This article is slated for publication in French translation in an upcoming issue of the Revue de droit canonique. English version: Copyright by author. Since the work has not been printed yet, I will be grateful for any comments or corrections from readers. FSP

Gratian’s Thirteenth Case and the Composition of the Decretum

Frederick S. Paxton

For the evidence to support the conclusions of this article see the Excel files on the author's web site

First at the Tenth International Congress of Medieval Canon Law in 1996, and most recently in The Making of Gratian’s Decretum, Anders Winroth has argued that certain manuscripts of Gratian’s text, up to now considered abbreviations, preserve in fact its first recension.[1] This first recension represents, in Winroth’s opinion, Gratian’s authentic and original work. It is a bit more than half as long as the final recension, of which Emil Friedberg’s 1879 Leipzig edition is the most familiar example, and considerably more cohesive in structure and argumentation. So cohesive is it that Winroth is hesitant to ascribe the final recension to Gratian at all. Its proliferation of canons too often overloads, confuses, or breaks up the tight argumentation evident in the shorter version.[2]

Winroth has uncovered substantial textual evidence of the priority of the text in his first recension manuscripts. There are, for example, a number of instances in which the first recension preserves a purer form of Gratian’s formal source than does the second, and none in which the opposite is the case.[3] Moreover, building on the research of Titus Lenherr, Winroth has shown conclusively that Causa 24 was composed in two stages, corresponding neatly to the versions of the text in the first and second recensions.[4] As Table 1 makes clear, 83% of the canons in the first recension of C. 24 derive from only two formal sources: Ivo of Chartres’ Panormia and Gregory of St. Grisogono’s Polycarpus. [5] 83% of the canons added in the second recension, by contrast, derive from two collections not used at all in preparing the first recension: the pseudo-Ivonian Collectio Tripartita and the anonymous Collection in Three Books (3L).

Winroth has also fully analyzed the 110 canons of Causa 11, questio 3. The results of that analysis, however, are less clear than in the former case.[6] On the one hand, the overall argument of C. 11 q. 3 is, like C. 24, more cogent and cohesive in the first than in the final recension. On the other hand, the pattern of source use is not the same. As Table 1 also reveals, whereas the canons in the first recension of C. 24 derive almost entirely from the Panormia and Polycarpus, only 3 (6%) of those in the first recension of C. 11 q. 3 have their source in the Panormia, and none at all in the Polycarpus.[7] Of those that remain, 11 derive from the Tripartita (21%) and 6 from Anselm (11%), and a further 33 (62%) have no clearly identifiable source.[8]

As for the additions in the second recension of C. 11 q. 3, roughly the same percent come from the Tripartita (24%) and the Panormia (3%) as in the first recension, and 2% come from Anselm. The composer of the second recension of C. 24, therefore, returned to the sources of the first recension for no more than 6% of the new material, while the composer of the second recension of C. 11 q. 3 did so for 30%.[9] There is one feature common to both C. 24 and C. 11 q. 3. They both draw heavily on 3L in the second recension. Table 1 shows that, of the 57 additional canons in the second recension of C. 11 q. 3, 58% are from 3L. The corresponding percentage for C. 24 is 57%. Thus, while the composer of the second recension of C. 11. q. 3 added canons from 3L at the same rate as for C. 24, he used also returned to the sources of the first recension much more frequently.

The original goal of this project was simply to see whether or not a comparison of Causa 13 in Winroth’s first recension manuscripts with Friedberg’s edition would confirm Winroth’s findings. Of the five manuscript witnesses to the first recension, one preserves a fragment of C. 11 and two end at Causa 12. Of the other two, Florence, Biblioteca Nazionale Centrale Conv. Soppr. A.1.402 (Fd), preserves the better text.[10] Thus, I initially set to work comparing the text of Causa 13 in Fd with the version in Friedberg (Fr). At the time, however, I did not reckon with Professor Carlos Larrainzar’s recent work on a manuscript in Sankt Gallen, Stiftsbibliothek 673 (Sg). Larrainzar believes that the St. Gall manuscript is a “first draft” of Gratian’s text that predates the first recension Winroth has uncovered.[11] Thus, in what follows, I will analyse Gratian’s thirteenth case as preserved in Sg, Fd and Fr. Although the results are still provisional, I hope, nonetheless, that they will contribute something to this fascinating problem and the lively debate it is generating.[12]

