Rhetorical Minds

Meditations on the Cognitive Science of Persuasion

Todd Oakley

• Publisher: Berghahn Books
• Year: 2020
• ISBN:
  9781789206708
• Edition: 1
• Topic: Social Sciences
• Subtopic: Cultural & Social Anthropology
• Index: Social Sciences

Minds are rhetorical. From the moment we are born others are shaping our capacity for mental agency. As a meditation on the nature of human thought and action, this book starts with the proposition that human thinking is inherently and irreducibly social, and that the long rhetorical tradition in the West has been a neglected source for thinking about cognition. Each chapter reflects on a different dimension of human thought based on the fundamental proposition that our rhetoric thinks and acts with and through others.

PART I

Theoretical Prerequisites

CHAPTER 1

Starting Points

Paradigms, Models, Theories, and Metaphysics

Human meaning construction relies on mental spaces, one of the key analytic frameworks of Cognitive Linguistics. The notion of a mental space, first introduced in Fauconnier (1994, 1997) as a way of systematically tracking referential practices in discourse, has developed a wider range of applications, and thus definitions thereof have become notoriously slippery within cognitive science and Cognitive Linguistics. It is at once a conceptual packet deployed as we think, talk, and comprehend; it is a mental scenario used to frame a situation; it is a domain for disambiguating reference, which solves or dissolves many problems specific to the philosophy of language.¹

It is all these things but, in essence, a mental space is one way to capture our modes of engagement with the world and others; it captures the notion that we exist in a lifeworld of pragmatic actions and situations, much of which can be both pre-reflective and reflectively used and referenced. Elements of one space can be projected into another, thereby generating a new “blended” space. For instance, audience members can play along with the fiction that Clint Eastwood is interacting with President Obama without being deluded that he is manifest onstage all the while experiencing the speech as a lively exchange. Such fictive interactions are often best described as conceptual blends. More subtly, Kennan’s “long telegram” blends the registers of a diplomatic report with the compressed language of telegraphy, with the result being a substantial treatise bearing the characteristics of an urgent message.

Mental spaces are useful ways of modeling natural languages and discourse, but mental spaces are not linguistic in any narrow sense of the term. Linguistic material is organized by mental spaces, not the other way around.

The slipperiness of mental spaces is symptomatic of the studies of meaning, as there is a welter of terminology in the bramble of meaning construction, not the least of which is the meaning of “meaning.” My immediate task is to disentangle some of the ambiguities inherent in these enterprises, for it is not unusual for mental spaces to be regarded at one time as a theory, at another as a framework or vocabulary, and still another as a model. (I prefer model.) What is more, Cognitive Linguistics is sometimes referred to simultaneously as a “theory,” “school,” and “movement.” To add to the confusion, we hear of ‘theories’ of embodiment (e.g., Johnson 1987; Lakoff and Johnson 1999; Shapiro 2011), and of the enactive “paradigm” in cognitive science (Stewart et al. 2010). To compound the confusion, practitioners in the cognitive sciences take notions like “mental representation” as axiomatic and without the need for specification. Finally, most cognitive scientists and cognitive linguists embrace some form of “naturalist” (a.k.a. “materialist” or “physicalist”) metaphysics without argument.²

Without presuming to be definitive, I hope to offer at least one possible path through the thicket of terminology in the hope that we can avoid fruitless wrangling. At the end of this chapter, readers should at least know what I mean by the following contested terms: paradigm, model, theory, naturalism, realism, and representation.

Three Waves of Cognitive Science

A mature science, so it is thought, goes through phases of alternating normality and revolution, with normality being a period of consensus on what counts as proper scientific research. During periods of normality, exemplars of good science epitomize what it means to be scientific, and thus, these exemplars are paradigmatic of the entire enterprise. During revolutionary periods, however, no such consensus is readily available, and everything is up for grabs (Kuhn 1962). In the cognitive sciences, the question of where the mind stops and the world begins is at the heart of programmatic debates over the boundaries of cognition, or what some call “the mark of the cognitive.”

