 | Studies in bibliography |  |
9. Numbers
"Follow the money" is an excellent adage for book historians. The production,
distribution, and consumption of texts are almost always co-determined
by commercial and/or economic factors. Whether we are calculating the area
of pastureland needed by a medieval monastery operating a scriptorium in
order to maintain a sufficient number of animals for parchment production
(surplus to requirements for food to feed hungry scribes), or are factoring in
the cost of severance packages when Bertlesmann has acquired a publisher of
scientific journals in North America, we cannot think very far about the
transmission of texts before we bump into matters of money because, invariably,
money matters. Books are business. It seems strange, then, that so few
book historians possess economic knowledge, while the rest of us know almost
nothing about prices, income, capital, markets and consumer behavior, financing,
taxation, money and banking, and other rudiments of elementary
macro- and microeconomics.[81]
And who among our number knows very much
respective periods? If commercial motivations dominate the world of print
production and dissemination, then how is it that we so often give them such
scant attention?[82] Perhaps for a start we should all resolve to read G. R.
Hawke's Economics for Historians (1980).[83]
In studying a business in which capital outlay was so substantial, credit
was routinely extended to customers for six months or more, return on investment
was slow to accrue, and bankruptcy was common, it is particularly
important to recognize how changes in financing and credit arrangements
fostered or hindered growth. This is especially true if we are to understand
how adjustments in capitalization drove the shift from bookselling to publishing
in the late-eighteenth and early-nineteenth centuries. In a similar
vein, we have yet to develop satisfactory accounts of the book trade's behavior
in—and varying responses to—the well-documented fluctuations of national
economics.[84]
Although there is a dearth of surviving business and banking
records for the book trade in the long eighteenth century, most of what we
do have has not yet been subject to a thoroughgoing financial analysis. Nor
has much comparative or synthetic work been done either to document the
finances of agents in the trade,[85]
or to develop more complete models of how
businesses actually worked.
Most scholars in the humanities, including many book historians, are
virtually innumerate, a liability that affects far more than our general inability
to deal adequately with economics and with business records. Much
book-historical work manifests a statistical innocence that impoverishes otherwise
valuable research. If we are to understand scholarly writings that presuppose
the reader knows the rudiments of statistics, then we need to progress beyond
the definitions of mode, median, and mean.[86]
Most of us are not at home
variables, or continuous independent variables, much less multiple regression
models. Who among us can interpret a chi-square test, something a first-year
undergraduate biologist or sociologist would be expected to do? Roderick
Floud's An Introduction to Quantitative Methods for Historians should be
required reading.[87] It would also be salutary for more book historians to be
familiar with SPSS, the Statistical Package for the Social Sciences, the most
commonly utilized computer program for basic statistical analysis.[88] More
advanced scholars examining large bodies of data, such as the ESTC, would
probably benefit from developments in sampling theory, though in such
cases close collaboration with a professional statistician is probably required.[89]
One of the chief methodological lacunae in Anglo-American bibliography
and book history also seems to stem, in part, from our discomfort with
numbers. This oversight in need of redress is our failure to interrogate the
bibliographical tools at our disposal—most especially the ESTC—so that we
know their strengths and limitations. Virtually all book historians routinely
employ enumerative bibliographies for their research, but many apparently
give little thought to the details of inclusion criteria, to problems raised by
varying survival rates, or to economic constraints and other important factors
among the circumstances that produced such research resources.[90]
A
common scientific adage is: "All measurements are wrong, but some are more
wrong than others. The key is to know how wrong your measurement is."
