Here Be Dragons (13 page)

Much of the relevant middle is conspicuously empty, however. While it is littered with names, there are only a few iconic map elements indicating terrain or buildings, roads or rivers. On the Shire map, the blank spaces correspond to a norm that does not need to be mapped. Here, the white emptiness, portrayed on the map by name only, is not a stand-in for fields too obvious to map, nor is it treeless, flat heaths, or desert, tundra, grassy plains, or any other one type of terrain. It is not even simply “wilderness.” On the general map, the fields of the Shire are just as blank as the desolate Brown Lands; the grasslands of Rohan are empty white, as is the broken wasteland of the Plateau of Gorgoroth. White is not the unknown or unmapped, nor is it a specific type of landscape: it is landscape that it is irrelevant to map. The only relevant features of this unmapped whiteness are the numerous names by which its areas are known. All those names—in English as well as Elvish, and providing translations between the two—communicate the importance of language
in itself. The map shows us a world defined by names and created by language, thus confirming that Middle-earth is a creation centered on language rather than nature—on the creation of new language and the translations involved in understanding it. Given Tolkien's love of languages and philology, it comes as no surprise that one of the principal messages delivered by the map of the western part of Middle-earth is how his fictional world is ultimately a linguistic creation.

• • •

At least since the 1970s, critics have observed the importance of maps to the fantasy genre, although no comprehensive studies have been published before now. Obviously, no study, quantitative or qualitative, can be all-encompassing, and my own contributions to the investigation of fantasy maps in this chapter should be seen as the first tentative mapping of an alien country, not as the definitive exploration of all its blank spaces. As in many studies, time and money came to be unwanted constraints. Maps proved to be more uncommon in the genre as a whole than I originally thought (occurring in at most two fifths of the books rather than in at least half, which a cursory pilot study had led me to believe would be the figure). Their scarcity in the book sample thus led to a smaller map sample that, in turn, resulted in large margins of error. To be certain that the largest margin of error could be at least halved, however, the book sample would need to be considerably larger, requiring almost 360 maps. The advantage of such a large map sample would, of course, be the appearance of more rare features (for instance, oriented any way other than northward; showing a subway system; or portraying modern cities of the primary world); even if a certain phenomenon did not appear in my sample, it might still be found on as many as one fantasy map in twenty-five. (With 360 maps, something not found in the sample would be rare enough only to appear on one map in every hundred in the fantasy-map population.)

My survey made plain that the vast majority of maps mapped settings in secondary worlds. A valid question that remains for a future study would thus be to what extent high-fantasy novels come with maps. Such a study would tell us the proportions of maps in high to low fantasy, but it would require either the books in the sampling frame or (probably more feasibly) those in the sample to be separated into high-and low-fantasy works (something for which there was no time in this survey). Future studies may also want to include more children's and young adult fiction in the sampling frame. My own sampling frame is biased away
from fiction for young readers because such fiction was separated from fantasy fiction in the database I used, with the result that the survey largely concerned fantasy for adult readers. Whether there are any major differences between maps in fantasy for adults and in fantasy for children and young adults remains to be investigated, for instance by using another sampling frame.

Certain typical features of fantasy maps were indicated by the survey, features we can expect to find in at least half of all fantasy maps. In brief, a typical fantasy map portrays a secondary world, a compass rose or similar device showing its orientation with north at the top. It is not set in any given hemisphere (not necessarily in a spherical world at all), although there are reasons to believe that clues in the text would indicate north as the direction of colder climates. Apart from topographical map elements such as rivers, bays, islands, and mountains, such a map would also contain towns and other artificial constructions. The hill signs used are typically pre-Enlightenment (either profile or oblique).

Even this brief list reveals the mixture of modern and historical map features. Like much high fantasy, the secondary-world maps follow a pseudomedieval aesthetic according to which dashes of pre-Enlightenment mapping conventions are rather routinely added to a mostly modern creation. Whether this is because of careless research, genre conformity, lack of imagination, or a desire to give the reader the easiest possible access to the map and the world it portrays is hard to say. If the map is meant as an aid for reading (and writing) the story, as a paratext on the threshold between the actual world and the unknown geography of the secondary world, maybe the map should simply challenge the reader's map conventions as little as possible.