One dominant narrative in the cognitive sciences goes like this: there was a revolutionary period in which behaviorist doctrines of psychological explanation were shown to be bankrupt; at the same time, computer engineering demonstrated the proof of concept that complex human thinking can be explained algorithmically. The mind is not like a computer: it is a computer. This insight led to new normality of computational and component models of mind. This first-wave cognitive science takes as axiomatic the notion that the boundaries of cognition are intracranial.

It soon became apparent to many researchers that the computation theory of mind was not delivering on the explanatory promises made at the height of the cognitive revolution. It became increasingly apparent that the things we call cognitive emerge from bodies at rest and in motion; and that much of the abstract features of cognition and conceptualization has a bodily basis; most of our prized concepts can be traced to a “body moving in space.” So was initiated the paradigm of “embodied cognition,” manifest most conspicuously in the George Lakoff and Mark Johnson’s conceptual metaphor theory and related programs. This second-wave cognitive science takes as axiomatic the notion that the brain represents the body and the body, in turn, represents the world.

As the embodiment approach became increasingly influential in linguistics and psychology, another revolution began to take shape in artificial intelligence and robotics employing subsumption control systems. Intelligent behavior emerges from relatively simple sense-act sequences, obviating the need for extensive central planning (i.e., planning is subsumed by these sense-act sequences). For roboticists such as Rodney Brooks, intelligent behavior need not (and indeed cannot) depend on architectures whereby every movement and action is planned out and modeled internally before its execution. It makes little sense to have organisms, even complex organisms like us, operate according to an internal model when, in fact, the world provides the information necessary for the organism to interact successfully with it.³ For much intelligent behavior, as Brooks famously put it: “it turns out to be better to use the world as its own model” (1991: 139).⁴

Brooksian robotics fits broadly (but cannot be fully identified with) the third revolutionary wave of distributed cognition, whereby cognition is not only embodied but that the very notion of mind itself is “extended” into the environment. Andy Clark and David Chalmers (1998) are perhaps the most renowned proponents of the programmatic view that cognition needs to be extended out into the world. But there are other exemplars as well, most notably, Anthony Chemero, Edwin Hutchins, Daniel Hutto, Richard Manery, Erik Myin, Alva Noë, Mark Rowlands, Kim Sterelny, Evan Thompson, and Michael Wheeler, all of whom, despite their differences, draw inspiration from the phenomenological tradition, and each of whom proffers a version of enactive cognition.

For enactivists, perception is not inferential and passive, but habitual; it is something we do and not something that happens to us. Proponents of enactivistism take as their founding presupposition that sensory experiences (i.e., “qualia”) are consequences of the organism’s exploratory activity. Moreover, this engagement always occurs in tandem with a repertoire of past experiences and valuations that guide its interaction with the environment. Our “smooth coping” with the world is constituted by a history of our “structural coupling” with the manifold versions of the world into which we are “thrown.” More importantly for these purposes, it is something that human beings also do together, a point not sufficiently captured in this approach.

To summarize, first-generation cognitive scientists revolted against the normalizing practices of both methodological and logical behaviorism, making it once again respectable for cognitive psychology and allied disciplines to posit mental representations as the necessary machinery of mind. Then, the second generation of cognitive scientists revolted against the new normalizing practices of what John Haugeland (1985) dubbed “good old-fashioned artificial intelligence” (GOFAI) to reassert the importance of human bodies to human thought. The way we think is intimately conditioned by the way we exist as embodied agents. The third generation of cognitive scientists then revolted against the notion that cognition and mind can adequately be explained by body and brain alone. The mind has to extend or expand beyond the skull and skin. Good cognitive science must presuppose that bodies in rich artifactual and social environments are constitutive of minds. Taking their inspiration from Hilary Putnam (1981), brains in vats do not constitute minds; at best they merely enable them.

