"GL1" - читать интересную книгу автора (vol03)

first; it is indeed possible, and would not be in any way uncharacteristic,
that my father began work on II while he was still composing at a later
point in I. II is essentially an expansion of I, with many lines, and blocks
of lines, left virtually unchanged. Until the second version is reached it
will be sufficient to refer simply to 'A' and 'B', the manuscript and
typescript of the first version.
The manuscript A consists of two parts: first (a) a bundle of small
slips, numbered 1 -- 32. The poem is here in a very rough state with many
alternative readings, and in places at least may represent the actual
beginnings, the first words written down. This is followed by (b) a set of
large sheets of examination paper from the University of Leeds, num-
bered 33 ff., where the poem is for the most part written out in a more
finished form -- the second stage of composition; but my father wrote in
line-numbers continuously through (a) and (b) -- lines 1 -- 528 in (a), lines
528 ff. in (b). We have thus one sole text, not two, without any overlap;
and if (a), the slips, ever existed in the form of (b), the examination
sheets, that part has disappeared. In part (b) there are many later
emendations in pencil.
Based on this manuscript is the typescript B. This introduces changes
not found in A or its emendations; and it was itself emended both in ink
and pencil, doubtless involving several movements of revision. To take a
single line as exemplification: line 8 was written first in A:
Lo! Thalion in the throng of thickest battle
The line was emended, in two stages, to
Lo! Thalion Hurin in the throng of battle
and this was the form in B as typed; but B was emended, in two stages, to
Lo! Hurin Thalion in the hosts of war
It is obvious that to set this and a great many other similar cases out in a
textual apparatus would be a huge task and the result impossibly compli-
cated. The text that follows is therefore, so far as purely metrical-stylistic

changes are concerned, that of B as emended, and apart from a few
special cases there is no mention in the notes of earlier readings.
In the matter of names, however, the poem presents great difficulty;
for changes were made at quite different times and were not introduced
consistently throughout. If the latest form in any particular passage is
made the principle of choice, irrespective of any other consideration,
then the text will have Morwin at lines 105, 129, Mavwin 137 etc.,
Morwen 438, 472; Ulmo 1469, but Ylmir 1529 and subsequently;
Nirnaith Ornoth 1448, but Nirnaith Unoth 1543. If the later Nirnaith
Onroth is adopted at 1543, it seems scarcely justifiable to intrude it at
lines 13 and 218 (where the final form is Ninin Unothradin). I have
decided finally to abandon overall consistency, and to treat individual
names as seems best in the circumstances; for example, I give Ylmir
rather than Ulmo at line 1469, for consistency with all the other occur-
rences, and while changing Unoth to Ornoth at line 1543 I retain
Ornoth rather than the much later Arnediad at line 26 of the second
version -- similarly I prefer the earlier Finweg to Fingon (I975,
second version 19, 520) and Bansil, Glingol to Belthil, Glingal
(2027 -- 8) . All such points are documented in the notes.