Show Less
Open access

Rhétorique et cognition - Rhetoric and Cognition

Perspectives théoriques et stratégies persuasives - Theoretical Perspectives and Persuasive Strategies


Edited By Thierry Herman and Steve Oswald

Ce volume met l’accent sur le lien entre démarches cognitives et art du discours qui a toujours été un des enjeux de la rhétorique. Sans ajouter une nouvelle couche à l’examen critique des sophismes, les contributions de cet ouvrage n’ont pas pour but de dénoncer les effets de certains schèmes argumentatifs que d’aucuns jugeraient fallacieux, mais d’étudier leur fonctionnement et leurs effets cognitifs hic et nunc. Quels sont les mécanismes qui expliquent la « performance » des arguments réputés fallacieux ? Comment fonctionnent les stratégies rhétoriques à l’intersection entre cognition, sciences du langage et société ?

This volume gathers contributions from two disciplines which have much to gain from one another – rhetoric and cognitive science – as they both have much to say in the broad realm of argumentation studies. This collection neither condemns the fallacious effects of specific argument schemes nor adds yet another layer to fallacy criticism, but studies how argumentation and fallacies work, hic et nunc. What are the linguistic and cognitive mechanisms behind the «performance » of fallacious arguments? How do rhetorical strategies work at the interface of cognition, language science and society?
Show Summary Details
Open access

A case for emotion awareness: Evgenia Paparouni

A case for emotion awareness

Evgenia PAPAROUNI, Université Libre de Bruxelles


Blinded by passion, jealousy or anger? Inspired by love for freedom, getting courage through cherishing a purpose, self-abnegation and sacrifice for others? Be politically indignant? Adapt in society, remember, pay attention, interpret, justify, take a decision and reason? According to recent research findings (Damasio 1995, Blanc 2006, De Houwer & Hermans 2010), all these involve some form of emotion.

Albeit pervasive, emotion is controversial for argumentation studies as some theories of argumentation account of it as an obstacle to reasoning or as an instrument of ‘manipulation’.

The ultimate purpose of this contribution is to deal with both these claims on emotion from the point of view of a rhetorician. To do so, I start by presenting the divide over emotion in argumentation studies (section 2). Although the term ‘manipulation’ seems to be potentially argumentative in the eyes of a rhetorician, I try to come to terms with the claim of manipulation, taken to mean the unsatisfactory results of a debate, portrayed as deceptiveness by the participants in it (section 3). I focus in particular on the efficiency of emotion. Then I make a bibliographic review of contemporary research in psychology and theory of mind concerning the relation between emotion and cognition, to see if the divide existing in argumentation studies also exists among different schools of thought in experimental psychology and in theory of mind. Even studies which contrast logicality to emotion, in the 80s do not allow us to conclude that emotion is fallacious (section 4.3), as it sometimes appears to improve the agent's performance in the experimental task. From that, I draw attention to a contemporary paradigm shift in cognitive science, as far as the study of emotions and the perception of human rationality is concerned. According to the new paradigm, inaugurated by the seminal works of Damasio (1995), ← 129 | 130 → rationality is composite and it includes emotions, which confirms ancient rhetoric’s intuition (section 4.4).

As verbal interactions allow for conflict by definition, attempts to subvert a composite rationality can be accounted for in the new paradigm. Emotions can be involved in such attempts/ unsatisfactory outcomes, as they involve a rapid appraisal of a situation and do not always avail themselves of propositional content (section 4.5).

On the basis of such an extended view of rationality, I propose an alternative way out of the ‘manipulation’ claim conundrum, through debate on emotions themselves. In such a debate emotion is not necessarily contrasted to logicality, but to another emotion, more suitable (section 5). Rhetoric can remain faithful to its descriptive vocation and still avail itself of a pedagogical project. It is through familiarisation with rhetoric, theory and practice, as well as through awareness of our own emotions that we can become more satisfied with our communication practice (5.1). It is not a matter for the rhetorician to whistle blow manipulation. Claims of manipulation are tested in the communication practice itself (5.2).

2.Emotion and argumentation theories

Despite the contemporary revival of the interest in emotion in various fields of the natural and social sciences, no unanimity exists, in argumentation studies, on the legitimacy and treatment of emotion. From this point of view, we could schematically classify argumentation theories as follows: (a) theories that are not interested in emotion, as they exclusively study the logical structure of arguments; (e.g. S. Toulmin 1958, J. Woods 2004) (b) theories that consider pathos as a means of proof, like classical rhetoric and most new rhetoricians (e.g. Plantin 2000, 2012); (c) theories that see the appeal to emotion as a fallacy or as strategic manoeuvring (e.g. Van Eemeren & Grootendorst 1994; 2006); (d) theories that aim at evaluating emotions' legitimacy under specific conditions (e.g. Walton 1992).

Aristotle qualifies pathos as one of the three legitimate discursive proofs and gives ample instructions on its usage in his third book of Rhetoric. ← 130 | 131 → Quintilian more strongly recommends the use of pathos for the sake of efficiency. Contemporary rhetoric studies meet again with the classical tradition.1

In the fifties, Neo-Rhetoricians tended to neglect pathos.2 Nowadays – and probably in parallel with the changing paradigm in all social sciences – rhetoric turns to the Aristotelian pathos in order to account for the role played by the speaker’s and the interlocutor’s emotions in persuading an audience and modifying an attitude. Emotions are not only a means of proof, but also a matter of debate in itself, since their individual appropriateness to the situation is questioned by participants in argumentation (Plantin 1998, Micheli 2010). Pathos is a central analytical category that translates speakers’ subjectivity in situated speech.

