Narrative Grammars

1. An Interesting Comment

On an earlier post on genre, Once More Into the Fray, Jakob Schmidt posted an interesting comment that I thought I'd deal with up here, since I didn't have time to respond to it on the actual entry, what with Christmas and all:

To be honest, I'm not too convinced by your set-subset model of fantasy and sf ...

If I'm understanding it right, your argument at this point rests upon the notion that

a) sf and fantasy narratives have certain "elements" (dragons, ftl, huge hairy monsters)

b) These notions can either be rationalized with regards to our contemporary paradigm of the world (sf) or not (fantasy, which would make it the more general set).

But I'm not sure if the distinction between the elements and their rationalisation holds, or, put differently: why don't we regard rationalization itself as an element, alongside dragons, ftl and huge, hairy monsters?

If we would do so, however, a distinction based upon the presence or absence of this one particular element would seem even more arbitrary. But that might turn out as a win rather than a loss, because we would be forced to look more closely at the narrative structure of fantasy and sf stories instead of the elements they feature (for example how a certain type of fantasy narrative is typically about the return of the "old" order of things, how certain types of sf are about establishing a "new" order, how certain types of horror are about realizing how fucked up things are in the first place). The question, after all, is not if something essentially IS sf or fantasy (or both at once), but why certain narratives strike us as fantasy and others as sf - and why this perception may vary from reader to reader. This may be less due to the presence of certain elements and more about the question of how a narrative is structure (and how "competent" we are individually in detecting and recognising certain structures).

What I'm getting at is actually John Clute's model of narrative grammars of the fantastic genres, which is less interested in making clear distinctions between them, and more in how narrative grammars of different types of the fantastic are structurally similar - and still very distinct in certain regards. He recently gave a very short summary of this model:

http://www.johnclute.co.uk/word/?p=15

I'd recommend taking the time to follow the link through and give Clute's essay a read. Go on. It'll only take a few minutes.

OK?

OK. So...

2. Fantasy as a Subset of Fantastika

Really, my earlier post is not so much an attempt to distinguish genres as it is an attempt to analyse the distinctions already at play, to unpack the politics of the argument(s). My own take on it is that the set-subset model of Fantasy and SF is just one model, that it's in conflict with the alternative subset+subset model in which SF and Fantasy are both contained within a larger superset. I may not have been clear enough in expressing my own belief both models are flawed. Truth is, I think the whole debate is fucked up beyond all reason, degenerated into endless disputes over semantic boundaries because the basic terms are overloaded. We end up talking at cross-purposes when we use these terms; some will mean the marketing category, some will mean works of a specific formula associated with that category, some will mean works of a particular aesthetic form they consider characteristic of that category, some will mean works of a multitude of aesthetic forms that are all published in that category, and some will mean the sum of all works in all those varied aesthetic forms, whatever category they were published in.

Clute acknowledges that the terms SF, Fantasy and Horror are problematic, but in accepting that we're stuck with them I think he fails to deal with this problem. In taking the subset+subset model as a given, treating SF, Fantasy and Horror as subtypes of the larger set of fictions he refers to as "fantastika", he presents his narrative grammars as the additional A, B and C which distinguish these genres from each other. But this approach, directly linking each label to a particular aesthetic form, is at odds with many who would use the terms as signifiers of a loose aggregation of works in multiple aesthetic forms, or who would see the underlying aesthetic form as something much less rigidly circumscribed in terms of narrative.

Essentially, in fact, the way he describes fantastika is indistinguishable from the way proponents of the set-subset model would describe fantasy -- as "fictional works whose contents are understood to be fantastic" -- with the word fantasy referring simply to work which utilises that literary technique. The problem of overload is pressing enough that there's a tendency to use capitalisation to distinguish Fantasy as a market-defined genre (upper case) from fantasy as a literary mode (lower case), but for many that's enough. As far as they are concerned Clute is simply relabelling the literary mode they would happily call fantasy, and it's hard to see much value in choosing his coinage over the more straightforward term given that it comes from the same root, carries the same denotative and connotative baggage and is distinguished only by the elevated register that it garners from the exotic ending -- fantastika rather than fantastic. How does that make it any more apt as a label when it comes to "fictional works whose contents are understood to be fantastic"? Why should we not refer to these works characterised by their fantastic content as works of fantasy written by fantasists, if the alternative on offer is simply a rather rarified way of saying the exact same thing?

But from that perspective, in which fantasy is simply those "fictional works whose contents are understood to be fantastic", there's a deeper problem with Clute's approach to "fantastika" and "Fantasy". Regardless of the fact that his model is grounded in narrative grammars rather than tropes, Clute describes Fantasy entirely in terms of a particular aesthetic form characterised by all those tired cliches of heritage ("the old ways"), noble blood ("the hero... the king"), the pre-industrial pastoral ("the harvest... the Land"), and military spectacularism ("battle after battle"). In Clute's model Fantasy signifies what others would say is very specific subgenre variously referred to as Epic Fantasy, High Fantasy or Heroic Fantasy.

His narrative grammar holds a lot of interest for me as a critical approach. And there's no question that this type of Heroic Fantasy is commonly perceived to be characteristic of the genre. But if we question the validity of that common perception we also have to question the validity of using the term Fantasy in that context, just as someone widely read in the field of SF would question a view of SF that characterised it purely in terms of Space Opera.

This is the core of the problem for me: when you have so much of what Clute calls fantastika published, referred to and thought of as fantasy, when the dealer's room at the World Fantasy Convention has way more of that sort of fantastika than it has Heroic Fantasy, when the fantastika actually seems more representative of the field (or at least its cutting edge) than the works which fit Clute's narrative grammar, when the fantastika features so prominently in nominations for the World Fantasy Awards (c.f. the 2006 awards where the shortlist included Ellis's LUNAR PARK, Murakami's KAFKA ON THE SHORE and Joyce's THE LIMITS OF ENCHANTMENT) -- then Clute's approach becomes, for me, as problematic in its exclusion of that type of work as a notion of SF that excluded everything that wasn't stereotypical Sci-Fi.

