About CA
Contact CA

Search Journal

Editorial Board

Permission to Reprint


Site Map


The Value of Fictional Worlds (or Why 'The Lord of the Rings' is Worth Reading)

  James Harold

Some works of fiction are widely held by critics to have little value, yet these works are not only popular but also widely admired in ways that are not always appreciated.  In this paper I make use of Kendall Walton’s account of fictional worlds to argue that fictional worlds can and often do have value, including aesthetic value, that is independent of the works that create them.  In the process, I critique Walton’s notion of fictional worlds and offer a defense of the study and appreciation of fictional worlds, as distinguished from the works of fiction with which they are associated.

Key Words
aesthetic value, criticism, fan, fiction, fictional world, genre, Walton

1. Introduction

It is hard for serious readers and critics to understand why certain works of fiction are as popular and successful as they are.  For example, from the time of its publication through today, the critical response to J.R.R. Tolkien’s trilogy, The Lord of the Rings,  has been resoundingly negative: critics derided it for its plodding, pedantic style, its overly complex plot, its heavy-handed self-importance, and its simplistic character psychology.[1]  Yet The Lord of the Rings’ popularity is enduring. It has won several readers’polls for the best book of the twentieth century and even of the second millennium.[2]  So The Lord of the Rings is not merely popular; those who like it think that its popularity is deserved.  What accounts for the divide in opinion between ordinary readers and critics? 

One might argue that, considered as a literary work, The Lord of the Rings has been underrated by critics who look down on genre fiction, perhaps particularly on science fiction and fantasy.  There certainly might be something to this.  Works of genre fiction, and particularly of “pulpy” and juvenile genres, are not given a fair chance by some critics, and no doubt some gems have been overlooked. [3]  However, this is not the line of argument I will pursue in this paper.  I concede that The Lord of the Rings, considered merely as a literary work, is not one of the gems.  However, I will argue, it should not be judged merely as a literary work; it should be judged also in terms of its contribution to the development of a fictional world.

My view is that critics and fans tend approach works like The Lord of the Rings very differently.  The critics evaluate works of fiction on their own merits considered as art objects, while fans evaluate some  works of fiction in terms of their contribution to a larger project: establishing and exploring fictional worlds.[4]  Many fans of Star Wars, for example, were disappointed by the “prequel” films released in the 1990s, not just because they were bad films (though they are), but also because they seemed to diminish or distort the world of Star Wars , which had been created and explored in earlier films.  It is the purpose of this paper to explain how it can be reasonable for fans to evaluate certain artworks according to the artworks’ creation of and contribution to a fictional world.  It is necessary to say something about the kind of value that fictional worlds can have so that we can explain how some works of art and related artifacts (books, television programs, comic books, video games, etc.) are appropriately valued in terms of their relationship to these fictional worlds. 

The structure of the paper is as follows:  First, I describe the relationship between fictional works and fictional worlds, with particular focus on a class of central cases of works where the associated world seems to take on an independent importance.  The central cases I have in mind are mostly works of science fiction and fantasy, such as the novels, television programs, films, comic books, and so on, that are collectively grouped together, such as Star Trek, Dune, the X-Men,  and the like.  However, the central cases are not always works of fantasy or science-fiction.  The world of Sherlock Holmes, for example, may be included in this class.  In the second section, I show that these works (and sometimes others) are valued for their relationship to fictional worlds.  I do so by exploring three objects of fan interest: non-artwork objects that allow fans to interact with, represent, or imagine fictional worlds (in Walton’s sense, “props”); fictional puzzles and debates; and other creative forms of fictional interaction besides the work itself, particularly what is called “fan fiction.”

In the third section I explore a central challenge to this sort of evaluative attitude: why should works be valued in terms of their relationship to worlds, rather than the other way around?  Insofar as some fans do value The Lord of the Rings because of its relationship to Middle-Earth (the fictional world of that work), why would we not say that they are confused, that they are looking for value in the wrong place?  I argue that it is eminently reasonable for fans to value fictional worlds as ends – as abstract objects in their own right. Thus, it is reasonable to value some works of fiction as mere means to the end of exploring a fictional world. 

2. Central Cases: Fictional Works That Fans Love

Before we begin, it is necessary to get clear on a few important terms.  We need to have at least some rough account of what a fictional world is in order to discuss its value.  In this paper I make use of Kendall Walton’s well-known account of fictionality.  Walton notes that whether or not a fictional world is constituted by a set of propositions, it is at least strongly associated with a set of propositions, viz., the set of propositions made fictional by a particular representational work (or works).  So, the books of The Lord of the Ring trilogy make it fictional that there are elves, dwarves, and hobbits, that Gandalf and Saruman are wizards, that Aragorn is a descendant of ancient kings, and so on.  These propositions, and others like them, are made fictional by these works insofar as the works function in a game of make-believe.  Walton puts it this way:

It is the function of a representation to be used as a prop in certain sorts of games.  Function in this case might be thought of as a matter of there being rules or conventions about how the work is to be used.  Appreciators are supposed to be playing certain sorts of games with the work.… So we can say that what is fictional in a work is what appreciators of it (qua appreciators of it) are to imagine.[5]

The world of a work has to do with what a work makes fictional, and a world makes certain things fictional by prescribing audiences to imagine in accordance with a set of conventions. 