One thing that Winroth and Larrainzar agree on is that Fd represents the penultimate stage on the way to the final recension of the Decretum.[13] In other words, whether or not it represents the one and only first recension or is the product of the more complicated development Larrainzar is arguing, it is not an abbreviation of the standard text, but an earlier form of it. Moreover, Causa 13 is essentially the same in Sg and Fd. Both include the same dicta and canons, and both end with Fr’s canon 15 (Table 2). This basic structural agreement between Sg and Fd means that we can regard them in general as witnesses to the first recension of C. 13, and regard Fr as the main witness of the final recension. Sg and Fd are not, however, identical in every respect. Consequently, I will first discuss C. 13 as it appears in Sg and Fd in contrast to the standard version in Fr. Then, I will turn to how Sg differs from Fd and Fr. In conclusion, I will consider what the results suggest about the composition of the Decretum, and Gratian’s hand in its creation.

Causa 13 is more tightly argued in the first than in the second recension, just as  C. 24 and C. 11 q. 3 are. The additions in Fr are precisely that. They do not alter the substance of the argument at all. The slightly longer text of questio 1, dictum post canon 1 in Fr is clearly the result of a later addition.[14] As for questio 2, Fd and Sg present only 10 of the 30 canons and 2 paleae in Fr, but most of the additional canons in Fr comprise the second half of questio 2 in Fr (cc. 16-30 and the two paleae). While related in a general way by subject matter, they do not address the issues of the case, which are fully treated by the end of the Causa as it stands in Fd and Sg (that is, by q. 2 c. 15 in Fr). As for the first 15 canons of q. 2, the second part of c. 2 and cc. 3 and 13 are logical later additions, and cc. 9-11 were already implicitly recognized as such by the anonymous decretist who wrote the Summa Parisiensis ca. 1160-70.[15] Whoever is responsible, it certainly looks as if Fd and Sg preserve an earlier version of Causa 13 than does Fd.

The results of the source analysis represented in Tables 2-3 strongly suggest, moreover, that the composition of C. 13 did not follow either the very clear pattern of C. 24 or the looser pattern of C. 11 q. 3. [16] The canons in the first recension appear to derive from the Collection of Anselm of Lucca, the Tripartita, and the Collection in Three Books (3L).[17] That is, each of these collections is the most likely formal source of one or more of the 8 canons common to Sg and Fd whose sources can be identified.[18] As was the case with C. 11 q. 3, the Polycarpus is not the most likely source of any of them. Unlike both C. 11 q. 3 and C. 24, however, C. 13 preserves evidence of the use of 3L in the first recension.[19] Of the 21 additional canons in Fr that are absent in Fd and Sg, excluding the paleae, only 2 remain unidentified (Table 3). Of the rest, 2 most probably derive from Anselm of Lucca. 1 most probably comes from Ivo’s Panormia, and 1 other could have. 3 most probably come from the Tripartita, and 1 other could have. 13 most probably come from 3L, and 1 of those (13.2.9) may have originally been taken from 3L and then altered on the basis of a reading in a Tripartita manuscript.[20]

