Printer Friendly

Chaucer's tell-tale lexicon: romancing seinte cecyle.

Most scholars agree that Chaucer's increasing use of Romance vocabulary can roughly date his works relative to each other and - based on the few works with generally accepted dates - help establish a chronology of his literary production. This procedure can also date extended parts of Chaucerian texts through analyses of the percentages of Romance vocabulary usage in each. Drawing on modern statistical methods and insights from Chaucerian poetics and other fields, lexical analyses in this study will reveal the complex history of the composition of the Second Nun's Prologue and Tale.

Joseph Mersand undertook lexical analysis of the entire Chaucerian canon in his 1939 study, Chaucer's Romance Vocabulary. His purpose was "to investigate Chaucer's Romance words, not only to provide definite, reliable quantitative estimates, but also to demonstrate the value of such statistics for the solution of problems of interpretation, chronology, and textual accuracy" (2). For each of Chaucer's works, Mersand calculated the "(1) number of Romance words per line in usage; (2) number of Romance words; (3) percentage of Romance words in vocabulary; (4) percentage of Romance words in the text; (5) number of Romance words used per line; and (6) percentage of Romance words that rime" (98n27).(1) The category of interest to this article is the fourth: the percentage of Romance words used in the text.

Mersand came to ten conclusions, two of which are central to this study: his first, that "Chaucer added new Romance words to his vocabulary as he advanced in his literary career" (137), and his sixth, that the "proportions of Romance words show wide variations in the separate works. The variation may be employed to verify the accepted chronology of his works and to suggest a few changes" (137). Mersand's thesis has been challenged by David Burnley and Norman Davis, but the two conclusions affecting this essay have been strongly corroborated elsewhere.

Both Burnley and Davis contend that Mersand's analysis ignores the important difference between established French loan words and those introduced by Chaucer. Burnley argues that "Statistical statements to the effect that Chaucer's vocabulary contains 51.8 per cent of Romance loan words [static total, not usage] are of little use in assessing the impact of Chaucerian style, and indeed simply ignore the crucial factor of the contemporary perception of the status of the words" (135). In fact, Mersand had no intention of assessing Chaucerian style. His much narrower goal was the calculation and analysis of the number and percentage of Romance words and usage in ten categories in each of Chaucer's works, and the chronology of the works these data reveal.

Burnley discusses the social implications in fourteenth-century England of adopting French words, which, he states, "was associated with social elevation," as "French words and phrases were thought to give elegance to English expression" (134). He goes on to point out, however, that since French loan words were common in romances generations before Chaucer,

a simple etymological distinction between the words of French and English origin would be a misrepresentation of the way in which Chaucer's compatriots would perceive their language. At the least, we must distinguish between established French borrowings, which have become incorporated into the common core of the language, and those which are new and are still felt as foreign. (135)

Burnley's reasoning here is, of course, correct, although he ignores Mersand's discussions of established borrowings and Anglo-French on pages 14-20, 30-31, and 40-41. Nevertheless, Burnley's objections do not affect the validity of Mersand's data, the total percentages of Romance usage, variously calculated, nor his conclusions about the relative chronology of Chaucer's works based on the increase in his Romance usage.

Davis also objects to Mersand's method. To Mersand's statement that "The only way to determine the nature of Chaucer's vocabulary is to count every word he used, to investigate the etymology of every word, and, finally, to arrange, add and compute carefully" (37), Davis responds,

These last activities should not come "finally" at all. The significance of Romance words varies infinitely. We need to know not only the bare fact of etymology but the associations and status of every word, and whether specific applications of it would seem to contemporary hearers in any way out of the ordinary. (71)

In his emphasis on Mersand's "finally," Davis tries to distract the reader from his earlier pronouncement about the importance of investigating the etymology of every word - exactly what he has done throughout his essay. It is no doubt true that Davis, with the resources of The Middle English Dictionary now available, has far more etymological information than Mersand had in the 1930s, but this bounty does not weaken Mersand's method.

As Davis points out, the data in The Middle English Dictionary do supersede those in the first edition of The Oxford English Dictionary (which Mersand had at his disposal), and this limitation should not be overlooked. Yet this new information does not influence Mersand's two conclusions with which this article is concerned: the increase in Chaucer's Romance vocabulary and the use of these increasing percentages to date his works. Facts not available in 1939 have certainly altered some of the accepted dates of composition, yet, in general, most of the dates suggested by earlier scholars with which Mersand agrees, such as the Prologue to The Legend of Good Women and The Parlement of Foules, have since been corroborated by other evidence.

The Second Nun's Prologue and Tale has five parts of quite different lengths. As Helen Cooper notes, in the Ellesmere and Hengwrt manuscripts the Prologue is divided by the insertion of two Latin titles, "Invocatio ad Mariam" and "Interpretacio nominis Cecilie quam ponit Frater Jacobus Januensius in Legenda" (359). These insertions divide the Prologue into three parts of differing lengths: (Pr1) Introduction, lines 1-28; (Pr2) Invocation to Mary, lines 29-84; and (Pr3) Interpretation of the name Cecilia, lines 85-119. As Sherry Reames has shown ("The Sources"), the Tale proper comes from two different versions of the Vita of St. Cecilia, and the break after line 343 produces two parts of relatively similar lengths. Part 1 of the Tale (T1), lines 120-343, has 224 lines, and Part 2 (T2), lines 344-553, has 210.

For this study, first I downloaded the five-part poem from the Labyrinth website onto the spreadsheet in Appendix I, identified the Romance words, and calculated the percentage of Romance words per line. These line-by-line percentages appear on the spreadsheet in Appendix I solely to illustrate a point in the following paragraphs. Then I analyzed the Romance vocabulary in two steps: first, the preparation of the initial data to be used in the statistical analysis, and second, the production of confidence intervals for all five parts of the Second Nun's Prologue and Tale. Confidence intervals, which indicate the likely range of numbers within which data lie, were used in this analysis as a tool in comparing the percentages of Romance vocabulary of the five parts.

