I
During much of the nineteenth century "bibliography" was understood
to mean what we would now regard as "reference bibliography"[7] (or "enumerative
bibliography")—that
is, it was concerned with the intellectual content of books, with preparing
lists of books on particular subjects, with the classification of knowledge
and the arrangement of libraries.[8]
Thomas Hartwell Horne, in An Introduction to the
Study of Bibliography (1814), went somewhat further and
included discussion of the materials of books and the history of printing; but
when he referred, in his preface, to "the infant science of Bibliography,"
he obviously meant nothing more than "the classification of books as a field
of knowledge."
[9] The use of
"science" in the general sense of "systematic knowledge" recurs in most of
the nineteenth-century discussions. It is explicit, for example, in Reuben A.
Guild's
The Librarian's Manual (1858), which defined
"bibliography" as "the Science or Knowledge of Books" (p. 3). That he
equated this science primarily with checklists is evident when he went on
to say, "In Great Britain Bibliography as a Science has received less
Attention than upon the Continent, although valuable Works have been
produced by Horne and Lowndes, Dibdin and Watt" (p. 5). His view of
bibliography as a "
practical Science" (p. 5) was still
essentially
the
same two decades later when he wrote an article entitled "Bibliography as
a Science," in which "bibliography" really means "librarianship."
[10] Similarly, E. Fairfax Taylor, in
the ninth
edition of the
Encyclopaedia Britannica (1875), though he
included some account of printing history under "Bibliography," defined his
subject as "the science of books, having regard to their description and
proper classification"—using "description," in the sense standard at
the
time, to mean a recording of the basic facts considered necessary to identify
a book (those we would now think appropriate for a checklist).
[11]
It is generally recognized that a new meaning for "bibliography"
developed in the last third of the century from the work of William Blades,
Henry Bradshaw, and Robert Proctor on incunabula. Blades's The
Life and Typography of William Caxton, published in 1861-63,
attempted to classify and date Caxton's books on the basis of a close
examination of their typography; according to T. B. Reed, writing in 1891,
this book "marked a new epoch in bibliography, and disposed finally of the
lax methods of the old school."[12]
Blades himself had commented a few years earlier on the achievement of
Bradshaw, whose important classified indexes of incunabula began with his
work on the de Meyer collection in 1869:
From an early period he perceived that to understand and master the
internal evidences contained in every old book, the special peculiarities of
their workmanship must be studied and classified, much in the same way
as a botanist treats plants, or an entomologist insects. This he called "the
natural-history system." . . . To make his work more effectual and
scientific, he did that which many a bibliographer has to his great loss
omitted to do—he made acquaintance with the technicalities of
book-making.
[13]
In remarks of this kind, both "bibliography" and "science" are obviously
used in a different sense from the way in which Horne or Guild had
understood them. "Bibliography" here means what we would now call
"analytical bibliography," with the emphasis on physical evidence, and
"science" refers not to systematic knowledge in general but to the
examination of empirical data. The movement initiated by these men is
what lies behind Henry Stevens's statement in 1877 that bibliography "is
fast becoming an exact science, and not a whit too soon. It is high time to
separate it from mere catalogue-making"
[14] —a statement which was
echoed in
remarkably similar language by W. A. Copinger in his "Inaugural Address"
before the newly formed Bibliographical Society fifteen years later.
