University of Virginia Library

IV. Authoritative F Readings

The discussion so far has established a useful starting point for identifying readings in F which are likely to be authoritative. The annotation, we suggest, had the authority of the promptbook and was in places detailed and careful; moreover, we can identify where those places are. This does not leave us in a position mechanically to determine the authority of individual readings. Textual and even literary judgements will remain at the forefront. So, for example, Q's furbish at 373 (1.3.77) is undoubtedly richer and more unusual than F's mundane furnish; Q should be followed, even though the line occurs within a corrected area. In other cases the pattern of correction, combined with the critical argument, presents a convincing case for the reading in F. But the argument from corrected areas will not prove sufficient in itself in the case of truly indifferent readings.

Whereas some scenes are almost entirely corrected and others scarcely corrected at all, 3.2 shows a fairly representative mixture. From 1359 to 1387 (ll. 1-27) and from 1416 to 1473 (ll. 61-115) it contains no evidence

of correction; the remaining areas, 1388-1415 and 1474-1580, contain variants from several categories suggesting annotation. What the tabulated list of corrected areas does not tell us is that after 1500 (l. 141) we enter a grey area where the evidence is thin and where two Q3-F agreements against Q1 suggest the absence of close annotation (1502, 1530).[44] In practice, the fifty lines after 1530 (l. 170) are indeterminate. It is worth spelling these details out, for this scene contains the most remarkable series of variants in the text.

The relineation and added Offence at 1492-3 (ll. 133-134) have already been put forward as attributable not only to the annotator, not only to the promptbook, but to the hand of Shakespeare himself. A variant of similar status allows us to plunge into the indeterminate area and claim at least some of it for the annotator:

If Q had read, let us say, waile theyr passing woes, F could be satisfactorily explained as an anticipation of presently in the following line. Even so, one would have to admit that present makes excellent sense and gives effective word-play on presently: a happy accident. Q reading as it does, such an explanation of F will not suffice. The assumption that an actor or the annotator himself would take a dislike to the Q reading and substitute that in F is fanciful, if not perverse. The logical agent is again Shakespeare himself. If anyone is likely to indulge in textual changes which appear to us inscrutable, it is the author. Yet here the change is perhaps more scrutable than it at first appears. Editors have compared the line in Q with another Shakespeare had previously written, but without drawing any inference from the comparison: One might fairly surmise that the motive for the Folio change was an author's motive: to eliminate what had been an unconscious borrowing from an earlier play (and an earlier personal style).

There may be a similar instance within 3.2:

This variant has also been treated to inattention, or at least silence. The points to observe are obvious enough. Q's and could not possibly be misread Losse, nor, as we usually understand the psychology of the process, could the variant arise through substitution. F's reading is more concrete, forceful and specific; the metre is locked firmly into the sense, where a reading of Q must either thump and or hiccough to an end. Either Q has been weakened in transmission, or F has been improved. Given the example of sit and waile theyr woes, one may set Q by Richard III 4.4.409, TLN 3200: The extent to which Shakespeare was reworking the same ideas is even clearer in the light of the following line in Richard II: 'The worst is death . . .'. The reading F supplies not only strengthens the line, but also varies it where the resemblance with Richard III is most specific. Q drops from the formulaic to the merely repetitive; F does not. It is the variants themselves that pass judgement on Q's repetitions, and suggest that Shakespeare was not satisfied with them.