Similarly, Nicholas Wolterstorff’s account of fictional worlds is that a work projects a fictional world by indicating or making understood that certain propositions are to be understood as holding in that world.[6]  Central to his view is also the idea that the author uses the work to indicate to audiences what they are to imagine about the world.  One difference between Wolterstorff and Walton is that, according to Wolterstorff, fictional worlds consist of propositions and hence of states of affairs, whereas Walton avoids making this metaphysical commitment.  By understanding worlds in terms of what is fictional, and what is fictional in terms of the prescriptive rules that govern imaginative games, Walton does not offer an ontology of fictional worlds.  Worlds are not constituted by objects or states of affairs but by a set of (mostly implicit) agreements that govern how we understand fiction.  In what follows, I use Walton’s terminology, though the differences between Wolterstorff’s and Walton’s theories do not affect the argument of the paper.

The examples I consider here come from what I will call central casesEvery work of fiction (even those that are not considered works of literature, such as Francis Bacon’s New Atlantis[7] ) projects a fictional world, and any fictional world might, in principle, have the sort of value I describe here.  The central cases are those in which a certain kind of evaluation, which emphasizes features of the world independent of the work, becomes dominant.  The central cases tend to be works of mass popular art rather than works of “high” art.  This is a contingent fact, however, and there are exceptions – the Dublin of Joyce’s Ulysses has attracted a great deal of the kind of interest I describe here, despite the fact that Ulysses is considered to be great literature.  What follows are general principles, not universal laws.  It is, however, helpful to focus on the class of cases that have most often tended to give rise to fascination with fictional worlds.  The central cases are interesting because the fictional worlds (such as Middle-Earth) tend to be evaluated much more highly than the corresponding works (e.g., The Lord of the Rings).  This is not the case, for example, with Joyce’s Dublin and Ulysses.   Since Ulysses is generally considered to be one of the greatest works of literature written in English, no one wonders why Bloomsday is widely celebrated or why Joyce’s depiction of Dublin holds such fascination for so many admirers.

Central cases tend to have three features:

(1) They are works that, taken together, are best understood as telling us about a single world.  Walton assumes for the sake of simplicity that there is a one-to-one correspondence between works and worlds, but it is clear that this is often not the case.  Even if we count the three volumes of The Lord of the Rings as one book, Tolkien wrote other books, most famously The Hobbit and The Silmarillion, which take place in the same world.  Not only do the individual episodes of the 1960’s television series Star Trek all take place in the same fictional world, but other television series (Star Trek: The Next Generation, Star Trek: Deep Space Nine, and so on) also take place in that same world.  So we can speak of sets of works where every member of that set describes the same world.

To establish that a set of works describes a single world, it is not sufficient to show that the works were created by a single artist in a characteristic style.[8]  Compare Sir Arthur Conan Doyle's Sherlock Holmes stories with the novels of Jane Austen.  Austen's books (except perhaps Northanger Abbey) share a great deal in common with one another: they have similar settings, character types, themes, and moods.  But no specific characters, places, or fictional events from one book recur in any of the others.  Emma does not make it fictional that there is a Mr. Darcy or a place called Pemberley.  It does not specifically deny these things, but it is not part of the world of Emma that such people or places exist. 

By contrast, in the Sherlock Holmes novels and stories, facts established in one work are normally taken as facts in later works.  Characters recur: not only Holmes and Watson, but Inspector Lestrade, Mrs. Hudson, and many others.  Places, facts, and events discussed in one story are overtly or obliquely referenced in later stories.  For example, in Doyle’s “The Adventure of the Greek Interpreter,” we are told that Mycroft Holmes has an important job with the British government.  A later story, “The Adventure of the Bruce-Partington Plans” makes reference to this previously established “fact.”  (This is not to say that any two works that share a world thereby share a cast of characters.  That Mycroft Holmes is part of Doyle’s London does not imply that he is part of Hound of the Baskervilles.)  Similarly, Virginia Woolf’s character Mrs. Dalloway appears in both the eponymous novel and The Voyage Out, suggesting that at least these two works share a world.[9]  Common authorship and stylistic similarity are not sufficient to establish that different works share a world. 

Nor are a common author and style necessary.  Two works of fiction might be quite different stylistically, even created by different authors, and yet share a common world, as Michael Chabon’s novel, The Final Solution, demonstrates.  The Final Solution includes Sherlock Holmes as a character (although he is not named) and clearly is meant to take place after a number of events in Holmes’ life that we know about from the original stories, but the book introduces distinctly twentieth-century themes, and its mood is different from that of Doyle’s own stories.  Nonetheless, the world of The Final Solution is the same as the world of Hound of the Baskervilles.

