An enactivist account of the dynamics of lying

Abstract

Enactivist accounts of communication have focused almost exclusively on honest, cooperative communication. However, much of human life involves deception and lies. Using the generally agreed upon definition of lying, we here develop an enactive account of the dynamics of lying. At face, lying poses a problem for enactive theories of cognition since lying seemingly requires the ability to represent counterfactual states of affairs and implant those representations in other agents' belief systems. On our account, lying involves the active manipulation of the short- and long-term dynamics of social cognitive systems so that agents have access to different sets of affordances from the one’s they counterfactually would have had access to without the lie. Representing truths and falsehoods are replaced with competency within social-cultural and material practices.

Introduction

Enactivist accounts of cognition have almost exclusively focused on honest communication. However, a large part of human life involves lies and deception. Many scholars argue that learning to lie and deceive are important healthy developmental steps that should be achieved in childhood (Ding et al., 2018; Jakubowska & Białecka‐Pikul, 2020). In fact, research suggests that people in the United States lie 1.6 times per day on average (Serota et al., 2010). While lies and deception are an important part of our lives, the growing framework of enactivism has so far not provided an in-depth account of the dynamics of lying or deception. This is in part because lies and deception are said to be so called “representation hungry” processes (Clark & Turibio, 1994; Edelman, 2003). To lie requires that the liar somehow “knows” the truth and can insert the falsehood into a network of beliefs, and agentive practices—in this respect lying is an ability that seemingly requires a lot of wit, metacognitive ability, and representational capabilities. Enactivism, however, is a non-representationalist theory of cognition. Thus, developing an account of lies and deception is a challenging and wanting project for enactivists.

Two caveats before we begin. First, we are not developing a new definition of what it means to lie. Rather, staying with a common philosophical definition of lying, we provide an enactivist description of how lying takes place cognitively. Second, it is important to understand that deception is a broad umbrella category containing several kinds of deceptive acts including but not limited to distraction, fabrication, omission, bullshit, catfishing, and of course lying. In this paper, we focus on the enactive dynamics of lying. To provide an account of the dynamics of lying, we argue from the perspective of autopoietic enactivism (Di Paolo et al., 2017; 2018; Gallagher, 2017; 2020; Varela et al., 1991). This perspective allows for the integration of heterogeneous data from different fields in one overarching framework. For example, although our intention is not to provide a theory of lie detection in this paper, we do look at the empirical literature on lie detection since this literature provides theories of the cognitive production of lies. Furthermore, enactivism is an embodied theory of cognition, and the lie detection literature has long looked at the embodied factors involved in producing lies. In addition, lying is an intersubjective phenomenon. Thus, we need to look across agents to understand the cognitive processes that underpin lying. Put differently; lying is a relational process between at least two people (perpetrator and victim). Consequently, we look at the cognitive processes of the liar and their victim(s).

Enactive accounts of cognition oppose representational explanations. Although there are many different conceptions of mental representation (Ramsey, 2007; Rowlands, 2017; Smortchkova, Dołrega & Schlicht, 2020), there are some common characterizations that define the concept. Mental representations are often understood as:

[Mental] objects with semantic properties (content, reference, truth-conditions, truth-value, etc.), which fulfill some conditions of satisfaction (truth, accuracy, etc.) and thereby determine the meaning of a thought (Rucińska, 2016, p. 3, p. 3)

Representations are said to be involved in syntactical operations akin to the logical operations of computers to produce content and other representations that may control behavior and issue in embodied actions (Carruthers, 2006; Clark, 2016; Fodor, 1975; 1984; Kirchhoff & Kiverstein, 2019; Stanley, 2011; Vesper et al., 2017). This is true of older forms of cognitivism as well as the newest, most cutting-edge accounts of predictive processing and sensorimotor control (Kirchhoff & Kiverstein, 2019; Ramstead et al., 2020; Shepherd, 2021). Although there are deflationary accounts of representations (Egan, 2020) and conceptions of representations that are not as heavily semantic—for example, neural covariation, action-oriented representations (Clark, 1997)—the more traditional semantic, truth-conditional version seems more relevant to the issue of deception. At first glance, it seems that lying requires the use of mental representations since it involves creating or simulating some version of events that is counterfactual and comparing it against true facts about the world before producing an utterance.

To move away from an account of lying that relies on the having of mental representations in the form of “beliefs” or internal mental states, we develop an account of the dynamics of lying that appeals to affordances, embodied intentions, habits, and the non-linear dynamical systems that span brain-body-environment. In this context, we argue for an account of enactive “sense making” in which lies are thought of as multifaceted speech acts. We are generally committed to the following:

• Lying is an embodied linguistic action (speech act) that perturbs the short- and long-term dynamics of an intersubjective communicative system. Deception is thus part of a holistic physical process “in the world.”

• Lying functions by preempting affordances or creating affordances for other agents that would not emerge without the deception. As a consequence, the liar and the victim end up with different fields of affordances.

• Lies are created to fulfill one or more distal intentions in the speaker.

• The ability to lie is an embodied skill that manipulates other agents’ dispositions to act.