In contrast to Neo-Rhetoricians, normative approaches to argumentation adopt a suspicious stance towards emotion, a stance that ranges from total rejection to acceptance under conditions. First-wave pragma-dialectics considers appeals to emotion to be fallacious (van Eemeren & Grootendorst 1994). The extended theory allows for strategic manoeuvring provided the rules of critical discussion are complied with. Persuasion is tolerated as long as it does not take priority over critical discussion (van Eemeren & Houtlosser 2006). We owe the most systematic normative approach to emotion to Walton (1992: 27-28). Walton focuses specifically on emotion trying to give it some legitimacy through a comprehensive theory rather than dealing with it as an obstacle or a side product (and necessary evil) in argumentation. He proposes a case by case evaluation of emotion in argument along a double pragmatic criterion: its effects on the argumentative goal of the participants and the relevance to the debated matter. If the goal of argumentation is promoted, through revealing deep commitments and through motivation for action, appeals to emotions are legitimate and beneficial. If the dialectical goal is diverted and pathos leads to suppression of evidence and entrenched positions, emotions are presumed to be fallacious.

This balanced approach has the merit of contextualising emotion and linking it to six variations of critical dialogue (1992: 19-22) which remotely ← 131 | 132 → remind us of ‘genre’. Nevertheless human communication offers a huge amount of ‘genres’ that cannot be reduced to the six variations of critical dialogue proposed by Walton.3 More importantly, an approach that defines argumentation as problem/conflict resolution excludes broader definitions of argumentation that do not take scientific discourse as a model. This exclusion does not answer the question why people engage in argumentation over problems they know to be unsolvable or without entertaining any expectation to convince. People may indeed engage in argumentation to assert their position, to bond with others or to underline difference of opinion which has to be resolved through voting.4 As a result, emotion in ‘genres’ outside the six variations is unnecessarily disqualified.

This schematic overview shows the division of argumentation studies over the role of emotion in argument. Before I turn to cognitive science for some support on the role of emotion in argument, I will touch upon another non-consensual topic in argumentation studies: Manipulation in terms of normative versus descriptive theories.

3.The claim of manipulation

Rhetoricians, in general, abstain from offering systematic norms to identify ‘manipulation’ in the name of a descriptive stance. They are even cautious towards the term manipulation as they consider it to be a metalinguistic argument, one pointing to the frame and not the content of the matter. C. Plantin underlines that one will never attribute the manipulator’s role to oneself (2002: 239). He considers the claim of fallacy to be an argument in itself too, that he ironically calls ‘ad fallaciam’ after the long series of fallacies historically identified under their Latin name (1995). ← 132 | 133 →

Nevertheless, the derogatory term ‘manipulation’ exists in language and we are all familiar with claims of deceptiveness on behalf of an observer of the situation and of the ‘victim’ oneself with the benefit of hindsight. Some of these claims target speech that may involve the use of emotions. Take for instance the following examples:

(a)This new car really matches your social status. (need to feel and seem superior);

(b)Immigrants steal our jobs. (fear, feeling of unfairness);

(c)We are the best. (pride).

All these phrases may be inaccurate or subjective. They can also elicit in the addressee in potentially desirable or undesirable behaviours, in his eyes and/or in the eyes of society. The same phrase may incite both a desirable and an undesirable behaviour: (c) does not lead to the same result if the ones who are the best are a soccer team or a nation on the verge of war. The car buyer may be absolutely satisfied with his choice, although in our eyes he spent a fortune for nothing. And whether we like it or not, Europeans are not unanimous on the number of immigrants a country can assimilate: We see that the addressee and an observer may not agree on whether the phrase has been deceptive.

Besides, when the claim of manipulation is made on behalf of an observer, we can infer the observer’s own (opposite) position; we can also infer (semantically) that the person who makes the claim (often a representative of a normative theory of argumentation) portrays the addressee of the manipulative phrase as a victim in need of protection, probably through some norms/discourse ethic on what is inappropriate in argumentative speech.5 Moreover, we infer that if such a discourse ethics prevailed and its rules were respected, people would agree more often among them, they would not live beyond their means and they would not be xenophobic or imperialistic any more. Due to these semantic inferences from the term manipulation, I consider it to be argumentative in the sense of Anscombre & Ducrot (1983). ← 133 | 134 →

In a nutshell, I believe that the above statements (a) to (c) are not sufficient per se to prove that there is manipulation. They have to be interpreted in a specific context. This is why I am referring to claims of manipulation. I do this only to the extent that such claims are used by scholars and laymen to dismiss the explicit or implicit appeal to emotions in discourse.6Now, what is it that makes emotions so powerful and alarming?

The following questions are open:

(i)If the use of emotion in argument is fallacious according to pragma-dialectics, why are audiences still swayed by it?7

(ii)How does a self-poised descriptive/rhetoric theory of argumentation account for manipulative scenarios, i.e., situations where agents themselves recognise that they have been driven astray by skilful communicators? Does their proclaimed ‘descriptive’ stance amount to permissiveness? I will come back to the issues under (ii) in section 5 of this article.

Rhetoricians circumvent the questions under (i) by adopting a descriptive approach (not reasoning in terms of fallacies), by admitting pathos to be a legitimate means of proof and by referring to psychology for the study of the cognitive mechanisms behind persuasion. This approach though does not come to terms with an empirical phenomenon: the situation where audiences feel themselves either misled by skilful communicators through the use of/appeal to emotion or trapped by their own emotional predispositions, independently of their interlocutor’s intention, or both things at the same ← 134 | 135 → time.8 In more general terms, we would like to test rhetoric’s answer to the users’ critics on the unsatisfactory outcome of a debate where emotions are involved.

Apart from answering such legitimate complaints, a second reason to deal with the claim of ‘manipulation through emotion’ is the need to test a theory’s implicit model of human rationality. Theories of argumentation in the past were much too often influenced by the Cartesian dichotomy between the body and the mind, in the sense that they perceived reasoning to be emotionless and that they aimed at isolating speech from the speaker and the situation, to retrieve pure argument. This dichotomy has been reinforced in the era of informatics through the so called computer metaphor of the mind. Nowadays, the dichotomy is revisited and criticised (Damasio 1995, Blanc 2006).

