"Canonicity II": Has Yours Changed?

[William Cloud Hicklin]

I think there is a tendency often to very much exaggerate the degree of Authorial Intent to be attached to the published text of the Lord of the Rings, as if "This is FINAL and OFFICIAL and represents Tolkien's consciously determined LAST WORD." The more I work with Tolkien's manuscripts, the clearer it is to me that this is not at all the case; that the text published in 1954-55 represents merely a contingent state, a transverse cut across an ongoing development that became "fixed" not because Tolkien had decided that it was finished, final and complete, but because it was physically taken out of his hands so that he couldn't continue to monkey with it. Even in the course of the typescript for the printers - weeks and months past deadline - he was rewriting things. Even on the galley proofs (which for normal authors exist to correct typos), he was rewriting things. Had not Rayner Unwin practically put a gun to his head and made him give up the papers, he probably would have spent the last two decades of his life continuing to alter and reshape The Lord of the Rings just as he did with The Silmarillion.

[Galin]

Yes but William, what about the art of subcreation? Something has to be the story, and even Tolkien bowed to that.

And if JRRT wants to niggle away for years, no problem, but he's still working toward a story to be published. And if he could change his mind about something right up to the point that that something was physically taken from him -- I'd say all the more reason to consider author-published work as canon. At least we know "that much" about a given thing: it prevailed into the Secondary World.


And even when JRRT just couldn't help himself, he still realized that stepping on already published work is simply not the same as continuing to niggle or revise aspects of the subcreated world that nobody knows about. The Secondary World exists on bookshelves for a once and future readership.

"It may be suggested that whereas my father set great store by consistency at all points with The Lord of the Rings and the Appendices, so little concerning the First Age had appeared in print that he was under far less constraint. I am inclined to think, however, that the primary explanation of these differences lies rather in his writing largely from memory. The histories of the First Age would always remain in a somewhat fluid state so long as they were not fixed in a published state; and he certainly did not have all the relevant manuscripts clearly arranged and set out before him."

Foreword, The Peoples of Middle-Earth

To me, this is another way to say that Tolkien (and quite naturally in my opinion), treats fixed work as canon.


Ursula Le Guin did some fancy dancing with Earthsea, for another example I've used in the past. But she herself published the later books of course, leaving no question as to whether she truly wanted to shine such a new light on Earthsea.

[William Cloud Hicklin]

To me, this is another way to say that Tolkien (and quite naturally in my opinion), treats fixed work as canon.

Or is Christopher saying that, without a bound book ready to hand and his manuscripts in disorder, his father's memory couldn't keep it all straight? This would apply especially to the 'late writings' 1968-73, since not only was Tolkien elderly, but due to an unfortunate accident his papers had become hopelessly jumbled during the move to Bournemouth. To take one example of his fading mental powers: the Glorfindel essays state that he was now stuck with the name because it had appeared in print in the Lord of the Rings, therefore Glorfindel of Rivendell must necessarily be the same person as Glorfindel of Gondolin and an explanation of how this was possible was necessary. But of course, there was nothing whatsoever preventing him from just renaming the Gondolin hero; it seems he was getting somewhat confused (or he could have let it ride, as he did with Legolas and Galdor)

[Galin]

Originally Posted by William Cloud Hicklin

Or is Christopher saying that, without a bound book ready to hand and his manuscripts in disorder, his father's memory couldn't keep it all straight?

I think he's saying that as well -- after the first sentence about consistency with already published works

Which brings up another issue: if Tolkien simply forgets something already in print and "steps on it" in a late text, has he truly, consciously revised that something?

This would apply especially to the 'late writings' 1968-73, since not only was Tolkien elderly, but due to an unfortunate accident his papers had become hopelessly jumbled during the move to Bournemouth. To take one example of his fading mental powers: the Glorfindel essays state that he was now stuck with the name because it had appeared in print in the Lord of the Rings, therefore Glorfindel of Rivendell must necessarily be the same person as Glorfindel of Gondolin and an explanation of how this was possible was necessary. But of course, there was nothing whatsoever preventing him from just renaming the Gondolin hero; it seems he was getting somewhat confused (or he could have let it ride, as he did with Legolas and Galdor)

I think the late Glorfindel text II contains a good example of Tolkien following his canon: in this text he negates the idea that Glorfindel might have been Sindarin -- negating it due to what is already in print -- if Glorfindel of Gondolin is supposed to be the same person as Glorfindel of Rivendell of course. In other words, The Lord of the Rings comes first, or in other other words, JRRT will naturally try to figure out the scenario giving top priority to already published description.

In other other other words, don't break the enchantment here.

And in my opinion, in this late text Tolkien is simply choosing to see if a satisfactory answer can emerge without altering the name of the Elf of Gondolin, and if not, he has the option of altering it.