W. W. Greg has calculated that, of the 11,000 works entered in the Stationers'
Register for 1557-71 and 1576-1640, "6,100 are identified as extant, a percentage
of almost 55.5." Reasoning that "there does not appear any very
strong ground for supposing that entrance or non-entrance affected survival,"
he nevertheless concedes: "it is possible . . . that in the field of ephemeral
publications, where survival is least likely, the proportion of copies entered
may have been somewhat lower."[92] D. F. McKenzie's work indicates that what
the Wing Catalogue (1641-1700) includes "may not be as high as even 60 to
70 percent of the titles and editions actually published."[93] In his examination
of survival rates for ABCs, psalters, and primers from the Stationers' stock in
sample years from 1660-1700, John Barnard has documented enormous
losses: although some 14,000 psalters were printed annually, only four copies
are found in Wing. The loss rates for primers is also astonishingly high: for
the 1676-77 fiscal year, "84,000 passed through the Treasurer's hands," yet
they "are represented in Wing by a single 16mo black-letter copy in the
British Library, dated c. 1670."[94]
My own calculations for British books in the eighteenth century suggest
that, for approximately ten percent of the titles and editions published, not
a single copy exists.[95]
Of course, by definition, none of these lost works or
editions is represented in the ESTC, a fact that surely should have a significant
effect on many of our searches.[96]
Because loss rates vary widely according
is far off the mark. A great deal more work will need to be done before we
can calibrate the tools we use and, hence, employ them with greater fidelity
to the historical phenomena under investigation.
One way to begin developing a sufficient understanding of book-trade economics is
to gain a deeper sense of the value of money and, hence, of the prices being charged—see
John J. McCusker, How Much Is That in Real Money? A Historical Commodity Price
Index for Use as a Deflator of Money Values in the Economy of the United States, 2nd ed.
(Worcester, MA: American Antiquarian Society, 2001). Historians of the book in Britain
will find McCusker's on-line calculator useful for converting an historical sum in pounds, shillings,
and pence into today's money: <http://www.eh.net/ehresources/howmuch/poundq.
php>. There is also a calculator for converting historical U.S. dollars into current money.
Of course, there are notable exceptions; see, for example, James Raven, "The Book
as a Commodity," in The Cambridge History of the Book in Britain, Volume 5, 1695-1830,
ed. Michael F. Suarez, S.J., and Michael Turner (Cambridge: Cambridge Univ. Press, 2007),
chap. 3; Raven's forthcoming history of the English book trade, 1450-1900 (2007); and
William St Clair, The Reading Nation in the Romantic Period (Cambridge: Cambridge
Univ. Press, 2004). Robert Darnton's The Business of Enlightenment: A Publishing History
of the Encyclopédie (Cambridge, MA: Harvard Univ. Press, 1979) is exemplary in this
regard.
For a noteworthy exception, see John Bidwell, "American Papermakers and the
Panic of 1819," in Barker, ed., A Potencie of Life, pp. 89-112; see also J. J. Barnes, "Depression
and Innovation in the British and American Book Trade, 1819-1939," in Carpenter,
ed., Books and Society in History, pp. 231-248. A useful guide for an especially significant
period is R. C. O. Matthews, "The Trade Cycle in Britain, 1790-1850," in Derek H. Aldcroft
and Peter Fearon, eds., British Economic Fluctuations 1790-1939 (London: Macmillan,
1972), pp. 97-139.
A noteworthy exception is Judith Milhous and Robert D. Hume, "Playwrights' Remuneration
in Eighteenth-Century London," Harvard Library Bulletin n.s. 10 (1999 [Aug.
2001]), 3-90.
Hugh Amory, "A Note on Statistics," in The Colonial Book in the Atlantic World,
504-518, is right to caution, "precision should not be mistaken for accuracy" (p. 515), but it
is nevertheless disconcerting that the only treatments of the data presented are simple percentages
and moving averages. More dismaying still is the fact that none of the evidence he
offers has been statistically tested to gauge its reliability.
Roderick Floud, An Introduction to Quantitative Methods for Historians (Princeton:
Princeton Univ. Press, 1973). Other helpful guides include: R. Darcy and Richard C. Rohrs,
A Guide to Quantitative History (Westport, CT, and London: Praeger, 1995); Lauren Haskins
and Kirk Jeffrey, Understanding Quantitative History (Cambridge, MA: MIT Press,
1991); Konrad H. Jarausch and Kenneth A. Hardy, Quantitative Methods for Historians:
A Guide to Research, Data, and Statistics (Chapel Hill: Univ. of North Carolina Press, 1991);
and Michael J. Moroney, Facts from Figures, 3rd rev. ed. (1969; rpt. London: Penguin Books,
1990). The more discursively inclined may wish to begin with Pat Hudson, History by
Numbers: An Introduction to Quantitative Approaches (London: Edward Arnold, 2000
[New York: co-published in the U.S.A. by Oxford Univ. Press, 2000]).