In short, on our account, the dynamics of lying involves the active manipulation of the short- and long-term dynamics of social cognitive systems so that agents have access to different sets of affordances from the one’s they counterfactually would have had access to without the lie. The liar keeps their field of affordances unmodified while the victim has a different field of affordances post-lie.

Deception, especially lying, requires an awareness of the “truth.” We regard deception as a nexus of embodied and material processes rather than a representational process. Thus, we also de-mentalize the concept of “distal intention.” A distal intention is a helpful heuristic that describes a mesh of interactive and embodied processes that allow agents to be intentionally directed at non-present objects or events. Often when agents are acting on affordances in their immediate environment those same affordances also activate neural and sensorimotor processes that enact imaginary experiences or re-enact past experiences. Such re-activation in turn can keep the agent dispositionally sensitive to non-present objects and events. Finally, skillfully partaking in narrative practices also allows for human agents to be directed at non-present objects and events (Dings, 2019; Hutto, 2008; Rucińska, 2016).

Lying takes place in many settings; in dyadic interactions, in political speeches, in board meetings, etc. We here focus on the interactive in-the-moment lie, as it is typically delivered between people in conversation. For example, “I met Idris Elba last night and he said we should run away together.” Such lies are qualitatively different from rehearsed speeches meant for mass communication; “My colleagues, every statement I make today is backed up by sources, solid sources… These are not assertions. What we’re giving you are facts and conclusions based on solid intelligence” (Collin Powell, 02/05/2003, to the UN Security Council). The first example happens on the fly within the dynamics of a conversation and should be understood across a short-term time scale. The second example happened in front of a global audience and is more akin to a rehearsed performance. Due to space constraints, we here focus on the dynamics of lying one-on-one. In later research, we will use this framework to scale up to cases of lying via mass communication.

The conceptual landscape: Deception, bullshit, and lying

Before we dive in, we must do a little conceptual cleanup—namely, by distinguishing deception, bullshit, and lying. In this paper, we look only at the dynamics of lying—one specific kind of deception. In addition, since lying by definition involves falsehood, we must also get clear on our definition of “falsehood.”

Deception

To deceive is simply to get someone to believe something that is false. Thus, deception can be performed in a variety of ways, many without the use of speech. For example, by dressing up as Dolly Parton I might convince someone that they have seen Dolly Parton walking down the street. In that case, I haven’t lied, but I have deceived. Lying necessarily needs a speech act, whereas deception doesn’t necessarily involve a speech act.

Related, is the act of misleading. Misleading involves true statements that are nonetheless acts of deception. In contrast, lying involves uttering a false statement. One can mislead via actions or words. Here is an example from Bernard Williams: A neighbor says “Someone has been opening your mail” which is true but the “someone” is actually the neighbor. So, the neighbor has misled you with a true statement (Williams, 2002). Typically, in cases of misleading, the statement made will be true by the speakers light and so would not count as a lie. Lies require that the speaker say something they know to be false. What is said to mislead has a conversational implicature intended to lead the audience to form a false belief (For an overview on implicature see Borg, 2009). Similarly, omitting information is also a way of misleading without stating something that is false. For example, you might actually have met Idris Elba last night and he did in fact tell you to “run away together” but it was because he was rehearsing a line for a new movie, and you were simply delivering coffee on set. Without being an outright lie, the omission of information in the speech act still leads the victim to believe in something that is not the case.

Bullshit

Bullshit is conceptually close to misleading but is neither a lie nor misleading. Rather, bullshit is the attempt to implant a belief regarding the speaker’s virtuosity in the listener. It does not matter to the bullshitter if they are speaking the truth, misleading, or deceiving. What matters is that the listener come to believe that the speaker is competent, knowledgeable, and generally virtuous in the subject domain (Frankfurt, 2009). For example, on April 23^rd, 2020, president Donald Trump told constituents to inject bleach into their arms to kill coronavirus (McGraw & Stein, 2021). At the time the coronavirus was still being studied and a vaccine was not available. At the moment of utterance, it may or may not have been true that injecting bleach kills SARS-CoV-2. However, the statement was made in an attempt to signal that the president and his staff had the raging coronavirus pandemic under control. Bullshit is an act that is concerned with optics; making the receiver believe that the sender of the message is competent.

Furthermore, a common bullshitting strategy is making the receiver think that a message contains profoundness or technical depth (Evans et al., 2020; Kolber, 2018; Pennycook et al., 2015). The speaker strings together words so that the meaning of the sentence is vague, opaque, or vacuous but has a veneer of hidden profoundness. For example, Pennycook et al. used the sentence “Wholeness quiets infinite phenomena” to tests subjects’ ability to detect bullshit (Pennycook et al., 2015). The bullshitter aims to have the listener look for profoundness in their sentences and assume that the flaw is in themselves when they do not find that profoundness.

Lying

We have seen how non-verbal deception, misleading, and bullshit differ from lying. Bullshit is about optics, misleading utilizes true statements or omissions, and non-verbal deception in the broadest sense is about implanting false beliefs (using whatever means including gestures and the like). At a minimum, what is distinctive of lying is saying something you believe to be false. However, on many (but not all) accounts lying also involves the intention to deceive (Fox Krauss, 2017; Michaelson & Stokke, 2018). Although there is no consensus among philosophers on how to define lying, on what can be considered a common or traditional account, A lies to B if and only if:

(i) A says p to B.

