The 33 manuscripts which contain all or portions of the First Part of Ockham's Dialogus may be classified into 4 basic groups:
A. Bb* Fi* An* Ce Na Sm Vd* Ca Lc La Un Ax
B. Vg* Va* Lb Pa Pb Pc Vb Ar Sa Ko
C. Ba* Di To Es Fr
D. Vc* Vf* Av Ox* Br We*
The asterisks indicate manuscripts fully collated for our critical Latin text of 1 Dial. 6.1-15.
The main reason for assigning a manuscript to a specific group is the quantitative and/or qualitative preponderance in that manuscript of readings which are peculiar to the group in which the manuscript has been included, notwithstanding the presence of occasional conflations and/or idiosyncracies in a particular witness. Each manuscript has its own genetic peculiarities, which must always be kept in mind in the context of a conveniently simplified ordering system. Fr, for instance, does not belong to C in quite the same fashion as To or even Ba, and We does not belong to D in the same manner as Vc or Av (see further below).
The complex nature of medieval manuscript copying (and the consequent difficulties in tracing definitive and clear-cut affiliations and developing consistent stemmas) is well exemplified by manuscript Na. This witness has been dated by Cenci to the 14th century, but independent examination by Prof. Doyle of Durham University (Offler Archives, Dialogus folder) has not fully confirmed such an assumption. My own analysis indicates that the manuscript's reproduction of the First Part of Dialogus apparently relied on exemplars of groups A and B, the former being of an intermediary type which was beginning to exhibit certain characteristics of C. The utilization of these sources did not always seem to be properly co-ordinated. Thus in 1 Dial. 1.3, where groups A and B usually describe the authors of the Canon Law as "viri acutissimi", and groups C and D as "viri eruditissimi", Na has "viri eruditissimi acutissimi" (as do e.g. Ca Vd La Fr). In the same chapter, where group B has "indices" and some exemplars of group C "determinares", Na has "indices et determinares" (while groups A and D normally prefer "intimares"). In 1 Dial. 2.21, Na has the same major omissions that we find only in some group B (Pa Pz Ly Sa Lb Ko) and in a few late group A (La Lc Vd) texts. In 1 Dial. 5.34 Na reflects an error of transcription, adding (as do Pb La and Sa) a superfluous "Discipulus" in the middle of a sentence of the Master (whereas Bb Ca Un Pa Vb Lb Lc have an equally superfluous "Magister" in the same place). This particular addition seems to have subsequently influenced the editor(s) of group D (but not of C, except for Ba which here borrowed from D) to write in what he (they) felt was an appropriate equivalent of the "missing" text, which in effect merely provided a redundant statement as to substance. In most other contexts Na follows the tradition of group A rather closely (with frequent points of convergence to Lc and Vd). All this suggests a very complex process of textual development and copying, and the gaps noticeable throughout (e.g. in 1 Dial. 6.1) possibly indicate an intermittent lack of interest in or capacity for proper editing.
Na is but one instance of what can be observed in many other manuscripts. Va for example (copied in 1437) is a B group text with multiple but not comprehensive integrated corrections from a D group exemplar. From time to time (for instance in Pb and Lb) we may even catch these adopted corrections "in statu nascendi" as marginal and interlinear glosses. On the other hand, C group manuscripts (the most idiosyncratic of the 4 basic groups) sometimes have a significant number of readings in common with the B group tradition (as in 1 Dial. 6.1-15). The same kind of relationship may be discerned (though not systematically or universally) between groups C and D.
The leading exemplar of the C tradition (Ba, a mid-15th century manuscript) shares some very specific readings with Va (cf. the apparatus for 1 Dial. 6.1-15 at chs. 1 /thrice/, 5, 14 /four items/), readings which otherwise are only known to exist in the D tradition, and which cannot be found in any other exemplars of B or C. This demonstrates that important elements of the D textual tradition were available by 1437.
If we focus on the extant manuscripts of the D tradition (leaving aside the Br fragment which John Scott links to Vc) we may recognize three sub-groups: (1) We (2) Av Ox (3) Vc Vf. We is a multiple copy-hands compilation of the later 15th century put together on the basis of at least two, and perhaps three, distinct manuscript sources (it has a "second prologue" in common with Fr, as mentioned below), but with highly visible and sustained though not quite comprehensive affinities to Av Ox Vc Vf. Both the Av Ox and the Vc Vf groups go back to the same original through at least one intermediary, and common place names in Av Ox Vc Vf (cf. 1 Dial. 5.22-24) associate this original with the March of Ancona (which, of course, suggests but does not necessarily imply that this is where the original was transcribed). Each "Ancona" sub-group has been edited further, and the immediate ancestor of Av Ox has provided chapter headings analogous to those of Ly. Vc and Vf are products of the 1470's (Vc was the copy of Pope Sixtus IV), while Av and Ox are slightly earlier (Ox existed in 1444). Their common original, with its Table of Contents and Admonition to the Reader (both authored by one or more extremely competent though unfortunately anonymous editor or editors), would thus appear to be at least contemporary to the early Council of Basel. There is no evidence to indicate the existence of an earlier D text independent of Table or Admonition.
