1. Introduction
In this article, I intend to provide a new definition of a significant subset of fantasy literature. Fantasy should be regarded as a literary order which conjectures “impossible” worlds that are regulated by a subjective principle. It presents these to readerships as catalysts for the development of new structures of feeling. In motivating my reader’s acceptance of this proposal, my main argumentative strategy will be to show that this is an implication of thinking about fantasy as belonging to the “literature of the impossible”. Although my aim is to deepen the implications of that critical approach, I nonetheless do wish to pluralize the impossible and to challenge a prevalent understanding of science fiction in the process. Once we see that both some science fiction and much fantasy literature belong to the same class, a natural question is whether anything can serve as a criterion of differentiation capable of distinguishing between the various literatures—plural—of the impossible. The difference between subjective and objective principles provides just such a criterion, but it also points towards a functional difference between much fantasy and some sci-fi. If some science fiction conjectures imaginary objectivities—impossible laws of a new physics—in order to present a cognitive novum, is it not possible that much fantasy literature, which speculates about imaginary subjectivities—impossible desires, beliefs, feelings or needs—does so to present an affective novum? Magic, after all, is often a way to represent subjective states as if they were directly effective material powers. Rather than dismissing this as “mere wish-fulfilment”, however, it should be grasped as a symbolic expression for desires, beliefs, feelings or needs that are presently “unthinkable”, “illegitimate”, “deviant” or “forbidden”—in short, impossible.
Before entering into the detail of the argument, let me foreshadow its basic strategy while outlining the structure of this paper. The kind of fantasy literature that I want to discuss can be uncontroversially illustrated using a few uncontentiously paradigmatic texts, which illuminate why this is, in general, a kind of literature where magic works or demons exist. That feature has led to the emergence of a widening agreement in the critical field of fantasy studies that fantasy is “the” literature of the impossible. What is meant by that is that something impossible in the real world is represented in the imaginary universe as possible, normally in a fictional context where some special magic operates. That is equivalent to the modal definition of fantasy literature as a type of writing where the marvellous mode is a dominant structure. Discussions of fantasy often contrast it with science fiction, thought to be a literature of the possible, and with realist fiction, manifestly a literature written in the mimetic mode. A moment’s consideration of the impossible/marvellous connection, however, reveals an entire body of speculative fictions, such as nonsense literature, literatures of paradox, utopian speculations, science fictions which do involve the impossible, and so forth, alongside this significant body of fantasy literature. Fantasy is not “the” literature of the impossible, but merely “a” literature of the impossible. Furthermore, while many science fictions may deal with realistic possibilities, some paradigmatic ones represent real-world impossibilities as imaginary-universe possibilities. Science fiction, or some important part of it, is also “a” literature of the impossible. What, then, differentiates fantasy literature from science fiction?
I propose that we should look to the nature of the referential domain constructed by the fictional text, in terms of the objective, subjective, logical or social principle that regulates its imaginary universe. Some science fiction conjectures an imaginary universe regulated by a law of physics that is presently impossible in the real world, but possible in the imaginary universe (e.g., FTL travel, personality uploads, superintelligent AIs). Much fantasy literature conjectures an imaginary universe regulated by a principle of subjectivity that is considered impossible in the real world, but possible in the imaginary universe (e.g., emotions directly affect the material world through magic, a divinity intervenes in natural reality in response to its concerns, imaginative unruliness directly manifests as the caprice of faery). If this suggestion is allowed, there are some interesting consequences. Science fiction is rightly thought of as involving a “cognitive novum”, that is, it conjecturally introduces some new objective principle into the real world to produce its imaginary universe, and this provokes reflection on the nature of objective reality. Application of the same logic to the idea of an imaginary universe regulated by a subjective principle yields the suggestion that fantasy involves an “affective novum”. The point of adding a subjective principle to the real world is to conjecture what life would be like were this aspect of subjectivity not considered impossible. For example, JK Rowling’s Harry Potter series is a conjecture about how there would be less anxiety, and therefore less violence, were individuals reconciled to the acceptance of mortality and tranquil about the reality of their own death. For example, JRR Tolkein’s The Lord of the Rings is a speculation about how affirmation of trust in divine providence is a crucial part of the fortitude that makes resistance to domination possible, just as the proud assertion of mastery is the root of domineering attitudes. Thus, by “subjective principle” I do not mean an imaginary universe regulated by arbitrary caprice, or authorial fiat. I mean, the author takes something from the subjective world—that realm of mental experience to which the individual percipient has privileged access—but something that is regarded by society as “impossible” (because, for instance, forbidden)—and conjectures a universe within which this feeling becomes a causal power (symbolizing this as “magic”, for instance, or “demons”). The “affective novum” offers a new insight into the subjective world of modern individuals, something which is proposed by the fantasy author as an enrichment of the domain of needs and desires, beliefs and feelings.
2. Describing Fantasy
By way of introduction to this discussion, I am going to start by illustrating what I mean by the particular subset of fantasy literature that I intend to speak about and clarifying the basis for the claim that it is already widely considered a “literature of the impossible”.
Fantasy literature as a whole is a kind of literature which includes both “literary” and “popular” strands, and which generally involves magical powers, supernatural beings and enchanted objects. Its twentieth-century manifestations are illustrated by JRR Tolkein’s
The Lord of the Rings trilogy (1954–1955) and Fritz Leiber’s
Fafhrd and the Gray Mouser series (1958–1970). Together, these represent major, “high fantasy” and “low fantasy” (respectively), lineages within late-modern fantasy literature, where I use those terms describe the heroic or satiric tenor of the texts, not whether or not they conjure a complete secondary world. Probably the high fantasy, “swords and sorcery” variety is the one best known to contemporary readers, mainly as a result of the post-1960s Tolkein boom (
James and Mendlesohn 2009). Enthusiasm for Tolkein on US college campuses during the Vietnam War inspired dozens of imitators, who combined “Tolkeinesque” conventions with “low fantasy” inspired by, for instance, Robert E. Howard, Fritz Leiber, Jack Vance, Lin Carter and Michael Moorcock. The period from 1970 through to 2010 witnessed a kind of generic institutionalization of Tolkeinesque fantasy, culminating in Robert Jordan’s 14-book tetradecalogy,
Wheel of Time (1990–2013). Other Tolkein imitators include Terry Brooks (
Shanarra series), Terry Goodkind (
Sword of Truth series), Stephen Donaldson (
Chronicles of Thomas Covenant series), Margaret Weiss (
Dragonlance series), David Eddings (
Belgariad series) and Anne McCaffrey (
Dragonriders of Pern series). More recently, George RR Martin’s incomplete
A Song of Ice and Fire (1996–), popularized through HBO’s
Game of Thrones TV series (2011–2019), led to Martin being described as “the American Tolkein”, indicating the continued influence of the Tolkeinesque. Fantasy literature used to be described as “swords and sorcery”, or “wizardry and wild romance” (
Manlove 1975;
Moorcock 2004). But that is no longer completely accurate, as fantasy novels have increasingly moved away from medieval settings and stereotypical magic, with what I have called the Tolkeinesque receding in prominence. In the twenty-first century, at last providing an influential high-fantasy alternative to Tolkein’s legacy, JK Rowling’s
Harry Potter series (1997–2007), a short, 7-book septalogy, can be regarded as indicative of fantasy today. Other celebrated contemporary fantasy writers include Brandon Sanderson, NK Jemisin, Philip Pullman, and Lois McMaster Bujold, as well as Robin Hobb, Cassandra Clare, China Miéville and Susanna Clark. Emergent fantasy genres now include, alongside “high fantasy”, “low fantasy” and “weird tales”, urban fantasy, grimdark, mythpunk, paranormal romance and the “new weird” (
James and Mendlesohn 2012). Additionally, under the inclusive description that I provide in this article, fantasy literature would also include horror fiction, when this unambiguously contains supernatural elements. Stephanie Meyer’s
Twilight (2005–2020) series and Anne Rice’s
Vampire Chronicles (1976–2018) would qualify, for instance.
