University of Virginia Library

Section 1: Memorial Reconstruction in Q Henry V

To seek verifiable evidence for or against memorial reconstruction, I designed a computer-assisted analysis to help compare the two versions of Henry V.[3] My analysis is based on the fluctuating quality of Q, observed by

many scholars, for the parts of certain characters seem to correspond more closely to F than others; the analysis is designed to isolate the lines with the highest correlation in the two versions and to identify both the speakers and the characters whose actors witnessed these closely parallel lines.

To implement the analysis, I first underlined each matching word in corresponding segments of my handmade parallel text, then marked each line with a code reflecting the degree of correlation. If every word in the line appears in the parallel segment of the other text (ignoring word order, lineation, and spelling), I marked the line "A" (for all). If more than half of the words correlate, the code is "M" (for most); if half or fewer words match, I marked the line "S" (for some). Lines that paraphrase the content of a parallel segment—but contain no matching words—were marked "P" (for paraphrase), and lines with no correlation were marked "X". Next, using computer typescripts of Q and F, I transferred the correlation codes from my parallel text to the computer texts. I also marked each line with codes to identify the speakers and the characters on stage.[4]

For each of the two texts, I used a database program to isolate and count the lines that each character speaks and the lines his or her actor witnesses while on stage. In addition, I determined the degree of correlation between the two texts for these spoken and witnessed lines, using the five correlation categories, A through X. Tables A, B, C, and D in the appendix record this basic information. Then I organized the data with the help of a spreadsheet program.[5]

Tables A and B in the appendix, which record the number of lines of each character in various correlation categories, show that actors playing 26 of the 51 roles are very unlikely candidates for a reporter. Even if a reporter

had doubled in some of these roles, there is often too little information to draw any conclusions about his part in a possible reconstruction. Beaumont, Berri, and the Second French Ambassador have no lines in either version, making them the first to be eliminated. Clarence, Gebon, and the Lord have no lines in F; most of the lines spoken by them in Q are spoken by other characters in F, making them equally unlikely reporters. The Chorus, Ely, Westmorland, Isabel, MacMorris, Grandpre, Jamy, Bedford, Britanny, Rambures, Erpingham, and the English Herald have no lines in the Quarto version, though some lines have been reassigned. Court has one line in Q, only 2 in F, while the French Messenger and Salisbury each speak only 3 lines in Q (7 and 9 lines respectively in F); parallels between the lines in F and Q for each are not striking. The roles of the Dauphin, Bourbon, and Burgundy are radically different in Q, eliminating them as likely reporters. The Dauphin's role is reduced from 117 lines in F to 22 in Q; Bourbon is assigned some of these lines, as his role expands from 9 in F to 29 in Q. Burgundy's role is trimmed from 68 lines in F to only 4 in the Quarto. Warwick's and Gloucester's much smaller parts are also altered in Q: Warwick has 1 line in F, a line that does not appear in Q, but is reassigned 7 lines spoken by others in the Folio; Gloucester's part includes 5 lines in F, 11 in Q, again reassigned from the roles of other English lords in F. Thus these 26 unlikely candidates have been eliminated as possible reporters.

Table 1 shows the percentage of lines in the Folio with a high correlation to parallel lines in Q, for lines spoken and witnessed by the remaining 25 characters. We would expect that a reporter—if one existed for Q Henry V —would have remembered his own lines more accurately than those he witnessed, for it seems likely that an actor would recall his own part more fully than the words of others on stage with him. Using this criterion, I have eliminated as possible reporters the 14 characters listed toward the bottom of Table 1, for in each case the lines witnessed in the two texts are more closely parallel than those spoken by their actors. The only possible exception among this group may be Mistress Quickly, for the difference in correlation between spoken and witnessed lines is very slight. I believe her actor is an unlikely reporter, however, because of the relatively low proportion of lines in the "A" category and because her spoken lines are not more accurately rendered than those her actor witnessed.

In addition to remembering his own lines more fully than those he witnessed, we would expect a likely reporter to recall his own role with an accuracy greater than the average, shown near the center of Table 1. Orleans is at the median whereas Alice and Kate fall below both median and mean (average), making their actors unlikely reporters.