I think a good example of Tolkien's shaky memory here might be the late detail that Gondolin was occupied by a people of almost entirely Noldorin origin, which is closer to the original conception of Gondolin's folk than description found in The Grey Annals.

[Saurondil]

To make a Biblical analogy: for most Christians - with the significant exception of Protestantism - there are 14 chapters in the OT Book of Daniel. In the Jewish Bible, there are 12. Protestantism very early adopted the Jewish reckoning. Chapters 13 and 14 (and about 2/3 of chapter are Greek additions to the body of the book, which is in Hebrew. These "Additions to Daniel" are 3 of the works commonly known to Protestants as "the Apocrypha". Catholics call the 12 Hebrew chapters "proto-canonical", and the Greek additions "deutero-canonical" - all 14 chapters are recognised as equal in canonicity, but the Hebrew parts of the book enjoy a certain priority over the Greek additions.

Similarly here. IMHO:

1) TH, LOTR, and the 5 works in the Silmarillion count as fully canonical. When editions of a work disagree or are in error, I take the most recent as (unless otherwise indicated) the surest guide to the author's latest canonical intentions - later ideas not published or not published as books, do not have the same authority. They would count as protocanonical, with the possibility of a gradation even within those works.

2) In second place come unfinished pieces such as those in UT. I think of them as canonical in a lesser degree - they are canonical, in so far as they agree with, or at least do not contradict, the works in group 1. I would reckon some of them as at least deuterocanonical.

So the works in UT about Numenor would count as deuterocanonical - the discussion of Celeborn and Galadriel, might not. The Disaster of the Gladden Fields disagrees with Of the Rings of Power and the Third Age over why the disaster happened. As ORPTA makes an error about the relative dates of the Finding of the Ring & the ending of the Kings in Gondor, I accept ORPTA as canonical, but not as always.correct; and I would supplement it by TDGF, and treat TDGF as (in this respect) more reliable. A work of generally more authoritative status can therefore be supplemented, and even corrected, by a work of lower status.

The three incompletely consistent deaths of Isildur, in LOTR, ORPTA & TDGF, are rather like the three incompletely consistent deaths of Antiochus IV in the Apocrypha; or like the two deaths of Judas Iscariot in the NT.

How many (legitimate) Rulers of Numenor were there ? According to the 1977 Sil and Appendix A of LOTR, 24; according to UT, 25. The list in UT explains and corrects those in Sil & LOTR, so I go with UT. Aldarion and Erendis adds a lot of detail, including many names, to the canonical info about Numenor; so I accept those details as accurate and canonical.

3) In third place I would put the Book Of Lost Tales. I don't regard any of it as canonical. It is all certainly of great interest, but as evidence of the development of Tolkien's imagination and of the development of the stories; not as a source of lore about the feigned history and the world in which it is set. The archaic literary style is not what separates it from the first 2 groups. What makes it different, is that many of the ideas "are in serious disharmony" with ideas in those other books.

4) After that, I would put HOME volumes 3 to 12.

If groups 1 and 2 are like the parts of the Catholic canon of the Bible, perhaps groups 3 and 4 are like the Jewish legends and speculations that grew up around the Biblical material - not canonical material, not even secondary to that, but in a circle of ideas even further removed from the "epicentre" of full canonicity. If full canonicity is like the epicentre of a splash in a pool, group 4 shares in the same imagination as created the first 3 groups; but is of much lower authoritativeness.

I think the question of canonicity is well worth discussing; but I think it cannot be separated from questions of authoritativeness and authenticity.

[William Cloud Hicklin]

I'm not sure about ranking The Silmarillion with such priority. It was after all NOT published by JRRT; and Christopher, quite openly, had to make a large number of decisions about which version or which text to include or exclude, some of which he later came to regret.

Just for one: Is Gil-galad "canonically" the son of Fingon? I would say no.

[Findegil]

Posted by William Cloud Hickling:

For me, anyway, my attitude is very process-oriented. Tolkien's corpus, published and un-, is a collective artifact which was the work of one author's lifetime. Sometimes he changed his mind. Sometimes he made mistakes. That is what I find fascinating. I just don't see much point in efforts to rather artificially determine One Truth when there is no underlying truth- it's all fiction.

