The greater part of the material cited or described in this book is found in the papers that remained behind; but the third section of the book (called ‘The Third Phase’) constituted a difficult problem, because in this case the manuscripts were divided. Most of the chapters in this ‘phase’ of composition went to Marquette in 1958, but substantial parts of several of them did not. These parts had become separated because my father had rejected them, while using the remainder as constituent elements in new versions. The interpretation of this part of the history would have been altogether impossible without very full co-operation from Marquette, and this I have abundantly received. Above all, Mr Taum Santoski has engaged with great skill and care in a complex operation in which we have exchanged over many months annotated copies of the texts; and it has been possible in this way to determine the textual history, and to reconstruct the original manuscripts which my father himself dismembered nearly half a century ago. I record with pleasure and deep appreciation the generous assistance that I have received from him, and also from Mr Charles B. Elston, the Archivist of the Memorial Library at Marquette, from Mr John D. Rateliff, and from Miss Tracy Muench.
This attempt to give an account of the first stages in the writing of The Lord of the Rings has been beset by other difficulties than the fact of the manuscripts being widely sundered; difficulties primarily in the interpretation of the sequence of writing, but also in the presentation of the results in a printed book.
Briefly, the writing proceeded in a series of ‘waves’ or (as I have called them in this book) ‘phases’. The first chapter was itself reconstituted three times before the hobbits ever left Hobbiton, but the story then went all the way to Rivendell before the impulse failed. My father then started again from the beginning (the ‘second phase’), and then again (the ‘third phase’); and as new narrative elements and new names and relations among the characters appeared they were written into previous drafts, at different times. Parts of a text were taken out and used elsewhere. Alternative versions were incorporated into the same manuscript, so that the story could be read in more than one way according to the directions given. To determine the sequence of these exceedingly complex movements with demonstrable correctness at all points is scarcely possible. One or two dates that my father wrote in are insufficient to give more than very limited assistance, and references to the progress of the work in his letters are unclear and hard to interpret. Differences of script can be very misleading. Thus the determination of the history of composition has to be based very largely on clues afforded by the evolution of names and motives in the narrative itself; but in this there is every possibility of going astray through mistaking the relative dates of additions and alterations. Exemplification of these problems will be found throughout the book. I do not suppose for one moment that I have succeeded in determining the history correctly at every point: indeed there remain several cases where the evidence appears to be contradictory and I can offer no solution. The nature of the manuscripts is such that they will probably always admit of differing interpretations. But the sequence of composition that I propose, after much experimentation with alternative theories, seems to me to fit the evidence very much the best.
The earliest plot-outlines and narrative drafts are often barely legible, and become more difficult as the work proceeded. Using any scrap of the wretched paper of the war years that came to hand – sometimes writing not merely on the backs of examination scripts but across the scripts themselves – my father would dash down elliptically his thoughts for the story to come, and his first formulations of narrative, at tearing speed. In the handwriting that he used for rapid drafts and sketches, not intended to endure long before he turned to them again and gave them a more workable form, letters are so loosely formed that a word which cannot be deduced or guessed at from the context or from later versions can prove perfectly opaque after long examination; and if, as he often did, he used a soft pencil much has now become blurred and faint. This must be borne in mind throughout: the earliest drafts were put urgently to paper just as the first words came to mind and before the thought dissolved, whereas the printed text (apart from a sprinkling of dots and queries in the face of illegibility) inevitably conveys an air of calm and ordered composition, the phrasing weighed and intended.
Turning to the way in which the material is presented in this book, the most intractable problem lies in the development of the story through successive drafts, always changing but always closely dependent on what preceded. In the rather extreme case of the opening chapter ‘A Long-expected Party’, there are in this book six main texts to be considered and a number of abandoned openings. A complete presentation of all the material for this one chapter would almost constitute a book in itself, not to speak of a mass of repetition or near-repetition. On the other hand, a succession of texts reduced to extracts and short citations (where the versions differ significantly from their predecessors) is not easy to follow, and if the development is traced at all closely this method also takes up much space. There is no really satisfactory solution to this. The editor must take responsibility for selecting and emphasizing those elements that he considers most interesting and most significant. In general I give the earliest narrative complete, or nearly complete, in each chapter, as the basis to which subsequent development can be referred. Different treatment of the manuscripts calls for different arrangement of the editorial element: where texts are given more or less in full much use is made of numbered notes (which may constitute an important part of the presentation of a complex text), but where they are not the chapter proceeds rather as a discussion with citations.
My father bestowed immense pains on the creation of The Lord of the Rings, and my intention has been that this record of his first years of work on it should reflect those pains. The first part of the story, before the Ring left Rivendell, took by far the most labour to achieve (hence the length of this book in relation to the whole story); and the doubts, indecisions, unpickings, restructurings, and false starts have been described. The result is necessarily extremely intricate; but whereas it would be possible to recount the history in a greatly reduced and abbreviated form, I am convinced that to omit difficult detail or to oversimplify problems and explanations would rob the study of its essential interest.
My object has been to give an account of the writing of The Lord of the Rings, to exhibit the subtle process of change that could transform the significance of events and the identity of persons while preserving those scenes and the words that were spoken from the earliest drafts. I therefore (for example) pursue in detail the history of the two hobbits who ultimately issued in Peregrin Took and Fredegar Bolger, but only after the most extraordinary permutations and coalescences of name, character, and rôle; on the other hand I refrain from all discussion that is not directly relevant to the evolution of the narrative.
In the nature of the book, I assume conversance with The Fellowship of the Ring, and comparison is made throughout of course with the published work. Page-references to The Fellowship of the Ring (abbreviated FR) are given to the three-volume hardback edition of The Lord of the Rings (LR) published by George Allen and Unwin (now Unwin Hyman) and Houghton Mifflin Company, this being the edition common to both England and America, but I think that it will be found in fact that almost all such references can be readily traced in any edition, since the precise point referred to in the final form of the story is nearly always evident from the context.
In the ‘first phase’ of writing, which took the story to Rivendell, most of the chapters were title-less, and subsequently there was much shifting in the division of the story into chapters, with variation in titles and numbers. I have thought it best therefore to avoid confusion by giving many of my chapters simple descriptive titles, such as ‘From Hobbiton to the Woody End’, indicating the content rather than relating them to the chapter-titles in The Fellowship of the Ring. As a title for the book it seemed suitable to take one of my father’s own suggested but abandoned titles for the first volume of The Lord of the Rings. In a letter to Rayner Unwin of 8 August 1953 (The Letters of J. R. R. Tolkien, no. 139) he proposed The Return of the Shadow.
No account is given in this book of