Here Be Cartographers: Reading the Fantasy Map

Monday, 18 April 2011 — 11:14pm | J.R.R. Tolkien, Literary theory, Literature

It is hard to imagine a world without maps.

Now stop—and diagram that sentence. Break its syntax apart. You can parse it in at least two valid and meaningful ways:

  • It is hard | to imagine | a world without maps. The use of maps is so embedded in our daily lives, so essential to our normal functioning, that the idea of a pre-cartographic society is as alien as the thought of a pre-literate one. On top of this, our idea of what it means to be a mapped society is itself confined to our familiarized expectations of what maps are like. How did people get by without maps—or rather, without the sorts of maps we know and understand?

  • It is hard | to imagine a world | without maps. Maps govern the way we think about space, and that extends to imaginary or hypothetical spaces. Without a graphic representation on paper or in our heads, our plans for things not yet built—homes, roads, electric circuits—may be cloudy and ambiguous. They may lack precision in the same way we have trouble with describing things that are outside our linguistic abilities. This is a negative definition of maps as a form of language: to be without a map is to be without language, and it impedes us from communicating ideas in the mind—to others, yes, but also to ourselves.

In both of these senses, maps of fictional places are remarkably challenging texts.

One of my chief interests in fiction, along with art in general, is how it presents itself as evidence of the way people receive the existing cultural data around them before they process it and spit it back out. (In literary criticism you will encounter words like allusion and intertextuality, but I think of them as subtypes of a broader cognitive activity.) When an author plans out a story’s setting in place, or when a reader attempts to reconstruct it from the words alone, the maps they produce tell us not only how they imagine the depicted geography, but also how they imagine the idea of maps. Furthermore, the author/audience distinction isn’t always sharp: some privileged readers, such as the illustrators at a publishing house or manuscript historians like Christopher Tolkien, participate in the interpretive stage as well as the official construction of the space for everyone else.

So when we open up a novel to find a map, we can think of the map as an act of narration. But what kind of narration? Is it reliable narration or a deliberate misdirection? Is it omniscient knowledge, a complete (or strategically obscured) presentation of the world as the author knows it? Or is the map available to the characters in the text? If it is, then who drew up the map, and how did they have access to the information used to compose it? If it isn’t, then through what resources do the characters orient themselves in their own world? And finally, does anyone even bother to think about these questions before they sit down to place their woodlands and forts?

In the post that follows, I am going to informally sketch out a theory of fictional maps, which is to say that I will put up a lot of pretty pictures from novels and talk about why they are neat. There is likely some academic work on this somewhere—I would be astonished if there weren’t—but I’m not aware of any, and certainly nothing that has accounted for modern critical approaches to the history of cartography. Map history and the comparative study of commercial genre literature are niches within niches as it stands, and my aim is to entwine them together.


Let’s begin with something familiar.

Depending on how you look at it, this map is one of the following:

  • The map that J.R.R. Tolkien drew up for The Hobbit, which appears in the endleaf of the original 1937 edition as well as most (if not all) of the English editions still in print today.
  • A map drawn by the dwarvish king Thrór depicting the environs of Erebor, the Lonely Mountain. Elrond deciphers the runes in Chapter III (“A Short Rest”).
  • A reproduction of Thrór’s map, copied and translated by the hobbit Bilbo Baggins.
  • A reproduction of Bilbo’s copy of Thrór’s map, received and delivered by one J.R.R. Tolkien from There and Back Again, the first part of the discovered manuscript known as the Red Book of Westmarch.

The complexity of the document is that it serves as all of these things at once. As Tolkien’s map, which we recognize to be a fictitious construction along with the rest of the text, the map is a device to orient the reader in an imagined world. But if we dive inside the fiction, the map is also Tolkien’s way of reporting to his readers what Bilbo and Thorin were looking at—no different than if your copy of the book came bundled with a replica of Bilbo’s sword, Sting.

What’s more, in style and technique the map is fully believable as something put together by the dwarves (apart from the lettering in English, which we can think of as Bilbo’s translation if we wish to suspend disbelief). Notice that rather than being a high-fidelity communiqué of how Tolkien imagined Middle-Earth, the map is a minimalist sketch of the world according to the dwarves. The sparsely chosen landmarks appear in relative (not absolute) position, the illustrations are abstract, and the inscriptions allude to people and events that would have been known to Thrór. Scale doesn’t even enter into the equation. (Tolkien’s original draft, which I’ll say more about later on, was even sparser: aside from the runes and text, its only graphic elements were the Running River and a top-down outline of the Lonely Mountain.)

Not to be neglected, of course, is that the map also functions as a two-layered riddle. In The Hobbit, we learn that while the runes on the left (in red above) are directly visible—”five feet high the door and three may walk abreast,” they read—the runes in the centre only reveal themselves when Elrond holds the map up to the light of the moon. (This custom-made replica demonstrates the effect.) The moon-runes provide a further clue: “Stand by the grey stone when the thrush knocks, and the setting sun with the last light of Durin’s Day will shine upon the key-hole.”