(2) Such works tend to be episodic.  Like Doyle’s Sherlock Holmes stories or the individual episodes of the Star Trek television program, the separate works each tell individual, narratively complete stories that are chronologically ordered in relation to one another.  This order might sometimes be difficult to make out and need not correspond to the order in which these works were created.  One interesting feature of an episodic set of works of this kind is that it is made fictional that there are gaps between the times covered by distinct works during which the characters continue to act and interact.  They continue about their lives both before and after each work (unless, of course, it is made fictional that they die). 

(3) These works usually, though not always, describe worlds that are relatively distant from our own in important ways, such as scientifically, technologically, geographically, culturally, or linguistically.  The worlds described in such works are very unlike ours and, as such, there is a great deal to tell the reader in a work about this world about what is and is not the case, culturally, technologically, and so on.  Many of the best known sets of works I discuss here belong to the genres of science fiction and fantasy.

These three features are important because they help to nurture a certain type of interest in audiences.  The first feature is particularly important in getting fans interested in works as a means of accessing a world.  The existence of multiple narrative works telling different stories set in the same “place” generates questions about that place.  The second and third features are perhaps less important, but the episodic gaps and fantastical elements can both focus audience attention on what is part of the world but not part of the story.  These three features, however, are not meant to be necessary or sufficient.  The idea is that the presence of these three features imperfectly tracks the sets of fictional works that have attracted fan interest.  (As noted earlier, there are exceptions; Ulysses has at most one of those features, and yet has a deep fan base.)  Fan interest and engagement are important because, as we shall see, the collaborative activity of fans is part of what makes the world worth caring about.

Prominent examples include the books, movies, and other fictional works that describe the fantasy world called Middle-Earth, the first of which were authored by J.R.R. Tolkien; the many books, films, and television series taking place in the futuristic world of Star Trek, the first of which were created by Gene Rodenberry; the Harry Potter books by J.K. Rowling and films set in Hogwarts and related magical lands; and the various stories, novels, and films by Sir Arthur Conan Doyle (and others) which are set in a rather dark and eccentric version of Victorian London; the various books, games, and films describing Frank Herbert’s Dune; and the various works contributing to Joss Whedon’s “Buffyverse.”  There are often arguments among fans about which works properly belong in the set.  The Star Trek novels are not generally considered to be proper members of the “canonical” set, while the films are; the reverse is true with The Lord of the Rings

I talk of fans rather than audiences because it is self-described fans who most often make the sorts of judgments that seem so at odds with those of the critics.  “Fan” is an abbreviation for “fanatic,” and no doubt some fans improperly and obsessively spend their time imagining and discussing the worlds of fictional works.  It is no part of my aim here to defend extreme fanatical devotion to fictional works or to discuss other dangers (moral or psychological) associated with obsessive interests in works of art.

However, I do want to defend one approach to evaluating art which is characteristic of fan interest: fans often desire to experience certain artworks because they wish to gain imaginative access to the work’s corresponding world.  What is fan interest in these works like?  How does it differ from a critic’s approach to such a work or from a more conventional approach to judging works of fiction?

3. Valuing Fictional Works Instrumentally

Fans value some works of fiction instrumentally.  Of course, some philosophers think that all who value artworks do so instrumentally, because they think that all artworks are valued for the pleasure that they produce.[10]  But even if this were so, fan interest would be instrumental in a different way.  Where in other cases the object of the audience’s attention is on what is made fictional by that particular artwork (or performance thereof), the central object of fan interest is the larger fictional world in which the story of the artwork is embedded.  The fan’s pleasure arises from attention to the fictional world, not just from attention to what the particular fictional work makes fictional in that world.[11]  To judge a fictional work as a fan is to attend first and foremost to what the work makes fictional reveals about the fictional world, where the fictional world is thought of as having an independent existence.  Here are three pieces of evidence which suggest that fan interest in the central cases often focuses primarily on the works’ world, not on other, perhaps more literary, elements of the work. 

(1) Fans are willing to spend significant amounts of time and money to acquire objects which are ancillary to the originating artworks, but which can serve as props for exploring the larger fictional worlds.  After the 1977 film Star Wars was released, many fans bought dolls and games representing not only the main characters (e.g., Han Solo, Princess Leia), but also, more surprisingly, characters who barely appeared in the film.  Some of these characters (e.g., Hammerhead) were not even given names in the film itself, and appeared for only moments in the background.  Nonetheless, the toys representing them became popular.  Those who bought these toys to play with them did so not in order to reenact the storyline told in the movie in which these characters barely featured, but in order to imagine new stories that take place in that world.

