Here She Be — The Battlements

     The Structure of a Pilgrimage: The Ellesmere Order in Opposition to Bradshaw

     The ordering of The Canterbury Tales has been a problem since the work was first put into publication after Chaucer’s death. J. S. P. Tatlock speculates that early scribes and booksellers “made changes and adjustments in the manuscripts (MSS.) … for the contentment and convenience of readers, and no doubt with business motives” (107). While Tatlock cannot prove this, it does not seem unlikely. Looking at the discrepancies between the eighty-three surviving MSS. and the six early printings exposes the many ways in which our working canon for the Tales is simply a mess, and we must accept that we can never know exactly what Chaucer intended (Hannah 1118). Even now, the closest scholars can come to mutual contentment is ten tale fragments. So rather than trying to discern authorial intent, we must use careful analysis to determine which ordering is the most preferable. This paper will show that by this process, the logical conclusion is that the Ellesmere manuscript (El) contains the most desirable arrangement of Geoffrey Chaucer’s The Canterbury Tales. This will be accomplished by taking a look at what links the fragments together in the text, and more importantly by extensive comparison between El and the Chaucer Society order.
     The Chaucer Society order has sparked a good deal of interest, and is favored by several scholars. The order was developed by Frederick J. Furnivall, who suggested moving Fragment VI up in the El order, and Henry Bradshaw, who argued for moving Fragment VII forward as well. George Shipley goes so far as to call this arrangement of the Tales “an answer that is good, as far as it goes, for all time … [and] at present accepted as being the most satisfactory solution possible” (132). It seems only reasonable that any order claiming to be the most preferable is placed in opposition to this one. Still, Bradshaw’s suggestion is made in order to reconcile geographical problems in the text, and it appears he was busy thinking like a mapmaker rather than a literary analyst. Furnivall’s own words call his claims into question: “A happy hit! and it sets us free to alter the arrangement of any or all of the MSS, to move up or down any Groups of Tales, whenever internal evidence, probability, or presumption, requires it” (Pratt 1142). Internal evidence is a reasonable way of making alterations to the text’s structure, but “probability” or “presumption” is more than a little dubious. Such an eager attitude toward reworking Chaucer’s masterpiece, as well as with so little care as to the effects, should make any sincere analyst wary.
     Norman Blake has been working with other scholars on an online edition of the Tales, and both El and the Chaucer Society order have certainly made him cautious to consider all the evidence (“Editing”). Fortunately, he has the breakthroughs of many brilliant literary analysts, as well as his own research, to guide him. However, for every new detail research uncovers in the MSS. there seems to be a new fragment arrangement proposed, drawing further debate. People remain interested and the question of proper order persists.
     The continued concern about the Tales’ order comes from the real source of the debate: what effects changing the arrangement of the fragments can yield. The importance of the order most visibly relates to the outer structure of the pilgrimage. As the Tales progresses, the pilgrims develop relationships with one another and both cruel and kindhearted quips take shape in and around the storytelling contest. In this same way, the inner-frame of the stories themselves directly impact one another and also the reader’s perceptions. The tale order has a good deal of control over interpretation. For example, the Miller’s vulgar fabliau is a drastic change of pace from the noble Knight’s tale that precedes it, and that affects how the readers view each of the stories in hindsight, as well as the work as a whole. Changing the order in any way can dramatically affect this delicate system. And summing up this intricate web of intra-textual relationships as simply a matter of geographical consistency is the desire for an easy fix for a minor textual discrepancy. The moving of Fragment VI and Fragment VII causes more problems than it fixes. To determine appropriate structure for the Tales, perseverance and a careful eye are what are needed. We need to see beyond the physical locations to the context the Tales create in conjunction with one another. This can be achieved from looking at the text, as well as viewing how the extant MSS. order the Tales. El is the ideal model, and herein analysis shall show why its arrangement of the fragments from one to ten, in order, is preferable to other structural models.
     The logical place to begin is with Fragment I, and how we know that it is followed by Fragment II. Fragment I contains the General Prologue (GP), which lays the foundation for the Tales structure. The introduction provided by the GP assures Fragment I’s position, but the positioning of Fragment II is a bit more of a mystery. The “Cook’s Tale” at the end of Fragment I is very brief, and feels incomplete. The start of a story is there, but the tale ends on “And hadde a wyf that heeld for contenance / A shoppe, and swyved for hir sustenance,” which is still only establishing characters with no real elaboration that could be called a story (1.4421-422). This presents an obvious problem. Still, when the text cannot provide a clear enough answer, we can turn to the physical, extant MSS. Shipley points out that “almost all the MSS. have this order” (133). Furthermore Germaine Dempster did extensive work with Manly and Rickert’s tale groups, which look at how the Tales has been arranged historically, and come up with five orders that have existed over time. Once the “Tale of Gamelyn” is removed, as it should be because the author has been determined not to have been Chaucer, all of Dempster’s groups are arranged with Fragment I followed by Fragment II (“Fifteenth” 1123). Robert Pratt after experimentation states that “no other position is satisfactory” (1143). Both the tradition of the MSS. and a respected analyst going through and testing different locations for Fragment II have shown that this curious fragment should follow Fragment I. While this is somewhat inconclusive, the evidence that is available supports the El MS.