The most important development in the last three decades of studies in fantasy has been the emergence of a widening agreement that fantasy literature belongs to the “literature of the impossible”. Edward James and Farah Mendlesohn report that “the major theorists in the field—Tzvetan Todorov, Rosemary Jackson, Kathryn Hume, WR Irwin and Colin Manlove—all agree that fantasy is about the construction of the impossible, whereas science fiction may be about the unlikely, but is grounded in the scientifically possible” (
James and Mendlesohn 2012, p. 1). That convergence has been building for the past fifty years, since definitions of fantasy as a literature of the impossible were first proposed in the 1970s by Colin Manlove (
Manlove 1975, p. 1), John Irwin (
Irwin 1976, p. 9) and Eric Rabkin (
Rabkin 1976, p. 12). In the 1980s, following the rise of structuralism and the influence of Tzvetan Todorov’s work on the fantastic as a mode (
Todorov 1975), theoretical research continued to gravitate towards the impossible as a description of what that mode represented (
Brooke-Rose 1981. pp. 77–78;
Jackson 1981, p. 80). Authoritative interventions in the critical field by John Attebery (
Attebery 1992, p. 7) and John Clute (
Clute 1997, pp. 335–337) consolidated this definition into a standard reference; Clute’s entry in
The Encyclopedia of Fantasy states unequivocally that “a fantasy text is a self-coherent narrative [that] tells a story which is impossible in the world as we perceive it” (
Clute 1997, p. 336). Indicating just how well accepted in the field this definition is, Farah Mendlesohn’s and Edward James’s recent, highly influential survey,
A Short History of Fantasy, casually reports that “the most obvious construction of fantasy in literature and art is the presence of the impossible”, before moving on to develop specific considerations (
Mendlesohn and James 2012, p. 3). In this article, I will describe this emergent agreement as a “potential consensus”, while noting that every consensus is susceptible to contestation.
The potential consensus on fantasy literature as belonging to the literature of the impossible is really quite striking, for two important reasons.
The first of these is that this is a convergence arrived at from startlingly different directions. Leading approaches to fantasy are modal, rhetorical or generic. I discuss these approaches in some more detail in a moment. But let me notice right away that from the epistemological perspective, since every theory constitutes its object domain by describing the phenomena in particular ways, it is rare to find such agreement. Really, we should not expect modal and rhetorical approaches to describe any given text in the same way, or even to agree that the text in question qualifies as fantasy. And indeed, in relation to boundary conditions and borderline cases, there is wide disagreement, not monumental concord, with, for instance, the texts in Mendlesohn’s “liminal” and “irregular” categories of fantasy (
Mendlesohn 2008, pp. 182–272) being exactly the sorts of texts that populate Jackson’s “fantastic” category, which is an order of texts she thinks entirely distinct from fantasy literature written in the marvellous mode (
Jackson 1981, p. 33). But that just makes the existence of a very large set of texts on which there exists a potential consensus—in the example given, for instance, Mendlesohn’s capacious “portal”, “intrusion” and “immersion” categories, which are populated with the sort of texts that fill up Jackson’s “marvellous”—all the more remarkable.
The second of these springs from the first, namely, that this region of agreement can be satisfyingly described a “fuzzy set”, defined around paradigmatic, or “taproot”, texts (
Clute 1997, pp. 336, 921–922). Attebery’s original proposal, in this respect, describes the constitution of these sets through processes of exemplification and similarity that define family resemblances (
Attebery 1992, pp. 12–14). Without for a moment contesting the idea that what counts as fantasy blurs at the boundaries into science fiction, horror literature and paranormal romance (for instance), Mendlesohn has since proposed to pluralize this into several fuzzy sets of fantasy literature (
Mendlesohn 2008, p. xvii). But here is what is remarkable in this context. Even though distinct critical approaches define their objects differently, in effect, most of them have managed to agree on the existence of what mathematicians call a “core set” of “clearly-in” members of the fuzzy set/s of fantasy literature. So, even when the taproot texts defining modal and rhetorical approaches are entirely different,
nonetheless, very many critics agree that whatever it is that is in their fuzzy sets, its core is defined by the representation of the impossible.
The notion that fantasy literature represents impossibilities, in combination with the idea that this potentially defines a core set of clearly-in fantasy texts, with the proviso that what is involved is the logic of fuzzy sets, makes it possible to develop some generalisations about fantasy.
3. Fantasy, Magic and Impossibilities
Let me now briefly exhibit the connection between impossibility, fantasy and magic (or demons) that the discussion in this article develops, by way of anticipation of the detailed argument that follows.
Describing fantasy as belonging to the literature of the impossible highlights its imaginative character as a utopian literature of desire (or a dystopian literature of dread). This lets us focus on the politics of fantasy, because a utopian (or dystopian) literature that expresses prohibited impulses (or that suppresses legitimate desires) represents a negation of existing social and cultural arrangements (i.e., a critique of them) (
Jackson 1981). Additionally, fantasy can be read as a suggestion about alternative arrangements (i.e., a utopian vision), involving catalysts for reform, revised perspectives on the world, and consolatory or disturbing illusions (
Hume 1984). That does not mean that the critique is necessarily valid or that the utopia is automatically worthwhile—after all, some impulses, such as those resulting in incest and murder, are anti-social and rightly prohibited (
Jackson 1981, p. 78)—but it does mean that the literary form has considerable critical potential.
Now, by “impossible” I mean, with Hume, a deliberate deviation from the consensus definition of reality (
Hume 1984, p. 21), something which in modernity is dominated by scientific conceptions of nature and technological applications of the natural laws it discovers. Some fictions, such as Gene Wolfe’s
Latro in the Mist (2003 [1986, 1989]), which imaginatively explores the ancient world in which reality was perceived as spiritually populated, are therefore not necessarily representations of the impossible. Conversely, there are many novels which present worlds inspired by indigenous worldviews which are, because of their defiant refusal to inhabit the real world defined by science, fantasy critiques of modernity’s limitations that deploy the impossible critically. The implication of this definition is that we should think about fantasy literature as belonging to a literary mode, that is, a broad class of non-realist writing generally known as the “marvellous”, and specify this in terms of a representation of impossibilities of a certain kind.
The imaginary universes of the core set of fantasy literature are ones where something like magic works and demons exist. Indeed, I take it that this is the entire point of the literary order, as well as the defining characteristic of its core set. By “magic” and “demons”, I mean any causal power in the imaginary universe that represents something excessive to natural causation in the real world, something which would have to be explained under today’s consensus definitions of reality as “supernatural” or “metaphysical”. Usually, as in
Harry Potter, “magic” works in the imaginary universe just as it would were it possible in the real world: magic in the imaginary universe is a supernatural source of causal power, directed by thinking or feeling and operationalized by some symbol (spell, rune, word, icon). Sometimes, however, as in
Lord of the Rings, “magic” in the imaginary universe is a providential dispensation, or a physical manifestation of spiritual power. This is the result of thinking by some deity, demon or angel (or cognate being) and it is operationalized by self-expression, the revelation, in the natural world, of a supernatural essence “as it is upon the other side” (
Tolkien 2004, p. 217). In both cases, there is a causal power in the imaginary universe that is obedient to thinking or feeling by its wielder, one which depends on the mastery of subjective capacities for its exercise.
Magic should not be conflated with science—Arthur C. Clarke’s remark that “any sufficiently advanced technology is indistinguishable from magic” notwithstanding (
Clarke 1973, p. 36).
Certainly, both magic and advanced technology, which work in imaginary universes, are impossible in the real world (under current consensus definitions of reality). But one often happens because something in the subjective domain of the imaginary universe works differently there than in the real world, while the other happens because something in the objective domain works differently. The wonders of (imaginary) technology depend on a speculative universe within which natural causes, impossible in the real world, can be harnessed, by virtue of some conjectured knowledge of the objective domain. By contrast, the marvels of magic depend on a speculative universe within which thinking and feeling can be harnessed by virtue of arcane symbols or supernatural essences. In Harry Potter, for instance, the success of the lethal spell “avada kedavra” depends on truly wishing your opponent dead. By contrast, in Star Trek, the success of the warp drive depends on the discovery of “supraluminal folding”, not on Captain Kirk’s heartfelt desire to escape the Klingons. Magic in the imaginary universe is often a representation of something impossible in the real world, which is that desires, feelings, beliefs or willpower are directly physically effective as causal powers excessive to nature. In many fantasy worlds, then, desires, feelings, beliefs or willpower are directly effective in the imaginary universe—generally, as “magic”. This expresses the idea of a world regulated by subjectivity—whether this subjectivity is that of the magician or the deity.
