Ethos, Hexis, and the Case for Persuasive Technologies

Steve Holmes, George Mason University

(Published November 22, 2016)

In this essay, I want to call attention to a new—if unorthodox—definition of digital rhetoric: the idea that behavior change is rhetorical (323). In Persuasive Technology, the computer scientist B. J. Fogg argues for an extension of rhetoric's historic roots in effective argumentation to include forms of nonconscious behavioral reinforcement monitored by computational algorithms and real-time feedback. Fogg founded Stanford University's Persuasive Technology Lab (PTL) in 1998, and he inspired the International Conference on Persuasive Technology (ICPT) in 2000. Highlighting his successes in these areas, the 2015 ICPT website states, "[Fogg's] students created Facebook Apps that motivated over 16 million user installations in 10 weeks" (Persuasive Technology Conference). The creators of other popular social media interfaces (e.g., Instagram, Friend.ly) also have participated in Fogg's designer boot camps (Wynia).

While his work has received little attention from digital rhetoricians, the idea of a persuasive technology usefully describes many of the ways in which digital rhetoric has become aligned with behavior change in the past few years. Many technology journalists heralded New Year's Day 2015 as the beginning of the "Track-Everything Revolution" (see fig. 1) to reflect the release of countless apps designed to aid Americans at the traditional time when many of us resolve to form new diet or exercise habits (Peterson).

Fig. 1. Image from Helen Peterson's aptly titled Buzzfeed article "Big Mother is Watching You"

A sample of newly monitored behaviors includes tracking an elderly relative's health (Lively), measuring blood alcohol content or electrodermal activity (Apple Health), evaluating beverage selection (Vessyl), and "master[ing] your fertility" for females (Glow). Alongside the ubiquitous FitBit watches are a growing number of interfaces that have habit as a specific object, such as HabitRPG, Chore Wars, Task Rabbit, Go Fucking Do It, or SuperBetter.

Degrees of what I would simply call "Foggian rhetoric" can be found in many networked interfaces. It is no small coincidence Amazon.com's recommendations feature is one of Fogg's key examples of a persuasive technology (2). Amazon.com's algorithms monitor users' aggregate browsing habits to tailor additional purchase suggestions. Similar efforts exist in various nation states' attempts to influence citizen behavior. With the rise of what Tim O'Reilly calls "algorithmic regulation," algorithmic tracking promises an objective mapping of the causes of various citizen behaviors. Algorithmic regulation is often coupled with real-time computational adjustments, such as smart parking meters in San Francisco that adjust their rates depending upon traffic flow and space availability. Echoing French theorist Michel Foucault's late lectures on "governmentality," rhetorical subjects are individually or collectively produced by their conversion into a measurable and governable population through various algorithmic forms of behavioral monitoring (Chaput; Davies; Holmes, "Rhetorical"; Pruchnic).

These developments confirm that a persuasive technology should not be viewed only as a neutral design techne, but as an essential element of the biopolitical construction of rhetorical subjects' procedural habits in the present moment.1 At the same time, I want to suggest that it is important for digital rhetoricians not to respond to this emerging cultural milieu by entirely rejecting or avoiding Fogg's thinking. An equally compelling aspect for me is the way in which Foggian rhetoric casts into relief certain unacknowledged elements of our conceptions of digital rhetoric. Zappen's essay lists Fogg as an unproblematic figure alongside other digital rhetoricians such as Barbara Warnick and Laura J. Gurak (320). Yet Zappen's inclusion of Fogg does not address a necessary tension between rhetoric and behavior change. For example, Kenneth Burke would likely view Fogg as importing the terms of "nonsymbolic motion" to characterize the world of "symbolic action" ("(Nonsymbolic) Motion"). In fact, Fogg's subsequent critics among game designers and computer scientists would agree entirely with a Burkean reading. They universally accuse Fogg of advocating nonrhetorical forms of coercion (Knowles et al.) or manipulation (Bogost; Larson).

While these criticisms are correct in many regards, I want to entertain the admittedly counterintuitive suggestion that existing criticism of Fogg's work has missed what may be his most significant albeit unintentional contribution to digital rhetoric. Unlike his critics, Fogg unabashedly refuses to believe that an audience's embodied habits and behaviors are a priori coercive or nonrhetorical. He therefore offers digital rhetoricians an opportunity to clarify the extent to which we theorize the specifically rhetorical role of physical forces of embodiment such as habit.2 Intriguingly, many definitions or applications of digital rhetoric remain in the realm of argument, classical rhetoric, consciousness, social context, or epistemology and do not directly address nonconscious forms of behavioral affect.3 It is worth noting that most variations of epistemic rhetoric are grounded in Cartesian-Kantian dualism and therefore require digital rhetoricians to agree with Fogg's critics by default. As a consequence, the digital-symbol-using and -abusing animal's actions remain far more rhetorically salient than the ways in which the use of digital tools shapes our bodies, minds, and habits through noncognitive and nonsocial forces.

While our field is seeing a growing number of texts that engage digital rhetoric and technology through rhetoric's embodied and material character, this is far from a dominant digital rhetoric paradigm.4 Furthermore, many of the insights of past and contemporary researchers remind us that theorizing digital rhetoric's embodied and material character—rhetoric's "ontological weight" is not as simple as grafting these elements onto existing frameworks for digital rhetoric (Bay and Rickert 213). Rather, integrating an approach like Fogg's alignment of habit and rhetoric requires recasting entirely the agentive and ontological grounds for how rhetorical beings interact with one another through technologies to make meaning in the world.

