Why did Pseudo-Isidore interpolate the Hispana?

In 1885, Friedrich Maassen showed that the Collectio Hispana Gallica which serves as a vessel for the False Decretals was no ordinary version of this collection, but rather a special Pseudo-Isidorian recension, complete with inauthentic adulterations. He also showed that this adulterated Pseudo-Isidorian Hispana survives mostly or entirely separately from the decretal forgeries, in Vatican, Biblioteca Apostolica, Vat. lat. 1341. Maassen himself and others since have been fond of writing that Pseudo-Isidore interpolated the Hispana by way of preparing it to receive his decretal forgeries. I too am guilty of writing this way. The problem with this view is that it makes no sense.

First of all, we must ask ourselves: Why did the Hispana have to be prepared at all? If indeed its contents required revisions, these could have been carried out alongside the more major changes, that is to say the massive expansion of the authentic Hispana decretals with forgeries. Not only are we supposed to think that Pseudo-Isidore wasted time and parchment producing a fair copy of this lightly retouched collection; he or somebody else also took steps to circulate it. Only one complete copy of the interpolated Hispana survives today, but even if we confine ourselves to the ninth century we have evidence of two further copies (the initial folios of Vat. lat. 630 and the s. IX med. additions to Berlin, Hamilton 132). In total, six or seven medieval copies of the Pseudo-Isidorian Hispana have left traces.

Second, the interpolations to the Hispana do not really mirror the agenda of the decretals forger. Instead they address a variety of miscellaneous concerns, in much the same way as the False Capitularies of Benedictus Levita tend to do. The decretal forgeries stand apart from both in presenting far less of a miscellany, with a much tighter focus on procedural matters.

What can this mean? In a forthcoming article for Deutsches Archiv, I show with many tables and parallel columns and so forth why all three books of the False Capitularies, together with the first three Additiones, must predate the False Decretals. So we began to get an earlier picture, of miscellaneous concerns in the Hispana interpolations and in the capitularies; and a later picture, of increasing preoccupation with matters of procedural law.

In fact, and just as an aside, you can by and large determine the relative chronology of Pseudo-Isidorian items by their degree of procedural hysteria. Take for example the Capitula Angilramni: There are strong source-critical reasons to assign them to the same era as Book III of Benedictus Levita and just prior to the False Decretals. They focus exclusively on procedural protections for bishops. The same is true of Additio IV of Benedictus Levita, which was assembled very late, after drafts of at least some decretal forgeries had been drawn up.  Early items in the Pseudo-Isidorian library, however, like Benedictus Levita Book I and Book II, have some of the procedural material, but they balance it against various other concerns.

Third, further on the matter of chronology and its implications. It is plain to me that Pseudo-Isidore revised the Hispana at the earliest stages of his project, alongside Book I of Benedictus Levita. At one point early on in Book I, Benedictus is forced to resolve a textual problem with a decretal of Leo transmitted to him in corrupt form by the Hispana. The Hispana interpolator encounters the same corruption and applies a different, independent solution. Conclusion: Benedictus had before him the unredacted Hispana Gallica and not yet the Pseudo-Isidorian recension.  

By the time Benedict got to Book II, however, the interpolated Hispana was more or less finished, because Book II includes unmistakable excerpts from it. These excerpts are extensive, so much so that they allow some very precise conclusions about what the Benedictus Hispana looked like; it was slightly different from what we have. After Book II, the forgers still had to compile Book III of the False Capitularies, the most complex stretch of the capitulary forgeries by far, where procedural concerns come to the fore more and more. Then they had to assemble the Capitula Angilramni, and only then the decretal forgeries. These were then finally embedded in the Hispana.

The distance between the interpolation fo the Hispana and the forging of the decretals is thus very great. The former is not immediately prior or preparatory to the latter. 

In sum: Nobody has explained why Pseudo-Isidore needed to prepare the Hispana in the first place; the adulterated Hispana circulated on its own and was not merely an internal draft; the Hispana interpolations align with early Pseudo-Isidorian concerns and do not have the developed procedural hysteria of the False Decretals; and in fact the interpolated Hispana was produced well before the decretal forgeries.

There is a powerful case to made that minor items like the Capitula Angilramni were indeed preparatory to the False Decretals. The interpolated Hispana is totally different, and we are driven to ask why Pseudo-Isidore bothered with it at all.

Radbert’s Autograph?

If Radbert is Pseudo-Isidore, we ought to find similar secretarial practices underlying the acknowledged works of both. If we fail to find similarities, and instead find substantial differences, that would constitute another grave mark against this theory.

The ninth-century Corbie library is well preserved and a great part of it has been digitized. This has made it possible to study actual codices that the great theologian must have consulted personally in constructing his Matthew commentary.  I have now gone through several likely candidates. I have pondered many notae indeed, but I have not found a single one that has any plausible connection to Radbert.

I will confess that I sort of expected this result. Scholars have long talked about Radbert’s corrections or annotations, but never as far as I know in relationship to his sources. Instead, if Radbert is to be found anywhere, it is said to be in the earliest copies of his works, specifically in a Corbie codex of his Expositio in Matheo, where we find somebody making minor corrections and expansions here and there that have an authorial flavor to them.

Note that this is the precise inverse of the situation with Pseudo-Isidore. For our august forger, we find suspicious activity in the sources, but the earliest Corbie copies of Pseudoisidoriana betray no privileged access to the sources of the forgeries or the agenda of their fabricator at all. As for Radbert, meanwhile, we have author-tier redactions in a Corbie copy of his Expositio, but nothing that I can find or that I know of in any source codices.

The question of Radbert’s hand goes back to 1975, when T.A.M. Bishop gave a talk at Oxford discussing what he thought were autograph notes of Eriugena and Paschasius Radbertus. Twenty years later, Édouard Jeauneau and Paul Edward Dutton wrote a whole book on Eriugena’s hand. Meanwhile, rumors of a Radbert autograph kicked around here and there but as far as I know, were never more fully developed.