A large part of the issue here is political. It's all too common to see this deeply narrowed (one might even say blinkered) view of Fantasy accepted implicitly, especially on the SF side of the debate, common enough that I'm sure Clute's conflation of Fantasy with Heroic Fantasy would go unchallenged by many people, his narrative grammar of Fantasy seeming broadly valid. But there's a whole host of writers and readers for whom the term fantasy has sod all to do with heritage, noble blood, the pre-industrial pastoral and military spectacularism, for whom Fantasy is a much broader field basically equivalent to Clute's fantastika, and for whom that preconception of Fantasy as Heroic is a stereotype to be taken apart. For them the set-subset model is a stance against that stereotype of Fantasy, and against the attendant recategorisation of any Fantasy that doesn't fit the stereotype.

If anyone should get this, of course, it's the SF community. The application of a stereotype to a genre and the subsequent relabelling of every work that doesn't fit that stereotype as not really part of the genre? Where have we heard that one before? Cause, yeah, Science Fiction is all about the robots and the aliens and the spaceships, right? That's what Sci-Fi is -- sorry, Science Fiction, whatever. What's that you say? NINETEEN EIGHTY-FOUR, you say? Well, you can't really call that Science Fiction. It's just fiction.

So, yeah, that's where the idea of Fantasy as a subset of fantastika is going to be a problem for some.

3. The Fantastic and the Strange

Still, as much as I find that stereotyping infuriating, as much as I think it's hypocritical, disingenuous and (perhaps wilfully) ignorant of the range of approaches within Fantasy, I don't blame people for buying into that stereotype. No more than I blame those who have the exact same attitude to SF. If you want to go back in time, create a pulp genre called War Fiction which is dominated by generic, formulaic dross, and then sell CATCH-22 alongside Commando comics and novelisations of The Sands of Iwo Jima (maybe even in a specialised shop that carries Action Man figures and toy soldiers too!), you can't expect people who aren't fans to have the remotest inkling that "War Fiction" is anything more than the stereotype they're familiar with. Especially when generations of writers, readers and critics have spent decades arguing over whether or not CATCH-22 actually counts as "War Fiction" given its lack of military plausibility.

This is the real crux of it, that the confusion inherent in our current critical vocabulary -- in which one man's fantastika is another man's fantasy, and one man's fantasy is another man's Fantasy, and one man's Fantasy is another man's Heroic Fantasy -- just leads us into a tiresome fuckload of genre slapfests born of incompatible terminologies. This is why I'd rather torch the whole shithouse of rotted names and start from first principles, ditching the overloaded label of Fantasy completely, and using the term strange fiction instead, to refer to that broad set of "fictional works whose contents are understood to be strange".

Of course, you could query this on the same level I've queried Clute's fantastika. Why should we not refer to these works characterised by their fantastic content as works of fantasy written by fantasists, if the alternative on offer is simply another way of saying the exact same thing? Well, the apparent inability of anyone who doesn't consider themselves a fantasy writer or reader to get to grips with the notion that fantasy does not necessarily equate to magic and elves and dragons and shit is one purely pragmatic reason. Hell, the fact that even the genre's most respected critic applies the term "Fantasy" to a narrative grammar defined entirely in terms of the Heroic has to be some sort of indication that the word is not going to lose those associations any time soon.

More importantly though, I'd argue that "strange" offers a distance from those associated meanings that are introduced whenever we talk of the content of Fantasy, SF and Horror at a genre-specific level. When we talk of the "novum" or the "uncanny" it may be more obvious than when we talk of the "fantastic" but all three terms carry their own set of associations -- desire, dread, wonder, hope, novelty, practicality, impracticality. Maybe the real problem with the wider use of the term "fantasy" is that for many it is simply indelibly coupled with a sense of desire, of yearning, with compensatory daydreams and wistful reveries. In that context I think the logical end result is the exact deadlock of conflicting definitions we find in the SF/Fantasy debate.

I'm less concerned with fighting a side in that debate than I am with breaking the deadlock by identifying the exact point(s) of contention, so for me, the term "strange" offers the neutrality of a fresh slate. It nixes these associations. It carries no further proposition, explicit or implicit, about the nature of the unusual / abnormal / anti-mimetic "contents" by which we can characterise this type of work . Or rather, to be more accurate, it carries no implications as to how we respond to these contents. It simply says that they breach our expectations that the narrative will function as a representation of a pseudo-reality modelled on our own.

4. Shifts in Subjunctivity

While I talk about fantastic elements and rationalisation in that earlier post, these are parts of the inherited terminology of the SF / Fantasy debate that I've tried to get away from elsewhere. Even Clute is defining his fantastika in terms of "contents" that are "understood to be fantastic". But as much as I do think that the conceit plays a hugely important role in strange fiction, I'm more interested in how these are constructed out of breaches of expectation than in treating them as tropes (dragons, FTL, etc.), as "ingredients" that suddenly transform the genre of a narrative just by being dropped into it. Or, for that matter, that suddenly transform the genre of the narrative again if we only add another "ingredient" of rationalisation. Essentially, I'd reverse the polarity, treat those "fantastic elements" as the end results of a literary technique of estrangement, the effects of strangeness rather than the cause. Where I talk of fantastic elements, I don't mean genre tropes -- dragons, spaceships, magic, FTL -- as things which, in and of themselves, make fiction strange. Rather I see those as the epiphenomena of an underlying process where the sentences of a fictive text function as propositions with subjunctivities.

This is riffing off Delany's essay on genre, "About Five Thousand Seven Hundred and Fifty Words", where he defines genre in terms of subjunctivity:

Suppose a series of words is presented to us as a piece of reportage. A blanket indicative tension informs the whole series: this happened. That is the particular level of subjunctivity at which journalism takes place.

Delany contrasts the representation of actualities, events that "have happened", the level of subjunctivity we find in reportage, with the representation of possibilities, events that "could have happened", the level of subjunctivity we find in naturalistic fiction. He goes on to characterise Fantasy as the representation of impossibilities, events that "could not have happened", and speculative fiction as the representation of non-actualities, events that "have not happened", even going so far as to suggest in a note that naturalistic fiction can be seen as a subset of speculative fiction.; after all, the set of events that "have not happened" includes not just the set of events that "have not happened yet" (the hypotheticals of SF) but also the set of events that "have not happened but could have" (the counterfactuals of Alternative History), which is basically another perspective on the level of subjunctivity we find in naturalistic fiction where events "could have happened (but did not)".

