Bibliography and Statistics
- From: "Jerry" <geralddowns@xxxxxxxxx>
- Date: 22 Mar 2007 21:56:53 -0700
Two or three years ago I took an interest in a bibliographical topic
involving both statistics and probability when I decided that an old,
unquestioned assumption was false. But I have a shortage of tools
for working out the problems. Perhaps someone on this group can
provide an insight or two.
A Shakespeare Editor broaches the subject of the survival of books:
The survival of thirteen copies [of the 1609 1st edition Shakespeare
Sonnet Quarto], most in good condition, suggests that the volume
did not undergo the enthusiastic thumbing that destroyed hundreds
of early copies of Shakespeare's earliest published poem, "Venus
and Adonis. . . . We might also compare the fate of another 1609
publication, _Troilus and Cressida_ . . . of which only four copies
survive. Assuming that the size of the print-runs was the same,
the evidence of survival-rates suggests that _Troilus and Cressida_
was three times as popular among readers of _Shakespeare's
Sonnets_ (8).
1) Edition size varied but was limited then to 1,200 copies by rule,
so taking 1,000 copies per edition is a convenient working number.
2) the first edition of Troilus survives in more than four copies, but
my
concern is with the math. 3) I believe it should be obvious (but is
not
to everyone) that survivals of 13/1,000; 4/1,000; and 1/1,000 do
not easily lend themselves to statistical analysis. 4) "Thumbing"
(repeated reading), has never been a leading destroyer of books;
fire, water, fire-water, children, servants, religion, heirs, bugs,
and
general loss of interest are more important. 5) For that reason,
a 13/4 ratio cannot be made meaningful only for "thumbing" --
or for anything else -- especially when the loss ratio is 987/996.
I've traced the 'thumbing' rule of thumb to an offhand 19th century
remark, and find no truly systematic study of the survival of books;
however, the idea that fewer extant copies means a book was more
popular is taken seriously. Before my questions, a few more notes.
Larger books tend to survive in greater numbers. Smaller & unbound
books disappear more completely. For example, roughly 240 copies
exist of the large, 1623 folio collected works of Shakespeare, while 2
copies of the smaller, 1603 1st quarto Hamlet survive. Schoolbooks
and other ephemera survive least. Exceptions to these rules occur.
In respect of the smaller books of literature, I note several trends
that I'm not sure how to treat: 1) Most survive to the present day
(if at all) in very small numbers; as a general rule, the small books
printed around 1600 rapidly declined. 2) Even though these numbers
approach zero, most books do survive (in the sense that at least one
copy still exists from an edition). 3) The more editions of a book,
the fewer copies survive of each edition. The result of these real
tendencies is low numbers of copies from particular editions that
can be compared with each other in seemingly meaningful ways.
For example, an authority wrote (a hundred years ago):
[Shakespeare's plays of Richard 2, Richard 3 and 1 Henry 4],
represented respectively by three, four and three copies of the
first edition, went through five or six editions apiece before
1623, whereas [2 Henry 4], of which nineteen copies survive, was
never reprinted in quarto. It is disconcerting to find that, on the
theory this suggests, so fine a play as [Merchant of Venice]. . . of
which 17 copies survive, comes next to 2H4 at the wrong end of
the list [By 'wrong end' the writer means that 17 surviving copies
of
1,000 printed is somehow a large number that indicates the play
was not popular]. We are forbidden, however, to regard the survival
of so many copies as accidental [when the disappearance of 983
is accidental] by the fact that the play was not reprinted until
1619.
In the same way in the case of Much Ado, of which fifteen copies
are extant, we may take the absence of any quarto reprint as
confirming the suggested deduction that the First Quarto was not
very successful.
My belief is that these kinds of books often survived only because
of collectors -- both of the deliberate and the pack-rat type. In the
early days of printing, 1st editions were not valued more than later
ones; so with ten editions enlarging the number of copies to 10,000,
the total survivors still approach zero, and it is surprising how this
holds true. Of course, repeated editions do indicate high sales,
but does a larger number of surviving single-edition copies indicate
low popularity, or does it reflect the availability to preservers of
only
one source of supply? Another researcher on pre-1640 books:
It was easy to find half a dozen copies from an edition of [100],
but it was difficult to find half as many [from a single edition?]
when a total of 10,000 was known to have been printed in the
various editions. Since copies from the largest editions are now
generally much less common than those from smaller editions,
one may conclude, somewhat paradoxically, that the less there
are, the more there were: that books were destroyed by the
diligence of their original readers.
The first conclusion is valid (with exceptions). The second I think
is inherited but mistaken opinion. Along these lines, which lead
to my questions, another observation:
Thus when only two or three editions were printed, all will survive,
and at least half a dozen copies can easily be found from each
edition. But when there were twenty or thirty editions, only a few
of them [editions] will survive, and those only in one or two copies
of each edition.
A final quote (I have more) about lost Tudor era (1500's) books:
The [Short Title Catalogue, listing all English books] . . . revised
letter M, taken as a sample . . . shows nearly 30% of entries with
only a single copy. More than half of these (180) would have been
completely lost [in any edition] without this fortuitous
survival . . .
Now, if I can formulate some reasonable questions. To begin with
an analogy: If an ocean liner with one lifeboat sinks, the number of
survivors correlate to the lifeboat, not to the number of passengers.
The almost complete disappearance of books numbering in the tens
of thousands (from multiple editions) alongside near disappearance
of smaller numbers from single editions implies a total destruction
of the smaller groups without intervention in the process.
This in turn implies 'single lifeboat' agents unrelated to the causes
of disappearance, but working against them in limited capacities
to preserve small numbers of each book, often stretched over many
editions, and often near zero.
For such events can statistics alone verify the need for agents to
account for the preservation of items within so narrow a range? In
other words, do I need to analyze the habits of collectors, libraries,
and other 'friends' of books to know that someone has intervened
in the destructive process -- or can these agents be deduced as
responsible for the blips of low numbers in the graphs?
Can any relative probabilities be measured by ratios of surviving
books when their initially varied populations have been almost
completely decimated? Can this be expressed in nonhuman terms?
For example, English Quaker Librarians interested in preserving
their literature decided in 1800 that
. . . it remains to separate the duplicates, such as can be
conveniently taken away, which will reduce the stock; so
many of a sort being an encumbrance . . .
Printed in the thousands; down to duplicates, which will be
reduced to single copies, but not by thumbing. Another example:
But, of course, unexplained variations must be expected in
these small numbers, and it goes without saying that statistical
probabilities alone could never allow one to correct the record
without examination of the copies themselves. A well-known
example is Thomas Heywood's A Woman Kilde With Kindnesse,
of which a third edition has at least 18 copies, while only one
earlier edition survives, and that only in a unique copy.
Don't numbers like this fall within the range to be expected from
thousands of titles? Can any meaningful probabilities be derived
from such beginnings? The problem with analogy and anecdote
is that authority in the Humanities derives less from empiricism,
mathematics, or demonstration than from hand-me-down opinion,
which can overpower disjointed questioning. Are any principles
besides common sense available? Does anyone know of a similar
topic that has been properly studied? Thanks for any response,
and I apologize for the length of the post.
Jerry Downs
.
- Follow-Ups:
- Re: Bibliography and Statistics
- From: Nick
- Re: Bibliography and Statistics
- From: Gerry Myerson
- Re: Bibliography and Statistics
- Prev by Date: Re: Pattern between cubes
- Next by Date: Re: The Collatz discrete primes!
- Previous by thread: Re: schoolkids construct extraordinary magic square
- Next by thread: Re: Bibliography and Statistics
- Index(es):
Relevant Pages
|
