Tuesday, November 8, 2016

Lights out

As I begin to produce a preliminary electronic edition of the Pseudo-Isidorian decretals, it's time for this blog to move. Pseudo-Isidore now has his own website. See you there.

Wednesday, August 31, 2016

Patzold and the Origins of the C Recension: III

The third installment in a series of observations about Patzold's new book. The beginning is here.

A great part of Patzold’s case for the ninth-cenury date of the C recension rests upon the canonical citations of Hincmar of Reims. Among other things, he raises the possibility that Hincmar drew on some collection like C for his Opusculum LV capitulorum, or his treatise in 55 chapters (55C). In this work from 870, Hincmar refuted the legal arguments of his nephew and namesake, the bishop of Laon:
In the conflict with his nephew, Hincmar adduced in his 55-chapter treatise...a passage from a forged letter in the name of Pelagius II (JK +1051)—from a text, therefore, which is contained not in the A2-Version, but only in one of the long versions of the [False Decretals] (including the C-class). In the same work...Hincmar referred pointedly again to the Breviarium of Liberatus of Carthage.
On the same occasion, moreover, Hincmar cited four letters of Pope Leo I, which he could have taken neither from the Quesnelliana nor from the Hispana (JK 475, 482, 495 and 496). Only one of these four letters (JK 496) can be read in the Leo Dossier of the A1 recension; all four, however, occur in codices of the C class. Still more: Precisely in the argumentative context where Hincmar borrows from the Breviarium of Liberatus of Carthage, he also cites two of these exclusive Leo letters (JK 495, 482), which, in Pseudo-Isidorian context, are transmitted only in the C class. (49-50)
Put another way, Hincmar cites diverse items in 55C that the C recension alone gathers together in one single book, these texts being the Breviarium of Liberatus and a Pseudo-Pelagius forgery, or various rare Leonine decretals and the Breviarium. To weigh the strength of this argument, we must remember that the 55C is a dense legal treatise that gathers hundreds upon hundreds of canonical citations, no few of them rare or unusual in some way. It is a monument to Hincmar’s deep learning. The C recension is, likewise, an enormous compendium of canon law, surely one of the longest and most comprehensive collections of the early medieval period. Additionally, C has close ties to the Reims province, discernible in its antecedents (A/B, likely from Corbie) and in its sources (the Collectio Sangermanensis). Hincmar himself was the metropolitan of Reims province. This means we should approach any overlap in content with caution. Hincmar and C were both comprehensive and they both breathed the same air. If they have the same stuff here and there we would do well to avoid leaping, straightaway, to any theory that posits C as Hincmar’s source.

Despite these words of caution I find myself intrigued by chapter 23 of the Opusculum LV capitolorum, where indeed Hincmar cites the Breviarium and then some rarae aves from Leo’s epistolary corpus. If we want to imagine that Hincmar got these citations from one book, that book begins to look an awful lot like C. But before we get carried away, we should pause and ask why this argument is framed so narrowly.

Was it only in composing the 55C that Hincmar might have turned to C for help?

Where did Hincmar ordinarily go shopping for his Leo letters?

How well did Hincmar know Leo anyway?

On all of these matters, Jean Devisse has helpful remarks.[1] To begin with, Devisse is able to document Hincmar’s longstanding and lively interest in Leo’s correspondence. Everything suggests that Hincmar was on the lookout for Leonine extravagantes throughout his career, and that his knowledge of Leo’s epistolary leavings grew over time. Hincmar, Devisse argues, cited Leo’s letters from multiple sources, including the Dionysio-Hadriana, the Quesnelliana, and the Hispana. For the odd piece he used the Dacheriana and also rarer collections, namely the Collectio Colbertina (which perhaps survives solely in Hincmar’s personal manuscript, Paris, Bibliothèque nationale, Ms. lat. 1455?) and the Collectio Arletanensis (which survives only indirectly).

Hincmar cites a great number of Leo letters that end up in Pseudo-Isidore, but in most cases Hincmar seems to take his borrowings from Pseudo-Isidore’s sources, rather than Pseudo-Isidore himself. In only one case is Devisse fairly certain that Hincmar’s Leo citation comes from a False Decretals manuscript (the decretal in question is present both in A1 and in C). In later years, however, and especially from 870, Devisse sees an increasing textual influence from the False Decretals upon some of Hincmar’s Leonine borrowings.