Fictional objects, such as swords, and even whole characters in the on-line fictional world of Everquest, for example, can be bought and sold, sometimes for thousands of dollars.  The merchandising associated with books like Lord of the Rings or films like Star Wars is immensely profitable, but money is not the only measure of worth.  Fans also create their own costumes, swords, maps, and other props to expand their interaction with these fictional worlds, spending a great deal of time and effort in the process.

(2) These fictional worlds often contain puzzles for fans to solve.  Of course, many fictional works contain puzzles also: mysteries and crime fictions call for their audiences to find the killer, for example.  However, fans also puzzle over different sorts of problems.  For example, fans disagree over how many wives Dr. Watson had over the course of his life.[12]  References” by Dr. Watson `in different stories sometimes appear inconsistent, since he was apparently made a widower at least once, but appears to have been married both before meeting this wife, and afterwards.  The work (or set of works), taken as a whole, is not merely silent on these details, but confusing.  Fans of the series offer differing theories about how many wives Watson had and when he married each.  There are a number of distinct solutions that are consistent with the evidence given in each story.  The puzzle arises in part because the world of Holmes seems to transcend the sixty individual stories and novels; fans want to bring these various glimpses of the world into a consistent whole.  Also, because of the episodic nature of the stories, by imposing a chronology on the world with periodic gaps fills in some details while omitting others.

The existence of such puzzles in the fictional worlds, not just in the works, creates a cognitive interest in the worlds themselves.[13]  Individual works of art are mostly consistent, so the puzzles would not arise unless fans concerned themselves with the worlds beyond those works.  While the presence of large numbers of direct contradictions tends to decrease fan interest in such worlds, the presence of a few generates opportunities for creative problem-solving, allowing fans to offer up new fictional possibilities to reconcile apparently conflicting features of that world.[14]  This, in turn, increases the complexity and even the “realism” of these fictional worlds, since our own real-world experiences of course contain many apparently contradictory and puzzling features.  Fictional worlds seem to be valued most when they pose enough such problems to generate interest, while creating in general a coherent and internally consistent world.

(3) These fictional worlds afford opportunities for imaginative participation well beyond the imaginative games proper to the works that generate them.  It is a familiar idea that works of fiction are valuable in part because they direct audiences to imagine the events and characters described in them.  But certain fictional worlds also seem to invite audiences to imagine other stories that are not described in those works but rather are suggested by features of the world of the work.  In some cases, other artists may even create new works of art meant to take place in a fictional world that is already established (as Michael Chabon did).  More commonly, however, people use these worlds as backdrops for more informal game-playing, writing, or creative expression that might not qualify as art-making.  The internet has volumes of “fan fiction” that take place in the worlds of Star Trek and Lord of the Rings.  Role-playing games (such as Dungeons and Dragons) offer participants the opportunity to populate fictional worlds with new characters, and to create new fictional places, events, and objects that become part of that world.  Fans paint pictures to accompany works of literature or tell stories that fill in gaps between events portrayed television episodes.[15]  Rich fictional worlds do more than afford opportunities for these creative experiences; in some cases, the original artists take steps to encourage such participation.

Michael Chabon notes that both Sherlock Holmes stories and The Lord of the Rings share this quality:

Readers of Tolkien often recall the strange narrative impulse engendered by those marginal regions named and labeled on the books’ endpaper maps, yet never visited or even referred to by the characters in The Lord of the Rings.  All enduring popular literature has this open-ended quality, and extends an invitation to the reader to continue, on his or her own, with the adventure.  Through a combination of trompe l’oeil allusions [sic] of imaginative persistence of vision, it creates a sense of an infinite horizon of play, an endless game board; it spawns, without trying, a thousand sequels, diagrams, and web sites.[16]

A world rich with detail, scientific, religious, cultural, historical, or technological, can never be so rich that it prevents audiences from expanding the world further.  In creating Middle-Earth, Tolkien created several new languages, a detailed mythology, history covering thousands of years, and even a cosmogony.  Minor characters, as well as larger ones, were provided with a rich back-story and genealogy.  This richness of detail, however, serves to generate more possibilities:  each proposition made fictional suggests other possibilities not yet settled.  The details provided create interest and the details inevitably missing encourage creative contributions to the world.  So fictional worlds, especially those that are generated by a series of separate episodic and fantastic fictional works, are well-suited to encourage active creative expression on the part of the audience.  It is for this reason that many of these fictional worlds can be found described in a variety of different media:  film, television, song, comic, novel, role-playing game, video game, short story, musical, and more.  The explosion of works about these worlds is evidence of the worlds’ tendency to promote creative exploration. 

4.  It Can Be Reasonable to Value Fictional Works Instrumentally

If the practices just described are sometimes reasonable and not merely a waste of time, it must be because the fictional worlds to which fictional works are just means are rightly and reasonably valued for their own sakes.  What makes it reasonable to value fictional worlds for their own sakes is that these worlds themselves have aesthetic value.[17]  A fictional world can be rich, elegant, cohesive, chaotic, mythic, or serene, and they can be rightly valued for these reasons. 