If substantiveness comes in degrees (the word perhaps misleadingly suggests that it does not), there are several other Q/F variants which could be called highly substantive. Some F readings will be rejected: hand for wound (l. 139; TLN 1498) is clearly a Folio compositorial error influenced by Hands in the previous line; at 1399 (l. 43) F describes how the sun 'darts his Lightning through eu'ry guiltie hole', where Q more meaningfully and metrically reads Light; Beares for beards (l. 112; TLN 1470) is, as suggested above, an annotator's folly, an easy manuscript misreading which produces a line of delightful nonsense ('White Beares haue arm'd their thin and hairelesse Scalps'). Variants such as power/friends (l. 35; TLN 1391) and partie/Faction (l. 203; TLN 1562) require more serious attention. At first sight they could be substitutions in either direction, though not of the commonest kind. But if compositorial substitution accounts for these variants, Q is more likely than F to have introduced them. Q's power and partie share in their neutral attitude. Folio friends is not only the more specific of its pair, but when set beside substance ('in substance and in . . .') it makes a meaningful distinction between material and human resources in a way that power does not. Furthermore, power occurs twice in previous lines (l. 27 and l. 28; TLN 1387 and 1388), whereas friends is nowhere in sight. Faction, too, is more precise and expressive than Q's partie; as may also have happened at Richard III 3447 (5.3.13), one explanation for the variant would be that manuscript factiō was misread partie, an error plausible in this direction but not in reverse.[45] With 'Is not the Kings name twenty/fourty thousand

names?' (l. 85; TLN 1442), it is again F which is less easily explicable as error, for 20000 has already been cited (l. 76; TLN 1433). F's fourty continues the progressive exaggeration begun with 20000 itself; the first figure cited is twelue thousand (l. 70; TLN 1426). The Q compositor might (like modern editors) have corrected fourty as apparently a self-evident error. 'Awake thou coward/sluggard Maiesty thou sleepest' (l. 84; TLN 1441) in isolation is more difficult both to explain and to judge. Q's coward gives the easier antithesis with Maiesty, but cowards are not usually characterised by a propensity to sleep. It will be noticed that the two last-mentioned variants occur in adjacent lines; taken together, they suggest nothing so much as further instances of minor authorial alteration. Certainly the influence of Shakespeare's own hand best explains why so many significantly different variants occur in one scene—or, speaking more precisely of the valid F readings, in the corrected parts of one scene.[46] To reserve final judgement on Shakespeare's responsibility for a variant in a single instance such as 1410 (l. 55) is no contradiction of this view. Q's imperfection there is almost certainly illustrated by unmetrical and redundant off ('Can wash the balme off from an annointed King'); F's removal of the word looks like restoration, not revision or sophistication.

Only in the concentration of its attractive F variants is 3.2 unique. Examples of F giving a more specific or harder reading can be readily found elsewhere. Three F readings rejected by editors may be considered together.

In each of these lines, Q has a simple, acceptable and uncontroversial reading. One tells a tale, the wind blows, sorrow is sullen. But in each case F gives an unusual idiom which is nevertheless, to say the least, acceptable. Fall at 2305 gets away from a cliché that has already been used three lines above: tell the [i.e. thee] tales. Richard relates himself to a specific genre of narrative, the Falls of Princes, and in doing so he heightens the extent of his self-dramatization. Grew and sudden both concentrate upon a momentary transformation instead of a state of being; this more dynamically (one might almost again say 'more dramatically') gives rise to the tear, or the death caused by sorrow. At 520, Q's sullen is appropriate to sorrow as received, not as inflicted, whereas F's sudden, though a more unusual epithet for sorrow, is exactly appropriate to acts such as Richard's banishment of Bolingbroke.

Other variants have been attributed to the actors, but are much more likely to be authorial. In the line:

F's variant downe for up probably represents, as Stanley Wells suggested in the New Penguin edition, a change in the envisaged staging. Editors follow F in correcting an historical error at 2802 (5.6.8), but they do not attribute the same authority to the F correction of Woodstockes to Glousters (1.2.1; TLN 218). There is admittedly a significant difference between these Q readings in that one is an error and the other no more than an inconsistency. Elsewhere in Richard II the character is called Gloucester, and Woodstockes is probably a first thought coming early in the play, when the play Woodstock, usually regarded as an antecedent to Richard II, would be still uppermost in Shakespeare's mind. It should probably be regarded as a foul-papers inconsistency which Shakespeare later eliminated.