In the first step, I have followed the approach for evaluation of evidence in content analysis recommended by Barron Brainerd in Weighing Evidence in Language and Literature (128). This approach uses blocks of poetry lines instead of individual lines to produce the initial or "raw" data for the subsequent statistical analysis. The three parts of the Prologue, Pr1, Pr2, and Pr3 are so much shorter than the two parts of the Tale, T1 and T2, that six-line blocks are used for the analysis of the parts of the Prologue and twenty-line blocks for the parts of the Tale. For a given block, the percentage of Romance words is calculated as raw data. Thus, the five parts of the Prologue and Tale, Pr1, Pr2, Pr3, T1, and T2, produce 5, 9, 6, 11, and 10 blocks of data and can be seen in Appendix II. For each part, these blocks of raw data are used to calculate the mean (or average), indicated by [Mu], and also the standard deviation (SD), the quantity that describes the distribution of data around the mean.

The purpose of using blocks of lines is to produce data with small standard deviations. Both the number of lines in a given block and the number of blocks should be as high as possible in order to obtain the smallest possible standard deviation. Otherwise, the amount of scatter (or dispersion) in data, corresponding to higher values of standard deviation, would render the statistical analysis useless for deriving clear conclusions. Thus, in this case, larger sections of poetry yield more precise statistical parameters which in turn will produce conclusions with higher probability. This is analogous to the smaller margin of error in public opinion polls: when larger populations of people are polled, poll results contain smaller margins of error. If, in this analysis, lines were used instead of blocks of lines, the amount of scatter would be too large to produce any meaningful analysis and would yield no firm conclusions. This is obvious in the line-by-line percentages in Appendix I, in which the percentages range from 0 to 60. In order to further reduce the amount of scatter, I have followed a common practice in statistical analysis and discarded the lowest and the highest data in Pr2, T1, and T2. This procedure will have little effect on the mean ([Mu]) but will produce smaller values for the standard deviation. Pr1 and Pr3 have too few raw data to undergo this treatment.

In the second step, I produced confidence intervals of varying probabilities for all five parts of the Prologue and Tale. It would, of course, be easier to use the mean ([Mu]) as a tool in the comparison of the percentages of Romance words in the five parts. However, in Statistics in Language Studies, Anthony Woods, Paul Fletcher and Arthur Hughes explain that a confidence interval is more accurate than a point estimator, e.g., [Mu], which "is only a single number calculated from a sample of data and used to estimate a population parameter," and "is of limited usefulness by itself. It is therefore preferable to provide estimates which take into account explicitly this sampling variability and state a likely range within which the population value may lie. This is the motivation behind the idea of a confidence interval" (96).

Confidence intervals can be calculated at any percentage of confidence or probability, but analyses with the highest possible probability are, of course, the strongest. In an ideal statistical sample with a large amount of data all distributed symmetrically around the mean, various percentages of confidence are defined as follows: in a 68% confidence interval, 68% of all data lies in the interval (or range of numbers) of [Mu]-SD to [Mu]+SD; in a 95% confidence interval, 95% of all data lies in the interval [Mu]-2SD to [Mu]+2SD, and in a 99% confidence interval, 99% of all data lies in the interval [Mu]-3SD to [Mu]+3SD. Stated another way, in 68 cases out of 100, or 95 out of 100, or 99 out of 100, the data will fall within the respective intervals. These formulas show why smaller standard deviations reflect data located in a relatively narrow interval (or range) centered around the mean. For a non-ideal statistical sample such as this, the confidence interval also depends on the size of raw data available. The tables in Figures 1-3 show the confidence intervals of percentages of Romance usage in the five parts of the Tale with 90%, 70% or 60% confidence, that is, in nine, seven, or six cases out of ten. In the first part of the Prologue, lines 1-28, for example, in nine cases out of ten the percentages of Romance vocabulary will fall in the interval between 4.5 and 24.2%.

Mersand finds an 11.1 percent Romance usage in the Second Nun's Prologue and Tale (98) and points out the similarity in the number of Romance words per line of the Second Nun's Tale and The Parlement of Foules, an acknowledged early work (97). He does not distinguish, however, between the two parts of the Tale, even though he is aware that they have two different sources (48). Mersand also acknowledges Carleton Brown's conclusion that at least part of the Prologue was written later than the Tale (71), but he ignores the poem's internal divisions. When the percentages of Romance vocabulary in all the parts of the Prologue and the Tale are analyzed separately, as I have done, great differences among the percentages emerge. Figures 1-3 include graphs showing these intervals.

[Mu] = Mean; SD = Standard Deviation Pr1: Introduction; lines 1-28: [Mu] = 14.4%; SD = 7.0% Pr2: Invocation to Mary; lines 29-84: [Mu] = 14.9%; SD = 5.0% Pr3: Interpretation of the name Cecilia; lines 85-119: [Mu] = 11.2%; SD = 5.7% T1: Tale Part 1; lines 120-343: [Mu] = 7.5%; SD = 1.7% T2: Tale Part 2; lines 344-553: [Mu] = 11.4%; SD = 1.9%

90% Interval Comparison: T1 vs T2: No Overlap 89% Interval Comparison: T1 vs Pr2: No Overlap 70% Interval Comparison: T1 vs Pr1: No Overlap

(1) 90% confidence intervals to compare T1 and T2: %Romance vocabulary in T1: 5.8 to 9.3 ([Mu] = 7.5) %Romance vocabulary in T2: 9.3 to 13.2 ([Mu] = 11.4)

Thus the percentage of Romance words in the first part of the Tale (T1) and the second part of the Tale (T2) have a difference of 3.9 with 90% probability (i.e., there is no overlap in their 90% confidence interval).

(2) 89% confidence intervals to compare T1 and Pr2: %Romance vocabulary in T1: 5.9 to 9.2 ([Mu] = 7.5) %Romance vocabulary in Pr2:9.2 to 20.5 ([Mu] = 14.9)

Thus the percentage of Romance words in the first part of the Tale (T1) and the second part of the Prologue (Pr2) have a difference of 7.4 with 89% probability (i.e., there is no overlap in their 89% confidence interval).