All of that is according to Devisse, whose analysis is at times less than tight and in many cases wanting in citations. This, however, is a point of absolute certainty:

The problem of Leonine citations in Hincmar that cannot be traced to sources beyond C is larger than the four citations in the 55C, and I do not fully understand why Patzold has framed this problem in such narrow terms. Across his entire oeuvre, Hincmar cites nine Leonine decretals that are also on hand in C, but that otherwise are very rare, occurring only in unlikely places like the Collectio Grimanica. 

Where Hincmar got these nine letters is a deep problem. It is so deep that any theory of a ninth-century C recension seems inadequate to answer it. Consider this list of each decretal and the earliest attestation in Hincmar, insofar as I know it:

1. JK 431 (n. 37), cited only in Hincmar’s De fide Carolo regi servanda from 875.

2. JK 432 (n. 38), cited only in the De praedestinatione from 859/60.

3. JK 464 (n. 84), cited only in De una et non trina deitate, 856/7.

4. JK 475 (n. 95), first cited in De praedestinatione, 859/60 (and later in 55C).

5. JK 482 (n. 105), first cited in a letter of Hincmar from 863 (and later in 55C).

6. JK 487 (n. 111), cited only in De praedestinatione, 859/60.

7. JK 495 (n. 119), first cited in De una et non trina deitate, 856/7 (and later in 55C).

8. JK 496 (n. 120), cited in OpusculumLV capitulorum (the 55C) in 870.

9. JK 532 (n. 156), first cited in De una et non trina deitate, 856/7.

Res ipsa loquitur: A great many of these rare Leo letters, available in C but in no other widely circulated formal source, entered Hincmar’s world in the course of his quarrels with Gottschalk. A third had already arrived in 856 or 857, if we are to believe Devisse’s arguments on the date of De una et non trina deitate.[2] Such an early date makes it hard to see how C can have been Hincmar’s source for many of these citations. At the beginning of this essay series we discussed a distant antecedent of C, namely the collection of Leo’s correspondence in New Haven, Beinecke Rare Book and Manuscript Library, Ms 442. The Leo collection in this manuscript is associated with a papal list that concludes with the pontificate of Nicholas I. All of which is to say that C has a terminus post quem and, and it sits sometime after 858.

When Hincmar wrote his 55C in 870, I suppose that he could have drawn the Breviarium and some of his rare Leo citations from C. Anything is possible. But we have to admit that he did not know all of these letters only from C, for he had cited one of them (JK 495) before C ever took shape, during his attempt to do away with the trina deitas in 856 or 857. And he had demonstrated knowledge of a second (JK 475) in the course of composing an extended statement on predestination in 859/60. Though theoretically possible, this still strikes me as an uncomfortably early date for C, given that several stages of redaction and expansion had to intervene between the Leonine collection in the Beinecke manuscript and the Leonine collection in C.

As I come to the end of this post, I cannot help wondering if we have not grasped this question the wrong way around. Hincmar had a deep and unusual knowledge of Leo’s epistolary corpus, one that according to Devisse was by and large independent of the False Decretals. This knowledge extended to some of the letters that eventually, after Hincmar demonstrated knowledge of them, found their way into the C recension. Among those responsible the C recension was, likewise, somebody with a fairly deep and prevailing knowledge of Leo’s correspondence, if the 102-item Leonine dossier in C is any indication. If Hincmar's citations do not proceed from C, perhaps some of the material in C proceeds from Hincmar?

That will be a question for next time.

[1] Devisse, Hincmar: Archevêque de Reims, 845-882 (Geneva, 1975-6). At 3 vols., not for the faint of heart. 

[2] See Devisse, Hincmar, 1:163-6. His argument is widely accepted in modern scholarship as far as I know.

Friday, August 19, 2016

Patzold and the Origins of the C Recension: II

A multi-part ode upon aspects of a recent book and the C recension of the False Decretals. First post here.