In what follows, I take a loose deconstructive approach to defining digital rhetoric through a particular form of embodiment (i.e., habit) and materiality. In "Structure, Sign, and Play," the poststructuralist theorist Jacques Derrida locates an unstable element of play that a given meaningful structure (e.g., digital rhetoric, epistemology) explicitly or tacitly excludes (e.g., persuasive technologies, embodiment) in order to stabilize a ground for meaning. By reincluding the excluded element of persuasive technology, Fogg enables us to gain new insights into digital rhetoric's unavoidable and fundamental grounding in habit. Yet, I only wish to retain Fogg's insistence on aligning behavior change and rhetoric while complicating his understanding of the relationship between the two terms. I suggest that Fogg offers an unwieldy juxtaposition of a philosophical conception of rhetoric and a "mechanistic" understanding of habit expressed in Immanuel Kant's thinking. Despite its actual influence on individuals' beliefs and actions, Kant argues that habit remains an inauthentic form for constituting (rhetorical) subjects' transcendental reasoning capacities.

Fortunately, Kantian mechanism is only one lens for discussing habit in relation to rhetoric. The relationship between rhetoric and habit can also be productively complicated by drawing on a post-Aristotelian lineage of habit found in thinkers as varied as Felix Ravaisson, Henri Bergson, William James, Maurice Merleau-Ponty, Pierre Bourdieu, Gilles Deleuze and Felix Guattari, and, more recently, Catherine Malabou, Thomas Sparrow, Elizabeth Grosz, and Clare Carlisle. For these thinkers, habits are not mechanistic but "ontological." Aristotle's ethos exceeds a mode of artistic proof in a digital interface and also describes an ongoing, fluid, and dynamic process of habituation of the rhetorical self (i.e., second nature) to form a hexis (i.e., habit, state, disposition, bodily comportment). Hexeis (plural) guide phronesis and rhetorical practice through various ethical forms of social and environmental affectivity. Ethos and hexis confirm that Fogg is not incorrect in arguing for habit's important role in digital rhetoric. At the same time, Aristotle demonstrates that digital rhetoricians must come to see habit as a far more fundamental part of digital rhetoric's condition of possibility than Fogg imagines.

While habit offers great potential for theorizing elements of digital rhetoric, a double movement ultimately characterizes deconstruction in that the reincluded element (i.e., habit and persuasive technology) cannot be allowed to offer a new metaphysical or foundational ground for embodied rhetorical meaning. As a result, theorizing digital rhetoric through habit comes at a considerable cost. We gain a theory of embodied rhetorical action in the world (presence), but at the expense of the fundamental autonomy of the Cartesian cogito (absence). To echo Gabriel Markus and Slavoj Žižek's comment on Hegel's discussion of habit, habit remains our pharmakon: the very poison and cure that Derrida has previously employed to describe writing. Even in Aristotle's original conception, hexis equally enables virtue and vice. Thus I conclude that Foggian rhetoric forms an imperfect but necessary part of defining digital rhetoric in the present—one that we must learn to work through even as we seek to diminish and negotiate some of its negative aspects in algorithmic regulation or biopower. I close by examining an older artistic example of a non-utilitarian persuasive technology: biomedia artist Eduardo Kac's Teleporting an Unknown State. Kac's installation effectively demonstrates how digital rhetoricians might develop new persuasive technologies to cultivate Aristotle's hexis of care for the environmental relations that support our rhetorical habits of existence in the world.

Deconstructing Persuasive Technologies

Because Fogg's work enjoys few citations among digital rhetoricians, it is useful to offer a brief overview of some of the possibilities of and problems with his idea of persuasive technologies. Fogg fully defines persuasive technology "as any interactive computing system designed to change people's attitudes or behaviors" (1). He includes any digital system (e.g., Amazon.com, eBay, smart toothbrushes, police roadside speed limit displays) that attempts to produce a behavioral response as a persuasive technology (4, 42). Despite offering an intriguing thesis, Fogg's lengthy book offers only a half-page sidebar on the history of rhetoric (24). Nevertheless, it is quite clear that his understanding of rhetoric is limited to a traditional philosophical (i.e., Platonic) view of rhetoric combined with operant conditioning and more contemporary research in behavioral psychology (49). In the context of Amazon.com's recommendations feature, Fogg draws a line between offering a service (e.g., delivering books) as nonrhetorical and the recommendations feature as rhetorical. In other words, rhetoric's domain is restricted to what Dilip Gaonkar memorably calls a "sweetener" to nonrhetorical forms of logical or empirical reasoning (77). While this bizarre juxtaposition of behavioral stimulation and rhetoric may conjure up Stanley Kubrick's immortalized conditioning scenes from A Clockwork Orange, Fogg also lists fairly basic activities, such as a business sending a customer a thank you note after a purchase, as nondigital examples of positive reinforcement.

Despite his simple definition, Fogg's work can be confusing for rhetoricians because he makes no concerted effort to explain why nonconscious behavioral prompts should be rhetorical in the first place. To offer a point of comparison to the rhetorical tradition, Fogg follows Aristotle's precedent in offering numerous taxonomies of various rhetorical techne for designing persuasive technologies, but skips entirely the latter's lengthy metaphysical explanations of how rhetorical beings and language come to exist in the world. Both his book and the PTL website offer numerous design heuristics to induce behavior change, such as the "Behavior Wizard" or the "Behavior Grid." The Behavior Grid contains design tactics such as the BluePath Note, which instructs a designer to reduce a given habit one time by applying a "de-motivator" (Stanford, "Behavior Grid"). For example, city officials in Buenos Aires experimented with removing saltshakers from restaurant tables to combat a cultural norm of oversalting food (CNN Wire Staff). The de-motivator consisted of having to take the extra step of asking the serving staff to bring the salt to the table. City officials attempted to make diners more conscious of the fact that oversalting food was not a naturalized or normal behavioral practice. Their efforts are undergirded by a tacit Foggian appeal to a behavior psychology principle of "default bias": users are more likely to accept preset offerings than to expend additional cognitive or embodied energy to ask for new choices (or saltshakers) (Shurger and Sher).

