Conceptual Integration and Fictive Interaction

By: Todd Oakley, Case Western Reserve University


Conceptual integration (also known as Conceptual Blending) is a general cognitive operation underlying virtually all acts of human creativity. First explored by Gilles Fauconnier and Mark Turner and subsequently by a range of scholars (see for a list of works), we can posit conceptual integration as a “cognitive” universal underlying linguistic and literary phenomena. More specifically, Fauconnier and Turner identify Double-Scope Conceptual Integration (see below) as a panhuman singularity, which is not shared by even our closest living primates; at least there is little to no evidence double-scope blending among non-human primates or cetaceans, either in captivity or in the wild.

That double-scope conceptual integration is a cognitive universal is fairly uncontroversial and ultimately not very interesting as a claim in itself. What is of more interest are the kinds of generalizations to be made about the literary typologies that conceptual integration theory (CIT) can develop based on the general principles developed therein. The fact that human beings “blend” is a universal, and that some outcomes of this blending process, such as primary metaphors (e.g., AFFECTION IS WARMTH, IMPORTANCE IS SIZE, and SIMILARITY IS SPATIAL PROXIMITY; see “Metaphor Universals in Literature”) and other forms of “simplex” blending may be universal or near-universal, it is most likely the case that specific products of conceptual integration are culture-specific, for blending is first and foremost a process for solving local conceptual problems. Floating in the sea of variation, however, is one candidate for a universal blend: fictive interaction. Put succinctly, fictive interaction is the idea that speech can be a construed as a property of non-actual human actors as well as non-human actors, be they animate or inanimate, natural or artificial.

In this article, I will describe the model of conceptual integration and link details thereof to the question of literary universals. I propose the blending phenomenon of fictive interaction (Pascual; Pascual & Sanders) as an implicational universal and propose future research on the various manifestations of narrator types.

General Principles

Conceptual Integration Theory begins with the basic assumption that human beings think in scenes and scenarios. Gilles Fauconnier calls these scenes as scenarios mental spaces, making CIT an outgrowth of mental spaces theory, the term for a model of natural language that solves several puzzling referential and related phenomena that are difficult for model-theoretic and truth-conditional semantics, such as referential opacity, presupposition, and analogical counterfactuals. CIT adds the observation that human beings not only integrate information within mental spaces but that many if not all conceptualizations involve the networking of multiple mental spaces. In a canonical form of the model, a CI network involves at least two “input” spaces that, in addition to creating cross-space mappings between input elements, selectively project information from each input into a third, “blended” space. This space, in turn, develops its own emergent structure as well as develops according to its own unique logic. Examples abound (see Fauconnier and Turner, Oakley and Pascual, for extensive overviews and examples), but perhaps we can illustrate it with this little quip from the narrator when describing the thoughts of Little Chandler in “A Little Cloud”: “He tried to weigh his soul to see if it was a poet’s soul” (83-4).

Here we have a network of two input spaces, one for SOUL (broadly construed) and one for WEIGHTS & MEASURES. The dominant element of the SOUL space is “disposition” or “personality” (which is quickly identified as “melancholy”). The WEIGHTS & MEASURES space remains unspecified, as no attempt is made to specify “ounces,” “pounds,” “tons” or whether such units are imperial or metric. The blended conceit that emerges is the expectation that one’s soul has a measurable “weight” that can be “converted” into some vocational ratio of creativity, none of which is forthcoming. In this instance, the emergent structure of the WEIGHING of SOULS blended space, not available in either input space, is the conceit that creative dispositions are measurable and can be “objectively” assessed, just as one can objectively assess the goods according to standardized units.

However vainglorious and whimsical Little Chandler’s conceit may be, it illustrates a universal feature of human conceptualization: the propensity to create scenes that variously “compress” or “decompress” relationships of time, space, cause-effect, change, identity, analogy/disanalogy, part-whole, representation, category, intentionality, and uniqueness to a human scale (Fauconnier and Turner; Oakley and Pascual, 427-8). Human beings live at a scale where it takes considerable ingenuity to conceptualize that which is diffuse and that which is condensed over time and space: we prefer to think about these topics in ways that comport with our human scale experiences. It is therefore tempting to emotionally manage the vague ambition of becoming a great poet by integrating it with a familiar scenario of anyone experienced with industrial, metropolitan life. Little Chandler hopes that a sufficient “quantity” of melancholy provides sufficient weightiness to constitute a poetic sensibility, just as a sufficient quantity of gold constitutes wealth.