C receives material from the Collectio Sangermanensis. That means something. Unfortunately, the Sangermanensis is also very hard. The best extant manuscript, Paris, Bibliothèque nationale de France, Ms. lat. 12098 (hereafter, P), indeed hails from Corbie. It was copied in the third quarter of the ninth century and its margins have received various annotations. Patzold wonders whether these annotations are evidence that the Pseudo-Isidorians or the compilers of C took an interest in the volume. Yet P coexists with Vienna, Österreichische Nationalbibliothek, Ms. 397 (hereafter, W). And a third copy was once known at Beauvais but has since been lost. Finally, Hincmar of Reims demonstrates knowledge of some of the items assembled in the Sangermanensis

For Patzold, this suggests that Hincmar had access to a version of the False Decretals very much like C. That is a possibility, but at this stage the blunter and simpler approach is simply to acknowledge that we have four distinct witnesses to the Sangermanensis from the ninth century: three of them direct (P, W and the lost Beauvais codex) and one indirect (Hincmar). Of these four, three hail from the archiepiscopal province of Reims (P, Beauvais, Hincmar). To argue that the Sangermanensis points to Corbie is therefore not wrong so much as it might be overly precise—a neglect of the forest for the trees. The synoptic view must be that the Sangermanensis has ties to Reims generally. The Reims province, of course, is not only Pseudo-Isidore’s stomping grounds, but the epicenter of his ecclesiastical concerns and the environment in which the False Decretals first circulated and where they were the most used.

Readers at Corbie have annotated the folios of P. You can study the annotations yourself. So far I have found nothing obviously Pseudo-Isidorian about these annotations. That is to say, P does not have the distinctive Pseudo-Isidorian marginal cipher that Klaus Zechiel-Eckes discovered in the source codices directly exploited by the forgers. Furthermore, I cannot spot any direct relationship between marginal annotations in P and material appropriated by the C recension. I am open to being corrected on this point, but until I am, I must write that I see no evidence that ties Sangermanensis material in the C recension to P specifically. Patzold’s argument on this front resembles somewhat his assertions surrounding the Bobbiensis and the Grimanica: It is possible to lean upon the evidence such that a Corbie connection becomes arguable, but that connection does not emerge unforced from the sources.

The Collectio Sangermanensis can be divided into four parts:

1. A substantial fragment of the Codex Encyclius. The Codex Encyclius, issued by Emperor Leo I in 458, consists of letters from the pope and metropolitans from across the empire on the orthodoxy of the Council of Chalcedon and the case of Timothy Aelurus, patriarch of Alexandria. Timothy’s predecessor, Proterius, had been driven from his see and killed by anti-Chalcedonians, who then installed Timothy as an ideologically acceptable replacement. The Codex Encyclius is a resounding statement against Timothy and in favor of Chalcedon. It was translated into Latin under the aegis of Cassiodorus at Vivarium. 

2. The Breviarium of Liberatus of Carthage, essentially a polemical history in defense of the Three Chapters that opens with the consecration of Nestorius as patriarch of Constantinople in 428, and that concludes with the Second Council of Constantinpole in 553. The Breviarium draws heavily on the Historia Tripartita of Cassiodorus/Epiphanius.

3. Five items that seem out of step with the rest of the Sangermanensis. These include 
i) Prosper of Aquitaine’s Contra collatorem, a treatise on the necessity of grace written in refutation of John Cassian’s Conlatio XIII;
ii) a letter of Aurelius of Carthage on the condemnation of Pelagius (Clavis Patrum Latinorum no. 395);
iii) excerpts from Augustine, Ep. 186, also relating to Pelagianism; 
iv) Pope Gelasius I to Euphemius of Constantinople, Quod plena cupimus, JK 620, in which Gelasius rebuffs the attempts of Euphemius to restore communion during the Acacian schism in 492; 
v) Damasus to Paulinus of Antioch, Per filium meum, JK 235, in the usual form with anathemas appended from the Roman council of 380. 
There are two reasons to suspect these pieces might not belong: The dominant concern of the second half of the Sangermanensis, from the Breviarium in Part 2 through the letters in Part 4 below, is a defense of the Three Chapters. The items in Part 3 address other matters, particularly, in i-iii, Pelagiansim. Moreover, some of them seem to hail from the Collectio Quesnelliana or a related collection (ii, iii, v). The rest of the Sangermanensis manifests no Quesnelliana connection.

4. A concluding sequence of letters and other brief items purporting to bear, like the Breviarium of Liberatus, on the Three Chapters.

