“A Fragment, Detached: The Hobbit and The Silmarillion”

John D. Rateliff - 2012
Theoretical ArticlesTheoretical Articles: A comprehensive knowledge of the works of J.R.R. Tolkien is needed to fully understand articles in this category, the subjects treated being studied in minute detail by their authors.
A paper for the International Congress on Medieval Studies of Kalamazoo (USA), May 2012.

Une traduction française de cet article a été publiée dans Tolkien 1892-2012, le numéro hors-série de L'Arc et le Heaume.

Tolkien 1892-2012

A french translation of this essay was published in Tolkien 1892-2012, a special issue of Tolkiendil's magazine, L'Arc et le Heaume.

In January 1938, just months after the initial publication of The Hobbit, Tolkien was asked a series of questions about his sources for the book. Specifically, he was queried on three separate points:

Tolkien replied no to the furry pygmies, and no to the sinister fairy-tale hobgoblins. Expanding upon the larger question of fairy-tale influence, he explained that his tale is “derived from (previously digested) epic, mythology, and fairy-story—not, however, Victorian in authorship, as a rule to which George Macdonald is the chief exception.” Beowulf he ranked “among my most valued sources,” although he added, rather disingenuously, “it was not consciously present to the mind in the process of writing, in which the episode of the theft arose naturally (and almost inevitably) from the circumstances. It is difficult to think of any other way of conducting the story at that point. I fancy the author of Beowulf would say much the same.”

So far as I know, no one has attempted to trace Huxley's lecture, probably because of Tolkien's disclaimer that despite his Bloemfontein birth this herring is red. And although Tolkien expressed curiosity as to the name and publication history of the vaguely referenced tale about hobbit bogeymen, this was never satisfied in his lifetime; a quarter-century after his death it was finally identified by Douglas Anderson as “The Hobyahs”, from Joseph Jacob's More English Fairy-Stories [1894]. Tolkien's debt to MacDonald is confirmed and comes into clearer focus in a comment Tolkien made about a year later, in one of the drafts for his classic essay “On Fairy-Stories”, when he wrote that “George MacDonald has depicted what will always be to me the classic goblin. By that standard I judge all goblins, old or new” (OFS drafts, Ms. B; Flieger-Anderson p. 250). As for Beowulf, whole dissertations have now been written on its influence on The Hobbit and The Lord of the Rings1) what the original readers of this letter in The Observer in February of 1938 could not have known was that Tolkien wd have considered Beowulf itself a fairy-story in the sense he uses that term in OFS: the adventures of men on the edges of the Perilous Realm. To borrow for a moment Tolkien's metaphor of The Soup of Story, MacDonald is a minor and Beowulf a major ingredient in the stock from which Tolkien drew when creating Mr. Baggins' story. There are of course different kinds of soup, and different kinds of stories: here 'epic, mythology, and fairy-story' tell us what sort of soup we're being served up: not Victorian mock-turtle but something harkening back to older, wilder elements—like, for example, Beowulf.

Having answered all three of the inquirer's original questions, Tolkien closed by volunteering the following remarkable confession:

My tale is not consciously based on any other book — save one, and that is unpublished: the 'Silmarillion', a history of the Elves, to which frequent allusion is made. I had not thought of the future researchers; and as there is only one manuscript there seems at the moment small chance of this reference proving useful.

That is, Tolkien explicitly states that The Hobbit is “consciously based on… the 'Silmarillion' …to which frequent allusion is made” [emphasis mine]. This unsolicited statement seems emphatic and straightforward. And yet just two months earlier, in a December 1937 letter to a reader, he had been at pains to distance The Hobbit from his created mythology:

I don't much approve of The Hobbit myself, preferring my own mythology (which is just touched on) with its consistent nomenclature — Elrond, Gondolin, and Esgaroth have escaped out of it — and organized history, to this rabble of Eddaic-named dwarves out of Völuspá, newfangled hobbits and gollums (invented in an idle hour) and Anglo-Saxon runes. My elves have a more gracious and cunning alphabet which appears on the pots of gold in one of the coloured illustrations of the American edition.2) …And Smaug is of course connected with smjuga since Icelandic was in a foolish moment substituted for the proper language of my tales.3) — JRRT to G. E. Selby, December 14th 19374)