Little Chandler’s soul comprises what CIT calls a double-scope blending network, for the two organizing frames governing each input space issue from incommensurate and potentially clashing conceptual domains. The fact that the blended conceit lasts only for a single sentence means that it is abandoned immediately after its composition so as not to be completed or elaborated, such that the narrator would then have to manage the inevitable clashes between the two input spaces. In other words, this is a conceptual blend that colors the narrative without playing a significant role in advancing the story. In the parlance of CIT, it exemplifies the process of composition without the subsequent processes of completion and elaboration (Fauconnier and Turner, 89-113; Oakley and Pascual, 428-9).

Perhaps most important for the empirical testing of CIT is the fact that the model posits six optimality constraints[1] to underscore the context-dependent nature of conceptual integration. That is to say that conceptual blends like the WEIGHING OF SOULS develop and succeed or fail through processes of “satisfying” or “relaxing” principles or constraints. Conceptualization is, therefore, best seen as a process whereby concepts develop by an uneven, ad hoc adherence to specific constraints.

The Integration constraint stipulates that a blend should form a coherent scene. Little Chandler’s mental machination of weighing properties of his soul, however fantastic, forms a coherent scene, as the experience of mental assessment is habitually construed in these terms.

The Topology constraint stipulates that it is optimal to preserve elements in the input spaces in the blend. That is, the blend should share as much information with the input spaces as possible. If we are Cartesians, the substance of souls has no discernable terrestrial properties. It is fairly easy to see the task of weighing souls as fanciful, since the two domains are often construed as quite distinct, despite it being a deeply entrenched cultural gestalt. In this instance, the topology constraint is relaxed, except that we are invited to construe “melancholy” as a measurable substance, and, in the next sentence, are entrained to regard “resignation” as a countervailing substance.

The Web constraint hypothesizes that the cross-space mappings between inputs should be maintained in the blend. The only cross-space mapping in this blend is the relationship between “melancholy” and “measured substance,” as there is no development of any other properties of the soul and their corresponding units of measure elaborated in the narrative.

The Unpacking constraint states that it should be relatively easy to infer the structure of the input spaces from the blend itself. (This constraint works at cross purposes with topology and integration.) Because the two input spaces share little to no conceptual structure, it is relatively easy for readers to unpack the two spaces without difficulty. Perhaps Joyce wishes readers to see this conceit as fanciful and unrealistic.

The Metonymy constraint stipulates that elements in the network that are related through contiguity or relations of similarity collapse to identity in the blend. There is no evidence of what we call “metonymic tightening” in this version of WEIGHING OF SOULS other than the possibility that “melancholy” comes to stand for the poetic sensibility. That said, metonymic tightening is a pervasive process of compression to human scale, and it is especially important in cases of pictorial blending (see Oakley and Pascual, 436-7).

The Good Reason constraint holds that any element that happens to emerge in the blend will be given significance, no matter how incidental or tangential. The fact that melancholy is the only significant personality trait mentioned after the weighing of soul quip creates pressure to regard it as a measurable substance.

Fictive Interaction as a Universal Form of Conceptual Integration

The staggering variety of blends human beings construct, both ephemeral and perduring, makes it difficult to identify a typology that is not culture-specific. One such candidate of blend type that seems has achieved maximal generality is the phenomenon known as Fictive Interaction (hereafter, FI). Developed most extensively by Esther Pascual, FI names the process of using the entrenched frame of a conversation to structure all manner of states of affairs that are not, in themselves, conversations. For example, in Catalan the phrase, El cos diu prou, is commonly used to express “exhaustion,” by literally saying, “The/my/your body says stop” (Pascual and Oakley, 349). Notice that exhaustion is not ipso facto a conversational state of affairs; rather the speaker structures the description of a physical state in terms of a performative speech act, where the body becomes a speaker issuing a command. What is more, native speakers of Catalan often substitute cos [body] with other body parts, such as cor [heart], ulls [eyes], and peus [feet] to express slightly different or emotional states. [2]