There is an important qualification that I need to add here. Although I have spoken about the core set of fantasy literature as defined by the representation of the impossible, I now wish to speak only about a subset of the core. That subset is the subset of fantasy where the imaginary universe represents the impossible as regulated by a principle of subjectivity, as exemplified, by (
Tolkien 2001), Rowling and (as we shall see) Brandon Sanderson and NK Jemisin. Granted, there exist fantasy universes that are also clearly-in which are not regulated by a principle of subjectivity, but rather by some other principle, which is not a principle of objectivity, normativity or logic (as discussed in a moment). Two big categories can be identified straight away: (1) imaginary universes in which the regulating principle is aesthetic, such as the worlds of Gene Wolfe, Ursula Le Guin and Samuel Delaney; (2) imaginary universes in which what seem supernatural powers are the result of a conjectural nature which is entirely objective, but difficult to describe as an alternative physics. Category (2) might include ones with psionic powers, e.g., EE Doc Smith’s
Lensman series, or Frank Herbert’s
Dune series, or seeming magic that may in fact be psionics, such as R. Scott Bakker’s
Prince of Nothing and
Aspect-Emperor series. I do not have a good theory of these categories, and so, beyond noting that they are clearly-in but rather different to the subset I wish to talk about, I cannot say anything further about them in this article.
Rather different is the case where the problem of meaningfully discussing fantasy is created by trying to retrofit definitions to include texts because they have been decreed to be fantasy (or science fiction). This particularly happens, I think, when the texts in question belong to the literary fantastic, not to the core set of fantasy literature or science fiction. It is not entirely obvious to me that Jackson’s “fantastic” (a.k.a. gothic fiction, e.g., Maturin’s Melmoth the Wanderer) and Mendlesohn’s “liminal” (centered on Mervyn Peake and Hope Mirlees) belong to the core set of clearly-in fantasy texts. That is significant, because these texts do not conspicuously or unambiguously include magic or demons, and so might be thought a counter-example to the thesis I am proposing. Jackson is explicit about this, excluding from the category of the fantastic anything that violates its “epistemological hesitation” statute, which is that natural and supernatural explanations of the action must remain both ontologically undecidable and mutually exclusive. Here, “magic” and “demons” are both actual possibilities and mere illusions, which means that the text represents a contradiction (which is why Jackson, rightly, does not foreground the impossible or claim that these belong to the marvellous). Peake’s Titus Groan (2007 [1946]), for instance, relegates the gothic to a pervasive ambience in which what is impossible is the utter isolation of the castle, not supernatural powers and entities. That is marked out by textual uncertainties regarding the distance of the mountain from the castle and the possibility that the entire region, including the mountain, are sometimes subject to flooding. Here too, other kinds of impossibility (social kinds: it is a dystopia of oppressive routinization and insular isolation) are in play than the kinds which generate the core set of fantasy literature. Mirlees’ Lud-in-the-Mist (2018 [1928]) seems exuberantly marvellous, until we realize that the entire action is susceptible to a reading according to which it is a symbolic allegory that can be decoded with reference to a combination of anthropology with psychoanalysis. This too is not clearly-in, which is why Mendlesohn, rightly, describes the category as “liminal”, i.e., belonging to the fuzzy set of fantasy literature, but somewhere at the edge, not at the center.
I suspect that this problem happens a lot because we lack a differentiated language with which to describe literatures which represent impossibility. We need, in other words, to pluralize impossibility. It is to this that I now turn.
4. A Literature of the Impossible
The concept of a “literature of the impossible” involves the relation of representation. What I mean by this is that something that is impossible in the real world is represented as possible in the imaginary universe. Focusing a description of the core set of fantasy around the representation of the impossible entails conceptualizing the field as belonging to a specific kind of writing, that is, as belonging to the literary mode known as the “marvellous”. In this section, I briefly define the marvellous, before noticing that this is not the only way that fantasy literature might be construed.
Thinking about the core set of fantasy literature in terms of the representation of the impossible means constituting it as an object of inquiry through a modal definition of narrative classes as analytically distinct representational kinds. From this perspective, narrative fictions can be thought of as resemblances to reality, grouped into classificatory categories that depend on the powers of action of the protagonist in relation to nature and society (
Hume 1984, pp. 1–52). According to this familiar principle of classification, abstracting from their social status and thinking only about their powers of action in relation to nature, protagonists can be greater than nature in kind, or in degree, they can be embedded in nature as agents, or subjected to nature (
Frye 2000, p. 51). These powers of action result in the “mythic”, “marvellous”, “mimetic” and “ironic” modes, respectively, with the “fantastic” mode (characteristic of gothic fiction) emerging as a “mixed” mode where the mimetic and the marvellous are in equilibrium (
Jackson 1981).
As the existence of the fantastic illustrates, all fantasy texts include ironic, mimetic, marvellous and mythic elements. The question is, which mode takes primacy in arranging the others? That is, technically, the question of which mode is the “structure in dominance” in the “complex totality” of narrative levels (
Jameson 1981;
2005). When a fantasy is an agent-centered, plot-driven narrative of a recognizable type (adventure, tragedy, comedy, satire), then the answer is easy. The dominant mode is the one that structures causation in the main plotline, and when this involves wielding extraordinary causal powers, ones impossible in the real world, then it is a marvellous narrative. In the core subset of fantasy I intend to consider, this plotline, by the way, almost always reduces to the acquisition of magical powers by the protagonist through a process of trials and tests, culminating in a triumph over adversity and the recognition of the specialness of the hero/-ine. But that is an issue of the type of plot (
ethos), not of its literary mode (
mythos) (
Denham 1978, pp. 59–131). To say, then, that the core subset of fantasy literature is narrative fiction in the marvellous mode is not to say that it is the only sort of fiction written in this mode or that fantasy texts consist exclusively of marvellous representations. But they should be considered written in the marvellous mode insofar as their major plot developments depend on supernatural causation, i.e., the protagonist wields an extraordinary power of action (“magic”).
Now, other specifications of fantasy are certainly legitimate. However, it is pointless to compare the modal approach directly with other approaches. Alternative definitions of fantasy literature not only describe the boundaries between fantasy and adjacent fictions differently—they populate the object domain with completely different kinds of things. Rhetorical approaches to fantasy literature concentrate on the narrative strategies that create effects of wonder and terror in the readership community, through the specific device of creating a threshold between ontologically different worlds (
Attebery 1992). What appears in the space of fantasy literature through this lens is not hierarchical arrangements of modes of representation, but tropologies such as “wrongness”, “thinning”, “recognition” and “return” (
Mendlesohn 2008, p. xv). Generic approaches populate the space of fantasy with conventions that are recognizable to readerships (
Manlove 1975;
Attebery 1980;
Siebers 1984;
Kroeber 1988;
Moorcock 2004;
Mathews 2011). These are customary ensembles of topics, typologies, and technics, which inform—as all forms of recognition do—judgments that something resembles a paradigm case sufficiently well to merit inclusion in a “fuzzy set”. Most of the confusion in debates about “what is fantasy?” arises because objects that seem to be recognizable as fantasies (generic definition) are held up for inspection from rhetorical or modal perspectives. Not surprisingly, they fail to satisfy the definition of the things that should populate the object domain, and the result is the critic’s despair, as the category of fantasy then seems to include only texts that defy classification. But that is the wrong way to go about it. From a rhetorical perspective, things are “fantasies” if and only if they are narrative strategies that cause wonder and terror through specific threshold-violating tropologies of “wrongness”, “thinning”, “recognition” and “return”. If, say,
Perdido Street Station fails to satisfy that definition, then two things are true: it is not a fantasy from the rhetorical perspective, and a generic or modal approach might be a better way to study that text.
