The entry for ‘hobbit’ in the Oxford English Dictionary has an unhelpful etymology:
hobbit, n. In the tales of J. R. R. Tolkien (1892–1973): one of an imaginary people, a small variety of the human race, that gave themselves this name (meaning ‘hole-dweller’) but were called by others halflings, since they were half the height of normal men.1
This alleged derivation of ‘hobbit’ from ‘hole-dweller’ can’t be the correct explanation for the origin of the word, as Tolkien created this derivation in the early 1950s as part of the ‘fiction of translation’ of The Lord of the Rings, in which Westron, the common speech of the western part of Middle Earth, is imagined to have been translated into modern English, and Rohirric into Old English. Appendix F of The Return of the King says:
Hobbit is an invention. In the Westron the word used, when this people was referred to at all, was banakil ‘halfing’. But at this date the folk of the Shire and of Bree used the word kuduk, which was not found elsewhere. Meriadoc, however, actually records that the King of Rohan used the word kûd-dûkun ‘hole-dweller’. Since, as has been noted, the Hobbits had once spoken a language closely related to that of the Rohirrim, it seems likely that kuduk was a worn-down form of kûd-dûkun. The latter I have translated, for reasons explained, by holbytla; and hobbit provides a word that might well be a worn-down form of holbytla, if that name had occurred in our own ancient language [that is, Old English, in which it would have meant ‘hole-builder’].
But none of this can have been on Tolkien’s mind when he began work on The Hobbit: at that point, even the idea of linking it to the world of The Silmarillion had not occurred to him, let alone the Second and Third Ages, the evolution of Rohirric and Westron from Numenorean, and the fiction of translation. For a more plausible account, we have to turn instead to Tolkien’s 1955 letter to W. H. Auden, in which he wrote:
All I remember about the start of The Hobbit is sitting correcting School Certificate papers in the everlasting weariness of that annual task forced on impecunious academics with children. On a blank leaf I scrawled: ‘In a hole in the ground there lived a hobbit.’ I did not and do not know why. I did nothing about it, for a long time, and for some years I got no further than the production of Thror’s Map. But it became The Hobbit in the early 1930s2
Tolkien’s biographer Humphrey Carpenter quoted another version of the story but (following his usual slipshod practice) without reference:
It was on a summer’s day, and he was sitting by the window in the study at Northmoor Road [in Oxford], laboriously marking School Certificate exam papers. Years later he recalled: ‘One of the candidates had mercifully left one of the pages with no writing on it (which is the best thing that can possibly happen to an examiner) and I wrote on it: “In a hole in the ground there lived a hobbit”. Names always generate a story in my mind. Eventually I thought I’d better find out what hobbits were like. But that’s only the beginning.’3
This must have been fairly soon after Tolkien took up his post as Professor of Anglo-Saxon at Pembroke College, Oxford, in 1925.
However, there is a complication to this tidy story. In The Denham Tracts: A Collection of Folklore by Michael Aislabie Denham (1895) we find:
[…] pigmies, chittifaces, nixies, Jinny-burnt-tails, dudmen, hell-hounds, dopple-gangers, boggleboes, bogies, redmen, portunes, grants, hobbits, hobgoblins, brown-men, cowies, dunnies, wirrikows, alholdes, mannikins, follets, korreds, lubberkins, cluricauns, kobolds, leprechauns, kors, mares, korreds, puckles, korigans, sylvans, succubuses, black-men, shadows, banshees, lian-hanshees, clabbernappers, Gabriel-hounds, mawkins, doubles, corpse lights or candles, scrats, mahounds, trows, gnomes, sprites, fates, fiends, sybils, nick-nevins, whitewomen, fairies, thrummy-caps […]4
How are we to reconcile Tolkien’s version of events with the appearance of the word in a list of supernatural beings in a 19th century work on folklore? There seem to be two possibilities: either Tolkien had in fact read the word in Denham (or some other source yet undiscovered) but had forgotten that he had done so; or the two occurrences of the word are independent inventions. The problem with the first of these possibilities is that Denham’s tracts are obscure. In The Ring of Words: Tolkien and the Oxford English Dictionary, Gilliver et al. describe his work thus:
Denham was an amateur folklorist who published many books and pamphlets, including twenty Minor Tracts on Folklore (1849–c. 1854). The majority of these Tracts were collected in an edition prepared for the Folklore Society in the 1890s, and the word hobbit appears in the second volume (1895) of this edition.5
In the absence of any evidence that Tolkien read Denham’s work, Gilliver et al. have to resort to speculation about how he might have done so:
The 1895 version would have been readily available in university libraries accessible to Tolkien (there is a copy in Oxford), and he was interested in folklore. Alternatively, he could have seen a list copied out by one of his friends or colleagues.6
But this is hard to believe. By 1925 the Bodleian Library had well over a million books, and there is no reason to suspect Tolkien of having found this particular needle in that vast haystack.