The autograph was supposedly to be found in Laon, Bibliothèque municipale Ms. 67: one of two ninth-century copies of the first four books of Radbert’s Matthew commentary. I think there was the expectation that Beda Paulus would look into the whole matter with his edition of the Expositio in Matheo. But, when CC Cont. med. 56 finally came out in 1984, Paulus could not have been more circumspect. He provided an extremely cautious description of corrections to Laon 67 that seems here and there almost at pains to avoid raising the question of their origins and authorship. From Paulus’s introductory discussion, however, and also from no little paging around in the edition and the digitized mircrofilm, I have gained some idea of why we might think Radbert worked in this codex:

In the beginning there was Paris, BnF ms. lat. 12296: a post-843 Corbie copy of in Matheo I-IV (it calls Radbert abbot of Corbie), which spent the Middle Ages at the monastery. Laon 67 was then copied from Paris lat. 12296. Like its parent, the Laon codex was produced at the monastic scriptorium, but it was intended for export to the Laon cathedral library. Radbert seems to have used the production of this second manuscript as an opportunity to lightly revise his commentary. Laon 67 thus received corrections, sometimes from the main hand and sometimes from others. A subset of these corrections must have been authorial, because at some later stage they were implemented in the exemplar, namely Paris lat. 12296: the local Corbie copy of Radbert’s commentary. They thus represent a general update to the first books of the Expositio. Importantly, there exists only one other codex of in Mattheo I-IV, copied from Paris lat. 12296 in the twelfth century. The absence of any other independent tradition means that the earlier readings, if the corrections to Lat. 12296 have obliterated them, are gone.

Now, I have not completed my study of this problem, and what follows is anyway only a very small sample of the evidence, which will perhaps become an article somewhere. So far, I am convinced that some of the corrections and expansions to Laon 67 surely reflect Radbert’s own changes. But it is not always clear to me that these changes were made with Radbert himself holding the pen. It seems to me that there are several different hands involved. Some of the corrections might even come from the primary copyist. Perhaps there is evidence somewhere in the codex tying one of these correctors to Radbert. In fact, Bishop must’ve thought he had such evidence, but right now I can’t imagine what it would be.

That is not really a problem for my purposes though. I am far less interested in Radbert’s handwriting than I am in the broader picture of secretarial activity surrounding an acknowledged Radbertian work from the era of the decretal forgeries. Remember: According to Zechiel-Eckes’s chronology, the first four books the Expositio predate Radbert’s activity as Pseudo-Isidore. Our earliest copies of the Expositio, however, both postdate Pseudo-Isidore, again according to Zechiel-Eckes. You could not ask for a more appropriate case study: the text and its manuscript witnesses seem to bracket the entire phenomenon of the False Decretals, as Zechiel-Eckes conceived of it.

We will start at the beginning, at an amusing erasure in Laon 67 at fol. 4v, also noted by Paulus (CC Cont. med. 56 p. xi):

Laon 67 4vOriginally, Radbert had promised to add marginal sigla to indicate his sources. Maintaining them in the proper position was however apparently too much even for a codex copied under Radbert’s supervision, and so the device was discarded in all extant manuscripts, including Laon 67. The corresponding discussion of these sigla in the preface has thus been erased.

An aside: I would even dare to posit this little story: That Radbert told his scribes he would add the sigla later himself, because knowing precisely where they ought to go in a new copy was a great problem. (The sigla, famous token simultaneously of Carolingian-era ambitions in the field of learning and the inadequacy of many of their methods, have a tendency to slide up and down the text through subsequent copies, as scribes can’t tell to which precise passage they are attached.) But, given the finished copy of Laon 67, Radbert recognized the enormity of the task before him, realized he had himself forgotten a lot about his sources, and with a great sigh rubbed out the naive promise he had made years before.

Now to the much more important matter and the primary point of comparison, namely tie signs. Remember that Pseudo-Isidore prefers a simple colon (:), with the upper dot a bit to the right of the lower one. 

Here, at fol. 34r, are some plausibly authorial additions to Laon 67, complete with tie signs:

67 34r

The colon does not appear. Instead, we get a common alternative, the dot and line (./). We find the same again at fol. 73r, where ut dixi is added in a manner that seems overtly authorial (as dixi speaks in Radbert’s voice):

Fertur autem \ut dixi/ haec stella sidereas vias non tenuisse…

67 73r autograph with tie sign also missing originally in C

Again, a dot and a line. But, do not despair, there is a colon somewhere here! The very same ut dixi is also added to the exemplar, Paris lat. 12296 fol. 51r, with the two dots:

12296 fol 51r ut dixi

I would think everybody can agree that this is not the same hand as in Laon 67.

Another tie sign in Laon 67 occurs at fol. 22r. Again the addition is plausibly authorial:

laon 67 22r

Again, a dot and a line. Of course a lot of marginal additions are simply corrections to scribal mistakes. When this happens the tie signs are often rather different, for example a line with two dots:

67 191v tie sign.png

Quod sicut was omitted by accident and is here supplied a little differently from the authorial changes above. The same sign occurs in several other cases where the addition is not Radbert’s revision but the result of scribal error.

A substantial revision to Laon 67, which must surely represent Radbert’s expansion of the text, comes at fol. 35r:

Laon 67 fol 35 recto

To understand what is happening here, see CC Cont. med. 56, p. 69, lines 258 to 260. Somebody is expanding here on the definitions of various names. He is doing so with the help of an anonymous Carolingian-era tract by a convert from Judaism who knew Hebrew that was once wrongly ascribed to Jerome, the Hebraicae questiones in libro Paralipomenon (Migne 23: 1365-1389: see col. 1371). The only place in the entire Expositio in Matheo where these Quaestiones are ever cited, is right here in this added text. So I will venture this hypothesis: While Radbert was reading over the freshly copied Laon 67, he came to a bit where he realized he had new information and he added it spontaneously in the margin. This would be a light indication that Laon 67 was copied after work on the entire Expositio had been completed, since this novel pseudopatristic source occurs nowhere else in the commentary.

