University of Virginia Library

IV

If ever a genealogical problem needed sorting out, it is that of the relationship
between the Quarto (1600) and Folio texts of Shakespeare's 2 Henry IV. In his

Lyell Lectures of 1959, Fredson Bowers remarked that 'at present, Hamlet and 2
Henry IV appear to be very seriously snarled in controversy',[28] thus summing up
two decades of work on the latter by four prominent scholars. Although M. A.
Shaaber, Alice Walker, J. Dover Wilson, and W. W. Greg had been able to agree
that the Folio restored eight passages (presumably cuts) omitted from Q cor-
rected and otherwise altered many of Q's stage-directions and speech-prefixes,
removed its profanity and colloquial or vulgar language, added its own mis-
lineations to two in Q, contained numerous differences of other kinds, and must
have derived from a manuscript of a 'literary' character, they had not been able
to agree whether F was typeset directly from that manuscript or from an example
of Q annotated by reference to it and containing both its variant readings and
possibly some initiated by the annotator.

In a long appendix to his Variorum edition, Shaaber had supported, not
infrequently with negative evidence, the position taken by the Cambridge edi-
tors (1864) that F's source was a 'thoroughly overhauled' transcript of Shake-
speare's manuscript: he had argued, among other things, that the 32 supposed
errors shared by Q and F (and rejected by at least a majority of editors) could
be reduced to three either because they were defensible, lacked an agreed-on
editorial emendation, had a precedent elsewhere in F, or could have derived
independently from a common source. Walker, however, had put forward the
contrary view first in an article and a note, and then more fully in her book on
quarto copy for the Folio; in her usual positive and persuasive manner, she had
cited eight shared errors in wording and other 'common errors' in punctuation
and word forms, including 'maner', as evidence that F was set directly from an
annotated copy of Q. The responses to Walker's essays had been swift but differ-
ent. Wilson, who in his 1946 New Shakespeare edition had followed Shaaber, in
1952 had added at the end of his discussion of the textual problems two sentences
that suggested he had accepted Walker's conclusions (as set out in her article)
and would thoroughly revise for 'a second edition'; he cited three passages for
which notes had already been revised. Shaaber had reacted almost immediately,
employing the strategies seen in his Variorum edition to neutralize Walker's case
but adding some new observations.[29] Almost simultaneously, however, Greg had

recorded a mixed response: he had credited Shaaber's Variorum analysis of F's
so-called 'massed entries' and other features with showing that they could not
have been drawn directly from the prompt-book, and had found his case for F's
direct dependence on manuscript persuasive, but had been impressed by the
common verbal errors (which he thought likely to number more than Walker's
eight), the shared nonsense, Walker's odd 'maner' spelling, and the instances of
mislining, all of which inclined him to believe that F was typeset from annotated
Q. 'And so the dispute continues without any certain conclusion in sight', Greg
wrote somewhat prophetically, not having seen Shaaber's article, published the
same year. [30]

By the end of the 1950s, then, discussion of 2H4 in the traditional terms of
textual criticism had ended in a virtual stand-off, with the main lines of differ-
ence regarding Folio copy defined, the proponents clearly identified, and with
Greg and Shaaber agreeing independently that more study was needed, either of
the kind seen in Philip Williams' analysis of Q and F Troilus and Cressida (Greg,
p. 272) or of'the working methods of compositors' (Shaaber, 'The Folio Text',
p. 144). In the editions that have followed and in other scholarship, whether in
support of these editions or independent of them, there has been much careful
analysis of the problem as defined by Shaaber, Walker, and Greg, but (with one
exception) no significant new evidence brought forward.