(ii) A believes p is false.

(iii) A intends to deceive B by saying p.

What does A intend to deceive B about? It is natural to think that A intends to deceive B about p. For example, when Kerri tells Christian that there is no chocolate in the pantry (knowing full well that there is one piece left) so that he will not take the last piece of chocolate, Kerri intends to deceive him about chocolate supplies. But lying might also have the effect of deceiving someone about one’s own state of mind. In telling Christian that there is no chocolate in the pantry, Kerri may deceive Christian in to thinking that she believes there is no chocolate in the pantry.

Although the traditional account captures a lot of cases, we might identify as lying, it doesn’t seem to work for other cases. There seem to be cases where people say something they know to be false but do so without the intention to deceive. Philosopher such as Roy Sorensen (Sorensen, 2007), Tom Carson (Carson, 2010), and Don Fallis (Fallis, 2009) have offered counterexamples to the claim that the liar must intend to deceive their audience. Sorensen, for instance, offers the “bald-faced” lie as an example of lying where there is no intention to deceive. Bald-faced lies are instances in which the lie is not hidden, everyone involved knows that the utterance is false. For example, “We will be young forever!”

For the sake of our discussion here we will not consider bald face lies. To understand the enactive dynamics of lying, we will stick with the traditional account of lying. At a minimum, lying involves speaking contrary to one’s mind with the intention to deceive. The speaker says p but does not believe that p with the intention of deceiving.

Here, a quick note on falsehoods is warranted since lying requires false statements. It is outside the scope of this paper to wade into any debates regarding the nature of “truth.” Thus, we take a pragmatic approach to false statements. For a person to lie, they simply need to believe that p is false. Whether the statement is “metaphysically false,” that is, whether the statement bears correspondence to the state of affairs in the world, is, for our purposes, irrelevant. A person is lying when they say something they believe to be false. Accidentally telling the truth while meaning to lie does not change the embodied cognitive dynamics of lying. We are investigating the dynamics of lying; thus, pragmatically, to utter a false statement is to believe the statement does not correspond with the world. In sum, to lie, is to intend to deceive by uttering a statement one does not believe to correspond with the world. As an approach, an enactivist account of the dynamics of lying is not interested in what lies “are” ontologically but rather what lies are “for”—namely, to produce targeted behavior change in the victim.

As a caveat, this paper only tackles cases in which the victim wholly believes the lie. That is, we are investigating the dynamics of perfectly successful cases of lying. In future research, we will address the issue of lies and deception in which the victim is only partially convinced and cases in which the listener is suspicious of the speaker. Just as theories of truthful communication often begin with the best-case scenario, we here begin with the worst-case scenario—when the listener is fully convinced by a lie.

Cognitivism and lying

On the traditional account, lying is an intentional action in which one person misrepresents a fact by uttering a false statement meant to change the representations of facts within the other person’s mind. The body, the physical environment, human culture, and other factors are mostly side characters in such a story. Deception in its various forms have traditionally been considered games of mental gymnastics. The lack of holistic thinking in the literature has made some psychologists argue that we need a more ecological account of deception—especially in relation to ontogeny (Jakubowska & Białecka-Pikul, 2020; Reddy, 2007).

Cognitive demand theory

One popular theory of lying that relies heavily on internal mental representations is “cognitive demand theory” (De Villiers & Pyers, 2002; Petty & Cacioppo, 1984; Reyna & Brainerd, 1995). On this view, lying takes up increased cognitive bandwidth and therefore results in atypical communication patterns in the liar (Vrij, 2015; Vrij et al., 2022). For example, the cognitive demand theory argues that liars have been observed to blink less frequently than truth tellers; the more complex the lie the slower the blink rate (Immanuel et al., 2018; Leal & Vrij, 2008; Siegle et al., 2008). One explanation is that liars have a lower blink rate since they actively and attentively try to control their body language, gestures and so forth, adding to cognitive demand, while attempting to execute the lie (Burgoon et al., 1990; 2003; Burgoon & Buller, 2008).

Furthermore, it has been shown that asking certain kinds of questions can put increased pressure on the speaker bringing out detectable differences when the speaker is lying (Vrij et al., 2011; 2022). The aim of this technique is to induce additional cognitive load in the respondent when they choose to tell a lie (Vrij et al., 2017). However, decades of the development of lie detection techniques, based primarily on the assumptions of cognitive demand theory (Masip, 2017), have proven to be mostly inconclusive.

The essence of cognitive demand theory is computational and representational. The detectable differences between liars and truthtellers come from the fact that “the brain” has to process both the made-up statement and compare that statement to its model of reality. The mind has to keep the actual state of affairs and the counter factual simulated model of the world side by side simultaneously. When comparing the two models the mind has to calculate whether the counterfactual model of the world will likely achieve the agent’s deceptive aims. For example, will this statement make my friend believe that Idris Elba wants to run away with me? As its name indicates, cognitive demand theory posits that liars must do more representing and computing than truth tellers and that this is difficult. For example, cognitive demand theory further argues that liars tend to have slower speaking on set times and speak slower because they need to internally process the lie first (Walczyk et al., 2005).