Laon 67 fol 35 recto

This is a hand that has nothing in common with the glossator in Paris lat. 11611 that we looked at yesterday.

Of course, as we have learned to expect from authorial additions in Laon 67, the same text has been added to Paris lat. 12296:

Paris 12296 fol 24 verso marginal addition

The hands are different but they have more in common with each other than it seems at first. Here we are at the limits of the resolution of our images but behold this hasty comparison:

 

 

Other marginal signs occur here and there in Laon 67. Here for example is a recurring note, the significance of which I do not yet understand:

67 63v marginal sign.png

Finally, the primary copyist has his own repertoire of rather more elegant and visually appealing tie signs. Here is how he supplies an eyeskip at fol. 65r:

67 65r tie sign eyeskip omission.png

At fol. 69v a corrector supplies another omission like this:67 69v another tie sign


I will state it plainly: Nothing here looks anything like Pseudo-Isidore at all. And you do not even have to take my word for that. Some years ago now, Zechiel-Eckes himself surveyed Corbie manuscripts for signs of Pseudo-Isidorian activity. His census is online here. He likewise discovered nothing Pseudo-Isidorian in these codices.

But he should have, right? Because that would seem to be what the Radbert theory predicts.

There are a lot of ways to rationalize this problem: Radbert used different secretarial teams as a forger and as an exegete (but in both cases we are talking about the scribal resources available to Radbert as abbot); it is an apples-to-oranges comparison, contrasting annotations in source codices with redactional activity in drafts (but tie signs are tie signs); the Pseudo-Isidorian forgeries are earlier and this is later (but not really, and at the extremes not nearly enough to expect a new generation of scribes or new practices in the scriptorium).

The differences between the authorial personae of Pseudo-Isidore and Radbert are reflected perfectly in the totally different literary practices associated with their respective works. Radbert is not Pseudo-Isidore, and everywhere you look you find new ways to show it.

Pseudo-Isidore’s Autograph?

To date we know of at least four manuscripts that Pseudo-Isidore used directly:

St. Petersburg, Russian National Library, F.v.I.11: Copied at Corbie and kept there throughout the Middle Ages, a famous example of a-b minuscule.

Paris, Bibliothèque nationale de France, Ms. lat. 11611: Copied at Saint-Denis; later booklists place it at Corbie. According to Zechiel-Eckes it came to Corbie in the 830s at the latest; as evidence of this he cites Maurdramnus-type annotations (see idem, Fälschung als Mittel politischer Auseinandersetzung, p. 10).

Vatican, Biblioteca Apostolica, Pal. lat. 1719: origins perhaps in Western Germany? No known associations with Corbie.

Vatican, Biblioteca Apostolica, Vat. lat. 3803: a Hadoard-style Corbie codex from the mid-ninth-century.

All four codices carry a series of subtle marginal notae that attest to Pseudo-Isidore’s interest in specific passages. These notated passages very often recur in the forgeries. In rare cases the annotations direct the copying of specific passages, or appear to anticipate some of Pseudo-Isidore’s falsifications. Thus in these codices we can observe the autograph of a person directly affiliated with the forgery enterprise. In my work I have spent hours and hours on the textual and source-critical problems of these notae. Here, I want to explore their graphical form.

Behold a typical cluster of notae between columns in the St. Petersburg codex (fol. 63r: reproduced from Zechiel-Eckes’s Francia 2001 article: Abb. 5 after p. 68):

petersburg fol 63r.jpg

Even with the poor quality of the reproduction you can see that all four notae were probably not added at the same time. Instead, it looks like the more graphically confident nota hic came first. Then beneath it you have the more cryptic signs, which it seems plausible arrived later on. This phenomenon recurs often enough to be a pattern. Frequently the dotted n stands alone, but in other cases it looks a lot like somebody else’s notes on interesting passages attracted more specific annotations, perhaps indicating their nature or significance to Pseudo-Isidore’s program.

vat. lat. 3803 fol 31r

Of course no pattern is totally consistent. This is from Vat. lat. 3803, fol. 31r (forgive here and elsewhere the watermark of the Biblioteca Apostolica). The dotted n occurs as a constellation of two or three signs (a q over a t precedes it).

This is as good a time as any to point out that a distinct notational personality prevails in each of the codices. Combinations of multiple signs recur in the Petersburg codex, while in Vat. lat. 3803 this is the only instance of notae in constellation. A lot of the marks in the St. Petersburg codex are fairly subtle and there are no glosses on the content. Again the annotator of Vat. lat. 3803 is even more restrained. Some of his notae have been erased and most of them occur only singly.

Paris lat. 11611 is very different. There the notae are much clearer and very plentiful: much more so even than in St. Petersburg F.v.I.11. The one fairly clear rule — namely that Pseudo-Isidorian interpolations and falsifications are nowhere to be found in the source codices, but were added at some later stage — comes very near to being broken in the case of Paris lat. 11611. At two points, apparently Pseudo-Isidorian glosses are added directly to its margins; both of them generalize from specific discussions about the competence of the bishop of Rome. The Pseudo-Isidorian forgeries then take up the annotated passages from Chalcedon, complete with interpolations that seem to be based on the glosses.

Why is the annotator less restrained in Paris lat. 11611?

One theory –it is not necessarily mine — could be that the margins of this codex, a copy of the Rusticus version of the acta of Chalcedon, are already full of Rusticus’s annotations to specific passages that the main scribe has carried over from the exemplar. The Pseudo-Isidorian annotator perhaps felt he could add his own notes here and there without standing out too much.