On the other side, even if we take as a basis a broader rationality which includes emotions, we can still admit the possibility of a subversion of it, be it intentional or accidental, be it at the production or at the reception side of communication, especially as verbal interactions allow for conflict. The concession to the possibility of subversion of rationality may look strange after the misgivings I expressed on manipulation at the beginning of this section. We will see in section 5 how it is possible to come to terms with attacks to rationality with no need for the analyst to adopt a normative stance.

For these reasons, instead of circumventing the questions under (i) and (ii) I will visit reported findings of cognitive experimental psychology in association with insights coming from the philosophy of mind on the relation between emotion and cognition9 to answer them. ← 135 | 136 →

4.Emotion and Cognition: from reciprocal influence to consubstantiality

4.1.   Some key ideas

In cognitive science, the return to subjectivity and emotion is a relatively recent but dominant trend; the current literature on emotion is bulky, controversial, non-conclusive but promising.10 I will limit myself to concisely underlying points of convergence that are relevant to the purpose of this chapter:

(a) It is admitted today that emotions are plural and composite. They consist in an appraisal of the situation, a preparedness to act, and several somatic and representational responses, i.e., they involve a cognitive, a motivational, and a physiologic component, all three acting simultaneously, in reciprocal influence. Appraisal concerns in particular the relevance of the stimuli for the individual and to the task, as well as the possibility to cope and to control. We can describe emotions through variables such as intensity, (positive or negative) valence, duration, familiarity and frequency. This definition fits the rhetorical use of pathos: adapt the speech to what is relevant for the audience, namely strong preferences, positive or negative.

(b) “The brain as a network of simultaneous and interactive operations replaced the computer as the dominant metaphor of the mind” (Chapelle in Blanc 2006: 63). Cognition is incarnated. Operations are parallel and not sequential. Emotion is not ‘noise’ in the system but part of it, meaning that we cannot escape the presence of some form of it. It seems to me that ← 136 | 137 → according to this new paradigm and contrary to the computer metaphor, emotionless entities are corpses, autists or machines. When I am sitting calmly in my office looking at the sky, I do have some form of emotion not necessarily intense or recognisable, but still inseparable of what we call thoughts. As a consequence, we cannot deal with emotion in terms of present/absent. We have to deal with it in a continuum of intensity and description/qualification. This element looks important to me to argue that the attempt to banish emotion from the argumentative practice is lost labour.

(c) Lower and upper cognition limits are blurred by the fact that all levels interact. Memory and attention are seen as selection of information according to a relevance criterion. This fact underlines the subjective input of evidence used for such high-level tasks as priority setting, reasoning and decision making. Points (b) and (c) are crucial for argumentation theories, as each of them presupposes a model of human rationality which does not in all cases include emotions.

(d) Despite the emphasis on the importance of stimuli appraisal for the causation of emotions, automatic/non-conscious mechanisms attract more scientific interest at the moment (Blanchette & Richards 2010).11 Automaticity and complexity explain that agents may disclaim ownership and control over the effects of emotion. In my view this explains historically the awe and suspicion of scholars and laymen towards emotion. A firm belief may be equally or even more erroneous and harmful than an emotion, but contrary to what happens in the case of emotions, we have the impression to control our own beliefs. We also can avail of more time to form a belief or an opinion, while emotion is spontaneous, rapid and unpredictable. These traits make emotions’ strengths and weaknesses at the same time. Speed is important in appraising the relevance of a situation, but as we will see in section 4.5, appraisal is not knowledge.

(e) Individuals can provide verbal accounts of emotions and also (mostly narrative) scripts involving their causes and consequences (Blanc 2006), i.e., they have representations of their own, and other people’s, emotions. Thus, what may start as an automatic/non conscious reaction can then be formulated into words, questioned and debated. This point is important for ← 137 | 138 → what I claim in section 5, namely that formulation into words and debate diminishes the risks associated with emotion.

(f) Emotions are a challenging test for scientific theories, terminology, models and research methods. Despite the role played by parallel, partly automatic and interactive operations in the causation of emotions (see points (b), (c) and (d)), research models, methods, parameters, models, and not least terminology, are still inspired by the old analytical paradigm that separates emotion from cognition.12 For instance, the relationship between emotion and memory is studied in terms of linear causes and effects, in spite of significant evidence for consubstantiality. As a consequence, it is very difficult to talk about consubstantiality of emotion and cognition, even among scholars, as we are imprisoned in old analytical and linguistic categories.

4.2.   Emotion and logicality in experimental psychology

A good example of ambivalent relation between cognition and emotion (viewed as two separate phenomena) is the relation between memory and emotion. The effect of emotion on memory can be both distorting and enabling.

It has been experimentally shown that there is a link between emotional states and memory performance. According to these, emotion distorts, reconstructs and flaws memory. In general, emotional information attracts attention and is easily recalled. On the opposite side, a protective perceptual defence may be activated in the case of traumatic experiences. There is experimental evidence that past experience is recalled and interpreted according to present moods and attitudes (Bower et al. 1978, Mac Farland and Ross 1987, Spiro 1977, Uranowitz in Blanc 2006: 74, 91). Moreover, it is possible to induce false memories (Loftus 1978 in Blanc 2006). Cases have been reported of convicts who came to believe their own fake confessions (Blanc 2006: 96). ← 138 | 139 →

On the other hand, suppressing emotion worsens performance in memory tasks (Gross 1998 in Blanc 2006), creating a catch twenty-two for the seekers of emotion-free memory. A possible explanation for these observations is that emotion selectively orients attention to information. Subjects prefer inferentially coherent stories to fragmented information and, as we saw above, they avail themselves of narrative accounts/patterns when dealing with emotions.

Memory and attention influence higher order cognitive functions. There is evidence that interpretation of ambiguous stimuli is conditioned by the person’s affective state, with anxiety-congruent interpretations being the most salient example (Blanchette and Richards 2010: 278). Empirical studies have also documented that affects influence risk perception (Johnson and Tversky 1983, Constans in Blanchette and Richards 2010).13

Emotion is shown to impair logicality and deductive reasoning, probably because it diverts attention from the task when it is incidental to it. On the contrary, when emotion is inherent to the experimental task (integral), performance improves, perhaps due to the agent’s better concentration. According to Blanchette and Richards (2010: 300-303) this paradoxical result has to be further studied. A rhetorician is sensitive to the role of emotion as a catalyst in performing a task as it looks obvious to him/her that emotion, as a link to the person’s values and interests, plays a crucial role in persuasion.

