Towards a Theory of Pseudo-Isidore: VI

Previous parts linked in the side-bar here.

Six years ago, on the old blog, I started pounding out a series of posts under the rubric Towards a Theory of Pseudo-Isidore. What I was trying to do, with all that blathering about the interpolated Hispana and those rare forgeries that draw heavily upon it, was demarcate one small morsel of the Pseudo-Isidorian soup that I thought was indeed especially early. I wanted to do that because, at that point in 2013, I had just stumbled upon reasons to believe that everything else in the world of Pseudo-Isidore didn’t have the enticing early dates that Zechiel-Eckes had proposed. Indeed, the False Decretals were probably super late — every inch as late as Paul Hinschius had told us they were.

Like many things over there (and many things over here), I didn’t get as far as I wanted to — but in this case, not because I gave up. Instead, I decided the basic idea was important enough for an article. I was an assistant professor writing lecture after lecture on late medieval history and without a good library, so the article took a while. It finally appeared in 2017:

Ebo, Pseudo-Isidore, and the Date of the False Decretals, in: Speculum 92 (2017) p. 144–183.

You can read it, but here’s the potted version: The decretal forgeries of Isidorus Mercator strive in very specific ways to condemn procedures used against Archbishop Ebo of Reims at the 835 Council of Thionville. Some of these same tendencies are visible in other Pseudo-Isidorian compilations, like the False Capitularies of Benedictus Levita, but they are never as clear or pronounced or repeated as often as they are in the world of the pseudopopes. I also realized that four decretal forgeries relax ancient canonical prohibitions on episcopal translation (the movement of a bishop from one see to another see), in a way that is obviously designed to justify Ebo’s transfer to Hildesheim after 845. They also deplore bishops who occupy the sees of deposed bishops while those unfortunate deposed bishops are still living. These lines are obviously about Hincmar, who took office at Reims in 845 while Ebo was still alive.

For me, there is no alternative to accepting that the decretal forgeries took shape in the years between 845 and 851, when Hincmar was at Reims and Ebo at Hildesheim. After Ebo’s death on 20 March 851, there can have been no point in making these arguments. This dating, post-845 to 851, applies even to the decretal forgeries of the A2 recension, as in fact three of the four decretal forgeries that address translation occur within A2; and one of the codices with Pseudo-Isidorian notae discovered by Zechiel-Eckes even has excerpt marks around a passage that came to serve as an important source for Pseudo-Isidore’s arguments on episcopal translation and that is used in no other connection. The whole lot of it postdates 845 and predates 851.

I think these aspects of Pseudo-Isidore destroy all of Zechiel-Eckes’s broader theories. Skeptics somewhere might propose complex arguments to subvert the proofs that I have brought, but that would be a grave error. The work of placing Pseudo-Isidore’s arguments in their proper historical context, the long neglected project of scholars like Weizsäcker over 150 years ago, is far from complete and much has been forgotten. There are other subtle political tricks and traps lurking in the forgeries. I have since found one or two more. Cleverer scholars after me will surely uncover more, The further this work advances, the more adherents of early dates will have to fear, because the political backdrop will inevitably turn out to be that of the 840s or early 850s time and again.

I promise.

The forgeries surely have close ties to the library of Corbie — here Zechiel-Eckes will always be right. That’s why I’ve preceded these words with a long review (not yet complete!) of Firey, which is also in its way a defense of Zechiel-Eckes where he saw further than anybody before him had.

But I now think he was mistaken as to the broader view. Pseudo-Isidore was not Paschasius Radbertus, and the forgeries cannot be called an immediate reaction to the 835 Council of Thionville (a distant reaction to it: sure), and the False Decretals have nothing to do with political opposition to Louis the Pious.

At the end of that Speculum article, I tried to put all those reheated Wasserschleben notions asideand think about the origins of the False Decretals in a simple and direct way, given what I now consider to be the certainty that the traditional dating of Paul Hinschius, Emil Seckel and Horst Fuhrmann is right; and given, too, that we are now certain that Pseudo-Isidore’s library was at Corbie.

I suggested that the forgeries might have something to do with the clerics Ebo controversially ordained during his brief restoration to the see at Reims in 840/1, and that the earliest moments of their reception are bound up in the controversies surrounding the career of Wulfad, a favorite of Charles the Bald and eventual archbishop of Bourges. Vat. lat. 630, the manuscript whose text of the False Decretals I am slowly editing here, is closely associated with Wulfad and also comes from Corbie. I think here we are very near to the forger himself with this manuscript.