[15]
However inexact the term "exact science" might be, it served the
rhetorical purpose these writers had in mind: an effort to contrast the
methodical inspection of evidence found in books with the dilettante interest
in old books merely as antiquarian objects. It is important to note that the
putative inexactness which the new "exact science" would replace did not
lie in the pursuit of enumerative bibliography but rather in the casual
attitude of book collectors who—like the "new" bibliographers
themselves—regarded books as physical objects. Since Stevens and
Copinger both made a point of saying that bibliography—in their
sense—was distinct from "mere" cataloguing, they may have given
the
impression that cataloguing was inexact work and that analytical
bibliography was the exact work that had developed from it. But such an
interpretation is actually an illogical mixture of two concepts which lie
behind their statements: first, that "bibliography" in the sense of listing or
cataloguing is a separate activity from "bibliography" in the sense of
attention to books as physical objects; and, second, that the "exact" pursuit
of the second kind of bibliography (examination of physical evidence) is
replacing the "inexact" (vague dilettante interest). The first concept
concerns definition of the field of activity; the second concerns the degree
of seriousness with which the field is pursued. When the definition of an
activity shifts in the middle of a discussion of a particular attribute of that
activity, only confusion can result, and this kind of confusion could be
regarded as the motif running through the whole history of attempts to link
the words "science" and "bibliography." It is no wonder that Olphar Hamst,
as early as 1880, felt that the word "bibliography" had so many meanings
as to be useless for "any scientific purpose."[16] One further point may be noted
about
these early descriptions of analytical bibliography: the
use of the phrase "natural-history method," like that of "exact science," was
meant to be suggestive, not precise. Obviously Blades knew that books,
being man-made objects, could not be studied in exactly the same way as
plants or insects, but there was no reason for him to make that point, since
he was concerned only with a general analogy between two examples of the
use of empirical observation, in order to contrast that method with one
which did not involve systematic observation at all.
In the years which followed, the Bibliographical Society and several
younger bibliographical societies continued to advertise the "scientific"
nature of physical bibliography. The title of Falconer Madan's "On Method
in Bibliography," read before the Bibliographical Society in 1893, is
characteristic of the concern of these groups that their field should be
systematic.
[17] Speaking before the
Edinburgh Bibliographical Society in 1899, John Ferguson said that
bibliography "has nothing to do, in the first instance at least, with the
contents. They may be good, bad, or indifferent, but they do not concern
the bibliographer. If one may so say, he is not a book-ethicist, but a
book-ethnologist."
[18] Ferguson's
choice of ethnology as his scientific analogy skillfully suggests that
bibliography has both an objectivity of method and a concern with the
human; even more revealingly he called bibliography "the biography of
books" (p. 9). He did not develop the point, but his recognition that
physical bibliography is a form of history probably accounts for his
unwillingness to
label bibliography flatly as a science; it is, he said, "the science or the art,
or both, of book description" (p. 3). Although he went on to concern
himself principally with the enumeration of books, his few comments on
physical bibliography constitute an intelligent revision of the scientific
analogy. Another speaker (J. Christian Bay) before another bibliographical
society (the Bibliographical Society of America) observed in 1905,
"Bibliography, as taught and practiced in the circle to which I address
myself, ranks now equal to, if not among, the exact
sciences"—phraseology which makes plain the metaphorical nature
of the
statement.
[19] And Victor H. Paltsits
at about the same time compared bibliography to anatomy in its concern
with analyzing the "component parts" of a book.
[20] Both writers, however, went on
to
confuse the issue somewhat by making the inevitable contrast with "library
routine" and the compilation of
lists. The whole tendency of this period to glorify the "scientific" aspects
of bibliography, in contrast to what went before, is well summed up in
James Duff Brown's
Manual of Practical Bibliography
(1906):
If once it is recognized that bibliography is really the index and guide
to all past and existing knowledge, . . . then there will be some hope of the
science being set in its proper place as a key to the knowledge stored, and
too often hidden, in books. At present we cannot hope for this recognition.
It has become crystallized in the public mind—if it ever considers the
matter at all—as a dull, repulsive game for snuffy and cantankerous
old
men who spend most of their time buying books from ignorant booksellers
at one twentieth part of their market value, in order to stow them away on
musty bookshelves, there to accumulate a further value in the course of
time. Book-hunting, indeed, has almost become synonymous with
bibliography in the minds of a great many persons. But, luckily, a more
advanced, more reasonable, and more scientific spirit is awakening, and
many modern practical exponents of the new bibliography have completely
repudiated the traditional view of the limits of the science. (pp.