(3) 70% confidence intervals to compare T1 and Pr1: %Romance vocabulary in T1: 6.5 to 8.6 ([Mu] = 7.5) %Romance vocabulary in Pr1: 8.7 to 20.0 ([Mu] = 14.4)

Thus, the percentage of Romance words in the first part of the Tale (T1) and the first part of the Prologue (Pr1) have a difference of 6.9 with 70% probability (i.e., there is no overlap in their 70% confidence interval).

With high confidence, therefore, it can be concluded that the difference in the percentage of Romance vocabulary between the mean of the first part of the Tale, lines 120-343, at 7.5%, and that of the second part of the Prologue, the Invocation to Mary, lines 29-84, at 14.9%, is 7.4. This difference is the largest among the five parts of the Prologue and Tale, and, in accordance with Mersand's thesis, suggests that the first part of the Tale is the earliest of the five parts Chaucer wrote, and that the second part of the Prologue, the Invocation to Mary, lines 29-88, is the latest. The application of Mersand's thesis to my data produces the following chronology of the parts of the Second Nun's Prologue and Tale, with the earliest first:

T1: Tale Part 1; lines 120-343: [Mu] = 7.5% Pr3: Interpretation of the name Cecilia; lines 85-119: [Mu] = 11.2% T2: Tale Part 2; lines 344-553: [Mu] = 11.4% Pr1: Introduction; lines 1-28: [Mu] = 14.4% Pr2: Invocation to Mary; lines 29-84: [Mu] = 14.9%

These results invite an analysis of the range of the differences among the means to establish the relative chronology of the parts of the Prologue and Tale. Rigorous statistical analysis supports these three chronological classifications:

earliest: T1, the first part of the Tale middle: Pr 3 and T2, the third part of the Prologue and second part of the Tale latest: Pr1 and Pr2, the first and second parts of the Prologue

When Mersand's theory is applied, these numbers suggest that Chaucer finished the poem by writing the second part some considerable time after the first. That is to say, this lexicon-based chronology supports Reames's opinion about the relative date of the second part of the tale: "I have long felt that Chaucer might have written the second half of the tale considerably later than the first half, given certain stylistic differences between the two" ("Recent Discovery" 347n29).

If Chaucer's Romance usage increased at a steady rate throughout his writing career, and if no other factors affected this increase, dating the writing of the parts of the Second Nun's Prologue and Tale and most of Chaucer's other works would be a straightforward mathematical procedure. However, as readers of Chaucer know, few of mortals' works are steady or constant. The frustration Mersand occasionally shows indicates he might have had to learn this lesson again and again. Throughout his book, he emphasizes factors which affect Chaucer's Romance usage such as "source, metre, period of composition, and subject matter" (138). A close look at Mersand's numbers also provides a warning to treat these percentages with caution.

The percentages of Romance usage in each of Chaucer's works from lowest to highest among the sixty titles Mersand analyzes range from 5.9 in "Chaucers Wordes unto Adam, His Owne Scriveyn" to 24.0 in Womanly Noblesse (153). Mersand points out that

Chaucer's poems exhibit a wide variation in this respect. We have the Book of the Duchess with a percentage of 7.14 per cent and Womanly Noblesse with 24 per cent. . . . In some cases the rime scheme will determine the percentage. Thus the excessively high percentage of 24 per cent the largest in all of Chaucer's works - in Womanly Noblesse, is occasioned by the necessity for finding thirty-one words that will rime with -ance or -esse. (82-83)

Twenty-four per cent is indeed rare, for the highest to appear in the Canterbury collection is 14.9, in the Parson's Prologue and Tale, which, as Mersand notes, "is replete with theological terms of Romance origin" (83). Only the Parson's Tale and the Clerk's Prologue and Tale, with 15.4% Romance usage in Mersand (75-77), have a higher Romance usage than my 14.9% for the second part of the Second Nun's Prologue (Pr2).

The ten titles with the lowest percentage of Romance usage are "Chaucers Wordes unto Adam, His Owne Scriveyn" at 5.9; "Proverbs" at 6.9; "A Complaint to his Lady" at 7.0; the Reeve's Tale at 7.0 (plus the Prologue at 7.5); The Book of the Duchess at 7.2; Troilus and Criseyde at 8.5; the Miller's Tale at 8.5 (plus the Prologue at 8.6); An Amorous Compleint at 8.7; the Prologue to The Legend of Good Women at 9.3; and the prologue to the Wife of Bath's Tale at 10.2. This evidence contradicts the generally accepted rough chronology of Chaucer's works, as three titles on this list are almost universally accepted as some of the latest Chaucer wrote: the tales of the Reeve and Miller and the prologue to the Wife of Bath's Tale. Mersand struggles unsuccessfully to explain these low percentages (82, 109-11); Benson suggests the cause of the anomaly to be an oral source (842), but Cooper's explanation seems the strongest. She concludes that "the words depend on the [Miller's] Tale's relationship to the Knight's," which is one of "marked contrast" (97-98), implying that by the time he writes the tales of the Miller and Reeve and the Wife's Prologue, Chaucer is in such firm control of his diction that he is able to modify it to characterize the speaker. He knows that no real miller would use "holiday and lady terms," nor a reeve or a west-country cloth maker highly Frenchified diction.

The evidence does establish that the second part of the Second Nun's Tale was written substantially later than the first. Happily, the date of the first part can be roughly calculated because it is mentioned in another poem, one with a generally accepted date. Among Chaucer's titles listed by Alceste in the Prologue to The Legend of Good Women is a "lyf of Seynt Cecile" (F. 426, G. 416), which establishes that Chaucer must have started translating the Vita of Cecilia before 1386-7 (Cooper 358). The low percentage of Romance usage found in the first part of the Tale, 7.6%, corroborates the early date.