The dramatically different assumptions about how Foggian rhetoric operates for digital technologies can be seen directly through a comparison to Bogost's notion of procedural rhetoric. According to Bogost, procedural rhetoric is the unique ability of videogames and other computationally expressive media to make persuasive arguments by modeling interactive real-life systems (3). For example, The Arcade Wire: Windfall asks players to earn a profit by simulating the process of creating clean-energy wind farms (see fig. 2).

Fig. 2. The Arcade Wire: Windfall

Windfall procedurally rewards players if they build windmills in smart locations, such as away from local citizens' homes. Energy can be sold as renewable energy credits to earn profits in order to build more windmills. Windfall even simulates local citizens' protests in support of procedural rhetoric's goal to accurately model real-life systems as a unique form of videogame rhetoric.

By contrast, consider an environmentally themed persuasive technology called Polar Bear, developed as part of an experiment called StepGreen, by researchers in Carnegie Mellon's SCS Green unit (Dillahunt et al.; see fig. 3).5

Fig. 3. Polar Bear from StepGreen

While behavior or attitude change is also the end goal of Windfall, Polar Bear causes users to physically enact the behavior rather than modeling an argument through an interactive pisteis. Polar Bear employs the Tamagotchi, a popular Japanese virtual pet, as a genre prototype. Users are provided with a virtual polar bear and fifteen behavioral actions related to sustainability that can be employed to interact with their polar bear (e.g., "take the stairs instead of the elevator"; "restrict length of shower to five minutes or less"; "take an environmental sustainability flyer and give it to a friend"). Users' positive actions increase the polar bear's ice floe size and improve its overall temperament. Unlike procedural rhetoric, this persuasive technology does not try to have the user learn a procedural argument about global warming. Rather, Polar Bear ties the bear's emotional happiness and survival to the ways in which users engage in forming real-world sustainability behaviors.

Illustrating several key elements of persuasive technology, Polar Bear's behavioral prompts are designed to mesh seamlessly with individuals' daily habits. As a "suggestion technology," Fogg claims that kairos can also describe how designers cue behavioral triggers at opportune moments in a user's preexisting habits (41-43). Polar Bear's designers do not want to encourage users to form dramatically life-altering habits such as "join Greenpeace and picket British Petroleum oil in a foreign country for a month" or "go two weeks without showering to cut down on water waste." By contrast, any user who works in a building with an elevator has access to stairs. To invoke a metaphor from the popular behavioral economists Richard Thaler and Cass R. Sunstein, Polar Bear's designers hope that their persuasive technology will simply "nudge" users toward forming new sustainability practices grounded in their daily habits. Unlike the persuasive game Windfall, Polar Bear refuses to start with the assumption that cognitive or dialogical interactive arguments, such as scare tactics or facts, are an effective form of producing a change in users' sustainability habits. Rather, a user's daily behavioral patterns and spatial situatedness form a superior kairotic terrain.

Responses to Persuasive Technology

While he has an intriguing thesis, Fogg's lack of engagement with the rhetorical tradition leaves him open to a number of predictable criticisms. For example, Karlyn Kohrs Campbell reminds us that behaviorist paradigms of rhetoric—which assume that persuasion occurs "not on an intellectual but rather on [an innate] motor level"—are, in contrast to the "rationalist" tradition following Plato, by no means unfamiliar to the rhetorical tradition" (Brigance, qtd. in Campbell 101). Campbell documents numerous objections that rhetoricians have raised to behaviorism in the past, including Burke's observation that not all forms of symbolic action can be productively attributed to motives stemming from the satisfaction of base biological drives. Along different lines, I see a great deal of Foggian rhetoric as embodying part of what media studies theorist David Golumbia dubs "computationalism": the unquestioned and often top-down political assumption that computers are a priori better at solving individual or collective problems than other democratic forms of deliberation. Furthermore, Fogg's use of streamlined and selective empirical studies to support his claims in Persuasive Technology often fails to acknowledge the existence of counterarguments. As a case in point, the progressive education theorist Alfie Kohn highlights an alternative set of empirical studies to argue that external computational forms are actually less effective at forming long-term or lasting habits than other means of education. In partial support of Kohn's complaint, research firms estimate that FitBit's continuing strong sales for new devices disguise the fact that close to one-third of FitBits become abandoned in a few months (i.e., fail to form a habit), and only half of FitBit's 20 million registered users are still active as of 2015 (Goode).

In recognition of these and other issues, Fogg has provoked a number of hostile reactions from computer scientists and game designers. Bogost's criticism in Persuasive Games offers a representative and frequently cited complaint:

Perhaps [persuasive technologies] offer valid ways of using technology to alter behavior. But not one of them deploys rhetoric. Instead, all of Fogg's techniques use technology to alter actions or beliefs without engaging users in a discourse about the behavior itself or the logics that would recommend such actions or beliefs. (60-61, my emphasis)

Persuasive technologies like Polar Bear only engage what Bogost calls "low process intensity" because they do not simulate complex forms of decision-making through effective procedural argumentation in the mode of a persuasive videogame like Windfall (61-62). Hence, he argues that a persuasive technology is better dubbed a "manipulation technology" (62, emphasis original). Similarly, the game designers of BARTER, another persuasive game that attempts to persuade Lancaster, UK, citizens to spend locally, add that the "question of whether persuasive technologies are manipulative has escaped serious scrutiny by the computer science community" (Knowles et al. 2). Along similar lines, technology columnist Jordan Larson complains,"Persuasive technology can be found … on the digital homes of tech giants like Amazon and Facebook, where behavior-oriented design persuades us to … stay logged in (manipulating social media news feeds)." These examples share Bogost's point of criticism: persuasive technologies can only be manipulative or coercive and nonrhetorical because they work on the body (i.e., form habits) and not the rational mind.

