The Trouble with 'Canon'

Tolkien, Myth, and Historiography

In a recent interview shared on Amazon Prime’s The Rings of Power social media channels, Corey Olsen (“The Tolkien Professor”) made a claim regarding Tolkien’s work which has caused an ineluctable wave of outrage in the Tolkien ‘fandom.’

“First thing to specify,” he states, “is that there's no such thing, really, as canon in Tolkien.” While the editors of the clip (which is a mere minute and forty seconds) include some soundbites in which Olsen contextualises his statement, this comment has provoked a mind-blowing amount of online criticism. Canongate has begun.

As a Tolkien scholar who often speaks about the question of ‘canon’ in Tolkien’s legendarium - and what it might teach us about our own approaches to primary-world history and historiography - I wanted to share a few words about this topic.

Criticisms of Olsen, which have indeed been abundant (as seen here, here, here, here, etc.), have largely centred around characterising him as a “shill” for Amazon - an allegation that has also been frequently levied against myself, as well as any other Tolkien scholars and content creators who have been even moderately supportive of the show. Armchair critics claim that comments like this are not only heretical, but also a tacit endorsement of any and all changes or lore augmentations introduced by the Rings of Power showrunners in their adaptation. But there is, in fact, a very good argument to be made for this perspective.

Olsen rightly points out that “Tolkien’s ideas were ever-evolving”, and this is perhaps the most basic argument to be made for an anti-’canonical’ perspective on Tolkien’s legendarium. But it actually goes much deeper, and unfortunately much of the online discourse surrounding this topic has completely ignored Tolkien’s own opinions on the differences between ‘canon’ and ‘history’.

It is, in general, far more useful to think of Tolkien’s legendarium 'historiographically' than 'canonically’. He positioned his mythic project as a pseudo-historical corpus, composed ‘in-universe’ by countless individuals during the early ages of our world. This is all part of Tolkien’s framing narrative(s), i.e. his ‘found manuscript conceit’, whereby works like The Lord of the Rings are positioned not as novels written from the omniscient perspective of a wholly reliable narrator, but as pseudo-historical ‘primary sources’ written by the characters themselves, or by lore-masters of Middle-earth recounting the historical (and legendary) events of their own past.

For instance, both The Hobbit and The Lord of the Rings are positioned - within Tolkien’s framing narrative - as texts which were chiefly prepared by the hobbit protagonists of the tales themselves. The former is based on Bilbo’s travel diary, titled (by him) There and Back Again, while the latter was prepared largely by Frodo Baggins, who called his work The Downfall of the Lord of the Rings and the Return of the King. Both are contained in the (fictitious) multi-volume Red Book of Westmarch, positioned as the main source text to which Tolkien had access. The Red Book had additional contributors as well, Samwise Gamgee being the most notable, while additional writings and research were contributed by Meriadoc Brandybuck and Peregrin Took, as well as Sam’s descendants (the ‘Fairbairns of Westmarch’). In sections of the tale which were not directly witnessed by the hobbits (like those centred around Aragorn, Legolas, and Gimli), information was collected by the hobbit chroniclers from the individuals themselves. Some further details, particularly in the Appendices, were later added by (mostly) Gondorian historians of the Fourth Age, including an excerpt of ‘The Tale of Aragorn and Arwen’ composed by Faramir’s son, Barahir, while Gimli is credited as the source for the materials on ‘Durin’s Folk’. Within this framing narrative, all of these materials were originally composed in Westron, the lingua franca of NW Middle-earth at this time, which was then ‘translated’ and edited by Tolkien himself, who (acting as a ‘translator/editor/historian’ within the framing conceit) equipped his edited translation with a prologue and additional appendices (namely Appendices E and F).

It is a dizzying scheme, and this truly only begins to scratch its surface. Tolkien provides us not only with details of the Red Book’s original composition, but also of its complicated transmission lineage - including various versions of the collection and the unique features associated with each redaction.

Why would Tolkien do this, you might ask? The reasons are complex. Tolkien himself noted that he was “historically minded” (Letter 183), and that he intended for his work to represent a body of pseudo-historical legends dealing with the distant past of our own world. He was, furthermore, staunchly opposed to the positioning of fantasy as ‘allegory’, writing in his Foreward to the second edition of The Lord of the Rings, “I cordially dislike allegory in all its manifestations […] I much prefer history, true or feigned, with its varied applicability to the thought and experience of readers. I think many confuse ‘applicability’ with ‘allegory’; but the one resides in the freedom of the reader, and the other in the purported domination of the author.”