To complicate things further, when the dwarves first lay out the map in Chapter I (“An Unexpected Party”), Tolkien makes an authorial interjection in the text:

“There is one point that you haven’t noticed,” said the wizard, “and that is the secret entrance. You see that rune on the West side, and the hand pointing to it from the other runes? That marks a hidden passage to the Lower Halls.” (Look at the map at the beginning of this book, and you will see there the runes in red.)

Not only that, but it will say this in the text whether your edition has the map printed in red and black or not! (Now that we’re in the age of paperback dominance, it’s unlikely that this is the case for you.)

So the multifaceted nature of the map isn’t limited to its plurality of authors, each at a different level of absorption into the fictional world; also in play is a plurality of potential readers. Tolkien’s real-world readers aren’t expected to go in knowing how to decipher the runes (though nowadays, you’d be surprised at how many of them do). But the further concealment of the moon-runes tells us that within the narrative, the mapmaker had a restricted audience in mind.

If the map is an act of narration, what kind of narration is it? We have a good system for answering this type of question with respect to prose. First, there is the distinction of person—first, second, or third—which is largely a question of using pronouns to position the narrator and reader in relation to the action. The concept of person doesn’t map neatly onto cartographic works, however, unless there are pronouns involved. A better apparatus for distinguishing between the possible authors of a fictional map is what literary scholars call levels of diegesis—a technical way of delineating whether something is outside the text, inside the text, or inside a text within the text.

At minimum we are always dealing with three layers of reality, though they are not always separate: the author, the narrator, and the characters. In non-fiction, for instance, we observe no distance between the author and the narrator, and we assume that the inhabit the same plane. In the most basic form of first-person narration, we assume that the narrator is among the reality of the characters, even if he or she is far removed from the action. If we think about maps in diegetic terms—if we ask whether the documents and their authors belong to the world in the book, or if they come from outside—we unlock two of the most powerful concepts for thinking about perspective: omniscience and reliability.


What’s the appeal of a fantasy map, anyway? I doubt too many would disagree when I assert that maps add to the sense of immersion. If they come directly from the author—keeping in mind the intermediaries of the publisher and illustrator, too—a map tells the reader that the creator of the fictional space has really thought this place out. Storytelling always happens in façades, but evidence of the author’s forethought fills out the setting’s illusion of depth.

If we see a map of an imaginary land, we feel like we know more about the place. Like any supplementary material—timelines, family trees—it satisfies our latent curiosities. Also present, I think, is an element of bowing to the world-builder’s authority: by looking at a map we don’t simply know more about the world—we know more about how the author imagined the world to be.

But here’s the trap: the attitude of wanting to know more about a world—and moreover, believing that a map can draw us closer to it—leads the audience to default to a certain passivity. To a certain extent this is the criticism that has always been made of illustration, cinema, or any kind of embellishment beyond mere words alone: that when something is imagined on our behalf, we are robbed of our duty to reconstruct the textual reality for ourselves.

It’s strange, for instance, that the Yoknapatawpha novels of William Faulkner are notorious for compelling readers to cobble the logic and action together from jumbled scraps of unreliable narration, but with the right edition in hand we get the geography delivered to us on a plate:

From the perspective of literary history, I very much admire Faulkner’s maps: as you can see from the University of Virginia’s collection of Faulkner manuscripts, his sketches were instrumental in developing the sense of overlaying his novels on top of each other like an eternal palimpsest. And I don’t have a problem with sitting back and letting a provided map do some of the work for me, myself. The real danger lies in the assumption that authors’ maps can do this for us at all.

Regardless of whether a map depicts a fictional or real-world space, we always have to ask ourselves whose knowledge the map represents. This notion is intuitive when we think of real-world maps, where we really never have access to an omniscient perspective that lies outside our own reality. All that we have to work with are the source materials and surveying techniques that are known to us, and the rest is speculation. Increasingly, scholars of cartography now recognize that maps embed our political dispositions and socio-economic practices like any other texts: even in the satellite-enabled age of Google Earth and GPS, we start out from inside the world and set specific agendas for what we wish to represent, and how. Only superficially are maps ever “objective” views of the land.

Fiction throws a serious wrench into the way we think about maps because of our familiarity with omniscient points of view. In narration, we typically think of omniscience in terms of the narrator’s godlike reach into the interior experience of the characters, but the notion also applies to whether the narrator has access to information that the people within the fictional world do not.

In the case of an extradiegetic map, a map explicitly outside the text that serves as a direct conduit from author to audience, we tend to assume that the information it communicates is a sort of objective truth. We don’t mind so much if the liberties taken with the style of illustration—the medium, the lettering, the colour—would not have been available to the people who inhabit the imagined setting; if anything, such extravagances add to the romance and mystique.

But with maps that purport to speak from within the tale—maps with here-be-dragons blanks that delimit what the people in the fiction perceive or care about—we run into the same paradox that authors who deal with far-flung places frequently encounter with language. The author feels a certain duty to communicate to the reader with maximal precision, but the higher the fidelity of the message, the more it draws on our present-day idioms and conventions, and the more it strains plausibility.