The similarity of the Romance percentage in The Book of the Duchess (7.2%) to that in the first part of the Second Nun's Tale also suggests an early date, for Duchess is usually dated early. The evidence for an early date for the first part of the Tale is also strengthened by the similarities in percentages in the other works written in rhyme royal. Because of the reference to Troilus in the Prologue to The Legend of Good Women, Benson dates the Legend to not much earlier than 1385 (1060), and Troilus to 1382-85 (1020). Mersand devotes several pages to a discussion of the rhyme royal poems (97-99), and concludes that when Chaucer translates from a Latin source into English (as in the tales of the Clerk and Second Nun), he uses a "higher percentage of Romance words in rimes than when he was translating from a French source" (97), as in the Man of Law's Tale. In discussing the Romance use in Troilus, which has an average of 8.5 percent, Mersand comes to the interesting conclusion that Chaucer uses fewer Romance words when he translates from Boccaccio than when he writes his own lines (106). This apparent anomaly is, however, understandable in light of the self-conscious tendency of a translator to prefer the non-Romance word when translating from a Romance source. It is also entirely consistent with John Fisher's conclusion that Chaucer's

poetry became more complex and this complexity was made possible by the reduction of his linguistic self-consciousness and the use of more and more words from the administrative, learned, and cultural French with which both he and his audience was so familiar, until he ended up in The Canterbury Tales with a French vocabulary of 51.8% [static total, not usage]. (43)

When compared to the percentages Mersand gives for individual tales, the 7.5 percentage of Romance usage I found in the first part of the lyf is the second lowest in the Canterbury collection, excepting only the Reeve's Prologue and Tale at 7.0 (75-77). The second part of the Prologue, the Invocation to Mary, at 14.9% has the highest percentage except for the tales of the Clerk and Parson. My results therefore show that the percentage range of Romance usage in these two parts of the Second Nun's Prologue and Tale spans very nearly the entire range found in the complete Canterbury collection. Thus the writing of the five different parts of the Second Nun's Prologue and Tale probably spans almost the entire interval Chaucer was at work on the whole collection, and perhaps even longer; Chaucer may have started the lyf quite a while before he even began thinking of the Canterbury frame, and the latest part of the Second Nun's Prologue and Tale, the second part of the Prologue, is statistically almost identical to that of the tale with the highest Romance usage - the Parson's - which is literally the final tale.

Mersand finds that fourteen of the tales and the General Prologue have a Romance usage between eleven and twelve percent, and he calls these tales "typical of Chaucer at this point in his literary career" (83); he also finds a range of about ten per cent in the tales of the. Nun's Priest, the Franklin, the Canon's Yeoman and the Manciple (83). The last two are of special interest because the Canon's Yeoman's Prologue and Tale shares Fragment VIII (G) with the Second Nun's Tale, and the Manciple's Tale undoubtedly follows in the next fragment. According to Mersand, the Canon's Yeoman uses a 10.8 percentage of Romance words, while the whole of the Second Nun's Prologue and Tale uses 11.1, and the Manciple 10.4 (77). Mersand's figure for the entire Second Nun's Prologue and Tale is a little higher than those of its adjacent tales, but my figure for the second part of the Second Nun's Tale, 11.4 per cent, is almost identical to the those of the tales of the Canon's Yeoman and the Manciple, other works at the very end of the Canterbury collection.

The striking similarity of the Romance percentage of the second part of the Second Nun's Tale to that of the The Canon's Yeoman's Prologue and Tale, 11.4 and 10.8 respectively, indicates that the two works were probably written at very nearly the same time and perhaps as companion pieces. In fact, this similarity challenges Mersand's insistence that the language of the sources affects the Romance percentages. The source of the tale told by the Canon's Yeoman, in Benson's opinion, most probably is personal experience, derived from Chaucer's work in 1390 at the King's Chapel at Windsor where a canon was reputed to be a teacher of alchemy (947-48). Both the Yeoman's tale and the second part of the Nun's tale were written at the very end of Chaucer's career, when, as Fisher suggests, he was far less likely to be influenced by the language of his sources and far more likely to use diction appropriate to his speakers.

This rough dating of the parts of the tale may explain why Chaucer uses two sources. The evidence indicates that very late in his career, in the mid-1390's, Chaucer found another source to complete the lyf which he had started translating many years before, in the early 1380s, about the time when he was released from charges of rape by Cecilia of Chaumpaigne. In addition, a recently discovered document related to Cecilia's release reveals an attempt by Chaucer to hide evidence about the specific charge against him that appeared in the frequently consulted coram rege rolls (Cannon 89-94), an endeavor typical of Chaucer's avoidance strategy. I propose that Chaucer began translating the lyf in the wake of Cecilia's release to deflect negative reactions by his family, among his fellow civil servants, and at court, and then, in the 1390s, he finished the poem and put it in the mouth of a nun to provide counterpoint to the Canon's Yeoman's Tale in the Canterbury collection.

Chaucer's familiarity with avoidance strategies to deflect negative reactions is evident in the fictitious cause given for the fictitious poet's writing of The Legend of Good Women, a work already connected to both the Cecilia poem and Troilus and Criseyde. In the Prologue to The Legend of Good Women, the God of Love charges Chaucer with heresy against the law of Love in his telling of the story of how Criseyde forsook Troilus (Prologue G 264-6). Chaucer avoids the God of Love's initial harsh sentence through the intervention of Alceste, who convinces the God to agree to a reduced penalty, a writing project of the legends of good women. Through this penalty, Troilus and Criseyde is linked to the Legend as the cause of its writing, and, in the Legend, Alceste mentions Cecile's lyf Both Troilus and the lyf of Cecile thus must precede the Legend, and perhaps closely.

Troilus and Criseyde and the Prologue to The Legend of Good Women, also associated with the first part of the Cecilia poem because of similar Romance percentages, seem to be thematically linked to the Cecilia poem. Is the aim of the project of writing poems about good women a penance for his sins against them, echoed in Chaucer's translation of the life of Saint Cecilia? And why does he choose St. Cecilia? Out of the hundreds of saints' legends, hundreds of saints' lives and hundreds of saints' passions, why does Chaucer choose to translate the life of a virgin martyr, Saint Cecilia? Is it just a stunning coincidence that the one saint's life Chaucer writes concerns a virgin martyr with the same name as the woman who undoubtedly caused him great personal and financial pain when she released him from all charges "in respect of her raptus as well as of any other matter" (Benson xxi)?