Fictional worlds are strange candidates for bearers of aesthetic value.  They are: (1) non-perceptible and (2) in the typical case, at least, a kind of byproduct of art making rather than its central aim.  Either of these facts may be taken as a reason for thinking that fictional worlds cannot have aesthetic value.

(1) Some philosophers have claimed that only perceptible objects can have aesthetic value.  Nick Zangwill, for example, claims that beauty and related predicates can only apply to sensory objects.[18]  This leads him to deny that a mathematical proof or any other abstract object can be beautiful.  Indeed, literary content (as opposed to the sensory character of the language) cannot be beautiful, according to Zangwill, and so cannot have aesthetic value, though it can have other artistic values.  The use of terms like “beauty” to describe non-perceptual objects is metaphorical.  He argues for this claim by drawing a distinction.  Abstract objects are only called “beautiful” because they also serve some other predetermined purpose, whereas perceptual objects are called beautiful without reference to any purpose.

What could motivate us to go one way or another over the question of whether our application of aesthetic terms to abstract objects, like proofs, theories, or chess moves, is metaphorical?  One argument for saying that they are applied metaphorically is that in all these cases, the abstract entity has a purpose.  The point of a mathematical or logical proof is to demonstrate a truth on the basis of other truths.  The point of a scientific theory is to explain the data.  And the point of a chess move is to win.  Our admiration of a good proof, theory, or chess move turns solely on its effectiveness in attaining these ends, or else in having properties which make attaining these ends likely.[19]

There are two replies to Zangwill’s argument.  First, it is far from clear that fictional worlds, considered as abstract objects, have fixed ends, as scientific theories and logical proofs do.  Perhaps Zangwill would reply that the end of say, Middle-Earth, is to enrich the experience of readingThe Lord of the Rings.  But if so, that is an aesthetic purpose, not an epistemic or technological one.  And thus it is not so different from the sort of purpose that any aspect of an artwork may be said to have.  Aristotle tells us that, in tragedy, the purpose of song and spectacle is to serve the plot, but one would not want to conclude that neither song nor spectacle can be beautiful.

Second, suppose that the use of aesthetic terms is metaphorical in the case of abstract objects.  So what?  The point is simply that it is reasonable for fans to value fictional worlds aesthetically, and to value fictional works insofar as they promote access to these fictional worlds.  If the application of aesthetic terms and evaluations to fictional worlds is metaphorical, as Zangwill maintains, that need not mean that it is inappropriate or unreasonable. Fictional worlds, like mathematical proofs, are rightly called beautiful when it makes good sense to value them in something like the way in which one values an artwork or natural scene: with contemplative enjoyment and delight.

(2) The second worry is that these abstract objects, fictional worlds, are not proper objects of admiration because they were not intentionally created.  Now in some cases this objection does not apply.  In J.R.R. Tolkien’s essay “On Fairy-stories,” he describes a process of creating fictional worlds that he calls “Sub-creation.”[20] (Tolkein is alluding to the act of creation described in Genesis).  Sub-creation is the intentional creation of an abstract object, a fictional world, through designing various props (maps, genealogical charts, etc.) that are not themselves artworks.  Tolkien quite deliberately set out to design features of his imaginary world that he never intended to appear in any work of fiction. 

However, in many (perhaps most) cases, the idea of creating a fictional world beyond what is needed for the story may never cross the creator’s mind.  This seems to have been the case with Doyle’s London.  In such cases, it may seem very odd to value a fictional world in the way one values an artwork.  That is not to say that an object that has not been intentionally created cannot be valued aesthetically:  natural objects are reasonably called beautiful if anything is.  But a fictional world is a kind of non-perceptual byproduct of an otherwise intentional act. 

It is of course strange to value a byproduct of the art-making process but it is not thereby unreasonable.  A fictional world is in some sense a facsimile of the actual world.  And our world gives rise to all kinds of intrinsic interests:  anthropological, linguistic, biological, cultural, moral, and scientific.  These interests are not always instrumental.  Often we just want to understand our world better and we delight in learning about it.  What is perhaps surprising is that we can also delight in “learning” about fictional worlds.  While there is no epistemic value in learning about fictional worlds (unless the exercise of cognitive faculties is itself indirectly epistemically valuable), there is good reason to think that the enjoyment we find in exploring worlds should not depend on their epistemic value.  Aristotle wrote:

Also (ii) everyone delights in representations .… The cause of this is that learning is most pleasant, not only for philosophers, but for others likewise (but they share in it to a small extent).  For this reason they delight in seeing images, because it comes about that they learn as they observe, and infer what each thing is. …[21]

The pleasure that Aristotle points to here lies in recognition, and there is no reason that such recognition cannot happen in fictional worlds.  On reading Tolkein’s Silmarillion, one recognizes features of Middle-Earth that are alluded to only briefly in songs from The Lord of the Rings