Dualism in Epistemic Rhetoric

These pervasive invocations of "coercion" and "manipulation" as criticisms of persuasive technology are actually what initially prompted me to reconsider the value of Fogg's work. I could not help but hear in these complaints a new iteration of Socrates' classical distinction between good rhetoric ("an attempt to disclose and communicate the truth in unbiased, reasoned speech") and bad rhetoric ("the use of weighted language and specious argumentation to conceal facts and bolster one's cause") (Consigny 35; see also Lanham).6 Simply put, certain forms of digital rhetoric are a priori good (e.g. dialogical modeling, persuasive games) and others are bad (e.g., embodiment, persuasive technology). These new iterations presuppose that an arhetorical realm of digital rhetoric exists apart from thought and meaning wherein the communication of facts or claims can be effectively immunized from bad rhetoric (see Lanham; Lynch and Rivers 7). Furthermore, good digital rhetoric relies upon Cartesian dualism in that rhetoric can only appeal to the disembodied cogito and not to the body (see Dolmage 1).

By contrast, few digital rhetoricians would advocate a Platonic criticism of Fogg. Indeed, many of us would point to Burke's universally cited definition: "Wherever there is persuasion, there is rhetoric. And wherever there is 'meaning' there is 'persuasion'" (Rhetoric 172). While variations of epistemic rhetoric differ, following Robert L. Scott's seminal treatment, a general thesis common to many accounts is that the contingent effects of language games and arbitrary symbolic systems mediate any attempt to ground a representation of reality independent of discourse (Cherwitz and Darwin 192; Scott 13). "Everything," Edward Schiappa claims, "or virtually everything, can be described as 'rhetorical,'" even absent a specific argumentative appeal (260). In other words, the goal of immunizing good rhetoric from bad rhetoric is undermined if rhetorical forces inflect any form of symbolic action.

Arguably, epistemic rhetoric is central to a great deal of research in digital rhetoric. Let me offer three brief examples. While Zappen does not offer a definition, his examples of digital rhetoric, such as Johndan Johnson-Eilola's examination of collaborative hypertext authorship, confirm that it is rhetors' cooperative symbolic actions that are the most important rhetorical elements to analyze; James E. Porter defines digital delivery as an examination of "how audiences are likely to access, engage, and interact with information" (27); and, finally, beyond any procedural argument that a given videogame might seek to offer, Matthew S. S. Johnson documents how "gamer-authors" engage one another in discussion online beyond the immediate activity of play, often revising the texts that they create in order to recruit new members (271). Despite their different subject matter, each of these three researchers share a focus on instrumental human agencies or cognitive forces and do not specifically seek to consider how digital media shape bodies, minds, and environments.

For many researchers, epistemic rhetoric remains an important framework because it enables us to examine the "social-political structures" within various forms of digital media writing (Moberly). Losh's Virtualpolitik argues that concepts such as "freedom" expressed through digital rhetoric are connected to social ideology and gender, race, sexuality, and class (56). To be sure, many of these theorists acknowledge that digital rhetoric is grounded by technology. For instance, in Electric Rhetoric, Kathleen Welch extends Walter J. Ong's "secondary orality" and Marshall McLuhan's "the medium is the message" axiom to claim "[the canon of] delivery is the medium" (1-28). Yet, in asking her students to "[interpret] an apparently ordinary electronic text, such as a Coca-Cola advertisement," Welch does not elaborate the potentially posthuman nature of her argument. She offers only a medium-specific focus on "camera and actor positioning, editing, lighting, production qualities related to camera and actor position," while the primary focus is "probing its ideological positioning as it emerges from the rhetorical canon of delivery" (Contemporary Reception 162, 160; see also Brooke 174). In Welch's description, ideological corresponds to social-construction or social-epistemic approaches to rhetoric.

To be clear on this point, social-epistemic forms of rhetoric are critically important for avoiding Platonic reductions of rhetoric as well as explaining the persistence of ideological worldviews such as patriarchy in digital media. As a recent case in point, the popular American football simulation videogame Madden 16 conceded an ongoing procedural form of patriarchal exclusion in that its August 2015 release was the first version to include female fans as part of the stadium crowd (Newhouse). My larger claim is that epistemic definitions of digital rhetoric do not automatically include the additional role of embodiment or materiality that produce symbolic actions. With respect to the bracketing of the body or materiality, Daniel J. Royer argues that epistemic conceptions of rhetoric presuppose Cartesian mind-body dualism and Kant's theory of the active mind as a knower of reality. Stemming from previous definitions offered by Ernst Cassirer, Barry Brummett, and James Berlin, the epistemic thesis that reality is a mental or social construction grows from a "reduction" of Kant's more nuanced and complex separation of noumenal and phenomenal dualism (Royer 286). In Ambient Rhetoric, Rickert similarly argues that many current epistemic, social, or discursive paradigms of rhetoric invariably limit embodiment or materiality to a passive backdrop or exigency through which cognitive agents employ symbolic action (11; see also Selzer 4; Dolmage 1). In a symptomatic example of this formulation, Berlin claims that, while he does not want to avoid addressing the "force of the material" in rhetoric, "Only through language do we know and act upon the conditions of our experience" ("Poststructuralism" 21). To be clear, rhetoricians such as Royer and Rickert are not alone in raising these criticisms. In Volatile Bodies, feminist theorist Elizabeth Grosz adds that Descartes' separation of the soul from nature in dualism "establishes an unbridgeable gulf between mind and matter … rationalism and idealism are the results of an attempt to explain the body and matter in terms of mind, ideas, or reason" (7).