Two transpositions may be set side by side:

In 2032, the Q reading is metrically possible in that Henry was, in early Shakespearian usage, occasionally trisyllabic. Apart from this example and one in Richard III, all such instances are found in the Henry VI trilogy. A deliberate revision would therefore remove an outmoded metrical form. At 2421 Q has an acceptable hexameter. F gives a more compact line with a more unusual word order. Whereas a compositor might drop a word to create space—F's line fills the rule—a complete reshaping of the line for this reason is most unlikely, especially as the end of the line could have been turned up or down. The changes do not point to compositorial error in Q in that they are more extreme than one would expect, though this does remain a possible source of error. More likely, this is also an instance of authorial adjustment.

The variant at 1538 (3.2. 178), in which Q echoes an earlier line, is not unique. Remarkably similar is:

Shakespeare used exactly the same words at 2624 (5.3.122) as Q gives here. Although this has been noticed before, it has not been connected with the variant; editors have preferred to imagine that an actor or annotator introduced the F reading on religious grounds (though different views have been advanced as to why Q was found unacceptable).[47] It is more reasonable to suggest, especially in the light of the example of sit and waile theyr woes, that Shakespeare changed his text to avoid repetition. The superiority often attributed to Q's reading must be weighed against the fact that it perpetrates such a repetition, while Faith is certainly defensible on the grounds that it emphasises the supposed meaning of an equivalent word. Unlike Shakespeare, his editors are generously reluctant to find fault with the repetition; they therefore cannot admit that F could supply an authorial improvement.

One may compare the laugh/mocke variant at 1759 (3.3.171) with partie/Faction discussed above. In both instances F has the more emotionally precise expression. F's Queene (5.1.78; TLN 2340) is doubly advantageous over Q's wife in that it avoids a repetition from five lines above and encompasses the connotations of pompe as well as the concept of wife:

When Richard tells Northumberland that his observation is confusing: are the wicked men those like Northumberland, or those like Bolingbroke to whom they are obsequious? Richard intends both: that is, the love between wicked men. This is made clear in F, where men becomes friends.

Three substitutions involving units of time in fact vary in the certainty with which they can be identified as authorial alterations. At 162 (1.1.151) Q has:

In F month becomes time. It may be suspected that this again is a deliberate alteration of the text, made to remove the unfortunate suggestion of menstruation that the original line may have been thought to elicit. An apparently similar variant is yeeres/dayes (5.3.21; TLN 2517), but this variant is probably connected with the presence of heere at the end of the following line. Either Shakespeare emended a jingle out of the text, or heere stimulated a compositorial substitution in Q. Another instance may be more safely attributed to compositorial error in Q. Ure recognized that Q gives an illogical sequence in: In F the second and third terms are transposed. Or, to express the situation more plausibly, F has the required text and Q contains a compositorial transposition.

With raise/reare (4.1.145; TLN 2065) F arguably gives the less usual idiom; it is reassuring to find that the variant ocurs in a carefully annotated area of the text. A similar case is throwne/stricken (5.1.25; TLN 2286). This is not in a corrected area, but F has a more significant advantage in that it completes an irregular nine-syllable line in Q.[48] Another metrical emendation supplied by F is tell me/tell (5.3.1; TLN 2497), where Q's me anticipates my two words later (and is in any case an easy interpolation).

In the last two instances we move from F readings which possibly or probably derive from Shakespeare's revising hand to those which are much more likely to preserve an original text which Q has corrupted. The preceding discussion has established or at least suggested the validity of some F readings which editors customarily reject, but there is a persistent difficulty in distinguishing between what must be Shakespeare's second thoughts in F and corruption in Q. Fortunately we have some detailed knowledge of the work and habits of the compositor who set most of the text in Q1: Simmes's Compositor A. Alan Craven has written a series of articles analyzing the types of error this compositor was prone to introduce, and the implications of such habits for an editor.[49] Contrary to the usual assertion that Q gives an accurate and reliable text, Craven observes that Simmes's Compositor A was particularly likely to introduce error that involved a change in wording: in Q2 Richard II, for example, which he set in its entirety, he introduced 63 substitutions and 30 omissions, as against 25 literals errors. Yet when Craven comes to discuss probable compositorial error in Q1 he covers very different

