II. Evidence
All evidence was obtained from examination of two copies of the
Folio, the one belonging to the library of the University of Wisconsin
—
Milwaukee (of unknown provenience, but purchased from Heffer,
Cambridge, in 1952) and the one belonging to Dr. Cyrus Hoy (formerly the
Fairfax of Cameron copy).[3] The
kind of evidence used, the terminology adopted for it, the means contrived
to display it, and the interpretations put on it have been influenced, as the
reader will recognize, by Dr. Charlton Hinman's Printing and
Proof-Reading of the First Folio of Shakespeare (1963), although
there are some divergences from Hinman's methods chiefly because of
technical differences between the Shakespeare and the Beaumont and
Fletcher Folios. Yet the general similarities between these books and the
kinds of evidence available from both are so strong that it has been possible
to refer the reader often to Hinman's study, where he will find many
matters
relevant to this investigation discussed admirably and at length.
It seemed convenient to work through the section quire-by-quire and
page-by-page, considering in turn evidence relating to presswork, to the
identity of compositors and the scope of their work, and to the method of
composition they employed. The evidence used and the nature of the
inference from it are governed by the relationship between the three distinct
stages in the cyclical journey of pieces of type from the type cases into the
formes (composition), through the press (presswork), and from the
wrought-off formes into the cases again (distribution).
Presswork
Because information about presswork, specifically the order of the
formes through the press, is relatively easy to obtain and is based on
evidence that is the least controvertible, the analysis of each quire
began with proof of the order of the formes, which was made on these
grounds:
- 1. Evidence from the components of the skeleton-forme,
such
as running-titles and rules, and from center rules, taking into account the
testimony of priority given by disfigurations, dislocations, and other
peculiarities.[4] It was
assumed,
in the absence of indications to the contrary, that two skeleton-formes
would have been used alternately.
- 2. Evidence from type matter reappearing within the
quire. When a piece of type, a scene head, an ornament, or a rule
not
a part of the skeleton-forme appears twice within the same quire, it is
apparent that one of the formes of the quire was printed and distributed (at
least in part) before composition of the entire quire was completed. It was
assumed, lacking contrary evidence, that two formes containing the same
piece of type matter would not have been machined concurrently, for had
they been, a delay in presswork would have resulted.
- 3. Evidence from embossing. The priority of the
formes printing and perfecting a single sheet can be determined from the
embossing of the inked paper by the type of the perfecting forme,[4a] but because embossing resulted
largely
from the pressure exerted by the pressman's pull, there are nearly always
differences in the degree of embossing in different examples of the same
sheet. Because one of the copies of the Folio used for this study is not
generally accessible, the evidence from embossing is not reported. It may
be understood, however, that the implications of this kind of evidence as it
appeared in the copies examined never contradicted the conclusions drawn
from other evidence.
The Order of Composition and
Distribution
It is reasonable to assume that a temporal relationship held between
composition, presswork, and distribution — that, in general, the first
forme set was the first forme machined and the first distributed. As soon
as this is said, however, qualification must be made. If the press
was temporarily out of commission, printing other material, or otherwise
unavailable for work on the Folio, two or even more formes might be
machined later in any convenient order. Yet when operation proceeded
smoothly, "the order in which the various formes were printed was
ordinarily the the same as that in which they were set."
[5] Thus, to prove the order of
printing is
usually to prove the order of the composition of the formes. Such a firm
relationship did not, however, exist between presswork and distribution, for
if two wrought-off formes were available for distribution nothing seems to
have compelled the compositor to distribute first the one that was machined
first.
[6] Once more, though, there was
a tendency to distribute material in the order of its machining; and,
regardless of its implications for priority, evidence of distribution indicates
at least that composition and machining had taken place. Such
evidence is, therefore, not irrelevant to problems of composition, although
it has to be used circumspectly.
Composition (in which term I include all the compositor's duties)
seems, in fact, to have been governed by presswork. When the press stood
idle, the printer could not profit; hence, when local conditions permitted it,
an effort evidently was made so to regulate the speed of composition that
the press would be supplied with a steady flow of material to be printed off.