A Rehabilitation of Foggian Rhetoric

One central problem with explicitly or tacitly bracketing the body from rhetoric is, as a number of rhetorical theorists have argued, that an audience's ideological beliefs are inseparable from their habituated embodied and spatial contexts regardless of the types of rhetorical forms that are privileged (e.g., effective argumentation, cognitive modeling, logic, symbolic action).7 For example, Marshall Alcorn argues that "libidinal attachment and loss" motivate us to act more than rational arguments (qtd. in Jenny Rice 97). Alcorn confirms that individuals do not just receive a deposit of an ideological idea that can be removed when they encounter an idea that causes them to revise their terministic orientations in the same sense that they withdraw money from a bank because ideological dispositions are unevenly entangled with our embodied practices. It is for this reason that Jenny Rice summarizes Alcorn as confirming that "in order to change people's minds, you must also have an effect at an affective level, which is much more challenging" (97; see also Mays). In terms of development rhetoric, the subject of Rice's book, individuals in rural communities may very well know that they should support local businesses and not shop at big box stores. However, an appeal to change a belief does not just present itself to the transcendental reasoning faculties of a disembodied cogito who then wills a means-ends decision (in Kant's sense). The desire to shop locally will be influenced by countless other actors and networks, including public transportation, work and family schedules, automobile ownership, weather, parking availability (and proximity), cost, political orientations, and income. Shopping habits are enacted through bodies, spaces, and technologies and are not innate. This example confirms why I remain drawn to Fogg's desire to include habit as an unavoidable element of digital rhetoric. As a result, step one of my deconstructive thrust is to claim that our definitions and conceptions of digital rhetoric cannot avoid engaging an element of Foggian rhetoric in digital rhetoric. Rather, digital rhetoricians are obliged to rehabilitate his thinking by articulating a more complex understanding of the relationship between habit and rhetoric.

Intriguingly, Fogg and behavior-tracking apps tend to work from a perspective of Kantian mechanism. According to Kant, habit (Gewohnheit or Angewohnheit) could only be a "physical inner necessitation to continue behaving in the same way we have behaved because it detracts from our freedom of mind; moreover, it leads to thoughtless repetition of the same action (mechanical uniformity) and so become ridiculous ... as a rule, all habits are objectionable" (39, 42). Habits function as part of our self-incurred tutelage and are only an inauthentic expression of self. Stemming from mind-body dualism, the Kantian approach to habits do not produce thought, will, freedom, or intellection, but only interfere with our ability to use the cogito's full reasoning capacities. A mechanistic conception undergirds many of our commonsense views of habit, including Marcel Proust's famous description of habit as a "heavy curtain [which] conceals from us almost the whole universe, and prevents us from knowing ourselves" and Parmenides' ancient warning: "let not habit do violence to you in the empirical way of exercising an unseeing eye and a noisy ear and tongue" (Proust 621; Parmenides 58). Many of the related goals of behavior-tracking software promise to transparently reveal the bad (nonconscious) individual or collective habits that interfere with our ability to lead the good life.

As an example, consider Jane McGonigal's gamified habit-tracking app SuperBetter. While her previous discussion of "gameful design" complicates gamification in productive ways (Brown Jr. 40), she does claim for SuperBetter that behavior tracking through gamification can help us "de-bug" ourselves or make us "scientists over our lives" (e.g., make bad habits more visible to a the cogito) ("Welcome"). I am in no way suggesting that SuperBetter is ineffective in using gamification to help some individuals "make an enemy of [their] symptoms" and improve aspects of their lives. However, the use of these metaphors in the context of persuasive technologies inadvertently figures habit as a line of bad code in the otherwise perfectly functioning computer program of the Cartesian theater. Neither good nor bad habits constitute the authentic (de-bugged) self. Rather, the presence of good habits is what confirms that the self is able to realize transcendental capacities for reason, action, or, to use SuperBetter's privileged term, "resiliency" in the world independent from habit.

However, Kantian mechanism is not the only way to figure habits in relation to rhetorical subjectivity. Fogg fails to observe in Persuasive Technology that the relationship between rhetoric and habit as articulated in Aristotle's discussion of ethos predates him by millennia. A great deal of digital rhetoricians in particular are familiar with the definition of ethos from Aristotle's Rhetoric: a manufactured proof or an artistic techne (1356a4). It is a form of appeal that does not depend upon the prior nature or history of the speaker. In standard (epistemic) formulations, Losh discusses how a web designer maintains the "ethos"—"the character or image of rhetorical credibility" (24)—of a legislator's website, and Jeremy Evans articulates how a persuasive game designer creates a "procedural ethos" that a player believes or disbelieves (see also Colby; Sierra and Eyman).

Yet the past few decades have seen a number of scholars reclaim ethos as form of habituation over the course of a rhetor's life.8 Drawing on Heidegger's "Lectures on Aristotle," Hyde writes:

Abiding by this more "primordial" meaning of the term, one can understand the phrase "the ethos of rhetoric" to refer to the way discourse is used to transform space and time into "dwelling place" (ethos; pl. ethea) where people can deliberate about and "know together" (con-scientia) some matter of interest. Such places define the grounds, the abodes or habitats, where a person's ethics and moral character take form and develop. (xiii, emphasis original)

In a passage that supports Hyde's view, Aristotle quotes the poet Evenus: "[Ethos] comes, my friend, by practice year on year—and see: At last this thing we practice our own nature is" (Nichomachean 1152a30-33). According to Aristotle, an individual's "first nature" (immature beliefs, natural temperaments, biological characteristics) is transformed into "second nature" (mature beliefs, virtues) through the process of interacting with various social and environmental factors over time.