Now, if Tolkien seems overly dismissive of The Hobbit here, it's well to remember that he's often disparaging of his own work—indeed, within the same letter he describes his illustrations for The Hobbit as “bad” (“with the possible exception of the dust jacket”) [ibid] and elsewhere even described his masterpiece, The Lord of the Rings itself, as merely “good in parts” [Letters p. 349]. Context is always important when considering Tolkien's remarks about his work: in this case, his letter to Selby was written during the week when Tolkien had learned that Allen & Unwin would not be publishing The Silmarillion and instead wanted a new story about hobbits — which indeed Tolkien started by the end of that week, writing that complaint to Selby on Tuesday (the 14th) yet reporting the completion of the first draft of the first chapter on the following Sunday (the 19th) [Letters p. 27]. While acknowledging that elements of the mythology found their way into The Hobbit, Tolkien here seems concerned to distance Mr. Baggins' tale from that mythology — what has since by general consensus come to be called the legendarium.

This distancing reached its apex a quarter-century later in a 1964 letter to Christopher Bretherton in which he bluntly asserts that…

…by the time The Hobbit appeared (1937) this 'matter of the Elder Days' [i.e., The Silmarillion] was in coherent form. The Hobbit was not intended to have anything to do with it … It had no necessary connexion with the 'mythology', but naturally became attracted towards this dominant construction in my mind,5) causing the tale to become larger and more heroic as it proceeded. Even so it could really stand quite apart, except for the references (unnecessary, though they give an impression of historical depth) to the Fall of Gondolin … the branches of the Elfkin … and the quarrel of King Thingol, Lúthien's father, with the Dwarves …
The passage … relating [Elrond] to the Half-elven of the mythology was a fortunate accident, due to the difficulty of constantly inventing good names for new characters. I gave him the name Elrond casually, but as this came from the mythology (Elros and Elrond the two sons of Eärendel) I made him half-elven. Only in The Lord was he identified with the son of Eärendel, and so the great-grandson of Lúthien and Beren, a great power and a Ringholder. — JRRT to Christopher Bretherton, July 16th 1964; Letters p. 346–347 (emphasis mine)

Since this letter has sometimes been taken as Tolkien's definitive statement on the subject, it's worthwhile taking a moment to unpack its implications.

…one of those people whose fathers came into the strange stories of the beginning of history and the wars of the Elves and goblins, and the brave men of the North. There were still some people in those days who had both elves and heroes of the North for ancestors, and Elrond the master of the house was one. (HoH.115 & 121–122)

— a clear reference to the heroes of the War against Morgoth; cf. HME IV.38. While not as explicit as in the 1926 Sketch of the Mythology and the 1930 Quenta, what's said of Elrond's background here in The Hobbit (circa 1930–31) is entirely in keeping with those earlier, more detailed accounts.8)

Similarly, the ”(originally) quite casual reference” to the Necromancer was, from the draft of the very first chapter of The Hobbit, already linked to Thû the Necromancer, the foe of Beren and Lúthien (cf. HoH.82¬–83); a deleted reference in the manuscript has the wizard telling the dwarf-leader that there's no need for them to seek for revenge against the Necromancer for his father's death because

his castle stands no more and he is flown to another darker place
— Beren and Tinúviel broke his power, but that is quite another story. (HoH.73; emphasis mine)

— a specific reference to the events in “The Lay of Leithian”, canto IX, lines 2820a–2822 (HME.III.254¬–255):

for Thû had flown
to Taur-na-Fuin, a new throne
and darker stronghold there to build

Not only is Thû explicitly called ”that necromancer” earlier in the Lay (line 2074; HME.III.228)—a term that seems to be restricted in the legendarium to Sauron (as Thû is renamed in the later stories) and his servants—but the forest to which he flees is elsewhere called Mirkwood; cf. the 1937 Quenta Silmarillion: “it was called by the Gnomes [Noldor] Taur-na-Fuin, which is Mirkwood” (HME.V.282)

That is, at the time Tolkien wrote the opening chapter of The Hobbit, he already knew who “The Necromancer” was, and in the manuscript draft explicitly identified him with a character in the pre-existing mythology, one who played a key role in the Great Tale closest to his heart. And even though he removed the explicit reference to Beren and Lúthien,9) there's nothing to indicate that he changed that identification. The natural assumption would be that the two Necromancers are one and the same, yet another link between the mythic background and the present tale.