Eduard Schwartz, editor of the Sangermanensis, argued that Part 3 did not belong and he excluded it as a later accretion from his edition. This third part is nevertheless present in both extant witnesses to the Sangermanensis, P and W. We cannot know for certain whether it stood in the lost codex from Beauvais. What textual evidence there is, however, suggests that the lost Beauvais witness aligns textually with P as against W, so it would not be surprising to find that this lost manuscript did attest to all four parts. For what it's worth, Hincmar, our indirect witness, cites Propser's Contra collatorem (from Part 3?) and he also knows the Breviarium in Part 2.

Did the architects of C also have access to a four-part Sangermanensis like that described above? Well, maybe. They appropriate both the Breviarium in Part 2 and the concluding letters in Part 4, along with the entirety of Prosper’s Contra collatorem., this later very probably from Part 3. They also include Gelasius to Euphemius, JK 620, another constituent of Part 3, though there is some question about their source for this letter.[1] Patzold concludes that "Whoever was responsible for the C-class must have had, as a basis for his work, a codex like that which we can still grasp in the two aforementioned manuscripts" W and P (37). 

For Patzold, it seems especially probative that the architects of C knew not only the hard and fast constituents of the Sangermanensis, but also the accretions in Part 3. Yet this is precisely where it gets hard. The majority of Sangermanensis content in C is appended to the Leonine dossier that we discussed last time. We get 102 letters of Leo the Great, followed by the Breviarium of Liberatus (Part 2 above), followed by the greater part of the concluding letters (Part 4 above). It is only later on, disconnected from Leo and from these appropriations, that we find possible traces of Part 3: Prosper’s Contra collatorem comes after the decretals of Silverius in C; and JK 620, Gelasius to Epiphanius, stands among the Gelasian decretals in C. Schwartz, who took into account the Pseudo-Isidorian tradition when editing the Sangermanensis, did not think that the items in Part 3 belonged and he preferred to read them as a later accretion; and he was confirmed in this by the Pseudo-Isidorian tradition on hand in C, which appears to leap directly from the Breviarium in Part 2 to the concluding letters in Part 4. Schwartz either did not notice or he did not care that the architects of C might well have known some of these supposed accretions, though they inserted them elsewhere.

It could surely be a coincidence that C is arranged in such a way as to provide Schwartz with a text-critical argument to exclude Part 3, which he wished to read as extraneous to the Sangermanensis for separate, internal reasons. Keep in mind, however, that my four-part division of the Sangermanensis is artificial and calibrated to the purposes of this discussion. In manuscript this collection presents itself as an undifferentiated sequence, beginning with the Codex Encyclius, continuing with the Breviarium, followed immediately by Prosper, then Aurelius, then the anti-Pelagian statement from Augustine, then Gelasius, then Damasus, and then—again without any differentiation—the final fund of letters that return the focus to the Three Chapters. How likely is it that the architects of C could have neatly excluded, from their appendix to the Leo letters, all the material that strikes Schwartz, and perhaps us as well, as a later addition? How likely is it that their knife should have excised those letters associated, at a formal level, with the Quesnelliana? If I have learned anything from Gratianus, it is that I should pay attention whenever a formal source leaves the building. Even an abiding interest in the Three Chapters could not account for C's leap from Part 2 to Part 4, for the connection of various items in Part 4 to the Three Chapters is less than obvious, requiring some explanation on Schwartz’s part.[2] 

For all of these reasons, I would find it far from crazy to argue that the C recension draws on a different tradition than P, the Corbie copy of the Sangermanensis—one that provided the accretions in Part 3 not between Part 2 and Part 4, but perhaps in some other connection.

In conclusion, and by way of affirming that there is reason to believe that the architects of C knew the subcollection in Part 3 in some form, I provide an overview of how the material in Part 3 of the Sangermanensis is handled by Pseudo-Isidore/C overall:

i) Prosper, Contra collatorem, ends up in C, attached to the decretals of Silverius.

ii) Aurelius of Carthage on the condemnation of Pelagius ends up in the anti-Pelagian appendix to the collection of Leo’s letters that, last time, we called 71L. This is Stage 4 in Chavasse’s progression. This appendix is replaced sometime after Stage 5, so that its material does not occur in the Leonine dossier of the C recension.

iii) Excerpts from Augustine, Ep. 186, likewise flow to the disappearing anti-Pelagian appendix in 71L.

iv) Gelasius, JK 620, ends up in C, attached to the rest of the Gelasius decretals on hand there.