Shaaber's view of Folio copy has prevailed, though in various permutations.
A. R. Humphreys' 1966 New Arden Edition set the pattern in its thorough review
of the 'exasperatingly ambiguous' evidence and its general conclusion. Running
a variation on Bowers' 1953 suggestion that F's copy was a transcript of an
annotated Q,[31] Humphreys hypothesized a scribal manuscript combining 'con-
currently' Q and an independent transcript 'showing some cognizance of stage
practice'—this despite his recognition that such a theory required 'mediation'
(transmission) of typographical details through both transcription and composi-
tors and despite his decision to prefer 'about eighty' of F's variants to Q's.[32]
David Bevington, revising Hardin Craig's Complete Works, tentatively approved
the New Arden's manuscript and Bowers' transcript as likely scenarios, without

mentioning annotated quarto, and G. Blakemore Evans, textual editor of the
Riverside Shakespeare, summarized the dispute and seemed to defer to the New
Arden's position.[33] In an overview of the play's textual condition, George Walton
Williams found no credible bibliographical links between Q and F, suggested F's
setting copy was the 'fair copy of the foul-papers made ca. 1598' which Walker
believed had been used to annotate Q, and proposed that it was 'a companion
piece to the manuscript from which the 1 Henry IV quarto was printed', both
of them made 'to prove to Oldcastle's angry posterity' that his name had been
removed from the play.[34] Peter Davison, uncomfortable with the New Arden's
adoption of 80 Folio readings, with another 100 being of equal merit to Q's, and
with F's 'excision, wholly or in part, of some twenty-five Quarto stage directions
that are superior to those remaining in the Folio', produced his Penguin edition
on the theory that F's copy was 'a transcript … made with the aid of actors'
parts, despite the trouble and expense' but 'with the Quarto at hand, an excellent
guide to the order of speeches'.[35]

More recently, manuscript as Folio setting copy has remained the preferred
scenario, but its precise character and the role of the quarto (if any) in creating it
has continued to generate multiple hypotheses. Eleanor Prosser, in a book-length
study now most valued for its observations on the compression F's text underwent
in its first quire (g) and the expansion in its second (χgg), thought that the manu-
script conflated with Q was not a transcript but Shakespeare's own foul-papers.[36]
The long-anticipated Oxford Edition, though most notable for its view that six of
the eight passages that F supplies and Q wants represent Shakespeare's revisions,
ruled out annotated Q, partly on the basis of new statistical evidence, speci-
fied a scribal copy of the prompt-book, but stipulated consultation of Q by the
scribe.[37] The New Cambridge Edition rejected many of the Oxford's arguments,

propounded its own theories regarding an ur-Henry IV play, the pre-publication
manuscripts, and the various agents involved in the transmission of their texts,
but ended up agreeing that for Folio copy the idea of an 'intermediate transcript
by an interfering scribe' was 'more plausible' than annotated Q, prompt-book, or
a 'transcript put together from actors' parts'.[38] Finally, the Oxford Shakespeare's
one-volume edition has also rejected the Oxford Edition's analysis of F's eight
unique passages, reverted to the traditional position that they were 'integral'
to the play 'from the beginning', revived Shaaber's argument (via Prosser) that
marking Q would have been 'a near impossible task', and concluded that Q and
'a post-1606 expurgated prompt-book' were 'collated in' a private literary tran-
script that later became printer's copy for F.[39]

Compositor B's speech-prefixes may shed some light on the central question
of Folio copy for this play. In his pages there are well over 50 forms that should
be useful for identifying his setting copy. On the whole, his long speech-prefixes
in these pages, which generally conflict with his tendency to set abbreviated
forms, indicate that certain Quarto forms—and a significant number of them—
somehow found their way into F. But the value and precise implications of the
evidence can only be properly assessed against the background of B's handling of
speech-prefixes throughout the Folio, especially in 1 Henry IV, and more narrowly
through analysis of the context in which the particular forms occur.

As to the larger background, B's pages of 1H4 on the whole confirm that in
the Histories (as in the Comedies and Tragedies) his long speech-prefixes, and
especially his full forms, should generally reflect the variable forms of his copy.
There is every reason to expect that such practices should have continued in B's
next play, 2 Henry IV, which was begun on sig. f6v—the forme-mate to f1, the
last page of 1H4 to be put into type—and which occupies the next two quires.
At those points in 2H4 where speech-prefixes attributable to his copy appear in
B's pages, almost all find precedents in the Quarto.