Rhetorical and generic approaches are substantively distinct from modal approaches, whereas many seemingly different, “brand name”, methodologies are just a variety of interpretive foci overlaid upon modal, rhetorical or generic assumptions. Most psychoanalytically-influenced commentary, for instance, departs noiselessly from modal or rhetorical definitions before posing the question of the psychological significance of representations of the impossible (
Klinger 1971;
Manlove 1983;
Palumbo 1986) or of the pleasures of reading fantasy topics (
Freeman 2000;
Mathews 2011). The “secondary worlds” approach, inspired by Tolkein (
Swinfen 1984), but recently developed in contexts of psychology (
Oziewicz 2008) and possible worlds theory (
Traill 1996), is straightforwardly modal, because its question is not about the effect on the reader of threshold violation but the mythological or ontological implications of kinds of narrative layering. Historical materialist criticism, like psychoanalysis, tends more to read for the plot than its effects, and does so by interpreting ontological hierarchies as material (rather than psychological) symptoms (
Kroeber 1988;
Monleón 2016). Of course, there is also a fair bit of criticism that is methodologically blurry, but I do not discuss it here because its relevance to this question is doubtful.
5. The Rationality of the Impossible
I now want to develop a relativistic understanding of what is “impossible”, and then explain why I think that there are different types of impossibility. From the modal perspective, in both fantasy literature and science fiction, what is impossible in the real world becomes merely extraordinary in the imaginary universe. “Magic” works. “Advanced technology” exists. But it is rare. Even when magic is common, Magick is rare. Even when superintelligent AIs and supraluminal hyperspace travel are ho-hum, the Technological Singularity of the Excession is rare. Etcetera. The term “extraordinary” indicates that what is improbable (rare) is not entirely implausible (incredible), because the imaginary universe is structured so that the improbable seems credible. Why is it so? Because magic (and technology) follows the speculative rules of the imaginary universe. It is not plausible in real-world terms, because it is impossible in reality, but it seems plausible in the imaginary universe, albeit improbable, because it follows the conjectural principles that govern that ontology.
For instance, time travel is impossible, but in a science fiction novel in which it becomes possible, the plausibility of time travel will depend on how the relations of causality in the imaginary universe evade some familiar paradoxes. Again, for instance, reanimating the dead by means of a magical incantation is impossible, but in a fantasy fiction in which it becomes possible, the plausibility of zombies will depend on the implicit rules governing necromancy. Necromancy, by the way, generally turns out to involve the expenditure of some sort of magical potency that is related to the will (or the wish) to dominate over life, and even to master death itself. In both of these cases, the question is not just the imaginary rule that makes time-travel or necromancy possible, but also, how that rule relates to the rest of the imaginary universe—which, for the most part, obeys laws that resemble those operating in the real world.
This is an important point. It means that the imaginary universes of speculative fiction tend to be ontologically consistent (
Traill 1996). George MacDonald already specified this constraint in 1893 and MacDonald-influenced “self-coherent narrative” remains central right through to Clute’s influential definition in the
Encyclopedia (
Clute 1997, pp. 336–337). That might seem surprising, for three reasons. The first is the fixed idea that impossibility violates logic, so worlds where this is represented must be inconsistent. But the relation of representation involves the transformation of impossibility into possibility through ontological speculation, the purpose of which is to imagine a universe consistent with what remains impossible in reality. (I deal with paraconsistent logics separately, in a moment.) The second is that there still exists a type of critic who thinks that every imaginary universe of speculative fiction is a playground for authorial caprice where “anything goes”. But the name that we give to intrusive manipulation of plot developments through the arbitrary introduction of transcendent powers is “bad writing”, not speculative fiction. The third argument is that writers are not logicians, so consistency should not be expected. What can I say in response to such condescension? Perhaps this: that since Aristotle, analyzing the marvellous writing of the ancient world, it has been known that “things plausible but impossible [in reality] should be preferred to the possible but implausible [in the text]” (
Aristotle 1995, pp. 124–125). What is this, if not a requirement of consistency? A “literature of the impossible”, then, is one in which what is impossible in the real world becomes possible in the imaginary universe, because that universe is consistent, that is, it is governed by a speculative principle that regulates these extraordinary events. The task, then, is to locate this speculative principle and the extraordinary events it makes possible within a consistent framework that relates extraordinary powers to mundane events through plausible rules.
Let me offer some reassurances. In support of the claim that there are several varieties of impossibility, I am not about to make recourse to some hair-raising speculative ontology that involves exotic logical commitments. I am not interested in philosophical irrationalism. My philosophical support for the “several impossibilities” claim is that stolid rationalist, Jürgen Habermas. My theoretical source for trans-consistent thinking is the logical research of Graham Priest. But it is important to notice that the two claims are hierarchically arranged. The several varieties of impossibility are objective, subjective, social (intersubjective) and logical, and these yield different orders of speculative fiction. I need to invoke the idea of consistent thinking about impossibility qua contradiction to preserve the validity of just one of these boxes. That is the logical one—which includes such things as nonsense literature, postmodern fictions based on ontological violations, and imaginary dreamscapes—because I have maintained that speculative fictions involve consistent ontologies, or “self-coherent narratives”.
Logical impossibility means something that contravenes the laws of logic as grasped by science and philosophy
in the present. It is worth noticing that logical impossibility is historically variable. For the medieval world, following the theologian Augustine, it is impossible for humanity to think the infinite. For the modern world, after the set-theoretical reconstruction of the foundations of mathematics undertaken by Ernst Zermelo and Abraham Fraenkel from 1922 onwards, mathematics itself is grounded on thinking about infinity. For the Enlightenment world, the “excluded middle”, the simultaneous truth of both “A” and “not A”, is impossible, because it violates the law that contradictions are illogical. But the late twentieth century logical field of transconsistent logics shows that it is possible to think contradictions in a logically rigorous way, as points of exception governed by the “inclosure theorem” (
Priest 1987;
1995). Explorations of logical contradiction that happen in fiction through ontological boundary violations between text and context can most certainly be consistently described (
McHale 1987;
Hutcheon 2013). Here, the critical rubric is not “inconsistency” and “arbitrariness”, but “violation” and “paradox”, where these terms refer to local instances of the included middle that create self-reflexive puzzles within the imaginary universe. Hopefully this digression has dissolved the prejudice that the representation of impossibility involves irrationality.
6. Varieties of Impossibility
Impossibility can be defined logically, objectively, socially and subjectively. This by no means exhausts all of the impossibilities that might be conceptualized. The basis for this idea is the pragmatic theory of language proposed by Habermas, which I have discussed in relation to literature in some detail (
Boucher 2021). According to this theory, statements make implicit truth claims against referential worlds that are constituted through intersubjective agreement. These worlds are analytically distinct insofar as they are constructed by specific procedures for defending the validity of different kinds of statements, such as scientific procedures, legal procedures, aesthetic procedures, and so forth. Assertions about the objective world make implicit claim to the accuracy of their representation of states of affairs that can be verified scientifically. Evaluations of the social world make implicit claim to the rightness of their normative judgments of conduct, relative to intersubjectively agreed standards. Expressions that refer to the subjective world of the speaker make claim to privileged access to a domain of beliefs and desires, needs and feelings, based on truthful reports on subjective states that a community thinks relatively normal. The subjective world, be it noted, is
rational to the extent that the speaker can be held to account for delusional or untruthful reporting and for deviant (or pathological) states. The symbolic coherence of meaningful formulations is assessed intersubjectively through the analysis of the logical consistency of statements. Habermas also discusses the expressive logic of aesthetic claims, which might be useful in relation to one of the categories of clearly-in fantasy texts that I mentioned earlier; however, I am not ready to develop that conjecture in this article.
These referential worlds—which exist analytically, not ontologically, for the real world of actual experience fuses these together—have types of impossibility associated with them. Logical impossibility has already been discussed. An impossibility in the objective world is something excluded by consensus definitions of objective reality, that is, in modernity, a natural- or social-scientific falsity. Imaginary universes within which the natural world has an impossible new rule added to it (e.g., anti-gravity, faster-than-light engines, time travel) are representations of (objective) impossibilities. In relation to science fiction, objective impossibilities are violations of what is possible according to the laws of physics—as they are presently conceptualized. The twentieth century’s multiple revolutions in physics—special relativity, general relativity, quantum electrodynamics, quantum chromodynamics, string theory, quantum loop gravity—have left consensus definitions of the objective laws governing the real world deeply provisional. Challenges to contemporary physics—say, for instance, science fiction involving conjectural universes in which faster than light travel is possible after all—are knocking on the open door of a growing acknowledgement that “impossibility” is susceptible to historical correction.