Still, even here the annotator is not totally confident. Here is what looks like a note to a copyist, usque hic, which could not be fainter or more retiring (fol. 20r):

11611 fol 20r

Beneath it you see another dotted n.

Here (fol. 154v) we have the first of the apparently Pseudo-Isidorian glosses:

11611 fol 154v

In the center of the image you see one of the Rusticus annotations in the hand of the main copyist. You can see how it is like a tree under which the Pseudo-Isidorian activity shelters. To the left of the annotation, you have another of these dotted-n moments. And under all of that, in rather different script, you have this gloss:

(quo)d missi apostolici primo s(ubscripserunt).

As I said, the gloss is interesting because it anticipates Pseudo-Isidore’s falsification of the same passage from Chalcedon. Highly unusual in the Pseudo-Isidorian context is the Tironian note for subscripserunt. Usually the annotator avoids Tironian abbreviations altogether, except generally those for hic and et. This one is of course not that obscure but this is still an atypical moment.

Here is the second seemingly Pseudo-Isidorian gloss in Paris lat. 11611 (fol. 176r):

11611 fol 176r

The pattern is the same, namely a phrase inc. quod… that anticipates how Pseudo-Isidore will falsify the associated passage:

Q. quod apostolicae sedis missi prius semper debeant iudicare. 

Finally, there is the single marginal note in Pal. lat. 1719 (fol. 48r), another direction to a copyist:

pal lat 1719 48r wide

The annotator’s dark ink makes it easy to recognize him. The two dots mark the point of interest in the text (Ut sicut…) and tie to the marginal note: hic usque in finem. Probably the same person has also added the point after caritas.

Now more detail:

pal lat 1719 48r detail.png

Some meager conclusions:

Of four known annotated codices, one (Pal. lat. 1719) has no obvious association with Corbie; the other (Paris lat. 11611) indeed came to Corbie eventually but was copied at Saint-Denis according to paleographers. It is precisely both of these codices that have received the most confident annotations.

The dotted-n-annotator highlights many passages of interest to Pseudo-Isidore, but his marks often seems graphically distinct from the more clearly Pseudo-Isidorian annotations.

There appears to be no obstacle to identifying the hand that glossed Paris lat. 11611 with the hand that annotated Pal. lat. 1719.

And some questions:

Typing all of this up convinces me once again that Paris lat. 11611 is enormously important for the abundant and anomalous nature of its annotations and the curiosities of its provenance.

Excerpts from this codex primarily feed the so-called Nonnullae sanctiones, a minor Pseudo-Isidorian forgery that circulated exclusively as an appendix to the False Decretals.

How did this codex get to Corbie? When did that happen? Is it all that absurd to suggest that Pseudo-Isidore brought it there? Who is the “Maurdramnus-type” annotator and what is he doing?

One of these questions might even turn out to be answerable.


This post lightly revised after I realized that the gloss in Paris lat. 11611 fol. 154v must begin with quod (partially illegible in the gutter), not a d as I had first thought. Schwartz (ACO 2.3.2 p. 156: apparatus) transcribes incorrectly: d(efininiunt) missi apostolici primo, so I had the in my head somehow. 

Paschasius Radbertus was not Pseudo-Isidore

Clara Harder has written a new article: “Pseudo-Isidorus Mercator,” in: Great Christian Jurists and Legal Collections in the First Millennium, ed. Philip L. Reynolds (Cambridge, 2019),  397–412. Everybody should read it.

I did, and I found the following remarks, at 398f. (please see the original for the citations, which I haven’t reproduced):

There has also been controversy as to when the work of forgery began and what were the forgers’ intentions. Some argue that there had been general dissatisfaction with the rule of Louis the Pious among the Carolingian clergy since the late 820s. The unjust dismissal of bishops, in particular, was already a topic of discussion. Work on the forgery, therefore, could have begun in the early 830s. The decisive moment might even have been in 830, when, after a brief uprising against him, Louis the Pious seems to have deposed bishop Jesse of Amiens and exiled abbot Wala of Corbie. Others have suggested recently that the nineteenth-century scholars Hinschius and Seckel were correct in assuming that the forgery was not carried out before the late 840s. According to this view, archbishop Ebo of Reims was a central figure in work on the forgery.

The crucial consideration pertains to the forger’s intentions. One can find textual support for several different theories by selectively analyzing the vast amount of material contained in Pseudo-Isidore’s collection. But any theory about the forger’s identity has to take account not only of the content of the collection, but also of additional evidence acquired through detailed textual and comparative analysis. Zechiel-Eckes showed how one could do this by compiling all the points of correspondence between the forgery and Paschasius’s known writings. Although Zechiel-Eckes could not definitively demonstrate that Radbert of Corbie was the guiding spirit of the forgery, therefore, he makes a very strong case. The involvement of this monk in the Pseudo-Isidorian project seems at least to be very likely, and to date no one has convincingly refuted the thesis or proposed a more convincing one.

In Kuhnian terms, the little world of Pseudo-Isidorian research is between paradigms right now. This is why it is not stable and to informed observers the research appears “highly technical and extremely controversial.”  (Well, at least it explains the controversial part: the technical problems are unfortunately down to Pseudo-Isidore himself.) What happened is that the question of Pseudo-Isidore’s origins, at the very end of the 20th century, had become a dead field. The Hinschius/ Seckel/ Fuhrmann framework for understanding the forgeries had no more active supporters because the entire question of who Pseudo-Isidore was and when he worked had no active scholars. This paradigm collapsed as Zechiel-Eckes revived the field with his quite different theories. Early on it seemed that these theories would become the new paradigm, but they have not held, and so we are where we are now.