This is, I think, a valid view for what I will call studying the Tolkien corpus as literature. But it does not help if you are more interested in the study of the 'Legend' itself rather than its literary value. You are of course fully right that there can not be a 'canon' since there is no true story behind - it's all fiction. That is why I do not like the name canon and canonicity for discussions like this one, but as yet nothing better had come up. Nonetheless in my point of view, the Legend of Middle-earth would never have been such a success in publication, if it would not have provoked the reader to look ever deeper into it. It is for sure part of Tolkien's successful literary techniques to create that kind of depth behind the told story - most of it in a kind 'real', since he had already written the background stories (e.g. Elrond recalling the End of the First Age or Aragorn singing part of the Lay of Leithian), some feigned since he only later drafted them (like the Cats of Queen Beruthiel or the Five Wizards). I think that only a fraction of the people out there that have bought one or more of Christopher Tolkien's editions of his father's earlier or later works have done so to study JRR Tolkien's corpus as literature. Another fraction (and it might be the greater one) were more interested in the stories itself. When ever I was at a meeting of Tolkien enthusiast, I could determine these fractions. You might call them LIT. and FAN. (you can choose if it is short for 'fantasy' or 'fanatic' or both). Even if the meeting was focused on LIT. in the audience you would find some FAN. and anyhow by some speaker you would mark that they have started their interest in Tolkien or fantasy as genre as FAN. even so they have come to speak now as LIT. (So you could argue that LIT. would be more grown ups while FAN. would be more childish or that for LIT. the 'spell' has been broken while FAN. do everything to maintain it - but both would be polarising more than seems necessary.)

Let's come to my approach to what is here call 'canonicity'. I am with William Cloud Hickling about The Silmarillion, but for a quiet different reason. For me there are only 4 sources in priority 1 - books published by JRR Tolkien:
- The Lord of the Rings (including the Appendices
- The Hobbit
- The Adventures of Tom Bombadil
- The Road Goes Ever On

It is a strange mix and the books have even in themself some failures never corrected (e.g. Ghan-buri-Ghan counting the Rider of the Mark) and some inconsistencies from one to the other (e.g. Thorin and Co needing much less than a day ride from Mitheithel to Trolls while Strider needs several days or Galadriel have set a ban on her return or not between LotR and RGEO).

Prio 2 is sources given out to a restricted public by JRR Tolkien e.g.:
- Letters by JRR Tolkien (not so much what he sent to his family or his publisher, but more so what he sent to readers asking questions.
- Parts of The Lost Tales that JRR Tolkien published in today arcane publications. This includes some of the poetry from that period.
- Parts of The Lost Tales that JRR Tolkien read to some public audience (so not the Inclings or similar private groups). Here I think mostly of Tuor and the Fall of Gondolin.
...

Prio 3 are sources published in a documentray style by Chirstopher Tolkien and others e.g.:
- Unfinsihed Tales
- The History of Middel-earth
- Beren and Lúthien
- Tuor and the Fall of Gondolin
- The History of the Hobbit
- [b]Param Eldalamberon[/(b]
...

Only in Prio 4 will be found books published by Christopher Tolkien as belles-lettres:
- The Silmarillion
- The Children of Hurin

And to come back to the topic of the thread: Yes, this has changed over time. As you may guess the above priority are a mix of attitude of the author against the text (ready for full publication or for restricted audience, a draft, ...) and level of information we have about the content of the text and the circumstances of composition. Earlier my view on 'canonicity' was rather based on the time of composition modified a bit by 'completeness' of the given information (a later rewritten small detail would only change that detail and not render the full described older story un-valid). It was like looking on a pastiche-picture: In some parts the original canvas with its first painting would still be seen, in other there was layer upon layer of new material. Some overlapping each other, some extending the picture. With each new layer covering what was beneath (a bit like the First Lord of the Ring map). But some layers would only be like a thin net: fine threads of narrative drafts with some knots where more substantial information is given, while other would be like a piece of new canvas glued on the old picture (full retelling of a tale).
Today we have to add some transparency to that picture: the higher the priority given above the more 'dense' that piece of pastiche is. Thus, with in the same priority time of composition is still the sorting criterium. But looking through the more transparent parts, they would look 'denser' if the layer beneath shows the same and more blurred if it is different. A later low priority source could thus still have an effect on an earlier high priority source, but it is no longer covering it. But sources of high priority will cover the deeper (older) layers well enough, and these layers may only peep trough where the high priority sources leave some gaps.

To take up the discussion from above: A FAN. will study the sources to discover what parts are still to been seen (looking form atop the pastiche) or he would make a parallel (horizontal) cut (lifting up some layer) and look on a the remaining layers for the fascination of that layer itself. A LIT. would rather make a crosscut to follow the development of some elements that return in many layers or he would cut out a single layer to analyse the technique used in that layer.

I hope this makes some sense at all, but the example of the pastiche was the best I could come up with. And for sure for greater clearness, the LIT. and FAN. characterisation is painted much more black and white than it is in real life. So if anyone found this characterisation embarrassing, please take this as an apologise. It was not meant in any harmful or disrespecting way. It is just a difference a percieve that often leads to misunderstanding on both sides, especially in canon-discussion like this.

Respectfully
Findegil