The textual archaeology of tradition D in 1 Dial. 6.1-15 intimates at first glance that either its original relied here on an exemplar of A which had very little in common with tradition B, or else that, in this context, the "Ancona" editor(s) of D (unlike tradition C in the same passages) deliberately chose not to draw on tradition B for assistance. Analysis of further segments where "Ancona" (whose contacts with C have been well-documented by John Scott) obviously and repeatedly borrows from B (cf. below at 1 Dial. 7.42-51 and at 1 Dial. 7-65-73) indicates the greater plausibility of the latter hypothesis. The clearcut and abundant evidence that both C and D do rely on tradition B as convenience dictates, becomes a powerful if not yet conclusive argument in support of the view that C and D are in many respects derivative and conflated traditions rather than representations of a neglected and subsequently rediscovered Ockham original. Certainly, neither C nor D as we have them can be anterior to B, even if we were to postulate that B was only sporadically available to the editors of C and D. There is also growing evidence that tradition C is closely linked to some later manuscripts of tradition A, and may well have been a continuation and "correction" thereof (cf. Introduction to 1 Dial. 6.51-67).
In any event, none of these groups is able on its own to provide an exclusive basis for reconstructing the text whence stem all of our extant witnesses. Furthermore, the fact that (unlike the case of Summa Totius Logice or of other purely philosophical and theological Ockham works) we do not possess a single manuscript of Dialogus which may securely be dated to the author's lifetime (our oldest is Bb, a mid-14th century group A exemplar which belonged to the Basel Dominicans) raises special issues of authenticity. All groups to a more or less evident extent share a common text in the incipits and explicits of individual books of the First Part, as well as of the treatise as a whole. The systematic reference in this common text to "Dialogi" (plural) rather than to Ockham's preferred "Dialogus" (singular) strongly suggests that the original of all our surviving manuscripts was edited sometime after Ockham's death by a person or persons who (as many other contexts indicate) was or were not always totally familiar with the Venerable Inceptor's intentions. There is no need to doubt that the reproduced text remains substantially faithful to Ockham's unavailable autograph, but the presence therein of occasional uncorrected errors (cf. for instance our Introduction to 1 Dial. 7. 65-73), as well as of additions, adjustments, or improvements some of which go back to the very beginning of its textual history should perhaps make us more vigilant as to yet further "improvements" demonstrably or potentially attributable in the various groups to a number of subsequent editors. The most curious of these are doubtless the "Frankfurt prologue" (in Fr) and its homologue in We, which I discovered in 1975 and 1991 respectively. The text of this spurious if interesting prologue has been edited by my Australian colleagues.
Group A contains the oldest manuscripts, but these have some defective peculiarities and significant verbal omissions. Group B represents the 14th century tradition which evolved into the printed editions (Paris 1476 and Lyons 1494 [the latter reprinted by Melchior Goldast in 1614]) and is therefore the one most familiar to historical practitioners of Dialogus. It is the group to which belonged the lost manuscript by reference to which Pierre d'Ailly composed his abbreviation of Dialogus. This group's text also has many defects. Groups C and D have some reasonably good late exemplars (though the text of C is on the whole obviously
inferior to that of ABD), but their tradition cannot be traced back much further in time than the beginning of the 15th century (the first "building blocks" of C are discernible in late 14th c. A tradition texts such as Ax and Ca), and the frequent excellence of the provided text (especially that of D) needs to be balanced by concerns for authenticity which cannot in all instances be positively resolved. We can demonstrate that Simon de Plumetot corrected his group B Dialogus exemplar (Pa, originally copied in 1389) by reference to a group C text sometime in the second or third decade of the 15th century. We know that Henry of Zoemeren's Epithoma Dialogi (c. 1460) was also based on a group C text. But we lack any convincing evidence for the early existence of this tradition. There is even less evidence for the antiquity of the D tradition. Therefore we cannot always fully identify the "good" readings shared by C and D with Ockham's authentic words. Strict analysis indicates to me that only four of the "significant" CD variants reviewed by John Scott in his most useful
It should also be pointed out that the textual adequacy of group D is not uniform or consistent, as the apparatus of 1 Dial. 6.1-15 (and that of 1 Dial. 7.65-73) clearly reveals. In many cases the "common text" of D has not been adopted, either because it is obviously defective, or because it is superfluous (sometimes awkwardly so, as in 1 Dial. 6.14). Group D is particularly strong (though not infallible) in the recording of Biblical, Canon Law, and Patristic citations, and is frequently our best general witness, yet it needs to be supplemented and corrected by the other groups if the original text of Ockham's published Dialogus is to be adequately or nearly adequately recaptured. While the notion that traditions C and D may in fact have been as close in time to Ockham's original as A and B, and that their 14th c. intermediary texts have been "lost", is not entirely impossible, this seems a rather improbable ad hoc solution to the vexed problem of textual continuities, and the challenge of demonstrating the antiquity of the glaring systematic defects of C is daunting to say the least. In the current state of the evidence, and in the absence of clear indications to the contrary, it is much safer and much more probable to accept that a given tradition begins with the direct and demonstrable sources of its earliest extant exemplars.