This evidence is naturally of various weights, as the earlier review of B's
work throughout F would suggest. But here its value is especially affected by the
compression and expansion which characterize the pages of this play in particu-
lar and which Hinman referred to generally as 'page justification'. Hinman has
shown that 2 Henry IV was unusually subject to such page justification because
the Histories were printed out of order, beginning with all of King John and most
of Richard II before the Comedies had been completed, then jumping ahead to
Henry V and most of the Henry VI plays, and then returning to 1H4 and 2H4 be-
fore the remainder of the Histories (the end of 3 Henry VI, Richard III, and Henry
VIII) were set. Since H5 had begun on sig. h1, the end of R2 and all of 1H4 and
of 2H4 were to be packed into four quires (d–g), and this obviously was found
to be impossible as the printing proceeded. The adjustment of type matter to
alloted pages became critical during the composition of 2H4, and consequently
the compositors were under special pressure, as copy was cast off quire by quire,
to fit text to the assigned pages, first by crowding as much of it into quire g as
possible, and then (after the decision was made to create an eight-leaf quire to
accommodate the rest of 2H4) to see to it that enough text was left to make xgg7v
a proper part-page. As the more experienced of the two typesetters, apparently,
much of the responsibility for this adjustment fell to Compositor B, who set up
not only the entire second half of the last quire (xgg5–8v)—where the play was to
be made to end part way through xgg7v, and then eked out with an epilogue (xgg8)
and a list of actors (xgg8v)—but also xgg1, which he composed as forme-mate to
his xgg8v before copy for the rest of the quire was cast off.

Following on Hinman's general suggestions regarding 'page justification' and
his more specific analysis of the production of quires g and xgg, Eleanor Prosser
has attempted to reconstruct precisely the circumstances and sequence of events
that led Jaggard first to compress and then to expand text in the two quires

that almost wholly comprise 2H4. In the process, she has indicted B for liber-
ally altering the wording of the play in order to perform his master's bidding.
Although these attempts are not entirely successful and the arguments regard-
ing B's treatment of wording particularly shaky, there can be little doubt about
the general validity of her observation that the compositors, especially B, were
under extraordinary pressures to adjust the length of lines and thus of pages for
this play.[40]

In the first quire, most of B's pages (sigs. g1–3) exhibit some compression,
though the first two he set up (forme-mates 3v and 4) appear relatively normal.[41]
In the second quire, however, sigs. χgg1 and χgg5–7 betray signs of expansion,
which ceases in sig. χgg7v, the part-page that, once reached, signalled success
in filling the eight-leaf quire. How strong was the pressure to adjust lineation
and vertical spacing to fit the text of the play into the allotted pages remains
debatable. But mechanical matters, such as the length of speech-prefixes and the
placement and spacing of stage-directions, would have been especially subject
to these pressures, when they affected the number of lines of actual text that
would occupy a column or page. These factors must certainly be considered in
an assessment of the quality of B's speech-prefixes as evidence of the nature of
his setting copy for 2 Henry IV.

In about 40 cases B's full speech-prefixes indicate that his copy for 2 Henry IV
must have had forms identical to those in Q. Some of the best evidence is in
quire g (see table 1). In sigs. g3v and g4, B's initial pages of the quire, it would
appear that B's work was generally less affected by considerations of linear spac-
ing than elsewhere in the play, and there are proportionately more full forms
here than in the other pages of this quire.[42] His general preference for 'Prin.' B
abandons six times for the full form, which Q has uniformly throughout II.ii. A
few of these forms may be affected by justification of the lines in which they occur
(TLN 873NL most likely, and perhaps 852L and 888L). The others, however, seem
clearly to have been set independently of such concerns (800L, 879, 898L). The

Table 1. B's full speech-prefixes (Folio quire g)

                   

Quire	TLN	Folio	Quarto	Riverside Act.Sc.Line
g3	616	Hostesse.	Hostesse	II.i.1
g3	622	Snare.	Snare	II.1.7
g3v	800L	Prince.	Prince	II.ii.9
g3v	847L	Pointz.	Poynes	II.ii.65
g3v	852L	Prince.	Prince	II.ii.70
g4	873NL	Prince.	Prince	II.ii.92
g4	879	Prince.	Prince	II.ii.98
g4	888L	Prince.	Prince	II.ii.106
g4	898L	Prince.	Prince	II.ii.117

single 'Pointz.' in sig. g3v (847L) recalls B's practice in 1H4.[43] But what is most
intriguing about this speech-prefix is that this single instance of B abandoning
his standard 'Poi(y)n.' occurs where Q fails to have the 'Poy(i)nes' speech-prefix
otherwise found throughout this scene. The parallels with 1H4 are striking and
tend to confirm the view that B's 'Pointz.' in 2H4 reflects the peculiar state of his
Quarto copy at this point, in much the same way that it did earlier.