It has also been argued, in contrast, that liars have “an acceleration of overall speaking tempo” due to the cognitive demands of delivering a lie (Kirchhübel et al., 2013). In the same vein, the deliberate use of pauses correlates more with truthful speech. In other words, liars tend to deliver their utterances without pauses (Benus et al., 2006). Here, think of the movie trope in which a lying character suddenly speaks faster and in a higher pitch to indicate to the viewer off screen that the character is lying. Again, the cognitive demand of lying is supposed to make deceptive speech harder which results in a rushed delivery rather than a natural sounding performance.

Additionally, fMRI studies have shown that lying is associated with increased activity in the prefrontal cortex and that “attentional orientation systems involved in visual target and novelty processing as well as working memory systems involved in contextual processing are active during deceptive behavior” (Langleben et al., 2005, p. 12). The idea here, is that lying involves an increased monitoring of oneself and the target of the lie. In other words, more attention is needed when lying to insure the successful implanting of the lie. On such accounts, bodily and behavioral differences (whatever they happen to be—blink rates, speaking tempo, etc.) are simply by-products, or “tells,” that result from the real work that is going on inside the head.

Expectancy violation

Theories involving the detection of lying may tell us something further about the nature of lies. The “expectancy violation” framework (Burgoon, 2009; 2015) is a general theory of intersubjective communication which argues that human beings are constantly cognizing the explicit and implicit social expectations of their context, and monitoring whether these expectations are being violated by themselves or others (Burgoon, 2015). On such a view, liars, despite their best efforts, often produce differences in gesture, posture, facial expression, and other embodied cues that break with the expectations of their context. While these might not be detectable with the naked eye, they are detectable with computerized technology such as facial expression recognition software (e.g., Imotions). The higher the difference between the baseline of an embodied cue and its observed divergence the more likely the utterance is to be false. For example, if a speaker always uses the same hand gesture when underscoring their points, but on a given occasion uses a wildly different hand gesture, expectations have been broken and there is reason to suspect foul play. The expectancy violation approach combines computerized discourse techniques and audiovisual analyses to uncover differences in non-verbal cues (Burgoon et al., 1990; Kim et al., 2012; Knapp et al., 2013; Mandal, 2014; Mehu & van der Maaten, 2014; Morgan et al., 2017; Sporer & Schwandt, 2007).

Here the general story is the same as with cognitive demand theory. Lies are produced internally and it is the difficulty of their production that creates externally measurable differences. Lies break with expectations if they are not processed well enough. Thus, the talented liar is good at internal processing. Within this approach, one way to measure the difference between lying and truth telling is to compare a subject’s baseline of word choices and sentence construction patterns against the sentence in question (Fuller et al., 2013).

Avoidance approach

Finally, in the “avoidance” approach to lying it is hypothesized that liars tend to distance themselves from the lie (Fuller et al., 2013). Within linguistics various analyses of word use have demonstrated that liars tend to use fewer pronouns when uttering lies. Especially first-person pronouns referring to the liar themselves tend to be dropped in deceptive utterances (Pennebaker, 2011; Pennebaker et al., 2015). Pronoun drop is a frequent strategy when liars try to linguistically distance themselves from the lie (Hancock et al., 2007). In fact, the way that liars use grammar tends to significantly differ from truthful sentences (McNamara et al., 2014).

In a similar vein, liars often deploy syntactic simplicity and an increase in word concreteness (Windsor & Bowman, 2018). Put differently, liars often try to simplify their speech to get the point across. This increase in simplicity has been hypothesized as an implicit strategy since complex speech is harder for the listener to understand and remember (Reyna & Brainerd, 1995). The liar augments their linguistic strategy to increase their chance of successfully implanting the lie. It has also been suggested that syntactical simplicity may be a way to balance the cognitive complexity or demand involved in lying (Bedwell et al., 2011).

Enactivism

We have surveyed some of the cognitive science literature on lying. Our goal here is to develop a new theory of the dynamics of lying that can accommodate the multimodal empirical data without relying on cognitivist notions such as mental representations or information processing. Neither humans nor their brains are computers and treating them as such has led to many theoretical and practical dead ends (For an overview see Hutto & Myin, 2013, Hutto & Myin, 2017). Following the enactive alternative, we develop an account of lying dynamics framed in terms of an integration of bodily, action-oriented, cultural, social interactive, ecological, and habitual factors.

First, we will outline the general enactivist framework of cognition and then proceed to summarize its account of truthful communication. From there, we demonstrate how this framework reinterprets the literature surveyed above to create an enactivist account of the dynamics of lying.

Enactivism is a theoretical and scientific perspective within the more general 4E cognition framework (For a comprehensive introduction to 4E cognition see Newen et al., 2018). Autopoietic enactivism thinks of life itself and cognition as being co-extensive (Thompson, 2007). The very basic processes of life are cognitive processes—where there is life there is cognition. On this model, cognition means the constant gross and minute physical sensorimotor actions taken by an organism to stay in optimal equilibrium with its environment (in humans to be in “equilibrium” is highly culturally determined). “Cognition” broadly conceived is synonymous with adaptive embodied sensorimotor action. Organic bodies are always in the process of acting on their environment to optimize their attunement to those environments. At the most abstract level this means controlling how the environment asserts force onto oneself, by asserting force onto the environment (Di Paolo et al., 2017).