An impossibility in the social world is a moral or legal impossibility, relative to consensus definitions of intersubjectively binding normativity—impossibilities such as state-mandated euthanasia or the destructive harvesting of cloned organs from slave donors. Imaginary universes within which these become possible (e.g., Soylent Green, Never Let Me Go) are representations of social impossibilities. Social impossibilities are often solidly rooted in popular prejudices (and vested interests). Things such as revolutions, totalitarianisms, or the complete absence of laws, are supposed to contradict human nature. That nature, however, is moral, not biochemical or subatomic. Generally speaking, the socially impossible is what is deemed morally illegitimate (because it is supposedly abhorrent) or what is thought utopian to truly aim for (because it is regarded as idyllic).
Finally, things which are subjectively “impossible” are so because they are prohibited subjective states. An impossibility in the subjective world is a subjective state that should not exist according to consensus definitions of normal beliefs, desires, needs and feelings. These deviant subjective states are regarded as disgusting and terrifying, or they are thought excessively demanding for human willpower, or supposed to be so desirable and wonderful as to be impossible to achieve. For instance, human happiness and its opposite, generalized torment, are thought to be impossible extreme states, defining the continuum of real experience as lying in-between these two poles. By subjective, what is meant here is not what is colloquially meant by saying that something is “subjective”, meaning, arbitrary, capricious, willful or opinionated. What is meant by “subjective” is subjectivity, the inner psychological states to which the person experiencing them has privileged access, as well as what is meant by character, and the ways that character connects up psychological motivations with social orientations. Subjectivity also includes the soul, which from an historical materialist perspective is a way of thinking about the uniqueness of individual character.
Impossible subjectivities are formations of subjectivity, or character structures, which are deemed impossible because they involve extreme states of dread and desire, beliefs which are alien to the society, or feelings that motivate highly unconventional conduct. Tolkein’s fiction, for instance, consonant with the orientation of the Inklings as a whole, speculated about religious beliefs, under conditions of an industrial society which denied the legitimacy, let alone the relevance, of such faith. The imaginary universe of
The Lord of the Rings represents as possible the characters’ belief in an interventionist deity, which modernity had rendered impossible (absurd and false) in the real world. Most of all, what is subjectively impossible involves forms of subjectivity which define human needs in contradiction with the human needs permitted by society. These can range from the demand for human happiness, through the desire for immortality, or the wish to truly speak with the nonhuman animals, to the hope for a cosmopolitan peace, or universal social equality. They also include the character structures that go with longing for a new identity that is no longer despised by society, and which fits with a yearning for respect, when disrespect is the norm. Finally, they include speculations about those motivations and orientations that make it possible to work, love and play, without having to subject oneself to the mutilating forms of psychological repression needed to conform socially within some real-world social arrangements. NK Jemisin’s
The Broken Earth trilogy is a good example of this latter (
Jemisin 2015–2017). It begins by literalizing the metaphor of the power of the powerless as the capacity to move mountains (something impossible as a literal event in the real world). It does this by conjecturing an imaginary universe that includes “orogenes”, seemingly powerless victims of discrimination who are truly able to cause massive earthquakes and move gigantic stones. To regain their world-transforming power, however, the central orogenes must not only rediscover their repressed rage against injustice, but also reconnect with their suppressed capacity to care for one another and for the earth. Real-world discrimination, Jemisin implies, mutilates the subjectivity of the oppressed, turning their rage against themselves and muting their capacity to care, making their “moving mountains” impossible. But the imaginary universe of the fiction speculates about the world-redemptive implications of a liberated (and liberating) subjectivity, which would re-equilibrate rage and care within a framework of self-respect and the desire for justice.
7. The Class of Speculative Fiction
The best description for what has so far been called the “literature of the impossible” is “speculative fiction”, because this designation refers us to the idea of a set of conjectures about alternative forms of existence. Speculative fictions are thought experiments that depart dramatically from an unsatisfying or limited reality and make radically transformative conjectures about how something crucial might be different. I now want to develop a discussion of the four orders of the class of speculative fiction, which can be derived from the four different ways that the impossible can be defined.
By describing speculative fiction as a “class”, consisting of four “orders”, defined by four types of impossibility, I am making reference to the concept of a taxonomy of forms. Familiar from the biological sciences, the standard hierarchy involved is as follows: domain, kingdom, phylum, class, order, family, genus, species. Notice that this system of forms works with differences, rather than similarities; within the function of the reproduction of animal life, for instance, its categories are “differentially defined”. In practice, that means that sorting out what group a newly discovered lifeform belongs to is a question of deciding what its differences are with known entities.
The function that speculative fiction performs is communication. The branch of a literary taxonomy that would include fantasy literature should be as follows: communication (domain); narrative (kingdom); narrative fiction (phylum); speculative fiction (class); fantasy literature (order); fantasy genres (family); subgenres of fantasy (genus); works of particular authors (species).
If speculative fiction is a class defined by its representation of the impossible, then it is a class with four orders, defined by the kind of impossibility that they represent: logical, objective, social and subjective. The function of speculative fiction as a literary class is communication, and specifically, the communication of a conjecture about consensus definitions of what is thought to be impossible. The implication is that fantasy literature and science fiction, in particular, are thought experiments in narrative form, with the culturally indispensable role of interrogating the limits of present agreements about reality.
Let me now run through my construction of the different types of impossibility and the kinds of speculative fiction their negation makes possible. In a later section, I will confront the objection that science fiction is a literature of the possible, whereas fantasy literature is a literature of the impossible.
The literature of logical impossibility is a literature of paradox, including so-called “nonsense fiction”, such as characterizes
Alice in Wonderland, for instance. It includes experimental fictions which violate ontological boundaries between the book and the world, or between levels of the imaginary universe. In these novels, the characters speak with the author, or characters in books in the novel communicate with the characters who read the books, and so forth. On the theoretical description of postmodern fiction provided by Brian McHale, a lot of postmodernism fits into the “logical” order of speculative fiction (
McHale 1987). The literature of paradox embraces logical impossibility as a representational possibility in a fictional context, without diminishing the force of the logical dilemma that results.
By contrast with the literature of paradox, which embraces logical impossibility as a representational possibility, some science fiction speculates about conjectural laws of physics, which transform real impossibility into imaginary possibility. Paradoxes are avoided—for instance, the whole point of many time travel novels is not to present an insoluble logical paradox. Instead, many time travel novels turn the paradoxical causality of temporal loops into a chain of imaginary events that “solves” the paradox through multiplying worlds. More satisfyingly, some conjectures about time travel avoid logical paradox through the qualification that nature nonetheless resists causal inconsistencies, and so sometimes the intended time is missed by the traveller. The recent series of books by Connie Willis, which include To Say Nothing of the Dog, The Doomsday Book and Blackout/All Clear, are excellent examples. It should be pointed out that the technological obsessions of science fiction novels—spaceships, artificial intelligences, plasma torpedoes, autonomous robots, biological implants, digital selves—spring from this conjectural physics. A conjectural physics is seldom presented through imaginary disquisitions on arcane topics, beyond deliberately bewildering mentions of the “Kleshnev Field”, and suchlike. Instead, its presence is represented through the operationalization of its strange new laws, via the technological manipulation of its wonderful (or terrifying) regularities. The point of a technology is that its augmentation of human (or nonhuman) agency in the world depends solely on the objective properties of that world, and not at all on what those agents think, or how their society is organized. Indeed, a lot of science fiction spends a lot of time imagining how societies and mentalities would be made entirely different, if such-and-such a technology were possible, by virtue of the effects of this-or-that conjectural law. The novels of Iain Banks’s Culture series are excellent examples: despite the seemingly magical technology of a fully-fledged Kardashev Type I civilization, the texts are at pains to present faster than light travel as a consequence of manipulation of the Energy Grid, super-intelligent shipminds as consequences of AI research, a sybaritic and xenophilic culture as the consequence of energy abundance, and so forth. Here, society and subjectivity are consequences of a conjectural physics, not causes of it.