Affective states are reported to have an influence over the strategy deployed in reasoning (Worth and Makie 1987 in Blanchette and Richards 2010). According to literature on ‘depressive realism’, sadness seems to be a predisposition for systematic processing (focus on the argument), while positive-valence emotions and anger favour heuristic processing (focusing on the source of the message and its superficial features). These last findings do not raise the question whether systematic processing is advantageous to heuristics and, if so, why.14 ← 139 | 140 →

I must confess that I am not satisfied with the subtlety of this evidence. Although the literature review I draw this information from was published last year (2010), the studies summarised were realised in the eighties, and thus could not anticipate recent theoretical evolutions in the field. A first layman remark is that the need for clear-cut experimental parameters leads to levelling complexity. In fact, many of these results reported are open to manifold interpretation. Behavioural psychologists alternate narration, pictures, music and semantic variants of the same term to induce emotion, but there is still room for research on the incidence of specific discourse ‘genres’ in emotion triggering (Blanc 2006). From the point of view of a rhetorician, the information that an upset person has poor results in mathematical tasks is of little use. Although D. Walton would probably find in this information a support for his view, namely that emotion is acceptable only when it is relevant to the objectives of the debate, in our opinion it is not easy to delineate relevant and irrelevant emotions, not least to do it objectively. Although distraction of attention by an emotion incidental to the task is an interesting phenomenon, mathematical calculations are very different from the texts which usually constitute a rhetorician’s corpus of study. Besides, a rhetorician would tend to be more interested in the role of emotion as part of the experimental task, as such a role would echo much better the role of pathos in rhetoric.

My impression is that theories on emotion are much more advanced than models of observation. Researchers themselves agree that it takes time to incorporate theoretical insights into experiments and models (Blanc 2006). Finally, although experimental psychology does not explicitly adopt a normative stance on emotions, experiments echo views and prejudice of their designers over the concept of rationality. It will take some time for behavioural experimental psychology to incorporate the recent theoretical views of psychologists in models of observation. For the time being, it is difficult to assimilate the above mentioned findings in a rhetorical framework, because most of the times they are the product of a restricted vision of rationality, modelled on the computer metaphor of the mind. Although these studies consider emotion and logicality to be separate ← 140 | 141 → variables, they do not confirm the view of the first wave of pragma-dialectics (1994) that emotions are fallacious (section 3 (i)). This is not surprising. These studies have an empiricist impetus and do not set themselves a model to confirm or a normative task to perform. Instead, they underline the importance of emotions to seek coherence and relevance and its involvement in high-order cognitive tasks.

4.3   Emotion and bias in the work of J. Elster

A far more convincing interpretation of bias is proposed by the philosophy of mind. A cognitive-conative approach interprets these irrational emotions as an incongruent set of beliefs and desires (Elster 1983). We owe to this tradition the exploration of such phenomena as self-deception and wishful thinking. Such approaches are compatible with psychological theories that emphasise the activation of parallel mechanisms in emotion triggering. It is also possible to link their research topics to the interest of appraisal theories for goal-incongruent emotions.

I consider Elster’s approach to be a balanced one, as it focuses on the role of emotions in indicating that values are at stake (2003) and not only in ‘subverting rationality’ in the traditional sense of it (1983). Elster maintains a terminology which opposes desires to beliefs, but allows for their cross-contamination. His account on emotion has evolved in time along the shift of paradigm presented in 4.4.

4.4.   A paradigm shift in the view of Rationality

A decisive shift in the study of emotion took place with Damasio’s seminal work (1995) on emotionally impaired patients and with the publications on emotional intelligence that followed (Mayer and Salovey 1995, Goleman in Blanc 2006). The contemporary view of psychology on emotions is that they are necessary for social adaptation. Practical reasoning and value negotiation are part of it. Cognitive psychologists are now interested in the study of behaviour in relation with the attribution of subjective meaning to the environment (Blanc 2006: 66). New theories are in search for new models ← 141 | 142 → and evaluative scales that do not associate emotion with logicality in the strict sense. This shift of paradigm probably accounts for the decline of experimental studies based on self-reporting. Rhetoric studies too advocate a paradigm shift that would lead to viewing rationality as encompassing emotion and subjectivity, in line with Damasio’s work. For instance Dominicy (2001) underlines the propensity for action sought in epidictic praise. Today potential for action is consensually considered to be the evolutionary functionality of emotions.

In accordance with this shift of paradigm, some philosophers of mind account for the functionality of emotions in terms of their unique association with values. Some philosophers consider emotion as an equivalent of perception that takes some value as its Intentional object (Tappolet 2000, Livet 2002). According to Livet, emotions are indices of values we cherish; they can be shared socially and they motivate action. Resistance to modifying beliefs when new data imposes reconsideration is related to the association of these beliefs with values through emotions.

Having strong feelings about the way things ought to be or ought not to be (which is a core issue in argumentation) is far from being irrational, but may bring agents to irremediable disagreement. For instance, an appeal to pity may be irrelevant to the discussed topic for some participants in the debate, while other participants, though, may consider compassion to be a value that overrides other imperatives. Notably, sacrifice and self-abnegation cannot be accounted for by optimal choice or utilitarian models (Elster 2003, 2010). Disagreement over values is not an index of irrationality, or of suboptimal judgment under the effect of these strong feelings, but a manifestation of incarnated preference.

Cognitive psychology and the philosophy of mind do not adopt a normative approach to emotion, one that would allow us to consider it exante and by definition fallacious. They offer neither any straightforward criteria nor means to exclude mistakes and to differentiate between bias and subjectivity either. It seems, instead, that contemporary research focuses on subjectivity, on the functionality of mental operations (Blanc 2006: 63), and on their mutual shaping, rather than on emotion as a source of bias and ← 142 | 143 → error.15 What seems to evolve is the kind of tasks used to evaluate the links between emotion and cognition. This could be a symptom of a changing view of what should count as rational. The account of such a composite rationality which encompasses emotions confirms ancient rhetoric’s intuition.

