The History of the Hobbit Read online

  —‘The Man Who Understands Hobbits’,

  Charlotte and Denis Plimmer, early 1967;

  Daily Telegraph Magazine, 22nd March 1968, pages 31–32.

  Carpenter

  I am not sure but I think the Unexpected Party (the first chapter) was hastily written before 1935 but certainly after 1930 when I moved to 20 Northmoor Road.

  —undated; quoted in Humphrey Carpenter,

  Tolkien: A Biography, p. 177.

  It is clear from these accounts that Tolkien did not remember the exact date, but he did retain a strong visual image of the scene. Two specific facts emerge: it was summertime, and the place was his study at 20 Northmoor Road. From this we can determine that the event took place no earlier than the summer of 1930, since it was early that year when the Tolkien family moved into the house from their former residence next door at 22 Northmoor Road (Carpenter, p. 113; Christopher Tolkien, Foreword to the fiftieth anniversary edition of The Hobbit [1987], p. vi).3

  This dating was challenged by Michael Tolkien, the author’s second son (1920–1983), who stated in his unpublished memoirs that he clearly recalled his father standing with his back to the fire in his study at 22 Northmoor Road and saying that he was going to start telling his sons ‘a long story about a small being with furry feet, and asked us what he should be called – then, answering himself, said “I think we’ll call him a ‘Hobbit’.”’ (quoted in Christopher Tolkien’s Foreword, p. vi). Father John Tolkien, the eldest son (1917–2003), was equally definite that the story began before the move from 22 to number 20 Northmoor Road: ‘The first beginnings of the Hobbit were at 22 Northmoor Road; in my father’s study, the room to the left of the front door as one looks at the house. I remember clearly the wood block floor, mats etc . . . [T]here were no family readings for us all in 20 Northmoor Road, where we moved early in 1930. I was 12+ & I think could read for myself! The room with its many bookshelves was not conducive to that sort of thing. As far as I remember the readings were always in the study . . . The Hobbit started with a couple or so chapters, to which if we were lucky a couple or more would be added at the next Christmas . . . I went to boarding school in September 1931 and so although very close to the family, all sorts of stories may have been told which I cannot date.’4 Carpenter, writing in 1976, notes that Michael and John Tolkien ‘are not certain that what they were listening to at that time was necessarily a written story: they believe that it may well have been a number of impromptu tales which were later absorbed into The Hobbit proper’ (Carpenter, p. 177).

  In support of his claim for an earlier origin of the book, in his guest-of-honor speech to the Tolkien Society’s Annual Dinner in May 1977 Michael described the stories he and his brothers and sister had written in imitation of The Hobbit.5 Michael recounts that these stories were populated by characters like Philpot Buggins, Ollum the giant frog, blokes (hobbits), smellers (wolves), the dwarves Roary, Borey, Gorey, Biffer, Trasher, Gasher, Beater, Bomber, Lammer, Throw-in (the chief dwarf), and young Blow-in and Go-in; Albert Bolger the troll, joshers, snargs, and the wizards Kimpu, Mandegar, and Scandalf the Beanpiper. Michael Tolkien dated his own contributions to this family apocrypha to 1929, when he was nine years old (Michael Tolkien, May 1977 speech; see also Christopher Tolkien, Foreword, p. vi), and thus argued that The Hobbit must have been begun by that date.

  While it is quite likely that many elements incorporated into The Hobbit came from family lore predating the book (see for example my commentary following Chapter VII), and The Hobbit was undoubtedly influenced by the other stories Tolkien read his children in the ‘Winter Reads’ (which, despite Fr. John’s comment, continued to at least 19376 and probably beyond), Michael’s own account provides evidence that the stories he describes could not have preceded the actual writing of the book; too many of the names are parodies of forms that only emerged at a later stage, well into the composition of the manuscript. For example, Scandalf the wizard and Throw-in the head dwarf are clearly modelled on Gandalf and Thorin – but for the first two-thirds of the story the wizard was named Bladorthin and for more than half of it the chief dwarf is named Gandalf, not Thorin; these two characters seem not to have received their now-familiar names until around 1932. Furthermore, Tolkien himself is quite clear on the point that he made up the name ‘hobbit’ spontaneously at the moment of writing it down – that is, that the word itself emerged in a written text.