This leaves eight possible reporters out of the 51 characters: Exeter, Gower, Pistol, Nym, Scrope, the Governor of Harfleur, York, and Williams. I believe Williams is an unlikely reporter, for the proportion of his lines in the "A" category is considerably below the mean. York has only two spoken lines and witnesses only three others in F; because of the small amount of data, it is difficult to determine if his actor was a reporter, although one of

Table 1: Folio Spoken & Witnessed Lines with a High Correlation to Q (expressed as a percentage of the character's total F lines)

                                                             

	Folio	Folio	Folio	Folio
	spoken	witnessed	spoken	witnessed
	A + M	A + M	A	A
Exeter	84%	49%	57%	22%
Gower	61%	55%	40%	21%
Pistol	65%	55%	35%	22%
Nym	67%	53%	22%	20%
Scrope	100%	55%	62%	28%
Governor	100%	0%	57%	0%
York	100%	67%	50%	33%
Williams	43%	39%	11%	13%
MEAN (AVERAGE)	39%	39%	17%	17%
MEDIAN	33%	33%	10%	10%
Orleans	33%	25%	22%	9%
Alice	30%	13%	19%	4%
Kate	25%	22%	15%	7%
---------------------------------------
Quickly	59%	60%	22%	24%
French Amb.	65%	88%	24%	43%
Grey	50%	59%	8%	32%
Fluellen	48%	60%	10%	31%
Canterbury	45%	59%	24%	26%
Henry	42%	48%	19%	22%
Cambridge	40%	60%	27%	31%
Montjoy	33%	75%	13%	40%
Boy	29%	43%	12%	17%
Constable	28%	35%	10%	18%
Bardolph	27%	62%	10%	22%
French Soldier	20%	60%	0%	30%
Bates	18%	21%	6%	3%
Charles	14%	31%	2%	16%

the reporters certainly could have doubled as York. The case for the Governor of Harfleur is similar, for though his 7 lines in F are closely parallel to his role in Q, his actor witnesses no lines in the Quarto; one of the reporters could easily have doubled as the Governor, but without lines witnessed by this character in Q, there is too little evidence to evaluate.

Five characters, then, remain as major candidates for reporter—Exeter (obviously the most likely possibility), Gower, Pistol, Nym, and Scrope—perhaps also doubling the smaller roles of the Governor and York. Table 1 shows why Exeter is often mentioned as a likely reporter, for well over half of his spoken lines are virtually identical in the two versions, while 84% are closely parallel.[6] Lines spoken by the other likely reporters show less correspondence,

though still considerably more than either the median or the mean. Equally important, the proportion of closely parallel lines witnessed by all five is also well above average, a key indication that Q did not begin simply as a transcript of these players' parts.

No single actor could have played all five of these roles in Henry V: although some doubling is certainly possible, each of the five appears with at least one of the others. Exeter could double Pistol or Nym, but this seems unlikely (apart from casting difficulties), because Exeter's scenes are generally more accurately rendered in Q than Pistol's and Nym's. Gower could double both Nym and Scrope; Pistol could also double Scrope, and any but Exeter could double the Governor and York. Three actors, then, could have reconstructed the Quarto: the actor playing Exeter, along with the actors playing Pistol and Gower, doubling Nym, Scrope and possibly the Governor and York. Thus Table 1, based on a quantitative analysis of possible candidates, verifies the widely-held impression that Exeter, along with one or two others, was responsible for reconstructing Henry V from his memory of performances, a significant confirmation of the memorial-reconstruction theory. This result is particularly important in light of the recent healthy skepticism of Steven Urkowitz and Paul Werstine, among others, concerning the validity of the theory of memorial reconstruction.[7]

Just as significantly, Table 1 indicates that the version the reporters apparently knew was a script linked to the Folio rather than to an intermediate abridgment. The proportion of closely parallel lines spoken by Exeter, Scrope, and the Governor suggests that the reporters attempted to reconstruct a version similar to the Folio, apparently abridging sections of it at the same time or shortly thereafter. Exeter's part in Q retains some 84% of his Folio lines with considerable accuracy, as noted above, while all 13 of Scrope's Folio lines and all seven of the Governor's reappear with equal accuracy. The Quarto, with 1629 spoken lines, includes only 50% as many lines as does the Folio (3253 spoken lines). If the reporters had known only an abridgment, their lines presumably would have been cut in such an abridgment in roughly the same proportion as the rest of the play. But Exeter's crucial part in particular is remarkably full as well as unusually accurate, a key indication that he was working from his memory of a longer, Folio-linked script. Though the Quarto version has obviously been abridged, probably deliberately, Table 1 presents

significant new evidence that the reporters began with performances directly related to the Folio version rather than to a lost intermediate abridgment.[8]