(2000), for example, tell us of the heroine’s entry into the guild and the conditions pertaining to that place. Later, Rani worries about a theft she has committed:
The theft made Rani nervous—she had been hired to clean the captain’s quarters and she would be the most likely suspect when the soldier discovered his loss. She had become enough of a fixture in the Soldiers’ Quarter that any of the girls who consorted with the guards would know to find her in Garadolo’s lair.…
Even now, she looked back to her life as an apprentice with blinding fondness. She’d been so lucky then, so privileged that her most difficult task had been scrubbing a whitewashed table. (184–185)
What should be already known to us, the context of the world, is delivered as memory, and more specifically, as story.10 Because we cannot stand outside the narrative, the “omniscient narrator” is compromised: it is able to tell us only what Rani knows. We both see into her mind but are not, as in the first-person narrative, inside her head. The “omniscient narrator” limits our vision while asserting that we have privileged insight.
Occasionally these reveries are expressed as mutual revelation, a device Deborah Tannen suggests in real life is intended to show rapport: “By this strategy, the speaker expects his or her statement of personal experience to elicit a similar statement from the other.” Unfortunately, authors rarely seem to remember that the “effectiveness of this device is dependent upon the sharedness of the system” (79). The single direction of information works instead to indicate the status-within-the-story of the speaker.
To steal yet again Clute’s idea of “making storyable,” I note that these reveries make storyable character and characteristics. Indeed, reverie and self-contemplation, far from creating depth, break the sense of immersion in a society, and are fundamentally antithetical to either character development or an immersive structure. It is a false mimesis that reminds us that we are in a narrated text and that the protagonist’s version must be true. To doubt the validity of the reverie would be to destroy the impermeable nature of the club discourse: either the reverie is “true” or the entire structure collapses. Such is the case even where, as in Tad Williams’s The Dragonbone Chair (1989), the reverie that describes the protagonist is actually that of another person; it is Rachel (31–38) who—with improbable detail—remembers Simon’s childhood for us. On the other hand, Robin Hobb, in the opening pages of The Farseer: Assassin’s Apprentice (1996), acknowledges the persistent doubts: “on that day they [memories] suddenly begin, with a brightness and detail that overwhelms me. Sometimes it seems too complete and I wonder if it is truly mine. Am I recalling it from my own mind, or from dozens of retellings by legions of kitchen maids and ranks of scullions and herds of stables boys as they explained my presence to each other” (2).
Memory does not have to be delivered this way. Lloyd Alexander gets to the heart of the matter in The Black Cauldron (1965), a quest fantasy that is also a rite of passage novel, but is not a portal fantasy inasmuch as Taran is always familiar with his world. When Taran tries to buy the cauldron from Orddu, Orwen, and Orgoch in the marshes of Morva, one suggested price is a warm summer’s day. Memory here is sense and feeling, yet this suggestion and its alternative—that they take one of his memories of Eillonwy—tell us much more about Taran than does a recounting of his childhood. Similarly, Tad Williams does not need to have his hero Simon think about being curious; his actions tells us that he is. Simon “could never understand how rooms that seemed as small as the doctor’s did from outside—he had looked down on them from the bailey walls and paced the distance in the courtyard—how they could have such long corridors” (The Dragonbone Chair 14). At this point Simon is in the world in which he grew up, and Williams demonstrates neatly that in this—pre-quest—world he is fully immersed.
The reverie is a commodification of memory: one aspect of this ritualization of memory is that it reduces characters to that which can be described in terms essentially photographic. Stephen R. Donaldson, in Lord Foul’s Bane (1977), provides us with a good example: “But where Lena was fresh and slim of line, full of unbroken newness, Atiaran appeared complex, almost self contradictory” (51). This kind of description tells us more than we can possibly know or that the observing character can possibly know. There are two aspects to this formulation. To begin with, the narrative structure of the portal-quest fantasy, in which we move through the map, posits many characters as mere signposts. In this context we do not have time to truly learn about people, any more than we would “learn” about the tree we passed. The second, and related, aspect is that we are therefore forced to rely on intuition. The portal-quest fantasy is thus often the last resting place of physiognomy, a tradition the sf writer Jeffrey Ford has mocked brilliantly in The Physiognomy (1997). This tradition is usually discussed in terms of race and fantasy, but the “racing” of heroes and villains is in part a consequence of a rhetoric that posits insight in terms of visual perception.
The narrativization of memory also affects the description of action. It is not action that is of interest but response; action has to be “seen” by the protagonist and the protagonist cannot see “his” own physical movements. In this form, expression and feeling cannot be interpreted; they must be described, pinned down for the reader. They must also maintain primacy in the narrative drive of the novel. In this mode of fantasy, action is there to carry emotive weight, so that fighting is “valiant” (see Brooks again) rather than “wild” (see Howard’s Conan the Barbarian). This emphasis on emotion is a distinct shift from, for example, the worlds of Conan the barbarian. The shift is from an externalized discourse of action—what we usually mean when we refer to the omniscient narrator—to the internalized discourse. Even the moment of recognition can be shaped by this demand. Toward the end of Lord Foul’s Bane, Thomas Covenant learns some of what was intended for him, but it is framed not as action, but as self-regard: “Then, with a sickening vertiginous twist of insight, he caught a glimpse of Lord Foul’s plan for him, glimpsed what the despiser was doing to him. Here was the killing blow which had lain concealed behind all the machinations, all the subterfuge” (353).
This raises the question as to whether this internal narrative is intended to parallel the external narrative of exploration and observation. Can one have an internal quest that does not require the protagonist to move through a physical and internal landscape? There are very few portal fantasies that are not succeeded by journeys.11 The implication of what I have set out above is that the journey through the mind, while not absolutely necessary to the format, is one way that has emerged to handle the need to create closed and reliable characters—reliable only in the sense that we must trust their own assessment of themselves in order for the veracity of the story to hold. Thomas Covenant is a twist, not an exception to the rule. His misinterpretation of his role, his refusal to accept responsibility, is the surety of his character’s narration. We believe him because Doubting Thomas is, in the final analysis, the most reliable witness—a double bluff.
This discourse and the insistence on the narrative and descriptive competence of the protagonist—even when we are told that they do not understand what is happening—thins the complexity of the world and makes of it a poorly painted stage set. The portal-quest fantasy by its very nature needs to deny the possibility of a polysemic discourse in order to validate the “quest.” There can be only one understanding of the world: an understanding that validates the quest. And yes, that is recursive, a point I shall discuss further on.
There are some exceptions: the original grail stories offer polysemic narratives and question the reality, desirability, and possibility of “goal,” as do some modern quest fantasies (which I shall discuss at the end of the chapter). More generally, however, this issue is extended into the world-building of fantasyland. Nonspecific landscape is unrolled like a carpet in front of the character. This landscape even embraces the contrived design of Romantic landscape painting.12 Note this comment from Raymond J. O’Brien: “Viewers, whether river travellers or gallery-goers, were commonly impressed with the importance of foreground. In landscapes,