Harder is a generous scholar and so she writes implicitly. But I will answer the charge of “selectively analyzing” as if it were directed at me.

From the beginning, Zechiel-Eckes developed his political theory of the forgeries on the basis of the relationship between procedural and related provisions in the False Decretals and what we know of Louis the Pious’s legal retaliation against his episcopal political opponents at the 835 Council of Thionville.

The False Decretals, or some core of them (the A2 recension) as a direct reaction to Thionville 835: this is Zechiel-Eckes’s thesis.

To explore how this thesis relates to the content of the False Decretals in a systematic (rather than a selective) way, I focused on Ebo specifically because most of what we know about Thionville is about his condemnation and deposition. It is thus the parameters of Zechiel-Eckes’s political hypothesis of the forgeries that determined my focus on Ebo. I pulled everything out of the decretal forgeries that could conceivably relate to Ebo and his problems, but then I also gathered everything I could find about episcopal trials in the Carolingian kingdoms between 814 (Louis the Pious’s accession) and 858 (when the False Decretals were surely in circulation). The point was to find circumstances attested only for Ebo that the False Decretals also addressed.

The result of this exercise was the Speculum article that I published. The exceptio spolii, an isolated passage in Pseudo-Alexander (J3 †49) on forced confessions, and four false decretals addressing episcopal translation: This was my harvest of things in Pseudo-Isidore’s program that are almost assuredly about Ebo in particular because there is not really anyone else in the historical record for them to be about.

This project yielded unexpected information about the post-845 date of the decretal forgeries. Naturally only some passages in the False Decretals address episcopal translation, in fact relatively few of them. But the thing about termini post quem or termini ante quem is that how selective the passages chosen to indicate them might be, does not matter. One textual moment can be enough to date an entire corpus. And the other thing about passages addressing Ebo in the False Decretals is that, selective or not, they help us determine Pseudo-Isidore’s ecclesiastical and political allegiances, which was my only point. The suggestion that Pseudo-Isidore has some connection with Ebo’s associates follows not from this internal reading, but from other points of evidence: in part, from Ebo’s own writings, which seem to demonstrate an advance knowledge of the forgeries; and in part also from the post-845 date itself, together with the circumstances of the earliest reception. I will elaborate more on this issue later, as it is a bit involved.

Anyway, that is how I would respond were anyone to suggest I have achieved the post-845 date and the association of the False Decretals with the Ebo clerics on the basis of a selective reading.

Now to the theory of Radbert-as-Pseudo-Isidore, which as the above quotation shows still has its defenders:

In the Speculum article, I tried to pen mostly a positive argument in favor of a different approach. I made my disagreement with the Radbert theory clear but I tried to avoid a comprehensive confrontation with it, because it seemed in poor taste to do so. As I wrote my more recent piece on Ennodius, the subject matter forced me to be a little more direct, but still I tried to play it down, addressing Radbert’s purported authorship of the False Decretals only insofar as it concerned Vat. lat. 3803, and only in the final pages of the article.

But Harder writes that there is “a very strong case” for the Radbert theory and also that “no one has convincingly refuted the thesis.” So, I will advance a much more direct argument here, and one that turns on three points.

I. The Radbert Thesis Is an Artifact of Zechiel-Eckes’s Impossible Dating Scheme

This is the logic of it: The False Decretals are brimming with material to the advantage of the episcopate. So people always thought the forger was some kind of episcopal actor. Zechiel-Eckes found that the forgers did some research at the Corbie library, and he decided that they had to be a monastic forgery. This in my view is where he went wrong; he had evidence placing the forgery at Corbie, which was not really the same as evidence identifying the forgery with the monastic institution of Corbie and hence its abbot.  However that may be, though, he then needed an explanation for the episcopal material in the False Decretals, because “monastic forgery at Corbie” meant “Abbot Radbert of Corbie is Pseudo-Isidore”, and yet Radbert was only ever an abbot and a deacon. Zechiel-Eckes provided a political explanation, glossing the apparently pro-episcopal material in Pseudo-Isidore as in fact a defense of Radbert’s political allies, namely bishops who had opposed Louis the Pious in 833 and who lost their sees at Thionville 835; and of course an attack from the perspective of the “Unity Party” (membership in which Zechiel-Eckes granted Radbert) upon Louis the Pious himself.

But, the False Decretals postdate Ebo’s transfer to Hildesheim in 845. They just do. And so if you continue to subscribe to the Radbert thesis, your explanation for the episcopal arguments that Pseudo-Isidore makes is broken. And, just to be clear: At the level of the False Decretals, the episcopal arguments are almost all of the arguments that Pseudo-Isidore makes.

II. The Radbert Thesis Was Never Fully Developed, Let Alone Proven

Harder writes that “any theory about the forger’s identity has to take account […] of […] evidence acquired through detailed textual and comparative analysis. Zechiel-Eckes showed how one could do this by compiling all the points of correspondence between the forgery and Paschasius’s known writings.”

If Zechiel-Eckes ever undertook such a comprehensive comparison, he never published the results. And so we are not obliged to refute anything.

Lest you think I am exaggerating, we turn to Zechiel-Eckes’s last statement on this matter[1], where he argued that Radbert was Pseudo-Isidore as follows:

1) Pseudo-Isidore and Radbert were both learned; 2) Pseudo-Isidore presumably worked on his forgeries between from the later 830s and this aligns with a caesura in Radbert’s literary activity from 831 and 849; 3) both authors know the rare works of Ennodius of Pavia; 4) Radbert at one moment in the second book of his Epitaphium Arsenii mentions giving Gregory IV a legal collection that proved the universal competence of Gregory IV to judge everyone, such that he was to be judged by nobody, and Pseudo-Isidore also has expansive views of papal jurisdiction; 5) Radbert forged a brief pseudopatristic tract inc. Cogitis me in the name of Jerome, and indeed “nach demselben Strickmuster” (p. 18) as Pseudo-Isidore, that is to say with the help of authentic sources that postdated the putative author (Jerome) and were therefore anachronistic; 6) and finally it would seem hard to believe that this great forgery enterprise could have operated at Corbie without Radbert’s knowledge.