The construal is fictive in the sense of being non-actual. The tired muscles in the body are not literally addressing the person controlling muscle movement, but it is not fictional in a strong sense; the states of affairs being referenced are real. One is simply using a frame of conversation to represent a real state of affairs that is not objectively a conversation. Pascual has provided significant evidence that FI patterns are found in all languages she has investigated, and at all levels of analysis, from discourse all the way down to morphology (29-58; 83-112; Van der Voort). Interestingly, FI seems to play a greater role in the structuring of oral languages than chirographic ones, and since there are far more examples of purely oral languages in existence than written ones, it is reasonable to conclude that FI is a form of linguistic universal based on the cognitive universality of the conversation frame.

With respect to CIT, FI as a general phenomenon forms a simplex network (Fauconnier & Turner, 39-59) in which a single mental space is organized by facets of the conversation frame. In the example above, it is organized around a fictive addresser, but there are versions involving fictive addressees, as well, as in My legs are not listening to me right now. Of course, many actual instances of FI can be readily completed and elaborated into fuller mirror, single-scope, and double-double scope integrations, as when the Cross speaks to a Thane in Dream of the Rood.

FI as a Narrative Universal

Two studies are worth presenting in some detail, as they show how the conversational frame is used as a discourse structuring device in distal cultural and historical formations. The first is Mingjian Xiang’s study, “Real, Imaginary, or Fictive?: Philosophical Dialogues in an Early Daoist text and its Pictorial Version,” offers extended treatment of Zhuangzi, the foundational Daoist text originating in the fourth-century B.C.E.; the second is Christopher Flint’s  article “Speaking Objects: The Circulation of Stories in Eighteenth-Century Prose,” which investigates the trend of using objects as narrators in fiction.

As with Platonic dialogues, Zhuangzi exhibits a series of dialogues that are meant to be read allegorically in order to illustrate Daoist principles. Unlike Platonic dialogues, however, the discussants extend not only to animals and deities but to inanimate entities, such as the wind, and abstract concepts (Xiang, 73). According to Xiang (who provides a detailed CIT analysis of these FI constructs of greater depth and precision than presented here), these FI structures are instances of yu yan (“imputed words”). One instance of yu yan appears when the author of the text has Confucius and his devotee, Yan Hui, engage in a dialogue wherein Confucius proceeds to espouse several Daoist propositions, which are, as Xiang reminds us, more often than not in opposition to Confucianism. Thus, we are confronted with a historically real personage, Confucius, having a non-genuine conversation. These forms of “adulterated speech” are commonplace across literary traditions.

More conspicuous are instances of FI between fantastical entities, my favorite being the exchange between shadow and penumbra:

Penumbra said to Shadow, “A little while ago you were walking, and now you’re standing still; a little while ago you were sitting, and now you’re standing up. Why this lack of independent action?” Shadow said, “Do I have to wait for something before I can be like this? Does what I wait for also have to wait for something before I can be like this? Am I waiting for the scales of a snake or the wings of a cicada? How do I know why it is so? How do I know why it isn’t so?” (quoted in Xiang: 73)

The fictional dialogue between these two personified entities is fictional in structure but serves a fictive purpose of illustrating a Daoist response to such philosophical questions as the nature of volition. It is an indirect means of expressing the writer’s position by means of “showing” rather than “telling.”

The Zhuangzi text is a tour de force of FI conceits involving both realistic but non-actual (disingenuous) speakers as well as fantastical, non-actual but genuine natural objects and entities. It is worth noting, however, that none of the speakers comes from the world of artifice.

Christopher Flint’s article offers up a concise and penetrating account of the underlying motives for a literary trend that began with the publication of Charles Gildon’s, The Golden Spy in 1709. A picaresque narrative told by a few “quarrelsome coins” who boast about their national origins and then relate their “transmigrations” across borders, these coins transform into first-hand witnesses of the exchanges of men, all the while highlighting the thoroughly itinerate and extemporaneous nature of their existences.