Both recensions of C. 13, thus, exhibit new patterns of sources. The first recension is based on a different set from either C. 24 or C. 11 q. 3. Where C. 24 used mainly the Panormia and the Polycarpus, and C. 11 q. 3 used Anselm, the Panormia, the Tripartita and one or more unidentified sources, C. 13 used Anselm, the Tripartita, and the Collection in Three Books. Furthermore, whoever expanded the case for the final recension returned to the original sources of C. 13 and drew on at least one new source, Ivo’s Panormia. A particularly striking characteristic of the treatment of sources in the second recension of C. 13 is that the composer seems to have shortened some of them. This is true of canons 3, 19, 26, 28 and 29 (Table 3). The fact that these five canons derive from both the 3L and the Tripartita, moreover, suggests that it was a conscious editorial decision and not the result of using alternative versions of those formal sources. This is all perfectly reasonable, but it is not the way C. 24 was composed, or C. 11 q. 3. Like C. 11 q. 3, the second recension of C. 13 draws on the same sources as the first, and at more or less the same rates (Table 1). Unlike C.24 or C.11, however, it looks as if 3L was a source for both recensions. The only thing that is consistent across all three, and which may become particularly significant as research advances, is the heavy use of 3L in the second recension. At Table 1 shows, 62% of the additions to C. 13 in the second recension derive from 3L, a percentage in almost perfect accord with the data on C. 24 (56%) and C. 11 q. 3 (58%).

Let us pause here to consider the details of the case itself. The problem Gratian presents is as follows. Driven from their homes by war, a group of parishioners have taken up residence in a new diocese. Although they continue to farm their ancestral lands, they pay tithes to their new church and choose to be buried there. After 50 years go by, the priests of the ancestral church bring suit over their right to tithes on the land within their jurisdiction. The case turns around two questions. First, should people pay tithes and hold funerals in the church of their ancestors or their baptismal church? And second, does the right to receive tithes and burials lapse over time? [21] 

In what is surely one of the most vivid examples of legal argument in the Decretum, Gratian presents the positions of the clerics of first the one church and then the other, slipping in each questio from third-person into first-person narrative. Those who brought the case to court speak first, citing a pseudo-Isidorian decretal that sets hard and fast boundaries for rights to tithes and cemeteries.[22] The other side responds, citing numerous Old and New Testament texts in support of the position that the priests who perform the sacramental duties for their parishioners should receive their tithes. The lawyer for the ancestral church shoots back (si uerba dedistis, uerba recipietis) that he does not wish to take what rightly belongs to the clerics of the new church, but only to receive what rightly belongs to his. He cites a ninth-century conciliar decree in support. (Here, Gratian breaks in, directing the reader to Causa 16 [q.1 c. 42] for the full text of the canon.)  Implicitly granting that point, the other side responds that, nevertheless, since more than thirty years has passed since the tithes have been paid to the ancestral church, the priests of the ancestral church have lost their right to collect them. They cite a decretal of Pope Gelasius to support their argument (q. 2 c.1). The clerics of the baptismal church counter with another Gelasian decree, and Gratian breaks in again, referring to Causa 16 (q. 4), where the issue is definitively settled.[23] The rights of a parish over what lies within its boundaries, so long as they are clearly established, are permanent.

The parties then turn to the question of which church has burial rights. The lawyer for the clergy of the ancestral church first cites patristic and biblical texts in support of their claims. The lawyer for the baptismal church responds with an equal number of counter examples from scripture. Then, he introduces the principle that, in the absence of a clear legal precedent, free will should come into play.[24] Since the law on burial is not clear, individuals are free to decide where they would like to take their final rest. The lawyer for the ancestral church counters that free will is limited by the demands of inheritors, including the Church, but that argument is set aside as conditional, and the case is settled. Although Gratian does not explicitly state it, it appears that the baptismal church will continue to collect tithes and oblations from its parishioners, the ancestral church will receive tithes on the land within its diocesan borders, and the parishioners themselves will decide where they wish to be buried.

It is here, at questio 2, dictum post canon 8, that the author of the Summa Parisiensis wrote that Gratian, “being on the subject of burials, has much more to say about the dead.”[25] In both Fd and Sg, one extra question is addressed, that is, whether or not money can be demanded for burials (cc. 12, 14-15). In the final recension, canons and short dicta on a number of other subjects appear: oblations for the dead (cc. 9-11), burial ad sanctos, the timing of memorial services, the propriety of mourning the dead, whether or not the dead are aware of the living, and where confessed criminals may be buried (cc. 16-32).