  The most specific proof may be found in a commentary Tolkien wrote on the text for the dust-jacket for The Hobbit and sent to his publisher accompanying a letter dated 31st August 1937, in which he remarked ‘My eldest boy was thirteen when he heard the serial. It did not appeal to the younger ones who had to grow up to it successively’ (cf. Letters p. 21). Since John Tolkien was born on 16th November 1917, the events Tolkien is recalling here could not have taken place before the end of 1930; furthermore, Tolkien notes that ‘the younger ones’ (Michael was born 22nd October 1920 and Christopher 21st November 1924 and were thus respectively about nine and five in the summer of 1930, while Priscilla was still an infant, having been born in 1929) showed little interest at the time. Michael’s account not only contains inconsistencies but directly contradicts both the evidence of the manuscript and the accounts set down by his father, both at the time of the book’s publication and many years later. Given these facts, we should feel fully justified in accepting the word of the author recorded closer to the event over the childhood memories of a member of the original audience set down some 45 to 50 years after the fact.

  If we grant a starting date of no earlier than the summer of 1930, is there any other evidence to help us narrow the field? In fact there is, in the form of letters and memoranda set down by C. S. Lewis, Stanley Unwin, Christopher Tolkien, and Tolkien himself. Early in 1933, Lewis wrote the following to his old friend Arthur Greeves:

  Since term began I have had a delightful time reading a children’s story which Tolkien has just written. I have told of him before: the one man absolutely fitted, if fate had allowed, to be a third in our friendship in the old days, for he also grew up on W. Morris and George Macdonald. Reading his fairy tale has been uncanny – it is so exactly like what we [i.e., Lewis and Greeves] wd. both have longed to write (or read) in 1916: so that one feels he is not making it up but merely describing the same world into which all three of us have the entry. Whether it is really good (I think it is until the end) is of course another question: still more, whether it will succeed with modern children.

  —letter of 4th February 1933 from C. S. Lewis to Arthur Greeves;

  They Stand Together: The Letters of C. S. Lewis to Arthur Greeves,

  ed. Walter Hooper [1979], p. 449.

  The ‘term’ Lewis refers to is the spring, or Hilary, semester at Oxford, which traditionally starts on or near St Hilary’s Day (13th January). Two points in Lewis’s letter that particularly stand out are that he refers to Tolkien’s story as having just been written, and that he criticizes the ending of the tale as not being as good as the rest of the story. From this we can conclude that Tolkien probably finished writing the Ms. over the 1932 Christmas break (that is, December 1932–January 1933) and, as was his habit, loaned it to his friend for criticism and critique right away. Furthermore, what Lewis read was a complete story, not a large fragment of one lacking the final chapters – not only would he have surely commented on being handed a tale that broke off at the most dramatic moment, but he specifically singles out that portion of the tale for criticism.

  This interpretation of events wins additional support from another contemporary document, the Father Christmas letters. Every year, Tolkien’s children received a personal letter from Father Christmas (the English Santa Claus) describing all the adventures Father Christmas and his companion, the North Polar Bear, had had since the last letter. Most of these adventures deal with various disasters which have prevented Father Christmas from sending all the presents the children had asked for (North Polar Bear’s falling down stairs on top of
packages, mixing up labels, and the like), but the letters for 1932 and 1933 represent a dramatic shift in tone. In them, the world of Father Christmas and his friends suddenly becomes very like that of The Hobbit with the introduction of goblins to the series, right down to details such as characters becoming lost in goblin-caves, being rescued by an ancient and magical bear, and finding themselves besieged by hordes of goblins – whom they defeat with a combination of Father Christmas’s magic, the combat prowess of a great bear, and the aid of their elven allies the Red Gnomes. What’s more, in the striking picture of Father Christmas, Cave Bear, and a leanish North Polar Bear exploring the goblin-caves that accompanied the 1932 letter (Plate VI [top left]), we can even see both Gollum and Smaug make a cameo appearance: Smaug appears on the wall of the first passageway to the right, while Gollum can be seen peeking around a corner of the same passage, near the picture of the mammoth (see Plate VI [detail]). At least four goblins lurk in the passages to the left, while the middle column depicts goblins on drasils, the Father Christmas Letters’ equivalent of the goblin wolf-riders encountered in the Battle of Five Armies.