The rate of printing was chiefly a function of the size of the edition, and the
rate of composition was chiefly a function of the nature and amount of the
text to be set up for each forme. The speed of printing was a constant, but
the speed of composition could be varied by several means, a favorite
among them being the employment of more than one compositor to set type
for a single forme or some other kind of collaboration to accomplish the
same purpose, such as the composition by one workman of a complete
forme while his companion set another forme or the employment of one
workman in distributing and performing
other chores while his companion set type. Flexibility was desirable, of
course, and these basic techniques may have been modified in many
different ways in order to adjust for the effect that particular circumstances
had upon the progress of a specific printing job. However, because a folio
forme contained a considerable amount of material and because a folio
forme could be machined in about the same time as one containing much
less (say a quarto forme) if the same number of copies of both were to be
made, it seems generally true that two compositors were required to work
simultaneously in folio printing.[7]
When two compositors collaborated
in setting a forme or some other subdivision of a quire, they would divide
the work to be done according to a more-or-less rational scheme consistent,
one supposes, with either the bibliographical structure of the book or the
physical characteristics of the copy. The actual division could, of course,
be affected by many things — the availability of workmen, their
speed
in composing relative to each other and relative to the speed of the press,
their competence, and so on — and the rationality of the scheme
adopted
may have been more apparent at the time than it seems now. Nevertheless,
the manner of dividing the copy is a factor that must be taken into account
when there are indications that more than one compositor was at
work.
Proof of the order in which the pages of a particular quire were set
can usually be made on the following evidence:
1. Reappearing types. As Hinman has shown in detail,
it
is possible to identify individual pieces of type which are distinctively
broken, bent, or otherwise marked and to trace their reappearances in quire
after quire.[8] There are, to be sure,
differences in the reliability of the identifications. Some types are battered
in so readily discernible a way that they can be recognized with no
difficulty. Others, however, are so deformed that the characteristic which
makes them distinctive can be obscured by variations in inking, in the
surface of the paper, or in the force of the impression. Still others are
virtually worthless as evidence because either through accident or through
some weakness inherent in the design of the letter two or more types will
have been damaged in such a way as to make them practically
indistinguishable; the ascenders of d's and the descenders of p's, for
instance, were very often bent or broken in almost exactly the
same way. Sometimes a letter which can be recognized in one copy of the
Folio cannot be recognized in another, and sometimes a pattern of
recurrences will indicate that a type must have been used on a certain page
on which it cannot be found.[9]
Furthermore, a compositor was rarely obliged to use
again immediately any type which was returned to his case; he may not
have needed it, or it may have become temporarily unavailable by being put
accidentally in an out-of-the-way place or simply by sifting down into the
box. Thus some types, as it were, go underground for several, sometimes
many, quires. On the whole, however, types were reused regularly; and,
when recognized, they provide very solid bibliographical evidence, although
the reliability of this evidence decreases, as will be seen, the smaller the
number of types giving testimony.
To find in Quire Y a type which had earlier appeared in Quire X is
generally to find that the part of Quire X containing the type had been
distributed in the ordinary way. But this conclusion is not always correct,
for some types moved into new positions not as a result of distribution but
as a consequence of an accident or irregularity, having been pulled during
inking, separated from other types when the skeleton was stripped from the
wrought-off forme or the center rule removed, or dropped during
distribution.[10] If one finds ten types
from X3a in Y4b, it is clear that X3a was distributed before the Y-column
was set and that Y4b was composed at the case[11] into which X3a type was
distributed, if
the column rather than the part-column was the unit of distribution. If,
however, one finds one type from X3a in Y4b, these matters are by no
means so clear; the one X3a type may be aberrant, having made its way
into the case
from which Y4b was set through accident rather than regular distribution,
or it may be quite genuine, being merely the only recognizable type from
X3a that happens to appear in Y4b. Thus when recognizable types are few,
it is necessary to use the evidence they provide with caution, and, if
possible, to confirm it by evaluating its consistency with other evidence.
Nevertheless, in spite of minor ambiguities, distribution is usually not hard
to prove, and it then follows that "whenever a number of types are
distributed into a certain case these types will necessarily next be used in
material set from that case."[12]
Evidence of case in conjunction with spelling evidence, as Hinman has
shown, is of major importance in establishing the limits of a compositor's
work in a particular forme,
the identity of the workman, and the relationship borne by composition to
presswork at any particular time.
[13]
Whether type drawn from one case was distributed into the same case
by the man who set it is a matter of some consequence. What little
information there is about this aspect of Elizabethan printing practice
indicates that sometimes, perhaps generally, compositors may be said to
have "owned" their types, so that if Compositor A set X3a from a certain
case, the reappearance of X3a type on Y4b will indicate that Compositor
A set that column as well from the same case.[14] In William Wilson's section of the
Folio,
however, this practice was not followed with any consistency; hence, the
concept of case as a category of bibliographical evidence is severely
restricted. It is not, however, invalidated. If one observes that X3a type
reappears in Y4b, Y4va, and Y4vb, it is
evident that the three
Y-columns were set from the same case regardless of which compositor set
or distributed X3a. But it does become necessary when there is a likelihood
of one
workman distributing another's type to distinguish between the testimony
of types which reappear immediately and those which do not —
which
are, so to speak, latent — the former being much more trustworthy
as
evidence than the latter. Suppose that a type found in Y4b was last seen in
W2a, a column known to have been distributed during the composition of
Quire X into Compositor B's case. The implication would be that the Y4b
type in question was not used in Quire X and that its appearance on Y4b is
evidence for the setting of that column by Compositor B. It is always
possible, however, that the type actually was used in Quire X but was
unobserved there, and if it happened to be used in an X-column that was
distributed into Compositor A's case, its appearance on Y4b means that that
column was set by A, not B. The special treatment accorded latent types
(discussed more fully below) is thus a protective measure adopted to guard
against one's inability always to see or to
recognize the types being used as evidence. The chief value of latent types
lies in the confirmation they can lend to the implications of other
evidence.