What does this mean? Essentially, an allegorical tale can ever only be about one thing - namely the thing that the author really intends for it to represent. An allegorical tale is not about itself, but a veiled commentary on something else. Particularly in a fantasy context, allegory makes use of fantastical and magical tropes and motifs to shed light on ‘real’ primary-world events and dynamics, often imbued with polemic arguments which are particular to the author and imposed upon the reader. Tolkien felt this to be an inferior kind of literature, and an egregious misuse of fantasy. If magic is only used to make a point about non-magical things, then it completely robs a tale of its capacity for enchantment. But this was not its only problem. An allegorical tale can only be accurately applied to a single set of historical dynamics - those determined by the author. It is limited in its scope, very much by design. By contrast, Tolkien wanted his work to be perennially applicable in a wide range of circumstances, and in myriad cultural and historical contexts. He wanted readers to be able to approach his work from manifold perspectives and get something out of it. In order to achieve this, the stories needed to be solely about themselves, not a veiled commentary on familiar primary-world events or peoples.

To make this work, Tolkien decided to compose his works as if they were grounded in actual histories - not merely ‘history’ in the sense of ‘things that happened in the past’, but also in the sense of ‘documents written about the past’. They are framed as works of history, with all of the complexities that this introduces. It may be useful here to draw a distinction between ‘primary’ and ‘secondary’ sources, though even this introduces some complications which need to be fleshed out.

In the context of history (as a discipline of study), ‘primary sources’ are generally comprised of historical documents written during, or shortly after, the periods and events they discuss. These could include memoirs, chronicles, letters, and other historical accounts written contemporaneously with the affairs which they document. In some cases, ‘primary sources’ can also consist of historiographic works prepared by historians (like Herodotus or Thucydides), even though they properly only serve as ‘primary sources’ for understanding a historian’s own time period (including perceptions of the past amongst their own contemporaries), rather than the periods upon which they are commenting.

‘Secondary sources’, by contrast, are commentarial works usually prepared by scholars to make sense of the past (usually based on primary written sources or other forms of evidence). Academic essays and books written by professional historians are the most familiar examples of secondary sources, representing an analytical interpretation of the past based on the evaluation of data.

Importantly, neither ‘primary sources’ nor ‘secondary sources’ are infallible. People lie about and misremember details, even of their own lives, and the mere writing down of ‘history’ is a distinctly literary craft that draws upon many of the same artistic devices used by storytellers of all kinds (including writers of fiction). Whether it’s an account of the Second World War or the rise and fall of the Roman Empire, all written histories make use of certain tropes and literary devices to create a compelling ‘story’ out of the affairs of the past. As Hayden White notes in his groundbreaking book Metahistory (1973), histories can be emplotted as romance, tragedy, comedy, or satire; they may highlight particular points of view on an event (i.e. “history is written by the victors”); and artistic devices like poetry and metaphor can be explicitly employed in order to direct a reader’s sympathies and opinions on a matter according to the historian’s own predilections. The writing of history is never neutral, nor is it objective, and this is something that Tolkien understood very well.

In Tolkien’s case, the Middle-earth legendarium contains an array of both primary and secondary sources, which take many different forms. There are prose accounts, memoirs, cosmogenesis myths, annals, poetic lays, academic essays, and many other kinds of literature. Some are firsthand accounts of historical events, while others are second- and thirdhand accounts based on hearsay, oral traditions, or translations of even earlier written sources. Many of these stories (including the affairs of the First Age) were often transmitted orally for millennia (e.g. as epic poems) before being written down, translated, and redacted numerous times. In many cases, the versions of the stories encountered in the legendarium are intentionally positioned as received accounts that passed through countless hands before reaching their ‘final’ form. For instance, the account of the Kinslaying at Alqualondë provided in the published Silmarillion is evidently based, at least in part, on a much older Elvish lament known as the Noldolantë, ‘The Fall of the Noldor’, composed by Maglor “ere he was lost” (Silmarillion, p. 87); while a more complete account of the Darkening of Valinor was prepared by Elemmíre of the Vanyar, known as Aldudénië (‘Lament of the Two Trees’) (p. 76). Neither of these original works were available to the compilers of the Silmarillion materials, who are simply recounting a condensed version of the tales based on the source materials available to them.

There are also proper ‘secondary sources’ in the legendarium, namely the academic ‘essays’ prepared by Tolkien himself - here acting not as the inventor of the Middle-earth corpus but as its scholar, translator, and historian. His essays on ‘The Istari’ (in Unfinished Tales) and ‘The Shibboleth of Fëanor’ (in The Peoples of Middle-earth) are, for instance, not imagined to have been written by some lore master in the first four ages of Middle-earth, but by Tolkien himself. But in these cases he is placing himself into the framing narrative as a part of the transmission lineage of the corpus. He is commenting upon the primary sources and attempting to shed light on them based on pseudo-historical ‘evidence’, rather than inventing ‘facts’ out of thin air.

This is, of course, all merely a part of his framing narrative and "translation conceit" - but Tolkien took this dimension of his work profoundly seriously, which is amply reflected in the materials with which he left us. There are, quite intentionally, countless inconsistencies to be found between the various rivalling ‘historical’ accounts - simply look at the differences between The Book of Lost Tales, The Quinta Noldorinwa, the Annals, and the published Silmarillion. ‘Externally’, we can certainly regard such incongruences as a function of Tolkien's own evolving ideas as the inventor of the tales. But 'internally', they can also be explained as arising from the variant accounts of past events written from an array of discordant perspectives. In short, Tolkien provides us with multiple ‘traditions’ of stories/histories without definitively identifying one as definitively ‘canonical’.

