Tolkien wrote an original Forward to
The Lord of the Rings which he later replaced in the second edition. This Foreward reads in part:
I have supplemented the account of the Red Book, in places, with information derived from the surviving records of Gondor, notably The Book of Kings; but in general, though I have omitted much, I have in this tale adhered more closely to the actual words and narrative of my original than in the previous selection from the Red Book, The Hobbit. That was drawn from the early chapters, composed originally by Bilbo himself. If “composed” is a just word. Bilbo was not assiduous, nor an orderly narrator, and his account is involved and discursive, and sometimes confused: faults that still appear in the Red Book, since the copiers were pious and careful, and altered very little.
This then can be used to explain much of the chronological and geographical discrepancies in
The Hobbit, in that the account is supposed to derive from the writing of a single person, possibly years after the events, with no help at the time from anyone who was with him on his journey. That errors are to be supposed to have been made is understandable, some by the original author, some by later copiests, and some by the modern teller.
Whether Tolkien originally intended this as an explanation for these problems I do not know.
Tolkien once in
The Lord of the Rings explains an error in his account, supposedly derived from Frodo, by this method. In a footnote to the first page of Appendix F Tolkien in the second edition:
In Lórien at this period Sindarin was spoken, though with an ‘accent’, since most of its folk were of Silvan origin. This ‘accent’ and his own limited acquaintance with Sindarin misled Frodo (as is pointed out in The Thain’s Book by a commentator of Gondor). All the Elvish words cited in Book II, chs 6, 7, 8 are in fact Sindarin, and so are most of the names and persons, But Lórien, Caras Galadhon, Amroth, Nimrodel are probably of Silvan origin, adapted to Sindarin.
This footnote is referenced by Tolkien in a further footnote in the chapter “Lothlórien” attached to the statement that:
Frodo could understand little of what was said, for the speech of the Silvan folk east of the mountains used among themselves was unlike that of the West. Legolas looked up and answered in the same language.
When Tolkien wrote this passage in his mind the Elves of Lórien did speak a Silvan tongue different from Sindarin and Tolkien later corrects this by making it an error attributed to Frodo.
There are various other apparent discrepancies in
The Lord of the Rings some of which might be explained by a metafictional assumption. See
http://tolkiengateway.net/wiki/Mista...kien%27s_works which includes most of them. Whether Tolkien intended them to be explained in this fashion I see as dubious. Cannot most of them be simple errors?
Besides these two examples I mention, the only other well-known example I know of is Tolkien’s statement in
Morgoth’s Ring that:
What we have in the Silmarillion etc. are traditions (especially personalized and centered upon actors, such as Fëanor) handed on by Men in Númenor and later in Middle-earth (Arnor and Gondor); but already far back – from the first association of the Dúunedain and Elf-friends with the Eldar in Beleriand – blended and confused with their own Mannish myths and cosmic ideas.
Tolkien then partly imagines a scientific version of his cosmos in which the earth rotates around the Sun and the Sun is as old as the Earth. But in his
Silmarillion story the Sun and Moon are created late in history from the Two Trees.
But this is only two versions of the history, a false but poetic mythological version and a supposedly historical version. Tolkien surely knew that genuine mythological traditions have stories that greatly contradict each other while he, except for the one case, continues to write a single version of his legendarium which changes. For example, his story of the Children of Húrin is a version of the tale that is consistent with itself, not like genuine mythological stories which have many variants.
Tolkien in general makes changes in his thinking which replace his earlier ideas across-the-board. Brljak states:
Tolkien’s mature fiction is centrally concerned precisely with this inability of the text to ever take us to that vanished, irretrievable “there”, from which even living memory was but the first remove.
That seems to me to be very wrong.
The Lord of the Rings and
The Children of Húrin work very much by taking us to what Brljak would like to see as a “vanished, irretrievable ‘there’”. So do individual genuine mythological works for the most part. Homer tells one version of a story, Apollonius Rhodes tells another, Ovid also tells another, and the stories often disagree when they overlap. Euripides’ plays sometimes disagree with one another when they touch the same story.
Tolkien also puts a strong emphasis on consistency. I see his works putting us
there as much as any author’s works do, whether the author is writing in an existent mythology as Shakespere does in
Troilus and Cressida and
A Midsummer Night’s Dream or an invented mythology as Mervyn Peake does in his Gormenghast books or Lord Dunsany does in
The King of Elfland’s Daughter.
That Tolkien has two versions of
The Silmarillion in theory comes, it seems to me, from is growing to dislike much of his
Silmarillion mythology because it breaks with science but still liking it for poetic reasons. Tolkien is attempting to have his cake and eat it too.