"GL3" - читать интересную книгу автора (vol03)

minor and as it were casual modifications to individual lines here and.
there.
After much experimentation I have concluded that to make a single,
text, an amalgam derived from the latest writing throughout the poem,
would be wholly mistaken. Quite apart from the practical difficulty of
changed names in the rewritten parts that do not scan in the old lines, the
later verse in its range and technical accomplishment is too distinct; too
much time had passed, and in the small amount that my father rewrote of
the Lay of Leithian after The Lord of the Rings we have fragments of a '
new poem: from which we can gain an idea of what might have been. I
have therefore excised these parts, and give them subsequently and
separately (Chapter IV).
A further reason for doing so lies in the purpose of this book, which
includes the consideration of the Lays as important stages in the evolu-
tion of the legends. Some of the revisions to the Lay of Leithian are at
least 30 years later than the commencement of the poem. From the point
of view of the 'history', therefore, the abandonment of the poem in or
soon after September 1931 constitutes a terminal point, and I have
excluded emendations to names that are (as I believe) certainly later than
that, but included those which are earlier.* In a case like that of
Beleriand, for instance, which was Broseliand for much of the poem in
B and always later emended to Beleriand, but had become Beleriand as
first written by line 3957, I give Beleriand throughout. On the other
hand I retain Gnomes since my father still used this in The Hobbit. '
The many small changes made for metrical/stylistic reasons, however,
constitute a problem in the attempt to produce a 1931 text', since it
is often impossible to be sure to which 'phase' they belong. Some are

(* This leads to inconsistent treatment of certain names as between the two long Lays, e.g.
Finweg son of Fingo1fin in The Children of Hurin but Fingon in the Lay of Leithian.
Finweg survived into the 1930 version of 'The Silmarillion' but was early emended to
Fingon.)

demonstrably very early - e.g. candle flowers emended to flowering
candles (line 516), since C. S. Lewis-commented on the latter - while
others are demonstrably from many years later, and strictly speaking
belong with the late rewritings; but many cannot be certainly determined.
In any case, such changes - very often made to get rid of certain artifices
employed as metrical aids, most notably among these the use of emphatic
tenses with doth and did simply in order to obtain a syllable - such
changes have no repercussions beyond the improvement of the individual
line; and in such cases it seems a pity, through rigid adherence to the
textual basis, to lose such small enhancements, or at any rate to hide them
in a trail of tedious textual notes, while letting their less happy predeces-
sors stand in the text. I have thought it justifiable therefore to be frankly
inconsistent in these details, and while for example retaining Gnomes
(for Elves or other substitution) or Thu (for Gorthu or Sauron), to
introduce small changes of wording that are certainly later than these.
As in the Lay of the Children of Hurin there are no numbered notes
to the text; the annotation, related to the line-numbers of the poem, is