I think it's a bit more complicated than that. The way I see it, all fiction performs acts of mimesis, each sentence presenting itself as an act of representation of an ersatz actuality. We know that the representation is an artifice, that the events described "have not happened", but for the sake of the narrative we suspend our disbelief. We read the sentence and play a game, pretending to ourselves that the narrator is not in fact breaching Grice's Maxim of Quality ("Do not say that which you believe to be false or that for which you lack evidence"). Unless the writer starts dropping hints that the narrator is unreliable, we take the text on face value, as a representation of events that could have happened. As long as these events are not impossible in a logical or nomological sense, as long as there's no inherent self-contradiction, no contradiction of the laws of nature as we understand them, these events are viewed as temporal possibilities. They contradict the actual history of the world, but that actuality is considered contingent. The events could have happened elsewhen.

All fiction constructs such an elsewhen in our imagination. Where fiction is seen as naturalistic, realistic, it is because that elsewhen is so closely modelled on the world we live in that most of the events described could have happened. All strange fiction contains naturalism in that sense, the mimesis of sentences that function on this level of subjunctivity. What makes it strange is that it also involves a shift of subjunctivity from "could have happened" to "could not have happened", wherever the narrative performs a sentence that cannot be read as simply mimetic. In some sense -- logical or nomological -- these events are just not possible. They contradict the very nature of the world as we understand it.

While Delany views each genre as having its own essential level of subjunctivity however, I think our suspension of disbelief requires that the sense of "could have happened" persists through the reading experience, reinforced by the mimesis even as it's disrupted by sentences with a subjunctivity level of "could not have happened". In this theory, all strange fiction involves a tension of subjunctivities, our suspension of disbelief at odds with our sense of incredulity, generating a pervasive estrangement, a cognitive dissonance.

Where genre comes in is largely a matter of what type of impossibility we ascribe to those events that "could not have happened". In the "literary" camps of both SF and Fantasy, and in allied territories like slipstream and postmodernism, where experimentalist approaches are par for the course, we may well see logical impossibilities, outright self-contradictions, but for the most part the impossibilities we're dealing with are nomological. The event shown contradicts what we think of as "how the world works", the set of contingent truths -- laws, principles and actualities -- that limit our physical and technical capabilities.

This is where, I suspect, many would seek to make a distinction between Fantasy and SF, arguing that the former allows physical impossibilities -- events that "could never happen" -- while the latter rules these out, allowing only technical impossibilities -- events that "could still happen", because although we currently lack the technical capacity this may not always be the case. We can't travel between the stars, talk with aliens or build sentient robots, but only because we lack the technology to do so at the moment. These are only impossibilities in terms of how the world works now.

My argument in the previous post (and elsewhere) is that this view is disingenuous, that only "Hard SF" is anywhere near so restrictive, that SF in general is happy to use physical impossibilities like FTL or jaunting, and that the Paradigm Shift Caveat (the idea that our understanding of "how the world works" may change radically) is essentially an act of wishful thinking that seeks to disguise such physical impossibilities as merely technical impossibilities. In fact, if we analyse the elsewhens of strange fiction as temporal displacements -- forward into the future, sideways into parallel worlds, or away into the secondary worlds orthogonal to both dimensions -- and the way these displacements manifest as hypothetical, counterfactual or metaphysical conceits, any honest appraisal of SF ultimately has to, I'd argue, admit of the presence of all three dimensions of strangeness.

Ultimately, I have no problem with the idea that SF requires a greater degree of plausibility than other forms of strange fiction, but I think it's important to distinguish the subjective perception of greater possibility from the objective reality of greater possibility. That subjective sense within the reading experience can be achieved by various mechanisms in the writing, only one of which is to limit the strangeness to technical impossibilities.

Disguising a physical impossibility as a technical impossibility, masking a metaphysical conceit (like FTL or jaunting) as hypothetical, by simply presenting it as a future reality, may be effective in its own right, given the general applicability of the Paradigm Shift Caveat (hence Bester gives us jaunting in the future rather than the present). We can literalise the "paradigm shift" by presenting the conceit as a product of an entirely alien culture (hence the prevalence of extinct ancient races leaving behind technology advanced beyond our understanding). If others have used the conceit before us then the conventionality it accrues also makes it easier to swallow, situating it in its own ersatz nomology (hence the acceptance of FTL as a tradition of how the world works within genre SF, and hence the growing popularity of tropes like wormholes, stargates and jump-points as a more recent tradition). We can create a greater sense of the conceit having reasonable foundations through the act of world-building, burying the metaphysical impossibility among hypotheticals (hence Bester puts jaunting in the context of spaceships and asteroid mining, and hence the common adage in SF that the writer is allowed "one impossible idea"). Working a little scientific or pseudo-scientific explication into that world-building doesn't hurt (so Bester and a myriad of other writers present a magical power like ESP and jaunting as a future-historical development, a "next stage in human evolution").

None of these techniques make the physical impossibilities any more rational as actual speculations -- which is why proponents of Hard SF scorn them -- but they do facilitate the rationalisation of these impossibilities. They offer the get-out clauses, cyclic arguments and cover stories by which we construct and sustain an artificial sense of contingency, of a potential alternative nomology. They persuade us into a further shift in subjunctivity, from "could not have happened" to "could have happened if..." where what follows that "if" may amount to a wholesale revision of how the world works validated more by self-delusion than speculation. So the strange becomes the novum of SF, with this conflict of subjunctivities creating a more palpable type of cognitive dissonance between the nomology in which the events presented "could not have happened" and the nomology in which they "could have happened". Where I talk of rationalisation as if it were another element, it's this process I'm really referring to, an act that takes place as much in the reading as in the writing.

The disparities between how far different readers are ready to rationalise the impossible, the differing effectiveness of different mechanisms for different readers, the multiple points of contention over where and how the nomological revisions become so wild as to be "just plain silly", where and how the sense of plausibility collapses -- these go a long way, I think, to explaining the disagreements between readers as to what does and what doesn't constitute an SF narrative.