On this model perception happens as the cognitive system acts on the world. In this process, enactivists contend that organisms enact their perceptual world (Varela et al., 1991). Only as a mental exercise of reflective analysis is it meaningful to separate action from perception. In everyday life, organisms are always busy acting on their environments which in turn brings forth those worlds (Gallagher, 2017; Varela et al., 1991). This is true of single-celled organisms as well as fully enculturated human beings.

On this view, cognition is performed by an entire agent, not just its brain. Note further, that in such a description of cognition there is so far no need to appeal to representations or a mysterious inner world of the mental. The mental simply is the whole body at work in direct commerce with the environment (Fuchs, 2018; Gallagher, 2017).

For the human, “environment” refers to the full Umwelt of the organism including its cultural world (Baggs & Chemero, 2019). While all organisms have a unique Umwelt, the Umwelt of human beings is distinctly cultural. Human beings, operate in dense cultural worlds (rich sets of social/cultural/normative practices) that they must learn to master in order to thrive. Maintaining equilibrium for human beings means becoming increasingly proficient at operating in rich cultural worlds (Bruineberg et al., 2021; Rietveld & Kiverstein, 2014).

The important takeaway for our purposes is that “the mental” is not relegated to some secret internal realm behind the physical body or without the world. Rather mental activity is the holistic coupling of processes across brains, bodies, and environments. This becomes important as we develop an account of the dynamics of lying that does not rely on internal mental representation.

Enactivism and honest communication

So far explicitly enactivist research has focused mostly on developing a rich model of successful, cooperative, and truthful communication (De Jaegher & Di Paolo, 2007; Di Paolo et al., 2018; Fingerhut, 2021; Gallagher, 2020; Hutto, 2008). We here provide the story that has so far been developed regarding enactivist conceptions of communication. We then proceed to demonstrate how we can develop an enactivist model of dishonest deceptive communication. As we shall see truthful communication is an embodied skill and so is lying.

Enactivism conceptualizes (truthful) communication as two or more agents becoming coupled through embodied synchronization processes in which each agent directly and reciprocally regulates the other and the coupling itself through the use of social affordances (De Jaegher & Di Paolo, 2007; Di Paolo et al., 2018).

Affordances are possibilities for action that exist between an agent and its encultured niche—that is affordances are action-possibility relations (Chemero, 2009; Gibson, 1977; 1979; Oliveira et al., 2021). Given the specifications of human bodies, the details of human cultural niches, and the skills and capacities of the individual, the agent stands in specific relationships of action possibility to the environment. For example, for able-bodied persons typically, staircases afford “climbing,” buttons afford “pressing,” and a card readers afford “purchasing.” In contrast, a card reader does not afford “purchasing” for a spider since spiders do not live in the right culture and do not have the right kind of body. However, given the spider’s body the card reader might afford “hide-ability.” Agents are always embedded within a dynamically ever-changing field of affordances specific to their bodies, skills, culture, and even current emotional and affective states (Dings, 2020; Gallagher, 2018; Kimmel & Rogler, 2018; Kronsted, 2021).

In human communication, agents provide and act on “social affordances” (Chemero, 2009; De Jaegher et al., 2010; Dings, 2020; Gibson, 1979; Raja & Anderson, 2019; Rietveld & Kiverstein, 2014). These possibilities for actions, conveyed through words, gestures, posture, facial expressions and more, directly regulate the agents in the interaction, their coupling, and relationship (Di Paolo et al., 2018). Typical examples include the outstretched arm affording the “high-five,” furrowed eye-brows affording “say more,” or two outward-facing hands affording a pause.

Naturally, words themselves present a wealth of highly intricate social affordances. For example, for a romantic couple; “Can we eat something greasy tonight?” affords “Order take-out,” “Talk to me about my bad day at work,” and “Hug me.” As this example demonstrates, social affordances are also highly specific to the history of interaction between agents. “Can we eat something greasy tonight?” does not afford “talk to me about my day” for all couples—minute social affordances often develop from years of interaction. Furthermore, affordances are highly regulated by our social status within various cultures and groups. Race, gender, class, and so forth augments how agent’s perceive each other including the affordances stemming from their bodies and their actions (Brancazio, 2020).

For human agents, any given environment contains a rich multitude of affordances. However, not all affordances have equal significance to the agent. A glass of water is highly soliciting to an agent who is dehydrated. The drinkability of the water will then standout to the agent while other affordances will recede into the background. Salty potato chips do not afford eat-ability to a person who is dehydrated. However, the chips might “scream come eat me” to the person who has the munchies—affordances are typically experienced as more or less soliciting given a number of biological, social, and personal factors.

Through words, gestures, body language and more, agents not only produce social affordances for each other that perpetuate the communicative interaction, they also change the available field of affordances for themselves and the other agent subsequent to the interaction. For example, asking one’s partner to order something greasy will increase the solicitation from affordances related to food ordering. Suddenly the pizza menus and burger menus in the junk drawer become highly salient and solicit the agent into action. Similarly asking one’s partner to get something specific from the grocery store means that the affordances from certain items in the store will be highly soliciting while others will recede into the background. Successful communication often is a matter of ensuring that other agents have a field of affordances that matches one’s intentions.