Overall, the results of the comparative analysis of Fd and Sg with Fr support the claim that they preserve a first recension of Gratian’s Decretum. The version of C. 13 common to Fd and Sg gives every indication of being complete in and of itself. While its initial source base and the expansions and additions in the final recension follow a different pattern than that which Winroth observed in C. 24 and C. 11 q. 3, there is nothing unreasonable about the pattern that they do follow. Additions were made in the second recension because it was a convenient place to address other canonical issues surrounding death, burial and the dead, but they add nothing of substance to the case as originally presented. That original version is, in fact, a splendid example of Gratian’s canonical method. Causa 13 presents a distinct legal problem, arguments for both sides, and the canons that support each of them. Gratian’s resolution of the case, moreover, grants a measure of justice, and some compensation, to both parties. Whoever is responsible for the second recension of the Decretum, I agree with Professor Winroth that the author of this case was the first to show how harmony could truly arise from dissonance in medieval canon law, and that we may as well call him Gratian.

But what about the differences between Fd and Sg? The most striking differences are in the incipits and explicits of many of the dicta. Often, Sg preserves a different word order, whereas Fd always agrees with Fr. For example, the initial dictum of the Causa begins with the familiar Diocesiani cuiusdam baptismalis ecclesiae in Fd and Fr, but Sg reads Cuiusdam ecclesiae baptismalis diocesiani.[26] In fact, all the initial dicta introducing the Causae are slightly different in Sg, as Larrainzar has discovered.[27] Similarly, the explicit of q. 2, d. a. c.1 reads sanctae romanae ecclesiae predia in Fd and Fr, but predia sanctae Romanae ecclesiae in Sg.[28] Sg’s version of q. 3 d. a. c. 2 exhibits different wording in both the incipit and explicit. Where Fd and Fr have Nunc autem de iure funerandi queritur…tumulandi uidentur, Sg reads Nunc queritur de iure funerandi…uidentur esse ponendi.[29]  Fd and Fr also agree on introducing every canon with a source citation followed by a brief statement of its subject matter, while Sg preserves only the source references. For example, all three introduce q. 1 c. 1 with the source citation Unde Dionisius Episcopus Seuero Episcopo, but only Fr and Fd include the following summary statement before the actual text of the canon: Singuli sacerdotes proprias debent habere ecclesias.

Such differences could be evidence of a prior stage in the composition of C. 13, but they could also be the work of an abbreviator or someone who had taken notes from a lecture on the first recension text.[30] None of them substantiate Larrainzar’s claims for the priority of Sg. Those claims are, however, supported by the overall structure of Sg. For what makes Sg unique, and uniquely intriguing, is that it has only one part. That is to say, unlike both Fd and the final version, there is no separate section in Sg devoted to distinctiones. Instead, there are simply 33 Causae. Absent are C. 24-26 and 28. Present is a Causa prima that contains canons and dicta from distinctiones 27-39, 45-6, 50, 54-5, 60-63, 68-72, 74, 77, 79, and 101. Larrainzar reports the introduction to Causa 1 in Sg as follows: “A certain literate layman had a concubine, whom he sent away upon becoming a subdeacon. Then he separated from his wife. A bit later, he ascended to the rank of deacon and was elected bishop. The first question is whether or not a married couple ought to be separated after a vow. The second is whether or not someone who has had a concubine ought to be elected bishop.”[31] A review of the first Causa in Sg, shows that the canons it contains address these two questions in turn, very cogently and with little extraneous material.

Sg has 33 Cases as one complete work. Fd has the familiar 36 Cases of Part II of the Decretum, preceded by the first Part, although not yet divided into distinctiones. In both parts there is much less material than in the final recension. Could Sg be an abbreviation of Winroth’s first recension? That is, would someone arrange selections from 36 of the 101 distinctiones into a Causa and cut out Causae 24-26 and 28 to make a pure casebook? It does seem unlikely. Why omit just those Causae? Why go to the trouble of excerpting a small group of distinctiones from Part 1 to make a new Causa prima? Nevertheless, it is possible.[32] It is also possible, however, that Sg’s Causa prima could have been a first draft of material that was later expanded into the 101 distinctions of Part I of the Decretum. [33] If it is, then there were at least three successive steps in the development of the text: the original 33 Cases in Sg, which were then expanded into the two parts of the Fd, which were then expanded again into the final familiar version of the text.