Flint, who does not couch his analysis in terms of CIT, claims that “A remarkably persistent feature in eighteenth-century fiction, the narrating object appears in a surprising number of satires published between 1709 and 1824, manifesting a particular cultural obsession with stories as things” (214). One could go so far as to suggest that this trend makes good on the saying “If these walls could talk,” as all the manufactured objects in question float easily and inconspicuously between machinations public and private. Examples of object narrators in his study include a goose-quill, a post-chaise (carriage), a pin, and a kite. Once again, such narrative integration networks optimize the metonymy constraint, insofar as they are information bearing artifacts. A goose-quill tells intimate tales; a post-chaise not only witnesses the happenings of its passengers but presumably has access to the contents of its mail; a pin finds itself fastened to all manner of intimate apparel, and a kite functions as an eighteenth-century equivalent of a surveillance drone. As proverbial “flies on walls,” these speaking objects satisfy the good reason constraint. In the situations of human exchange, each is typically regarded as incidental or a mere means by which something is done. Coins are a means of exchange; quills make writing and correspondence possible; pins hold fast that which is the object of attraction to eligible suitors, and kites amuse their terrestrial handlers. By shifting the perspective or vantage point to the objects themselves, their significance as bearers of sensitive information becomes apparent. We are forced then to see the mental spaces from their perspective.

Flint’s meta-literary argument is worth emphasizing here. The trend of using manufactured artifacts as narrators coincides the rise of print culture and the legal-institutional conditions of the author as producer of goods and services and its attendant alienating effects. Just as the coins have no control over their transmigrations and exchanges; just as the goose quill becomes complicit in correspondences not of its choosing; just as a post-chaise cannot choose its passengers or parcel; just as a pin does not choose that to which it fastens; and just as a kite is at the mercy of wind direction and speed, so are authors at the mercy of other agents, agencies and readerships in this age of mechanical reproduction. The author becomes a commodity. As Flint takes care to emphasize, the speaking object’s status as a commodity is likewise emphasized throughout these narratives. In The Adventures of a Cork-screw, the narrator often interrupts the story with descriptions of its users and the uses to which it is employed, and at other times, key moments in the narrative are interrupted by the fact that their owners, in order to satisfy some debt, reach in their pockets and hand over said cork-screw; as the conditions of ownership change, so does the story. One can see the attraction of using manufactured objects as such picaresque genres.

 Implicational Universals

We are now in a position to make some speculations about nature of FI blends across oral and literary traditions. FI appears to be an organizing scheme at work at various levels of linguistic analysis, which includes that of discourse and narrative structure. The impulse to endow non-human entities, such as deities, animals, and the natural world–especially boundary and orientation landmarks like mountains, rivers, streams, lakes, and seas–appears to be part of a cultural universal of myth-making. Animism as the dominant metaphysical stance of ancient and traditional societies, wherein practices of daily survival involve the interaction and worship of these objects and beings, leads us to conclude that FI with these types of entities is a mythic universal, one still available as a special trope for denizens of modern, industrial civilizations. The River Liffey, for instance, speaks in Finnegan’s Wake.[3] It is reasonable to speculate that “wind” is a preeminently common candidate to be a fictive speaker or narrator.

Flint’s investigation into the role of manufactured objects as narrators suggests that in industrial societies where strangers regularly interact through information bearing artifacts, there will be a noticeable shift in range of inanimate narrators to the world of the artificial, all the while maintaining a link to its traditional, animist past. Thus, we have an implicational universal, such that all known human societies and cultural formations engage in fictive interaction with nature, so that if a society has FI scenarios involving artifacts, then it can draw on FI scenarios involving nature; however, there should not be any societies in which their oral and literary traditions exhibit robust FI scenarios involving artifacts without at least having vestiges of FI scenarios with nature. Why might this be so?

While all human cultures produce artifacts, not all societies produce a critical mass of symbolic, information-bearing artifacts of the type that can “speak louder” or more authoritatively than those natural entities on which life, love, and worship depend.