5. Boulomaic Modality

Where it all becomes more complicated is in the way, I'd argue, additional subjunctivities may be introduced into the narrative and used to generate a sense of modality that shades the subjunctivity. In so far as subjunctivity is to do with levels of possibility, this may be stretching the meaning of the term, but in so far as our notion of nomological possibility, of "how the world works", is experienced as a sense of natural order, bound to and coloured by our aesthetics and ethics, of "how the world should work", I think it's worth our while to extend the term to boulomaic as well as epistemic modalities. If the events portrayed in strange fiction "could have happened" or "could not have happened" according to our sense of nomological possibility, then there is also a degree to which, according to our sense of aesthetic and ethical necessity these events "should have happened" or "should not have happened". I don't think it's hard to see how a reader's reaction to the strange may add exactly this sort of boulomaic modality, particularly with Horror, where the strange becomes the uncanny, where the transgression is as much moral as nomological, where the events not only "could not have happened" but "should not have happened"

This is where, I think, Clute's narrative grammars come into play and fit, for the most part, with this idea of strange fiction as produced by shifting subjunctivities and the tensions between them. So I'll start with the Stage One(s) that Clute identifies for Fantasy, SF and Horror:

1) Wrongness. Some small desiccating hint that the world has lost its wholeness.

1) Novum. Darko Suvin’s term for that aspect of the SF world which differs measurably from our given world.

1) Sighting. Some small sour lesion in the world is suddenly visible, even in daylight.

To my mind, Clute's Stage Ones for Horror and SF make a lot of sense, but even if we come to terms with the narrowing of scope as regards Fantasy, accepting Clute's narrative grammar as referring specifically to Heroic Fantasy (and to belabour my point about this type of Heroic Fantasy not being the full story, we only have to look at a work like GORMENGHAST to find a "Fantasy" narrative at odds with this), his identification of "wrongness" as the Stage One is, I think, mistaken. Rather I think it's the "wholeness" we should be focusing on, the sense of enchantment that comes from the extent to which the secondary worlds of Heroic Fantasy embody our desires and daydreams, our nostalgia for the simplicities of childhood and the past as they are reconstructed in our idealising imaginations. All of the features by which we're identifying this type of fantasy and by which many examples of this fiction are judged (by writers like Moorcock in his WIZARDRY AND WILD ROMANCE, for example) to be infantilist and reactionary -- heritage, noble blood, the pre-industrial pastoral, and heroic spectacularism -- fit neatly, I think, with an idea that the strange is used initially to construct a sense of the idyllic, a sense of "rightness" rather than "wrongness". What we're talking about is the introduction of a "should have happened" subjunctivity..

There's an interesting point here, I think. Where the strange kicks off the narratives of Horror (in the shape of the uncanny) and SF (in the shape of the novum) by establishing a tension, with Heroic Fantasy it simply sets the scene. To get to the "wrongness" which Clute takes as his starting point for Heroic Fantasy, you have to introduce something else that disrupts the base level strangeness of the idyll. In the grand-daddy of this type of fiction, THE LORD OF THE RINGS, it's another level of strangeness that disrupts the bucolic idyll of the Shire in particular and Middle Earth in general -- in the shape of Sauron, the Ringwraiths, the Nazguls, the whole host of Mordor, and the ring itself, of course. I don't think it's unfair to look at Heroic Fantasy in general as utilising the strange in the same manner as Tolkien, disrupting the idyll by introducing supernatural evil, "unnatural forces". What Clute calls "small dessicating hints that the world has lost its wholeness" might well be seen in the exact same terms he applies to Horror, as "small sour lesions on the world."

Where hobbits, dwarves and elves impart a subjunctivity level of "should have happened" in our yearning for the fabulous then, from the first grave warning of Gandalf through to the destruction of the ring the reader is constantly encountering the subjunctivity level of Horror, faced with the demonic events of past and present, that which "should not have happened", and the dreadful events of the future to be averted, that which "must not happen". In both modes of fiction the uncanny already is happening in one sense -- this is why we have these small dessicating hints of wrongness, these small sour lesions on the world -- that which must not happen has already begun. But it has also "not happened yet", not wholly, and so the possibility of prevention kick-starts the narrative.

(A quick note: It's tempting to suggest, when we look at THE LORD OF THE RINGS in particular, that some of these "small dessicating hints" or "small sour lesions" are open to a reading as a sort of novum -- or perhaps anti-novum would be a better term -- that in these worlds of Heroic Fantasy backdated to the agricultural era, the strangeness that the hero must face is to some extent the strangeness of modernity -- i.e. that Heroic Fantasy manifests the futureshock of the reactionary in the face of the novum we are confronted with in present reality, albeit rearticulated in the vernacular of the mediaeval peasant, as the dark mechanics of black magic. It doesn't seem too much of a stretch when you watch Peter Jackson's take on the trilogy with Saruman's transformation of his realm into an almost Blakean "dark satanic mill". But this is kind of off-topic, food for thought more than anything else. Anyway...)

If we find the uncanny as narrative trigger in Heroic Fantasy it is important to note that we may also find the idyllic as background in Horror. Given the degree to which Gothic Romanticism underpins much of the work published as Horror, we should not assume that the world in which the Horror narrative is triggered by the manifestation of "small sour lesions" will necessarily be modelled on normative reality. Rather we may discern Horror narratives in which the strange (initially) embodies our desires and daydreams as much as it does in Heroic Fantasy. In the fiction of Anne Rice or Poppy Z. Brite, for example, the foundation is not our given reality but rather the dark idyll of the vampire's night-world. If Heroic Fantasy can be critiqued as infantilist in its preoccupations, this type of Horror might well be critiqued as adolescent. Embodying perverse idealisations of sex and power as acts of transgression, it is the libertine's idyll, the fetishist's idyll, the decadent idyll of the teenage rebel at odds with moral authority but defined entirely in terms of that antagonism; but it is nevertheless an idyll. The impossibilities it presents as realities are just as much articulations of desire. We can clearly discern a subjunctivity level of "should have happened" at work here.

If we are seeking the key differences between Heroic Fantasy and Horror, then, in terms of the narrative grammars that structure them (viewing narrative as static composition), or in terms of the narrative logic that drives them (viewing narrative as dynamic process), I am not sure Clute's Stage One offers a clear distinction. Fundamentally, the hero in each is cast by Fate into the role of combat with unnatural forces that the narrative will gradually reveal as being behind all those "small dessicating hints" / "small sour lesions". In such a situation, in both Heroic Fantasy and Horror, the hero is faced with the challenge of that which "must not happen", but backed up by the fact that it "has not happened yet". It may be better to look at how that combat plays out as where the distinction between the two genres is born.