Other questions complicate the issue. For example, in C. 16 q. 1 d. p. c. 16, Gratian refers back to C. 13 q. 2 c. 6 in the following manner, citing a decree of the Council of Tribur: Ubicumque facultas rerum et oportunitas temporum suppetit, etc. sicut in eodem capite supra legitur in causa eorum, qui de diocesi ad diocesim transierunt. The trouble is, although the main body of the text of C. 13. q. 2 c. 6 equals the text that follows the incipit Gratian uses in C. 16, which is standard in all the known sources, the incipit in C. 13 is significantly different. It reads Ubicumque temporum uel locorum facultas tulerit. How could this happen? If Gratian wrote C. 13 first, which this evidence seems to suggest, since C. 13 contains the full text to which C. 16 refers, then why did he fail to correct the irregular incipit to C. 13 q. 2 c. 6 when writing C. 16? On the other hand, there are cross references in the initial dicta of C. 13 to C. 16, which suggest that C. 16 was already written when C. 13 was composed.[34]  Moreover, all these cross references are present in both Fd and Sg, and therefore in the oldest layer of the text that we know. What are we to make of such “untidy seams” at this textual level?[35] The most plausible explanation is that the first recension was itself the product of a prior development and still a work in progress. This accords with Titus Lenherr’s intuition, based in part on his own analysis of the beginning of C. 13, that the first recension texts of Gratian’s dicta have a long history behind them.[36] And why not? As Winroth’s research makes clear, the first recension could very well be the product of as many as 20 years of work by Gratian (1119-1139).[37]

Overall, the analysis presented here suggests that, however clearly the text in Sg and Fd preceded the text in Fr, there are still open questions about the method of composition of both recensions. Only time and more hard analytical and comparative work will decide the issue. It may, in fact, be necessary to analyze fully all the Causae of both recensions, along with the Distinctiones of Part I, before an overall pattern emerges, if there is one at all. For now, let me conclude with some thoughts on where Gratian’s own hand might be in all this.

Whether or not Sg preserves an earlier version of the Decretum than Fd, I agree with Winroth that the Causae are the real hallmark of Gratian’s mind. They are, after all, the common denominator of all textual traditions, and thus the most properly ‘original’ feature of his work, as John Noonan argued over twenty years ago.[38] If the 33 Causae of Sg are, however, Gratian’s work, then perhaps he had no hand even in the creation of the version in Fd. After all, there is already a loss of coherence in the Distinctiones of Part I, which are not as tightly organized as the Causae. But perhaps not. If the text of C. 13 in Fd and Sg ended before the question on burial fees, I would be more inclined to support the claim that Gratian would not have diluted his work with non-essential material. As it is, the evidence of a tendency for the work to grow by accreting material on issues tangentially related to the case at hand, which is common to both Sg and Fd, and so to whatever might be the oldest layer of the text, gives me pause. But then, who is to say that the author of the Causae was so jealous of his achievement? Perhaps he was simply compelled by his teaching duties to continually expand on his initial work. Weigand asserted that that was the reason Gratian developed the Causae out of loosely organized tractates.[39] The need for a compendium of canon law for teaching purposes could have exercised its influence over the casebook that was the Concordia discordantium canonum, forcing it to expand into the Decretum. Whatever the case may be, I see no way yet of knowing for certain exactly where Gratian’s personal contribution began and ended.

The mind and experience of a lawyer clearly lie behind the narrative at the core of Gratian’s 13^th case.[40] What remains unclear is where the line between a pure casebook and a canonical reference work lay in the mid-twelfth century, or even whether or not Gratian would have drawn it as we would today. The answer may lie in the manner of composition, but a definitive pattern has yet to appear. Thanks above all to Anders Winroth, but also to Rudolf Weigand, Peter Landau, Titus Lenherr, and Carlos Larrainzar, we know a great deal more about the making of Gratian’s Decretum than ever before. Nevertheless, much remains to be done.