Within societies exhibiting artifact-based FI scenarios, there should be a hierarchy of more or less likely narrator functions. Information-bearing artifacts, such as texts, coins, forms of identification, cards, and other means of exchange are particularly powerful, as it is natural to regard them as “speaking” through the process of metonymic compression. Articles of clothing and intimate apparel, as well as toys and other ludic gadgets, appear to be strong candidates. Expressive artifacts, such as musical instruments, are another likely candidate for widespread adoption. Add weaponry to the list of candidates.

In contrast to many natural FI entities that get their prominence from their status as stationary, immovable landmarks that can enable or frustrate human agents, artifact-based ones tend to gain prominence from their transmigratory dispositions. This is especially important in advanced, industrial and postindustrial societies that have elaborate configurations of public and private spaces. The fact that these entities can migrate seamlessly and inconspicuously between the domains of public performance and intimate interaction makes them fodder for all manner of narrative mischief.

At the same time, it is possible to regard some artifacts as less likely (but not impossible) candidates. Such artifacts would include meat and other perishable foodstuffs, and items that are not countable or easily individuated, such as sawdust, straw, and other raw material, in part because they are not easily integrated into coherent scenes and scenarios.

Future Research

For the most part, the future research suggested by the preceding hypotheses is straightforward. There are many cases of FI in other traditions. The first task for research in this area is to identify further instances, exploring the nature, varieties, and functions of FI through multiple instances. The second is to examine the necessity of natural FI entities as a step leading to the adoption of artificial FI entities. The third is to determine the frequency of classes of natural or artificial FI entities. There are also further research questions that arise in this context. For example, one set of questions concerns the diachronic study of cultural and language contact. Do we see a shift in oral and literary storytelling among erstwhile traditional societies that come into subsequent and sustained contact with industrialized and postindustrial societies? I have in mind aboriginal cultures of the Australian outback or Native American tribes, most of which live hybrid lives. What does that change over look like? Are money and other tokens of exchange a likely candidate for early adoption of artifact-based FI scenarios?

Works Cited

Fauconnier, Gilles. Mental Spaces: Meaning Construction in Natural Language. New York: Cambridge University Press, 1994.

Fauconnier, Gilles and Mark Turner. The Way We Think: Conceptual Blending and the Mind’s Hidden Complexities. New York: Basic Books, 2002.

Flint, Christopher. “Speaking Objects: The Circulation of Stories in Eighteen-Century Fiction.” PMLA 113.2 (1998): 212-226.

Joyce, James. “A Little Cloud.” In The Portable James Joyce. Ed. H. Levin. New York: Penguin Books, 1977: 80-96.

Oakley, Todd and Esther Pascual. “Conceptual Integration.” In The Cambridge Handbook of Cognitive Linguistics. Ed. B. Dancygier. Cambridge, UK: Cambridge University Press, 2017, 423-448.

Pascual, Esther. Imaginary Trialoques. Utrecht: LOT Publications, 2002.

Pascual, Esther. Fictive Interaction: The Conversation Frame in Thought, Language, and Discourse. Amsterdam & Philadelphia. John Benjamins Publishing, 2014.

Pascual, Esther and Todd Oakley. ‘Fictive Interaction.” In The Cambridge Handbook of Cognitive Linguistics. Ed. B. Dancygier. Cambridge, UK: Cambridge University Press, 2017, 347-360.

Pascual, Esther and Sergeiy Sandler, Ed. The Conversation Frame: Forms and Functions of Fictive Interaction. Amsterdam & Philadelphia: John Benjamins Publishing, 2016.

Voort, Hein van der. Recursive inflection and grammaticalized fictive interaction in the southwestern Amazon. In The Conversation Frame. Ed. E. Pascual and S. Sandler. Amsterdam: John Benjamins, 2016, 277-299.

Xiang, Mingjian. “Real, Imaginary, or Fictive?: Philosophical Dialogues in an Early Daoist Text and its Pictorial Version.” In The Conversation Frame: Forms and Functions of Fictive Interaction. Ed. E. Pascual and S. Sandler. Amsterdam & Philadelphia: John Benjamins Publishing, 2016, 63-86.


[1] They are variously referred to as “principles” or “constraints” throughout the CIT literature. I prefer “constraint” and will use it throughout.

[2] I thank Esther Pascual for bringing this example and its variations to my attention.

[3] I thank Patrick Hogan for calling this example to my attention.