Could the lyf of Seynt Cecile be a self-imposed literary penance resembling that assigned to Chaucer the poet by Alceste in the prologue to the Legend? Could this penance be undertaken in a truly remorseful spirit? Or could Chaucer's praise of Cecilie's virtue be consciously or unconsciously less than genuine? If so, it is similar to what Robert Frank called "Chaucer's mock role as a sinful penitent performing an act of penance" (210). Nothing in the entire Prologue and Tale is offensive or even disrespectful except, perhaps, for the line describing Cecilie's wedding night: "The nyght cam, and to bedde most she gon / With hire housbonde, as ofte is the manere" (VIII.141-42).(2) Literary penance for crimes against women was not exactly the rage, but as Derek Pearsall points out, "the pretext for writing a poem is itself a court game of a kind that Machaut had provided a precedent for in his Jugement dou Roy de Navarre; there too the poet is accused of defaming women, is arraigned before a lord sympathetic to ladies, offers his defense, submits to his literary penance" (191). Clearly influenced by Machaut's work in his other poems, Chaucer embraces this courtly position in the Prologue to The Legend of Good Women, and might have seen its usefulness earlier in his situation with Cecilia.

The conventions of Ricardian poetry might also explain why Chaucer would translate a poem about the namesake of the woman who caused him real harm. Indirection and/or the art of obliqueness abound in conventional Ricardian poetry, and a pattern of avoidance behavior has been found in Chaucer's life and works by R. T. Lenaghan, especially in "Complaint of Chaucer to his Purse" (45-46) and Envoy to Scogan (46-61), and by John Scattergood in Lak of Stedfastnesse (469-75). Scattergood concludes that conventional genres sometimes serve Chaucer as covers for personal concerns: "It seems to me that many of Chaucer's shorter poems are genre pieces in which personal statement emerges by way of a treatment of conventional matters[;] a traditional poem is invested with particular significance" (470).

Lenaghan and Scattergood have also shown how genre pieces, at least in Chaucer's hands, can develop into something quite different, as Envoy to Scogan demonstrates. It is certainly a begging poem, and in which Chaucer asks for help from Scogan, a fellow civil servant, with a problem common among civil servants, arrearages in payment of royal annuities. Six stanzas pass before the final Envoy, where Chaucer finally gets around to business. The poem begins, as Lenaghan notes, with the persona's "deadpan assertion of the preposterous" (48), that the recent heavy flooding has been caused by Venus' tears because you, Scogan, have given up your lady. Cupid will not take his usual revenge on fat old men like us, and, although you will say I am joking, you, Scogan, should remember me where it will help, that is, with the king. As this bare summary indicates, Chaucer's plea for financial help is easy to miss, and the summary includes none of the courtly finery. To make the poem even more dense, Chaucer uses the conventions of no fewer than four genres - the complaint, the ballade, the lyric, and the envoy or letter. Finally, the speaker shamelessly uses Scogan's unnamed and never-described lady merely as a literary conduit in his attempt to forge a relationship to Scogan, thence to the king, and ultimately to cold cash, assuming, of course, that the lady wasn't always just a fiction. This little poem does in fact raise the suspicion that in Chaucer's poetry the lady is always a fiction, a pretext for initiating a communication, a conduit through which homosocial desire is fulfilled, a pretext similar to that found in Renaissance poems by Eve Sedgwick in Between Men: English Literature and Male Homosocial Desire. Chaucer's petition through the agency of Scogan's "woman" is consonant with the male homosocial desire Sedgwick finds in Shakespeare's later sonnets and patriarchal discourse in general, "a form of desire to consolidate partnership with authoritative males in and through the bodies of females" (38). Chaucer's plea is through a woman, but to Scogan, an authoritative male, or at least a male with a position at court. The poem also displays a secondary characteristic of such literary traffic in women, that the "commodification" of women in this system makes the specific identification of any one woman irrelevant. Finally, Envoy to Scogan provides a clear example of Sedgwick's conclusion that, in patriarchies, "Levy-Strauss's normative man uses woman as 'a conduit of a relationship' in which the true partner is a man" (26).

With strategies similar to those in Scogan, "The Complaint of Chaucer to His Purse" baffles expectations when the persona addresses his purse as his lady, "dere" (2), but "light" (3), which is, according to Lenaghan, "a condition as deplorable in ladies as in purses, and with probably more than a knowing wink at the poetic meaning of 'purse' as vagina - which he wishes, again and again in the refrain, would be 'hevy agen, or elles moot I dye'" (46). The proximity of "purse" and "dye" suggests that Chaucer is unable to resist making that time-honored misogynistic reference, even when it only tangentially serves his financial purpose. In such ways, Chaucer's unorthodox uses of genre and convention in these two poems disguise his true purpose: dealing with conflict via strategies of deflection and indirection.

Evasion, deflection and oblique exploration, Chaucer's characteristic strategies for facing a conflict, certainly can be found in the ways - legal and literary - he deals with the repercussions of Cecilia's release. The original purpose of the lyf of Seynt Cecile may echo that of the Prologue to The Legend of Good Women as an oblique petition poem requesting relief, albeit a mock petition for an historically documented release, from charges of a crime in case his offer to settle out of court fails. T. F. T. Plucknett's article about the legal meaning of "raptus" and the evidence in the two 1380 documents referring to Cecilia explain that "raptus" almost certainly meant rape (34). As Plucknett reconstructs the events (some not documented), Cecilia threatened to charge Chaucer with rape, but they quickly came to terms - settled out of court; in the earlier document, she released Chaucer from actions pertaining to rape. Chaucer, however, did not have the agreed-upon amount at hand, and since she refused his personal bond, he had to find friends to put up bond for him (the amount of which is not in the two documents). The second document shows that the amount was so heavy that the three men had to refinance; Cecilia gave releases to Chaucer's friends, Goodchild and Grove, and they to Chaucer. Grove alone became bound for [pounds]10; at Michaelmas he paid. Chaucer must have given Grove a bond, which was paid when he sold his father's house the next year. He probably also gave Cecilia the balance at that time. No record exists of criminal action against Chaucer. We thus have records of the injured party relinquishing her right to make criminal charges - she accepted settlement out of court for a cash payment. Chaucer's settlement offer was apparently acceptable, explaining why no records of criminal action against Chaucer survive.