v) Damasus to Paulinus, JK 235, Per filium meum, with its appendix of conciliar anathemas, was available to Pseudo-Isidore from the beginning in the Hispana. C does not take it up anew.[3]

The contents of Sangermanensis/Part 3 are distributed throughout the world of Pseudo-Isidore/C in very complementary fashion. Nothing is wholly neglected; ii) and iii), which seem at first to have been neglected by the architects of C, were in fact taken up in an antecedent collection. And as v) reveals, nothing is ever duplicated.[3] Otherwise, in the world of C, any association between the supposed accretions in Part 3 of the Sangermanensis and material from Part 2 or Part 4 of the collection are very consistently lacking. Whether we are talking about the appendix to 71L, or the decretals of Silverius and Gelasius in C, items from Part 3 never appear with Sangermanensis content from Part 2 or Part 4.

Next, thoughts on Hincmar and the C recension.

[1] Detlev Jasper, "Papal Letters and Decretals Written from the Beginning through the Pontificate of Gregory the Great (to 604)," in Papal Letters in the Early Middle Ages (Washington DC, 2001), 64. The Gelasius dossier in the C recension reproduces exactly the order of the Gelasius dossier in the Collectio Frisingensis (Munich, Bayerische Staatsbibliothek, Clm 6243, s. VIII): "Whether this sequence has occurred by coincidence or is the result of the textual tradition remains unanswered." Relevant to this puzzle are Patzold's remarks in Gefälschtes Recht, 36-37, note 70.

[2] As a final proof, consider that the ninth-century table of contents in P does not demarcate the contents of Part 3 in any way at all. This account divides the context of the codex into the Codex Encyclius, the Breviarium of Liberatus, and Prosper’s Contra collatorem, and then finally a series of “epistolae sanctorum patrum et exceptiones quaedam satis utillimae.”

[3] This is in contrast to the handling of JK 235 in the A1 recension. As noted above, the letter is divided into two parts, an initial statement to Paulinus followed by the conclusions of a Roman council anathematizing various heresies. The conciliar anathemas also occur in Cassiodorus/Epiphanius, Historia Tripartita, IX.15-16. A1 folds both of these chapters from the Historia Tripartita into its Damasus dossier, and the curious consequence is that the A1 recension (and therefore Hinschius’s edition) provides the conciliar anathemas from 380 twice, once from the Hispana and once from Cassiodorus/Epiphanius. The Tripartita borrowing corresponds to an excerpt mark in Pseudo-Isidore’s codex of the Tripartita (Zechiel-Eckes, "Ein Blick in Pseudoisidors Werkstatt," Francia 28 [2001], 43).

Tuesday, August 16, 2016

Patzold and the Origins of the C Recension: I

Steffen Patzold, Gefälschtes Recht aus dem Frühmittelalter: Untersuchungen zur Herstellung und Überlieferung der pseudoisidorischen Dekretalen (Heidelberg, 2015)

Steffen Patzold has written a stimulating monograph on the origins of the C recension of the False Decretals. Francia Recensio asked me to review this book last fall, but the distractions of leave delayed my work. I have been a Bad Scholar. Then, when I began studying Patzold’s volume, I became so intrigued by the problems he raises that the review became still more delayed. Happily it is now on its way out, but I will spare you the suspense: I have always found Patzold to be an excellent historian and I have learned a lot from him. I really like his book, even though I find myself in fairly deep disagreement with its argument.

A primer on the manuscript tradition of the False Decretals and what is up with these recensions: Pseudo-Isidore’s decretal forgeries circulated in several different, well-defined forms, which were first subjected to systematic description and analysis by Paul Hinschius when he edited the Decretales Pseudo-Isidorianae in 1863. For Hinschius only the A1 and the A2 recensions dated from the era of Pseudo-Isidore. Three further recensions, which he christened A/B, B and C, were, according to him, all high medieval developments. Hinschius defined his recensions without the benefits of mature text-critical theory and without any deeper, common-errors based analysis of the Pseudo-Isidorian tradition. This is to say that A1, A2, A/B, B and C are not archetypes in the text-critical sense of the term; instead, they denote only roughly the different forms in which the False Decretals circulated. Modern scholars discovered that Hinschius misdated the key manuscript witness of the A/B recension, which in fact comes to us from the era of the forgeries. According to me anyway, A/B is probably the oldest recension out there. And while the manuscript tradition of C does not antedate the later twelfth century, scholars here and there have suspected that it might be older than its extant codicological witnesses. After reading Patzold’s book, I am convinced that these suspicions were right. Minus a few obviously later accretions, C is an early medieval recension and can no longer be excluded from the study of Pseudo-Isidore’s early reception and circulation.