So anyway, I got that article in press and then I decided to write a book about the bigger problem of Pseudo-Isidore and who he was. This was going to get the provocative title Who Was Pseudo-Isidore? My idea was to write an introduction to the forgeries (so far no such thing exists in English with the exception of Fuhrmann’s helpful article) that could also be a receptacle for some of my crazy theories. I even wrote a good part of this book at the American Academy in Rome.

But as I wrote it I began to sour on the project. For one, I had the feeling I was working at cross-purposes to myself. It is hard to provide a neutral introduction to the Pseudo-Isidorian forgeries and the theories scholars have developed about them, while at the same time arguing a particular point of view. I was always treading the fine line of being untrue to myself or unfair to others. Furthermore, I realized that my own ideas on some crucial points required more work: not ordinary book research, but very hard collation edition research, the kind that takes years. And finally, I was writing chapter after chapter under a title that posed the one question that has been on everyone’s minds since 2000 at least — Who was Pseudo-Isidore? — and I was not certain about the answer, or even clear that trying to find one was the right exercise at the moment.

I did most of this writing in the Vatican library, and as I worked on this footnote and that one, I came to wonder about Pseudo-Isidore’s manuscript source for the writings of the truly dismal sixth-century poet and bishop of Pavia known as Magnus Felix Ennodius. Unlike a lot of really terrible authors, Ennodius never produced a bestseller. Indeed, for much of the Middle Ages he was almost totally unknown. Thus Pseudo-Isidore’s use of his works, almost exclusively in the decretal forgeries, has intrigued many people for a very long time. Since the nineteenth century, Ennodius’s editors realized the close connections between Pseudo-Isidore’s text of Ennodius and one of three ninth-century copies of his collected opera: Vatican, Bibl. apost. Vat. lat. 3803. This is a Corbie codex — Corbie, where Pseudo-Isidore worked! — that Bischoff dates to 850 or shortly thereafter.

The conventional view was always that Pseudo-Isidore used the archetype of Vat. lat. 3803. With one exception, the entire medieval tradition of Ennodius also descends from Vat. lat. 3803 in some way, so this view put Pseudo-Isidore literally at the head of nearly the entire medieval Ennodius tradition.

But, I thought: Isn’t that odd? How are to understand this little literary history? Now that we are free of the Radbert theory of Pseudo-Isidore’s identity, we no longer have to think of the forger as a Corbie monk: Maybe he just had a library card, or borrowing privileges, or maybe he was some kind of episcopal dissident hiding out at the abbey, or whatever. The point is, perhaps Pseudo-Isidore had his own copy of Ennodius, which he maybe brought from his own library, wherever that was, and in the course of his other research at Corbie the scriptorium there made a copy of it because they found it interesting. I imagined other scenarios too, but this seemed the simplest way to explain how it was that Pseudo-Isidore knew the archetype of the Corbie Ennodius , rather than as you would expect either the Corbie Ennodius itself or some copy thereof.

Then I noticed another thing. This is that Ennodius’s editors have really terrible reasons for supposing that Pseudo-Isidore’s appropriations from Ennodius are prior to Vat. lat. 3803. Ideally, you’d want them to show instances where Pseudo-Isidore has a better or less corrupt or somehow earlier text, and Vat. lat. 3803 has the later or worse or more corrupt text. All the editors could do, however, was give two variants that didn’t prove anything at all. These learned philologists weren’t fools, but they were laboring under a bad manuscript date: Everyone at the time thought Vat. lat. 3803 was a tenth-century codex, but the early Pseudo-Isidorian manuscript tradition dated from the ninth century. So Pseudo-Isidore had to be earlier! And thus they found themselves scraping around for reasons that Pseudo-Isidore’s Ennodius had an earlier text than the Vatican manuscript.

These considerations drove me to consult Vat. lat. 3803. When I did so, I found that this precise manuscript — and not any archetype — was Pseudo-Isidore’s source for Magnus Felix Ennodius. Its margins carried the notae of a Pseudo-Isidorian excerptor, very much like those that Zechiel-Eckes discovered in the three source codices he identified. But there were also some subtle differences from Zechiel-Eckes’s annotator: The system employed was much simpler, in most cases consisting of a single note rather than a series of signs. There was the same limited knowledge of Tironian shorthand, whereby only widely known abbreviations for words like hic and et appeared. Some of the notae in Vat. lat. 3803 had been erased; in other cases, passages appropriated by Pseudo-Isidore were not marked at all. Most striking, though, were a great many notae demarcating texts that Pseudo-Isidore never seems to have used, at least not in his extant works.