Why Chaucer would - even in a wryly playful spirit - translate a poem celebrating a virgin saint whose name is the same as the woman who threatened him with charges of rape is perhaps not immediately apparent, but the answer might be suggested by Chaucer's position in fourteenth-century England. Paul Strohm points out that,

As an esquire, he was situated at a particularly volatile and ambiguous point in the social structure of his day. . . . Chaucer's social grouping was . . . an uneasy amalgam of aristocratic and mercantile elements. . . . The social documents we have just considered can help us to understand the particular senses in which Chaucer's contradictory experiences of his role were inherent in the role itself. . . . Given the transversal of this group of knights and esquires by conflicting vertical and horizontal allegiances, Chaucer nevertheless was situated at a more than ordinarily ambiguous place in this group. (10-11)

Given this ambiguous if not marginal position, Chaucer was probably acutely sensitive to his social situation and the social, political, and economic repercussions of a charge such as rape filed against him. Bonnie Anderson and Judith Zinsser note that "Fourteenth-century . . . manor courts in England freed men for rape and put them in the stocks" (115). Georges Duby adds,

It should be noted that this venereal pleasure (delectatio Veneris), of which the males of these families were so proud, was only given free rein either before marriage, during their "youth" and in the prolonged state of celibacy that was the lot of most of the canons, or after marriage, when the head of the family had become widowed. . . . Throughout a man's married life, such dissolute behavior may have occurred, but we do not hear about it, for it was not considered seemly to mention it. (95-96)

Chaucer was not a twelfth-century French nobleman, but such ideal patterns of behavior often linger for centuries, and Chaucer's behavior as shown in the surviving records suggests his strenuous attempts to contain the legal traces of the accusation of rape. Fourteenth-century London was a small town of perhaps only 60,000 inhabitants (Robertson 89); Chaucer was surely known by everyone of any consequence, if only by hearsay. In such a small city, public opinion is certainly powerful, and Chaucer's attempt to hide the accusation of rape is thus understandable.

The document recently discovered in the coram rege rolls by Christopher Cannon concerning this case "was meant to withhold the very information that the original release was designed to disclose," that is, the word raptus (92). The rolls of the Court of King's Bench "were well produced, in a careful hand, and they were frequently and easily consulted," compared to the close rolls - where raptus appears - "which were seldom (if ever) read" (93). It seems very much like a cover-up, although Cannon argues that the revision

need not have been part of any subterfuge, since the substitutions . . . did not really change the legal function of the document in any way. For the same reason these changes would not necessarily lay either the clerk who made them or the person or persons who arranged for them open to the serious charge of tampering with official documents. They may well have been made with Cecilia Champaigne's full complicity. (93)

If not cover-up, subterfuge and obfuscation immediately come to mind; however, none of them seem attributable to Cecilia. This new document is a reminder of the seriousness of the offense and the heavy financial burden Chaucer entailed, but, more importantly, it points to Chaucer's need to avoid scandal, to obfuscate every trace of the charge.

Another reason for Chaucer's urgent desire to erase all vestiges of the charge is suggested by Sedgwick's conclusions about homosocial desire. In order to continue his homosocial relationships - relationships which, considering his ambiguous social situation, would be of great importance - Chaucer needed to contain the possible notoriety the case would cause. Sedgwick's work helps explain Chaucer's motives in the writing of the poem about St. Cecilia. To reassure his friends, associates, and colleagues of his standing as a male respectable enough to be desired as a male bonding partner, he needed to adopt a persona simultaneously responsible and blase toward the whole situation. And if he couldn't contain all the consequences (his mother and wife were still living), he might deflect some negative ones by making an apparently genuinely penitential translation of the Latin life of St. Cecilia, with undertones of the comic implicit in the situation. A similar deflecting strategy can be seen in his handling, or non-handling, of the Peasants' Revolt of 1381. Sheila Delany complains that "Chaucer is notoriously silent about the important social movements of his day, and most conspicuously silent about the great rebellion in 1381" (90). The only time he possibly refers to it is in the widow's safe farmyard in the Nun's Priest's Tale, in an allusion to the noise of the mockepic attempted rescue of Chanticleer:

So hydous was the noyse - a, benedicitee! Certes, he Jakke Straw and his meynee Ne made nevere shoutes half so shrille Whan that they wolden any Flemyng kille, As thilke day was maad upon the fox. (VII.3393-7)

Derek Pearsall explains this passage:

The presentation and debating of urgent social and political issues in these transposed forms may be a model of the way Chaucer domesticates into literary narrative his response to the events of the day. . . . Chaucer's immersion in the events of his day may have been no less complete [than Langland's], but his method of communicating their impact and importance was indirect, whether because of temperament, or the political caution needful to someone of his rank and position, or because of a deliberate choice concerning the materials appropriate to high-literary vernacular poetry. (151)

Chaucer uses similar evasive strategies for dealing with personal problems and conflict. He never directly refers to them; he seldom meets them head-on, but obliquely explores and documents them through the fictitious personal, social, economic and political conflicts in his poetry. The Cecilia poem was probably started not long after 4 May 1380, the date of the first known legal document concerning Chaucer and Cecilia Chaumpaigne, or even after 19 June 1381, the date in Chaucer Life-Records when Chaucer sold his father's house (Crow and Olson 1-2) in an intense effort to meet some very large but unforeseen debt. For about fifteen months, according to Pearsall,

Chaucer was exerting himself to raise money principally by gathering in outstanding debts . . . . [T]he weight of this activity suggests an attempt to meet an unexpected increase in expenditure, which may have been more considerable than that accounted for by the immediate need to repay Grove for shouldering the payment of the settlement. (136-37)

The [pounds]10 Grove gave to Cecilia, which Chaucer presumably repaid, amounted to more than twice Chaucer's annual salary at the customs. Finally, Pearsall adds, "There may . . . have been other payments to Cecily" (137).