Among the many guides to SPSS are: David M. Shannon and Mark A. Davenport,
Using SPSS to Solve Statistical Problems: A Self-Instruction Guide (Upper Saddle River,
NJ: Merrill, 2001); Andy P. Field, Discovering Statistics Using SPSS for Windows (Thousand
Oaks, CA, and London: SAGE Publications, 2000); and Brian S. Everitt and Sabine Landau,
A Handbook of Statistical Analyses Using SPSS (Boca Raton, FL: Chapman & Hall/CRC,
2003).
For a straightforward and genuinely useful application of statistics to analytical
bibliography, see David Shaw, "A Sampling Theory for Bibliographical Research," Library
5th ser. 27 (1972), 310-319.
We may recognize a closely related case in the ways that bibliographers and book
historians have routinely used book auction catalogues for provenance information, even
though the commercial circumstances of their production most often makes them unreliable
sources; see Michael F. Suarez, S.J., "English Book Sale Catalogues as Bibliographical
Evidence: Methodological Considerations Illustrated by a Case Study in the Provenance and
Distribution of Dodsley's Collection of Poems, 1750-1795," Library 6th ser. 21 (1999), 321360.
Cf. G. Thomas Tanselle, "Tolerances in Bibliographical Description," Library 5th
ser. 23 (1968), 1-12.
Greg, "Entrance in the Stationers' Register: Some Statistics," in W. W. Greg,
Collected
Papers, ed. J. C. Maxwell (Oxford: Clarendon Press, 1966), pp. 341-348 [pp. 344, 348].
On factors affecting the survival rates of early printed books, see Joseph A. Dane and Rosemary
A. Roberts, "The Calculus of Calculus: W. W. Greg and the Mathematics of Everyman
Editions," Studies in Bibliography 53 (2000), 117-128 [pp. 124-127].
D. F. McKenzie, "Bibliography and History: Seventeenth-Century England." The
Lyell Lectures, delivered at the University of Oxford May-June 1988. Typescripts circulated
privately. Lecture 4, "[Matthew, Mary, and Samuel] Simmons," p. 4. (Although not published
together, two individual lectures have appeared separately in print; "Speech-Manuscript-Print"
and "Trading Places.")
Barnard, "The Survival and Loss Rates of Psalms, ABCs, Psalters, and Primers
from the Stationers' Stock, 1660-1700," Library 6th ser. 21 (1999), 148-150 [p. 149]. See also
John Barnard, "The Stationers' Stock, 1663/4 to 1705/6: Psalms, Psalters, Primers and
ABCs," Library 6th ser. 21 (1999), 369-375.
See Suarez, "The Quantity and Nature of Printed Matter: A Bibliometric Analysis
in The Cambridge History of the Book in Britain, Volume 5, 1695-1830, ed. Michael F.
Suarez, S.J., and Michael Turner (Cambridge: Cambridge Univ. Press, 2007), chap. 1. Survival
rates are highly variable, depending on format and genre, among other factors. An
overall figure of c. 90% of editions having at least one extant copy is an estimate based on
comparing titles and editions from eighteenth-century printers' ledgers and book reviews
with entries in the ESTC. Less than one percent of eighteenth-century jobbing printing
survives, though such ephemeral publications were the overwhelming majority of presswork—both
in terms of sheets printed and of items produced—and the financial lifeblood
of most firms, especially in the provinces; see Keith Maslen, "Jobbing Printing and the
Bibliographer," cited in n. 40 above.
One problem of information retrieval is that the user may easily detect the irrelevance
of some of the "hits" that result from a search, but most often has no way of knowing
what potentially relevant data is missing. In information science this is "the recall problem,"
where recall is "the proportion of relevant information that was retrieved." C. N. Ball,
"Automated Text Analysis: Cautionary Tales," Literary and Linguistic Computing 9 (1994),
295-302 [p. 295]. See also D. Blair and E. Maron, "An Evaluation of Retrieval Effectiveness
for a Full-Text Document Retrieval System," Communications of the ACM 28 (1985), 298299,
who note that users were confident in the system even when less than 50% of relevant
information was retrieved (cited in Ball, 295-296).
| Studies in bibliography | |