The literature of social impossibility should be very familiar: it is utopian fiction and dystopian literature. The utopias and dystopias in question depend neither on logical paradox, nor on a conjectural physics. Rather, they are the consequence of social arrangements that are contrary to human nature (as society defines it currently). In both cases, utopian and dystopian, social arrangements stifle human nature—either the society of the future shall deliberately strangle human nature with its totalitarian organization or its shocking entropy; or, the society of the future shall cast off the shackles of present ideologies and oppressions, letting human nature flourish at last. Although technological means (and sometimes magical devices) play a role in enslavement or liberation, they are not central to the imaginary universe, which centers rather on the question of what forms of social life are right for human beings. It is the idyllic and the abhorrent which is at stake in utopian and dystopian literature, not the possible manipulation of a conjectural environment.
Finally, the literature of subjective impossibility deals with the idea of a subjective state becoming directly effective in the extra-psychological environment, as an efficient cause. In the real world, it is impossible for a feeling, a desire, a belief or a need to be an efficient cause—subjective states are motivations that inform (ethically inflected) orientations, which guide actions (that are the efficient causes of events). Without the action (often mediated by a technology), the subjective state is impotent, a “mere wish”, a “frustrated impulse”, or the like. For instance, think about the falsity of the slogan of the National Rifle Association (USA), “guns don’t kill people, people do!” No: guns kill people. The technology of the rifled firearm, which operationalizes the laws of physics relating to ballistics, is intrinsically necessary to the lethal action-at-a-distance that happens when someone with the intention to kill squeezes the trigger at a remote human target. Without the implement, the desire to murder is just a vacuous, seething hate. By contrast, in fantasy worlds, “wands don’t kill people—wizards do!” is a true proposition. Nowhere is this more evident than in JK Rowling’s Harry Potter series, where there is much ado about wands, and objective Laws of Magic, including the much-cited Gamps Laws of Elemental Transformation. But nobody thinks that this is actually a science fiction universe within which a conjectural physics operates, which includes the capacity to kill, fly, conjure or abjure Dementors for the very simple reason that it is not one, as evidenced by the following. When push comes to shove, in the Potterverse, it is always a question of the intensity and intentionality of a subjective state (especially a powerful emotion such as love or hate), and never a question of, say, the combination of quantitative energy with qualitative directedness, multiplied by some technology (such as a wand or a potion). In the final analysis, what matters is not the wand, or the incantation, “Avadakedavra!” or “Crucio!” but the desire to harm animated by the spirit of hate; the wand is just a psychological prosthetic and the spell is simply a rote bash. In all of the most interesting magical combats, the characters no longer bother to point their wands or utter their curses—they just feel intensely and direct this feeling intentionally. The imaginary universe is directly responsive to a subjective state. In the paradigmatic instances of fantasy literature, then, the existence of “magic” means that the imaginary universe is responsive to willpower or to wish-fulfilment. Meanwhile the protagonist typically wields an extraordinary magic, which transforms them into someone recognized as special because of their superior capacity to direct the force of subjectivity.
The two biggest groups within the order concern the two main popular definitions of subjectivity: the soul; and, the emotions. In general, what happens in the literature of subjective impossibility is that the subjective world emerges directly into the natural environment as “magic” and “demons”. Flowing from this, society is reshaped by the existence of sorcerers and spirits, which is something that, generally speaking, leads to apocalyptic threats and redemptive scenarios. But in fantasy literature (including horror stories), the supernatural is never ultimately a conjectural physics involving magical laws, nor is it a mere result of social arrangements that do or do not conform to human nature. Instead, the supernatural is an independent dimension that is primordial and powerful, a force to be reckoned with and not a technology to be mastered, nor a set of institutions to be reformed. Technology and society are influenced by it, not the other way around. The supernatural in fantasy literature (including horror stories) turns out to express a speculative subjectivity: magic depends on directing powerful emotions from the extreme ends of the spectrum, such as love and hate; or, magic expresses the nature of the soul because it is a projection of spiritual qualities. Of course, there are many magical systems and demonic adversaries in fantasy literature, expressing desires (and wishes), dreads (and anxieties), idyllic and demonic definitions of the satisfaction of human needs, such as the desire for domination, or the need for protection, and feelings such as release and control, as well as love and hate. What this incredible diversity of speculative narratives all have in common—what makes them parts of an order—is what they are not. They are not the result of a conjectural physics. They are not the consequence of good or bad social arrangements. And they are emphatically not logical puzzles. Rather, the supernatural is something like subjectivity run wild; fantasy novels are thought experiments, in a symbolic (“magical”, “demonic”) key, about what it would be like to think and feel the world differently.
This is often true even when its authors seek to do something else. Brandon Sanderson’s “hard fantasy”, developed by analogy with “hard sci-fi”, sets out to evolve an imaginary universe that is objectively regulated, albeit by the paraphysics of magic, not a conjectural physics (
Sanderson 2007). On the logic of the description of fantasy that I have provided, that should land his work somewhere in “category 2”, alongside psionics, in the group of science-fantasy works that elude my special subset of “subjective principle” fantasies. Nonetheless, his
Mistborn series is an effort to realize this vision that instantly miscarries into the field of “subjective principle” fantasies. There is a complex architectonic of “allomantic”, “feruchemical” and “haemalurgic” magics, described as one might set forth the ontological semantics of a computer game (i.e., something objectively regulated) (
Sanderson 2006, pp. 645–647): allomantic raw materials provide magical powers in proportion to fuel load plus skill level, with determinate effects conditioned by the type of metal burnt. The fascinated reader soon discovers, however, that although allomantic metals are ingested into the stomach, they flare in the chest (
Sanderson 2006, p. 136). Their suspicions aroused, the reader cannot help noticing that the emotional correlate to the metallic material is an essential component of the supernatural effect. Indeed, once the main characters are completely out of fuel, pseudo-objective allomancy goes out the window, replaced by something more familiar: “The Inquisitor raised his axe to strike.
She loves him [thought Kelsier desperately]. Kelsier flared steel within, stoking it, raging it until his chest burnt like the Ashmounts themselves”; “Her pewter ran out.
He killed Kelsier [thought Vin desperately]. Anger, desperation and agony mixed within her, and the Pull became her only focus. The bracelets ripped free” (
Sanderson 2006, pp. 565, 630).
Accordingly, I want to describe a significant subset of the core of fantasy literature as an order of speculative fiction in which the imaginary universe is regulated by a subjective principle. But it is important to notice two things here. The first is that what I mean by “regulated” is that the subjective principle trumps all others, not that the world is exclusively responsive to subjectivity. Generally speaking, in fantasy literature, the mundane world is responsive to the same laws that regulate the real world. The point of the fantasy conjecture is that it is a “what if” thought experiment about varying this only at specific points—extraordinary events—where the subjective principle that ultimately governs the imaginary universe irrupts into the texture of human action. Sometimes, however, the mundane world is permeated by minor magic—cantrips, sprites and the like—representing ordinary instances of that subjective principle that turns out to be susceptible to extraordinary expressions too. The second is that while most contemporary fantasy thinks subjective principles in paraphysical terms, as a reservoir of something like emotional “mana” together with a set of activating rituals, the definition also includes religiously toned imaginary universes. Here, although the world is literally governed by a subjective principle—namely, the intentionality of the divinity, which legislates to nature—nonetheless, the standard distinctions between mundane existence and the extraordinary manifestations of that divine will still obtain.
9. Contrasting Sci-Fi with Fantasy
The differential nature of a taxonomy of forms can sharpen the distinction between fantasy literature and other forms of speculative fiction. The best starting point for this is a particular subset of science fiction, which has a considerable body of theory pointing to its status as a literature based on a conjectural physics, which represents real impossibilities as imaginary possibilities. But before I can discuss this, though, I have to confront a major problem. My constantly alert and permanently combative reader, a demon who used to belong to Professor Maxwell, has an objection to make. “Doesn’t science fiction exclusively concern itself with extrapolating present possibilities as future technologies? Don’t sci-fi authors often pride themselves on their knowledge of contemporary science? Face it: science fiction is a literature of the improbable, not a literature of the impossible. Your ‘class’ of speculative fiction is extraordinarily fragile, because its most important ‘order’ just refused to belong to its definition!”
