Norman D. Hinton
[On-line editor's note: tables given on p. 39 and p. 40 of the printed edition are found at the end of this essay. 08/00]
Recent Chaucerian scholarship and criticism by Malcom Doyle, A.I. Parkes, and Janet Coleman1 has taken a fresh look at the manuscripts of the Canterbury Tales which may produce fuller knowledge of how Chaucer's work was viewed by his contemporaries, and perhaps may culminate in a new look at the Canterbury Tales themselves.
These publications are recent enough, and the theory of the compilatio unfamiliar enough to literary scholars, that it seems in order to discuss the compilatio theory briefly before I present the computer analysis I have made of the contents of the Canterbury Tales manuscripts.
A compilatio is a work which arranges auctoritates in such a way as to produce materials for a discussion of moral or ethical issues: I am not aware of compilationes on other subjects, but would not be surprised to hear of them. The compilatio is carefully ordered. It is typical for compilatio manuscripts to furnish such aids as a table of contents (sometimes analytical), analytical titles, rubrics, running heads, and copious marginal glosses, sometimes with cross-references throughout the collection. Compilationes were often produced in professional manuscript establishments, and it is common for the work to have been parceled out to a number of scribes. In such cases, traces of the work of manuscript supervisors can often be seen, surviving in the form of marks, notes, underlinings, pointings, and the like, to indicate the stages of production and assembly of the completed compilatio, or to note corrections to be made.
In a brilliant piece of paeleographic research, Doyle and Parkes have shown that the Ellesmere and Hengwrt manuscripts of the Canterbury Tales bear many of the marks of compilationes: Ellesmere and Phillipps 6750 even end with "Heere endith the Tales of Caunterbury compiled by Geffrey Chaucer." They suggest that one could think of the Canterbury Tales itself as adapting the compilatio form, given the narrator's insistence on accuracy in quoting his sources and the suggestion that he is merely the organizer of the Tales.2 The Tales, then, would be the auctoritates, and the ordinatio of the compilatio, the "array" the narrator analyzes in the General Prologue.
Whether the theory of Chaucer as fictional (or even actual) compiler can be fully supported is still to be decided. One would like to see, for instance, all the Mss. glosses: Manly's treatment of them tends towards the perfunctory, and omits all pointers, etc., except for the Knight's Tale.3 This is only the first of many reasons why we badly need a resurvey of all the Canterbury Tales Mss., their glosses, marginalia, and so forth.
Doyle and Parkes' work deserves the highest commendation for its carefulness, its precision, and the extremely high level of paeleographic analysis it displays. It does, however, as they would be the first to admit, leave a great deal of room for further study. For one thing, the fact that Ellesmere and Hengwrt are "complete" collections of Canterbury Tales leaves open another fascinating question, to which this paper makes a tentative first approach. Assuming that we do, in fact, know the order in which Chaucer left the Tales, or at least that we can make good informed guesses which leave us with two or three closely competing theories of the order of the Tales, and assuming further
It must be noted that an auctoritas is not an auctor. Auctoritates are "authoritative" statements (in our modern sense of the term). Nor is a compiler an auctor: the definitive statement is probably that of St. Vincent of Beauvais in the Speculum Maius: "I have added little or nothing of my own. Theirs is the authority, our part only the organization.''5 The compilers' job is to take the auctoritates and re-order them into a new meaningful arrangement.
Perhaps some of the assemblers of Canterbury Tales Mss. thought of themselves as compilers. Chaucer would have been their auctor, but the Tales would be the auctoritates. At one time, an auctoritas was as succinct and pithy as a sententia: by the middle of the 13th century, there was a growing movement to provide auctores in toro. This could produce a complete version of the Canterbury Tales, but in the special case of the compilatio, an auctor in toto might mean an entire Tale. (I linger on the point because it seems important, at this early stage of thinking about Chaucerian compilationes, not only that we keep our options open, but that we know what the options may be.)
I am making two heuristic assumptions: first, that the makers of the Mss. of the Canterbury Tales, and/or the patrons who ordered the Mss., may have treated the Canterbury Tales as a work which could be mined for auctoritates; and, second, that "incomplete" or "re-ordered" Mss. (or to use Manley's pejorative terms, "disarrayed" or "multi-lated" Mss.) may show us not how incompetent or confused the scribe(s) may have been, but rather may imply or actually show how people who lived shortly after Chaucer used the Canterbury Tales, or what they thought the Tales meant or taught.
I further assume that it is worthwhile to make such a study as easy and as complete as possible, and to provide a tool for such study that is flexible, is suited to various critical and scholarly approaches, and will save us all the task of leafing and releafing through the only information generally available for this study: Volume I of the Manly-Rickert edition of the Canterbury Tales.6
Acting on these assumptions, I have programmed a database and driver (as computer jargon would label them), and stored the database with information about the Mss. of the Canterbury Tales. This paper will describe the database briefly (not in computer terms),7 discuss some of its capabilities, and present some tentative early conclusions, or rather suggestions, as to where in the mass of data we might look, and how we might order the information we can obtain.
