I have a pet theory. I would like to articulate it eventually through a scholarly monograph, but for the moment it guides my peregrinations through fantasy (and other forms of) fiction, and emerges in partial form in the things I say about it. It is roughly as follows: certain writers in the twentieth-century, predominantly of fantasy fiction, and most notably J.R.R. Tolkien, developed world-building (a discipline which every writer of prose fiction must attend to) into a literary form in its own right. Tolkien, and others including the linguist, role-playing game designer and sometimes novelist M.A.R. Barker, made the secondary world the primary focus of his creative efforts, his life’s work in fact. The two novels he set there were relatively incidental in his mind, although writing the second certainly spurred him to firm up much of the detail and history. The Lord of the Rings is not, to my mind, a good novel – yet it is one of my favourite, and one which I have re-read repeatedly. The affection I have for it, and I think its widespread popularity, are due not to its virtues as a novel, which are rudimentary, but to the breathtakingly coherent and poetic world-building which underpins it, and which continually ruptures the surface of its prose with intimations of a vast, deep history, and with vatic language of epic power. Few other writers have made world-building their primary focus, but some have understood some of Tolkien’s insights. The past is a well, and its deepest waters must be mythic, or their sun-warmed historical surface will be tepid. An individual, even a carefully particularised one (as Tolkien’s characters are not), is a product of their place, and a place is an archive of its past. And most importantly, cultural coherence is a function of linguistic coherence.
Frank Herbert knew all this, as the documentary evidence which he fabricates for his deep future history in Dune demonstrates. Ursula K. Le Guin understands it well, as can be seen in her remarkable Earthsea books for children, in which she both locates magic in the act of naming, and attends carefully to the specific phonological music of her proper nouns. G.R.R. Martin is also very clear on the relationship between a secondary world and the tenor of a narrative, and his A Song of Ice and Fire series presents a deep, well-thought history, and a coherent linguistic milieu from the outset. None of them appear to prioritise world-building over prose-writing, as Tolkien did for many decades, but having seen Martin’s doorstop Fire & Blood arrive in the library last year it was apparent that he has been enjoying that ‘secret vice’ (a term Tolkien used of his language construction, but equally apposite here). One also might speculate as to whether this explains why the fuck The Winds of Winter is taking so long to arrive, but given my interest in world-building I’m not about to begrudge Martin a decision to set aside his present-tense narrative in favour of a work of constructed history.
I’ve not read his much shorter The World of Ice and Fire, which I assumed would be a cash-in of the sort that attends any intellectual property of this general flavour, but having held Fire & Blood in my hands when it arrived as new library stock, and seen that it consists of nothing but chapter after chapter of densely written historical narrative, I thought it would be remiss of me not to find out what happens when this master of long-form adventure fiction turns his attention directly to world-building. Of course this is still story-telling, and much of world-building will always be story-telling, but it’s fairly clear that Martin is a lot more interested in stories than say maps, or languages. In fact, what this seven-hundred and thirty-six page monolith of dynastic struggle amounts to is a plot outline for another dozen series as large as the still incomplete A Song of Ice and Fire.
When Tolkien wrote history, none of which was published before his death, he adopted a mythic voice which was grounded in a deep appreciation of the music of English, and crucially, of its linguistic past. Fire & Blood is written consistently in a voice which is reminiscent of another of the several which Tolkien employs; it can be found throughout The Hobbit and LotR, but is most prominent in the opening chapters of the later book, which are set in the Shire, among the pastoral, allegorically English hobbits. It’s clear that the work is written in the voice of a fictional scholar within Martin’s world of Westeros, and the tone that it adopts, particularly in reported speech, is very similar to the affectionately patronising tone with which Tolkien reports the speech of ordinary hobbits, in their commentary on the events of his narrative. The effect here (it’s pretty irritating in Tolkien) is twee, and jars with the sometimes explicit content, which of course any medieval chronicler would have left out.
In fact there is a fundamental contradiction between Martin’s desire to situate the voice of his text within his world, and his equally obvious intention that we take this text as a definitive and objective account of his world’s history. What interests does his narrator serve? What does he leave out? What does he embellish? Why does he appear to be regarded as reliable, when we are asked to understand him as subject to the same social and political forces as a medieval chronicler, none of whom were ever remotely reliable? He even discusses the relative reliability of his own fictional sources! The convolutions of succession, dynastic struggle, and internecine warfare which Martin describes are an absolutely convincing reproduction of the complexities of medieval European history (give or take a few dragons), but those complexities have been reconstructed by historians from multiple unreliable and politically interested sources. Martin’s decision to put his historical speech in the mouth of a single narrator pitches it straight into the chasm which exists between the singular, indisputable, oral truth of Tolkien’s backstories, and the multiple, debatable, constructed documentary truth of written history.
The impossible subject position of the speech that results could be forgiven as an aspect of the fantastical whimsy which characterises this genre of writing, but Martin’s usual narratives, and indeed his world, are characterised by nothing so much as their grim plausibility (give or take a few dragons). This weakly situated narrative voice, compounded by the lack of a protagonist, decentres the storytelling – without the complex interplay of subjectivities that characterise Martin’s novels the text becomes flat, its functional value (whether that is to edify, to entertain, to inform or whatever) perpetually deferred, until the narrative concludes with the reader none the wiser as to whose story they have just read, or from what position they have received it.
It is undeniably interesting to read a text by a commercial fantasy author that is pure world building. Martin paces his storytelling as carefully as he does elsewhere, and puts in plenty of cliffhangers and ‘Chekhov’s guns’ to keep the reader engaged despite the rather broad scale of the narrative. However, for me, this kind of writing foregrounds literary style in a way that Martin is uninterested in engaging with – it invites the reader to encounter it as the writing, not of its author, but of their constructed world. When Tolkien writes the deep past of his Middle Earth he adopts a backward-looking style which no contemporary writer would regard as being of much literary interest, but it is a burnished and musical mode of textual production which reveals the hand of a master-craftsperson, and which is absolutely plausible as the speech of a denizen of his world. Martin is also a master, but of plotting, not of prose, and his version of Bygonese is a clumsy one. His book will be of great interest to fans of A Song of Ice and Fire and Game of Thrones, and I certainly enjoyed it, but it reads as though it was written to communicate his plot outline to a publisher or show-runner. It is not at all compelling as a metafictional text.
Oliver Arditi 