6. Epistemic Necessity

As that combat begins, even at the very start, the narrative may provide an optimistic or pessimistic representation of how those larger forces relate to the world, whether they embody an unacceptable lie (Heroic Fantasy) or an unacceptable truth (Horror). Through negations or validations of the authority of the uncanny (presenting it as alien and fundamentally illegitimate or alienated and fundamentally legitimate), combined with representations of the activity or passivity of the hero (presenting the hero as empowered or disempowered), a sense of epistemic necessity may be instilled in the reader, a sense that the outcome is inevitable, whether it involves the hero as victim, as in the case of Horror, or as victor, as in the case of Heroic Fantasy. In essence, when it comes to those events that "must not happen" but "have not happened yet", the Horror narrative is ascribing a subjunctivity level of "will happen" while Heroic Fantasy is ascribing a subjunctivity level of "will not happen".

We can look at these narrative grammars as narrative logics, generated from conflicts of subjunctivities, and involving the interplay of boulomaic modality and epistemic necessity. If we do so, I think Clute's Stage Two offers an intriguing parallel:

2) Thinning. The diminution of the old ways; amnesia of the hero and of the king; the harvest fails, the Land dries up; diversion of story into useless noise; battle after battle.

2) Cognitive Estrangement. Suvin’s term — modified from Vikor Shklovsky and Bertolt Brecht — for arguable and therefore structured defamiliarization of the world, which derives in part from the fact of Novum, and which allows the defectiveness of the ruling paradigm to be seen whole.

2) Thickening. The protagonist is mired deeper and deeper in the falseness of the world. The plot literally thickens around him. Fatally, he may think he understands himself, but in fact every move he makes deepens his amnesia, which coils through Thickening like fog; intensifies his resistance, his golem-like rigidity at the threat of change. It is a Gnostic phase: the truth is occluded, which allows us to lie to ourselves constantly.

In these stages, what Clute calls "thinning" and "thickening" is, I would argue, the manifestation of tensions designed to counteract the sense of epistemic necessity that the narrative logic imposes upon the outcome. Narrative requires at least an illusion of unpredictability, so that same narrative logic requires that we persuade the hero (and the reader) that the outcome might be exactly the opposite of that which, we are becoming increasingly certain, is inevitable. The more obvious it is from the start how the story will end, and the more obvious it becomes as the narrative progresses, the more the narrative must provide an insistent counter-argument in order to sustain our interest by sustaining the tension between that counter-argument and the epistemic necessity of the outcome, the more it must manifest a denial of the resolution that is coming.

So, where the authority of the uncanny is validated in Horror we have a pessimistic view in which, ultimately, the world itself is against the hero. The unacceptable truth is an intrinsic part of the underlying structure of the universe. Fate itself as part, of that Natural Order, is against the hero. So both hero and world must be strengthened by their conflict with each other, in order to raise the stakes with heightened drama in each encounter. The hero must become more entrenched, more defiant, even as the forces of the unacceptable truth are proven more and more authoritative in each encounter. So we have Clute's "thickening", the world mirroring the protagonist's increasing resistance.

And so, where the authority of the uncanny is negated in Heroic Fantasy we have an optimistic view in which, ultimately, the world itself is on the hero's side. The unacceptable lie is an extrinsic imposition upon the underlying structure of the universe. Fate itself, as part of the Natural Order, is on the hero's side. So both hero and world must be weakened by their conflict with the other, in order to raise the stakes by increasing their peril. The hero must be brought closer and closer to the very edge of defeat, the forces of the unacceptable lie made to seem more and more authoritative in each encounter. So we have Clute's "thinning", the world mirroring the protagonist's deteriorating resistance.

In both cases we can see these, I think, as logical responses to the epistemic necessity developing within the narrative, understanding Clute's grammars as emergent features of the dynamics of subjunctivities rather than as structural frameworks of plot / theme. From the dialectics of boulomaic modalities and epistemic necessities and of epistemic necessities and their counter-arguments, what we get is two potential paths of development through which we interpret a narrative as this genre or that rather than a narrative that essentially is this genre or that and that therefore simply has the basic architecture of that genre.

7. Restoration and Revelation

Following these narrative logics through to their resolutions, the outcomes for Heroic Fantasy and Horror in this model also map quite well to those in Clute's narrative grammar:

3) Recognition. The key in the gate; the escape from prison; amnesia dissipates like mist, the hero remembers his true name, the Fisher King walks, the Land greens. The locus classicus of Recognition is Leontes’s cry at the end of The Winter’s Tale (1610) on seeing Hermione reborn: “O she’s warm.”

3) Conceptual Breakthrough. Peter Nicholls’s term, from The Encyclopedia of Science Fiction (1979), for the thrust of release when a defective paradigm collapses and the new world — the true world — is revealed. A sense of wonder is often felt, sometimes in spaceships.

3) Revel. The story saves us. The rind of the world is peeled off, we see our true face in the mirror, carnival rules, what we see is what we get, the high are made low: where we belong. There is an almost infernal glee in learning the simple dreadful monistic clarity of the truth. Compared to the foggy parsimonious Marlow, Kurtz is pure glee.

Where Clute talks of "recognition" and "revel", I think "restoration" and "revelation" would be more apt, but otherwise his characterisations of the resolution stages of Heroic Fantasy and Horror fit quite nicely with a view in which these outcomes are understood as the authentication of epistemic necessity.

The end-point of Heroic Fantasy, driven as it is by the "should have happened" subjunctivity, is to restore us to the serenity of idyll, to make everything right again in the world, as it was in childhood, as it was in the past, with the land enchanted, the rightful king on his throne, and so on. So Clute describes this stage in terms of restored freedom ("the key in the gate"), restored memory ("amnesia dissipates"), restored identity ("his true name"), restored health ("the Fisher King walks") and restored world ("the Land greens"). Since the epistemic necessity being authenticated here is, to all intents and purposes, the victory of the "should have happened" boulomaic modality around which the world is initially constructed, it is only logical that it should be articulated in the symbolism of restoration, of return to the idyll.