Forming an ethos eventually leads to the cultivation of hexis, a term variously translated as "state," "disposition," or "bodily comportment" that characterizes both ethical and intellectual virtues. In one of the few contemporary treatments, Ellen Quandalh confirms that Aristotle's virtues are "characteristics of habits (hexeis) of feeling and action that develop through activities. Thus the name for moral virtue (ethike), is related to ethos" (15). Yet Quandalh's translation of hexeis as "habits" requires qualification precisely because of the ways in which our modern understanding reflects Kantian mechanism. The Latinization of hexis into "habit" actually grants habit its post-Kantian mechanistic definition: a repeated observable behavior pattern that is ingrained and nearly permanent (Carlisle 15). Yet Aristotle's notion of hexis was not reducible to a fixed set of behavioral patterns (e.g., biting one's nails, exercise) for FitBit to intervene in and solve. By contrast, hexis derives from the Greek verb echein. Echein is often translated as "to have," but it is better translated as an "active having" or holding oneself ready for potential ethical action that, if coupled with phronesis, results in a tendency toward virtuous comportment in the use of any technai, from navigating a ship to producing rhetoric (Gross; Malabou, "Preface" xvi).

As a result, hexis is not a temporary feeling (diasthesis) like coldness or anger, but a "dynamic equilibrium" (Lockwood 19) that produces ethical actions while nevertheless remaining the condition of possibility for varied ethical actions across different circumstances—"how an agent holds himself when he acts" (Aristotle, Nicomachean 1105a30-31). The hexeis are also not a list of moral laws to follow, such as the catechism in the Christian tradition. Aristotle says of hexeis that matters of ethics "have as little fixity about them as questions of what is healthful; and if this is true of the general rule, it is still more true that its application to particular problems admits of no precision…. agents are compelled at every step to think out for themselves what the circumstances demand" (Nicomachean 1104a4-10). For example, no one would accuse an elderly individual of a lack of effort (i.e., vice) for failing to meet FitBit's 10,000-step goal each day; nor would anyone accuse a marathon runner of excessive exercise for consistently reaching 30,000 daily steps in training for a marathon. The golden mean depends upon our unique configuration of ethos and hexisin relationship to our specific social and physical environments.

It is important to observe that hexisis not grounded on Cartesian dualism. Aristotle acknowledges that hexeis form through cognitive and noncognitive forces of embodiment and environment (see Lockwood 21-22). Eugene Garver adds, "the virtues are rational first energeiai of the irrational part of the soul" (96; see also Nicomachean 1102b31-1103a1). While other rational powers such as rhetoric allow a rhetor to argue both sides of a given topic, virtue cannot admit opposites. I can ignore my Vessyl's dire warnings and consume an entire bottle of wine for a meal, or I can choose not to. However, a temperate individual will have the habit of not drinking an excessive amount of alcohol relative to a specific social context. In either case, hexis is a fundamentally embodied mode of thinking. Aristotle elaborates on this connection, stating, "For Empedocles says those changing their bodily conditions (hexin) deem to change their thought (phronēsin)" (qtd. in Hawhee 57). M. A. Wright interprets this line from Empedocles as, "when men change their hexis, they change their thinking" (qtd. in Hawhee 58). Debra Hawhee takes Wright's translation a step further to suggest, "Hexis equals thought. Thought does not just happen within the body, it happens as the body" (58). In perhaps the earliest articulation of a non-Cartesian theory of rhetoric, Empedocles' hexis via Aristotle confirms that embodiment—habit formation—co-constitutes or shapes thought and cannot therefore be immunized or bracketed from digital rhetoric.

Habit as Rhetorical Ontology

Even through a cursory—and by no means comprehensive—outline, Aristotle confirms that habit (hexis) is not just an object of an instrumental design techne, but a fundamental element of digital rhetoric's condition of possibility that encompasses the myriad ways in which designers (rhetors) and users contract habits through specific bodies, discourses, and environments—digital or otherwise. Habit can be seen as part of what Rickert via Heidegger describes as a "rhetorical ontology." He maintains that "ambient rhetoric" does not just treat with the way human beings generate and negotiate knowledge, but with the way that "human beings are" (xv): "Rhetoric is an emergent result of environmentally situated and interactive engagements, redolent of a world that affects us, that persuades us prior to symbolicity" (Rickert 34).

For Aristotle, habit is a similarly important element of how we are affected and formed as rhetorical beings prior to symbolic action and epistemological forces. As a result, analyzing digital rhetoric includes not only the content of a given rhetorical practice, but the entire constellation of embodied habits (ethos and hexeis) that condition, produce, and constrain digital audiences and rhetors alike.

While it is beyond the scope of this essay to address in detail, it is worth noting that Aristotle's formulation remains influential among subsequent philosophers and theorists of habit. The philosopher Clare Carlisle observes, "in one sense, all [ontological approaches to habit] simply elaborate the Aristotelian thesis that moral virtue is 'the child of habit [ethos]', which 'ends up as our nature [phusis]'" (103). In a prominent example, Ravaisson's pioneering essay "On Habit," which influenced subsequent thinkers such as Bergson, Merleau-Ponty, and Deleuze, opens by citing a line from Aristotle's On Memory and Recollection, "Habit is a second nature," and declares in the first sentence, "Habit, in the widest sense, is a general and permanent way of being" (25). Dewey similarly challenges the Kantian separation between mind and habit by offering a profound description of habit's power: "habit does not preclude the use of thought, but it determines the channels within which it operates. Thinking is secreted in the interstices of habits" (The Public 158; see also Crick; Stroud). At the same time, Aristotelian conceptions of habit until very recently have yet to command much specific attention from researchers. Indeed, A History of Habit, the first comprehensive edited collection on habit in the Western philosophical and theoretical tradition, was published in only 2013. In the introduction, editors Tom Sparrow and Adam Hutchinson confirm that mechanistic conceptions have been the dominant post-Kantian trend: "the explanatory power of habit is immense. And yet, with some exceptions its ontological constitution is rarely investigated, or it is overshadowed by its physical description" (3).