The more that agents interact the smoother their interactions become as a history of interaction builds (for example, the case of romantic couples). However, such interaction is not only built over longer time scales. As human beings interact their bodies begin to synchronize across various physiological measures. A synchronized coupling happens as two or more systems, in this case human beings, fall into a steady physical interaction pattern with one another—in other words they become phase locked (typically, moving together “in-phase,” moving in opposition “anti-phase,” or “moving in a non-symmetrical pattern off-phase” [Pikovskij et al., 2001]). For human beings, the minute relationships of couplings take place as they spend increased time together. As human beings interact a slew of low-level embodied alignment processes couple and synchronize. For example, posture, gestures, tone of voice, proxemics, eye gaze patterns, and many other features begin to fall into increasingly stronger patterns of synchronization as human beings interact (Garrod & Pickering, 2009; Lakens, 2010; Paxton & Dale, 2013; Zimmermann & Richardson, 2016). Even neuronal firing patterns have been shown to couple and synchronize in interaction (Cheng et al., 2019; Dikker et al., 2017). So, as human beings interact their embodied processes fall into stronger alignment through multimodal synchronization.

Over time, as agents develop a rapport with one another their shared field of social affordances become increasingly nuanced making it easier to perpetuate the interaction. Richer synchronization means more affordances, and more affordances mean more nuanced synchronization. An easy way to think of this is to consider behavior at a cocktail party or a reception. At first small talking to people is typically awkward and driven by public conventions. However, as people get more familiarized with one another interactions become more fluent and typically less formal. On the enactive model of communication, there is an important embodied communicative feedback loop between sensorimotor processes and acting on social affordances. As agents act on each other’s social affordances their sensorimotor alignment processes strengthen their coupling and reach a higher degree of synchronization. As tighter synchronization is achieved the field of affordances becomes more nuanced allowing agents to further perpetuate the interaction by acting on each other’s social affordances. Affordances promote action, action regulates the system, the system promotes more or richer affordances.

What communicators are doing then is taking turns regulating and being regulated (De Jaegher & Di Paolo, 2007). Speaker One impacts the dynamics of the overall system by regulating Speaker Two through words, gestures, facial expressions, posture, etc. In turn, Speaker Two regulates Speaker One through words, gestures, facial expression and so on. Communication is a matter of action—each agent is acting on one another for the perpetuation of the communicative joint dynamical system.

Consider one important point about affordances that will become especially salient when we discuss lies: when someone acts on an affordance they are often also simultaneously intentionally directed at some non-present object or future state of affairs. Affordances are not only dispositions to act in the immediate environment, but they also lead us to be dispositionally inclined towards future actions. Such regulation includes setting, augmenting, and adjusting our plans and intentions. For example, when entering a parking garage a button with the sign “press for tickets” affords press-ability. In acting on the affordance, that is pressing the button, a future affordance become salient. Having pressed the button, parking ticket validation machines now become highly soliciting (you don’t want to pay full price for parking when you are just running to pick up two things in a store). In this simple case, acting on an affordance (pushing the button for a ticket) makes the agent disposed to act in a certain way in a future environment (validating the ticket inside the store with its parking validation machine). Significantly, affordances not only adjust our dispositions for action in the short term, but can also do so over longer time scales, as is often the case when someone successfully lies. Thus, in the next section we will look at affordances, distal intentions, and lies.

The role of intentions and narrative scripts in lying

In section The conceptual landscape: Deception, bullshit, and lying we outlined how on the traditional account lying requires the intention to deceive. Thus, we need to understand how “intending” functions in a non-representational framework. Contemporary debates in philosophy tend to use the distal-proximal- motor-intention framework—the DPM model (Mylopoulos & Pacherie, 2018; Pacherie, 2000; 2006). Motor intentions are rapid intentions that guide action within movement. Proximal intentions exist at the time scale of quotidian action, for example, deciding to get a beer from the cooler or intending to open the door when someone knocks. Distal intentions, exist across longer time scales, for example, deciding to get a PhD or planning to go see Romeo Santos in concert on Friday. Importantly, distal intentions are action guiding—they fix and organize lower-level intentions and help structure our actions (Bratman, 1987; 1999; 2007). To stay within the scope of this paper, we here focus mostly on distal intentions.

Setting various intentions structure an agent’s available affordances. If I “set” the distal intention to “turn my life around—workout every morning and eat healthy” then in the grocery store tofu and tempeh afford purchasing, whereas before I simply walked by barely registering these items. Similarly, the intention to “find my keys” will make the affordances related to key finding stand out from the background. Pillows now afford moving over, junk in the drawer affords pushing to the side, and the couch affords looking under.

While intentions structure our available affordances, affordances can also help structure our intentions. As we act on affordances, new situations present themselves which makes the agent set new intentions, or reorganize their existing intentions. For example, the parking garage ticket creates the intention to validate the ticket once I’m in the store, to avoid paying. In this way, there is a recursive loop between affordances and intentions. However, our truly long-term distal intentions tend to have more inertia, that is they are not easily changed. If someone truly intends “to be the very best that no one ever was” then that distal intention will not easily be changed and will generally have a high influence over the agent’s intention and affordance structures even as they encounter obstacles along the way.

