Interesting points made and questions raised by Tim Romano, Carrol Cox and
Tim Bray (under 'presentation' rubric).

First, to get it out of the way: 'blind / blind as a bat / ear'  in canto 2.
Cliche yes; but cliche recuperated, since bats see with their ears -- as
blind Homer saw the sea-surge (which is both visual and auditory), and saw
through the old men talk.

Yes, surely the laying out of the lines on the page matters supremely: it is
a visual notation to lead the voice  Eye and ear again, to work together.

And the indentations, with the degrees of indentation, determine a rhythm in
the reading -- the rhythm of the mind picking up the relations (syntactical
as well as musical) of the sequence of phrases.

'To compose in the sequence of the musical phrase' -- every word there is
necessary -- that is the key to Pound's verse from at least 1912 to the end.

The line-breaks are critical to the music, as marking the musical phrase;
so too are the indented lines or half-lines.
Punctuation is a secondary marker only -- it is grammatical, sometimes;  and
with that nearly always to mark a breathing or slight pause necessary to get
across the sense.

The thing to remember always is that Pound composed in his head and ear to
be read with the ear, not with the eye only (Hopkins' plea).  (See Mary de
Rachewiltz's memoir on his habit of composing.)

As for the mss.  They exist in considerable quantity, mostly in the Beinecke
Library at Yale; and there is a lot to be learnt from them.  Such as, that
from 1907 or '08 Pound was already in the habit of writing down his poems in
natural phrases rather than regular length lines.  The mss drafts of cantos
are all like that.  I.e. all those minute decisions to be taken about line
breaks, indentations etc. were being taken by him at the first stage of
composition.  Not always finally, but surprisingly often. There might be
successive mss drafts with revisions.  Then came the typescript stage, where
the format comes close to printer's layout and is what the printer worked
from. Tss and proofs exist and would be worth studying.  (I've been more
interested in the process of composition.)  Christine Froula's "To Write
Paradise: Style & Error in Pound's Cantos" (Yale U.P., 1984) is a first-rate
study of mss and tss, particularly for canto 4, with some copies of both.
My impression is that Pound wanted the printer to be faithful to his tss,
without being as bothered by their feeling the need to regularise as was
Robert Duncan.  (Duncan's care for the precise placing of his words on the
page was of course exemplary.)

Finally, re lines that over-run.  Pound did not like turned lines, as who
does, but where they were unavoidable his instruction to the printer was to
carry over a unit of sense or sound, and not just the excess word(s).  It
can be a teasing question to determine whether the phase isolated off to the
right is a separate line or part of the line before --but the ear can
usually settle which.

Post-finally, as to how to number and count lines in the Cantos ???   I find
it best to count each line as a line, except where it is evidently a
run-over.  I.e. no truck with 'half-lines' etc.  It seem to work that way.

ppf:  The Cantos should be printed with a more generous format than
publishers outside Paris seem able to imagine.  And I don't mean de luxe
editions at art auction prices.  Just a wide enough page to avoid turned
lines, and a balance of type and line-spacing to allow the words the white
space in which to register fully on the hearing eye.

David Moody