The end-point of Horror, driven as it is by the "should not have happened" subjunctivity, is to reveal to us the authority of the uncanny, to strip away this illusion of control, the lies of reason, the lies of sanity, to force upon us the unacceptable truth of our own damnation. So Clute describes this stage in terms of discarded illusion ("the rind of the world is peeled off") and revealed truths ("what we see is what we get"), truths about who we really are ("our true face"), how morality is a sham ("carnival rules"), and how what seems noble is actually base ("the high are made low"). Since the epistemic necessity being authenticated here is, to all intents and purposes, the surrender of the "should not have happened" boulomaic modality which occludes the reality of the world in its denial of the uncanny, it is only logical that it should be articulated in the symbolism of revelation and the overthrow of orthodoxy.

These differences in resolution could be traced back to the idyllic basis of Heroic Fantasy, if we consider the whole narrative grammar as fulfilling a consolatory function. In this analysis the use of the "should not have happened" subjunctivity -- the use of the uncanny -- could be seen as simply part of a larger narrative strategy, a means to an end. It is simply that in order to exploit the "should have happened" subjunctivity fully we must create a sustained tension, a state of thwarted desire until the inevitable resolution. In so far as the idyll is a signifier to which a real world referent can be attached -- pre-industrial culture, medieval society, etc. -- we might be dubious of a restoration that is essentially an act of validation.

In fact, in so far as the uncanny is also a signifier to which a real world referent can be attached we might be even more dubious. Where we have an idyll under threat from the uncanny we would be well to ask ourselves if there's a conflicting paradigm being demonised here, what exactly it is about this paradigm that makes it transgressive, whether we should truly dread it, and whether we really support the validation of the idyll. There is a large extent to which the narrative grammar of Heroic Fantasy is not so different from the narrative grammar of reactionary propaganda.

This risks, however, de-emphasising the complementary artifice of Horror, where Heroic Fantasy's dubious consolation of our desires made flesh -- threatened, weakened, but in the end restored -- is replaced by the equally dubious conviction of defeatism, of our dreads made flesh -- defied, denied, but ultimately triumphant. In this analysis, the strategy is to exploit the "should not have happened" subjunctivity directly, playing other subjunctivities off against it to maintain a state of rampant dread until the resolution of revelation. And in so far as the uncanny is a signifier to which a real world referent can be attached -- the bestial amorality at the heart of human nature, the id unbound -- we might be equally dubious of the acts of validation carried out in the revelations of Horror.

In fact, in so far as the uncanny presents its revelation as intrinsically unacceptable, we would be well to ask ourselves if the demonisation of a paradigm which is held to be true as nevertheless innately transgressive, innately dreadful, innately horrific, is not perhaps a little pathological. In presenting the "truth" as morally abhorrent, reifying aesthetic and ethical boundaries in order to exploit them, Horror may well open itself up to critique as an endemically neurotic form manifesting the very restraints it is arguing against. In its bleak insistence on the inevitability of failure and the ultimate authority of the monstrous and malevolent -- supernatural or natural, divine or human, diabolic or brutal -- we might well see it as a literature of despair as much as a literature of dread, the depressive's self-fulfilling prophecy of defeat.

If this sounds like I'm damning both Heroic Fantasy and Horror for embodying the self-delusions of optimism and pessimism respectively, well, I do think these narrative modes require to be critiqued on those terms, but at the same time I think Clute is right to put his narrative grammars in the context of a Romantic "world storm", a response to the Rationalism of the Enlightenment, to characterise what I'm calling strange fiction and its relationship to mimetic fiction in terms of a return of the repressed. It seems to me quite natural to link a fiction founded on the strange and its exploitation of desire and dread to the Rationalist antipathy to passion and its attempt to control it. When it comes to reason and passion, when it comes to the psyche, to control means to repress, and to repress means to render it Other -- alien, weird, strange.

In a psycho-political sense then, we might well argue that the restoration that takes place in Heroic Fantasy is at heart an assertion of the validity of desire in and of itself, while the revelation that takes place in Horror is at heart an assertion of the validity of dread in and of itself, that each narrative grammar is, in its own way, re-enfranchising a Romantic aesthetic disempowered by Rationalism, reclaiming a place in the psyche for passion, avowing the import of the irrational, and allying itself to an aesthetic of the sublime that has been marginalised for over a century as a result of the Enlightenment's valorisation of reason.

Actually, I think there's a lot more to it even than this... but we'll come to that.

8. The Narrative Grammar(s) of the Cryptic

So far I've been dealing entirely with Heroic Fantasy and Horror. With SF we have a form in which boulomaic modalities take a more complex role than they do in Heroic Fantasy and Horror, but there are many points of similarity to be discerned, and I think it's important to note that -- from its origins in the pulps, through the Golden Age, and right up to the present day -- SF has often had a similar dynamic of subjunctivities at play within its narratives.

Many of the novum which have been used to establish the fictive environments of SF -- space travel, cyberspace, the singularity -- carry a "should have happened" level of subjunctivity just as much as those which construct the Heroic Fantasy idyll. The pastoral may be translated to the frontier idiom of the Western, scaled up and out into space, the desire may be expectant rather than nostalgic, for the promised land of the future rather than the lost paradise of the past, but this is still the stuff daydreams are made on. In fact, as yesterday's novum are conventionalised by generations of reuse, I'd argue, they have become equivalently folkloric -- comfortably familiar as genre tropes. The result is that much SF exhibits an equivalent base stage to that of Heroic Fantasy, in which the strange is initially used, for the most part, to establish an elsewhen of manifest desires that the narrative will subsequently disrupt with a second level of strangeness.

Similarly, many of the novum of SF which are introduced to carry out that act of disruption and thereby kick-start the narrative -- doomsday devices, artificial intelligence, nanotech -- carry a "should not have happened" level of subjunctivity of the sort we find in Horror in their capacity to destroy civilisation as we know it, cause the end of humanity if not life itself, unravel the very fabric of the spacetime continuum, and so on. When we talk of "sense of wonder" in SF we are talking of the same desire and dread we find in Heroic Fantasy and Horror, of futureshock and awe.