ground from the present study. The reason is that Craven's evidence points consistently to the triviality of the compositor's many errors. 'The words thus affected are not usually nouns and verbs, but instead are connectives and qualifiers' ('Five Shakespeare Quartos', p. 56). Although Simmes's Compositor A was prone to substitute, he is unlikely to be responsible for the 'words of nearly equivalent meaning' noted by Ure and discussed above. The only comparable substitutions in the whole of Q2 are speak/said, fall/drop, sparks/sparkles, shock/smoke and slander/slaughter. If these were Q/F variants, it is unlikely that more than one of them would have been mentioned in this article: the first two appear completely indifferent, and the last three are self-evident errors of which the first might have been cited in passing as a metrical corruption. Simmes's Compositor S, who set the remainder of Q1, was 'much more accurate than A', though prone to errors of omission ('Five Shakespeare Quartos', p. 57,n. 9). Yet there are no indications that the variants discussed above are influenced in number or in kind by this compositor.

Assuming that the general tenor of the preceding critical comments proves acceptable, few or none of the variants discussed can be attributed to error on the part of the F compositors. But this assertion can be put on a more objective footing by considering the work of the F compositors elsewhere. Richard II was set by Compositors A and B, the latter taking the greater share of the work. Compositor B's efficiency in setting from printed copy has been examined by Paul Werstine, some of whose conclusions can be cited.[50] Although he sometimes substituted 'a more difficult word for the clear reading of his copy', Werstine's two examples—brawles/Broyles (Romeo and Juliet 91) and beastly/Beastlike (Titus Andronicus 2703)—are not remotely comparable with the variants in Richard II discussed above. Similarly, when Werstine attributes to Compositor B substitutions of 'words which are orthographically dissimilar to the words of his copy, but similar in meaning to the copy words' he is speaking in relative terms, and his examples again distance Compositor B's work from the distinctive substitutions in Richard II: smirched/smeered (Much Ado about Nothing 1796), curelesse/endlesse (Merchant of Venice 2051). These examples are drawn from a much wider body of work than his stints on Richard II. Transpositions prove to be Compositor B's rarest form of error, and interpolations almost as rare. Though he sometimes introduced metrical sophistications, he 'never added half-lines, adjectives, verbs, or expansions to his copy'. His omissions are 'too few to classify'. To summarize, the only discussed

variants that Compositor B could normally be thought likely to have introduced are the metrical emendations at 1410 and 2497; Compositor A actually set the latter.

Of Compositor A's work set from minimally-annotated printed copy we know less, for the only passage so set outside Richard II itself is just over a single page in Richard III (1569-1735). His work there introduced the following changes: five minor rewordings of stage directions, five obviously necessary corrections of copy error, two metrically-required elisions, and one minor substitution or literal error (Nor/No at 1581). His propensity to correct on his own initiative is of some interest, but there are no hints here of the kind of substitutions and other alterations found in Richard II. His work setting from manuscript copy suggests that he was more accurate than Compositor B.[51] But 3.2 occurs in Compositor A's stint, with the result that he set more of the variants attributed to annotation than Compositor B, even though A set less than half the text.

This returns us to the pattern of annotation. Only 3 (12%) of the variants we have discussed lie outside the corrected areas. These provide the slight sprinkling of exceptions to the pattern that, on the basis of the other information, one would expect. As it has not been felt necessary to exclude such readings from consideration simply because they fall outside the corrected areas, and as the annotation of the surrounding area has not been put forward as a criterion for justifying most of the F readings, the new suggested readings from F lend further weight to the interpretation of the promptbook annotation. Or alternatively, the pattern of correction gives added authority to the advocated F readings.

The present discussion does not exhaust the number of F readings which should or might be adopted. The conclusions are twofold and mutually supporting. F is a rich source of readings that are both authoritative and radically different from Q. And even in Richard II, perhaps the most straightforward instance of an F text printed from annotated quarto copy, there are unmistakable indications that Shakespeare himself was responsible for some of these variants. We see clear evidence of the author in the process of making final adjustments to his play.