There is the tantalizing possibility of another source for the word. A letter to The Observer a few months after the publication of The Hobbit in 1937, asked about the origins of the name:
Sir,—Dr. Julian Huxley, in one of his recent lectures, referred to the “little furry men” seen in Africa by natives and, although dimly in moonlight, by at least one scientist.
What I should like to know is whether these creatures provided the inspiration for Professor Tolkien’s attractive hobbit, the newest visitor to so many of our nurseries this Christmas. Naturally, I always read my children’s book before giving them to them, and I noticed that all the characters in the hobbit were nearly all drawn from real animal life or from real mythology. Few of them appeared to be invented.
On mentioning the hairy-footed hobbit, rather like a rabbit, to one of my contemporaries, I was amazed to see her shudder. She said she remembered an old fairy tale called “The Hobbit” in a collection read about 1904. This creature, she said, was definitely frightening, unlike Professor Tolkien’s. Would the Professor be persuaded to tell us some more about the name and inception of the intriguing hero of his book? It would save so many research students so very much trouble in the generations to come. And, by the way, is the hobbit's stealing of the dragon's cup based on the cup-stealing episode in Beowulf? I hope so, since one of the book's charms appears to be its Spenserian harmonising of the brilliant threads of so many branches of epic, mythology, and Victorian fairy literature. Yours, etc. ‘Habit’7
Tolkien replied, denying any knowledge of the “old fairy tale”:
I was born in Africa, and have read several books on African exploration. I have, since about 1896, read even more books of fairy-tales of the genuine kind. Both the facts produced by the Habit would appear, therefore, to be significant.
But are they? I have no waking recollection of furry pigmies (in book or moonlight); nor of any Hobbit bogey in print by 1904. I suspect that the two hobbits are accidental homophones, and am content* that they are not (it would seem) synonyms.
* Not quite. I should like, if possible, to learn more about the fairy-tale collection, c. 1904.8
The “old fairy tale” has not turned up in 80 years of searching, and so it seems that ‘Habit’s friend was mistaken in her recollection, perhaps at the distance of 34 years having confused ‘The Hobbit’ with a similar-sounding title. An example of a tale that could fit the bill is ‘The Bar-gaist, or Boggart’ which was printed in John Roby’s Traditions of Lancashire (1829).
So I think it is more likely that Denham’s and Tolkien’s ‘hobbit’ are independent inventions. There is a word ‘hob’ (an abbreviation of ‘Robin’) from which they could both derive:
hob, n.1 2.a. = Robin Goodfellow or Puck; a hobgoblin, sprite, elf.9
This gave rise to the compounds ‘hobgoblin’ (an imp or sprite; a terrifying apparition), ‘hobhouchin’ (an owl), and ‘hob-thrush’ (a goblin), so it could also have assisted with the coining of ‘hobbit’.
↩ This definition was supplied to the Oxford English Dictionary by Tolkien himself10 and I can only imagine that it is a kind of lexicographical joke.
↩ J. R. R. Tolkien (7 June 1955). Letter to W. H. Auden. Number 163 in Humphrey Carpenter, ed. (1981). The Letters of J. R. R. Tolkien. London: George Allen & Unwin.
↩ Humphrey Carpenter (1977). Tolkien: A Biography, p. 172. Boston: Houghton Mifflin.
↩ James Hardy, ed. (1895). The Denham Tracts: A Collection of Folklore by Michael Aislabie Denham, volume II, p. 79. London: David Nutt.
↩ Peter Gilliver, Jeremy Marshall & Edmund Weiner (2006). The Ring of Words: Tolkien and the Oxford English Dictionary, p. 146. Oxford University Press.
↩ Gilliver et al., p. 148.
↩ ‘Habit’ (16 January 1938). Letter to The Observer.
↩ J. R. R. Tolkien (20 February 1938). Letter to The Observer. Number 25 in Carpenter (1981).
↩ Oxford English Dictionary.
↩ J. R. R. Tolkien (11 September 1970). Letter to R. W. Burchfield. Number 316 in Carpenter (1981).