A truly plausible map is one that we could imagine being created with the techniques and materials available to the people within the world; one where we could see the world’s cartographers as having the inclination, purpose, and skill to create something that looks like the product in front of us. Thrór’s map in The Hobbit is a classic example. Jim Hawkins’ map in Treasure Island is another:

The depth soundings, the rendering of the coastline, the dated handwritten addenda, the note that this copy is a facsimile (thereby accounting for its tidy, finished look)—everything about this illustration plays the part of an authentic nautical chart.

Then again, Treasure Island has the benefit of being set in a recognizable culture and time period, and we have a historical point of reference by which to gauge its illusion of authenticity. This can’t be said of secondary-world fantasy, the stories that take place in lands that have little to no geographic relation to the planet Earth we know.

The fictional cartographer’s paradox of perspective is more apparent in a map like this:

This illustration comes from Star Wars spin-off novel Vector Prime. Published by Del Rey when they first acquired the license to the tie-in books, it was the first officially sanctioned map of the galaxy depicted in the merchandising empire that has grown out of the George Lucas films. It’s an attractive design: the simple spiral arms and the dashing lettering recall the retro-futuristic pulp-serial aesthetic that inspired the Star Wars series. (Far better, at any rate, than some of the soulless pseudo-3D charts that followed it.)

Yet you can see how its heart is in two places at once. It’s a two-dimensional printed work depicting a universe where there isn’t a shred of print anywhere to be found: Star Wars, you will remember, is famous for projecting any and all visual information into a hologram or wireframe schematic. There are no illusions here that this map originates from within the galaxy, as it clearly doesn’t. But the map also situates itself in a limited heroic point of view. It’s the world according to the protagonists, with its hyperspace trade routes and dotted-line expansion frontiers, its Unknown Regions and Wild Space—all of it condensed and flattened into the pancaked realities of the twentieth-century printed page.

In a way, the map’s abstraction of the galaxy reminds us that we are dealing with a comic-book reality, albeit a thoroughly developed one. It has the omniscient privilege of sitting outside the text, but instead of filling in the blanks it elides them with the flair of escape.


As it turns out, the pragmatics of publication have a noticeable effect on the kinds of maps we see in fictional works.

If we return to our first example, Thrór’s map of the Lonely Mountain, notice the orientation. The Iron Hills to the east are somewhere up top; the spiders of Mirkwood out west are down at the bottom. Now, as readers receiving The Hobbit as a freestanding text divorced from history, it’s easy to come up with all sorts of justifications for why the map is east-side-up. We could say that it’s a cultural peculiarity of the dwarves. We could even say that it befits the direction of Bilbo’s quest, upwards being the way forwards. But if we delve into how the book ran its course from manuscript to first edition, a different story emerges.

Tolkien’s original concept for Thrór’s map was a page in portrait orientation, with north on top and south on the bottom as we would typically expect. As we know from Tolkien’s correspondence of January 1937, the plan was to have the map inserted into Chapter I at the point where Gandalf first shows it to Bilbo and Thorin. In the original unpublished map, the moon-runes were inscribed in reverse on the back of the page so the reader could reveal them by holding the map to the light, just as Elrond does in Chapter III of the novel. (I unfortunately don’t have a picture of Tolkien’s original manuscript map at hand, but it’s in the Bodleian Library at Oxford (MS. Tolkien drawings 33); I believe it also appears in The Annotated Hobbit.)

The orientation of the map was rotated into its present form when Tolkien’s publisher, George Allen & Unwin, refused to produce it in the form that Tolkien intended, primarily for reasons of cost. Thus the map was redrawn as an endleaf in landscape orientation, with the dimensions of two pages side by side.

Where the plot thickens is when you look at The Hobbit in translation. Here is Mikhail Belomlinsky’s map from the Russian edition of 1976:

The illustrations in the Soviet Hobbit have a very distinctive look; Gollum, in particular, is fantastic. As for the map, Frank Jacobs of Strange Maps has already written an excellent analysis of its Slavic character, which I will not repeat here.

One thing to notice, however, is the map’s use of space—specifically, its complete spread over the page. Rather than adhering strictly to the canonical geography of Middle-Earth, which was by then widely known (outside of Russia, anyhow) thanks to The Lord of the Rings, Belomlinsky tucked something interesting in every nook and cranny. Look at how Mirkwood and the Misty Mountains snake along with the rivers. Spanning the whole adventure from Hobbiton to Smaug, west to east, bottom to top, one almost forgets to catch that the directions on the compass rose are wrong.

Observe: с, в, ю, зсеверо, восток, юго, запад—north, east, south, west. By preserving the orientation of Thrór’s map in the English edition but not adjusting the compass directions to match, we end up with an erroneous map where Hobbiton is to the south and the Lonely Mountain is to the north. Preposterous! Yet we arrived here at the first place because of a series of publishing decisions across multiple editions that determined if the map would be printed on one page or two.