As might be predicted, there are fewer long speech-prefixes in the remaining
pages of quire g, which were subjected to the crowding already discussed. Yet if
B in fact was trying to compress the text while setting sigs. g1–3, then the two
full speech-prefixes that do occur in these pages (those at 616 and 622) are very
good evidence of his copy's influence on the forms he set. Both are contrary not
only to B's general preference for shorter forms, but to his specific aims in these
pages. In particular B's 'Snare.' is an unmistakable instance of a copy-derived
form. The earlier 'Hostesse.' may exhibit not only the additional influence of Q's
catchword on the previous page, but also of the full name in the immediately
preceding stage-direction, though (as already shown) stage-directions rarely ex-
erted so strong a force on B as to make him depart from his usual inclination for
shorter forms in the absence of some other inducement. Although predictably
not numerous, the full speech-prefixes in these compressed pages are valuable
evidence of the presence of Q's forms behind F.

Compared to this quire, the next one, where B was generally under pressure
to lengthen his type pages, should contain more long speech-prefixes, and they
should be more suspect as reliable evidence of the influence of his copy. Such
forms are indeed more frequent in quire χgg (see table 2). Five of these full forms
(TLN 1693L, 2799L, 2809L, 2832L, 3254L) may be put down to B's expansionist
policy in this quire, and four others are perhaps suspect on similar grounds (2712,

Table 2. B's full speech-prefixes (Folio quire χgg)

                                                                 

Quire	TLN	Folio	Quarto	Riverside Act.Sc.Line
χgg1	1673	Wart.	Wart	III.ii.138
χgg1	1675	Wart.	Wart	III.ii.140
χgg1	1685	Feeble.	Feeble	III.ii.148
χgg1	1687	Feeble.	Feeble	III.ii.150
χgg1	1693L	Feeble.	Feeble	III.ii.156
χgg1	1700	Feeble.	Feeble	III.ii.163
χgg1	1705	Feeble.	Feeble	III.ii.169
χgg5	2671	Prince.	Prince	IV.v.138
χgg5	2712	King.	King	IV.v.177
χgg5	2757	Prince.	Prince	IV.v.220
χgg5	2764	King.	King	IV.v.224
χgg5	2768	King.	King	IV.v.227
χgg5v	2775	King.	King	IV.v.232
χgg5v	2778	King.	King	IV.v.235
χgg5v	2795	Dauie.	Dauy	V.i.8
χgg5v	2799L	Dauy.	Dauy	V.i.13
χgg5v	2804	Dauy.	Dauy	V.i.18
χgg5v	2809L	Dauy.	Dauy	V.i.22
χgg5v	2817	Dauy.	Dauy	V.i.29
χgg5v	2822L	Dauy.	Dauy	V.i.34
χgg5v	2826	Dauy.	Dauy	V.i.38
χgg5v	2832L	Dauy.	Dauy	V.i.43
χgg6	2905L	Iohn.	Iohn	V.ii.19
χgg6	2907NL	Iohn.	Iohn	V.ii.22
χgg6	2915L	Iohn.	Iohn	V.ii.30
χgg6	2930NL	Prince.	Prince	V.ii.44
χgg6v	3070	Dauy.	Dauy	V.iii.41
χgg7	3254L	King.	King	V.v.44
χgg7	3259L	King.	King	V.v.47
χgg7v	3309	Iohn.	Iohn	V.v.97
χgg7v	3315	Iohn.	Iohn	V.v.103
χgg7v	3318	Iohn.	Iohn	V.v.105

2757, 2804, 2826).[44] But the other full forms in these pages would appear to be
reliable evidence of the influence of his copy, rather than of page justification,
on B's work.

