NOTES
DATING BLAKE’S SCRIPT: a postscript
A question raised by Professor Bentley that reached me too late for the June Newsletter has enabled me to cast out an error in my data and therefore in my reckoning of the time of Blake’s ceasing to use his left-serif “g”.
A crucial mistake lurked in my description of the “single errant plate” (see top paragraph, page 10, June Newsletter) which I deduced from the reproduction of the Eagle Ballad In A.G.B. Russell’s The Engravings of William Blake (1912). My mistake was to assume that since the Eagle design in Russell was a fairly faithful reproduction of the 1805 engraving, the inscription beneath it must be a faithful reproduction of a lost issue of the 1805 Ballads. It should have given me pause that the lettering was, as I noted, much less highly finished than that of actual engravings examined, which I took to be of a “second state of the plate”.
Bentley had cautioned me that inscriptions are not always the work of the original engraver, but I had insisted that a leftward g-serif connoted Blake. I can now see that the rather clumsily lettered inscription in the Russell electro-engraving represents an effort at retouching in 1912; a bad imitation of the original Inscription and not an early state of it. And I see that the 1805 inscription Itself is the work of someone whose serif is shaped like a tear drop or raised comma, unBlakean. What of the leftward turning serif in the Russell copy? Its left turn misled me (had Russell’s engraver noticed the odd g’s in Blake?), but in shape it imitates the unBlakean tear drop and is attached to the center rather than the left edge of the head of the letter.
Freed of this ghost serif, the “g” hypothesis can assume the simpler and more comfortable pattern that the rest of the evidence points to. Between November 5 1802 and March 25 1804 Blake changed his “g” serifs from left to right and, so far as the valid data indicate, never looked back.
Bentley asks how I “know that ‘A Divine Image’ was ‘on the back of one of the plates’ of the Songs.” The point is that it turned up in two posthumously printed copies of the Songs, c and n, and the explanation that it did so because it was on one of the surviving plates seems more probable than the alternative, that it was on an independent piece of copper kept among the Songs plates.
Further discussion of the chief remaining problem, of the precise definition of the first transition, from conventional to idiosyncratic g’s, will be welcome.