19-20)
Brown's chief interest is obviously in reference bibliography, but his
contrast of the scientific present with the dilettante past
[21] is characteristic of the viewpoint
lying
behind the insistence on science in analytical bibliography as well.
By 1912 the tradition of comparing bibliography with science was
well established, and W. W. Greg had given enough thought to the matter
that he was ready to make what would be the first of an important series of
statements on it. He recognized that a general analogy with science could
be drawn, for bibliographers "are gradually evolving a rigorous method for
the investigation and interpretation of fresh evidence."[22] But what distinguished his remarks
from
previous ones is that he turned the scientific analogy into a criticism, saying
that bibliography was not yet a "satisfactory science":
In a sense every science is descriptive. But in so far as a science is
merely descriptive it is sterile. You may dissect and you may describe, but
until your anatomy becomes comparative you will never arrive at the
principle of evolution. You may name and classify the colours of your
sweet peas and produce nothing but a florist's catalogue; it is only when
you begin grouping them according to their genetic origin that you will
arrive at Mendel's formula. (pp. 40-41)
Like the writers before him, Greg contrasted enumerative and analytical
approaches but, unlike them, did not feel that the analytical had developed
far enough to provide cause for celebration; the scientific analogy, if it was
useful at all, apparently could serve to stimulate bibliographers to greater
activity. Still, Greg used the word "science"
in his own definition, which at the same time gave currency to another
element that would complicate the issue. His chief interest in analytical
bibliography, in contrast to that of Blades or Bradshaw, was the effect
which its discoveries might have on the establishment of texts, and he
defined bibliography (he called it "critical bibliography") as "the science of
the material transmission of literary texts" (p. 48). In effect, his definition
tended to make analytical bibliography the servant of literary study; and,
while he did not say that the editorial process—choosing among
variant
readings and correcting mistakes—was a science, his statement did
use
the word "science" and did mention "literary texts." He had entered a
fertile ground for misunderstanding, and it is not surprising that debates
about the scientific nature of editing would occur, especially after others
began to pronounce similar definitions.
Of course, one of the principal accomplishments of the
Bibliographical Society in its early years—reflected in the emphasis
of R.
B. McKerrow's "Notes on Bibliographical Evidence for Literary Students
and Editors . . ."[23] —was to
demonstrate the bearing of analytical bibliography on literary matters; and
it is understandable that A. W. Pollard, as he surveyed in 1913 the
Society's first twenty-one years, should have defined bibliography as
dealing with "the material mediums . . . through which the thoughts of
authors reach those who will take the trouble to gain a knowledge of
them."[24] George Watson Cole
followed in 1916 with another statement stressing textual
transmission—the "perpetuation of thought . . . by means of the
printing-press"—as
the domain of bibliographical study.
[25]
And Falconer Madan, in his Presidential Address to the Bibliographical
Society in 1920, saw bibliography as "the groundwork to which every
literary researcher and writer will instinctively turn"; like Greg, but more
elaborately, he had recourse to "science" in expressing the connection
between bibliography and literary study: "It is not too much to say that our
work bears, or ought to bear, the same sort of relation to literary subjects
of research as mathematics bear to natural science."
[26] By a curious shift, the
natural-history
analogy was now more indirect; bibliography was not compared to science
directly but instead to mathematics, as a tool employed in the sciences, thus
making bibliography a tool of literary study—a rather narrow view
of
both mathematics and bibliography. Whether this statement was intended as
a summary of the current situation or as a recommendation for
the future is not clear, but in any case its hint of some sort of exactness in
bibliographical work is unusually vague. This tendency to increase the
distance between analytical bibliography and science was furthered in
Pollard's Presidential Address to the Edinburgh Bibliographical Society in
1923. Entitled "The Human Factor in Bibliography," it pointed out that,
unlike botany and geology, bibliography deals with human productions, and
any analogy between bibliography and science was therefore somewhat
limited.