The speech-prefixes 'Feeble.' and 'Prince.' (1685, 1687, 1700; 2671, 2930NL,
and perhaps 2757) are especially valuable, because they contain more than four
letters, and such forms are among the surest signs of his copy's influence on B.
The latter particularly recalls B's retention of this form from Q5 copy in 1H4,
although a four-letter name, 'Iohn.' also constitutes reliable evidence the six times
it occurs. As already shown, B rarely uses such forms unless copy has them. More

specifically, in V.iv of 1H4 B usually follows Q5's variable speech-prefixes for
Prince John, reproducing exactly its single 'P.Ioh.' on f5 and its two 'Iohn.' forms
on f5v, though characteristically shortening one 'Iohn.' to 'Ioh.', probably as a
result of the need to justify a line (2973L).

In repeatedly setting the full name of 'Dauy.' in V.i of 2H4, B uses the form
found in Q throughout this scene. These eight complete names in χgg5v cannot
be easily explained away as the result of either line or page justification.[45] B also
reproduces this form in V.iii, despite the fact that he there reverts to his usual
tendency to shorten speech-prefixes, setting both 'Da.' and 'Dau.'.[46] In view of
his performance in the plays set from known quartos, most of these must be at-
tributed to the influence of copy. The speech-prefixes for this character, as well
as those for Feeble and the princes, in quire χgg look very much like copy-derived
forms.

Of all the speech-prefixes in Folio 2H4 perhaps the least reliable as evidence
for quarto copy are the seven for the King, which invariably occur in B's pages
in the form 'King.', as they do throughout Q. It is possible, as Howard-Hill at
one point suggests, that B had 'settled' on the full form of this title after quire b
of Richard II. But it is equally possible, as Howard-Hill also seems to recognize,
that B's inclination was to the 'Kin.' form and that the complete form exhibits
the influence of his copy's persistent 'King.' on his work.[47] Clearly this is the case
in LLL, where he followed Q's uniform 'King.' on his first page, began to shorten
to 'Kin.' on his next, and thereafter alternated between this preference and Q's
form. It may be inferred that his adoption of the full form midway through R2
is similarly traceable to his copy: both Q3 (printed by Simmes and probably set
by his compositor A) and Q5 (the other print believed by some to have provided
Folio copy) have 'King.' throughout. In 1H4 Q5's virtually uniform 'King.' would
have reinforced such influence, but the presence of one 'Kin.' (2709) more than
half way through B's work on this play suggests that he retained his preference
(however weakened by repeated exposure to 'King.') for the shorter form. There
is no clear evidence in the Histories of an actual preference for the full form, but
only of its domination of his pages, which presumably reflects its persistence in
his copy. Acquiescence (in this case) is not the same thing as preference.

Table 3. B's long speech-prefixes (Folio quires g and χgg)

                           

Quire	TLN	Folio	Quarto	Riverside Act.Sc.Line
g1v	332	Ch.Iust.	Iustice	I.ii.58
g3	654L	Falst.	Falst.	II.i.46
g3	661NL	Falst.	Falst.	II.i.54
g3	681L	Falst.	Falst.	II.i.78
g3	687	Falst.	Falst.	II.i.84
g3	729	Falst.	Falst.	II.i.132
g3	733	Falst.	Falst.	II.i.138
g3	739CW	Falst. (Fal.)	Falst.	II.i.143
χgg7	3217L	Falst.	Falst.	V.v.10
χgg7	3223	Falst.	Falst.	V.v.16
χgg7	3250NL	Falst.	Falst.	V.v.41
χgg7	3258L	Falst.	Falst.	V.v.46
χgg7v	3298NL	Shall.	Shall.	V.v.187

The 'King.' speech-prefixes in his pages of 2H4 may, then, be taken as some
evidence, though perhaps the least unequivocal, of the influence of the full forms
of Simmes' quarto on B. One could argue, of course, that such forms could have
been present in the left margin of a manuscript serving as Folio copy (or provid-
ing the basis for such copy), where its scribe, unlike a Folio compositor, would
have been under no pressure to fit them within a narrow column along with the
opening line of each speech. But a possibility, or at best a probability of some
indeterminate degree, is not a certainty. Q's invariable full form is an observable
fact and offers a credible explanation of the forms in B's pages that is consistent
with other evidence.