And yet, in a 1955 letter to W. H. Auden, Tolkien was even more emphatic that ”The Hobbit was originally quite unconnected, though it inevitably got drawn in to the circumference of the greater construction; and in the event modified it” [Letters.215; JRRT to WHA, 7th June 1955]. As we have seen, this is deeply misleading. Not only are Tolkien's borrowings from the legendarium many and ab origine (that is, present from the original drafting), but they are by no means clustered towards the end of the book; indeed, closer scrutiny reveals that that most occur in the first half of the book. And the evidence of The Hobbit manuscript(s) show this was always the case; these allusions to the older myths were not added in the process of revision, as we might expect from Tolkien's claim that it was “drawn in”; instead the story borrows more heavily from the Silmarillion tales in its early stages and becomes progressively less indebted as it goes along.

Why, then, have so many uncritically adopted the view that The Hobbit and The Silmarillion had only the slightest of connections? The answer, I think, is a tendency to confuse views Tolkien held decades later, in retrospect, with how he saw things at the time he published The Hobbit.10) In the end, we must reject Tolkien's assertions in his letter to Bretherton, that there was “no necessary connexion” between The Hobbit and the legendarium but only some casual borrowing of a few names, for the same reason that Christopher Tolkien rejects Tolkien's assertion, later on in the same letter, that his Numenor story was “originally unrelated … [to] the main mythology.”11) In that case, the evidence of the Lost Road manuscript itself disproves the later claim. So too with The Hobbit.

In the end, perhaps Tolkien best encapsulated the relationship between The Hobbit and legendarium in a 1955 letter to Harvey Breit of the New York Times Book Review. The original letter is unfortunately lost, but fragments of it survive, including the following quote:

“My work did not 'evolve' into a serious work. It started like that. The so-called 'children's story' (The Hobbit) was a fragment, torn out of an already existing mythology.” — JRRT to Harvey Breit, before June 5th 1955; Letters p. 218 (emphasis mine)

As it happens, this is also precisely the view of the first person outside Tolkien's immediate family to read The Hobbit, who was also the only person known to have read The Silmarillion (or at any rate significant chunks of it, including The Lay of Leithian) before reading The Hobbit in manuscript: C. S. Lewis. In a 1958 letter Lewis used almost exactly the same words:

The Hobbit is merely a fragment of his myth, detached, and adapted for children, and losing much by the adaptation. The Lord of the R[ings] is the real stuff. — CSL to Th. Howard, Oct 14th 1958; Collected Letters of C. S. Lewis, vol. III.981 (emphasis mine)

Lewis repeated this judgment, again using almost the same words, in his Times file obituary of Tolkien, apparently written about the same time:

The Hobbit (1937) was in origin a fragment of this cycle ['the private mythology'] adapted for juvenile taste — Times obituary of JRRT, published 3 September 1973; rpt in J. R. R. Tolkien: Scholar & Storyteller, pp. 11–15 (emphasis mine)12)

If we consider the Selby letter merely a grouse without overmuch significance, and subject the claims of the Bretherton letter about The Hobbit to the same scrutiny as its similar claims about The Lost Road, which we know to be deeply misleading, then something like a coherent picture emerges. In retrospect, when trying to explain the rather complicated interrelationship between The Hobbit and The Lord of the Rings and The Silmarillion, Tolkien downplayed The Hobbit's links to the mythology. But elsewhere, and much earlier, when one would expect his memories of the book's creation to be fresher and more reliable, he was quite clear in his insistence that The Hobbit was “consciously based” on The Silmarillion, and in many times and places he asserted that Bilbo's world was the world of the mythology.13)