Whatever the reason, a mixture of cartographic conventions from various time periods is found in the reading of the two maps from
The Lord of the Rings
as well. The readings also demonstrate that paying close attention to fantasy maps,
as maps
, can reveal information about the maps beyond the elements that were used in their construction. To find that center and periphery are set up against each other on both maps is hardly surprising; to set the familiar in focus, in the center of the map, is a traditional mapmaking strategy. But whereas the Shire map privileges the familiar over the unknown and communicates as its dominant message the control of the land (and the landscape) and the safety this brings, the map of western Middle-earth defines the world it portrays by naming it. The importance of the world's language and history is as
central to this map's message as is geography, and both maps contain a link between geography and story. Where the Shire map communicates control and safety, however, the map of western Middle-earth communicates a tension between cultural control and wilderness.

Apart from what the readings of the Tolkien maps tell us about
The Lord of the Rings
, they demonstrate how much any fantasy map can say about the work it belongs to. Each map, it is clear, relates in one way or another to the text. To ignore what the map communicates and only analyze the text means omitting a significant part of the work. Furthermore, as two maps from the same book, portraying parts of the same world, display such significant differences in the messages they communicate, it seems obvious that not only one map but all maps present in a book should be considered. Although the great majority of all fantasy novels that come with maps only include one map, a not insignificant number include two, and at least some come equipped with three or more. Each of these maps is a doceme that adds something to the document as a whole, and each map is a paratext that offers a particular threshold across which the text can be entered.

Whether the map is alone or one of several, typical or idiosyncratic, referring to a fictive map or situated firmly outside the diegesis, we should not dismiss it lightly. Instead, critics as well as readers should let the map do what is—ultimately—its job: to lead us into the fantastic world of the story.

J
ust like the actual world, all reasonably complex secondary worlds are divided into areas of various kinds. Divisions may be geographical or administrative in nature, with areas demarcated by, for instance, rivers, mountain ranges, beaches, hedges, ditches, dykes, or simply lines on a map. Crossing from one area into another may be fraught with peril, unexciting, or barely if at all noticeable.

In fantasy settings, whether primary or secondary worlds, other kinds of divisions and types of areas occur as well. Two areas, while side by side geographically, can have quite different rules for how—for instance—time, space, and causality work. A day in one place might be a year on the other side of the wall. In the middle of snowcapped mountains, there might be a valley of eternal summer. The magic power to change one's environment inside the forest might simply be superstitious nonsense outside. This chapter is devoted to an investigation of how demarcations between such dissimilar areas—domains—are constructed, how they reflect the domains on either side, and what their relevance is to the worlds where they occur.

In
The Encyclopedia of Fantasy
, the editors settle for
threshold
as the preferred term for the various dividing lines of fantasy landscapes and stories. The critic behind the “Threshold” article, John Clute, distinguishes between “physical” and “metaphorical” thresholds. The former type of threshold, the type relevant in this context, marks a “gradient between two places or states of being.”
¹
Although it would be possible to split hairs and wonder if a threshold is not, in fact, a line rather than a gradient, and to observe that in many cases, including the examples that follow, places and states of being are conflated, Clute's is a succinct and to-the-point definition. Clute proceeds to list four functions of the physical thresholds. First, they “normally form the spines of borderlands, demarcating regions which borderlands join together.” A borderland, he notes elsewhere, serves as a “marker, resting place or toll-gate between
two differing kinds of reality.”
²
Presumably, Clute's “normally” is not intended to imply that this is the most common function of a physical threshold in fantasy (I would find that hard to agree with) but that borderlands generally (“normally”) have a physical threshold as a defining feature around which they are situated. The function would thus be dual, both separating and joining two regions of different realities.