Annoying creature! I can see that you have been reading work by science fiction author Stanislaw Lem and sci-fi critic Darko Suvin. According to Lem and Suvin, there is an opposition between science fiction (as the literature of the possible) and fantasy literature (as the literature of the impossible). The intentions of both Lem and Suvin in making this contrast are to highlight the positive status of science fiction against the negative status of fantasy literature. They think that science fiction involves disciplined conjectures based on extrapolations about possible developments that are consistent with scientific knowledge. By contrast, fantasy literature, according to them, involves wild speculations based on the idea of a universe governed by subjective caprice (the arbitrary will of a god), or a wish-fulfilment world governed by magical thinking (the wielding of magic). They think that because fantasy deals with impossibility, “anything goes”, just as also happens in thinking where contradictions are permissible.
According to Lem, the imaginary universes of fairy tales and fantasy literature differ systematically from those of realist literature and science fiction (
Lem 1973). In the fantasy universe, the cosmos is positively oriented towards human beings and the novel is “happiness-giving”, whereas in science fiction, as in realist literature, the universe is neutral, even indifferent, towards human beings. Science fiction differs from realist literature only in that it describes “other points on the space-time continuum”, compared with the actually-existing world.
Suvin hones this assertion into a theoretical definition. Initially, Suvin states (a bit obscurely) that: “S[science] F[iction] is distinguished by the narrative dominance of a fictional novelty (novum, innovation) validated both by being continuous with a body of already existing cognitions and by being a ‘mental experiment’ based on cognitive logic” (
Suvin 1978). I think that what that means is that science fiction involves adding a scientifically credible cognitive conjecture to existing science—which would rule out speculations about impossibilities. He concludes:
what differentiates SF from the ‘supernatural’ genres or fictional fantasy … is the presence of scientific cognition. … Any tale based on metaphysical wishdreams—e.g., omnipotence—is ‘ideally impossible’ … according to the cognitive logic humanity has cumulatively acquired … It is this, and not positivistic scientism, which separates … supernatural fantasy from SF.
The difference between science fiction and fantasy literature is the difference between scientific knowledge (cognitive thinking about the objective world) and “metaphysical wishdreams”, also known as magical thinking. Suvin thinks that science fiction involves a conjectural physics, but only on condition that the conjecture involved as an extension of what is currently regarded as possible within contemporary physics. Indeed, he thinks that fantasy involves infantile omnipotence because its cultivation of impossibility means that it ends up governed by a wish-fulfilment principle. This makes fantasy literature the contemporary form of the fairy tale.
A major contemporary critic such as Fredric Jameson, who constructs the opposition between science fiction and fantasy literature on exactly these grounds (albeit with a more generous interpretation—fantasy is about alienation, not just infantile) shows just how lastingly influential this line of thinking has been (
Jameson 2005, pp. 58–64).
Historically speaking, Lem and Suvin represented efforts to turn science fiction criticism aside from the literature’s involvement with impossibility, as formulated for instance in the seminal
New Maps of Hell (
Amis 1960, pp. 16–21), and recruit it to a form of positivist culture everywhere opposed to “mere fantasy”. Kingsley Amis acknowledged the existence of two poles in science fiction, the plausible and the impossible, and differentiated these both from fantasy not on grounds of immature versus mature, but of speculation about the laws of nature versus speculation about something else (
Amis 1960, pp. 21–22).
By contrast, Lem and Suvin sought to then align the opposition subjective/objective with the opposition impossible/possible.
Two key examples will demonstrate what the problem is. Stanislaw Lem’s novel Solaris (1961) is paradigmatic of science fiction for Darko Suvin, just as HG Wells’ novel The Time Machine (1895) is paradigmatic of science fiction for Lem.
But both of these rely on a conjectural physics involving objective impossibility! The impossibility of time travel—relative to contemporary physics—probably needs no special discussion. Although (according to current knowledge) time’s arrow is reversible in mathematical models of physical processes, in actual events, the speed of light and increasing entropy seem to determine the momentum and irreversibility of time.
Lem’s sentient ocean, the planet Solaris, meanwhile, seems to generate communication and to manipulate distant events using neutrinos. In 1961 these were theoretical particles of zero mass and short range, whose interactions with baryonic matter were restricted to those governed by the weak nuclear force (which happens only over subatomic distances). Even in 2024, where it is now known that neutrinos travel incredible distances and have non-zero mass, the Standard Model still excludes the possibility of weak interactions involving neutrinos from generating durable macroscopic effects (like the appearance of replica human beings), such as happens in Solaris.
Is it laboring the point to notice that the novel therefore involves a conjectural physics which includes an objective impossibility?
The kinds of science fiction that speculate about forthcoming technologies often seem to belong to realism, however, as a literature of the possible. But they often only become science fiction by virtue of the present impossibility of realizing that technology.
I acknowledge that the claim that something is a presently impossible technology represents a significant concession, relative to the claim that science fiction depends on an invented force that is currently excluded by the laws of physics. The last of these two examples—the terraforming of Mars and the escape of a super-intelligent AI—are presently technologically impossible, but certainly not theoretically impossible according to current physics. It would contradict the facts to maintain that all science fiction is about objective physical impossibility, when a lot of science fiction is about the emergence of improbable, but objectively possible, things, through the invention of presently impossible technologies.
I therefore restate my position as the claim that there is an important part of science fiction that it involves speculation in which the imaginary universe is regulated by a conjectural physics, or by some presently impossible technological breakthrough.
10. Science Fiction and the Cognitive Novum
These reflections on the difference between science fiction and fantasy literature are highly suggestive in relation to the other part of Suvin’s discussion, concerning the “cognitive novum”. According to Suvin, “science fiction is a literary [order] whose necessary and sufficient conditions are the presence and interaction of estrangement and cognition, and whose main formal device is an imaginative framework, alternative to the author’s environment” (
Suvin 1979, pp. 7–8). By “estrangement”, Suvin means a technique of representation that presents something from the real world in an unfamiliar, yet recognizable, way—in a different light. By “cognitive”, Suvin means rational thinking about the world as a set of objects, such as happens in both natural and social science. The coordination of estrangement and cognition is summarized by Suvin as the “cognitive novum”, something new about the objective world of the imaginary universe. Suvin believes that the novum refreshes thinking about objective reality in the actual world.
Now, it is manifestly impossible to think about the cognitive novum without constituting science fiction as a literary order defined by its invention of an objective world, which can be considered cognitively. That means: scientifically—which is to say, as a set of regularities that can be summarized in physical laws, statistical tendencies, probabilities of events, and so forth; in shorthand, a conjectural physics. This consideration strongly suggests that the cognitive novum is an entailment of the definition of science fiction as kind of speculation whose imaginary universe is governed by objective regularities that are consistent with a conjectural physics (or technology).
But wait. Cognitive thinking, as modelled by the sciences, is not the only kind of rational thinking. Of course, the world can be considered as a collection of objects whose regularities can be detected by means of scientific hypotheses. But the social world can also be considered as a set of agreements between subjects, about how to behave and what arrangements to have. When subjects argue about morals and politics, about interpersonal interactions and collective values, they are not arguing about the social scientific description of these rules. Is it true or false that such-and-such a rule exists? That is not the question! They are arguing about the rightness of the rules and about their justice. Such debates are described as “normative” (they are about norms), and they involve rational claims about arguments which are right or wrong, not about hypotheses which are true or false.
Now, the inner domain of psychological experiences can also be debated rationally. Feelings and beliefs can of course be capricious, opinionated or idiosyncratic. But they can also be learnt about and modified, and they can be rationally justified. It is extraordinarily common for people to debate the validity of an expression of feelings: you may feel hurt, but that is because you are taking it too personally, when it was clearly not meant that way. This is not just an exchange of opinions, where there is no way to reach agreement or even acknowledge disagreement. It makes reference to evidence (it was clearly not meant that way) and to standards (you should not normally feel personally injured by an impersonal comment). The debate that emerges from this discussion of feelings is rational, and the kind of rationality is “expressive”, concerning the affective domain.