Flipped cardinal directions are hardly unique to The Hobbit, mind you. For a very long time, maps of L. Frank Baum’s Oz had the Munchkins to the left and the Winkies to the right, despite how clear it was to every Oz reader that Dorothy landed on the Wicked Witch of the East and made her way westward to Winkie country:

If we think of fictional maps as a study in the medium constraining the message, where it gets interesting is when you consider the maps that exist as documents within the story. Publishing considerations have the power to shape the reality of the imaginary place. Thrór’s map in the narrative plane of The Hobbit is oriented east-side-up because Tolkien’s map had to be so.

Stepping inside the fiction, we may ponder if the material constraints on mapmaking play a part in the story. In the context of these imaginary worlds, is there any suggestion of how maps are produced, transported, and preserved? Does their content spring forth from an overriding purpose or utility? These are considerations that buttress the depth and plausibility of a map, much as how Tolkien’s meticulously crafted languages and folkloric songs carry the impression of a bottomless history—the sense that this world and its people weren’t born yesterday.

Indeed, one of the finest touches in the Peter Jackson films of The Lord of the Rings is the scene at the exact midpoint of The Two Towers: Faramír, having captured Frodo and Sam and returned to his refuge at the Window to the West, goes over a map with his lieutenants. It’s a minor but effective embellishment on the book that sums up the various troop movements on the board while providing a brief visual treat for the Tolkien enthusiasts—and rather than displaying the map as an animated overlay with voiceover narration, it’s presented as a thoroughly creased parchment over which Faramír runs his finger. The map isn’t any old god’s-eye-view: like a real-world map, it’s a strategic instrument on the field. As a study in condensing an immensely challenging work for the screen, this subtle directorial decision stands out as a masterstroke.


One crucial point to take away from what I’ve said thus far is this: the coherence of a map with respect to the fiction it represents is not necessarily the same as its realism or level of detail. As much as we may desire to imprint an author’s vision onto our own with the utmost fidelity, I do not believe this ought to be the aim of fantastic cartography. As anyone who has been following the brief history of computer animation would know, verisimilitude isn’t at all the same thing as immersive reality. It is not the case that maps draw us closer to a believable fictional space with greater topographical accuracy or ever more lifelike terrain.

Take, for example, the six-book series by William Horwood that begins with Duncton Wood. The Duncton books revolve around communities of moles that worship the ancient stone circles of Great Britain, and some of the moles are capable of inscribing things in writing with their claws. One of the characters, Mayweed, is a talented navigator who sketches a map that appears (in translation, of course) in the third volume, Duncton Found:

Now contrast the deliberate, playful simplicity of this map with the one below, which appears in the second Duncton trilogy:

The second map looks substantially more attractive at first, thanks to the naturalistic depiction of terrain features. Unlike the original sketch, it looks like the sort of work that only a professional illustrator could pull off, and so it has the air of being the more polished, artistic piece. But it is also indisputably drawn from a human’s overground point of view—odd for a series where the moles largely keep to their burrows and humans are barely present apart from the odd occasion where they zoom along in their “roaring owls” (automobiles). And we might be fine with taking it as an omniscient artistic rendering from the publishing house, but the claim in the caption—”Based on Mayweed’s map found in Seven Barrows”—so brazenly contradicts the perspective of the map that it only serves to throw us off.

So as readers, we may appreciate the second map of Moledom as the superior illustration purely in terms of artistic merit—but it’s not clear at all that this map is better suited for the books. To the contrary, it presents a greater impediment to our ability to suspend disbelief.

The gravitation towards realistic detail in maps is especially remarkable if you remember that maps are inherently abstractions. The whole point of a map—of any variety, not solely the geographical kind—is to pack the chaos of information into a selectively delimited and instrumentally efficient container. Short of the 1:1 scale maps of Lewis Carroll and Jorge Luis Borges, cartography is a form of compression.

In terms of utility, the value of going from abstractions back to high-fidelity detail lies in how distinguishing visible landmarks—coastlines, mountains, fortresses—is useful for navigation. Before the proliferation of contour lines and coloured heat maps as methods for representing elevation, solutions for shading peaks and valleys led to the maps we now look back to fondly as exquisite in their artistic finesse. Satellite photography, which offers the highest fidelity of realistic representation we can achieve, is an excellent general-purpose tool because for a clientèle as diffuse as Google Earth’s, one never knows which topographic peculiarity might be useful at any given time.

Contrast that with maps of imaginary places, which are like their real-world kin in that all of them are works of art, but differ in that not all of them have a targeted function beyond serving, somewhat vaguely, as guidance for the reader (and often the author, as architect and city planner). Beyond looking pretty, framing the action, and setting the scene, the stylistic decisions often seem to lack any functional rationale.

Have a gander at this:

This is Ellisa Mitchell’s much-admired painting of the lands of Robert Jordan’s Wheel of Time, which for some reason is held up by some as one of the gold standards of fantastic cartography. When the up-and-coming fantasy novelist Saladin Ahmed called on Internet artists to produce a map for his forthcoming book, one that would outstrip the “very serviceable, basic, black-and-white line map” that his publisher could provide and “move beyond utility”, Mitchell’s map was one of the exemplars he had in mind.