2. Reappearing rules, display types, heads, and other
typographical matter not a part of the skeleton forme. In the
Shakespeare Folio it was
found that center rules were associated "with the type-page proper and with
the types used to set it. In fact they were treated very much as if they were
themselves only
types belonging to a special 'sort' . . . ."
[15] In Wilson's section of the
Beaumont and
Fletcher Folio, it seems that center rules were, in general, handled not as
types but as components of the skeleton-forme, which, of course, they
actually are not. This difference perhaps arises from the fact that in
Jaggard's shop the page was usually the unit of distribution and in Wilson's
the column, but the point is that in Wilson's work the center rules cannot
be used to prove the identity of the compositor, nor can other miscellaneous
typographical matter not a part of the skeleton forme. Initials, act and scene
heads kept standing, short rules inserted horizontally to set off heads,
ornaments, and display types seem to have been returned after use to a
central location from which any
workman could draw them subsequently. Thus the reappearance in Y4b of,
for instance, a short rule previously observed on X3a, Compositor A's
work, does not mean that Compositor A set Y4b; it means only that X3a
was off the press at the time Y4b was set.
3. Type shortages. Because double-column folio formes
contained a relatively large amount of type and because the repetition of
proper names or abbreviations of them demanded the same letter again and
again, type shortages sometimes occurred, particularly in the italics used for
speech-prefixes and stage-directions. These shortages caused the
compositors to substitute letter of a different sort, usually roman for the
customary italic, so that a speech-prefix for the character Zenocia in
The Custome of the Countrey, for example, sometimes
appears
as Zen. rather than as Zen. Deliberate
substitutions
of this kind can be distinguished from errors arising from such causes as
foul case by the frequency and consistency of the occurrence of the
wrong-font letter. It seems reasonable to suppose that, in general, the
compositor would use up his supply of regular type before he would begin
to substitute and would continue to substitute until his supply of regular
type was renewed by the distribution of a wrought-off forme containing
type of the required sort. On this supposition inferences can be made about
the priority of composition of certain columns or pages within a
compositor's stint. But workmen apparently did not always wait until their
regular supply was completely exhausted before they began to substitute,
and perhaps they occasionally raided other cases or broke new supplies out
of storage rather than obtaining types from distribution.
Moreover, it appears sometimes that during distribution the boxes
containing a depleted sort were deliberately fouled, a crude but efficacious
answer to the shortage and one which creates the impression of deliberate
substitution which did not take place. As a rule, the testimony of type
shortages is reasonably reliable when the shortage first comes into being,
but it tends to degenerate in worth beyond a quire or two.
The Identity of Compositors
The identity of compositors can be proved on the basis of their
habitual and occasional preferences or lack of preferences in spelling,
typography, punctuation, abbreviation, and so on, although the detection
of these preferences can be complicated by several factors, among them
the alteration of spelling to "justify" a full line of type, the
adoption of non-preferential spellings for visual rhyme, and, most
enigmatic of all, the influence of copy spellings on the compositor's
normal habits. A high degree of consistency in spelling evidence seems,
unfortunately, to be rare, and it is conceivable that two compositors in
the same shop could have preferences so nearly alike that their work is
for practical purposes indistinguishable. Yet very considerable aid and
comfort in the solution of problems of identity is available from
evidence of case and order, applied on the reasonable assumption that
two compositors could not set simultanously from the same case. If one
finds, for example, that Quire X was set from one case, it follows that
it was set by one compositor, unless it can be shown by variation in
spelling that he was relieved at the same case by another workman during
the course of composition. If one finds that Quire X was machined in the
order 2v:3-2:3v-1v:4-1:4v
and that X1-2v were set from one case and X3-4v
from another, it usually follows that two compositors were setting
simultaneously, barring once again the chance that either man was
relieved at his case. Even with the help thus provided, conflicts in
spelling evidence occasionally create uncertainties; but by and large
one finds that evidence of case and order accords with spelling evidence
in such
a way that the identity of the compositor is reasonably clear.[16]