The tales of the fall of Númenor are another great example of this. In addition to the materials included in the Appendices of The Lord of the Rings - which may be based on Gondorian materials and/or Bilbo’s research in Elrond’s library - we have three major accounts of the Númenórean cataclysm: The Fall of Númenor (written by Tolkien in the 1930s), The Drowning of Anadûne (1940s), and Akallabêth (1950s). While we can think of these as being three phases of Tolkien’s ‘external’ development of the story, Christopher Tolkien demonstrates in Sauron Defeated (p. 406) that they actually represent three internal traditions of Númenórean history: an “Elvish tradition”, a “Mannish tradition”, and a “Mixed Dúnedanic tradition”, respectively. Notably, Tolkien suggests that the last account, the Akallabêth, was actually composed by Elendil himself (Unfinished Tales, pp. 242, 246) - and numerous scholars have noted the presence of a certain degree of ‘propaganda’ in this work.

Dividing the various phrases of Tolkien’s stories into authoritative ‘canonical’ versions vs. mere ‘draft versions’ actually does a disservice to the grand vision of Tolkien’s creative project. The very presence of inconsistency and divergence across accounts actually affords the legendarium a deeper sense of historiographic ‘reality’. By modelling his corpus on primary-world histories and myths, he allowed for his legendarium to be engaged with dynamically as a repository of variable accounts of the deep past, thus supporting the cultivation of ‘secondary belief’. If we’re truly following his lead, then we must conclude that there is indeed no such thing as a definitive ‘canon’ because ‘canons’ simply do not exist in real-world history. The very writing of history (real or feigned) is just as much an artistic process as a ‘scientific’ one (using that term loosely), and often reveals just as much about the historians (again, real or feigned) who write them, and the times in which they live(d), as they do about the period and affairs being represented in their works.

Now, none of this is meant to suggest that adaptors of Tolkien’s work should have carte blanche to change anything they please. Even Tolkien did not take this approach with his own work. As he writes in Letter 180, “I have long ceased to invent […] I wait till I seem to know what really happened. Or till it writes itself.” He functionally regarded his process as being evidence-based, much like a reliable work of history, even though the kinds of evidence he used would not be the same as those accepted by academic journals (historians generally do not regard lucid dreams or revelatory experiences as reliable sources of evidence). Nevertheless, it was profoundly important to him that narrative ‘decisions’ were firmly rooted in logic and reason (albeit clearly enchanted).

We obviously cannot expect modern adaptors of his work to draw upon the same visionary fonts of inspiration that he had access to, but we can very much expect a similar commitment to the internal logic of his mythos. So it is certainly reasonable for critics to challenge some of the decisions made by the Rings of Power showrunners (and, for that matter, Peter Jackson) on these grounds. But reducing these conversations to a matter of ‘canon’ is quite disingenuous, and largely a function of our modern relationship with literature and media as ‘intellectual property’ and commodities.

It’s worth noting that mythology itself - which is the best descriptor of Tolkien’s art - has always been a highly mutable and plastic phenomenon. Myths grow and evolve over time, with characters and archetypes perennially reworked, repackaged, reinterpreted, and recontextualised across time, space, and cultures. There is no such thing as an authoritative ‘canon’ of Greek or Norse mythology any more than there is an authoritative ‘canon’ of the Revolutionary War. Is Snorri Sturluson’s Prose Edda the authoritative ‘canon’ of Norse mythology? Is Elias Lönnrot’s Kalevala the authoritative ‘canon’ of Finnish and Karelian mythology? Emphatically, no. Both are simply snapshots of mythologies in flux - unique to their own time and place and coloured in, as all works of literature are, with their authors’ own ideas and points of view.

Tolkien’s work is different, perhaps, only on an ‘external’ level, since it was the creative output of a single author rather than an entire culture (itself something truly extraordinary). But he intentionally crafted his legendarium as if it were a true body of mythic history formulated by countless hands and minds over many thousands of years. The inconsistencies and incongruences are themselves a part of that effort. The benefits of approaching Tolkien’s work in this way are manifold. Not only does it allow for greater dynamism and richness in our engagement with his work, it also guides us towards some poignant insights into the writing of history, and our perceptions of the past.

So, yes, I whole-heartedly agree with Corey Olsen’s statement, as (I believe) do all serious Tolkien scholars. Tolkien’s legendarium is not a ‘canon’, nor is it feasible to definitively divide the countless manuscripts with which he left us into ‘canonical’ and ‘non-canonical’ forms. Instead, it is a a pseudo-historical corpus comprised of fictitious ‘primary’ and ‘secondary sources’, written by unreliable narrators and loremasters whose accounts of events are inextricably coloured by their own perspectives, values, and pseudo-historical contexts. This is one of the key reasons why Tolkien’s work is so unique and compelling, and why it is capable of producing ‘secondary belief’ in its readers.