In the end, whether or not we consider The Hobbit part of the legendarium from its inception (as opposed to by later adoption) depends on how we define the legendarium. Certainly Mr. Baggins' adventures were not intended to be added as an extra chapter into the Quenta Silmarillion, so if the legendarium is defined only in its most restrictive sense (what Tolkien intended to include in the 1930 or 1937 or 1951 Silmarillion) then The Hobbit fails the test. But by that standard, The Lost Road and The Notion Club Papers could be excluded equally well, since neither takes place in the old world of the myth, what came to be called The First Age. Even The Lord of the Ring might find its status challenged by that heightened standard. And any challenge to The Hobbit's conceptional connection to the myths has to satisfactorily take into account Tolkien's assertions that he “consciously based” The Hobbit on The Silmarillion, and that the one was “a fragment, torn [from] an already existing mythology”.

If on the other hand we use 'Silmarillion' as a handy shorthand for the legendarium as a whole (as Tolkien himself sometimes used it), particularly those works set within the subcreated world that came to be called Middle-earth, then The Hobbit was obviously part of that personal mythology from the start. In any case, by the sheer fact of its publication The Hobbit gained autonomous status, and a place apart from the ever-shifting Silmarillion texts. And in time it became an authoritative text, so that the older works in the mythology had to be revised to take it into account rather than the other way around.

So in the end, the answer we get depends on the question we ask. If the question is, “Was The Hobbit part of the legendarium from the very beginning?”, then the answer is “Yes, obviously.” If the question is “Was The Hobbit intended to be part of 'the Silmarillion'?”, then the answer is “That depends on what you mean by Silmarillion”. I would say it instead marks the point at which Tolkien discovered that there was more to Middle-earth than telling and re-telling the Book of Lost Tales stories in various forms (poem, tale, lay, chronicle, essay): the point at which the mythology escaped beyond The Silmarillion in the narrow sense of that word and became The Legendarium.