Now, it’s obvious why Mitchell’s map is well received. It doesn’t take Sherlock Holmes to deduce that inveterate fantasy readers like Saladin Ahmed admire its colour and natural gradients, things you would never find in a mass-market paperback for reasons of cost alone. And embellishment for purely aesthetic reasons is certainly nothing new: back when printed atlases were engraved, colourists would fill in rivers and trace political borders on individual copies by hand. But look at how the version above compares to the black-and-white paperback map:

At first glance it seems like Mitchell’s painting is simply a more expensive and lovingly crafted reimagining of its monochrome counterpart. What it actually does, however, is emphasize topography at the expense of features that are more narratively functional, like political borders and roads. By foregrounding the terrain, the linear elements of the map recede.

(Doubly fascinating is that for all her good intentions, Mitchell unwittingly captures the trappings of the Robert Jordan series—indulgent top to bottom, muddled to the point of being unreadable, and plastered with a faux-medieval ethos that screams inauthenticity. I know, I know; I’m being much too harsh. Comparing the map to the books is incredibly unfair to the map.)

So here’s the question we should ask: why is it so attractive to conclude that the painted poster is the superior piece of work?

In part, this has to do with the realities of the mass-market fiction industry. Considering the practical limitations in the age of desktop publishing and (in many cases) direct-to-paperback, it’s uncommon that books will come with endpaper maps printed in multiple colours like the first edition of the The Hobbit. You’ll certainly never see historically accurate engravings in the mode of Treasure Island. In this environment, it isn’t altogether surprising that readers would treasure maps in other media for their relative rarity—even though in truth, maps for the flagship fantasy brands (particularly those adapted to film) are issued and sold en masse.

What this explanation doesn’t address, though, is why topography is the element that takes precedence to everything else in receiving what you might call the artistic treatment. In all probability, this is grounded in the desire for immersion; read any encomium for fantasy mapmaking and the first thing you’ll hear is that maps make the world more real. But realism is really just a code for saying that something is more in line with our embedded assumptions about what it means to perceive the world, and in the cultural value system we live in today—empirical, literal, photographic—it refers to the imagined experience of seeing the physical terrain with your own eyes.


The most striking thing about fantasy maps as a whole, especially the sort that dominates the industry of doorstopper mythopoeia that claims to be descended from J.R.R. Tolkien, is how rigidly they stick to convention. No matter how nice they look, structurally they reduce to orthogonal landmass drawings. Rarely, if ever, will you find visualizations of an imagined cosmology like the commonplace depictions of Midgard and Asgard wrapped around Yggdrasil in Norse mythology—and you would think high fantasy in secondary worlds is the genre that could use them most.

This is a problem that plagues much of the fantasy genre in its modern form, and why it has yet to fully escape the unfortunate stigma of juvenilia. The more you learn about history, and the more you are able to see through the anachronistic façades, the less imaginative conventional fantasy seems. We see this in action with language all the time: it’s almost expected these days to see gratuitous apostrophes as a desperate grab for an illusion of foreignness, not to mention unquestioning adherence to the “glottals ugly, labials pretty” phonetic valuation that Tolkien laboured to design. The same applies to maps.

(Ironically, children’s books are rather good at averting the implausibility trap, thanks to their embrace of figurative thinking. Witness the chart from The Phantom Tollbooth, featured at the top of this essay: like the novel it accompanies, it takes the abstractions of interior experience and projects them into real space.)

Indeed, as fantasy has entrenched itself in a self-propagating commercial set of norms, it has developed a reputation for being extremely conservative in form and politics. Jonathan McCalmont explained it well in his breakdown of the norms and values of “fat fantasy”: accessibility, immersion, and (less obviously) the safe escape of a reactionary aesthetics. All of this put together, along with the commercial considerations of how the publishing industry works, accounts for why a genre that benefits so much from cartography yields maps that are so extraordinarily literal.

For the commercial market, officially published maps are passive documents that serve as the easily readable evidence that the author thought things through. The maps embed our familiar expectations and unquestioned ideologies because the novels they accompany do the same. Perfectly content to assume that a map is the world, these fictions ignore the map as an instrument for grasping the world. Even if we step back from lofty cosmology and into functional lay-of-the-land geography, it is shocking how rarely fantasy maps explore the notion of the map as a visualization of a subjective and contingent worldview, a picture of an imaginary people’s collective weltanschauung.

Just to pluck one example out of thin air, in all of modern fantasy fiction, you will perhaps never find a map as fantastic as this:

The map contained in this 18th-century print by the Dutch publisher Pieter van der Aa is Giovanni Domenico Cassini’s Planisphere terrestre of 1696, which was meticulously assembled on the floor of the Paris Observatory from the most ambitious global survey of the time. (Working for Louis XIV assuredly had its benefits where funding was concerned.) The azimuthal projection centred around the North Pole looks warped to us today, accustomed as we are to the distortions we get from Mercator. But get past the initial unfamiliarity, and we can unpack all sorts of information from the structure of the map. We can see how it reveals its own construction in standard longitudinal slices, made possible by advances in astronomy and observational logistics; how the South Pole encircles the rim as the outer limit of human exploration.