I wrote an article about all of this:

Pseudo-Isidore’s Ennodius, in: Deutsches Archiv 74 (2018) p. 1–52.

Among other things, I find it important that Pseudo-Isidore used a codex from the very middle of the ninth century. This is another obstacle to the Zechiel-Eckes hypothesis of A2 decretals dating from the later 830s, because passages from Ennodius are all over the decretal forgeries, in no few instances they correspond to marginal notae in Vat. lat. 3803, and this codex was clearly not copied in the 830s. It was produced in the graphically distinctive style that Bischoff and other paleographers have associated with the name Hadoard. The earliest codices copied in this distinctive style date the 840s. So here is more confirmation that the False Decretals are late through and through.

But Vat. lat. 3803 also confirms another thing, which I’ll write more about in Part VII.

 

Firey Review, Part II/2: More on Hamilton 132 and ‘Collaboration’

Introduction here. 

Part I here.

Part II/1 here.

Firey advances four arguments to convince us that Hamilton 132 is the product of collaboration. First, she finds problems with the traditional explanation that Caroline scribes in Hamilton 132 are reworking a Dionysio-Hadriana to look more like a Hispana:

The Hispana has a set of Gallican councils not in the Dionysiana, and also a substantial sequence of about fifteen Spanish councils. … If the Caroline scriptorium were updating or expanding a Dionysiana to make it more like an Hispana, it would be the Caroline scriptorium that added the Gallican and Spanish councils to the manuscript. The Gallican councils, however, are in “a-b” script, and the Spanish councils are in both Caroline and “a-b” script. What the Caroline scriptorium added were the sorts of supplementary texts…that often get appended to canon law codices to enhance their intertextual value (45-46).

Three pages prior to these remarks, Firey acknowledges that the Collectio Sancti Amandi was a collection of Gallican and Spanish councils assembled to supplement the Dionysio-Hadriana, available in at least two manuscripts besides Hamilton 132.[1] When Gallican councils in a-b script or Spanish councils in a-b script occur in Hamilton 132, it turns out that they are always from the Collectio Sancti Amandi. This observation is the basis of the standard characterization of Hamilton 132 as a Collectio Dionysio-Hadriana cum Collectione Sancti Amandi that has been expanded with Hispana material. Again, just to be totally clear: Because the Collectio Sancti Amandi contains Gallican and Hispanic councils, a-b scribes sometimes copy Gallican or Hispanic conciliar texts, which they have from the Collectio Sancti Amandi. The Caroline scribes do add “supplementary texts,” but they also add a substantial amount of canonical material from the interpolated Hispana.

Firey has a second point in favor of collaboration:

The evidence that seems to controvert the possibility that the Caroline portions were inserted in an independent, later process is in the pricking and ruling of the codex. The pricking and ruling seem to reflect the existence of two, separate scriptoria. The ‘a-b’ leaves are ruled in a particular way: four pricks and lines for the center boundaries of the columns; the Caroline leaves, although ruled for the same text space and also for the same forty-two lines per column, show a much simpler form in the column boundaries (single rather than double lines). The matching text spaces and number of lines do, however, suggest co-ordination. Again, the Caroline leaves could have been planned later specifically to match the codex that was to be updated. At several points, however, the Caroline scripts are on leaves that are pricked and ruled in the ‘a-b’ style. That suggests that some parchment was passed from the ‘a-b’ scriptorium at the time of the ‘a-b’ transcription to the Caroline scriptorium for use in the same codex…. (46-7)

In a footnote that sometimes confuses which leaves in Hamilton 132 have a-b and which have Caroline text, Firey explains that the “a-b” ruling is evident on Caroline folios 31, 33 through 49, 40 through 51, and 56 through 61. After considering explanations for this phenomenon that do not involve “collaboration,” Firey concludes:

Simultaneous production in two different, co-operating scriptoria seems more likely, and fits with Bishop’s analysis of other manuscripts showing similar combinations of “a-b” and Caroline script.

