Cognitive Poetics: Discourse Worlds & Mental Spaces

Here I continue my response to Cognitive Poetics: An Introduction and Cognitive Poetics In Practice. The chapters on discourse worlds theory and mental space theory build on schema poetics, introduced in the previous chapter. Discourse worlds and mental spaces further the consideration of how context and meaning are reproduced in the minds of readers. These two theories are distinct and come from different traditions, but for cognitive poetics they compliment one another in both scope and depth.

Discourse worlds theory comes out of possible worlds theory from philosophy of language and pragmatics. Possible worlds theory is applied to logical sets and obeys certain logical rules, such as the laws of non-contradiction and the excluded middle. In Cognitive Poetics: An Introduction, Stockwell uses the term discourse worlds to describe the adaption of the philosophical theory to readerly interactions that have narratological and cognitive dimensions.

Discourse worlds can have counterparts in the actual world, as well as other discourse worlds. For example, Shakespeare’s Richard III is the fictional counterpart of the actual Richard III King of England. And King Richard in Kill Shakespeare by Connor McCreery is the counterpart of Shakespeare’s character. Alternative discourse worlds are more or less close to the actual discourse world but vary depending on accessibility to objects, time, nature, and language. A principal of minimal departure is at play, so that unless the text specifies otherwise the assumption is that alternate discourse worlds are the same as the actual.

Furthermore, discourse worlds theory can be applied within the domain of the text itself. Using dreams and flashbacks, a text can offer counterparts in the same discourse world. Also, fictional characters themselves can be understood to have their own virtual discourse worlds which they use to interpret their experiences. These virtual discourse worlds are of six types: epistemic, speculative, intentional, wishful, obligational, or fantasy. This particular application of discourse worlds theory to characters in a text lends itself to use in creative writing, and below I link to a short story I wrote making use of this type of analysis.

In Cognitive Poetics In Practice, Elena Semino’s discussion of possible worlds theory references Umberto Eco and much of the ideas from this theory were familiar to me from a 2008 Richard Ellman lecture given by Eco. Eco relates the ethical function of literature to the ontological truth values of fiction. Fictional truth, Eco claims, can teach us important lessons about truth in the real world. The theory of possible worlds was developed by philosophers to calculate the truth value of sentences. According to Stockwell, discourse worlds extend the philosophical consideration to dynamic readerly interactions with possible worlds . Trans-world identity and counterparts within fictional discourse worlds seem especially salient to the problem of alternative worlds arising from domains of race and class, central concerns of my own writing.

Semino’s analysis of A Very Short Story by Ernest Hemingway using possible worlds theory takes into account tensions between the actual world, the discourse world, and the virtual discourse worlds of the two main characters. Her analysis operates at a more detailed level than is typical for possible worlds theory, tracing the changes in discourse worlds through each paragraph of the story. What is revealed is a rich private world of interpersonal and intrapersonal relationships. Last month, I posted a short story that I’d rewritten using a framework suggested by discourse worlds theory and Semino’s analysis. My intention is to continue revising the story, and I will likely use mental space theory and conceptual blending.

Mental space theory is offered as a supplement to discourse worlds, one that is grounded in cognitive linguistics. There are four types of mental spaces: time spaces, space spaces, domain spaces, and hypothetical spaces. Our mental representation of reality is called reality space and memory, imagination, and prediction operate on the reality space to create a projected space. Each of the four main types of mental spaces is characterized by linguistic features called space builders that open or focus on a space.

Mental space theory can be extended to narratives using the associated notion of conceptual blending that introduces two new types of spaces, a generic space and a blend space. For example, the reality space and the projected space are mapped onto a generic space that shares nodes and structural relationships with both. Features that emerge from that mapping create the blended space.

Understanding how to manage alternative possible worlds has implications for literary characters, who share similar cognitive apparatus with ourselves. Discourse worlds theory and mental space theory are ways of modeling how people comprehend stories, applicable to people as readers and people as characters. The six different types of discourse worlds listed above and the four types of mental spaces from conceptual blending are useful for representing the complexity of characters’ experiences and perspectives.

My approach to these theories and cognitive poetics in general is as a writer, interested more in the creative act than interpretation per se. A couple of months ago, I wrote that I wasn’t sure if this perspective had been explored in the research literature. Of course, very shortly I began to find multiple sources for research in this direction. One is Keith Oatley, a Canadian novelist and professor of cognitive psychology at University of Toronto. I am currently reading Oatley’s book, Such Stuff As Dreams. Another researcher writing about creativity and cognition is Charles Forceville, a professor at University of Amsterdam, whose research pertains to conceptual blending and creativity.

Forceville’s paper, Creativity in Pictorial and Multimodal Advertising Metaphors provides a description of blending theory (conceptual blending), including background information on antecedents to Gilles Fauconnier and Mark Turner, as well as case studies of several commercials and advertising billboards. The reason Forceville focuses on advertising is because in that domain the goal of the creator is well defined.

The creative task of the advertiser can be divided into several sub-tasks related to conceptual blending:


  1. select attributes of product (target domain) to be emphasized,
  2. choose a source domain with those attributes,
  3. imagine a scenario that includes the source domain and the target domain, and
  4. select mode of expression.

What Forceville emphasizes is the second step, choosing a source domain, because this is the primary creative act in creating a blend. He points out that creativity is more closely associated with the source domain than with the target domain or blended space. Very often the blend is a starting point, as may be the target domain, and from these the source domain is reverse engineered. For example, if a writer wants her readers to feel sympathy for a character, then she may set up a conceptual blend with that character as the target domain and sympathetic emotions as qualities of a blend. Her task then is to come up with a source domain, someone or something that is obviously sympathetic, and to link the source with the target in the narrative.

I have not approached my own writing this way, at least not deliberately. My characters usually come to me as themselves. I put them into language and their they are. But I have approached whole stories using a process that has emotional outcome in mind. The characters and plot have started out as abstractions, mere bullet points in an outline. The actual writing is where I’ve come up with the metaphors and idiosyncratic details that make up the story. But my goal in writing a story is not usually so well defined as the examples Forceville considers; I’m not selling a product or even selling a character.

Metaphors and blends encompass entire stories. Just as often as I write stories from an outline, I write stories that express emotions which may be unarticulated until the story is told. I can’t reverse engineer the emotion backwards into a story. But if I narrow my attention down to one part of the story or one aspect of the writing, then the mechanics of creativity are more apparent. At the sentence or paragraph level, I do have specific goals. In a presentation Forceville gave at Hong Kong City University, he includes a study of creativity in comics balloons along with the advertising case studies. For example, he describes Alan Moore’s use of wavering lines around the dialog balloons for the character Rorschach in Watchmen as a conceptual blend that helps create the disturbing quality of the character. The comic artists’ varied use of balloon styles relate closely with components of literary style a writer might use in creative writing.

I am now passed the halfway mark in my response to Cognitive Poetics: An Introduction and Cognitive Poetics In Practice, and the variety of concepts so far presented are beginning to interrelate in motivating ways. My response to earlier chapters, such as figures & grounds and cognitive deixis, were full of interest, but each concept seemed more or less independent of the other. As I continue reading and responding, it is apparent that these concepts function together, extending and supplementing one another to generate models useful for reading and writing. Next month I will respond to the chapters on conceptual metaphor, a central concept of cognitive science and one that underlies the concepts presented in previous chapters.

Comments are closed.