Given the quiet proliferation of challenging fantastic fiction that is conscious of literary nuances in prose, one would hope that the cartography eventually catches up. Maps could do a lot more to dive into the perspective of an imagined land’s inhabitants, revealing how the people see their own world as well as their techniques and motivations for piecing that picture together. There is much room for subtlety in made-up maps, just as there is in lexicons, timelines, and family trees, and authors who do not feel bound to mainstream fat-fantasy conventions are in a unique position to explore the possibilities.

That said, it’s understandable why this hasn’t happened yet, and it has much to do with the specialization of labour along the pipeline of publication. With the rise of the e-book, prose content has become increasingly divorced from the totality of the book as an object, and authors are prose-module specialists more than ever.

One of the stories about Tolkien that has gotten a bit lost these days, I think, is the considerable control he exercised over his own work as a book designer who drew up his own cover art, runes and all. This isn’t to say that he had the visual talents of, say, William Blake, who remains the supreme English example of the all-in-wonder author/artist, but Tolkien’s relationship to his publisher is continuous with that tradition. This is the precondition that allowed a quirky thing like Thrór’s map of the Lonely Mountain to make it into the public sphere.

Outside of the vanity presses, this level of control is practically non-existent these days. Probably the best solution towards opening up the variety of maps we see is to develop the kind of author/artist collaboration for cartography that you see for graphic novels, where the writing doesn’t exercise any peculiar authority that confines the illustration to a secondary, supplemental role.


The question of authorship is a fascinating issue all to itself. Where exactly does a map (or indeed, an illustration of any sort) cross over from narration to interpretation? Is this solely a question of who is officially licensed to participate in the world-creation of an intellectual property? We can tear down the intentional fallacy all day from a theoretical point of view, and insist (quite reasonably) that the author has no better say than anyone else in the interpretation of the text; but in practical terms, reader communities who crave immersion derive their sense of a “canon” from officially sanctioned materials.

Consider this quandary: how would we mediate a discrepancy between Christopher Tolkien (who has privileged access to his father’s manuscripts and the legal authority of the Tolkien estate, and who delivered most of what we “know” about Middle-Earth in posthumous publications) and Barbara Strachey’s essential interpretive atlas, Journeys of Frodo? Do we rule in favour of the archival drafts, or do we side with the charts that are directly inferred from the final published text of The Lord of the Rings?

To return to what I said about omniscience early on, perhaps the demarcating factor that defines fictional cartography—what sets it apart from pictorial representations of real spaces—is the presence, however provisional, of the author as God. When we read non-fictional maps, we can perform a kind of analysis similar to what I’ve walked through here, looking at means of production, political agendas, and underlying worldviews; but it’s fairly unambiguous which plane of existence the cartographers reside upon.

Fictional maps introduce the complication of having, at minimum, two layers of authorship: the layer outside the text that has the power to dictate and reshape the world, and the layer that belongs to the reality of the world. It’s clear that the author is in the first and the characters are in the second, and that having the first speak for the second passes for a kind of ventriloquism or free indirect discourse. But these are not the only stakeholders in play. The “narrator” of the map, if it’s discernible as a separate voice, can belong to either layer or both. And once we introduce the other living participants—the readership and the publishing apparatus—determining who influences our perception of the fictional space becomes considerably trickier.

For one thing, it isn’t safe to take it for granted that immersion in a world means the same thing as immersion in the author’s mind, as if the goal of literature were some sort of telepathic telos of lossless communication. Among other problems, this attitude towards literary immersion as a matter of filling in the blanks has no way of dealing with deliberate ambiguity.

To see what I’m getting at, take a look at this:

This is the fold-out from The Discworld Mapp, a companion book to the Terry Pratchett series designed to the instructions of Pratchett and Stephen Briggs. It’s a fascinating map in view of what I’ve said thus far, not because of its superficial resemblance to the Cassini planisphere if you zoom out far enough, but because it achieves the look of an azimuthal projection via being a fiercely literal map in a two-dimensional rectangular coordinate system. Much of this, of course, is due to how the Discworld was conceived to be exactly that—a world where an azimuthal projection is a one-to-one circle-to-circle mapping, and therefore no transformation at all.

As far as companion publications go, The Discworld Mapp sounds par for the course—that is, until you look into how the map came into being. Stephen Briggs was a playwright who made a name for himself as an amateur Pratchett loremaster who adapted several books for the stage (and who, unlike Steve Vander Ark, knew better than to publish derivative works without legal sanction). When Briggs first decided to map the Discworld novels, beginning with The Streets of Ankh-Morpork, he did so with Pratchett’s support; but the project went against the author’s original insistence that Discworld was a fluid, aleatoric wander of the imagination that could not and should not be mapped.