4.5   Swayed away, in spite of one’s will?

This short overview of the relation between emotion and cognition, either in terms of reciprocal influence or in terms of consubstantiality, according to the chosen paradigm does not entail any conclusion about the need (or uselessness) of ringing an alarm of manipulation because of the use of/appeal to emotion in discourse. Thus, we have given a negative answer to the first part of question (i) and we can deal with the second part of it: the impression of deceptiveness and emotions’ impact.

Hence, I would like to turn to the triggering of emotion, in order to focus on two particular characteristics that seem to be widely accepted and may account for the problematic nature of emotion: automaticity and the possible absence of a propositional content.

Emotions are said to be triggered by beliefs, to have a cognitive antecedent that consists in the appraisal of a situation. Nevertheless, “appraisal is not knowledge” (Lazarus 1991: 152); it is considered to be a partially automatic process that “does not imply rationality, deliberateness or consciousness” (ibid.). This postulate reminds us of a well-known quandary of philosophy of mind about emotion, viz. the doubts on the existence of its Intentional object and on the status of its propositional content (Dominicy 2007). Philosophy of mind does not deal with cognitive mechanisms, but with the form of representations emotions consist in. It raises the question whether emotions that are endowed with Intentionality have a propositional ← 143 | 144 → content, like beliefs, or a non-propositional content, like perception. This dilemma is not so important for cognitive psychology, which conflates both types of Intentionality under the generic term of “representation” (Moors 2010: 27). But the philosophical distinction is important for the purpose of my intervention, as I believe that the putative absence of a propositional content and the partial automaticity of appraisal are the reason audiences may feel and be misled by skilful speakers, as I indicated under point (ii) above.

If compared to a fastidious and time-consuming evaluation of the same situation, appraisal is imperfect,16 as it is closer, in reality, to a rapid overview and perception than to an exhaustive analysis and cannot be submitted to the standards of conscious judgement. Even if it were a conscious activity, appraisal would be a strenuous task. Appraising agents have to deal with lack of evidence, conflicting evidence, conflicting and mutually incompatible popular goals, and conflicting goal priorities, all these giving rise to appalling situations that can be responsible for uncertainty, risk and stress.17 Again, as in the case of conflicting values, options are not definitely set, answers to questions are not clear-cut, and ambiguity reigns. Even if evidence is available and priorities straightforward, the means may lack and the alliances may be fragile. Contrary to what the computer metaphor suggests, the brain allows for ambivalence, non-conclusiveness and simultaneity of apparently incompatible beliefs and desires.18

To answer question (ii) we have to say that appraisal is not knowledge. As emotions involve automatic and parallel operations and, in some cases, are said to be devoid of propositional content, they may lead to what the agents themselves, with the benefit of hindsight, may consider as bias or ← 144 | 145 → unsatisfactory communication. Their fast and uncontrollable manifestation and their link with taking action expose agents to risks.

These observations bear consequences into the way an argumentation theory should provide for means of defence from what may be considered the subversion of an extended, composite rationality which encompasses both logicality and emotion.

If emotion is consubstantial to memory, attention and reasoning, if it is necessary for the vivid perception of incarnated values and social adaptation and if beliefs and desires cross-fertilise each other, there are not that many cognitive processes left for ‘cold knowledge’, certainly not practical reasoning which is mostly involved in argumentation. If emotion is widely present not to say inescapable, then we have to seek a way to come to terms with its presence in discourse instead of banishing it under the qualification of fallacious.

5.A pedagogical project for rhetoric

5.1.   Debate emotions for their own sake

I would like, now, to turn back to rhetoric and argue in a quite provocative way for the idea that rhetoric in the broad sense of verbal situated communication to an audience or verbal interaction among interlocutors is not (only) the problem, the source of manipulative emotion, as it is sometimes portrayed to be, by rhetoric’s opponents, but (also) the remedy for it. I am echoing here C. Plantin’s (2009) heart-cry to “let people speak”.

Emotions can be valid reasons to act, if associated with social adaptation, survival and values. Furthermore, emotions can themselves be the centre of a debate, as their appropriateness is defined according to a culturally defined Topic – i.e., a set of rules and associations (Plantin 1998).19 Appraisal ← 145 | 146 → theorists develop such a topic about specific emotions, around ‘if-then’ sentences (Lazarus 1991). Emotions are debatable: speakers give reasons why they feel them, insert them in narrative scripts and challenge their legitimacy/suitability. But the fact that emotions are debatable does not imply that they are effectively debated.

We have seen that humans do have the cognitive capacity of empathy and self-knowledge, i.e., they can have representations of emotions – their own emotions and those of other people – but this does not mean that they systematically have such representations. To debate about emotions, one first has to be aware of them, and second to be able to put them in words. These steps are intertwined but not conceptually identical: interestingly enough, although the (im)possibility of thinking without language has been a philosophical conundrum, we can comfortably accept the idea of vague sensations and we are, culturally, less demanding in terms of precise expressions of subjective feelings. We saw that a great deal of emotion elusiveness is due to the automatic and possibly non-propositional character of the appraisal. Agents may perceive physiological reactions while a part of the appraisal representations are not explicit. In addition to that, an emotion can be felt without being recognised as such, specifically identified, labelled or expressed. The reasons why the process may remain incomplete vary: emotion awareness can be superfluous, time- and attention-consuming or painful, as not all emotions are socially acceptable and self-image flattering.