Expressive reasoning typically connects up feelings and beliefs with standards and values. I am going to speak of the link between “motivations” and “orientations” as what is at stake in expressive rationality. A motivation is a ground for action in a subjective state (feeling, desire, belief, need). Motivations are “authentic” when they are genuinely experienced and when they are socially accepted. An orientation is a framework conditioning what goals of action are valid. Orientations are pre-normative: they are not the same as rules and norms; rather, they are what conditions which rules and norms are likely to be selected/operationalized. They are “legitimate” when the orientation in question is socially accepted as a good reason for selecting goals. “Being a good person” is a legitimate orientation in most societies (although what counts as good will vary), whereas “being as selfish as possible” is unlikely to be regarded as anything other than illegitimate. These debates are rational, but they are not cognitive or normative—they are affective.
What all of this suggests is that the cognitive novum should be supplemented by the normative novum and the affective novum. Is this not, in fact, exactly the role played by utopian fiction and fantasy literature, respectively?
Before drawing conclusions about the connection between the affective novum and some subset of fantasy literature, it is necessary to break, once and for all, with the prejudice that cognitive thinking exhausts the scope of rationality. The technical name for this mistake is “positivism”, the theory that rationality consists solely in testing hypotheses about the objective world, in light of factual evidence derived from the object domain. Positivism is extraordinarily widespread (it is the characteristic prejudice of a scientific-technical age) and it is connected to the reduction of the world (including other persons) to raw materials that can be profitably manipulated. From the positivist perspective, the world consists exclusively of objects that can be manipulated effectively, based on accurate knowledge about their objective regularities. Thinking consists of making calculations about how these materials can be most efficiently used as instruments by scientifically-informed subjects. That sort of thinking is known in Critical Theory as “instrumental reason”, and its manipulative and calculating reduction of norms and feelings is thought by Critical Theory to be linked to exploitation and domination (
Adorno and Horkheimer 2002).
Yet positivist thinking and instrumental reason are surprisingly common in literary theories about science fiction. The assumptions of instrumental reason show their hand in the opposition between cognition and irrationality: anything that is not cognitive, is irrational. As Carl Freeman unfortunately illustrates, this claim typically arises in connection with fantasy literature:
[Estrangement] refers to the creation of an alternative fictional world that, by refusing to take our mundane environment for granted, implicitly performs an estranging critical interrogation of the latter. But the critical character of the interrogation is guaranteed by the operation of cognition, which enables the science fictional text to account rationally for its imagined world. … If the dialectic is flattened out to mere cognition, the result is then ‘realistic’ or mundane fiction …; if the dialectic is flattened out to mere estrangement (or, it might be argued, pseudo-estrangement) then the result is fantasy, which estranges, or appears to estrange, but in an irrationalist, theoretically illegitimate way.
What seems “irrational” to the science fiction critic is that the imaginary universe of the fantasy novel is responsive to some subjective principle. Either magic works, responding to the characters’ wishes or willpower, or the universe is sentient, and it wishes or wills that things turn out well or ill for the characters. The first kind of fantasy resembles fairy tales. The second kind resembles religious myth. Often both are combined: the universe is ruled by antagonistic deities, who donate magical power to the protagonist and the antagonist; but, this power can only be operationalized as magic by sufficiently strong willpower, or fervent wishes, or emotional intensity, etc. Freeman, Lem and Suvin all complain about this, in terms that indicate that they think that this results in “anything goes”: the imaginary universe is “capricious”, “arbitrary”, and “irrational”. Its difference from the real world is “inexplicable” (
Freeman 2000, p. 43); it is “happiness-giving” or “suffering-causing” and therefore tilted into irrational religious beliefs (
Lem 1973); it imposes “anti-cognitive laws onto the empirical environment”, which confuses the supernatural with the natural in a way that “reduces the fantasy tale’s horizon to indisputable Death” (
Suvin 2000, p. 209). (I am not entirely certain what Suvin means by that last claim, but its implication—anything goes, so nothing matters, which flattens the imaginary universe onto the one remaining existential reality, death, which means that fantasy is trivial—seems clear enough.)
11. Conclusions: Affective Novum and Subjective Principle
Imagining a world responsive to wishes or will is not irrational when this involves speculations about fresh motivations and new orientations, under conjectural conditions where they are authentic and legitimate. The fantasy novel presents some new feeling, desire, belief or need for inspection, linking this to the way that it provides grounds for action and frameworks for assessing goals. The imaginary universe of the fantasy novel, in other words, is regulated by a subjective principle, and this representation of an imaginary universe presents an affective novum for inspection and evaluation. Consider the fantasy fictions of JK Rowling, for instance, as an example of the affective novum in a context where magic works.
In JK Rowling’s “Potterverse”, the imaginary universe of the Harry Potter series, magic depends on powerful emotions, especially love and hate. This is a cosmos, in other words, regulated by the principle that primordial subjective feelings are objective forces. For those with magical potency, powerful feelings, when properly structured by magical incantations, become objective forces that can be directed by willpower. Love in the Potterverse means familial love and friendship love more than romantic love, and it means self-love in distinction from these other-directed forms of love. Indeed, as the story develops, it becomes evident that self-love is opposed to other-love, and that self-love is actually the root of hate. That is because self-love, driven by survival instincts and expressed as self-preservation, precludes self-sacrifice on behalf of the other, which is precisely what other-directed love demands. Triumph over destructive hate involves renouncing self-love, and refusing the desire for mastery over death, accepting mortality and, with it, the fragility (and value) of sociability and solidarity. One element of the affective novum in the Potterverse, then, is that Harry Potter, and his friends, must learn to be motivated by other-directed love and to renounce the destructive desires—such as retribution—that arise from self-preservation.
The other element of the affective novum in this universe concerns orientations. Self-preservative love and other-directed love are in tension under conditions of the threat of death by violence. That threat arises because of the use of violence by social forces whose aim is to preserve or restore social hierarchies. Maintaining domination within social hierarchies rules out other-directed love because it relies on a lack of sympathy for the excluded other. Instead, structures of oppression rely on self-preservation to motivate self-interested behaviour, which depends on egocentric calculations about the advantages and disadvantages of performances of domination and submission. The motivation of other-directed love can only be effective when linked to an orientation to the deconstruction of social hierarchies, so that there is no bond of solidarity without social justice. What is “new” about that, what makes it a novum, then, is the strong link between motivation by love and an orientation towards justice. The novum of the Harry Potter series has nothing to do with boarding schools, teenage wizards or orphaned boys. To say that what is new about it is that the evil wizard, the dark lord, Voldemort, is a fascist, comes much closer to the truth. But it must not then be forgotten that Harry and friends are anti-fascists. Fantasy literature is seldom explicitly political in the way that JK Rowling’s imaginary universe is. Nonetheless, what is new about it is that, rather than motivating anti-fascism by a hatred of injustice, for instance, she insists that, to be effective as justice, opposition to fascism must be motivated by other-directed love.
Fantasy fiction is not “swords and sorcery”, or supernatural literature, but any work in which the impossible becomes possible within a universe regulated by the operation of a subjective principle. The purpose of this sort of conjecture is to present an affective novum, some new potential of subjectivity which is presently considered to be impossible in the real world. This potentially affects the motivations and orientations of fantasy readers, to the extent that they accept the suggested changes as rationally acceptable.
My critical proposal has a reformative but non-polemical intention. Seeing fantasy literature as a natural ally of science fiction, utopian literature and experimental postmodernism is intended assist with current efforts to shift the locus of its study out from the exploration of a special category, “the” literature of the impossible, and into productive dialogue with other sorts of speculative fiction. By providing a differentiated language for speaking about impossibilities and a theoretical construction of the potential effects of the different orders resulting from their representation, I am to facilitate this conversation. What I am proposing is a sort of “united front” across the study of speculative fiction, for it is high time, it seems to me, for critics and writers to be able to confidently state that the most popular literature in society today performs a vital cultural role that can by no means be reduced to mere entertainment.