[27] Pollard expressed a point
of view which has been heard often since then, but one cannot help feeling
that its target is a nonexistent argument, since surely no one who had called
bibliography scientific had believed that its materials of study were of
precisely the same order as those in the physical sciences. It was
bibliographers like Greg and Pollard, interested in the literary application
of analytical bibliography, who were finding
increasing reason to suggest qualifications of the scientific analogy; but
ironically their association of bibliography and literature helped give rise to
the misconception that bibliographers were attempting to put literary
criticism on a scientific footing.
[28]
During the 1930s two bibliographers in particular—Greg and
McKerrow—made comments about science which go to the heart of
the
matter. In each case they attempted to rectify certain fallacies which they
believed the comparison with science had led to, and one begins to feel that
they found the analogy more distracting than helpful. Greg, in his
Presidential Address of 1930, recognized that bibliography is essentially a
historical study; whether or not it is scientific thus turns into the question
of whether historiography is a science, and Greg answered in the
affirmative: "The knowledge of human events, and the methods by which
that knowledge is pursued, have just as good a claim to be called a science
as have any other body of facts and any other instruments of
research."[29] Whereas he had believed
earlier that bibliography was an immature science, he thought that it had
now moved to a new stage, and he pointed out the meaninglessness of the
often-used phrase "exact science":
Is not exactitude the aim of every science, which it approaches as it
gains in mastery over its material? . . . I think that the real distinction is
not between an exact science and any other, but between a mature science
and one that is still groping after its foundations, or else merely between
science and bunkum. (p. 256)
Although Greg was still calling bibliography a science, the implications
were now different, since it had been equated with history. And as history
it was an independent discipline, not "the slave of other
sciences" (p. 259).
[30] If from the
beginning analytical bibliography's subject had been described as (in Greg's
words) "human events"—as opposed to "natural
events"—there would
perhaps have been less misunderstanding about it. By the end of the decade
McKerrow seemed even more exasperated with the scientific analogy. In his
Prolegomena for the Oxford Shakespeare (1939) he observed
that the popular reputation of "science" caused people to wish "to bring
within its scope, at least in name, many subjects which cannot properly be
said to belong there" (p. vi). He admitted that science could be defined so
as to include bibliography: "Truth is truth and logic is logic . . . and in a
sense any honestly conducted enquiry may be termed scientific." But
science as "usually understood," he asserted, involves demonstration by
controlled experiment; in this sense the "textual critic"
[31] cannot be scientific:
. . . for scientific proof of his theories he must substitute arguments
based on what seems to him, from his "knowledge of human nature" and
from what he can learn of the procedure and habits of early copyists,
printers, and theatrical producers, most likely to have occurred, and which
can seldom or never be more than probably correct, even
though the probability may in some cases be of a high degree. (p.
vii)
Taken together, these statements of Greg and McKerrow cover the crucial
points: analytical bibliography is a form of historical investigation; its
conclusions are on a lower plane of probability than the inductive
generalizations of many sciences because of the impossibility in
bibliography of repeating past events as experiments; it can be thought of
as scientific only if "science" is taken in an extremely general sense. One
wonders what more needed to be said on the subject.