When someone lies intentions are affected in a tripartite fashion. First, the liar typically has some distal intention that they believe cannot be achieved without a lie (although some people do lie for sport). Second, the liar therefore intends to deceive. Third, through the speech act of lying the liar may aim to change the victim’s behavior, to cause the victim to augment, revise, or adopt some new intention (distal or proximal). This new or augmented intention of the victim may mismatch with reality or may make them act in a fashion that aligns with the liar’s intentions.

More concretely, having one’s intentions augmented by the lie means having one’s landscape of affordances changed. Let’s return to the chocolate example. Kerri lies to Christian and claims that there are no more chocolates left in the house. She is preemptively changing how intentions and affordances solicit Christian’s actions. Normally, Christian gets a sweet tooth after dinner and locations around the house that typically contain chocolate (pantry, drawers, fridge) become highly soliciting. By telling Christian that there is no chocolate left in the house, pantries, drawers, and the fridge no longer afford “chocolate-getting.” Christian will therefore not look in the pantry. The effective lie has changed Christian’s behavior by changing his field of affordances. Furthermore, Christian’s proximal intention to get chocolate changes since he now has the “belief” that the house is a chocolate desert.

Importantly, in a lie the perpetrator changes the victim’s field of affordances so that the two affordance fields mismatch. Lying involves changing the affordances of others in a manner that does not happen for the liar—to Kerri the pantry still affords chocolate getting. Additionally, there is now less of a chance that Christian will set a chocolate getting intention.

Often, the successful lie becomes meshed into the agent’s dispositional patterns across a longer time frame. In the harmless chocolate example, Christian generally may stop checking the pantry for chocolates. In a less harmless example, the agent believes that a presidential election was rigged, and begins carrying out actions that undermines democracy. We see then how lying often can manifest itself across entire patterns of behaviors and across longer time frames. Lies change the victim’s landscape of affordances, and often make victims set distal or proximal intentions that fulfill the aim of the lie. However, the perpetrator’s field of affordances may largely remain unchanged, or is changed to her advantage. Lying creates “asymmetrical” fields of affordances.

It is easy to think of intentions, especially distal intentions, in cognitivist terms as mental items that are processed in the head—however, we are here de-mentalizing the notion of distal intention. We take intention to be a helpful heuristic that refers to the combination of an agent’s skills, dispositions, beliefs, desires, self-narratives, and other relevant psychological features. We take a helpful point from Gilbert Ryle: to have a “belief” means being disposed to act in a certain way in specific contexts (Ryle, 1976). Similarly, to have an intention (distal or other) is to be disposed, with varying degrees of skill, to act in certain ways. Thus, when Joe has a distal intention to complete the video game Elden Ring on “ultra-hard,” it means that he is solicited by affordances related to getting better at the game such as watching tutorials on YouTube or discussing strategies with friends. Affordances within the game becomes nuanced as he gets better, acquires skill, and spends time playing the game rather than spending time on other things. In other words, intentions describe an agent’s behavioral patterns across multiple timescales. To intend to lie, then, is to have a disposition to act in a certain way; and lying is an action that, when successful, changes the behavioral regimes of other agents so that the liar can fulfill their own distal intentions.

One might think that one forms a distal intention by a process of deliberation. Deliberation too is a skillful activity, often shaped by habitual dispositions, some of which may be clearly bodily, for example, pacing the floor, adopting various postures, head nodding, gesturing (on the one hand, on the other hand), and sometimes intersubjective, as when our deliberation just is talking something through with another person (which may also be an imaginary, or internalized other). Deliberation too is embodied and enacted.

The core difference, between lying and other speech acts, is found in the intention to lie and the asymmetrical nature of affordance fields produced by lies. While truthful communication through participatory sensemaking creates a joint field of affordances, lying aims to create two similar, yet different fields of affordances (the two fields are different across the features relevant to the lie). The liar is intending to create dissimilar fields of affordances. The truth teller on the other hand intends a joint, correlative, or in some cases matching fields of affordances. Through the process of creating affordance asymmetry the liar coopts participatory sensemaking to fit their aims.

From a systems perspective the difference between lying and other speech acts is in the way lying attempts to grant the speaker autonomy over the interaction and future interactions. The intention to deceive and the use of false statements constrain and redirect the emergent properties of the communicative system in a direction that favors the liar, thereby increasing their autonomy over the system and future interactions. In cases of truth telling, the sensemaking process is genuinely participatory—both participants enter into and perpetuate the interaction cooperatively. In the case of lying, the perpetrator uses false statements to skew the system in their intended direction. In truthful communication from the interaction itself emerges new systems-properties that are non-reducible to the individual participants. Importantly, truthful communicators understand that cooperative communication leads to emergence that is outside of the communicators’ individual control. In the case of lying, the liar is attempting to control the emerging properties of the system in their favor, thereby making the interaction non-cooperative. Truthful communication aims for mutual autonomy. Lying aims at individual autonomy, and the diminishment of the other’s autonomy.

