A question of canonicity

[KingODuckingham]

I was thinking about the elf-orc thing again after watching FOTR and seeing Saruman tell Lurtz that the orcs were elves once, and my brain spun off a question I'm not sure has a good answer. I think that makes it good for collecting potential opinions on though. We know Tolkien went through many drafts and revisions of various aspects of M-e and its universe. Is there ever a time when we would take the word of younger Tolkien as canon over older Tolkien? We have an essentially omniscient narrator, so far as these things go, and while it is generally easy to accept the final word the narrator gives, there's no guarantee that a) he would not have changed his mind further/again, or b) that he was 'right' to change his mind in the first place.

It still seems to me that we would need something at least slightly extraordinary to justify taking an earlier version over a later version; what might such a situation be? Or do you know of one?

[Elenhir]

People do this frequently with some of the material in the early texts of 'Myths Transformed'. All the stuff about the Sun always existing is not something that Tolkien backed away from, but it does end up creating a more confusing and unnecessary mess than what he hoped to avoid by going that direction in the first place. So the pieces that don't break anything are taken forward, but the rest is usually ignored.

[Chrysophylax Dives]

I have a problem with the very idea of canonicity. Say I am interested in the magic ring of The Hobbit. Canon says this magic ring is the One Ring. But (and I think the movies demonstrated this) if you read the One Ring into The Hobbit you make a mess of the original story.

[Eldy Dunami]

I've become increasingly less enamored of the whole concept of canon over time. When I was first getting into Lore as a teenager, I really wanted to have a single, internally consistent version of Arda to mentally explore. Call it a Hobbitish impulse ("they liked to have books filled with things that they already knew, set out fair and square with no contradictions"; LOTR, Prologue). However, I think this approach is ultimately detrimental to appreciating the legendarium. Christopher Tolkien made this point in the Foreword to the 1977 Silmarillion. Even when assembling a single volume in which he tried to iron out many inconsistencies, he warned that:

A complete consistency (either within the compass of The Silmarillion itself or between The Silmarillion and other published writings of my father's) is not to be looked for, and could only be achieved, if at all, at heavy and needless cost. Moreover, my father came to conceive The Silmarillion as a compilation, a compendious narrative, made long afterwards from sources of great diversity (poems, and annals, and oral tales) that had survived in agelong tradition; and this conception has indeed its parallel in the actual history of the book, for a great deal of earlier prose and poetry does underlie it, and it is to some extent a compendium in fact and not only in theory.

It is tempting to take this concept--what Dennis Wilson Wise termed the "compilation thesis" in his essay on The Silmarillion in volume 13 (2016) of Tolkien Studies--and attempt to reconstruct a "true" version of Middle-earth underlying all the various inconsistent texts. This can lead to a grab bag approach where people construct their own individual "Silmarillion" continuities, which can be enjoyable, but is counter to the notion of canon as an authoritative, standardized frame of reference. It can also easily lead people to non-Tolkienian conclusions. While Tolkien wanted to put together a single consistent version of "The Silmarillion", we can only speculate as to how he would have ironed out certain inconsistencies, what ideas he would have recontextualized (or discarded entirely), etc.

I understand people's frustrations at, for example, not having an answer to the question of orcish origins, but I think a lot can be lost in the drive for a single answer. In the context of early stories, orcs were simply created by Morgoth, because the idea that evil couldn't create from nothing hadn't been introduced yet. Obviously we can't ignore such an important development (both thematically and in worldbuilding terms), but when going back and rereading pre-Lord of the Rings material, it's worth remembering what orcs were at that stage of Tolkien's creative process. (Also, since Tolkien never found a satisfactory answer to the orc question, there isn't anything to project backwards in this case, even for those so inclined.)

Towards the end of his life, Tolkien played around with working a similar principle into the internal structure of the legendarium. As Elenhir mentions, he became preoccupied with matters of scientific plausibility and developed a bunch of radically new ideas, but he didn't toss everything he'd previously written into the trash. Much as the second edition of The Hobbit did not obliterate the first--Gollum giving Bilbo the Ring is recontextualized as a lie Bilbo told because of the Ring's corruptive influence, rather than Tolkien pretending he'd never written anything like it--Tolkien introduced the idea that the legends of the First Age were a mixture of Eldarin knowledge and preexisting Edainic myth. The scientifically implausible elements (eg, vast forests growing in a world illuminated only by starlight) can be regarded as relics of primitive human ignorance, but the stories themselves retain both their artistic merit and their status as tales told by characters within Arda, passed down over thousands of years and ultimately incorporated into the Red Book.