The good news is that humans are not at the mercy of their subjective understanding of reality. Communication promotes awareness and sets the conventional/social boundary for the acceptance of an emotion. When argumentation does not apply, fiction and poetry take over, to show what cannot be coped with otherwise. Communicators are sensitive to this conventional play with discourse ‘genres’.20 Debating, questioning, acclaiming or refuting the adequacy of an emotion broadens the sphere of its propositional representation and promotes awareness and regulation of it. It is in this sense that discourse practice can offer a remedy to the problem of ‘manipulation through emotion’ that it is supposed to create. ← 146 | 147 →

The need to develop defensive skills and attitudes in communication is a reason to train young people in rhetoric, to raise awareness over potential pitfalls and our inescapable exposure to them due to the way our brain works. Familiarisation with rhetoric’s tools and practice is the best way to prevent some unsatisfactory outcomes in communication. This familiarisation does not mean proscribing analytical categories or demonizing emotions a priori, or constructing a theory of manipulation, but training people to defend themselves in communication itself. An example of the way I perceive such a defence is illustrated in section 5.2. As a matter of fact this is the way people do communicate: all I am suggesting is that the more they practice, the better they cope with it.

As a rapid appraisal of situations, emotions may lead us to conclusions we may revise and acts we may regret with the benefit of hindsight. The risk diminishes as our knowledge of our self and of other people rises. This risk is the price to pay for a mechanism which brings us in contact with our human nature, our positive and negative values, which are a crystallisation of past collective experience.

5.2.   Does a descriptive vocation equal permissiveness?

Now, how about permissiveness and standards?

Although rhetoricians typically abstain from reasoning in terms of manipulation and fallacies, in practice they use analytical categories that may inspire a critic. I would like to associate two of these categories to the use of emotion is speech. The first one is the Aristotelian triangle of logos, ethos, pathos. The second one is the so called ‘genre’.

J.M. Adam considers manipulation to be the relative hypertrophy of ethos or pathos to the detriment of logos (2002: 46). Although he does not propose this view as a formal definition of manipulation, a term he rather uses once in its non-technical occurrence, I would like to extend his view and use it myself as a functional definition/criterion to talk about potential manipulation (in the non-technical occurrence in question (ii)) in the case of the ← 147 | 148 → hypertrophy of one of the three proofs to the detriment of the others).21. ‘Genre’ can also help us to come to terms with emotion in speech. We can expect a higher amount and variety of emotion to be present in fiction than in an essay on a new molecule. Still, there is much room for hybrid genres and the introduction of subjectivity, even in science.

These functional criteria are indicative rather than conclusive about the existence or not of ‘manipulation through emotion’ in a specific speech. Nevertheless, the absence of a systematic conclusive theory on ‘manipulation through emotion’ in rhetoric does not mean ‘permissiveness’. It is rather an indication of a view of the role of the expert and of the way a debate is closed.

The researcher in rhetoric does not usually take a normative stance towards the speech as he considers that the outcome of a debate has to be decided by the participants and not to be proclaimed by a linguist-expert. It is a matter for the participants as well to decide on the appropriateness of the use of emotion, which is a metalinguistic topic among others to discuss in a specific debate. To give an example, statements (a) to (c) of section 3 could respectively be dismissed in the following way:

(a)So tempting! Can I afford it? / Do I really care for status? / The salesman is trying to do his job, but he won’t get me.

(b)In a world of free movement of capitals, we have to allow for some movement of people too. / You are inciting hatred. / Who else wants to become a waiter nowadays?

(c)We’ll beat you! / Nationalism in our region has caused two world wars and a regional one. Stop warmongering!

The above examples are artificial ones for the purposes of this article. In natural debates a refutation may take more than an exchange of a couple of sentences. Argumentative chains in verbal interactions are open to further refutations, not least when the topic is interwoven with clashes of power and interest and/or ethical options. The fact that sometimes the need to take a decision seems to close a matter (ideally by vote) does not mean that ← 148 | 149 → theoretically it cannot be reopened if new elements appear and somebody assumes the burden of proof. In such a model of ‘self-regulation’ and through an endless possibility of refutation, the rhetorician may opt for participating in a debate, without the claim to close it, as if he were a judge or a referee.

There are also cases where a rhetorician may feel strong on the stakes in a debate and where, as a result, he will unavoidably take stance. The correctness and efficiency of this stance though should also be a matter for the participants in the debate to tell, as they do not automatically stem from a better knowledge of the argumentative tools a linguist may have.

Although the educational project rhetoric could have should help the participants in a debate to be more satisfied with the outcome of this debate, it has in itself its own limits. Ethical dilemmas, conflict and subjectivity (expressed through emotion) are inherent in numerous debates. These debates may be opened, closed or reopened, but unlike problems, they cannot be ‘solved’.


Some theories of argumentation consider appeals to emotion to impair critical reasoning. On the opposite side, rhetoric studies pathos as a natural and legitimate means of proof. I tried to explore if the same divide can be found in cognitive science. Although in the past emotions and logicality have been studied as two separate variables, contemporary theories in cognitive psychology perceive emotion and cognition as mutually shaping each other, indicating that ‘cold knowledge’ may be an illusion. Furthermore they consider emotion to be vital for survival, social adaptation, perception of values and coping with tasks requiring subjective evaluation. This extended view of rationality confirms ancient rhetoric’s intuitions.

Nevertheless, as emotions involve automatic and parallel operations, a rapid appraisal of the situation and, in some cases, are said to be devoid of propositional content, they may lead to what the agents themselves with the benefit of hindsight may consider to be bias or unsatisfactory outcome of the communication. These observations bear consequences into the way ← 149 | 150 → argumentation theory should perceive rationality and provide for means of defence from what may be considered the subversion of it. In an extended view of rationality, which includes emotions, these are not necessarily contrasted to logicality, but to other emotions that are more suitable to the situation according to the participants in a debate.

Since emotions are embedded in our rationality (Damasio 1995, Blanc 2006), considering them fallacious is a vain struggle on the basis of a misperception of the mind. I propose to deal with emotions through ‘self-regulation’ in discourse practice as an alternative option to the ex-ante definition of acceptance criteria and conditions in a system. Emotions can be shown, told, debated and shared ex post through numerous discursive ‘genres’ and modes. Albeit debatable, emotions are not always effectively debated. Communication broadens the sphere of propositional representations of emotions by offering a natural remedy against ‘manipulation through appeal to emotion’.