In constructing the elsewhen of the future rather than reconstructing the elsewhen of the past, in adopting hypothetical rather than metaphysical conceits, and in presenting the threat of the unknown through the imagery of technological advance rather than supernatural incursion, SF may give the illusion of offering something that is more Rationalist mechanics than Romantic magic and monstrosity, but behind that illusion we can often see, if we only look a little closer, the same dynamics of desire and dread at play. In fact I'd argue that it's not at all unusual for the SF narrative to play out from this in the same way as it does in Heroic Fantasy or Horror, to exploit the exact same dynamics of boulomaic modalities. Which is to say, Clute's narrative grammars of Heroic Fantasy and Horror can also be applied directly to much that is classified as SF.

In SF, of course, the introduction of epistemic necessity that I'd consider crucial to those narrative grammars is entirely optional. SF includes works that are clearly in the Heroic mode and works that are clearly in the Horror mode, but it also includes works with narrative grammars that would best be described as Noir/Thriller or Mystery/Adventure. In these narrative grammars the world itself is riven with the tension between subjunctivities of "should have happened" and "should not have happened". The uncanny horror of crime is an integral part of the urban environment, but in its very seediness the underworld is also wonderland, the cityscape of Chandleresque Noir/Thriller fiction shot through with the same romanticism we find in vampire fiction, with a sense of the city at night as dark idyll, as decadent's playground, embodying our adult desires for sex and power. In Buchanesque Mystery/Adventure fiction, where the backdrop is not limited to the dark and urban but ventures out to far-flung exotic locales, the flavour is more orientalist than decadent, but the horror is scaled up in line with this, war replacing crime as the all-too human horror.

What these narrative grammars have in common is that in their worlds built from (broken, human) dreams, where the "small sour lesions" that disrupt them are understood as part and parcel of human nature, there is no fate to be for or against the hero, and so there can be no epistemic necessity, no certainty of the outcome. The Fisher Kings are phonies and the Carnival a scam; they may even masquerade as each other. The hero is on his own, with his own strengths and weaknesses, as any human being is, drawn by pure chance into the games of these Powers That (Would) Be. So here the strange manifests as the cryptic, as conspiracy and intrigue, at the root of which is that very doubt as to what those games are, who exactly is playing them, what they are playing for, and how it will all pan out.

It is not too hard to see, I think, where the downbeat idiom of Noir shades into the pessimistic certainty of Horror, where the upbeat idiom of Adventure shades into the optimistic certainty of Heroic Fantasy, to see how all four narrative grammars fuse into a single spectrum, but this should only enrich our understanding of the processes at work here, allowing us to better see the extent to which these different narrative grammars are fundamentally permutations of a single grammar of narrative, and offering us a framework for analysis . At the extremes of that spectrum we have that fiction where epistemic necessity is at its strongest and where counteracting tensions must be introduced to compensate. At the centre-point, meanwhile, in a region we might well call Thriller/Mystery, is that fiction where epistemic necessity is at its weakest, where we are least sure of the outcome and where uncertainty itself is sufficient to drive the narrative on.

What interests me in these narrative grammars of Noir/Thriller and Mystery/Adventure is that they are, I think, undeniably recogniseable in SF, with all technothrillers, much of cyberpunk and many contemporary Space Operas exemplifying that exploitation of the cryptic. Taken together, moreover, as Thriller/Mystery, they might suggest to us a second stage more comparable to the "thinning" and "thickening" Clute associates with Heroic Fantasy and Horror than the abstract "cognitive estrangement" by which he characterises SF. Here, to use Clute's metaphor, the process might be thought of as a struggle through inconsistency, a struggle to attain equilibrium. In this Stage Two the hero moves on in fits and starts, bogged down in the lie here, staggering out into the truth there, alternately lost in amnesia and dazed by illumination. The problem is not that the world is against him and he refuses to recognise it (the thickening of Horror), nor that he is too weakened by the fight to believe that the world is with him (the thinning of Heroic Fantasy). The problem is that half the world is for him and half the world is against him and he does not know which is which.

And the end result? The end result may be restoration or revelation, or both or neither. Neither is a dangerous and difficult outcome for a writer to try and pull off, to have the hero fail in the end, but I'm sure it can and has been done, though I suspect in most cases there is some sense of restoration and/or revelation for the reader even if not for the hero. When the narrative is driven by intrigue, by a sense of the cryptic, a lack of solution, for the reader, is likely to be experienced as a lack of resolution.

9. The Narrative Grammar of the Conceptual

All of this, I think, illustrates the key problem with Clute's schema, highlighting the implicit narrowing of focus in his labelling. With Fantasy the focus is narrowed to spotlight the Romantic mode of Heroic Fantasy and exclude modes of fantasy with other narrative grammars. But in the vampire fiction of Rice and Brite, and in other works classed as Dark Fantasy or Urban Fantasy -- the works of writers like Neil Gaiman, for example -- the very reason they are often thought of and talked of as some sort of hybrid of Fantasy and Horror is that they are essentially utilising the narrative grammar(s) of Noir/Thriller/Mystery/Adventure as freely as SF does. Regardless of the label slapped on them -- Fantasy or Horror -- they are not bound to the narrative grammars of Heroic Fantasy or Horror as outlined by Clute. Writers like Gaiman stand, I think, as evidence of a narrative grammar in Fantasy that is at odds with the one Clute labels "Fantasy" and identical to the narrative grammar to be found in much SF. I'm sure the same can also be said of Horror.

With Clute's narrative grammar of SF, meanwhile, the focus seems to be similarly narrowed to spotlight that mode where it is most "literary" in its Rationalism, excluding the Romantic modes where it is essentially, in terms of narrative grammar, equivalent to Heroic Fantasy or Horror. In the abstraction of the characterisation -- in terms of novum, cognitive estrangement and conceptual breakthrough -- I'm not sure I even see any recognition of the importance of the narrative grammar(s) of Noir/Thriller/Mystery/Adventure which I would argue are blatantly obvious in the field. I think there's an implicit inequity in approach here which smacks of the tiresome privileging of SF over Fantasy, whereby a blind eye is turned to generic Romanticism in the former and to everything but the most generic Romanticism in the latter.

