View source for Descriptive Bibliography, RBS (July 2010)

Introduction to the Principles of Descriptive Bibliography, Rare Book School (Charlottesville, VA) July 26-30; Richard Noble, David Whitesell

Making of a  book like the making of a sausage -- what cannot be in two places at once

Book as an assemblage -- assemblage of assemblages; things all the way down -- synchronic assemblages (forms, type) and diachronic assemblages (same piece of type can occur only once in a form, but can reappear later in other forms)

Book is ''modular'': early example of modular building

Book is fungible, but bibliography is looking for ''patterns''

Bibliography is an attempt to reverse-engineer the book

== Bibliography ==

'''pagination''': omit detailed records of routine omissions (chapter headings, etc.), because disguises actual interesting bits of omission (don't make something simple look complicated, or something complicated look simple); pagination should always come out even (foliation should ''generally'' come out even); use commas to separate sections

use commas sparingly

put number of unnumbered pages in italics in brackets

'''ideal copy''' conforms to issue --  is simply a reference standard, not a "perfect" copy (somewhat of an oxymoron); like an "average" copy

basic unit that we're trying to get back to is the '''form''' -- the printing surface; e.g., take any sheets cut from one part and stuck in another, and show that in the formula

assume printing begins with the first page of text

π1 A-C^2 [D]1

or?

π1 [=D2] A-C^2 [D]^2 (-D2)

example: Hoyle card games book; owners could bring in their copy of the book to get new pages added to the front, refresh their editions -- this is a concern of ''issue''

example: only "ideal copy" of Shakespeare's First Folio is Hinman's facsimile edition, since it contains only corrected sheets

alike as peas, or alike as buttons?

formula is basis for making meaning out of the structure of the book

relationship between printed formula and issue formula == analysis

formula is encoding relationships; not simply enumerative practice, but analytic

''edition'': all those iterations of a text from a given composition

("work", "expression", "manifestation", "item" -- cataloging terms that work for multiple media)

leaf is a binding phenomenon, not a printing phenomenon (that would be the form, which deals with pages) i.e., bibliographer's unit is the page (occupying a certain position in a form with other pages, binder's unit is the leaf

Greg saw collation formula as a way of reference; Bowers took this and said it was a report of a particular analysis

table of contents comes after pagination; important in understanding how important anomalies are in regards to the content

cancellations: H^4(+/-H3); H(H3) refers to the inserted leaf; H3 refers to the cancelled original

'''quasi-facsimile transcription''': some problems: no "W" in type font; English printers soldered together two "V"s; if totally separate, write them totally separate, if not, write "W" and indicate in note; 

antiquarian practice now, though (why do it?); can't get across white spaces or exact vertical relationships; often doesn't putll out important differences

== Paper ==

pre-1820, almost certain to be handmade;
post-1840, almost certain to be machine-made
1820-1840, transitional phase

laid and wove can be handmade or machine-made

in Western world, all paper up to 1760 is laid

handmade laid paper: 
around chainlines, paper pulp gathers more, creating '''bar shadows''' -- 1/4in. darker area 
-- incicates handmade paper

white/light spot on paper -- created from a drop of paper that forms a depression (''vatman's tears'')

mold side: meets the paper mold; has a kind of grid line characteristic
felt side: meets the felt; flatter, smoother

shadowless laid paper: modern handmade laid paper that doesn't produce bar shadows

Brits leaders in machine-made paper; by 1820s, very rare to find handmade paper in London

in German-speaking territories, handmade paper used well into 1840s

some early wove papers, you still get chainline-like impressions -- rib marks

machine-made paper: two ways -- pulp spread on continuous belt (Fourdrinier), or on revolving drum of wire mesh (cylinder and vat machine); other impressions can be made on it by running it under a dandyroll (e.g. fake chainlines, watermarks

deckle edge can be found on cylinder vat machine; usually cylinder creates continuous web, but if dropped onto rubber belt with felt shapes on it, pulp sticks to felt but runs away from rubber; can create different shapes 

watermarks:
* originally identified the papermill;
* by 1830, indicates paper size and quality
* by c18, find paper with no watermark at all
* location of watermark varies
* identifying watermarks not exact

describing paper:
* dimensions
* watermarks, countermarks
* thickness of paper
* thickness of book / number of leaves
* chainline patterns and widths
* number of wirelines in 3cm space

paper as bibliographical evidence:
* anomalies
* deckle edges, cut edges

== Printing ==

type is not an abstraction, not the printed impression, but an object - ''something you pick up in your hand''

diagram of type: page 9 in Gaskel, page 31 in RBS workbook

Phillip Gaskell, "A nomenclature for the letter-forms of Roman type," ''The Library'' 5.29 (1974): 42-51.

comparing typefaces:
* x-height; also relative relation of ascenders to descenders
* axis or stress; draw vertical line through "o" -- is it vertical or oblique?
* weight: thin and light letters, or thick and heavy?
* contrast (difference bw thick and thin spaces -- greater diff in 19c on)
* set -- how much relative space does the font take up?
* serifs

point size introduced in 18c; 72 points in an inch; measured ascender to descender

ascender and descender between two lines; if space between them, lines have been leaded to create space between them

list of common type sizes of hand-press period: pg 35 in RBS workbook

paper is printed wet; shrinkage may change height of type impression

some type may not be set solid (ascender hits bottom of body / descender hits top of body), but leaded (space between top of ascender / bottom of descender and body) -- i.e., body size is different from font size

15c, up to 300 sorts in a font

'''ligature''': joining of letters

'''logotype''': several letters together on one sort

'''kerned''': ascenders/descenders going above/below the body of the type

'''em''', '''en''': "M" is square; em is size of capital "M"; en the size of capital "N"

'''titling type''': height matches body size, doesn't account for ascenders; didn't include puncuation, so sometimes comma in titles will be set high (see pg. 33 of RBS workbook)

by 16c, typefounding becomes specialist -- number of available fonts goes down; within any one firm, there is standardization

Tanselle, "Identification of Typefaces" -- RBS exit reading list T50

== Bookbinding == 

until recently, tend to study bookbinding through decoration; in the last generation, others beginning to study the structure

David Pearson -- item B25 on exit reading list

bookbinding trade: low status, modest profits

in earlier period, more likely for folios to be nested; by 18c, more common for folios to be in 2s, and begin to see 2-on and 3-on sewing, in which several gatherings were sewed alternatively

can find trade/retail and custom/bespoke bindings at any end of the spectrum

printers try print most efficiently; paper most expensive thing that went into the book

printer furnishes the bookbinder with a kit of sheets, as well as instructions for binding; instructions had to be understood in multiple languages, so signature marks developed

start to see signature marks in 1470s (see them earlier, but often by hand)

manuscripts also have marks indicating binding structure

register: printer's explicit statement of what constituted the book -- how sheets are signed and arranged (16c); persisted longest in Venice; expanded signatures, may have been legal document showing book to be "collated and perfect" (see workbook pg. 20)

if printer doesn't tell bookbinder where the plates go, up to the bookbinder to determine it

cancelled leaves: can either leave stub on cancelled leaf, or the leaf to pasted to; sometimes hard to determine where the stub is

collational formula affected by bookbinder's decisions