That said, this was a relatively late idea. A great many "Silmarillion" texts, including post-Lord of the Rings material up until the late 1950s / early 1960s, was written within the framing device of Pengolodh and Aelfwine (neither of whom are Númenóreans or Hobbits), and I think this must be borne in mind when reading them. On the other hand, Tolkien may have decided to change the framing device without substantially editing the texts themselves. For example, he described the Akallabêth as a Númenórean-authored work even though the last version of it still made reference to Pengolodh, and Charles Noad suggested (in "On the Construction of The Silmarillion") that Tolkien might have decided to alter "Dangweth Pengolodh" to be Elrond addressing Bilbo rather than Pengolodh addressing Aelfwine. But there are valid criticisms to be made that the originally-conceived in-universe authorship of those works is essential to understanding them, and changing the names in the framing device without making substantive edits to the texts themselves is not always a graceful solution.

[Boromir88]

Very similar to Eldy, I started out thinking there was one clear, no contradictions answer to every Tolkien related question ever. He wrote a ton, there has to be a clear answer to everything. That task became harder and harder, and as I got a hold of Tolkien's letters I started feeling more reflective and open in answering lore questions. I cared less about canon and more just trying to learn something.

This might be a bit off from the topic of canon, but still relevant. The one thing I will always argue tooth and nail over is those who insist on allegories. I remember some tangled babble from one writer who insisted that Tom Bombadil was Tolkien's allegory for some sort of Stage Manager in a play. I mean, that's fine to an extent if that's what the person reasoned out from their personal reading. But it got to some Dan Brown Illuminati level of "Tolkien left all these hidden clues to who Tom Bombadil is, and every one was too stupid to not see it, except this one person in the world who uncovered the mystery! And now you're all too dumb not to agree with her." It was weird and extremely contradictory to what I think a person passionate about learning Tolkien should do.

So, yes, I tend to still be a contrarian to anyone who insists in saying Tolkien wrote allegorically. To the point that even if I sort of lean toward agreeing with the person's viewpoint, I'll still argue against their insistence that it's the "truth." As far as people presenting an argument, whether using a more strict canon view, or sort of an "anything Tolkien wrote goes and is canon," well I'm always for trying to learn something new I had not considered before.

[KingODuckingham]

Thank you very much for these responses, this is fascinating. I had considered (but not mentioned in the opening post) just such a mythological approach, wherein the stories, by being passed down and retold, often orally, achieve something of a kitchen sink approach where variations or even contradictions are accepted as par for the course (f.e. Ovid's Metamorphoses vs earlier renditions), but I had not quite realized that Tolkien himself preferred to think of the Silmarillion in these terms. I also agree, Boromir88, that it is quite another thing to equate this approach with a negation of any authorial intent, as if Tolkien might be writing whatever we want him to write.

[Flame Fried Ent]

Wasn't that a large part of why Tolkien wrote so much about Middle-Earth, though? To create a legandarium and mythology for the British Isles? From what I've read in mythology, and what I remember from HoME, mythology can be kind of nebulous when trying to pin it down with modern ideas, even at the best of times, let alone when trying to figure out what's what in a fictional one. Given that, I always thought later Tolkien was probably closer to the "final word" than early Tolkien, if only because he had spent more time refining the ideas.

[Troelsfo]

@simon and @Eldy Dunami, thank you for stating things so clearly!

I will put it in even stronger language. The very concept of “canonicity” creates a fallacy. The idea of a “canonical” representation of the Silmarillion legendarium represents the desire to create something that never existed in Tolkien's mind.

For instance, Gandalf was never a “Maia” in neither The Hobbit or The Lord of the Rings. In the former he was a funny old man (human – big people) who could do magic, in the latter he is an unspecified emissary from the Valar, but specifically not a Maia, since those only came into existence after Tolkien had finished writing the book (albeit, admittedly, before he finished editing it for publication). A “canonical” representation will, however, always see him as a Maia in both works, thus actually misrepresenting Tolkien's conception of the character as he was writing the story.

@simon has elsewhere spoken of how writing The Lord of the Rings in many ways changed Tolkien's view of his underlying mythology. This, however, means that The Lord of the Rings marks a transition in his concept of the mythology, and this transition can be seen in a number of inconsistencies within the published text, as I wrote about in my 2011 blog post, ‘The Lord of the Rings’ as a transitionary work.

[Túrin]

For instance, Gandalf was never a “Maia” in neither The Hobbit or The Lord of the Rings. In the former he was a funny old man (human – big people) who could do magic, in the latter he is an unspecified emissary from the Valar, but specifically not a Maia, since those only came into existence after Tolkien had finished writing the book (albeit, admittedly, before he finished editing it for publication). A “canonical” representation will, however, always see him as a Maia in both works, thus actually misrepresenting Tolkien's conception of the character as he was writing the story.