The antiquity of C is Patzold’s concern only for the first third of his study. He spends the final two-thirds of his book on arguing that the “core of the C recension” can be dated to the ninth century and ascribed to the Pseudo-Isidorian atelier at Corbie. Most of Patzold’s case depends upon many items associated with the collection of Leo’s letters in C, and that is what I want to look at here and in the next few posts. 

C presents 102 pieces of Leo’s correspondence, by far the largest and most complex Leonine dossier associated with Pseudo-Isidore. C appends still more material to the end of the Leo letters, the greater part of it relating to the Council of Chalcedon (451) and the Three Chapters controversy from the sixth century. Patzold is absolutely right to look for clues as to the origins of C in this canonical jungle. Many of the items that C associates with Leo's dossier are rare artifacts that suggest the resources of a highly distinctive and well equipped library. Patzold pushes very hard to place that library at Corbie, but it is far from clear that he is right. At points the relatively simple dimensions of his argument deny his study an appreciation of the complexity and potential meaning of the evidence before him.

So, 102 letters: By way of comparison, the A1 recension favored Hinschius has only 56 or 57 letters, and the interpolated Hispana and the A/B recensions include a mere 39. Patzold shows that nothing about the contents or configuration of the Leonine dossier in C suggests a high medieval date, despite some suggestions to the contrary. He also notes that C incorporates one decretal from the so-called Collectio Corbeiensis (Paris, Bibliothèque nationale de France, Ms. lat. 12097), a Corbie source also tapped by Benedictus Levita in the False Capitularies (39-40). Furthermore, the Leo letters assembled in C are derived from two early medieval compendia of Leo’s correspondence, namely the Collectio Bobbiensis (Milan, Biblioteca Ambrosiana, C.238 inf) and the Collectio Grimanica (Paris, Bibliothèque Mazarine, Ms. 1645). According to Eduard Schwartz, the Bobbiensis is excerpted from the Grimanica. True to its name, the Milan manuscript of the Bobbiensis was moreover set down in Bobbio, which Patzold reads as evidence that the Grimanica must also have been at Bobbio. Because Abbot Wala of Corbie served as Bobbio’s abbot between 834 and 836, Patzold sees the Grimanica and the Bobbiensis as clues of Corbie provenance:  
For the production of the C-class, rare antecedents were necessary, which were partly available in Bobbio and partly at Corbie. In the 830s, we see a close personal connection between these two monasteries, which was moreover supported by common Pseudo-Isidorian interests (42).
I am not sure that this is right. 

Patzold suggests that Ambrosiana C.238 inf is not the ur-manuscript of the Bobbiensis, but rather a copy of the lost exemplar. He seems driven to this hypothesis because he is convinced that the Milan codex is very late, perhaps dating to the early tenth century. (Bischoff dates Ambrosiana C.238 inf to s. IX 3/3.) Similar considerations apply to the Grimanica: While the Bobbiensis does indeed appear to derive from something like the notional Grimanica, post-Schwartz research holds that Milan manuscript does not itself descend from the Paris codex. Both rather proceed from a common exemplar, from which the Bobbiensis is perhaps, at a minimum, twice removed.[1] The origins and provenance of this antecedent—that is to say, the origins and provenance of the Grimanica in a more fundamental sense—are unknown. 

The upshot is that neither the Milan codex of the Bobbiensis nor the Paris codex of the Grimanica need have informed C directly. Indeed, Patzold seems to suggest that neither did. I do not see how the Bobbio connection can survive this observation, to say nothing of the Corbie connection that hangs beneath it. The “rare antecedents” necessary for the Leonine dossier in C are therefore without any discernible connection to Pseudo-Isidore's monastery, save for one item borrowed from the Collectio Corbeiensis. Otherwise, the best we can say is that C draws Leo’s correspondence from two rare repositories of unknown origin that, in their extant forms, have only Italian associations. Either before or after C took shape, one of these collections, the Bobbiensis, found its way to Bobbio. Beyond the fact that the sole Paris manuscript hails from northern Italy (specifically, Frioul), we do not know anything about where the Grimanica was known in the ninth century, or where it might have given rise to the Bobbiensis.
Thus far we have only considered formal sources, but there is another way to approach the fund of Leo letters in C. We can consider this dossier as the product of a long process of expansion and redaction that admits of partial reconstruction. The standard study, by Antoine Chavasse, holds that the Leo letters in C reflect at least six distinct developmental stages.[2] While not every aspect of Chavasse’s scheme has been proven beyond all doubt, the broader picture of a dense and multi-layered Leonine compendium cannot be disputed. Here is the story of C as Chavasse tells it, from beginning to end:

1) At the core of the Leo letters in C lies the original dossier of 39 letters that the Pseudo-Isidorians redacted from their most important formal source, the Collectio Hispana Gallica. As emended by the forgers, this 39-letter collection—we will call it 39L—is transmitted by Pseudo-Isidore’s own interpolated recension of the Hispana Gallica (Vatican, Biblioteca Apostolica, Vat. lat. 1341, usually called the Hispana GallicaAugustodunensis) and by the A/B recension of the False Decretals.

2) Either the forgers or their close associates expanded 39L with eighteen additional items, all but two or three from the Collectio Quesnelliana. The fifty-seven letter dossier is an important characteristic of the A1 recension of Pseudo-Isidore printed by Paul Hinschius. The greater part of the additions occur en bloc, at the beginning of the 39L from Stage 1.

3) In an anomalous subtype of A1, these eighteen additions are interspersed among the contents of 39L, while 39L is itself subjected to further correction and rearrangement (during the course of which one of the eighteen additions was withdrawn). The original exemplar of this subtype, complete with original scribal redactions, survives in New Haven, Beinecke Rare Book and Manuscript Library,Ms. 442 (N442). According to me, N442 is an early form of A1 that was originally copied without the supplements that were added to 39L in Stage 2 above. The extensive scribal redactions reflect an effort to bring the Leonine dossier in N442 up to date. After redaction, N442 gave rise to an entire subtype of the False Decretals that is known as the “Cluny Recension.” Despite this name we do not know where N442 was copied. We know only that the highly distinctive Leonine dossier in this manuscript experienced wide dissemination. The Leonine collection in N442 consists of 56 letters, and we will call it 56L. Very probably, 56L emerged sometime during the pontificate of Nicholas I (858-867), because the redactors of N442 contribute a papal list that concludes with Nicholas.

4) 56L was then rearranged and expanded with material from the Collectio Bobbiensis and other items, growing to a full 71 letters. This collection, 71L, circulated broadly in the High Middle Ages independently of the Pseudo-Isidorian decretals. Typically 71L concludes with an appendix of four additional pieces associated with a sixth-century collection of items on Chalcedon and the Three Chapters controversy that Eduard Schwartz christened the Collectio Sangermanensis. One of only two extant Sangermanensis mansucripts comes from Corbie; another comes from Reichenau and a third was known at Beauvais but has been lost. For Patzold, Sangermanensis material is the single most important indication of the Corbie origins of C. We have a lot more to say about the Sangermanensis in the next installment. For now, we need only note that this four-item appendix does not consist of Leo's letters, but rather a letter of Aurelius of Carthage and extracts from ep. 186 of Augustine. 

5) At some later stage, 71L and its Sangermanensis appendix received a further 26 additions. All but three of these additions come from the Collectio Grimanica, source of the Bobbiensis employed just above. This stage is a great mystery, as it is so far known to survive only in Merlin’s 1524 edition of the False Decretals.[3]

6) This expanded collection was then redacted a final time, yielding the Leo letters on hand in C. The architects of C, as we know it today, had before them A/B, and they had before them something like the 71L with Grimanica supplements in Stage 5. They adopted 39L from A/B outright. Afterwards they included all of the non-39L (that is to say, the non-Hispana) pieces from the 71L, altering the order in some cases. They either omitted or did not know the four-item appendix from the Sangermanensis. In its stead, they provided six additional Leo letters from the Collectio Quesnelliana and other sources. The sole decretal from the Collectio Corbeiensis, which for Patzold points to the Corbie provenance of the Leonine dossier in C, is to be found among these pieces. Finally, the architects of C added the greater part of the Grimanica material that was appended to 71L in Stage 5.