The evidence that Chaucer began the lyf of Seynt Cecile as a mock penance for an alleged crime against Cecilia de Chaumpaigne remains circumstantial but compelling. The percentage of Romance vocabulary in the first part of the Tale indicates a date of its writing close to that of his problems - legal, financial and otherwise - with her; his puns on names, even perhaps his own,(3) and of contemporary women - "faire White" for Blanche of Lancaster - suggest he was liable to write poems with thinly disguised allusions to acquaintances. In addition, the lyf is linked in several ways to the Prologue to The Legend of Good Women, primarily to what R. W. Frank, Jr. calls "Chaucer's mock role as a sinful penitent performing an act of penance" (210). Also, the poem's purpose echoes, albeit mock-seriously, the literary penance Alceste gives Chaucer the poet in the Prologue to The Legend of Good Women for his past literary offenses against women:

Now wol I seyn what penaunce thow shalt do For thy trespas, and understond it here: Thow shalt, while that thow lyvest, yet by yere, The moste partye of thy tyme spende In makynge of a gloryous legende Of goode women, maydenes and wyves, That were trewe in lovynge al here lyres . . . (Prologue G 469-75)

As if in answer to a similar request, the speaker in the Prologue to The Second Nun's Tale begins, "I have heer doon my feithful bisynesse / After the legende in translacioun / Right of thy glorious lif and passioun" (VIII. 24-26). These poems are also connected by Alceste's defense of Chaucer - that he had written a lyf of Seynt Cecile - against the God of Love's charges in the Prologue to The Legend of Good Women.

Further light is shed on the history of Chaucer's writing of the lyf of Seynt Cecile in V. A. Kolve's discussion of the iconography of St. Cecilia in works of art dating to Chaucer's day (132-51). Important to this poem is the early-fourteenth-century Tuscan altarpiece which Chaucer could have seen on his trip to Florence in 1373. It is typical of the representations of St. Cecilia in showing her martyrdom in a bathtub rather than the historically correct hypocaust, a room in a Roman villa with a space under the floor warmed by a furnace. An image of the saint in majesty in a large central panel dominates the Tuscan altarpiece, flanked by a sequence of narrative, sexually suggestive images. The marriage banquet is followed by the scene in the bridal chamber where Cecilia explains to her husband that her angel forbids that her body be touched in a carnal way. The huge but empty marriage bed significantly dominates this scene. In the final scene, her martyrdom, Cecilia stands in profile, naked in a tub heated by a fierce fire, raising her arms in prayer. Her executioner is about to strike. Kolve claims that most of the depictions of the Cecilia story show her death in a similar fashion, which he describes as erotic:

[T]he martyrdom in a bath is given to [Chaucer] by his sources. All versions known to me imply the symbolic identity of lechery with heat and fire. . . . Saint Cecilia . . . remained in that bath as if in a cold place . . . . But Chaucer takes this one step further, to internalize and psychologize the cold. in saying that she "sat al coolde and feelede no wo" he focuses not on the place as it seemed to her, but on how she felt within herself. . . . The coolness of Cecilia's chastity protects her from the fire of the bath as absolutely as it protected her from Valerian's demands on the night of their wedding. Saint Cecilia in a bath "al coold" represents a condition beyond the reach of carnal heat or physical fire. (143-49)

In this context, it is important to remember that Chaucer writes the lines about Cecilia's coldness many years after he first begins translating the Vita of Saint Cecilia. The erotic appeal of the Tuscan altarpiece might have served as one of the first connections Chaucer saw between Cecilia the saint and Cecilia the London baker's daughter. In a wryly playful spirit, and perhaps with a wink to his fellow civil servants, Chaucer probably begins the lyf as a spoof of a literary penance for his sins against Cecilia, celebrating her purity. Other work intervenes and, as with several other poems, including The Legend of Good Women, he never finishes it. Later in the 1390s, he takes a second look at the unfinished poem when he is adapting other earlier works for positions in the Tales. Since he no longer has his first source, he uses what is short and easily available: the Franciscan version Reames recently found. By the time he begins finishing it, the percentage of his Romance vocabulary has nearly doubled, he needs, as Russell Peck has noted (17), a counterpoint to the Canon's Yeoman's Tale, and he sees the possibilities of the other kind of flame, the fire of charity ("And brennynge evere in charite ful brighte" [VIII.18]). Finally, this identification of Cecilia de Chaumpaigne as the "inspiration" for the Cecilia poem explains why the one and only time Chaucer writes a saint's life, he chooses one about a celibate martyr with erotic associations, St. Cecilia.

Strategies of indirection and obfuscation, characteristics of Chaucer's reaction to conflict in other poems, emerge when he begins the lyf to deflect the problems and to obviate the negative repercussions of the Chaumpaigne release. Chaucer starts the poem with a mock-penitential purpose early in his career when he self-consciously chooses vocabulary with an artificially low Romance element. He completes the lyf much later, when he is comfortable using his normal diction with its nearly equal percentage of Romance and native words, but the differences in the language divulge the long interval which had elapsed between the dates of the poem's beginning and end, dates which produce tell-tale differences in Romance proportions.


I wish to express my gratitude to Dr. Bagher Bahavar for his help with the statistics in this article.

1 Categories 1 and 5 look similar but are not; number 1 calculates the number of Romance words per line in the total vocabulary of each work, while number 5 counts the number of Romance words used per line, even repetitions.

2 All quotes from Chaucer are taken from Benson, The Riverside Chaucer.

3 Lester Matheson suggests that Chaucer plays on his family name Malyn (or Malin): "Perhaps Geoffrey Chaucer . . . used Malyne [for the name of "that swete wight," the miller's daughter, in the Reeve's Tale 1.4236] and Malle ["a sheep that highte Malle" in the Nun's Priest's Tale VII.2831] as typically self-deprecatory references to his own family background" (182).

Works Cited

Anderson, Bonnie, and Judith Zinsser. A History of Their Own. Vol. 1. New York: Harper & Row, 1988.

Benson, Larry D., ed. The Riverside Chaucer. 3rd ed. Boston: Houghton Mifflin, 1987.

Brainerd, Barron. Weighing Evidence in Language and Literature: A Statistical Approach. Toronto: U of Toronto P, 1985.