     Moving on with linking Fragment II to Fragment III involves the first conflict between the El MS. and the Chaucer Society order. Whereas El follows the fragment order II – III, the Chaucer Society order proposes II – VII. This fragment rearrangement proposed in the Chaucer Society order stems from the recommendation made by Bradshaw. The basis of this change, often referred to as the Bradshaw shift, has reasoning that is both apparent and logical: Fragment III references a town called Sittingbourne, and Fragment VII references a town called Rochester; on the road to Canterbury, Rochester would be encountered before Sittingbourne (Pratt 1148). However, a more in-depth look at the text and extant MSS. provides more than sufficient evidence against Bradshaw, also revealing reasons to keep the II – III arrangement from El.
     A large problem with the Bradshaw shift stems from a complete lack of MS. support. No existing MS. agrees with Bradshaw’s theory. The MS. that Bradshaw used to bolster his argument is called Selden (Se). To see why this is ineffectual, we need only take a look at the fragment order: I – IV1 – III – IV2 – V1 – II – VII – VIII – VI – V2 – IX – X (Moore 117). Fragment II and Fragment VII do appear side by side, as Bradshaw suggests they should; however, Fragment IV and Fragment V appear to have suffered a not so happy hit. As Samuel Moore points out, Se is “unquestionably wrong (a) in placing the Clerk’s Tale … before the Wife of Bath’s prolog and tale; (b) in splitting group F by placing some 7000 lines of text between the Squire’s Tale and the Franklin’s Tale” (118). Aside from the bizarre discrepancies in Se’s ordering, there is one far more curious and important detail: Fragment III comes before the Fragment II and Fragment VII combination. The geographical inaccuracy of Sittingbourne preceding Rochester exists in the very document Bradshaw uses to make his claim. Throughout all the surviving MSS. there is not one that does show Fragment VII before Fragment III (Brown 34). This of course means that although Bradshaw claimed to have MS. support for his argument, he did not. In contrast, El shares its exact ordering with five other MSS. (Dempster, “Manly” 395). There are also an additional four MSS. that have been shown to be based on El (Hannah 1120). Unlike the Bradshaw shift, El does have MS. support (other than itself).
     Of course these were the choices of scribes, and the more authoritative source is the text of the Tales. Regarding the actual text, it seems most reasonable to first look at the reference to Sittingbourne in Fragment III, which creates the geographical discrepancy. The Somonour in the Wife of Bath’s Prologue says, “But if I telle tales two or thre / Of freres er I come to Sidyngborne / That I shal make thyn herte for to morne” (3.846-848). And at the end of his tale, the Somonour indicates that they are making an arrival: “My tale is doon; we been almost at towne” (3.2294). Robinson made the connection in his edition of the Tales that the words “almost at towne” indicate “Sidyngborne,” which the Somonour mentions earlier (Greenfield 52). The biggest problem with this assumption is that the Somonour has not told “tales two or thre” (3.846). He has told one. And although he explicitly mentions Sittingbourne, at the end of his tale he does not say which town they are coming to, only that they are in fact arriving at some town. It seems entirely possible that the reference to Sittingbourne indicates some town farther along on the journey and that the Somonour is going to continue to tell stories at the Friar’s expense. Since the story contest entails that each participant will tell two stories on the trip to Canterbury, the Somonour probably expects to tell his second story by the time they reach the town of Sittingbourne.
     An important realization comes from taking the line preceding “almost at towne” into account: “And Jankyn hath ywonne a newe gowne” (3.2293). The line about Jankyn winning a new gown for his scheming is the more important part of the couplet and “towne” is simply provided as a rhyming word. The need for a rhyme seems to indicate less thought towards the Somonour’s prior reference and more thought towards creating an effective couplet. Stanley Greenfield argues that the word “may not even have a real place as a referent” (52). The other likely possibility is that the group is coming to some other town on the road to Canterbury that is not Sittingbourne and that the Somonour’s reference from the Wife of Bath’s prologue is merely his way of saying “later on the trip.” Helen Cooper offers this in a simple statement: “There is however no logical reason why the Summoner’s remark indicating that Sittingbourne is still a considerable distance ahead … needs to follow the mention of Rochester” (277). Taking this into account, there is no evidence that provably discredits the Ellesmere fragment order or preferences the Bradshaw shift. In fact, recognizing the Somonour’s statement to mean “later on down the road” indicates an early placement for Fragment III, which coincides with its location in El.
     Scholars have made a variety of arguments regarding Bradshaw. For example, Pratt sets forth his own theory in support of Bradshaw:

One complete set … was piled up, arranged physically … close to or identical with the poet’s intention (the “Chaucerian” order). At about this time, however, the fascicles and leaves making up the sequence … (Fragment VII) cannot have been in [its] proper location. Fragment VII may have lain in the wrong position … it may simply have lain apart from the rest of the MS. and then been inserted in its wrong (“1400”) position by someone who knew no better. For example, the bulk of the MS. was perhaps in Chaucer’s chest while … [Fragment VII] lay on his desk. Once this misplacement occurred it was never rectified. (Pratt 1162)

     It is strange that Pratt overlooks the possibility that Chaucer was revising the Tales up to his death, and thus had not made the alterations to place names that would make the Tales geographically accurate. Furnivall’s “probability” and “presumption” cannot be given credence, nor can Pratt’s idle thought. This grasping at straws is not sufficient literary analysis. Carleton Brown pointed out scholars’ inability to provide reasonable support for Bradshaw by saying, “Editors have laid too much stress on the topographical allusions in determining the order of the tales”; Brown continues by citing Manly: “When he [Chaucer] changes his intentions with regard to the use of a particular tale, he did not always remove traces of its previous use” (34). Manly is known for his extensive reading and analysis of the MSS., and his judgments about the arrangement of the Tales have sufficient research behind them, especially in comparison to Pratt’s assumptions about how Chaucer liked to leave his work in progress scattered around his house. Cooper’s own research-based assessment follows that actually “there is negligible evidence that Chaucer was so interested in the geographical detail of the journey, or had revised the work sufficiently, to make such references a reliable indication of tale order” (277). The constant theme among scholars who use research to back their claims is that geographical sites are of little importance. The emphasis of the Bradshaw shift is poorly placed.
     In addition to the Bradshaw shift, Furnivall suggested moving Fragment VI up to follow the new position of Fragment VII. As noted earlier, literary analysts are wary of Furnivall’s recommendations and with good reason. His change makes the opening of the Tales appear in the following order: I – II – VII – VI – III. Although we can already see why Fragment VII should not follow II, we must still consider Furnivall’s placement of Fragment VI. According to Furnivall, the Pardoner’s line, “I wol bothe drynke and eten of cake” (6.322) indicates a morning hour, so Fragment VI should precede III with its reference to the time being around ten o’clock (Pratt 1142). Pratt, who gave evidence in support of Bradshaw, argues against Furnivall, saying that Furnivall was more interested in the preconceived notion that the trip took four days, plainly stating “no good has been effected by it” (1142). There is no reason to assume that drinking, presumably wine, and eating cake is a reference to breakfast. Furnivall’s conclusion is poorly supported, and should be given little, if any, consideration.
     Moving onto the next segment of the Tales, Fragment III, Fragment IV, and Fragment V, have clear connections both through the Wife of Bath and their theme. Fragment III begins with the “Wife of Bath’s Prologue” and the “Wife of Bath’s Tale.” Various references to the Wife and her tale arise in the “Clerk’s Tale,” which begins Fragment IV. The most notable is a direct reference: “For which here, for the Wives love of Bathe / Whose life and all hir secte God maintene” (4.1170-171). The second line indicates the Clerk has a working knowledge of the Wife of Bath’s life, which he could only have acquired from her prologue. This clearly shows that the Clerk comes after the Wife in the contest, and as his story begins Fragment IV, Fragment IV should follow Fragment III. The Merchant who tells his tale after the Clerk also mentions in the Wife of Bath: “The Wife of Bath, if ye han understonde, / Of marriage, which we have on honde” (4.1685-686). This also shows that the “Merchant’s Tale” (MerT) must follow the “Wife of Bath’s Prologue,” bolstering the argument that the proper order is III – IV, as found in El.
     The fifth fragment, as well as Fragment III and Fragment IV, are more closely tied together by theme. Donald Howard calls it them the “tales of domestic conduct” (247). Eleanor Prescott Hammond dubbed these fragments the “Marriage Group” in 1908, and Kittredge furthered the idea, as well, referencing the segment as “a discussion of marriage, characterized … as a dramatic unit treating a common theme” (Howard 247). Although there are variances in the analyses of these fragments, they have long been considered a single unit and their arrangement has gone undisputed. Howard describes each of the tales presented in these fragments with the different pilgrims’ viewpoints: “The points at issue are raised by the Wife, male hostilities are vented by the Friar and Summoner … Clerk’s ironic idealism … Merchant’s cynicism … the Squire’s aristocratic romanticism and the Franklin’s naïve optimism” (271). The effect of this trilogy of fragments is to offer different viewpoints on the theme of marriage without actually giving a solution. This section provides the reader with an opportunity to interpret different ideals for themselves and all at the same time. In this way, the order III – IV – V is supported largely by theme.
     However, this is not all that supports the idea that these fragments are a single unit. The epilogue to the Merchant’s tale at the end of Fragment IV and the introduction to the “Squire’s Tale” at the start of Fragment V make logical sense in sequence. At the end of his epilogue, the lines “Syn women konnen outen swich chaffare; / And eek my wit suffiseth nat thereto” (4.2338-339) suggest that the MerT was insufficient in clarifying questions relating to love and marriage. This ties in perfectly with the request in the introduction to the “Squire’s Tale”: “Squier, com neer, if it youre wille be, / And sey somewhat of love, for certes ye / Konnen theron as muche as any man” (5.1-3). If the Merchant’s story was less than adequate, the Host moves on down the line to the next most likely candidate. In this case he assumes that the Squire has some great knowledge of love. These two segments that frame their respective stories combine almost seamlessly together, showing good reason for Fragment IV to be followed by Fragment V, as per the El order.
     The unification of Fragment V and Fragment VI is based largely on theme. Dorigen’s list of maidens is an apparent theme in this portion of the Tales, and those women who prefer death to shame is a common bond between the “Franklin’s Tale” (FranT) that concludes Fragment V and the “Physician’s Tale” (PhyT) that begins Fragment VI (Cooper 247). This shared trait strongly suggest that these fragments, specifically because of FranT and PhyT, were meant to appear alongside one another. Were the underlying theme for these two stories vaguer, the tales would not seem connected; however, Dorigen’s list of maidens is extremely particular. This very specific concept supports the arrangement of V – VI. (By the end of this paper it will also have become apparent why Fragment VI cannot appear anywhere else in the Tales.)
     Actually joining Fragment VI to Fragment VII is easy to justify, as it is strongly supported throughout the history of the extant MSS. of the Tales. Once again referencing Dempster’s article on tale order, in which she provides five core MS. groups based partially on the research of Manly and Rickert, there is no existing series of MSS. that does not have the arrangement of Fragment VI followed by Fragment VII. Se does conflict with this arrangement, but its order is clearly unauthoritative. The likelihood of fragments VI and VII being one unit, based on the evidence, is almost indisputable; Tatlock and Skeat also support this conclusion (Moore 122-123). The vast majority of the surviving texts propose this order, and extensive research by many scholars has only yielded support for the argument. The logical conclusion is that Fragment VI is followed by Fragment VII, the same as in the El order, as well as almost every other ordering.
     Fragment VII’s placement in the tale order is also supported by a wealth of MS. evidence. Even the strange Se MS. places it towards the end of the tale fragments. The closing line of the epilogue that ends Fragment VII is “Seide unto another, as ye shuln heere” (7.4652). This line suggests that a likeminded tale is next in the order, and while the unity of the two tales is questionable, the next tale belongs to the Second Nun. The epilogue that ends Fragment VII comes from the Nun’s Priest. It seems logical that if another tale of similar mind is next, the reference would be to the Second Nun. The unity of Fragment VII to Fragment VIII stems from a unit amongst the characters.
     Finally, fragments eight through ten are the closing tales. There are several references to time that make it clear that fragments VIII and IX “come at the end just before the Parson’s Tale” (Howard 288). The link between fragments IX and X are what cement Fragment VII’s position, as we know that Fragment VII can appear nowhere else besides its current location once IX and X are shown as the appropriate ending.
     Fragment IX ties to Fragment X in the first line of Fragment X: “By that the Maunciple hadde his talle al ended” (10.1). Fragment IX consists solely of the “Manciple’s Tale,” so if it ends as the Parson begins at the start of Fragment X, Fragment IX must tie to Fragment X. The Host’s request to the Parson to tell the final tale cements Fragment X as the end of The Canterbury Tales: “Thou sholdest knytte up wel a greet mateere” (10.28). The Host is referring to the contest, and the Parson will tell the final tale. This brings the Tales to a close, reflecting the El order.
     After consulting all the evidence involved in the dispute between the Chaucer Society order, especially the Bradshaw Shift, and the Ellesmere order, the obvious conclusion seems to be that the preferable arrangement of the Tales is the El order: I – II – III – IV – V – VI – VII – VIII – IX – X.
     The Se MS., which was the original foundation for Bradshaw’s proposition, has been shown to be one of the poorest quality extant MSS. It does not even fix the problem of the Sittingbourne / Rochester reference that Bradshaw is dealing with.; it only provides an example where Fragment II and Fragment VII appear side by side — towards the end of the text where, according to Bradshaw, neither fragment should be placed. El, on the other hand, is highly acclaimed by a variety of sources, and is used as the basis for The Riverside Chaucer, which is arguably one of the best arrangements of the Tales to date. Critics discredit Furnivall’s work with good reason for its support of “presumption” and “probability,” but for the most part they lay foundations for their own arguments based on similar conjecture. Discovering the missing link in the ordering of Chaucer’s Tales would be an exciting endeavor for any literary analyst, but the truth is, based on what we have, there appears to be no missing link. We cannot know what Chaucer intended, and that means we can never be sure.
     The linked section of Fragment VI and Fragment VII should always be given sincere consideration and not written off as an error that all the scribes, except for the scribe responsible for Selden, must have made. There is much to suggest that El is as close to Chaucer as we can get, and there is little if any effective evidence to prove otherwise. The El is a superior ordering, overwhelmingly preferable to the Chaucer Society order and the Bradshaw Shift. The Sittingbourne reference appearing before the Rochester reference, which has so long been an issue, clearly does not indicate what has so often been concluded. Simply reading and paying attention to the text, in a way that any normal reader would, shows us that. The El MS. gives us a clear and effective ordering of the Tales supported by many extant MSS. as well as the context of the stories themselves.

Works Cited
Blake, N. F. “The Editorial Assumptions in the Manly-Rickert Edition of ‘The Canterbury Tales’.” English Studies (1983): 385-400. JSTOR. Ames Lib. 20 Nov 2007 .

Blake, N. F. “Editing the ‘Canterbury Tales’: An Overview.” The Canterbury Tales Project: The Occasional Papers, Volume 1. Ed. Norman Blake; Peter Robinson. 1993. Geoffrey Chaucer Online. April 2007. Office for Humanities Communication, Oxford. 22 Nov. 2007 .

Brown, Carleton. “Author’s Revision in the Canterbury Tales.” PMLA 57.1 (1942): 29-50. JSTOR. Ames Lib. 20 Nov 2007 .

Chaucer, Geoffrey. The Riverside Chaucer. Ed. Larry Benson. New York: Houghton Mifflin, 1987.

Cooper, Helen. Oxford Guides to Chaucer: The Canterbury Tales. 2nd ed. New York: Oxford University Press Inc., 1996.

Dempster, Germaine. “The Fifteenth-Century Editors of the Canterbury Tales and the Problem of Tale Order.” PMLA 64.5 (1949): 1123-1142. JSTOR. Ames Lib. 17 Nov 2007 .

Dempster, Germaine. “Manly’s Conception of the Early History of the Canterbury Tales.” PMLA 61.2 (1946): 379-415. JSTOR. Ames Lib. 20 Nov 2007 .

Greenfield, Stanley B. “Sittingbourne and the order of ‘The Canterbury Tales’.” The Modern Language Review 48.1 (1953): 51-52. JSTOR. Ames Lib. 20 Nov 2007 .

Hannah, Ralph. Textual Notes. The Canterbury Tales. By Geoffrey Chaucer. Ed. Larry Benson. New York: Houghton Mifflin, 1987. 1118-1122.

Howard, Donald R. The Idea of the Canterbury Tales. London: University of California Press, Ltd., 1978.

Moore, Samuel. “The Position of Group C in the Canterbury Tales.” PMLA 30.1 (1915): 116-123. JSTOR. Ames Lib. 25 Nov 2007 .

Pratt, Robert. “The Order of the Canterbury Tales.” PMLA 66.6 (1951): 1141-1167. JSTOR. Ames Lib. 17 Nov 2007 .

Shipey, George. “Arrangement of the Canterbury Tales.” Modern Language Notes 10.5 (1895): 130-140. JSTOR. Ames Lib. 3 Dec 2007 .

Tatlock, J. S. P. “The Canterbury Tales in 1400.” PMLA 50.1 (1935): 100-139. JSTOR. Ames Lib. 20 Nov 2007 .

The arguments made in this paper were presented at the 2008 John Wesley Powell Research Conference held at Illinois Wesleyan University.