The information is divided into two main sections. The first section contains information about which Tales are found in which manuscripts. The second section contains information about the order of the Tales in each manuscript. The interaction of these two sets of data allows us to retrieve a great deal of information about the Mss. and the choices
I must interpolate here that Manly used the old Chaucer Society method of identifying the Tales and Groups, and I have perforce followed that method to enter his material. (The database will soon make automatic alterations for other methods of ordering the Tales and Groups.) It should be particularly noted that the Parson's Tale forms a separate Group I and that the Retraction forms a separate Group R. In addition, the Mss. carry a Group designated X, a non-canonic Tale. Most often this is the Tale of Gamelyn: other spurious Tales in the Mss. include the Tale of Beryn and the Ploughman's Tale, but I have labeled all occurrences as "Gam" for quick reference.)
Figure 1 is the opening screen display, the index to the database: I show it to give some idea of the flexibility and range of the program. Other options, including some statistical analyses of the distribution of Tales and Groups, are programmed, or soon to be completed, and will be ready for display very soon.
Figure 2 is the first set of data which may be of interest: the frequency with which Tales were copied into these manuscripts. It may be distressing to some to see the Prioress' Tale in the number 1 position, but there is really little difference in the frequency of the first 24 Tales, as can be seen from the statistics at the bottom of the Figure. The low number of copies of the Retraction is interesting: Chaucer's near-contemporaries were apparently not
Figure 3 shows the distribution of Groups throughout the Mss., when all members of a Group are present. (Group A moves up to 2nd place in occurrence when the Cook's fragment is removed from consideration.) Group B,8 the Man of Law's Tale, is a single-member Group, of course, but this alone cannot explain its popularity, since the other single-member Groups H, I, R, and X are all at or near the bottom of the list.
Figures 4-5 show the relative placement of Tales in various Groups, whether the entire Group is present in a given Ms. or not. Group A is a very homogeneous Group, most often found where we would put it today, and rarely interrupted by Tales from other Groups. The other examples shown are handled quite differently in the Mss., however. Group D is often in the positions which our modern editions would predict, appearing as Tales 7, 8, and 9. But the next three positions are also likely in the Mss., and while Group D is often followed by Group E, it is also preceded by Tales from Group E on a number of occasions. Group F seems almost to have no specific position in the Mss.: Group F Tales are found in all but 9 of the 27 possible positions, and Tales from Group F are preceded or followed by Tales from every other Group save the Retraction and Group A.
Such statistics would seem to suggest that Groups C, D, and F lack strong internal coherence in the Mss., and this is indeed the case. The latest
It is in these sequences of individual Tales that we can perhaps begin to see some criteria applied by the compilers. The Clerk's Tale, for instance, is preceded by the Summoner's Tale and followed by the Merchant's Tale, as we would expect, 17 times. But 20 times it is followed by the Franklin's Tale, setting two tales of either 'love beyond the norm' or of 'love beyond reasonable expectations' or perhaps love according to strict religious principles followed by love which takes social standing into account. Once the Clerk's Tale is preceded by the Knight's Tale, and once by the Man of Law's Tale, which puts it in serious company and 5 times it is followed (but preceded only once) by the Wife of Bath's Tale.
The sequence Squire's Tale-Merchant's Tale, just the opposite of our order, occurs 29 times, and the Merchant's Tale is followed by the Wife of Bath's Tale 21 times. Our sequence Clerk's Tale-Merchant's Tale is found just 17 times, despite the echo of the Clerk's closing lines at the start of the Merchant's Prologue. The Squire's Tale-Merchant's Tale sequence seems at first glance to emphasize the courtly aspects of the characters in the Tales, whether for praise or for satiric purposes, and Merchant's Tale-Wife of Bath's Tale gives husbands a chance to complain about wives, and then for wives to complain about husbands.
Other sequences provide material for speculation: to mention only two, the order Franklin's Tale-Pardoner's Tale occurs 12 times, putting two of the more pious frauds in juxtaposition, and the Franklin's Tale is followed by the Second Nun's Tale
This "first pass" through some of the information provided by the Mss. shows that there will be much to contemplate.9 Eventually we will have to examine the Mss. for clues which Manly and Rickert do not document, for it was information which they were not seeking and which they might not have recognized, since they were assembling materials for a classic edition on generic principles.
The clues may point to new interpretations of individual Tales, of the inner coherence (or lack thereof) of a Group, or perhaps to new or restored readings of the Canterbury Tales itself. At this stage of the study, we can only conclude that careful study of the Mss. does seem to give us a glimpse of how 15th century readers viewed Chaucer. We can think of our own attitudes towards some writer we cherish who has recently died: as Auden said of Yeats, "he has become his admirers" and he has not yet become the property of critics and scholars. The Chaucer dimly hinted at in the last few paragraphs is fresh in his readers' minds: if he is not the Chaucer we have come to love in almost a century of intense study since the Six-Text Chaucer, he is a more straightforward, a more direct Chaucer. If we can retrieve that Chaucer, we should not hesitate to admit his existence, and to admit him to the company of the Chaucer we now know, to refresh us and to enlarge our insights into his character, and that of his greatest work.
Figure 1 [for illustrative purposes only: on-line ed. 08/00
A database by Norman Hinton
a for raw data on contents of Mss.
b for raw data on order of Tales in Mss.
c for number of Tales per Ms.
d for number of Mss. per Tale
e for groupings of Tales by content
f for orders of groups
g list of Mss. sigla
h number and order of Tales in any given Ms.
i Canterbury Tales Groups per Ms.
j for contexts of Tales of one group
k for contexts of single Tale
copyright 1984, Norman D. Hinton
Page 39: Chart
Page 40: Chart
Sangamon State University