Burnley, J. David. A Guide to Chaucer's Language. New York: Macmillan, 1986.

Cannon, Christopher. "Raptus in the Chaumpaigne Release and a Newly Discovered Document Concerning the Life of Geoffrey Chaucer." Speculum 68 (1993): 74-94.

Cooper, Helen. Oxford Guides to Chaucer: The Canterbury Tales. Oxford: Oxford UP, 1996.

Crow, Martin M., and Clair C. Olson, eds. Chaucer Life-Records. Oxford: Clarendon, 1966.

Davis, Norman. "Chaucer and Fourteenth Century English." Geoffrey Chaucer. Ed. D. S. Brewer. London: Bell, 1974. 58-84.

Delaney, Sheila. Chaucer's Sexual Poetics. Madison: U of Wisconsin P, 1989.

Duby, Georges. Medieval Marriage: Two Models from Twelfth-Century France. Baltimore: Johns Hopkins UP, 1978.

Fisher, John H. "Chaucer's French: A Metalinguistic Inquiry." Chaucer Yearbook 1 (1992): 33-45.

Frank, Robert Worth, Jr. Chaucer and "The Legend of Good Women." Cambridge: Harvard UP, 1972.

Kolve, V.A. "Chaucer's Second Nun's Tale and the Iconography of St. Cecilia." New Perspectives in Chaucer Criticism. Ed. Donald Rose. Norman, OK: Pilgrim, 1981. 132-51.

Lenaghan, R. T. "Chaucer's Envoy to Scogan: The Uses of Literary Conventions." Chaucer Review 10 (1975): 46-61.

Matheson, Lester. "Chaucer's Ancestry." Chaucer Review 25 (1991): 171-89.

Mersand, Joseph. Chaucer's Romance Vocabulary. Port Washington, NY: Kennikat, 1939.

Pearsall, Derek. The Life of Geoffrey Chaucer. Oxford: Blackwell, 1992.

Peck, Russell. "The Ideas of 'Entent' and Translation in Chaucer's Second Nun's Tale." Annuale Medievale 8 (1967): 17-37.

Plucknett, T. F. T. "Chaucer's Escapade." Law Quarterly Review 43 (1948): 33-36.

Reames, Sherry. "A Recent Discovery Concerning the Sources of Chaucer's Second Nun's Tale." Modern Philology 87 (1990): 337-61.

-----. "The Sources of Chaucer's Second Nun's Tale." Modern Philology 76 (1978): 111-35.

Robertson, D. W., Jr. Chaucer's London. New York: John Wiley & Sons, 1968.

Scattergood, John. "Social and Political Issues in Chaucer: An Approach to Lak of Stedfastnesse." Chaucer Review 21 (1987): 469-75.

Sedgwick, Eve. Between Men: English Literature and Male Homosocial Desire. New York: Columbia UP, 1985.

Strohm, Paul. Social Chaucer. Cambridge: Harvard UP, 1989.

Woods, Anthony, Paul Fletcher, and Arthur Hughes. Statistics in Language Studies. Cambridge: Cambridge UP, 1986.


Appendix II
Prologue Part 1 (Pr1): Introduction, lines 1-28

Line numbers per block Raw Data: (percent Romance words)

1-6 20.5
7-12 8.5
13-18 8.9
19-24 10.6
23-28 23.3

Number of blocks of raw data Mean SD

5 14.4 % 7.0%

90% Interval 80% Interval 70% interval 60% Interval

4.5-24.2 7.4-21.3 8.7-20.0 9.4-19.3
Prologue Part 2 (Pr2): Invocation to Mary, lines 29

Line numbers per block Raw Data: (percent Romance words)

29-34 14.3
3540 20.0
41-46 15.7
47-52 23.1
53-58 6.5
59-64 4.6
65-70 10.2
71-76 18.2
77-84 19.1

[High, 23.1, and low data, 4.6, discarded.]

Number of blocks of raw data Mean SD

9 14.9% 5.0%

90% Interval 80% Interval 70% Interval 60% Interval

8.9-20.8 10.7-19.0 11.4-18.3 11.9-17.8
Prologue Part 3 (Pr3): Interpretation of the name Cecilia, lines

Line numbers per block Raw Data: (percent Romance words)

85-90 20.0
91-96 13.6
97-102 7.0
103-108 6.0
109-114 14.3
114-119 6.5

Number of blocks of raw data Mean SD

6 11-2% 5.7%

90% Interval 80% Interval 70% Interval 60% Interval

4.0-18.5 6.1-16.4 7.0-15.5 7.6-14.9
Tale Part 1 (T1): lines 120-34

Line numbers per block Raw Data: (percent Romance words)

120-139 7.7
140-159 5.5
160-179 7.5
180-199 9.7
200-220 3.8
221-240 9.7
241-260 5.9
261-280 16.1
281-300 8.4
301-320 5.3
321-343 8.0

[High, 16.1, and low data, 3.8, discarded.]

Number of blocks of raw data Mean SD

11 7.5% 1.7%

90% Interval 80% Interval 70% Interval 60% Interval

5.8-9.3 6.3-8.8 6.5-8.6 6.7-8.4
Tale Part 2 (T2): lines 344-553

Line numbers per block Raw Data: (percent Romance words)

344-363 11.8
364-383 11.5
384-403 15.6
404-423 11.9
424-443 12.3
444-463 14.8
464-483 8.6
484-503 9.7
504-523 4.7
524-553 10.2

[High, 15.6, and low data, 4.7, discarded.]

Number of blocks of raw data Mean SD

10 11.4 1.9

90% Interval 80% Interval 70% Interval 60% Interval

9.3-13.2 99-12.8 10.1-12.6 10.3-12.4
COPYRIGHT 1997 Northern Illinois University
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1997 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Geoffrey Chaucer
Author:Weise, Judith A.
Date:Sep 22, 1997
Previous Article:Figuring subjectivity in 'Piers Plowman C' and 'The Parson's Tale' and 'Retraction': authorial insertion and identity poetics.
Next Article:Greimas, Bremond, and the 'Miller's Tale.' (A.J. Greimas; Claude Bremond)

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters