This is the second in a series of posts about the textual variants I’ve found in printings of The Silmarillion. In this post, I’ll cover six more changes to words in the text. This will finish up all the non-punctuation changes in the main text of the book.

Previously we introduced the notion of a locus and a pagination and introduced the text forms I’ve been looking at (each assigned a lowercase Greek). After introducing six new loci, in a similar manner to in the previous post, I’ll start to sketch out some history that seems to be hinted at by the textual variants.

The Next Six Changes to Look At

The next six changes are not obviously errors like in the previous post. One is a change in capitalisation, one a dropping of a definite article used with a name, one is an interesting change in the choice or word and three are changes to the numering of kings.

In summary:

  • Locus 27. the Mormegil > Mormegil (A217 B261 C199 D222)
  • Locus 24. Forest > forest (A203 B245 C186 D206)
  • Locus 22. makers > masters (A196 B236 C178 D195)
  • Locus 31. nineteenth > twentieth (A267 B322 C245 D276)
  • Locus 33. three > four (A270 B324 C247 D278)
  • Locus 32. twenty-second > twenty-third (A268 B322 C245 D276)

Locus 27. the Mormegil > Mormegil

  • known … ere the end that the Mormegil was none other than Túrin — αβγδε
  • known …ere the end that Mormegil was none other than Túrin — ζηθ

(A217 B261 C199 D222)

This change is a late one that only (of the texts I looked at) appears in the current HarperCollins hardcovers (ζ), the ebooks (θ), and the 2004 Illustrated Edition from Houghton Mifflin (η). Notably, the change was not made in the 2001 Houghton Mifflin hardcover (γ).

The change (like the next one) seems a little odd to me. Earlier in this very paragraph and in the two preceding paragraphs, the definite article is used. The only other time “Mormegil” is used without the article is the first time it is mentioned but even then, it is immediately used with the article:

So great was his prowess and skill in warfare on the confines of the Guarded Plain that he himself became known as Mormegil, the Black Sword; and the Elves said: ‘The Mormegil cannot be slain, save by mischance, or an evil arrow from afar.’

although there’s perhaps a nice enveloping if the first and last use lacking the article.

I would love to get to the bottom and when and why this change was made.

Locus 24. Forest > forest

  • looked, and descried the Forest of Brethil climbing green — αβγ
  • looked, and descried the forest of Brethil climbing green — δεζηθ

(A203 B245 C186 D206)

Here is another change that is even more odd to me, although appears in much earlier texts, going back to the 1979 Unwin Paperback (δ). Interestingly the change was not made in the 2001 Houghton Mifflin Second Edition (γ) but was made in the 2004 Illustrated (η) and in the ebooks (θ).

What is strange is that, in the ebooks, of six other times the expression “Forest of Brethil” appears, “Forest” is capitalised five times and lowercase only once. So we went from six instances of “Forest of Brethil” and one of “forest of Brethil” in the 1st Edition, to five instances of “Forest of Brethil” and two of “forest of Brethil” in the ebooks.

The fact the change was seemingly first made in the 1979 Unwin Paperback makes this even stranger.

I should note that I have only looked at the one locus. It is possible that texts along the way changed other instances of Forest only to have them reverted again at a later date.

Locus 22. makers > masters

  • Smiths and miners and makers of fire they brought — αβδε
  • Smiths and miners and masters of fire they brought — γζηθ

(A196 B236 C178 D195)

This seems to firmly be a “Second Edition” change. The change exists in the 2001 Houghton Mifflin hardcover (γ) and in all the versions published since that I’ve looked at.

The change is somewhat understandable, coming out of a desire to distinguish between actually creating fire and merely mastering it.

Locus 31. nineteenth > twentieth

  • And the nineteenth king took the sceptre — αδ
  • And the twentieth king took the sceptre — βγεζηθ

(A267 B322 C245 D276)

This is the first of three corrections made to the numbering of Númenórean (or is that Númerórean?) kings. It failed to make it into the 1979 Unwin paperback (δ) but did make it into the 1983 Unwin (ε) and the 1983 Houghton Mifflin paperback (β).

Locus 33. three > four

  • foundation of Númenor; and three and twenty Kings and — αδ
  • foundation of Númenor; and four and twenty Kings and — βγεζηθ

(A270 B324 C247 D278)

As with the previous locus, this king-numbering change made it into the 1983 Unwin (ε) and the 1983 Houghton Mifflin paperback (β).