Whether or not to adopt subsequent editorial improvements into our final critical text is, of course, a distinct issue. It is arguable that Ockham might not have been averse to the inclusion of such improvements into the text he was still working on in his scriptorium when death came calling. We have, after all, the evidence of the famous chapter 51 of Part I of the Summa Totius Logice as an indicator of his attitude towards such matters, as well as the concluding sentences of 1 Dial. 7.73.
Chapters 1 through 15 of 1 Dial. 6 have therefore been reconstructed here on the foundation of a full collation of the 11 "best" and most reliable exemplars of our 4 manuscript groups, with occasional references to other witnesses. The two printed editions of the 15th century have also been carefully examined, though only that of Lyons (Trechsel) has been fully reported.
The "reliability" pattern which emerges in this first published segment of book 6 is quite interesting, but cannot as yet be fully conclusive for the entire book, and even less for the entire treatise. The manuscript which is closest to our critical text in these opening chapters is We (with an 87% rate of variants convergence), closely followed by An (86%), Fi, Vc, Vf, Bb (all at 85%), more distantly by Ox (83%, not counting a large textual omission at 1 Dial. 6.15), with Va, Vd, Vg and Ba trailing somewhat (all between 74% and 78%), and the historic Trechsel Lyons printed edition bringing up the rear at 72%. We should note however that many defective variants are unique to each discrete witness, whose value in confirming or denying most standardized readings is not drastically impaired by such erroneous idiosyncracies. Nor should we forget that, on balance, variant units and/or clusters only affect some 15% of the total text. A "reliability" rate differential of merely 13-15% between "best" and "worst" within that narrow 15% is hence contextually minimal.
My colleagues John Kilcullen and John Scott have kindly reviewed the first pre-posted version of 1 Dial. 6. 1-10, as well as the first posted versions of 1 Dial. 6. 1-50 and of 1 Dial. 7.42-51, for which I am most grateful. They have, in some cases, convinced me to adopt tradition D (rather than my original options) as best for our critical text. I also thank my colleagues for verifying in a number of manuscripts the contexts of 1 Dial. 1.3 and 1 Dial. 5.34 mentioned earlier.
The anonymous 14th century scribe who copied our witness Vg recorded book 6 of the First Part as the "secunda pars" of this treatise. This is a useful perspective. For it is here, in this massive sixth book, that Ockham's conflict with Pope John XXII begins to spill over into issues of immediate practical relevance to the dissident Franciscans of Munich. Is the Pope above the law? If not, how should one proceed to verify whether he is a criminal? How should one punish him if it turns out that he is? The very title of book 6 is pregnant with political passion. The tensions and not always restrained fury of this historic confrontation still reverberate through these pages, and Ockham's powerful dialectic continues to fascinate and to inspire nearly seven centuries after the events to which it was applied.
The material presented by Ockham in this first segment of 1 Dialogus 6 had been utilized for doctrinal reconstructions in A.S. McGrade's The Political Thought of William of Ockham, at p. 19 n.38, p. 88 n.23, p. 94 n.38, and p. 107 n. 78. It had been utilized for the same purpose in my Political Ockhamism, at p. 28 n.21, p. 35 n.34, p. 50 n.128, p. 96 n. 229, p. 98 n. 233, p. 158 n.249, p. 238 n.7, p. 261 n. 98, p. 263, p, 268, and pp. 290-292. A new perspective may be added to these earlier analyses. It is now arguable that Ockham knew the theories of Jean Quidort ("Johannes Parisiensis", "John of Paris"), and may sometimes have quoted him verbatim in the Dialogus. The French Dominican thus plausibly joins Marsilius of Padua as a major source of the radical anti-papal doctrines discussed in 1 Dial. 6.6-9.
For the general context and meaning of 1 Dial. 6.1-15, see my Fragments of Ockham Hermeneutics, pp. 92-99.
George Knysh
May 2004