We have before us a many-layered phenomenon, and whether you buy Chavasse’s entire story or quibble with details (and I have smoothed over or omitted some important details, so please read Chavasse for yourself before you get any grand ideas), there is no disputing that the Leo letters in C reflect something more than a single campaign of expansion. 

What can it mean, from the perspective of the Leonnine dossier, to argue that "the core of the C recension" dates from the ninth century or that it was assembled by the Pseudo-Isidorians at Corbie? Must all the stages in this process be from the Pseudo-Isidorians at Corbie for the thesis to hold? Only the last? Here is the best that I can do: Stage 1 is Pseudo-Isidorian by definition, and its Corbie associations are proven by the extant manuscripts of the A/B recension. Stage 2 at least stands very near Pseudo-Isidore, and yet there are no demonstrable links between A1 and Corbie. Whether or not Stage 3 is Pseudo-Isidorian, it has nothing to do with Corbie, for neither the original copyists of N442 nor its many redactors can be associated with that abbey. In light of our earlier considerations about the extant manuscripts of the Bobbiensis and the Grimanica, the 71L of Stage 4 might seem Italian. Here the Sangermanensis appendix, with its Gallican and possibly Corbie-specific associations, is layered on top of the Italian material. These different associations persist through Stage 5 and Stage 6, where the Grimanica (source of the Bobbiensis and another collection with Italian associations) recurs alongside more items of Leo’s correspondence with Gallican associations, one of them from the Collectio Corbeiensis (Corbie again).

A related and probably more significant question would be how many different hands are stirring this soup. Stage 1 is Corbie and Stage 3 is not, so we are up to two at least. While the formal sources in play throughout these first three stages (the Hispana and the Quesnelliana) are typical of Pseudo-Isidore, from Stage 4 we are confronted with another world of sources that do not, as a rule, typify the Pseudo-Isidorian forgeries beyond C. Nobody has yet demonstrated that the authors of the False Decretals knew or used the Sangermanensis, the Bobbiensis or the Grimanica. Either the passage of time or the involvement of a third redactor would seem necessary to explain this shift. Relatedly, we have the use of the Bobbiensis in Stage 4, followed by the use of the source from which the Bobbiensis was derived, namely the Grimanica, in Stage 5. Another temporal interval or still another compiler would seem necessary to explain this progression. At Stage 6, we find the exclusion of some of the material assembled in Stage 4 and 5. Notably, the Sangermanensis appendix to 71L is set aside in favor of different material, including an item from the Corbeiensis. Possible Corbie associations are also reinforced with the return of 39L from the Hispana and/or A/B. Curiously, though Patzold argues that the Sangermanensis and the Corbeiensis both point to Corbie, we find them at odds in the Leonine dossier, with the Corbeiensis and its associates edging out the Sangermanensis appendix to 71L. It would seem reasonable to posit a new compiler at the level of Stage 6 to explain all of these changes, or once again a substantial temporal interval.

The Leonine dossier in C, then, probably reflects the work of more than two compilers, and perhaps also the passage of considerable time between Stage 4 and Stage 6. Because Stage 3 must postdate 858, I think it unsafe to ascribe the 102-letter collection of Leo letter’s in C to the ninth century--let alone the era of Pseudo-Isidore's activity--with any confidence. 

[1] So Benedikt Vollmann, Studien zum Priszillianismus (St. Ottilien, 1965), 106-7, who came to this conclusion in the course of editing one of Leo's letters that is present in both the Bobbiensis and the Grimanica. 

[2] Antoine Chavasse, "Les letters du pape Léon le Grand (440-461) dans l'Hispana et la collection dite des Fausses Décrétales," Revue de droit canonique 25 (1975), 28-39.

[3] This is as good a time as any to note that Merlin's editio princeps of the False Decretals, reprinted in vol. 130 of the Patrologia Latina, is generally said to have been printed from Paris, Bibliothèque de l'Assemblée Nationale, Ms. 27. Such goes, at any rate, the oft-repeated opinion of Hinschius, Decretales Pseuod-Isidorianae, lxxii-lxxiii. Yet this would seem impossible, for the Leo dossier in Ms. 27 is the standard 102-item collection described here as Stage 6 (Ms. 27 is, as far as I can tell, an ordinary C-recension manuscript), and Merlin seems to print an antecedent of this collection.