Rhetoric does not adopt a normative stance towards manipulation in accordance with its particular view of the linguist’s role and of the way a debate is closed. Nevertheless it can contribute to diminish the risks of manipulation through a pedagogical project of familiarisation of audiences with rhetorical tools. The better we know ourselves, our positive and negative values and the tools speakers use to convince, the less we are exposed to the risks of ‘manipulation’.

Furthermore, the claim against ‘manipulation through appeal to emotion’ is a legitimate metalinguistic argument in a debate, although its acceptance depends on the participants in the same debate. In the self-regulatory model described above, the participants and not an external expert/linguist get to decide whether they deem some emotions to be relevant and appropriate in a specific debate and whether they want to accept such a metalinguistic argument. ← 150 | 151 →


Adam, J.-M. (2002): “De la grammaticalisation de la rhétorique à la rhétorisation de la linguistique: aide-mémoire”, in Koren, R. & Amossy, R. (eds), Après Perelman: quelles politiques pour les nouvelles rhétoriques? L’Harmattan, Paris, 23-55.

Adam, J.-M. & Heidmann, U. (eds) (2005): Sciences du texte et analyses du discours. Enjeux d'une interdisciplinarité. Slatkine Érudition, Genève.

Anscombre, J.-C. & Ducrot, O. (1983): L’argumentation dans la langue. Mardaga, Bruxelles.

Aristote (1991): Rhétorique. Le Livre de poche, Paris.

Blanc, N. (2006): Émotion et cognition. Éditions In Press, Paris.

Blanchette, I. & Richards, A. (2010): “The influence of affect on higher level cognition”, in De Houwer, J. & Hermans, D. (eds), Emotion and Cognition: Reviews of current research and theories. Psychology Press, Hove, UK, 561-595.

Berne, E. (1996) [1964]: Games people play. Ballantine Books, New York.

Bouvier, A. (1995): “Les paralogismes d’un point de vue sociologique”, Hermès 16, 45-55.

Clément, F. (2006): Les mécanismes de la crédulité. Droz, Genève / Paris.

Damasio, A.R. (1995): L’erreur de Descartes. La raison des émotions. Trad. par. M. Blanc. Odile Jacob, Paris.

Danblon, E. (2002): Rhétorique et rationalité, Essai sur l’émergence de la critique et de la persuasion. Éditions de l’Université de Bruxelles, Bruxelles.

–   (2005): “Discours magique, discours rhétorique”, in Adam, J.-M. & Heidmann, U. (eds), Sciences du texte et analyses du discours. Enjeux d’une interdisciplinarité. Slatkine Érudition, Genève, 145-160.

De Houwer, J. & Hermans, D. (eds) (2010): Emotion and Cognition: Reviews of current research and theories. Psychology Press, Hove, UK.

Derbaix, C. & Grégory, P. (2004): Persuasion. La théorie de l’irrationalité restreinte. Economica, Paris.

Dominicy, M. (2001): “L’épidictique et la théorie de la décision”, in Dominicy, M. & Frédéric, F. (eds), La mise en scène des valeurs. La rhétorique de l’éloge et du blâme. Delachaux et Niestlé, Lausanne / Paris, 49-78.

–   (2005): “Langage, interprétation, théorie. Fondements d’une épistémologie moniste et faillibiliste”, in Adam, J.-M. & Heidmann, U. (eds), Sciences du texte et analyses du discours. Enjeux d’une interdisciplinarité. Slatkine Érudition, Genève.

–   (2007): “Sémantique et philosophie de l’esprit: les rapports de perception visuelle”, in Neveu, F. & Pétillon, S. (eds), Sciences du langage et sciences de l’homme. Actes du colloque 2005 de l’Association des Sciences du langage. Lambert-Lucas, Limoges, 65-82.

Dominicy, M. & Frédéric, F. (eds) (2001): La mise en scène des valeurs. La rhétorique de l’éloge et du blâme. Delachaux et Niestlé, Lausanne / Paris.

Eemeren, F.H. van & Grootendorst, R. (1996): La nouvelle dialectique. Kimé, Paris.

Eemeren, F.H. van & Houtlosser, P. (2006): “Strategic Manoeuvering: a synthetic recapitulation”, Argumentation 20, 381-392.

Elster, J. (1978): Ulysses and the Sirens. Cambridge University Press, Cambridge. ← 151 | 152 →

–   (1983): Sour grapes: Studies in the subversion of rationality. Cambridge University Press, Cambridge.

–   (2003): Proverbes, maximes, émotions. PUF, Paris.

–   (2010): Le désintéressement. Traité critique du monde économique. Editions du Seuil, Paris.

Harris, M. (1976): “History and significance of the EMIC/ETIC distinction”, Annual Review of Anthropology 5, 329-350.

Koren, R. & Amossy, R. (eds) (2002): Après Perelman: quelles politiques pour les nouvelles rhétoriques? L’Harmattan, Paris.

Lazarus, R.S. (1991): Emotion and Adaptation. Oxford University Press, New York.

Livet, P. (2002): Émotions et rationalité morale. PUF, Paris.

Micheli, R. (2010): L’émotion argumentée. Éditions du Cerf, Paris.

Moors, A. (2010): “Theories on emotion causation: a review”, in De Houwer, J. & Hermans, D. (eds), Emotion and Cognition: Reviews of current research and theories. Psychology Press, Hove, UK.

Pareto, V. (1968) [1916]: Traité de sociologie générale. Droz, Genève.

Plantin, C. (1995): “L’argument du paralogisme”, Hermès 15, 245-262.

–   (1998): “Les raisons des émotions”, in Bondi, M. (ed.), Forms of Argumentative Discourse. CLUEB, Bologne, 3-50.

–   (2002): “Analyse et critique du discours argumentatif”, in Koren & Amossy, Après Perelman: quelles politiques pour les nouvelles rhétoriques? L’Harmattan, Paris, 229-263.