Christopher Tolkien actually mentioned the king-numbering changes in the Preface to the Second Edition of The Silmarillion and refers to a note in Unfinished Tales. The Akkalabêth originally had twentieth and four in loci 31 and 33 respectively but Christopher “corrected” them based on attempting to resolve an inconsistency in the LOTR appendices. An alternative resolution came to light in the preparation of Unfinished Tales and so the original numbers were restored. This might be worth a whole post on its own at some later date.

Locus 32. twenty-second > twenty-third

  • For Ar-Gimilzôr the twenty-second king was the greatest — αβδε
  • For Ar-Gimilzôr the twenty-third king was the greatest — γζηθ

(A268 B322 C245 D276)

This third king-numbering change did not make it in to either of the 1983 paperbacks, Unwin (ε) or Houghton Mifflin (β) but did make it into the 2001 Second Edition (γ) and subsequent texts.

This change is not referred to by Christopher and post-dates Unfinished Tales so must have been identified later.

The Genealogy of Text Forms So Far

Let’s look at the 12 loci we’ve reviewed so far and which texts had which variant.

We’ve come across the following patterns:

  • αβ vs γδεζηθ (20, 35)
  • αδ vs βγεζηθ (30, 31, 33)
  • αβγ vs δεζηθ (21, 24, 37)
  • αγδε vs βζηθ (18)
  • αβδε vs γζηθ (22, 32)
  • αβγδε vs ζηθ (27)

It is clear that the first Unwin paperback (δ) and the first Houghton Mifflin paperback (β) had independent corrections.

  • Changes made in δ not β: 20, 35, 21, 24, 37
  • Changes made in β not δ: 30, 31, 33, 18

There were no changes made to both!

Next on my list to check out is the First Edition Ballantine. It will be interesting to see where that falls.

Subsequent Unwin paperbacks (ε) had (as might be expected) all the changes made in δ plus 30, 31, 33. That ε didn’t fix 18 suggests that the new corrections on top of δ were made independently of β.

The 2001 Houghton Mifflin Second Edition (γ) and presumably (it’s on its way, courtesy of Jeremy Edmonds) the 1999 HarperCollins still have errors already fixed in those paperbacks so almost certainly were not based on them.

  • Changed in δ or ε but not γ: Hador > Haldor (21); Forest > forest (24); Númeróreans > Númenóreans (37)
  • Changed in β but not γ: Leithien (18)

It did make some corrections (20, 35, 30) and the first two king-numbering fixes (31, 33); and also had as innovations the makers > masters (22) and the third king-numbering fix (32).

Christopher Tolkien says in his Preface to the Second Edition that he removed errors in the text (and index, which we have not considered at all here) which had “escaped correction in the hardback printings (only)”. The “only” is particularly interesting. It suggests corrections which had already been made in the paperbacks. As best I can now tell, those changes were at loci 20, 35, 30, 31, 33 (plus any index corrections).

Of the twelve loci we’ve looked at in our two posts, four changes had made it into a paperback which did not make it into the Second Edition hardcover:

  • 21 Hador > Haldor (already in Unwin paperbacks)
  • 24 Forest > forest (already in Unwin paperbacks)
  • 37 Númeróreans > Númenóreans (already in Unwin paperbacks)
  • 18 Leithien > Leithian (already in Houghton Mifflin paperbacks)

And the only entirely yet-to-come change was:

  • 27 the Mormegil > Mormegil

Across all 40 loci (not just the ones we’ve seen), the 2001 Second Edition Houghton Mifflin (γ) doesn’t have any changes that aren’t also in the 2004 Illustrated Edition, the 2006 HarperCollins hardcover, and the 2007 HarperCollins Deluxe. This suggests that all those come from the Second Edition (which is as is claimed).

It seems so far that, from the First Edition (α), at least three separate texts developed:

  • the Unwin paperbacks starting at δ and from that ε
  • the Houghton Mifflin paperbacks starting at β
  • the Second Edition γ from which the current HarperCollins harcovers (ζ), 2004 Illustrated Edition (η), and ebooks (θ) came.

But as we’ll see in the next post, there are some punctuation changes which add a little more nuance to the genealogy.