The scientific analogy, however, having become established,
continued to turn up. G. F. Barwick, sketching the history of the formation
of the main bibliographical societies, used "scientific bibliography" to mean
the examination of a book "as an entity" (as opposed to list-making) and
commented on various societies in terms of their attention to "scientific
bibliography."[32] Arundell Esdaile
considered bibliography to consist both of enumeration and analysis, the
first of these being an "art" and the second a "science."[33] And Stephen Gaselee, agreeing
that both
are legitimate aspects of bibliography, went farther than previous writers in
finding that both could be called scientific in the same sense—"both
are
a part of science, at any rate of that natural science to which bibliography
is ordinarily and reasonably compared."[34] In addition to comments of this
kind,
there was one event
in the 1930s which gave new force to the scientific analogy: the publication
of John Carter and Graham Pollard's An Enquiry into the Nature of
Certain Nineteenth Century Pamphlets (1934). A spectacular
instance
of answering a bibliographical question by recourse to the laboratory was
bound to become a classic illustration of "scientific bibliography." Yet no
one would be likely to argue that the laboratory analysis of paper is a
peculiarly bibliographical technique; it would be more accurate to say that
it is a technique from a field other than bibliography which proved to be
helpful in investigating a bibliographical problem. Microscopic analysis fits
the popular conception of "science," but bibliography does not achieve
scientific status merely through association with it. The way in which the
Enquiry could legitimately be said to represent a "scientific"
approach to bibliography
is in the frame of mind of its authors, whose objectivity in assessing
physical evidence led them to see the necessity for turning to another
discipline for assistance.
[35] A
carefully worded statement on the dust-jacket of their book clearly reflects
this distinction, saying that the book "introduces scientific methods which
have never before been applied to bibliographical problems of this period."
But probably most people, when they call Carter and Pollard's work
"scientific bibliography," are thinking of the microscope and do not reflect
on the fact that they are thereby attaching an additional meaning to an
already overburdened term.
Bibliography continued to be referred to as vaguely "scientific"
through the 1940s,[36] though two
important essays did appear—Madeleine Doran's "An Evaluation of
Evidence in Shakespearean Textual Criticism" and R. C. Bald's "Evidence
and Inference in Bibliography," both in the English Institute
Annual of 1941. These essays constitute the most serious and
extended treatment that had appeared of the implications of the scientific
analogy, following the lines of Greg's and McKerrow's comments; more
than that, they provided a direct examination of the nature of
bibliographical reasoning and demonstration. Bald, agreeing with Greg,
classified bibliography as history —or, more precisely, said that it
belongs among those "organized human activities . . . loosely known as
'history and the social sciences'" (p. 162). Just as history studies
"monuments" (material objects which survive) and "documents" (accounts
of events, liable to human error), so
bibliography, he reasoned, examines both books themselves, as physical
objects, and external evidence bearing on their production and distribution.
Because historical study involves human actions and because laboratory
experiments cannot recreate the past, the method of "proving" a case in
bibliography could be likened more appropriately to that followed in a court
of law than to that employed in a scientific investigation.
[37] "Bibliography," he summarized,
"cannot
claim for its conclusions the same universal validity as belongs to those of
the exact sciences" (p. 162). Miss Doran, in her essay, provided a concise
expression of this point of view:
It should be clear that we are in a realm where demonstration, in the
strict sense of the term, is impossible. For our method cannot be solely
deductive; nor do our problems admit of controlled laboratory experiment.
. . . The textual problem is always a historical one—an attempt at
recovery of what actually did happen; demonstration, therefore, is always
a matter of the establishment of probability. This is so great in some cases
as to amount almost to certainty; in others, so slight as to be
questionable.
[38] (pp. 98-99)
It would be hard to find a more compact and penetrating statement of the
case. Four years later the Bibliographical Society's commemorative volume,
Studies in Retrospect,
1892-1942 (1945),
naturally
gave some attention to the development of bibliography as a "scientific"
pursuit,
[39] but it included no comment
which brings together all the central issues as this one does.
From this point forward the most prolific commentator on
bibliographical theory has been Fredson Bowers, and his writings, as one
would expect, repeatedly touch on the "scientific" question.[40] However,
he uses the word "science" infrequently, and it is clear that he follows in
the line of those writers who find the scientific analogy somewhat facile.