Now these are certainly not all the arguments that Zechiel-Eckes made about Radbert’s authorship. (This is a blog post, I do not have to cite everything.) They are, however, the arguments that Zechiel-Eckes believed were sufficient to demonstrate Radbert’s authorship, and in fact they are an entirely fair representation of the case Zechiel-Eckes tended to make for Radbert-as-Pseudo-Isidore, with minor variations here and there.

As I think any neutral reader can appreciate, if you have a priori reasons for believing in Radbert’s authorship, the above will seem to be promising points of departure for further research; but if you are not convinced of Radbert’s authorship, or even perhaps agnostic, none of this has any force at all.

As for 1), I think that Pseudo-Isidore’s learning has been in many respects exaggerated. Work on the sources of the False Decretals, which has occupied me intensely this past year, has convinced me that it is a mistake to make too much of this. Also it is possible for two learned people to be in the same place, and at an august monastic institution like Corbie, even probable that such a thing would happen.

As for 2), there is no proof at all that work on the forgeries began as early as 831, so this neat alignment with the gap in Radbert’s known activity is a product of the assumptions some have made about Pseudo-Isidore’s dates, and not of any evidence. As for circulation after ca. 850, well, there is a lot more happening around this time than Radbert’s retirement: Ebo’s death at Hildesheim and Hincmar’s preparations for Soissons 853 are the two big Other Things. And then it is also the case that dates can just as easily be read against the Radbert thesis as for it: One could argue, for example, that the forgeries only begin to circulate after Radbert’s retirement ca. 850, because before then the monastic scriptorium (responsible for early post-850 copies of Pseudo-Isidoriana like Vat. lat. 1341 and Vat. lat. 630) had been under Radbert’s oversight as abbot and was thus closed to the forgers. That is of course not my argument: it is just an illustration that conclusions drawn about Radbert’s dates and the chronology of the forgery enterprise are conditioned by prior assumptions.

As for 3), I deal with this in my article on Ennodius. The reason Radbert knows Ennodius is that Vat. lat. 3803 was at Corbie when he was writing the Matthew commentary. Pseudo-Isidore was also at Corbie and also knew Vat. lat. 3803. Radbert cites just one line from Ennodius from the first folio recto of Vat. lat. 3803. Alas he does so without the distinctive marginal notae we have learned to associate with Pseudo-Isidore. Pseudo-Isidore demonstrates a far more extensive knowledge.

As for 4), and the whole matter of the libellus that Radbert claims he and Wala gave to Gregory IV in 833: This is bound up with arguments I made a long time ago and have since partially retracted (see the link if you want to know what is going on here). Harder repeats this argument several times in her book, but a) you could buy the whole story and even my old idea that the libellus was some proto-Pseudo-Isidorian forgery (as Zechiel-Eckes also seems to suggest at one point) and still we are only as far as point 3), namely that Radbert and Pseudo-Isidore shared a library. Otherwise, b), an expansive vision of papal jurisdiction by itself does not prove Pseudo-Isidore is at issue, which is why even in my old work I tried to bring the Gregory IV decretal into the mix as additional evidence that could bear on the problem of the libellus. Expanded papal jurisdiction may be in the 830s an uncommon thing to insist on, but the idea can be derived from widely circulated decretal texts like those in the Dionysio-Hadriana and other items like canons from the Council of Sardica. I don’t know what else to say: A shared theme, one widely available in authentic sources, is no way to argue for authorship.

On 5), there is nothing more likely to destroy belief in the Radbert thesis than to actually read Cogitis me. It is ed. Albert Ripberger, CC Cont. med. 56C, from p. 97. Other than the fact that it is a pseudonymous letter, it has nothing in common with Pseudo-Isidore at all. The source base is almost totally different. Many of the authors that Radbert draws on are nowhere in the False Decretals. It is a pseudepigraphical invention defending western devotion to Mary, a theme that could not be more remote from the Pseudo-Isidorian forgeries. And I type these next words a bit reluctantly, because I don’t want to be impolite. But as Harder has written of a “detailed textual and comparative analysis”, an endeavor in which she says Zechiel-Eckes has shown us the way, it is necessary to say this: Arguments like this one, but also points 1), 2) and 3) above too, are not the result of close textual investigations, but rather of frankly superficial comparisons. Radbert forged, Pseudo-Isidore forged; Radbert used authentic and anachronistic sources, Pseudo-Isidore used authentic and anachronistic sources. These kinds of arguments are a grave weakness and they are most of what the Radbert thesis is made of. In other articles, for example, Zechiel-Eckes referred to Radbert’s great Matthew commentary, and argued that this implicated him because Pseudo-Isidore cites Matthew more than any other book of the Bible. (But why is that?

And this brings us to 6), which is a possibility I am anxious to consider: Perhaps Radbert did know about the forgery enterprise at least in some capacity. Other people who were not Pseudo-Isidore surely had some idea of what was going on as well. This is a very different thing from arguing that Radbert was Pseudo-Isidore, but the whole question also turns on speculation. Nobody has direct insight into how scholarship and the library and the scriptorium were administered and what it was possible to do at a large institution like Corbie, for how long, with or without the abbot’s approval, and with or without his coming to find out about it.