There is, however, a further objection to attributing aesthetic value to fictional worlds, at least value not possessed by the fictional works corresponding to those worlds.  One may claim that the notion of a fictional world that is needed is simply not available.   To see this, we need to return to Kendall Walton’s account of fictional worlds.  If we understand “work” and “world” in the way Walton does, as sets of prescriptions to imagine, then whence do these prescriptions arise?  Walton considers two possibilities.  He distinguishes between what he calls “game worlds” and “work worlds.”  Game worlds are “worlds of games that appreciators play with representational works.”[22]  There are infinitely many game worlds for each work, as each appreciator can generate her own imaginative rules, and each can incorporate first person imaginings as well (e.g., “I am looking at the Gates of Mordor”).   These worlds will vary greatly from one appreciator to the next, and the prescriptions are an aggregate of those generated by the work, as well as the prescriptions that the appreciator herself deliberately or implicitly adopts.   A work world, by contrast, is the world generated by any person who follows all and only the prescriptions made by the work.  Walton notes that the work worlds will be the same for everyone who treats the work as a prop and imagines accordingly; game worlds will vary from person to person.  These, Walton claims, are the only two choices.  He argues,

 If work worlds are not distinct from game worlds in which the works are props, how are we to decide which of the worlds of the various games that different appreciators … play with La Grande Jatte is to be identified with the world of La Grande Jatte?  If this cannot be decided nonarbitrarily, we are forced to regard the world of the painting as a world over and above those of appreciator’s games.[23]

A world “over and above” those of the games people play would be a new entity for which we’d need an ontological account.  Walton’s theory aims to make such ontologies unnecessary.  In short, Walton argues that there are only two reasonable ways to delineate the presumptions that constitute fictional worlds.[24]  Walton’s solution is to say that the work world is related to the game world by being associated with all and only what is made fictional in all of the game worlds of that work.  The dilemma for fans who would value these works as means to an independently valued fictional world is whether that fictional world is a game world, which would be different for each person, or a work world, in which case it is not independent of the works that generate it?  Neither of these options yields a single world that could reasonably be seen as distinct from the originating work.

The worlds we have been talking about here are all worlds that are products of more than one work of art.  But we cannot simply modify Walton’s account by saying that the work world comprises all the worlds of all of the relevant works.  Such a modification would present two problems.  First, it is difficult to see how one could say which works are the relevant works without making reference to the worlds that those works describe, and thus without being circular.  Second, as we have already seen, the works that generate these worlds often contain conflicting prescriptions.  The Lord of the Rings makes it fictional that Glorfindel is an elf of Elrond’s house; The Simarillion seems to describe Glorfindel as a quite different person.  One cannot talk of the fictional world of these books by simply aggregating the particular works’ prescriptions about this world. 

The way out of Walton’s dilemma is to specify which rules govern what is to be imagined, where the rules are not merely an aggregate of those that govern each work world, and are at the same time more than the idiosyncratic game worlds of an individual fan.  And there is such a way.  Persistent communities of fans, sometimes in collaboration with artists, can create a relatively stable consensus about the prescriptions to follow in imagining a fictional world.  Communities of fans who share an interest in a set of fictional works come to agree among themselves on a set of standards about what is to be imagined with regard to the fictional world.  As when children collaborate to create a new make-believe game, the agreement on what is to be imagined evolves over time, and is both explicit and implicit.  Call the worlds generated by such methods “fan worlds;” like game worlds, they extend beyond the prescriptions made in the work itself but they belong and apply to communities of appreciators rather than to lone individuals.

Fans tend to defer to the original authors of the fictional works with regard to the prescriptions that define the fan worlds, but the views of the authors are not absolute, and fans may agree to set aside an author’s prescriptions in favor of a set of prescriptions that they think is more consistent or more exciting, given everything else that is to be imagined about the fan world.  Typically, a kind of reflective equilibrium takes hold between the world of the individual works and the fan world of the fictional work.  While some works, typically those written by the author or group of authors who originated the set of works, are regarded as “canon” and are not revisable, other, non-canonical works are revisable in light of other things we take to be true about the world.  Even elements of canonical works might be seen as revisable if they are contradicted by other prescriptions in other canonical works.  This distinction between “canon” and “non-canon,” while difficult to defend in principle, is common in practice. 

And of course, fans can and do split into smaller communities over such questions, with each community of fans adopting a slightly different set of prescriptions.  The fact that it will often be difficult to specify exactly what constitutes a fan community, or precisely which imaginative prescriptions hold in a particular fan world, however, does not undermine the view.  Work-worlds face similar difficulties, and philosophers and critics continue to argue about how works are to be interpreted and what exactly is to be imagined.  The central point is just that communities of fans can collaboratively imagine fictional fan worlds that are neither game-worlds nor work-worlds but something in between, sustained by mutual discussion and imaginative collaboration.