–   (2009): “Laisser dire: évaluer c’est encore argumenter. De la surabondance de règles sur la parole”, in Séminaire doctoral. Modèles de l’argumentation, ULB, 9 décembre 2009.

Plantin, C., Doury, M. & Traverso, V. (eds.) (2000): Les émotions dans les interactions. Presses Universitaires de Lyon, Lyon.

Searle, J.R. (1995): The Construction of Social Reality. Penguin Books, Londres.

Tappolet, C. (2000): Émotions et valeurs. PUF, Paris.

Toulmin, S.E. (2003) [1958]: The Uses of Argument. Cambridge University Press, Cambridge.

Walton, D. (1992): The place of emotion in argument. The Pennsylvania State University Press, University Park, Pennsylvania.

Woods, J. (2004): Argument, critical thinking, logic and the fallacies. Prentice-Hall, Toronto. ← 152 | 153 →


1For a substantiated account of the historical positioning of argumentation theories on emotion, see Plantin (1998, 2000), Micheli (2010).

2Perelman talks about values rather than about emotions.

3These variations are: critical discussion, information seeking dialogue, negotiation dialogue, inquiry dialogue, scientific inquiry, quarrel (ibid. p.20-21).

4‘Same sex marriage’, ‘taxing financial transactions’, ‘deregulation of labour market’ are topics on which the conflict is not reasonably expected to be resolved through argumentation. These topics are intertwined with values and interests.

5The typical example of manipulator in literature is the one of politicians that led their people in wars through demagogy. I consider this example to be politically argumentative as it absolves the citizens of the respective countries from their responsibility.

6Emotion can be explicit or implicit. Rhetoric discusses the emotion of the audience/eceiver under the category of pathos and the emotion of the speaker/source of the message, under the category of ethos. I don't make any differentiation between explicit or implicit emotion, present in the text, attributed or provoked, as these distinctions do not matter here.

7The first question reminds us of a famous saying by Pareto (1968/1916) according to which “logic explores why a syllogism is erroneous while sociology is interested in the reasons why it is frequently accepted”. A. Bouvier (1995) points out that this remark boils down to a distinction between two different research programs, as far as fallacies are concerned: the exploration of their cognitive and their affective causes, respectively.

8I use this rudimentary example, a piece of self-criticism produced by the participant in argumentation him/herself, for, as I already said, I consider the claim that somebody else is manipulated to be potentially argumentative in the sense of Anscombre & Ducrot (1983).

9Traditionally rhetoricians are concerned with the discursive manifestation of emotions and, in the name of a ‘clean job’, leave to psychology those matters that relate to speakers’ and audience’s representations and feelings. Nevertheless, coming to terms with manipulation implies a return to interaction and the perlocutionary effects of speech on the audience. These effects are not necessarily intentional, as the term ‘manipulation’ seems to imply: they may be accidental, unpredictable, and may vary according to the audience.

10Historically, emotion, when not completely discarded as a lower somatic expression in the dualist paradigm, has been studied separately from cognition for a number of reasons (Blanc 2006). It has been considered as a reminiscence of our animal past, a signal of alert in front of change that may paradoxically undermine performance and render our behaviour dysfunctional if it persists in high amounts. Cognitive science, in spite of its critic of this behavioural model, has also failed to come to terms with emotions, as it developed on the basis of a computational model of mind and, in parallel, considered individual functions (i.e., perception, memory and understanding) as lower and higher levels of a unique and linear processing. In addition to that, the difficulty to define emotions has complicated further a long-standing dichotomy between emotion and cognition.

11This implies reconsidering both the effectiveness of experimental methods such as self-reporting and the rise of expectations from investigating techniques in neurobiology.

12For instance, self-reporting is still used as an experimental method, although its efficiency to provide access to automatic processes has been questioned. In fact, it remains an important tool to access perceptions and the agents’ own representations of emotions.

13Anxiety congruent judgements probably appear at a higher extent when risk concerns the subject himself (Muris & van der Heiden 2006).

14The association of emotion and cognition along the differentiation between central and peripheral information processing echoes the views of Chaiken (1980). Other models underline the role of motivation, involvement and relevance to enhance the likelihood of elaboration of the message (Petty and Caccioppo 1981a; 1986 and subsequent publications). For a review of cognitive models of persuasion in advertising, see Derbaix & Grégory (2004).

15Talking about bias presupposes a possibility to choose between a biased and a non-biased version. To be able to talk about error (and about truth), we need a definition and a criterion (Dominicy 2005), even when knowledge of natural and institutional facts is concerned (Searle 1995). It does not make sense, though, to talk about error and truth in the case of epistemologically subjective experiences of objective realities (pain, joy) and in the case of preferences concerning deontic propositions (‘freedom over equality’).

16This is not to say that if agents took the time to precisely and explicitly evaluate goal congruence and their own capacity of control, the emotion could disappear.

17Danblon (2002) associates persuasion with a feeling of safety, provided in the past by magicians and priests. In her conception of multilayered rationality, modern genres compensate for old needs through the ‘as if’ mechanism. Rhetorical convention is supposed to be perceptible for agents.

18A subsidiary claim of my paper is that, at least in some cases, the search for straightforward stances and the cry against manipulation stem from a misrepresentation of the task to be performed. Not all problems have a solution, and not all tasks include problem-solving among cooperative peers. Conflict of interests and power struggles, ethical and political choices can fail to be resolved with the best argument.

19R. Micheli (2010) presents a corpus analysis of such a situation, where the legitimacy of emotions like pity and revenge was at the heart of the debate over the abolition of death penalty.

20Danblon (2006) stresses the importance of the conventional ‘as if’ mechanism, which allows for fiction to play the comforting role once held by magic.

21A counter-intuitive illustration of this idea would be a technocratic report about grants to the banking sector, set in an impersonal, emotionless mode of supposedly pure logos. Rhetoricians will recognise a variation of ad verecundiam in this evanescence of ethos in the name of impartiality. An implicit appeal to the expert opinion through an impersonal style gives an impression of objectivity.