The second function of the physical threshold is to “announce the presence, or intrusion, of a crosshatch,” that is, a place where “two or more worlds may simultaneously inhabit the same territory.”
³
This function is clearly connected to the first (thresholds as spines of borderlands), in that borderlands often provide a strip-like crosshatch region.
⁴
Third, physical thresholds “constitute the perimeter of polders.” And finally, “for those of peculiar talents, they may comprise a map of the land.”
⁵
Land
in this context is taken to mean “a secondary-world venue whose nature and fate are central to the plot: a land is not a protagonist, but has an analogous role.”
⁶
If the third function is the most clear-cut, the fourth is the most puzzling; neither Roz Kaveney's cross-referenced entry on “Maps” nor David Langford's on “Talents” offers much in the way of enlightenment.
⁷

The term
threshold
is undeniably versatile when taken as described in the
Encyclopedia
. As a word, however,
threshold
implies not only a dividing line but also the intended crossing of such a line. The
Oxford English Dictionary
gives the figurative meaning of
threshold
as “the line which one crosses in entering,”
⁸
and Clute's remark that a threshold “may not even be meant to be liminal, or passable”
⁹
only serves to broaden an already broad blanket term. For all the usefulness of
threshold
as a term, the geographical focus of this discussion requires it to be complemented by two more specific terms:
border
and
boundary
. A border corresponds to the first two functions of the physical threshold. It is a line (or gradient) that separates two places or areas, and it differs from a boundary in that the latter implies a perimeter or circumference. In other words, you can be on either side of a border, but inside or outside a boundary. A polder, for example, is surrounded by its boundary, while two adjacent domains are separated by a border. It should be noted that neither word implies any intended crossing, and thus cannot be fully subsumed under
threshold
.

TOGETHER APART: BORDERS IN BRUST, GAIMAN AND VESS, AND NIX

The fantasy genre offers a great variety of borders, and this section will investigate three examples, representing three different kinds. The first is the border between the land of the living and the land of the dead. Journeys to the land of the dead form a common theme in several mythologies and so-called taproot texts (texts that predate the emergence of generic fantasy but that include the fantastic and are of heightened significance to the genre
¹⁰
). The Sumerian goddess Inanna descends into the underworld only to end up captured there. The Japanese god Izanagi enters the dark realm of Yomi to bring back his spouse Izanami. Odysseus and Æneas both venture there for information, and Orpheus attempts to bring back his Eurydice. The Norse gods go down to Hel to bind Fenrir. Examples of how the land of the dead can be reached if one travels to the right place are common in fantasy literature as well. In J. R. R. Tolkien's Middle-earth, elves have to wait out the end of the world in the Halls of Mandos, which are located in the Undying Lands (and thus accessible to those elves who are still alive). Fritz Leiber has his heroes Fafhrd and the Grey Mouser ride to the land of death in “The Price of Pain-Ease” (1970). Raymond E. Feist's protagonists Pug and Tomas journey to the Halls of the Dead (albeit with some magical help, as it is a very long trip) in
A Darkness at Sethanon
(1986). With the Celtic king Urtha, Merlin enters the Ghostland, where the spirits of the dead reside, in Robert Holdstock's
Celtika
(2001). Jasper Fforde writes about how Thursday Next and her colleague Spike drive off the M4 motorway to a service station limbo beyond which the domain of death can be entered (
Something Rotten
2004). Later, I discuss the Deathgate and the Paths of the Dead in Steven Brust's Dragaera books, mainly
Taltos
(1998) from the Vlad Taltos series and
The Paths of the Dead
(2002) from the Khaavren Romances. Through their very different narrators and the close focus on the actual crossing of borders, these books offer an unusually clear example of the border between the realms of life and death. The Dragaeran realm of the dead is reached easily enough by descending a waterfall, leaving the land of the living at the top. Returning, however, requires special dispensation. The two trips described also illuminate how the border between the realms of life and death can be employed to present very different perspectives on Dragaeran society.