  The presence of the Cave-Bear, Elves, and a magician7 at the battle with the goblins argues that the final chapters were in progress at the time this letter was written and not, as Carpenter suggests, only set down shortly before the submission of the book to Allen & Unwin. Carpenter believed that

  . . . shortly after he had described the death of the dragon, Tolkien abandoned the story.

  Or to be more accurate, he did not write any more of it down. For the benefit of his children he had narrated an impromptu conclusion to the story, but, as Christopher Tolkien expressed it, ‘the ending chapters were rather roughly done, and not typed out at all’. Indeed they were not even written in manuscript. The typescript of the nearly finished story . . . was occasionally shown to favoured friends, together with its accompanying maps (and perhaps already a few illustrations). But it did not often leave Tolkien’s study, where it sat, incomplete and now likely to remain so. The boys were growing up and no longer asked for ‘Winter Reads’, so there was no reason why The Hobbit should ever be finished.

  —Carpenter, pp. 179–80.

  Unfortunately, this will not do. Certainly there was a pause in the writing – in fact, several pauses; see ‘A Note on the Text’, below. But there is no evidence that the story was abandoned in an unfinished state, and a good deal of evidence that it was not. One is the notable fact that none of the people to whom the manuscript was lent before its publication8 made any comment on the story’s having been incomplete – remarkable in itself if we believe with Carpenter that the final quarter of the book was missing. Carpenter’s account confuses the issue further by stating that ‘there was a completed typescript in existence (lacking only the final chapters) in time for it to be shown to C. S. Lewis late in 1932’ (Carpenter, p. 177); in fact, as we have seen, Lewis not only read but specifically criticizes the ending. Furthermore, Lewis’s letter to Greeves makes it clear that Lewis was not reading Tolkien’s story over the Christmas break – in the paragraph preceding the one already cited, he tells his friend ‘In the way of reading [,] Lockhart [i.e., John G. Lockhart’s Life of Sir Walter Scott] kept me going through the whole vac. [vacation] and I am still only at Vol. 8’ (They Stand Together, p. 448); the next paragraph introduces the new topic of what he had been reading ‘Since term began’ – i.e., The Hobbit.

  More evidence appears in the letter thirteen-year-old Christopher Tolkien wrote to Father Christmas in December 1937, shortly after the book’s publication, where he says

  He [JRRT] wrote it ages ago, and read it to John, Michael, and me in our winter ‘reads’ after tea in the evening; but the ending chapters were rather roughly done, and not typed out at all; he finished it about a year ago . . .

  —quoted in Christopher Tolkien,

  Foreword, p. vii.

  While Carpenter evidently interpreted this to mean that the final chapters had not been written at all but existed only in a hasty outline (what I have dubbed Plot Notes B and C), I suggest that we take young Christopher’s remarks literally and that by ‘roughly done’ he meant that the conclusion of the book existed only in his father’s handwritten manuscript, not typescript; then ‘about a year ago’ (in fact, in the autumn of 1936) Tolkien had returned to the text and at last typed out the final section in order to submit it to the publisher.

  Two additional pieces of evidence from the period immediately following upon the book’s publication help us complete our chronology. In a memorandum made by Stanley Unwin after a meeting with Tolkien on Wednesday 27th October 1937 to discuss a possible follow-up to the success of The Hobbit,9 Unwin notes in passing that ‘He mentioned that THE HOBBIT took him two or three years to write because he works very slowly.’ This detail coincides perfectly with the dates from our other evidence – i.e., that the story was begun in the summer of 1930 and finished in early January 1933, a period of two and a half years from first inspiration to final chapter. Finally, in a letter Tolkien wrote to the English newspaper The Observer in response to a letter of inquiry which had appeared in the 16th January 1938 issue asking about the sources for his book, he concluded with the following tease:

  Finally, I present the future researcher with a little problem. The tale halted in the telling for about a year at two separate points: where are they? But probably that would have been discovered anyway.

  —J. R. R. Tolkien, letter to The Observer,

  printed Sunday, 20th February 1938; see Appendix II.