This portrait of cognitive science is painted with very broad strokes and should not be taken as anything more than an impressionistic sketch designed to orient the reader. Nuanced detail has been left for later. One thing is clear, cognitive sciences are plural because there is no single consensus on what counts as appropriate cognitive science.

But here is the relevance to this project. As it will be clear in sub-sequent chapters, the account of Homo rhetoricus offered here fits most comfortably within the third wave of distributed cognition. If a reader is expecting an account of cognition to mirror that of the first wave and the related theories of massive modularity of mind, or of evolutionary psychology, then your expectations will be unmet. The same is also true if you think the human cognition is something that can be stuffed inside the head, such that our explanations of human cognition will ultimately be reduced to neurons, neurotransmitters, and synapses. Explanations that include the organism-and-environment as the proper unit of analysis seems at present to be the most fruitful form of second-person cognitive science.⁵ Models of human agency must expand to include multiple agents linked together in symbol rich environments.

Models and Modeling

Of central importance to science generally, models are continually being built, tested, compared, and revised within the cognitive sciences. There are computational models, developmental models, idealized models, heuristic models: you name it, we’ve got ’em.

What, precisely, is a model? Models are representative substitutes. They provide access to facets of the world, making them, in effect, “graspable.”⁶ Thorvaldsen’s Copernicus qualifies as a model insofar as it enables us to “grasp” the man and his idea. An empty chair is a model intended for us to grasp the essence of President Obama as a political persona. The saltshaker and peppermill on my dining room table can be a model of, say, the relative location of Terminal Tower in Cleveland, Ohio (the peppermill) to Lake Erie (the saltshaker).

All these are models, but here is the key point: They are only effective models in correspondence with a specific context and purpose.

Statuary is an effective modeling device for accessing persons only if such features plausibly substitute for the entity being modeled. An empty chair can be an effective model of a living person if circumstances permit. Saltshakers and peppermills can be useful models of terrestrial landmarks if the purpose of modeling is for accessing a geographic relationship between two static objects. The peppermill is woefully inadequate for modeling architectural features of Terminal Tower (unless the peppermill itself made to resemble it), just as the saltshaker is insufficient for modeling facets of a large body of water. Models are models for specific purposes; one cannot divorce a model from its practical contexts. They do not carry their own inviolable content.

Spoken and written language itself is perhaps best regarded as a modeling tool, par excellence. Mental spaces are the metamodeling products of language and other semiotic systems that enable effective disclosure of the world and its diverse projects. That is, higher-order human cognition issues from the practice of engaging the world not as isolated objects and events but as scenes and scenarios, and mental spaces are ways of modeling how we think and talk. This view is consistent with rhetorical theorists like Kenneth Burke and Chaim Perelman and Lucie Olbrechts-Tyteca, among others.

Here is an illustration of online modeling behavior.

When George Kennan sent his long telegram, he was acutely aware of his rhetorical situation—using the normally compressed mode of communication to send what would normally arrive by diplomatic pouch. Kennan may have modeled the typical scene in which such policy arguments are issued, which thereby enabled him to anticipate resistance from the intended readers. This is one explanation for the presence of the plea that they “bear with” him. Kennan operates according to a common model of discourse expectations, compares the present situation to that model, and uses the discrepancy between the two, prompting him to persuade them to read on, despite the unconventional length. Besides, Kennan also was working under a model of situations in which communication through regular routes is more likely to be ignored. The unfamiliar becomes an attentional advantage, he reckoned. Kennan could not act this way without a working model of how other agents interact in the niche they occupy. He has developed a network of relevant mental spaces (particularly the mental spaces for communication protocols of telegraphy (hurried) and diplomatic carriage (languorous) of the discourse situation. These two media serve as different contexts against which his ideas about the Soviets are to be staged.