Narrative scripts, the hermeneutical background, and lies

Communication is often (but not always) smooth—that is we often communicate with ease and know just how to behave in different communicative contexts. Rather than relying on the cognitivist notion of interaction smoothers (Vesper et al., 2010; 2017) we rely instead on narratives and cultural backgrounds (Gallagher, 2020; Hipólito et al., 2021; Hutto, 2008).

As human beings develop they learn to interact with ease by relying on culturally and socially inherited narrative scripts—that is we learn to understand the intentionality and intentions of others by slowly mastering social-cultural narrative scripts (Hutto, 2008). Through carefully scaffolded narrative practices such as bedtime stories or play-acting, children learn to follow cultural narrative scripts that eventually transition to explicit belief-desire reasoning. Children often learn to mix and match narrative scripts through interactive playtime with adults, other children, or alone. In such cases, they enact the script as a form of social competency practice (Rucińska, 2016). For example, through games children quickly learn about turn taking and reciprocity and the normativity that come with these virtues. When playing “Go Fish” the learned narrative script unfolds so that when someone asks whether the player has a certain card the player answers truthfully; “do you have any fives?”—“Go fish”—a script that is defined by the game, or more generally by the cultural context.

Human beings are always moving through various contexts and background structures in which practiced narrative scripts become relevant and “live.” Part of cultural competency is for the whole agent to always be in tune with the narrative practices that are relevant to their current situation (Gallagher, 2020). Often when a lie is produced it is “inserted” into the flow of the unfolding of a narrative script. Effective liars often use the current interactive narrative script that is in play to their advantage. When the child learns to lie they are manipulating the flow of the game by saying “Go fish” when really they did have a five! As a result, the affordances available to their co-player change and they now have to pick up a card. The liar is exploiting the general template of the interaction, to change the behavioral regime of their victim by changing intentions and affordances within the frame of narrative practices.

We have conceptualized lying dynamics as involving a relational action, practice or performance that exploits narrative scripts, intentions, and affordances—the lie changes the victim’s behavioral dispositions by changing their available affordances, and makes them create new intentions. The victim’s augmented behavioral dispositions now align more closely with the intentions of the liar. For example, if I tell my little cousin that ducks at the local park have been genetically modified by the government to eat young redheaded boys, then his field of affordances at the park, especially in relation to the ducks will be drastically different from the case where I did not tell the lie. His new duck-avoiding behaviors now fit with my intentions to go to the playground rather than the pond. Through the manipulation of long-term and short-term interaction dynamics, the liar ensures that the deceived is acting on a “counterfactual” set of affordances. The liar is reconfiguring how the victim is dynamically relating to the world—lying is about changing the action and interaction patterns of others to fit one’s own long-term interests. A person who has successfully been convinced by a lie now stands in a different dynamical relationship to the world than if they had not been interacting with the liar—ducks now afford running away rather than feeding. Lying is a skill that changes how others interact with the world skewing their sensemaking processes.

While jokingly lying about ducks might change someone’s interaction patterns in the short-term, powerful lies can radically change behavior over time because they become sedimented into the agent’s general patterns of worldly interaction. For example, much of the debate over cognitive penetration (whether or not perception itself is augmented by beliefs) centers around cognitive bias that comes from being told lies about population groups (Briscoe, 2015; Siegel, 2012; Stokes, 2013). In the United States, “The Big Lie” about election fraud has radically changed interaction patterns between population groups, their interactions with civic entities and government institutions, and increased the presence of white supremacist politics and groups (DiMaggio, 2022; Hawkman & Diem, 2022; Painter & Fernandes, 2022). “The Big Lie” has radically changed the behavioral regimes of its victims, by changing what people and institutions afford, what intentions people are setting (for example, to start an insurrection), and even the narrative scripts for acceptable ways of interacting during public discourse.

Conclusion: The multimodal future of deception research

Currently research on lie detection has not made much progress. Standard lie detection techniques are no more accurate than flipping a coin (Burgoon, 2015; 2018). While as we have seen there are several theories on the market, no current approach to deception detection takes into account the whole embodied agent. According to our embodied-enactive and interactive account of lying dynamics, future research on deception detection needs to take a holistic and multimodal approach. Without, taking the full embodied agent into account future approaches to deception detection will likely remain inadequate. While in the many academic debates about embodied versus cognitivist approaches, there seems to be little real-world consequence, in the case of deception and deception detection, one finds many real-world applications that can have dire consequences if approached without a better understanding. For example, pseudoscience lie detection and interrogation techniques have been sold to police departments around the United States for decades without any evidence of efficacy, in many cases leading to false prosecutions (Denault et al., 2020). A contextual and more embodied approach to lies and deception shows more promise for improving effective detection techniques.

For the enactive approach to lying, multiple research questions need further elaboration. On a more fine-grained analysis of lies, how do different types of lies manipulate affordances? For example, one prominent type of lying that needs further analysis is gaslighting. Furthermore, what happens to the embodied interaction dynamics when one interactor suspects foul play? For example, how are interaction patterns different when speaking with somebody who is known to be a habitual liar? More generally, research needs to further investigate uncooperative and combative forms of sensemaking.