Other traces of the forms of Q's speech-prefixes may be found in Folio 2
Henry IV. Although longer forms occur throughout, the significant ones would be
in the pages where B was not pressed to expand the matter (sigs. g3v, g4, χgg7v)
and especially in g1–3, where he was actually trying to compress (see table 3).
Perhaps the single most interesting speech-prefix in the whole play is the 'Shall.'
on B's last page (sig. χgg7v) in a one-line speech that nearly fills the measure
(3298NL). This is the only such form in B's work. Otherwise, he set 'Shal.' (the
norm) or 'Shallow.' (under special circumstances).[48] The length of this crowded
line would have encouraged B to use his customary shorter form. His unique
'Shall.' duplicates that in Q, which is, moreover, the only instance of this form
that B would have found there.[49] The coincidence is too much to sequester be-
hind the skirts of Fortune: F's unique 'Shall.' reproduces Q's equally odd usage.

Some other long, but not full, speech-prefixes in these pages may also indi-
cate F's dependence on Q. The frequency of 'Falst.' in B's g3 would suggest that
J/A—who on the evidence of 1H4 favored this and other longer forms—rather

than B, typeset the page. But this suggestion carries no weight once it is seen
that Q also has 'Falst.' here and could well have influenced B to vary from his
pronounced preference for 'Fal.', as exhibited in 1H4 and in B's first page of 2H4
(g3v).[50] Cumulatively, these 11 are the longer forms of greatest weight, but those
in χgg7 must largely be discounted because of the general expansion happening
there, even though they might not have occurred with such frequency without
inducement from copy.[51] On the other hand, the single most interesting 'Falst.'
is that in the catchword on g3, because earlier B had set the actual speech-prefix
'Fal.' at the top of g3v (739) in accord with his settled preference.[52]

Likewise, the many instances of 'Ch.Iust.' in these pages, which occur despite
B's inclination to set the shorter and simpler 'Iust.', are misleading at first blush
(and therefore not listed here). Some of them would have occurred in response
to annotations in Q. But the most significant longer form for this character—the
first 'Ch.Iust.' in the play (332), which is the last B set—was apparently induced
by Q's rare complete name ('Iustice') in combination with the preceding stage-
direction (see Appendix, Note C).[53]

Although more complex than others in 2H4, these longer speech-prefixes
reflect the state of the forms that B would have found in an annotated copy of
Q. These longer forms, therefore, provide some support for the view that Q was
B's basic copy for 2H4, but apart from 'Shall.', and perhaps his odd 'Ch.Iust.', the
best evidence remains the full forms in his pages.

This is so despite the presence of some anomalous ones, which prove to be
not altogether surprising given his practices in the Comedies and in 1H4. In 2H4
B set a total of 13 full-name speech-prefixes where Q has a shorter form, four
together on sig. g3, the remainder in quire χgg (see table 4). The four anomalous
speech-prefixes on g3, all 'Hostesse.' instead of 'Host.', occur in one sequence. On

Table 4. B's anomalous full speech-prefixes (Folio quires g and χgg)

                           

Quire	TLN	Folio	Quarto	Riverside Act.Sc.Line
g3	618L	Hostesse.	Host.	II.i.3
g3	621	Hostesse.	Host.	II.i.6
g3	626L	Hostesse.	Host.	II.i.13
g3	632	Hostesse.	Host.	II.i.20
χgg1	1654L	Shallow.	Shal.	III.ii.119
χgg1	1731L	Falstaffe.	Fal.	III.ii.196
χgg5v	2824L	Shallow.	Shal.	V.i.36
χgg5v	2850L	Falstaffe.	Falst.	V.i.60
χgg5v	2881L	Warwicke.	War.	V.ii.1
χgg7	3171L	Hostesse.	Host.	V.iv.1
χgg7	3212L	Falstaffe.	Falst.	V.v.5
χgg7	3216	Pistol.	Pist.	V.v.9
χgg7	3245	Pistol.	Pist.	V.v.39

χgg1, where B again stood in for J/A, 'Shallow.' and 'Falstaffe.' replace Q's 'Shal.'
and 'Fal.'. In the latter half of quire χgg, these three full forms reappear, along
with a 'Warwicke.' (Q: 'War.') and two instances of 'Pistol.' (Q: 'Pist.').