  If, as Tolkien told Unwin, the story took ‘two or three years’ to write but, as he noted to The Observer, that period was punctuated by two hiatuses of approximately a year each, then the actual writing of the book took place in several short, intense bursts – in fact, during the vacations between term-time – which I in this book refer to as the First Phase, Second Phase, and Third Phase. Such was, indeed, Tolkien’s regular habit of composition, as careful perusal of Letters and the History of Middle-earth volumes dealing with The Lord of the Rings manuscripts will reveal; see ‘A Note on the Text’ below for more information on the actual writing of the book.

  There still remains one unresolved crux: why did Tolkien tell Auden (in 1955) and the Plimmers (in 1967) that a gap of several years intervened between the writing of the first chapter (The First Phase) and the rest of the book, when his earlier testimony to Unwin and the letter to The Observer make it clear that in fact the hiatus could have lasted no more than a single year? The answer I think lies in Tolkien’s tendency to exaggerate the passage of time and date events before they actually occurred; as an event recedes into the distance, he will often assign an earlier and earlier date for it.

  A prime and unusually well documented example is the short tale ‘Leaf by Niggle’. In March of 1945, Tolkien had written to Stanley Unwin ‘. . . I woke up one morning (more than 2 years ago) with that odd thing virtually complete in my head. It took only a few hours to get down, and then copy out . . .’ (JRRT to Stanley Unwin, letter of circa 18th March 1945; Letters p. 113). The story was, therefore, written sometime in early 1943 or late 1942; Tolkien submitted it to the Dublin Review on 12 October 1944 (Letters p. 97; Hammond’s Descriptive Bibliography p. 348 notes that the editor had written to Tolkien soliciting submissions on 6th September), and it appeared in the January 1945 issue. Twelve years later, in his letter of 24th June 1957 to Caroline Whitman Everett (Letters p. 257), Tolkien tells much the same story:

  I have not published any other short story but Leaf by Niggle. They do not arise in my mind. Leaf by Niggle arose suddenly and almost complete. It was written down almost at a sitting, and very nearly in the form in which it now appears. Looking at it myself now from a distance I should say that, in addition to my tree-love (it was originally called The Tree), it arose from my own preoccupation with The Lord of the Rings, the knowledge that it would be finished in great detail or not at all, and the fear (near certainty) that it would be ‘not at all’.
The war had arisen to darken all horizons. But no such analyses are a complete explanation even of a short story.

  By 1962, however, Tolkien had began to shift the origin of the story to an earlier date; he told his aunt Jane Neave that the story ‘was written (I think) just before the War began, though I first read it aloud to my friends early in 1940’ (JRRT to Jane Neave, letter of 8th-9th September 1962; Letters p. 320). Thus, whereas the 1957 letter makes it clear that the war was already underway at the time the story was written, the 1962 letter moves it back to ‘just before’ the war. By the time Tolkien wrote the introduction to the 1964 collection Tree & Leaf in October 1963 (Hammond, Descriptive Bibliography pp. 183–4), he believed that ‘Leaf by Niggle’ and the essay ‘On Fairy-Stories’ had been ‘written in the same period (1938–9) when The Lord of the Rings was beginning to unroll itself . . .’ and that ‘The story was not published until 1947’ (Tree & Leaf, p. [5]), thus exaggerating the period between composition and publication from about two years to almost nine while pushing the date of actual composition back by some 4 to 5 years.10

  Like Michael Tolkien’s attempt to push the starting date of work on The Hobbit back into the 1920s, we must reject Tolkien’s later assertion of a gap of several years between the writing of the first line and resumption of work on the story – not just because it directly contradicts remarks he made much earlier, at the time of the book’s publication (when we might reasonably expect his recollection to be more accurate), but because it creates unresolvable paradoxes in the evidence. The simple fact is that if Tolkien began the story after the move to 20 Northmoor Road in 1930, then stopped for several years before proceeding further, and paused twice for a year or so during the actual composition (these pauses being attested by changes in paper in the manuscript itself), he could not possibly have loaned the completed tale to Lewis in January 1933 – yet we know he did. The external evidence of the date of the move and the weight of the contemporary documentary evidence (especially Lewis’s letter to Arthur Greeves and the 1932 Father Christmas letter) between them establish a consistent body of evidence which agrees with all the facts of Tolkien’s other recollections. Accordingly, we may state with some confidence that the story was indeed begun in the summer of 1930 and completed in January 1933.