We should not, however, assume that the forgeries were common knowledge at Corbie or that they represent anything like a collective enterprise of the monastery. A curious thing about early manuscripts of the forgeries copied at Corbie (Vat. lat. 1341, Vat. lat. 630) is how the scribes responsible for them nowhere demonstrate special knowledge of the nature of the forgeries. Underlying sources are never brought in to correct textual defects, for example; and in Vat. lat. 630 sometimes the scribes seem puzzled about the proper order of what they’re copying and where specific pieces ought to go. Then there are the extremely subtle marginal annotations in the source manuscripts we have discovered, very much in contrast to the bold and confident excerpt marks left by other Corbie scholars like Ratramnus. In Vat. lat. 3803 a few of the excerpt signs have even been erased.

III. The Radbert Thesis Has Led Nowhere

Now I come to the part of the argument that I have generally tried not to make, because it seems the least polite of all, but it is also in my view the most powerful reason to put all of this talk about Radbert aside.

The Radbert thesis has not proved helpful in reading the forgeries or in editing them or in understanding their early history or reception or anything like that. Nor, conversely, has the Radbert thesis proven helpful in reading the other acknowledged works of Radbert or in understanding their perspective or sources or purpose. In fact it looks like scholars of Radbert reject the idea that he had anything to do with the Pseudo-Isidorians.

Now of course scholars could be wrong. But the theory that Radbert steered the Pseudo-Isidorian enterprise is now old enough to drive a car, having been outlined in detail for the first time in 2001. And in all of that time, how has it advanced? The answer is that it has not advanced at all. No new proofs beyond the cursory suggestions of Zechiel-Eckes in his articles have been adduced for it. Worse still, adherents of this theory have not really behaved like they believe in it. I don’t mean to accuse others of insincerity, but it has always struck me as strange that Zechiel-Eckes thought Pseudo-Isidore and Radbert were in some sense the same person and that for a while many people including myself believed this, and some still do, but that their work on Pseudo-Isidore never once spilled over into study of Radbert. Years ago in 2010, when I thought Zechiel-Eckes was right, I even asked myself why I wasn’t doing more with Radbert’s acknowledged works. I sat down to read the great Matthew commentary and after returning to it on and off for three years, I learned why: The Matthew commentary has nothing to do with Pseudo-Isidore.

What would be most welcome is any study at all that compared the prose style of Paschasius Radbertus or the broad pool of mostly theological sources that Radbertus draws on in his work to the Pseudo-Isidorian forgeries. Rather than saying Radbert commented on Matthew and Pseudo-Isidore cites Matthew a lot, somebody might attempt the apples-to-apples comparison: Does Radbert also cite Matthew a lot when he is not commenting on Matthew? Or OK, perhaps we Pseudo-Isidorians are all legal scholars and uncomfortable in the Radbertian theological sphere. I accept that. So why not try to do the next best thing? Why not explain anything in Pseudo-Isidore (beyond a broad political orientation) according to Radbert’s concerns and circumstances and his favorite themes or authors in his other works? I do not mean the rough comparisons outlined above, I mean real work. If the Radbert theory were worth anything analytically, it should have proven fruitful by now. We ought to be smacking our heads and saying: What an excellent theory, it has helped us explain Pseudo-Isidore’s whole puzzling thing about chorbishops/clerical oaths/the consecration of Chrism on Maundy Thursday/whatever. But nobody has done anything like that at all. And in fact just typing out this careless list of puzzling things that come to mind shows again how far Pseudo-Isidore is from Radbert, and why nobody has undertaken this work, and why they never will.

The Radbert theory does not, like good productive theories, suffuse our texts and help us explain them. Adherents of this theory (as I was myself) read Pseudo-Isidore on his own terms and then in totally separate connections genuflect to Radbert’s involvement at semi-regular intervals.

Just look at the research: Zechiel-Eckes’s theories might have other problems, but they are no less valid if we take Radbert’s name out of them and replace it with “unknown monk(s) at Corbie” or the like. The same is true of Clara Harder’s publications, which insist on Radbert-as-Pseudo-Isidore but which also advance wide-ranging arguments about the forgeries that stand or fall on their own without need of the Radbert thesis at all. In fact, assuming a refuted Radbert thesis, I have trouble finding anything in Harder’s publications that is undermined, except her case for the Radbert thesis itself. And in looking over my old article on Gregory IV and Divinis praeceptis, which was even inspired by my belief in Radbert-as-Pseudo-Isidore, I am struck by how all of it works just the same, or does not work just the same, save for a few details, if I get rid of the Radbert thesis.

That the theory sits apart from work on Pseudo-Isidore and not within it is a powerful demonstration that it does not go anywhere, and it does not go anywhere because it is not true.

[1] Fälschung als Mittel Politischer Auseinandersetzung, Nordrhein-Westfälische Akademie der Wissenschaften und der Künste, 2011, esp. at pp. 17–19.


For further reasons to disassociate Radbert and Pseudo-Isidore, see this brief investigation of secretarial activity associated with Pseudo-Isidore, including what may well be Pseudo-Isidore’s autograph. The literary and intellectual processes illustrated there do not compare favorably to secretarial activity associated with Radbert, including perhaps an example of Radbert’s autograph.

Firey Review, Part III: Hamilton 132 and the 829 Council of Paris

Intro here; Part I; Part II/1; Part II/2

Firey concludes her remarks on Hamilton 132 by positing some connection between this manuscript and the great reform council that convened at Paris in 829. She writes that “…perhaps the most interesting question about the testimony of Hamilton 132 with respect to its historical context is that it may reveal something about the purposes for production of canon law codices in the early ninth century in the Paris basin.”