Though before LOTR, there were other spirits. In HoMe V: The Lost Road and Other Writings, they are called the Vanimor, and are described exactly how we understand Maiar: Similar to the Valar, but with less might. This doesn't discount your point that Tolkien's perspective on many matters was evolving through time, but I'm not sure that by the time of LoTR, equivocating the lesser spirits with the Valar with the eventual concept of the Maiar is that grievous a sin.

While on that topic, is this point about Gandalf discussed anywhere in HoMe? I'd love to take a look and read further.

[Chrysophylax Dives]

Some thoughts on this that are not very clear in my mind. It seems that the two views on canonicity relate to two sides of JRRT's imagination, which might be identified in terms of different scientific approaches to language: the philologist and the linguist. i see the philologist as concerned with words in history, teasing out the meaning of an obscure word at a specific moment, attentive to how the same word had different earlier meanings. we find just this side of Tolkien in his commentary on Beowulf. the linguist, by contrast,thinks of language as a system, and i think (though this is to go beyond my realms of understanding) that this kind of systematic thinking is exemplified in Tolkien's invention of languages.

i think that after he had finished Lord of the Rings the systematic side of Tolkien's thought became dominant. it seems to me (but this is an impression and not something i feel sure of) that the later HOME volumes show Tolkien resolved on systematizing his sub-creation and that the notion of canonicity fits well with the way the older JRRT looked upon his fairy stories (myths and legends). A website like Tolkien Gateway, which sets out an index of Arda and (ideally) assigns one meaning to each item in the index (so, for example, last time I looked there was no entry for 'magic ring' as it is subsumed by 'One Ring') seems to continue just this systematic side of Tolkien's imagination.

to my mind this systematic approach obscures much that is wonderful in Tolkien's stories and, perhaps more importantly, while the various later theological and metaphysical concerns that he grappled with in trying to establish a coherent system are fascinating, the real creative genius of this astonishing man is to be found in his earlier writings. but i accept that most 'Tolkien fanatics' do not hold this view.

finally, i think that the essay 'On Fairy-stories' contains a sort of justification of the systematic or canon side of Tolkien's imagination, namely in the insistence that a secondary world must be made 'credible', which translates into an author needing to harmonize all the parts so that a fantastic invention does not seem out of place. while this argument of 'On Fairy-stories' is obviously correct in some way, and to those who embrace canonicity is perhaps the last word, i have the sense that the various hesitations about canonicity in the posts above point to a limitation of the argument.

[Eldy Dunami]

finally, i think that the essay 'On Fairy-stories' contains a sort of justification of the systematic or canon side of Tolkien's imagination, namely in the insistence that a secondary world must be made 'credible', which translates into an author needing to harmonize all the parts so that a fantastic invention does not seem out of place. while this argument of 'On Fairy-stories' is obviously correct in some way, and to those who embrace canonicity is perhaps the last word, i have the sense that the various hesitations about canonicity in the posts above point to a limitation of the argument.

On this note, I believe Tom Shippey (among many others) made the point that the multitude of overlapping and contradictory texts that make up the legendarium actually enhances its believability as a pseudo-mythology, since organic Primary World mythologies are never internally consistent. It is ironic, then, that (as you note) Tolkien spent so much time and energy in the later part of his life considering how to remove inconsistencies. In many fantasy stories, I might consider it a worldbuilding flaw to not have a definitive answer to such a foundational question as when the sun was created, but in the context of the legendarium, having to say "it depends on which version we're talking about" is part of what makes the Secondary World so engaging.

As I mentioned in my previous post, I think Tolkien recognized this to some degree, and was willing to include contradictory versions of particular stories within the same overall framework. To use Númenor as an example, Christopher Tolkien discusses his father's proposed synthesis of multiple different versions of the tale, quoting a note he wrote (probably) in the 1960s on an envelope containing the texts of The Drowning of Anadûnê:

Contains very old version (in Adunaic) which is good -- in so far as it is just as much different (in inclusion and omission and emphasis) as would be probable in the supposed case:
(a) Mannish tradition
(b) Elvish tradition
(c) Mixed Dúnedanic tradition

Christopher notes that he believes the three "traditions" mentioned here correspond to three different versions of the tale written by his father: The Fall of Númenor (written in the 1930s), The Drowning of Anadûnê (1940s), and the Akallabêth (the final version of which was probably written in 1958; cf. the relevant chapter of HoMe XII). Tolkien used the former two as a basis for writing the third, but he also proposes an in-universe explanation for the relationship between these texts, suggesting that the Akallabêth was written by Third Age Dúnedain drawing on two different, older sources. I consider this an elegant solution to the dilemma of wanting to have both the richness of many different, contradictory texts and a definable "true" version of the Secondary World underlying them all.