To get a sense of how an ontological approach to habit can shift our conceptions of digital rhetoric, I want to offer a few concrete examples. Media theorist Ingrid Richardson suggests that while a given version of a mobile device (e.g., an iPhone) offers the same system protocols, interface, and physical design to each user, the unique ways that our subjectivities become folded through habits of use form our collective and individualized "technosoma" (Richardson para. 12; see also Pigg; Swarts). We engage in:

[activities] such as stooping, bowing the head to conceal the face and reduce audibility, shielding one's mouth with the hand to define a provisional private space, or deliberately not altering one's trajectory or visual/facial orientation, and directing one's gaze into the middle distance, as is the case with the more blatant Bluetooth pedestrian. (para. 12)

Echoing Dewey, Richardson argues that these embodied habits are produced both by the embodied and spatial contexts of use, including our own bodily practices and the physical design of technologies themselves, and by social forces. In articulating the idea of "emplacement," mobile media theorist Jason Farman offers a similar claim (22). While I may access the same Wikipedia content on a mobile phone on the DC Metro and a desktop PC at home, my embodied conditions of access differ at each site, which in turn will impact the particular hexeis that I form through use. Far from opposing the social and embodied spheres, Farman importantly observes that these modes of comportment will be inflected by differences of class, race, publicness, privateness, and gender. Embodiment is what can both produce and undermine these social relations.

It is critical to highlight that the conception of habit—ontological or mechanistic—matters a great deal in terms of how researchers theorize the relationship between technology and thought. Echoing previous positions argued by Walter J. Ong, Antonio Damasio, Catherine Malabou, or Andy Clark, Katherine Hayles comments, "research indicates that the small habitual actions associated with web interactions—clicking the mouse, moving a cursor, etc.—may be extraordinarily effective in retaining (or more accurately, repurposing) our neural circuitry, so that the changes are not only psychological but physical as well" (2). For Hayles, the unavoidable potential of technologies to shape and condition our minds and bodies is a largely positive phenomenon in that networked technologies form new habits of reading ("distant reading") and communication to enable new rhetorical practices. The notorious print literacy defender Nicholas Carr concedes a similar view, writing, "As the same experience is repeated, the synaptic links between the neurons grow stronger and more plentiful through both physiological changes, such as the release of higher concentrations of neurotransmitters, and anatomical ones, such as the generation of new neurons" (27). However, quite unlike Hayles, Carr implicitly doubles down on the Kantian conception of habit in claiming that certain (digital) habits are a priori negative obstacles to (print-based) reason. By extension, habit's ontological or mechanistic description functions as a useful index for determining how a given conception of digital rhetoric engages embodied and material forces in the world.

Despite habit's important capability to enable embodied action, many post-Aristotelian thinkers also highlight habit's potential danger. Even if it is necessary to disagree with Kant's reduction of habit to mechanism to argue that habits co-produce thought (i.e., presence), Kant is by no means incorrect in his concern that habit's mechanistic elements can produce nonconscious forms of forgetting (i.e., absence). For example, Dewey firmly acknowledged, "The influence of habit is decisive because all distinctively human action has to be learned … habits bind us to orderly and established ways of action because they generate ease, skill and interest in things to which we have grown used" (Public and Its Problems 149). To clarify, Dewey suggests that we develop habits specific to our localized ecologies in order to free up our minds and bodies to perform different or more complicated tasks, but these habits in turn can compel the performance of nonconscious actions without much conscious attention. As a result, Dewey is well attuned to habit's pharmacological danger even if he does not use this Derridean term. À la deconstruction, it is important to acknowledge that allowing habit to have a fundamental role in digital rhetoric does not allow us to re-center habit as a new moment of metaphysical clarity.

Conclusion: Forging New Persuasive Technologies

To bring my argument full circle, hexis illustrates why digital rhetoricians need to develop a more complex view of persuasive technologies and Foggian rhetoric as opposed to explicitly or implicitly bracketing habit's role. Playing a persuasive videogame such as Windfall may indeed cause "metanoia" (a change in mind), as Ken S. McAllister claims (54). Yet, thinking back to Alcorn's and Rice's respective observations about affect, habit confirms that digital rhetoric will invariably require contending with more than just cognitive or social-epistemic forces. Effective forms of digital rhetoric will also have to address the complex ways in which affective worldviews become habituated into second natures through our digital hexeis of virtue and vice. In support of this idea, I want to close by examining a more sophisticated version of Polar Bear: Kac's Teleporting an Unknown State. In this art installation, Kac encouraged direct participation from an audience in an interactive website that "combined biological growth with (remote) Internet activity" (221). Kac made a distant audience responsible for activating a light on a video projector to ensure a plant seedling's growth in conditions of total darkness. The teleportation of photons (light particles) became a "metaphor of the Internet as a life-supporting system" (223). Participants could find a nine-image grid online that displayed different locations of plants around the world (Slovenia, Vancouver, Paris, Moscow, Chicago, Antarctica, Sydney, and Cabo San Lucas) (see fig. 4).

Fig. 4: Eduardo Kac's Telepresence in an Unknown State

When users saw black rectangles, a given plant was receiving no light from the actions of any other users. In response, a user could select a plant's black rectangle and the image would stay lit only five minutes, thereby encouraging other users to keep continually checking the website to ensure that the plants had enough light to survive. Importantly, Kac did not place his audience inside a collective Skinner box that forced them to form new behaviors. Rather, what he offered was a persuasive-technology-like opportunity to cultivate a hexis of care. Similarly, individuals who lack access to goods like friends, Aristotle maintains, will never find happiness (eudaimonia) simply because they will have fewer opportunities to engage in virtuous practices of friendship to form a hexis (Nicomachean 1153b17–19).