The somewhat unusual…texts in Hamilton 132…are attested in the Council of Paris. … The Pseudo-Leo text on chorepiscopi (JK †551; fol. 95v)[1] is reproduced in the Council of Paris as cap. 37. The insertion of this text, transcribed in a Caroline hand, occasioned more disruption in Hamilton 132. The leaves in this portion of the manuscript are prepared differently from others, and could have been added even later than the other Caroline leaves, although the continuation of the text [onto pre-existing a-b folios] shows the same method of integration as other Caroline-inscribed leaves. The Pseudo-Leo decretal is replicated in Paris, Bibliothèque nationale, lat. 3838, where it is also attributed to Leo, on a flyleaf… It appears to be written by the same scribe as that who wrote it in Hamilton 132. In Paris 3838, the first three lines of the decretal are annotated with later interlinear neumes and, above the second column of the text, is an annotation Fiat, fiat, both suggesting that the Paris copy was perhaps later used for public acclamation, as at a council. …

First, a very minor point: The Fiat fiat thing in Paris lat. 3838 (it is actually fiat fiant) is a pen trial; it is followed by Tironian notes for qua or quia and in the middle of the page between columns the same doodler has written DEUS in adiut t t t. You can go to the photos of Paris lat. 3838 at Gallica yourself and see. I don’t see any reason to read these marks in light of Pseudo-Leo.

Otherwise, to further our understanding of Hamilton 132, the relative chronology of its materials, and its importance for the history of Carolingian church councils, I must point out that Firey is mistaken on several levels:

1) Paris 829 legislates on chorbishops at c. 27, not c. 37. Whatever relationship c. 27 might have to Pseudo-Leo, JK †551/J3 †1118 is as yet undetermined. Both Paris 829 and Pseudo-Leo make what you might call related arguments with reference to the same (genuine) patristic-era canons, but no decretal forgeries occur in the acta of Paris 829. If they did, our entire estimation of both the Pseudo-Isidorian problem and Paris 829 would be much different.

2) Right below Pseudo-Leo in Hamilton 132, we find our Caroline scribe adding not Paris 829, c. 27, on chorbishops, but instead the related c. 9 from the Relatio episcoporum of 829. The distinction is a minor one, of course, as this is the summary of c. 27 that got sent to Louis the Pious. And the upshot is the same either way, namely that this particular scribe must have copied after the promulgation of the Paris acta in 829. This scribe also goes on to add (from the Dionysio-Hadriana) the first part of Antioch, c. 10; then Ancyra, c. 12; Laodicea, c. 57; and then the final part of Antioch, c. 10. What we are looking at here is a miniature dossier to the disadvantage of chorbishops. The exact same dossier also occurs as an addition to Paris, BnF Ms. lat. 1453, so we now have a total of three manuscripts with this same sequence of texts: Hamilton 132, Paris lat. 3838 and Paris lat. 1453. In all three cases we are dealing with a manuscript of the Dionysio-Hadriana; in Hamilton 132 and Paris lat. 1453, the dossier is clearly a later addition, and in Paris lat. 3838 it occurs on a flyleaf.

I am not sure what all of this does for Firey’s arguments about the relationship between Hamilton 132 and Paris 829, but it might imply that she has gotten the chronology wrong when she wonders (as at points she seems to) whether the activity we witness in Hamilton 132 was in some way preparatory for the Paris council. I would think that the anti-chorbishop dossier in all three manuscripts postdates Paris 829, perhaps substantially; and also Hamilton 132 is merely one of three witnesses to these texts, so any relationship between this chorbishop dossier and this or that council is not specific to Hamilton 132.

3) If the Caroline scribe adding Pseudo-Leo is writing after Paris 829, what are we to make of Firey’s thesis of collaboration, given that paleographers date the a-b folios of Hamilton 132 to the years around 800 (Ganz, 50; Bischoff, 74; the CLA at VIII, no. 1047)? This consideration would seem to be why Firey allows that the Pseudo-Leo decretal might have been added to Hamilton 132 even later than the rest of the Caroline material. But how much later can it have been added, if this Caroline folio “shows the same method of integration as other Caroline-inscribed leaves”? Three decades later? Five?

4) Five decades would seem around the minimum necessary for Firey’s scenario to work, since Paris 3838 was copied in the third quarter of the ninth century according to Hubert Mordek[2], and if the scribe copying Pseudo-Leo in Hamilton 132 and Paris 3838 is indeed the same (more on that later, perhaps), we seem to be exactly where Bischoff said we were vis-a-vis the dates for the Caroline material in Hamilton 132: In the middle of the ninth century, as opposed to the beginning of it, when the a-b folios were copied.

[1] For Pseudo-Leo on chorbishops (J3 †1118), a forgery with Pseudo-Isidorian tendencies that is based on a canon from the Hispana Gallica (Seville II, c. 7) and shows a close relationship to the interpolated Hispana, see Fuhrmann, “Pseudo-Isidorian Forgeries,” in Papal Letters in the Early Middle Ages, ed. Hartmann and Pennington, p. 167.

[2] Hubert Mordek, Bibliotheca capitularium regum Francorum manuscripta,  MGH Hilfsmittel 15 (1995) p. 435.


In conclusion, a few observations:

Firey calls into question arguments that Klaus Zechiel-Eckes and Paul Hinschius made in two lengthy, sophisticated, and difficult articles. She discards the finer points of argument in both. In the face of Zechiel-Eckes’s analysis, she raises pedestrian concerns about the uncertainties of manuscript provenance that do not apply in the case of the St. Petersburg codex, that ignore a wealth of other evidence, and that have aged poorly as still more evidence comes to light.

Firey’s analysis of Hamilton 132, meanwhile, appears almost totally oblivious to Hinschius’s analysis and overlooks basic features of the manuscript. I am as nervous about palaeographical dates as anybody, but I also think it unwise to argue against the grain of too much established opinion on this front without good reason. If we accept Firey’s arguments for collaboration in Hamilton 132, we are forced either to reject the entirety of scholarship that dates the a-b hands in these folios to the years around 800 (moving a-b towards the middle of the ninth century), or to backdate the interpolated Hispana to the reign of Charlemagne (so that Caroline scribes can have collaborated with the a-b scriptorium in copying interpolated Hispana texts). There is no reason to do either.