That said, there are limits to this. From our perspective as readers, we can't appreciate, say, the full story of Gondolin without consulting the version from The Book of Lost Tales, but as far as I'm aware Tolkien never considered rehabilitating the Lost Tales as texts that existed within the Secondary World of the post-LOTR conception of Arda. There must have been some such text (and/or oral tradition) within the Secondary World, because Third Age characters are aware of Gondolin, but it's impossible to say what that the full version of that text might have looked like.

[Chrysophylax Dives]

As I mentioned in my previous post, I think Tolkien recognized this to some degree, and was willing to include contradictory versions of particular stories within the same overall framework.

Yes. I think you can see this recognition even within a story that we have only one version of, namely The Lord of the Rings. In the Prologue the book we have in our hands is said to be derived from the Red Book, which in Appendix B is said to have been deposited by Sam with Rosie at her home in the colony of Undertowers in the Westmarch - hence, 'The Red Book of Westmarch.' But we also learn that the original Red Book has been lost, but several copies were made, one of the most complete was made in Gondor, where it received many annotations, and this copy was then brought to the library at Great Smials in the Shire, and so on. So even here we have a picture of the book that we are reading being not quite that which was written by Bilbo, Frodo, and Sam and the possibility (even likelihood) scribal errors generating variations in the different copies. Which I think is just part of Tolkien's sense of what you refer to above, namely that variations (rather than a single canon) being at the heart of any actual body of received myths and legends.

[Troelsfo]

Returning after a somewhat busy summer ...

i think that after he had finished Lord of the Rings the systematic side of Tolkien's thought became dominant. it seems to me (but this is an impression and not something i feel sure of) that the later HOME volumes show Tolkien resolved on systematizing his sub-creation and that the notion of canonicity fits well with the way the older JRRT looked upon his fairy stories (myths and legends).
[...]
to my mind this systematic approach obscures much that is wonderful in Tolkien's stories and, perhaps more importantly, while the various later theological and metaphysical concerns that he grappled with in trying to establish a coherent system are fascinating, the real creative genius of this astonishing man is to be found in his earlier writings. but i accept that most 'Tolkien fanatics' do not hold this view.

Yes. Or, perhaps rather, how he would have liked to look upon his fairy stories.

There are, I think, two movements going on at this point.

One is systematising movement that you speak about, @simon, which I see as being strongly related to the desire to publish. This is related to two periods in particular. About 1950-1, when Tolkien hoped to publish The Silmarillion along with The Lord of the Rings, he turned to the Silmarillion with the aim to put it in a publishable form. When that hope was crushed, he turned to The Lord of the Rings, making that book ready for independent publication. With the success of The Lord of the Rings, publication of The Silmarillion was finally within reach, and so we see Tolkien again turn to this task about 1958-60.

You also touch on the other movement in words reminding me of how it is captured by Christopher Tolkien in his foreword to The Silmarillion as it was eventually published:
Moreover the old legends (‘ old’ now not only in their derivation from the remote First Age, but also in terms of my father’s life) became the vehicle and depository of his profoundest reflections. In his later writing mythology and poetry sank down behind his theological and philosophical preoccupations: from which arose incompatibilities of tone.
Christopher Tolkien, “Foreword”, in Tolkien, J. R. R.. The Silmarillion.


And not only this. Tolkien also used his legendarium as a vessel for trying out his changing ideas and tastes. Just as for his languages, so, too, his legendarium was in many ways “a private enterprise undertaken to give pleasure to myself by giving expression to my personal linguistic ‘aesthetic’ or taste and its fluctuations.”
J.R.R. Tolkien, to Mr. Rang, August 1967 (no. 297), in Carpenter, Humphrey (ed.), The Letters of J.R.R. Tolkien

Looking at Tolkien's work on his Silmarillion mythology in the years after he first completed The Lord of the Rings, it is easy to see which of these movements won out. It has been suggested that Tolkien himself was a perennial niggler, incapable of completing anything except under great pressure (was that Sisam? @geordie, do you remember?) , and though I think that might be taking it a bit too far, I also think that there is some truth to this. The concept of the internally consistent and complete representation of Tolkien's legendarium that underlies the idea of a ‘canon’ might thus be something that Tolkien at some points would have liked to attain, but which he was most likely incapable of producing, and which certainly never existed in his conception.

[Androthelm]

Although the conversation is long and gone (and I don't even have @Troelsfo's justification of having been busy with anything this summer), I'd like to throw my voice on the top of people who started out as purist/Hobbitist fans, concerned with bringing everything into alignment and creating one solid and internally consistent world, but who have grown to be far more content with inconsistencies and shifting truths. The points brought up by everyone here are good, and better than I can probably muster, but as one small note -- I think the shifting history which @simon raises in regards to the Red Book also probably speaks to Tolkien's experience as a scholar, and a love of manuscript tradition as a tradition and beautiful thing in its own right, and not merely a stumbling block on the path to full knowledge of the past.