The central cases are central for just this reason:  they are the sorts of works that are most apt to attract fan interest of this kind.  And it is communities of fans, working and playing together, who create and sustain the fan worlds, giving them a status and interest greater than, or at least distinct from, the originating works.  In principle, the world of any work can be valued in this way; in practice, few are.

Fan worlds sometimes have a prevailing tone which dictates the way in which causal or historical events must occur in that world.  In the world of Star Trek, every event that seems magical or divine must be given a technological explanation (even if that explanation is itself mystifying).  A novel taking place in the world of Star Trek in which a character used magic to rescue her ship from danger would violate the implicit rules and tone of that world, even if it did not expressly contradict any of the specific prescriptions made by any of the original fictional works.  These fictional worlds, and others like them, have ways in which things are done.  Works or games that contradict these “ways” are appropriately criticized.  This suggests that fictional worlds sometimes have the kinds of attributes necessary for aesthetic evaluation: Gibson’s cyberpunk world may be valued for its bleakness, Herbert’s Dune for its mythic character, the world of Star Trek for its optimism.  Other worlds, such as the world of Superman, are appropriately criticized for their glaring inconsistencies and their lack of any consistent mood or tone.  In Superman’s world, things can happen in just about any way you like.

However, one may object that while fictional worlds can and should be valued aesthetically, the aesthetic value of these worlds is merely a reflection of the value of the work(s) with which they are associated.[25]  Superman’s world is inconsistent because the comic books from which it originated are inconsistent; if the world of Dune is mythic, that is because the Dune novels are so.  If the aesthetic character of fictional worlds merely reflects the aesthetic character of the originating works, then how could Tolkien’s set of rather mediocre novels give rise to one of the most admired fictional worlds?

The aesthetic qualities of worlds are indeed dependent on the aesthetic qualities of the originating works, but this dependence is not perfect: sometimes the aesthetic qualities of worlds will differ from those of the originating works.  This is possible for a number of reasons.  First, even if the work and world have the same qualities, those qualities can be aesthetically approbatory in the world and disapprobatory in the work, or vice versa.  For example, The Lord of the Rings includes excerpts from, and references to, epic songs and legends detailing ancient history that is not directly pertinent to the story.  The work has a superfluity of historical and cultural detail.  From the point of view of the story, this may rightly be considered a demerit.  The story becomes stilted and confusing, because these details add nothing to one’s understanding of the characters or themes in the story.  However, this same superfluity of historical detail is one of the best qualities of the fictional world.  The richness and complexity burden a story but liberate a world, creating more possibilities for imaginative exploration.[26]

Second, many of the qualities of works are not thereby qualities of worlds.  A story told in verse will have many poetic qualities, but none of these will be qualities of the counterpart world.  Third, when we are dealing, as in the central cases, with a number of different works, often in different media, all of which describe the same worlds, it is plausible that these works will have a great variety of aesthetic qualities.  In some cases, there is a canonical work or works whose qualities are most important.  But in many cases this is not clear.  The aesthetic qualities of the world will therefore be shared with some but not all of the associated works.

Fourth, as we have already seen, fan worlds are created not just by works but by communities of fans, and their implicit agreements about what is to be imagined.  Fans may go beyond the original works to posit fictional truths, which in turn give rise to aesthetic qualities not prescribed by the original works.  The works, however, still play a role here.  Some works of fiction, particularly those in the central cases, have what Chabon called an “open-ended quality,” so that the features of the world are importantly underdetermined by the work.  It is true that this “open-endedness” can be done well or badly, and that whether it is done well or badly is relevant to the literary value of the work.  But even poorly handled open-endedness can have a positive effect on the value of the world.  As discussed earlier, many of the most interesting puzzles and features of Doyle’s London seem to be the result of errors on Doyle’s part, which may rightly be seen as flaws in the original short stories. 

For these reasons, the value of fictional worlds cannot always be simply reduced to the value of the originating works.  Some works, like The Lord of the Rings, fail as literature but succeed as worlds because the literary flaws either do not affect the world’s value, or even affect it positively because the work’s open-endedness vastly increases the possibilities for value in the imagined world.

Because some fictional worlds are appropriately distinguished from the work-worlds of the individual works that give rise to them, it is not surprising that we should find that these fan worlds attract more attention and interest than others do.  In these cases, the fictional works sometimes play a merely instrumental role.  They are valued because they provide access to an imagined world which has rich cognitive, creative, and aesthetic value, and not because the fictional work has great value in its own right.  This is not to say that every world that fans like is thereby valuable: a world is valuable if fans rightly or reasonably value it.  Some worlds might be very popular but not worthy of that popularity.  Sometimes, however, fans are drawn to certain works because of the real value of the world that these works describe.  While critics may focus on the flaws of the works qua artworks, many fans look instead to the enormous richness of the world that lies beyond the particular work.[27]

James Harold
James Harold is Associate Professor of Philosophy and Director of the Weissman Center for Leadership and the Liberal Arts at Mount Holyoke College.