As for Bishop, we have seen that he explicitly rejected collaboration between the a-b and Caroline scriptoria at Corbie. But however that may be, there is nothing inherently “a-b” about double bounding lines, beyond their association with a-b scribes in this specific codex. With that point in mind, the worst we can do to Firey’s argument is put it in simpler terms: According to Firey, the a-b scribes always write on folios with double bounding lines. Also according to Firey, the Caroline scribes generally write on more simply ruled parchment, but sometimes their parchment is also ruled with double bounding lines. You may decide for yourself whether and to what degree traditional views of Hamilton 132 are refuted by this observation.

Firey has a third argument:

Finally, as evidence for apparently simultaneous production, fol. 44r has two corrections, which appear to be made at the same moment, to judge by the system of symbols for insertion of the correction, the size and pen-angle of the scripts, and the ink. One, however, is in Caroline minuscule, and one is in ‘a-b.’ Either they were made by a person who could switch script styles readily, or by two persons who may have passed the same quill back and fort at a final phase of proof-reading’ (48).

From Firey’s own reproduction of fol. 44r (fig. 1 at page 76), we can see that a Caroline scribe has supplied one omission in the upper margin, while an a-b scribe has supplied another in the lower margin. The tie signs are different, the module is different, and the pen-angle is different. As for the ink, the relative colors are hard to judge in black-and-white. Even if all these things were the same, though, would we be compelled to imagine script switching or quill passing? What has become of Firey’s skepticism in the face of other scholars’ argumentation? Here, it seems that the only thing we can say with certainty is that the Caroline corrector and the a-b corrector worked after the main text  on fol. 44r had been written.

Firey has an ultimate point:

Furthermore, in several instances, the Caroline hands transcribe material from the Dionysiana that is not in the HGA as represented in Vat. lat. 1341: at some point, the ‘a-b’ exemplar would seem to have been passed to the Caroline scriptorium for final revisions to Hamilton 132.

This is another observation that does not compel anybody to posit collaboration in the production of Hamilton 132. Otherwise, the phenomenon Firey notices has an explanation, but it does not lie in the putative circulation of exemplars between scriptoria. We will look at one instance in which Caroline scribes copy Dionysio-Hadriana text, highlighted by Firey in support of this final point, to explain how it came to be that Caroline scribes, adding Hispana material to a Dionysio-Hadriana manuscript, found themselves in the position, occasionally, of recopying Dionysio-Hadriana material. Here I am dependent upon Paul Hinschius’s study.[2]

At fol. 106, Caroline scribes copy J3 1182, a decretal of Simplicius from the Dionysio-Hadriana that is not present in the Hispana. Why do they do so? We must begin by recognizing, along with Firey, that Hamilton 132 is badly misbound. The folios that ought to precede 106 are therefore not the folios that actually precede it in the present binding; rather, the immediate sequence begins at fol. 96 and continues through fols. 97, 104, 105 and then finally 106. All but fol. 106 are a-b folios. On fol. 96 we have the beginning of the Hilarius decretals from the Dionysio-Hadriana (though these letters recur in the Hispana as well), and this sequence continues through fols. 97, 104 and 105. Yet the Hispana has more Hilarius letters than the Dionysio-Hadriana, and so at fol. 105 our Caroline redactors erase the beginning of the Simplicius dossier that originally followed Hilarius in their Dionysio-Hadriana, in order to create more room for Hispana supplements. These supplements have to go here if the chronological order is to be maintained and the Hilarius material is to be kept together, all in one block. The Caroline scribes then continue to add two pieces of Simplicius correspondence from the Hispana, directly after their Hispana supplements to the Hilarius decretals. Then, finally, they recopy the Simplicius decretal from the Dionysio-Hadriana, J3 1182, that they obliterated to make way for their expansions to Hilarius. They do this on fol. 106, which is tipped in.

The process of supplementing chronologically arranged canonical collections, in other words, occasionally required both cutting and pasting to maintain the pope-by-pope sequence of decretals. This implies nothing about collaboration, though in this case it does show that the Caroline scribes wished to retain Dionysio-Hadriana material alongside their supplementary Hispana texts.


[1] See Kéry, Canonical Collections of the Early Middle Ages (Washington DC, 1999), 84.

[2] Die kanonistischen Handschriften der Hamilton’schen Sammlung im Kupferstich-Kabinett des königlichen Museums zu Berlin, Zeitschrift für Kirchengeschichte 6 (1884), 193-246, at 228.