Terry Pratchett’s opinion on cartography still appears on his official website:

There are no maps. You can’t map a sense of humor. Anyway, what is a fantasy map but a space beyond which There Be Dragons? On the Discworld we know There Be Dragons Everywhere. They might not all have scales and forked tongues, but they Be Here all right, grinning and jostling and trying to sell you souvenirs.

This is a very sensible position, and as much as I love cartography I am somewhat inclined to agree. There is something completely bounded about maps. We all know that the data they present is highly selective, as is also true for prose, with strategic omissions and gaps left for future exploration. Yet the boundedness of maps commits to an illusion of having reached a roughly finished state, as if everything inside has already been fixed and everything outside is still untouched and malleable.

From the perspective of someone involved in creating a world, particularly in a series that continues to emerge over time, a map intended to serve as an aid may also be a suffocating constraint. Mapmaking does not seem to permit carefully targeted ambiguity with the same flexibility as prose alone. With other forms of book illustration, one always gets the sense that the visual depictions could always be replaced or re-envisioned some other way. Maps exert a stronger form of authority: any improvements or revisions by readers or in future editions take place within the author’s borders as if they were immutable, objective truths.

It is a strange twist indeed that we are less liable to accept in fiction than in reality that cartography is a form of language: a medium for our perception of place, not to be confused with place itself. If there is a remedy for this, it may resemble the solution we developed for language, and take the form of self-conscious experimentation with maps as narrative voices—subjective, perspectival, and often unreliable. Literary writing deserves a literary map.

As for the alternative, let’s defer to what Lewis Carroll said in The Hunting of the Snark:

He had bought a large map representing the sea,
Without the least vestige of land:
And the crew were much pleased when they found it to be
A map they could all understand.

“What’s the good of Mercator’s North Poles and Equators,
Tropics, Zones, and Meridian Lines?”
So the Bellman would cry: and the crew would reply
“They are merely conventional signs!

“Other maps are such shapes, with their islands and capes!
But we’ve got our brave Bellman to thank”
(So the crew would protest) “that he’s bought us the best—
A perfect and absolute blank!”

Further reading

I am certainly not the first to gush over fantasy maps in the blog format. The most thorough look at fantastic cartography that I found while researching my post is “Beyond the Aryth Ocean”, a four-part series (1, 2, 3, 4) by Jason Denzel, who manages the biggest Robert Jordan community online and who therefore holds opinions that differ substantially from my own. He knows a lot of genre fiction I don’t, however, and his series on cartography is definitely worth a look.

Brian Sibley, who co-authored several map books with the eminent Tolkien painter John Howe, wrote a post about fictional maps with an inside perspective on the community of artists who work on these kinds of projects.

Of the dedicated cartography blogs on the Internet that cover fictional maps from time to time, two in particular stand out. The first is Frank Jacobs’ Strange Maps, which you probably know about already if you’ve followed some of the links in this essay or are interested enough in cartography to have made it this far down at all. (I also recommend the book.) The second is Jonathan Crowe’s The Map Room, which devotes equal attention to historical geography and modern surveying, and features an extensive subsection dedicated to imaginary places.

The Tumblr site entitled Fuck Yeah Fictional Maps really speaks for itself.

The Map History site maintained by Tony Campbell is an extremely comprehensive resource, mainly directed at serious scholars but certainly useful for enthusiasts of early maps.

For those of you with academic journal access through JSTOR, Imago Mundi is the history of cartography’s inexhaustible resource par excellence. The journal dates back to 1935, and skimming the tables of contents alone will give you a picture of how cartographic scholarship has developed over the past century.

Speaking of academic map history, one name to look out for when it comes to the literary turn in cartographic studies is J.B. Harley, best known for his collaboration with David Woodward on a compendious multivolume history of cartography that remains to be finished. (Harley and Woodward are both now deceased.)

Late in his career, Brian Harley became the radical subversive postcolonial deconstructionist of map historians, but he isn’t nearly as scary or loopy as that iconoclastic description makes him sound. In fact, the sort of critical theory that literary scholars overworked to the point of Sokalian fashionable nonsense seem positively fresh and sensible when applied to maps (see for yourself), and it’s rather puzzling that in the twenty-year history of the literary study of maps, fictional cartography has remained largely untouched.


submit to reddit

24 responses to “Here Be Cartographers: Reading the Fantasy Map”

  1. Two additional essays on maps that I somehow missed, despite them both being in blogs/publications I read on a regular basis: Johan Jönsson’s “The Reader and the Map”, and Matthew Cheney’s subsequent refutation, “Maps and Fantasy”. Thanks to Jonathan Crowe for pointing them out.

    Tuesday, 19 April 2011 at 1:36am

  2. Elliot McIntire

    In 1982 I published a short article on imaginary landscapes, drawing largely on sci fi examples. Here’s the reference.

    Elliot McIntire, Imaginary Landscapes, Yearbook, Association of Pacific Coast Geographers, Vol. 44, pp. 93-108, 1982.