A mental space is best understood as a model (rather than a theory), but it is not necessarily a model that has to fully embrace either of the above social cognitive accounts of other minds. That said, a mental space is, however, a representational model, insofar as mental spaces must account for meaning construction as a consequence of natural language and other systematic semiotic activities. Mental spaces are at once the post hoc models some investigators use to understand human meaning-making and a reflection of a significant portion of the actual processes used in meaning construction. I contend that this approach will be on firmer ground if we take a significant portion of “mental” beyond the cranium.

Theory and Theories

Both rhetorical theory and cognitive science are awash in theory. I mean the term, “theory.” Among rhetoricians, you can find books on neo-Aristotelian theory, feminist theory, and so on. We encounter the dual coding theory of vision or the prototype theory of categorization. Linguistics and related disciplines serve up multiple courses of conceptual metaphor theory, conceptual blending theory, theory of perceptual symbols, and, from a different school, principles and parameters theory, relevance theory, and so on. The work itself is valuable and influential, but “theory” is used ambiguously in these endeavors as a placeholder for an entire intellectual activity, or as a specific outcome of that activity. This book is a work of theory if by the term we mean something like the following definition: a research program with identifiable intellectual characteristics. Second-person cognitive science is not a specific theory if by the term we mean something like the following definition: a set of predictions, hypotheses, and methods for testing specific hypotheses.

With this distinction in mind, let us focus once again on the twin notions of mental spaces and conceptual integration. Neither of these is a theory in the narrow, scientific sense. For a theory to be a theory, it has to generalize our understanding, which can also lead to generating predictions. Again, a mental space is best understood as a model that presents significant improvements in our understanding of natural language and other semiotic phenomena as encountered in realistic situations. In this regard, problems long associated with the philosophy of language (referential opacity; role value readings; presupposition; conditionals, counterfactuals, and analogical counterfactuals, and so on) find an elegant solution within the mental spaces model and, thus, contribute to theorizing in the first sense stipulated above.

While it is true that mental-spaces analysis yields general theoretical insight, mental-spaces analyses, in themselves, are not predictions about specific phenomena that can then be tested according to research hypothesis, the results of which are then compared to an alternative and null hypothesis. For instance, referential opacity (e.g., “Pussy Galore thinks that James Bond is debonair” vs. “Pussy Galore thinks that Agent 007 is debonair”) is a phenomenon elegantly modeled by mental spaces. Thus, the account of referential opacity as involving two distinct referential scenes, each with a contrasting epistemic status (e.g., either Pussy Galore knows that James Bond is Agent 007, or she does not). A mental spaces model of referential opacity can then be used to generate specific testable hypotheses on how language users cognize references according to particular epistemic states. For instance, researchers can design referential tasks based on the manipulation of the participant’s background knowledge. But it cannot make a specific prediction about specific referential processes in the future beyond statistical likelihood. The upshot is that theory in the narrow sense tends to operate principally, though not exclusively, in the realm of experimentation.

Mental spaces are models; blended mental spaces—the integration of two or more mental spaces—are also models, but neither are theories in the narrow sense as bequeathed from physics and chemistry. These models do not, in themselves, offer specific predictions or generalizations. Paradigmatically, both Gilles Fauconnier (as the originator of mental spaces) and Mark Turner (the co-originator with Fauconnier of conceptual integration or “blending”) reject strong modularity of mind hypothesis, yet many who apply or reference this model to the study of specific phenomena also have embraced stronger modularity of mind hypotheses. There is no contradiction in using a model for diverse paradigms that posit different theoretical stances ...

Table of contents

• Cover
• Title Page
• Copyright Page
• Contents
• List of Figures
• Acknowledgments
• Introduction
• Part I. Theoretical Prerequisites
• Part II. The Evolution and Development of Homo Rhetoricus
• Part III. Discourse and Social Ontology
• Concluding Remarks. Homo Rhetoricus and the Mark of the Cognitive
• References
• Index