Of the nine in quire χgg, all but two seem to be traceable directly to the
policy of expansion which B was following in these pages and which Prosser has
explored at some length.[54] She notes particularly the 'Falstaffe.' forms as products
of this strategy. But on these same pages, each of the 'Shallow.' forms allowed B
to overrun the matter into an extra line, as did the second 'Pistol.' and the 'War-
wicke.' on χgg5v.[55] Hence, only the earlier 'Pistol.' (3216) and a 'Hostesse.' (3171L) in
this quire appear to be genuinely anomalous, and both may well reflect B's gen-
eral expansionist tendencies in this part of the play. The latter, however, is like
the earlier examples of this full form in that it might be related to the preceding
stage-direction. Indeed, this and the three other aberrant speech-prefixes on page
χgg7 are all immediately adjacent to changes found in the Folio's text. The first
'Pistol.' (3216) precedes the deletion of a single word presumably struck to censor
oaths, but beside the other three are stage-directions that had undergone major
alteration and would have been heavily marked up and potentially distracting
had B's copy been an annotated Q.[56]

In contrast, the four 'Hostesse.' speech-prefixes on g3, early in II.i, cannot be
attributed to such general pressures on B, since if anything he was compressing
in this page. Yet the brief sequence, which ends with her long speech beginning

at 635, is comparable to the cluster of 'Prince.' forms on sig. f2 of 1H4, where B
was called upon to stand in for J/A. On g3 B seems to have been temporarily
influenced, not by considerations of space, but by the three full forms that he
found in the opening stage-direction (615) at the bottom of Q's sig. C1, in the
catchword there, and in the first speech-prefix of the scene (616) at the top of its
C1v, which he reproduced.[57] In short, if we except the seven complete forms in
quire xgg that are almost certainly traceable directly to B's efforts to expand his
copy, the remaining aberrant instances parallel almost exactly those in B's pages
of 1H4, even if we ignore the markings that would have been present in an an-
notated copy of Q.

There can be little doubt that by and large the longer and full speech-prefixes
in B's pages of 2H4 exhibit the influence of Q's forms. Or, to put it differently, a
significant number of B's speech-prefixes in this Folio play are remnants of those
set in Valentine Simmes's shop, presumably by his compositor A, some twenty
years earlier. As such they may be regarded as the kind of direct bibliographical
links between F and Q that Williams sought, unless a scribal manuscript that
preserved Q's forms intervened between the two books.

The purpose of this theory of intervening manuscript, which posits a docu-
ment incorporating not only the major Folio 'additions' that could not be written
upon the leaves of a copy of Q, but also all the other variants drawn from the
prompt-book (or other theatrical manuscript) as well as the Quarto's words and
forms, would be to account for those readings unique to F that a critic or an edi-
tor wished to reject as not deriving directly from the theatrical manuscript (or
from the Folio compositors). To the extent that Q's speech-prefixes have been
transmitted to F in a manner similar to that observed in the seven control texts
typeset from identified quartos, this theory faces a serious impediment. For their
survival through an intervening manuscript would have to be credited to the
slavish accuracy of its scribe, who cannot then readily be blamed for changes in
the actual wording (as opposed to forms) that it was his business to reproduce,
and whose reputation as an 'interfering', 'cavalier', and 'overhauling' workman
would therefore require considerable rehabilitation.

As genetic evidence of the dependence of F upon Q, the speech-prefixes of
Compositor B in 2 Henry IV may, then, be taken to bear considerable weight
in attempts to sort out the question of the precise printer's copy for the Folio
typesetting, as well as the more important one of its text's derivation from that
of Q. One Shallow, however, does not make a summer, and Taylor's statistics
for round brackets, hyphens, and exclamations remain to be reckoned with, not
least because they lack contextual analysis. On the other hand, should evidence
similar to B's speech-prefixes (say, his spellings and typographical styling) con-
firm the implications of these 'appurtenances', the case for annotated Quarto as
Folio copy would rest on a less ambiguous and more substantial basis than can
be provided by the substantive readings and anomalous features so long cited
and debated.