On Tolkiendil

1) For example, Bonniejean Christensen's Beowulf and The Hobbit: Elegy into Fantasy in J. R. R. Tolkien's Creative Technique [Univ. of S. Calif, 1969], Roberta Albrecht Adams' Gollum and Grendel as Cain's Kinsmen [M.A. Thesis, Stetson Univ., 1978], and Wm. H. Green's The Hobbit and Other Fiction by J. R. R. Tolkien: Their Roots in Medieval Heroic Literature and Language [LSU, 1969].
2) I.e., “Conversation with Smaug”, which would appear in both the American edition (March 1938) and the second printing of the Allen & Unwin edition (January 1938); see Hammond-Anderson's Descriptive Bibliography, pages 15 & 17.
3) This final sentence is a marginal note, marked for insertion at this point.
4) An excerpt from this letter appears in Christopher Tolkien's Foreword to The Return of the Shadow (HME VI.7); a transcription of the full letter appears in the 1987 Marquette exhibition catalogue, p. [4]. There is one error of transcription in the latter: the passage printed as “I am afraid this is rather a scrawl, but I am feeling rather better” should in truth read ”… but I am feeling rather tottery”. Cf. his use of the same phrase to his 16th Dec. 1937 letter to Stanley Unwin, written just two days later (“I have been ill and am still rather tottery”; Letters.26).
5) Compare Tolkien's comments to Stanley Unwin regarding his legendarium's tendency to colonize his other works:
…the Silmarillion and all that has refused to be suppressed. It has bubbled up, infiltrated, and probably spoiled everything (that even remotely approached 'Faery') which I have tried to write since. It was kept out of Farmer Giles with an effort, but stopped the continuation. Its shadow was deep on the later parts of The Hobbit. It has captured The Lord of the Rings… — JRRT to Stanley Unwin, February 24th 1950; Letters p. 136
If there was influence, then, he argues that it was one-way borrowing, from the mythology to The Hobbit.
6) For examples of The Hobbit influencing The Silmarillion in turn, see my discussion of The Arkenstone and the Silmarils in HoH.607 & 721.
7) In a parallel case, it is often asserted that just as the inclusion in Roverandom of the episode in which Rover the sea-dog is taken to the edge of the world by Uin the great whale to see Elvenhome (Faery) does not make that tale part of the legendarium, so too the inclusion in The Hobbit of many elements from the Silmarillion tales does not make it part of the legendarium, except through The Lord of the Rings. This argument wholly misunderstands Tolkien's conception of Middle-earth, which is our world in a mythical past. Although now disenchanted, the setting of his tales and our modern world are one and the same: we live in Middle-earth in a later age.† This is the whole premise of The Lost Road and The Notion Club Papers, whose status as part of 'The History of Middle-earth' no one has ever challenged: although the world has changed, it is the same world at different times in its history (cf. “Kortirion Among the Trees”). And what is more, the old mythic world has left behind elements that are still with us (according to the Prologue of The Lord of the Rings, hobbits are still with us, if now shy and fugitive) and is still accessible under the right guidance (cf. Ælfwine of England) or conditions (cf. The Notion Club Papers).The presence of Uin (a distinct individual memorable to all who have read The Book of Lost Tales) in Roverandom, and the appearance in the latter tale of The Man in the Moon (who had flitted through several of the earliest Silmarillion poems), like Uin's guided tour of Rover and his friend through the same precincts as Eärendel voyage are thus wholly in keeping with the argument that Roverandom takes place in a latter-day Middle-earth which is also recognizably a fictionalized version of our own world.>† I argue this point at greater length in my essay ”'And All the Days of Her Life Are Forgotten': The Lord of the Rings as Mythic Prehistory” in the Blackwelder Conference volume.
8) As far back as The Sketch of the Mythology (1926), we are told that at the end of that War “When … the Elves return to the West, bound by his mortal half he [Elrond] elects to stay on earth. Through him the blood of Húrin (his great-uncle) and of the Elves is yet among Men, and is seen yet in valour and in beauty and in poetry” (HME.IV.38; see also the parallel passage in the 1930 Quenta, HME IV.158). As I observed in Mr. Baggins, Elrond is remarkable in that among all the characters in the legendarium his “name and genealogy remained unchanged through all the various texts that comprise the Silmarillion tradition. From the first he is the son of Eärendel and Elwing, saved by Maidros or Maglor when the Sons of Feanor destroyed the refugees of Gondolin and Doriath” (HoH.121).
9) While regrettable from the point of view of students of the mythos (those “future scholars” Tolkien lightly refers to in his letter to The Observer), the removal served two purposes: it resolved a chronological anomaly of having had Beren & Lúthien's exploits occur within the relatively recent past (the last century or so) while the subsequent Fall of Gondolin took place, we are told, “many ages ago” (HoH.115), yet at the same time it kept the Necromancer present as a sinister force on the edges of the adventure, casting a distant shadow on Mr. Baggins' tale.
10) Unfortunately, we have no information on his view of the book at the time of its writing, other than the manuscript itself.
11) See HME.V.8 for Christopher's rejection of his father's claim that The Lost Road was “originally unrelated” (“the conclusion seems to me inescapable that my father erred when he said this”), and HME.VI.6–7 for his acceptance of the same claim made of The Hobbit on that basis of similar passages in the same letter.
12) Lewis's authorship of this obituary has been challenged, but the evidence strongly supports the traditional ascription of it to CSL. For details, see the discussion of this point in my essay “Inside Literature: Tolkien's Explorations of Medieval Genres” (forthcoming).
13) Perhaps most memorably in one of his letters to Arthur Ransome, when he responded to a query about the book:
You tempt me grievously to a mythological essay; but I restrain myself …For the history of the hobbit must come before many who have not before them the exact history of the world into which Mr. Baggins strayed… — JRRT to Arthur Ransome, circa Dec. 15th or 16th 1937; cf. HoH.p874 (emphasis mine)
Paired with this, his Letter to Waldman confirms the identification:
The generally different tone and style of The Hobbit is due, in point of genesis, to it being taken by me as a matter from the great cycle susceptible of treatment as a 'fairy-story', for children … this is a study of simple ordinary man … against a high setting[.] — JRRT, Letter to Waldman, late 1951; Letters p. 159 (emphasis mine).
A comprehensive survey of all Tolkien's (often contradictory) statements on The Hobbit's relationship to the larger myth would go beyond the scope of his paper; those interested will find more references (often briefer or more allusive) in Letters pages 17, 21, 24, 26, 38, 145, 298, and elsewhere.