replacing a living bishop.[18] John Dillon has analyzed Gratian’s use of sources and the formulation of the case narratives introducing each ‘causa’, leading him to posit the existence of a series of short treatises that Gratian included into the Decretum.[19] More research into the prehistory of the first recension is desirable, although any conclusions remain speculative until solid manuscript evidence can be found.

Against this background, it is easy to agree with Larrainzar’s insistence on more stages than two in the textual history of the Decretum, but is it possible to relate three of those stages to stages in the production of specific manuscripts?

Central to Larrainzar’s discussion of the second and third stages is the Florence manuscript of Gratian’s Decretum. In an article published in 1998, he argued that this manuscript ‘not only contains a first and reduced Concordia of Gratian, but is the original manuscript, in which the author of the work construed the later redaction of his more ample Concordia, which was later known as the Decretum of Gratian’.[20]

In other words, Florence was the working manuscript of the author and it contains many notes in his own hand. This is why the uncorrected Florence manuscript corresponds to the second stage, whereas the text after it has been corrected corresponds exactly to the third stage. Indeed it is, quoting Larrainzar, ‘the direct and immediate source of the manuscript tradition of the vulgate Decretum’.[21]

Larrainzar fails to prove this contention, as I have pointed out elsewhere.[22] Today, I shall give only one example of his arguments. If the Florence manuscript indeed was the original manuscript of the second recension, it must predate 1150, when this recension demonstrably was in circulation. Earlier scholarship has, on paleographic grounds, dated the manuscript to around the 1170s. Larrainzar endeavors to prove that it indeed was produced in the 1140s by claiming that one of the scribal hands in the manuscript, hand C, worked in 1148 exactly.

According to Larrainzar, hands A and B wrote most of the texts of the first and second recension, respectively. Their work was corrected by hands Gα and Gτ1, which belong to the author of the Decretum himself; G stands for G(ratian).[23] Subsequently, hand C added the division into distinctions. The same hand C also added two canons from Pope Eugenius III’s council at Rheims in 1148. Therefore, Larrainzar concludes, any hand that wrote in the Florence manuscript before hand C ‘undoubtedly’ wrote before 1148.[24]

This is a conclusion that simply does not follow. Larrainzar has confused the date of the council of Rheims with the date when hand C copied decisions of that council into the manuscript. Those dates may be far removed from each other. This is easy to recognize if one considers that hand G on folio 5 of the Florence manuscript added a canon from the council of Braga of 675. This does not make hand G into a seventh-century hand. The Florence manuscript is not the author’s own working copy.

In Larrainzar’s reconstruction, the first stage of the composition of the Decretum is represented by manuscript 673 in the monastic library of St. Gall in Switzerland. When Alfons Stickler examined it in 1958, he concluded that it contains an abbreviation of Gratian’s Decretum.[25] In 1999, Larrainzar pointed out that it contains a version of the first recension.[26] I am convinced that he is right, and this is an important discovery. He further argued that the manuscript is not an abbreviation; it is a version preceding the first recension, or ― in other words ― Gratian’s first draft (‘borrador’). He fails to prove convincingly this contention, and careful study of the manuscript shows that it is wrong.

One of his arguments for the St. Gall text being an early version of the Decretum focuses on C.15 q.3 d.p.c.4, which contains a reference to ‘huius operis initium’, ‘the beginning of this work’:

C.15 q.3 d.p.c.4: ‘Sed sicut circa huius operis initium premissum est, tociens legibus inperatorum in ecclesiasticis negociis utendum est, quociens sacris canonibus obviare non inveniantur’. Aa Fd Friedberg

The reference is to the D.10 c.1: ‘Lege imperatorum non in omnibus ecclesiasticis controversiis utendum est, presertim cum inveniantur evangelice ac canonice sanctioni aliquotiensobvi-ari’. AaBc Friedberg

In the dictum in C.15, Gratian even echoes some of the language of that canon. I have bold-faced such echoes.

In the St. Gall text, no passages from the first twenty-six distinctions are included. As could be expected, C.15 q.3 d.p.c.4 in the St. Gall manuscript lacks the reference to the non-existent D.10, but the sentence is otherwise almost unchanged. In other words, the echoes of D.10 c.1, are still there. I have marked them in boldface:

C.15 q.3 d.p.c.4 in Sg: ‘Sed totiens legibus in ecclesiasticis ecclesiasticis [sic!] negotiis utendum est quotiens sacris canonibus obuiare non inuenientur’.

To Larrainzar this is proof that the St. Gall text preceded that of the first recension: ‘A simple comparison of the two “dicta” [i.e., the d.p.c.4 in St. Gall 673 and the same “dictum” in the first recension] demonstrates, unequivocally, the “precedence” of the redaction of Sg in comparison to the text of the other manuscripts [i.e., the manuscripts of the first recension]’.[27] In this view, the lack of the explicit reference to ‘initium huius operis’ proves that the St. Gall text is an earlier version of the Decretum.