Echoing Aristotle's advice, Kac confirms that individuals need opportunities to enact new hexeis through digital interfaces beyond the narrow utilitarian aims of nation states' behavior tracking or individualized habit-forming apps. With O'Reilly's algorithmic regulation as a case in point, these tracking apps often function to place the responsibility for habit formation on individuals and therefore abstract the habits that produce individual failure from any larger political-economic relations (see also Wiegel; Davies). For a similar reason, Dewey maintains that habit only threatens to become a form of mindless Kantian repetition if it is "the product of conditions that are uniform because they have been made so mechanically—as in much school and factory work" (Logic 109). Indeed, implicit in Kac's installation is the claim that we need to develop new means to shape a more creative range of habits that will support non-instrumental forms of dwelling with the environment. As a general guideline for his artistic endeavors, Kac argues, "More than making visible the invisible, art needs to raise our awareness of what firmly remains beyond our visual [and sensory] reach but, nonetheless, affects us directly" (236).9 Teleporting an Unknown State confirms that persuasive technologies can in fact play a role in these critical revelations.

In some ways, Kac's experiment would be even more successful in the era of social media in which audiences are habituated to continually hit "refresh" to receive new updates and posts. We keep virtual farms alive in the social media game FarmVille 2, so why would digital rhetoricians not seek to embody more complexly Fogg's kairos-as-suggestion-technology by attempting to connect these habituated forms of procedural identification to caring for physical plants as well? In support of this aim, for example, the free online trivia game Freerice takes advantage of precisely these habituated forms of action (United Nations). Answering each new set of questions requires a player to refresh the interface page, thereby increasing advertising revenue. Freerice donates its advertising revenue to purchase rice for impoverished countries.

Speaking to habit's ontological character as pharmakon, hexis highlights precisely the power of our digital habits to trigger and compel various nonconscious forms of repetition that can make habits so difficult to change. Nevertheless, "there is," Catherine Malabou affirms via Derrida, "responsibility in the double movement of giving and receiving form" through plasticity—another contemporary articulation of hexis (What Should We 30). Since our habits shape our rhetorical practices, we have some say in terms of which habits to cultivate, even if we cannot will all of the habits of digital rhetoric that enable us to act in the world. Habits are always potentially in formation with each word we type on our laptops or each step we take while walking to work. While a given procedural habit may threaten to become recalcitrant second nature, habit simultaneously offers room for optimism because by definition a habit is always changeable. As a result, it is my suggestion that the enduring contribution of Fogg to our definitions of digital rhetoric lies precisely in demonstrating that we need to better engage the not purely social or epistemic dimensions of habit as an important part of digital rhetoric's ethical project in the current moment.

• 1. My book project, Procedural Habits: The Rhetoric of Videogames in Embodied Practice, explores in detail the rise of Foggian rhetoric and rhetorical and philosophical conceptions of habit in the context of videogames, gamification, and other game-like procedural media as well as the post-Aristotelian lineage of ontological approaches to habit.
• 2. These dimensions also resonate quite well with rhetoric and composition's past and current engagements with ecology (Cooper), material rhetorics (Biesecker and Lucaites), postcomposition (Dobrin), Latour (Lynch and Rivers; Jeff Rice), object-oriented rhetoric (Barnett; Rivers and Brown), ambient rhetoric (Rickert), embodied rhetoric (Fountain; Pigg; Swarts), and rhetorical transformation (Gries).
• 3. In fact, my colleague at George Mason University, Douglas Eyman, the coeditor along with Cheryl E. Ball of our field's first digital rhetoric journal, Kairos: A Journal of Rhetoric, Technology, and Pedagogy, reminded us of the impasse between symbolic action and nonsymbolic motion in the opening presentation of the 2015 Indiana Digital Rhetoric Symposium (Eyman, "Defining"). He declared his firm—which is not to say uncritical—support of a human-centered and largely social-epistemic definition of digital rhetoric. Elsewhere, Eyman does offer a commendably open-ended definition, stating, "the term 'digital rhetoric' is perhaps most simply defined as the application of rhetorical theory (an analytical method or heuristic for production) to digital texts and performances" (Digital Rhetoric 44).
• 4. Examples include Thomas Rickert's Ambient Rhetoric, Byron Hawk's A Counter-History of Composition, Alex Reid's The Two Virtuals, Jeff Pruchnic's Rhetoric and Ethics in a Cybernetic Age, Sid Dobrin's Postcomposition, Paul Lynch and Nathaniel Rivers's Thinking with Bruno Latour in Rhetoric and Composition, Laurie Gries's Still Life with Rhetoric,Jeff Rice's Digital Detroit, and Collin Gifford Brooke's Lingua Fracta. As a case in point, while Anne Frances Wysocki, along with coeditor Kristin L. Arola, "do[es] not argue that digital technologies profoundly change our possible embodiments," she does acknowledge this aspect (Wysocki 8). In the introduction to Composing(Media) = Composing(Embodiment), Wysocki concedes that she and Arola "[largely] leave these arguments to others" (8).
• 5. Discussion of this example can be found in Dillahunt et al.'s "Motivating Environmentally Sustainable Behavior Changes with a Virtual Polar Bear."
• 6. My book project offers a more detailed analysis of these conversations, including using Richard Lanham's weak defense of rhetoric as a lens through which to read these dismissals of Fogg's work. I also argue in "'Can We Name the Tools?' Ontologies of Code, Speculative Techné and Rhetorical Concealment," which is revised in the book project, that the philosophical division between "good" and "bad" rhetoric is central to software studies' treatments of the materiality of code as well.
• 7. Thomas Rickert's Acts of Enjoyment and Diane Davis's Breaking Up (at) Totality draw on the philosopher Peter Sloterdijk's "cynical reasoning" to highlight precisely this point.
• 8. See, for example, work by Susan Jarratt and Nedra Reynolds, Michael J. Hyde, Jodie A. Nicotra, Kristie S. Fleckenstein, and Thomas Rickert (Ambient Rhetoric).
• 9. My book project offers a sustained analysis of Heidi Ray Cooley's Finding Augusta and the Augusta app. Cooley's work is foremost among media studies scholars in utilizing big data tracking and GPS algorithms to help individuals shape new procedural habits.