Published March 10, 2010


[1] See Edmund Wilson, "Oo, Those Awful Orcs," The Nation (April 14, 1956); or, more recently, Richard Jenkyns, “Bored of the Rings,” The New Republic (January 17, 2002).

[2] See O’Hehir, Andrew, “The Book of the Century,” Salon.com (June 4, 2001).

[3] For arguments in this vein, see the first two chapters of Noël Carroll’s A Philosophy of Mass Art (Oxford: Oxford University Press, 1998).

[4] As will become clear shortly, I borrow this notion from Kendall Walton.  See his Mimesis as Make-Believe: On the Foundations of the Representational Arts (Cambridge, MA: Harvard, 1990).

[5] Ibid., pp. 60-61.

[6] Wolterstorff, Nicholas, Works and Worlds of Art (Oxford: Oxford University Press, 1980), esp. pp. 126-131.  The view is more complex than indicated, since Wolterstorff distinguishes between two strands of what is projected.  One of these strands includes those propositions that would have been believed to follow from those indicated by the author and the other includes those that do in fact follow from what the author directly indicates. 

[7] I am grateful to an anonymous referee for suggesting this point.

[8] I am grateful to Peter Lamarque for raising this question.

[9] I am grateful to an anonymous referee for making this point.

[10] Stecker, Robert, “Artistic Value,” Chapter 12 of his Artworks: Definition, Meaning, Value (University Park, PA: The Pennsylvania State University Press, 1997), pp. 247-268.

[11] That is, the fan pays attention not just to what is made fictional by the particular work in front of her.  She may, for example, also attend to facts made fictional by other works in the same set of fictional works.

[12] It is perhaps important that many of the most interesting of such puzzles seem not to have been intended by the authors of the relevant works.  It is the nature of episodic creation that it is easy for the author to lose track of details of the fictional world she has created.  It is in fact difficult to think of a fictional world that does not contain some such puzzles, and in many cases these works contain straightforward contradictions.  (Watson was apparently shot by a single bullet in both the leg and the shoulder.)

[13] For an exploration of the psychology of engagement with fictional worlds, see Skolnick, Deena, and Paul Bloom, “The Intuitive Cosmology of Fictional Worlds,” in The Architecture of the Imagination: New Essays on Pretence, Possibility, and Fiction (Oxford: Oxford University Press, 2006), pp. 73-86.

[14] For example, see Michael Chabon’s discussion in “Fan Fictions: On Sherlock Holmes,” in his Maps and Legends: Reading and Writing along the Borderlands (New York: Harper Perennial, 2008), pp. 23-46.

[15] Peter Ludlow offered a series of fascinating examples with careful analysis in his “Truth in Fanfic,” presented at the National Meeting of the American Society for Aesthetics in Los Angeles, CA, November 7, 2007.

[16] Chabon (2008), p. 44.

[17] This sort of talk might seem to bring with it a commitment to realism about value.  To say that something “has aesthetic value,” however, is not the same as to say that the object has the mind-independent property of being aesthetically valuable.  Expressivist or projectivist accounts are consistent with such talk, so I make no presumption of value realism here.

[18] Nick Zangwill, The Metaphysics of Beauty (Ithaca, NY: Cornell University Press, 2001).  For a critique of Zangwill, see Berys Gaut, Art, Emotion, and Ethics (Oxford: Oxford University Press, 2007), pp. 124-27.

[19] Zangwill (2001), p. 141.

[20] J.R.R. Tolkien, “On Fairy-stories,” in Tolkien on Fairy-stories (London: Harper-Collins Publishers, 2008), ed. Verlyn Flieger and Douglas A. Anderson, pp. 27-84.  See esp. pp. 59-60.

[21] Aristotle, Poetics: with the Tractatus Coislinainus, a Hypothetical Reconstruction of Poetics II, The Fragments of the On Poets, trans. Richard Janko (Indianapolis, IN: Hackett, 1987), p.4.

[22]Walton (1990), p. 58.

[23] Ibid., p. 59.

[24] Walton has a second argument that I will not discuss here.  He also argues that a game world must admit first-person statements about the audience’s observation of the world, and such statements cannot be admitted into a description of what the work itself makes fictional.  I agree with his point, but I believe that the account I defend here can accommodate this point as well as his does.

[25] This line of argument was suggested by an anonymous referee.

[26] Similarly, qualities that may make a fictional work better as literature may make it worse as philosophy or vice versa.  Eryximachus’ speech in Plato’s Symposium adds nothing to the philosophical value of the dialogue, but it is very funny and adds dramatic structure when the dialogue is read as literature. 

[27] Two anonymous readers for this journal provided very helpful and detailed comments, which improved the paper in many ways.  An earlier version of this paper was presented at the annual meeting of the American Society for Aesthetics in 2007, and I am grateful to my commentator, Peter Lamarque, for his thoughtful comments, and to all the participants in that session.  I particularly want to thank Rachel Zuckert for her suggestions and encouragement.