His position, as set forth in "Some Relations of Bibliography to Editorial
Problems,"
[41] is that, although there
has often been an "inferential identification of bibliography with textual
criticism," the two are separate, and analytical bibliography can be pursued
independently of any possible application to textual matters. Since analytical
bibliography deals with physical evidence, it lends itself to logical,
systematic procedures; "strictly bibliographical evidence," Bowers says,
"crosses the line of probability into something close to the field which in
science would be regarded as controlled experiment capable of being
reproduced" (p. 58).
[42] Textual
criticism and editing,
on the other hand, require cirtical insight: "the great emendations have been
inspired art and not systematic science" (p. 45);
[43] evidence in this area "can seldom
if ever
afford more than a high degree of probability, and this is essentially
different from positive demonstration" (p. 57). Bowers makes clear, both
here and in succeeding essays, a point which some of the later writers on
"science" in bibliography do not seem to recognize—that analytical
bibliography, while it may at times invalidate a literary argument through
a factual demonstration, cannot (and does not claim to) eliminate the need
for judgment and critical acumen in editing. Indeed, Bowers repeatedly
defends the authority of informed critical insight, when coupled with an
understanding of the extent to which analytical bibliography can contribute
to the solution of a given problem:
The scientific method should have its valued place in humane studies,
but as a servant, not the master. The current exaltation of the scientist in
other fields should not lead to his domination of the humanities. Yet the
processes of logical and material demonstration which the more scientific
bibliographical methods bring to literary studies cannot be idly surveyed
from an ivory tower or they will eat away its foundations and topple
it.
[44]
In a concise statement of the point he says, "Bibliography endeavors to take
as much guesswork as possible out of textual criticism, and the literary
method endeavors to inform bibliography with value judgments as a check
on mechanical probability."
[45]
Although Bowers does occasionally apply the word "scientific" to analytical
bibliography,
[46] therefore, he is
careful not to use it to describe editing; and he has done more than any
other bibliographer to give substance to the word, by examining at
length—in his 1959 Lyell Lectures
[47] —the nature of
the evidence which analytical bibliography produces and the soundness of
the conclusions drawn from that evidence.
We shall have occasion shortly to return to those lectures. But, first,
it is worth noting that the use of the word "science" in connection with
analytical bibliography—as a brief historical sketch of this kind
reveals—has developed in two phases. First came the enthusiastic
phase,
in which bibliographers found science a useful analogy to help them
advertise the fact that their field was a serious and systematic study, not a
dilettante pursuit. Exaggeration was probably inevitable;[48] but however strongly they claimed
bibliography to be science, these bibliographers generally did not examine
in detail the implications of such a comparison but instead used it in a
vaguer way for its suggestive value.[49]
The second—or critical—phase began when bibliographers,
taking these
scientific claims more literally, recognized that a comparison of
bibliography with "science" (that is, in the usual sense of "physical
science") involved pointing out many differences, perhaps as many
differences as similarities. Leading bibliographers of the past fifty or sixty
years have taken this second position and have stated over and over various
distinctions between bibliography and "science." At the same time, through
both phases, the issue has been complicated by shifting terms, with one
person talking about a different kind of "bibliography" from another, or
using "science" in a different sense. One begins to wonder whether the
whole matter was not a red herring from the start. Presumably the point of
the analogy is to define bibliography,
and definition by analogy can sometimes be illuminating, even when the
supposed analogy serves as something to be reacted against. But when the
comparison involves a concept as complex as "science," it may do more to
confuse than to clarify. Whether bibliography can be defined as a "science"
or as something else is of less importance than understanding, in a direct
way, what in fact it does, what its methods of procedure are, what its
strengths and weaknesses may be. More direct discussions of such matters
might have promoted greater understanding than that which has resulted
from the perennial concern with the "scientific" quality of bibliography.
The course of these "scientific" comments over the years is not an inspiring
one and appears to be leading nowhere; the last word on the subject would
seem to have been said, and said repeatedly. But apparently Bradshaw's
concept of a "natural-history method"—and all that follows from
it—is
so intriguing to bibliographers that they
cannot let go of the analogy, for it remains a matter of discussion.