It is not unfair and not inaccurate to distinguish the SF narrative by the more palpable cognitive dissonance at the heart of the novum. And in so far as strangeness in the form of novelty is not intrinsically valuated as positive or negative, does not automatically accrue a boulomaic modality of "should have happened" or "should not have happened", it is by no means unfair or inaccurate to say that the SF narrative is capable of exhibiting an entirely different narrative grammar to any of those outlined above. The novum of SF, in this view, may not present us with the emotional crisis of an idyll under threat and/or an irruption of the uncanny, and where they do not we should not expect to find the narrative grammar of Heroic Fantasy or Horror. They may not (even if they often do) function as the mysteries and MacGuffins at the heart of the narrative grammar(s) of Noir/Thriller/Mystery/Adventure. But the same can be said of Fantasy if we simply refuse to relabel the "literary" fantasies of VanderMeer, Ford, Link, Joyce, etc., as magical realism or fabulism or postmodernism or slipstream or whatever. And why should we if we are not doing the same for an SF which fits the narrative grammars of generic Romanticism as often as not?

It is, I would say, entirely possible for the SF narrative trigger to be situated wholly in the novum, in the degree of measurable difference from our "given world", for this to generate a wholly conceptual conflict, a conflict of nomologies, played out through cognitive estrangement to conceptual breakthrough. The construction of an SF elsewhen out of novum can be an intellectual ("arguable" and "structured" -- i.e. Rationalist) reimagining of normative reality rather than an emotional (irrational and unstructured -- i.e. Romantic) act of fabrication or disruption. SF can and often does function as analysis through conceit, the future re-presenting the present, defamiliarising it in order to expose its faults, to allow "the defectiveness of the ruling paradigm to be seen whole". The narrative grammar at work in SF can be the articulation of an argument between the ruling paradigm of the world as is -- or as we think it is -- and an alternative paradigm of the world as it is reconstructed by and filtered through a matrix of conceits, extended and concretised metaphors, the points of difference highlighting, figuratively or by exaggeration, the precise defects of that ruling paradigm. But the same can be said of Fantasy if we look at the work of Peake, say, where the elsewhen of Gormenghast is perhaps the best example I can imagine of a defamiliarising re-presentation, brutally exposing the deficiencies of the ruling paradigm of British society in the early 20th Century through the conceit of the "Big House".

The resolution of SF can be and often is the conceptual breakthrough at the completion of this argument, the synthesis of antithetical nomologies as we come to fully understand the alternative paradigm and relate it back to the ruling paradigm, understanding the re-presentation and understanding that it is a re-presentation. We might well call this stage realisation, for it is the point where the last piece of the puzzle is put into place, the conceit finally fully realised, and it is also the point where the hero (and the reader) grasp the reality of what is going on, realise the root of the struggle that has shaped the narrative. But again the same can be said of Fantasy.

There is one point of difference, within this narrative grammar, between some SF and Fantasy. There is a subset of SF -- dystopian fiction such as Orwell's NINETEEN EIGHTY-FOUR, for example -- where the extrapolative novum read as caricature rather than metaphor, capable of being read as literal predictions even if they are not meant as such but rather as exaggerations designed to focus our attention on the real defects they represent, defects neglected in the ruling paradigm. So the world of Big Brother, Newspeak and so on caricatures the surveillance and control culture of Stalinist society in particular and totalitarianism in general. And so these novum can be read as either symbolic warnings or literal predictions of what might be.

The plausibility of the novum is crucial in this type of SF because the argument being made in these narratives is that the ruling paradigm, the accepted nomology in which the events of the narrative "could not have happened", is false, and that the hypothetical nomology, the alternative paradigm in which the events "could have happened", is in a very real sense, a truth of temporal possibilities. In the straightforward polemics of dystopian fiction, the ruling paradigm is the complacent thesis that "it couldn't happen here and now", the alternative paradigm is the antithesis where an absolute dystopia is a very real possibility in the future, and the synthesis is the revised paradigm in which we recognise the real possibilities of the future represented by the novum because we recognise a part of them is possible here and now. It is the conceptual breakthrough where we realise that, in reality, there is a Winston Smith in us that could well be made to believe that two plus two equals five.

But with only a minority of SF's novum, I'd argue, meant to be read as literal prediction, the majority of these hypothetical conceits meant to be read as figuratively as the metaphysical conceits of Fantasy, I'm not convinced that plausibility has much impact on the narrative grammar. Whether the strangeness takes the form of a novel and plausible conceit or simply that of a conceit has little effect on its capacity for re-presenting the world. Through the looking glass or through a scanner darkly, the effect is still a deliberate, sustained and coherent estrangement designed to re-present the world in a warping lens and to lead us to a resolution of realisation.

Ultimately then, it seems to me that this narrative grammar can be found in both Fantasy and SF, and as such to call it a narrative grammar of SF is a misnomer. In fact, how far it actually even constitutes a narrative grammar when compared to those outlined above seems somewhat arguable. Stripping away the aspect of defamiliarisation as a means of commentary on reality, how much does it really tell us about the narrative over and above a basic model where equilibrium is disturbed, struggled for, and finally re-established in a revised form? If we take away the distinctive feature of the strange (novum or otherwise) used as extended and concretised metaphor to integrate plot and theme, do we have anything more than a model of narrative itself? Couldn't pretty much any plot be fitted into that narrative grammar?

Which is, I would argue, exactly how it should be for a type of fiction which is being identified entirely by its conceptuality. Because ultimately I would argue the real meaning of the term "literary" as it is applied to SF, Fantasy and Horror is just that -- conceptual. It indicates that the strange is being exploited as a conceit rather than for the affect that it produces -- idyllic, uncanny, cryptic -- that the narrative is not therefore bound within the narrative grammars of Romanticism but rather has adopted a Rationalist narrative logic of realism. Or that where it is working with the Romantic grammars it is simultaneously subverting them through that conceit and/or the narrative logic of realism, creating a hybrid form that is essentially Modernist or Post-Modernist.

In essence then, I think Clute's schema of narrative grammars is flawed in where it places its labels and in what it disregards because of that, but I think it opens up two distinct (and distinctly useful) lines of approach to a text, one focused on the conceptual use of the strange in all of three "genres" of fiction, the other focused on a set of narrative grammars crudely understandable as a sort of spectrum running from Horror through Noir/Thriller and Mystery/Adventure to Heroic Fantasy but open to a more sophisticated view in so far as those genres are viewed as superficial interpretations of an underlying dynamics of subjunctivities, boulomaic modalities and epistemic necessities.