    Tuesday, 19 April 2011 at 11:30pm

  3. What a fantastic and thought-provoking article. One of my passions is game design, and am very much interested in the distinction that exists there between an author’s ‘map’ of the game (the rules) and the game itself. It’s great to read an article on another topic and find so many great connections and inspirations. As someone once said…

    “The map is not the territory” — Alfred Korzybski

    Wednesday, 20 April 2011 at 5:47pm

  4. Still digesting this one (wonderful piece, by the way), but my initial thoughts:

    I’m pleased to see that Harley and Woodward’s efforts are being taken up, in part, by Mark Monmonier, of From Squaw Tit to Whorehouse Meadow fame. It’s always nice to see cartographers enjoy the baudy or mean aspects of their trade, just like most linguists love studying swear words.

    When I was young, I loved reading Jacques’ (RIP) Redwall series, and early on always blew by the maps. But as I grew older and more familiar with the canon, I came to really enjoy the maps in each book (see here), as they’re of the infinitely more interesting highly-stylized variety.

    Cartography in fiction, in many respects, is simply art with less sturm und drang; the Mitchell map you mention above strikes me as beautiful but without meaning beyond the literal; the Tolkein map seems a very obvious inspiration for the later Redwall maps–that is, communicating something about scale of import rather than scale of physicality.

    Thursday, 21 April 2011 at 5:52pm

  5. Pfly

    Thanks, that was fascinating.

    Friday, 22 April 2011 at 11:25am

  6. You might be interested in “Mapping Darkover”, an article on constructing maps of Marian Zimmer Bradley’s Darkover post hoc. Internal evidence makes it clear that Bradley herself was not working from a map at all: there is a published map in some of the books, but it is obviously and massively wrong in many ways. Indeed, the only way to get even a semblance of consistency is to assume that in the text of certain books of the Darkover series, east and west are reversed!

    Saturday, 23 April 2011 at 7:41pm

  7. Siobhan Carroll

    What a wonderful overview! Thanks for taking the time to write and post this.

    Thursday, 28 April 2011 at 6:48pm

  8. Thanks for the responses and especially the additional readings. Here’s another one that I missed, which covers many of the same examples—uncannily so, in fact—and shares my concerns about fictional maps clinging to the passive and literal when the fictions themselves are anything but.

    Sunday, 1 May 2011 at 5:25am

  9. Nicholas – many thanks for taking the time to write such a detailed, and well argued, essay. It was a pleasure to read.

    You wrote: “Outside of the vanity presses, this level of control is practically non-existent these days.” It is exactly for this reason that I chose not to present my constructed world to others through a set of novels … although the telling of stories (including in book form) is a part of the whole.

    For me, the maps of my conworld were the well from which the rest of the world could be drawn forth. Because I took the decision to display the world as a sort of online encyclopedia, where the reader is a part of the constructed world, I chose to make my maps as atlas-like as possible.

    I found my first map on the inside cover of “Winnie the Pooh”. I still love its simplicity of purpose and form. It puts my maps to shame for its sparse ability to tell the story.

    Monday, 16 May 2011 at 12:12pm

  10. David Harmon

    You touch briefly on “cosmological maps” as with the Norse depictions of Yggdrasil and it’s surrounding realms. As it happens, I so recall one example of such a map, from Gene Wolfe’s The Wizard Knight (which in fact is vaguely Norse-themed). While there’s no giant tree, the map does depict the nested worlds of the duology’s universe, and IIRC the text specifically states the map as the work of the protagonist/narrator.

    Tuesday, 17 May 2011 at 3:45am

  11. Well how’s that for synchronicity.

    Less than a month ago I landed my first book deal for a fantasy novel called “Magickless” I’m in the process of freaking out and celebrating all at once right now, and the one thing that has kept me sane while I wait for my editor to finish his copy edit has been creating a map.

    I also follow many blogs, and so you can imagine my joy when I followed one link to his article and found more than I could ever hoped to have searched out. Thank you for this brilliant article. You made so much clear about what I wanted for my own map.

    If anyone is interested, have a look at where I’m up to in the process….

    I’ve linked back to this article of course…

    Thank you again — the internet needs more articles like this.


    Tuesday, 17 May 2011 at 1:43pm

  12. Salim

    The Art History majors at my school have been passing around this article for some time. Your article is the bomb.

    I especially appreciate your discussion on the layers of authorship, worlds-within-worlds, and the ways of seeing (is the author inside the text or outside the text? Where is the reader located?) Great questions! I will be mulling these over every time I read a new book or play a video game (Everquest came with a cloth map that I still have)

    Thanks for posting the additional reading links. I will get to them in due time…

    Saturday, 14 July 2012 at 7:31am

  13. I absolutely love this post. Going to bookmark it :)

    Friday, 24 August 2012 at 1:11pm

  14. John Leland

    To me one of the most fascinating fantasy-mapping examples is in the first hardcover edition of Fletcher Pratt’s Well of the Unicorn. There is an overall map of the world (or more accurately region) of the story in the front, but then the heading for each chapter includes a smaller but more detailed map of the scene of the chapter. All the maps are in what we now call “fantasy” style (with little castle shapes for towns etc.) and beautifully drawn.

    Tuesday, 13 August 2013 at 6:44pm

Say something interesting: