ISSA Proceedings 2006 – Who Is Afraid Of Emotion In Argument?
The relationship between emotion and argument can be found confusing. Normally we always feel something as we keep living and also as we are arguing and discussing. Does this matter for the course and the outcome of our argumentation? Does it interfere? Whenever we accept or refuse an argument or a conclusion emotion may have a certain influence. We all know that the judgement about a sentence as being right or wrong can be accompanied by something like a feeling of rightness or wrongness. Is this kind of feeling important for acceptance, is it essential for it or is it the final grind of our intellectual understanding? Let me distinguish three possible ways in which emotion can play a role for argumentation: Emotion instead of Argument, Emotion as Argument, and Emotion in Argument.
For introduction I will present a few considerations about emotion instead of argument. We should be aware that this role is extremely important: During the past 1 ½ decades the world has seen several (more or less) democratic states waging war against several other (more or less) democratic states. Acceptance for a war waged by a democratic state has to be based on the reasoned consensus of the population majority; it has to be won by argument; and, as we all know, it was mainly won by arousal of emotion: 1990 it was pity with Kuwaiti premature babies, 1999 pity with Albanian refugees, 2001 anger about Afghanistan’s government, 2005 fear of Iraq’s weapons of mass destruction. At present the so called Atomic crisis with Iran is again a triggering of emotions against a state which counts as an “evil” state and could as such eventually be attacked in the end. As far as the manipulation of emotions via mass media is concerned the political class of today has learned its lesson from Hitler’s propaganda minister, Mr. Goebbels. This powerful and sophisticated playing on emotions is a fact which can cause sorrow and fear for the future of democracy on the globe. It is a shame and a danger that it works even in most important issues. Yet it poses no major theoretical problem. The theoretical problem arises when emotion interferes in a clearly argumentative process.
In the following paper I will present some considerations about that problem. I will be concerned with the second and third of the above mentioned ways: In section 1. emotion as argument shall be tackled and in section 2. emotion in argument (or accompanying argument). Of course every treatment of the problem depends on what “argumentation” and what “emotion” means. Therefore one has to do (or to pick up) some conceptual work. As to “argumentation” I will very shortly sketch the approach which I have been developing with my research group in Hamburg for 15 years. Concerning “emotion” I will mainly propose a difference between “raw emotion” and “elaborated emotion”. The first section will go into this difference, its motivation and its consequences. The result will be that raw emotion can never function as a real argument whereas elaborated can, if it is judged to be appropriate. The second section will take up some aspects of the general relationship between emotion and argument. Here I will consider the emotional quality of every orientation and then muse about the possibility that the ”unrefutedness” of a thesis on one hand and the insight into its meaning on the other can fall apart. As far as this maybe due to emotions viz. emotional attachments, the whole system of orientation and its continuation are at stake. This is where the ”principle of transsubjectivity” enters the stage.
1. Emotion as argument
The central question of this section is whether an emotion which is appealed to or which appears in an argumentative dialogue can function as a genuine argument.
(Intellectual) common sense usually excludes this possibility because emotion and argument are counterparts. Argumentation is verbal, public and rational, whereas emotion is sensual, private and irrational. There is a dualism between the two which is a component of the dualism between mind and body. Their messages are heterogeneous or even incommensurable. Despite this they happen to conflict with each other from time to time and then the dualism becomes an antagonism and calls for a hierarchy. Traditionally, of course, argument is superior to emotion, but this has lately been questioned in the humanities and for some authors the answer seems at least to be open (Goleman (1995)). Whether or not an emotion can function as a real argument has, however, not been decided so far. In order to approach this specific question let us have a closer look and take up some examples.
“Ad Misericordiam” is a figure which contains an appeal to the emotion of pity. Let us suppose, that the appeal is successful and that pity takes place. We then could construct a very rough Premise-Premise-Conclusion argument scheme of the following shape:
P2: Pity → help!
Here the appearance of pity seems to be a real premise. The second premise would be a conditional normative proposition connecting pity with help. It could be formulated like this: ”If somebody feels pity he/she should offer help.” The respective argumentation would express that if somehow pity arises in an addressee’s emotional state, this could count as a true premise for the normative conclusion, that help should be offered.
Looking for material in the real world we can get an easy example for this kind of “argument” by considering a commercial of a charity organisation, presenting information like pictures of starving poor children. It may be important to mention that the normative content would somehow be inherent in the rise of the pitiful emotion. To feel the emotion of pity is here not the result of a decision but it happens to me, it is a befalling (German: ”Widerfahrnis”). Therefore: whenever the commercial can trigger my feeling of pity the ”argument” would run by itself.
As we all know the ”Argumentum Ad Misericordiam” counts as a fallacy and this is so because of the said dualistic principle that arguments should be rational and emotions are not. Meanwhile a lot of people have already pleaded that this view should be revised. Douglas Walton has, based on the analysis of many examples, proposed that certain conditions have to be distinguished and watched; conditions that concern as well the appeal, as the emotion itself, as the mundane context. (Walton (1992), 109-142) He argues that if someone takes care, the appeal can be a perfectly rational move. In one of his examples he takes up a dialogue between a mother and a daughter about visiting old aunt Tillie. This aunt has done a lot for the girl and now she expects to be visited from time to time. Mother’s demand to do so culminates in the appeal “I think you should have certain feelings”. (Walton op. cit. 109) Walton’s analysis of the case concludes that “the appeal to pity would appear to be just the right means for the mother to use in trying to convince her daughter to act in a certain way” (Walton op. cit. 110). But why is this so? In Walton’s view it seems to be due to the specific goals of this dialogue. It is classified as an ”action-oriented persuasion dialogue” – and this seems to be a kind of communicative effort in which such an emotional pressure can make sense. But for the sake of theory I would like to ask: When is it good, when bad; and why is it good in the present case?
Certainly the answer to this question depends on what one expects from argumentative dialogue at all and what one denominates as an ”argument”. In Walton’s approach “appeals to emotion are typically weak arguments, but they can give an argument that added little push needed to make it swing to one side of the disputed issue” (op. cit. 115/6). Yet the appeal to the emotion of pity can go wrong if it pushes “pity beyond its reasonable weight as a claim to consideration (op. cit.115). Walton concludes his treatment of Ad Misericordiam with the proposal to come away from the simple two-valued view of fallacy vs. non-fallacy and to replace it by a scale of 5 degrees, of which only the deepest one is definitely fallacious. It stands for cases in which the appeal to emotion “is used to prevent critical questions” (ibid.) about circumstances and consequences.
I am afraid I do not completely agree with Walton’s view. I personally have not been socialized in a tradition where a list of fallacies was established and an author could win some reputation by showing that a certain fallacy is not under all and every condition fallacious. Without this tradition in the back, however, the case of the Ad Misericordiam would simply appear like this: There is an emotion aroused which shall function as an argument and I am in the opponent’s role to find out whether it really can work as one. My crucial question would be: Does it contribute to show the validity or invalidity of the thesis? If it does, it works as an argument and if not, it has to be taken as a communicative side phenomenon. Well then: Is it possible that an emotion contributes to the construction or deconstruction of a thesis in an argumentation?
As I have stated above, the answer to this question depends on the meaning of the involved terms ‘argumentation’ and ‘emotion’. About my view on the key term ‘argumentation’ I will give some information in section 2. For the moment the following rough characterisation must do: An argumentation is a step-by-step effort to build up or tear down a thesis; and an argument is a part of an argumentation. So it is something which has to function as a step in the said kind of effort. But what about the second term: ‘emotion’? There is a lot of psychological theory available concerning emotion. (Zimbardo 1979, Chap. 10). The need of conceptual clarity, however, is not very well served. Usually the gap against reason and a lot of examples for different emotions is all we can find.
In philosophy concept clarification is a recognized task and we can find more or less helpful definitions. I take up some “circumscription” of the German philosopher Heiner Hastedt (Hastedt 2005, p.21) which says that emotion is an inner state of experiencing a kind of involvement with something or somebody. If this kind of involvement is lasting it can become stable and is then a bond or an emotional attachment. (Walton’s favorite term for a stable attachment is ‘commitment’ which seems to be too strongly accentuated with obligation and duty.). At any rate in this conceptual framework an emotion would be a fact. If a fact shall serve as an argument if it needs to be theorized: It has to be embedded in theories which provide for a verbal expression allowing to distinguish the fact from others and which provide for conditional sentences, putting the verbalized fact into a relationship with other circumstances. An emotion however is an “inner fact”. This is something special and it asks for some major theoretical effort. I will try to explain: As long as I simply feel something I am imminently involved in a relationship with something or somebody. Per se it is neither clear what the feeling precisely is – in contrast to other feelings nor whether and how it relates to the special circumstances and/or the special person who are part of the situation. This blurred state of affairs clears up with theoretical effort. Theorizing feeling is a process in which a ”gap” between the self and the feeling is opened up. The feeling develops into a conscious and verbalized and understood emotion. As such it is no longer a merely subjective inner state but a quasi-objective (viz. intersubjective) affair.
Let me refer to another example: A small boy is invited to a birthday party and is full of anticipated joy for several days. Then the party is cancelled. The boy becomes very mute and sad and his mother might tell him something like: “Oh my poor darling, you were so full of anticipation and joy. And now you are deceived.” This is a paradigmatic situation for the child to learn the emotion of deception. The feeling itself surely emerges spontaneously but it takes up a clear and recognisable form because it is given a name and it is connected to certain well identified circumstances (the sudden lost of anticipated joy). This is how the boy is able not to be completely overwhelmed by the feeling viz. to win or save his self-knowledge in the waves of feelings. In the best case he would be able to express something like: ”I am Harry and I am now affected by a deep deception.” Deception would now be something like a recognisable affair. It has become a special emotion, a species of sadness occurring in the context of joyful anticipation which had been staying for a while and was then broken off. It expresses itself in a specific behaviour of muteness, depression and lack of energy. In this best case the gap between self and emotion would be perfect. Yet as we all know this gap is usually much weaker and the emotion is then not so clear as in the ideal case. Despite this we can learn some general lesson from the ideal case:
An emotion can become clear and quasi-objective if its appearance can be theorized, which means that it performs three steps:
(a) Verbalisation (giving the emotion a name to distinguish it from others)
(b) Conditionalisation (establishing connections between the emotion and certain circumstances on one hand and between the emotion and certain expressive behaviour on the other)
(c) Distanciation (establishing self and self-knowledge opposite to the emotion)
The more and the deeper an emotion gets theorized the more we dispose of experiences and knowledge which can be articulated in the following kind of inferences:
C ⇒ E
E ⇒ B
In this diagram E stands for ‘emotion’, B for ‘behaviour’, and C for ‘circumstances’ or ‘context’. The double arrow stands for a theoretical connection, which allows to expect E under the condition of C, and B under the condition of E.
A feeling which is not in this sense theorized, is a “raw emotion”. This means that we simply feel something, not being aware of any discriminating and differentiating quality in the feeling. Raw emotions (like hunger, fear, desire, anger) are like black and white pictures. They are few, they are simple and they circle around lust and pain. In the development of emotions, however, they become “elaborated”, i.e. we become more and more conscious of them, we construct theories around them, experience them in the light of those theories and as a result they become refined, differentiated and numerous.
Let me go back to the examples. Walton’s analysis of Ad Misericordiam has brought out some conditions for its functioning as an argument, mainly concerning relevance (of emotion) and information (about circumstances). Obviously this is compatible with my statement that not emotion as a sheer fact but only emotion as embedded in theory can reasonably appear in argumentation. But Walton refrains from conceptual work about emotion and therefore seems to miss a clear view of the development of emotions and the role of theory therein.
Embedding emotions into theories will primarily consist in realizing, investigating and establishing the two types of inferences shown in diagram (2). This can also be illustrated with the help of Walton’s example: Mother thinks that in the given context (Aunt Tillie having benefited the daughter) an emotion of gratitude and pity towards the old lady is appropriate. She expresses this by: “You ought to have such feelings” (Walton (1992), 109). Further on mother presupposes that, given the emotion, a visit would be the appropriate behaviour.
I think that inferences of this kind are indeed established. We more or less know them and can refer to them explicitly. This means that we have established the above said gap between ourselves and our feelings. The gap is even stabilized with theoretical knowledge about emotions and all this is an essential part of our civilization. Whenever an emotion can function as an argument, this is because the respective inferences are established and can be correctly applied. Those inferences viz. the theoretical embedding of the emotion supply us with the theoretical basis to judge whether the emotion and/or the respective behaviour is appropriate. All this works, when emotions are not simply left in their raw form, but have been elaborated via verbalization, conditionalisation and distanciation.
Elaborated emotions are no more heterogeneous or incommensurable with rationality. The dualism between emotion and argument seems to be based on the view that normally emotions are “raw”. With respect to the possibility of elaborating emotions Walton is right in stating: “Emotion should not be (categorically) opposed to reason – even though appeals to emotion can go wrong or be exploited in some cases.” (op cit. 257).
Usually we can claim that a civilized person may have certain emotions. This is part of a cultivated life: Communication would hardly be possible without “emotional reciprocity”, which means that if certain conditions are given, we can expect certain emotions in our partner and so can our partner in us. But in order to serve as a serious argument an emotion has to be theoretically elaborated – because only then a theoretical basis is available to judge about its validity. Even those authors who properly emphasize the importance of emotion for argumentation (Gilbert 2004) seem to be not yet completely aware of this.
These considerations can be driven one step further. Let us look again at the inferences in diagram (2). Obviously the two can be put together and then the emotion can be eliminated.
C ⇒ E ⇒ B
C ⇒ B
This is not only a formal possibility but it illustrates again our civilized practices with emotions. Let us suppose the daughter in Walton’s example, when demanded “to have certain feelings” would sincerely reply: “Well mother – whatever kind of feelings I ought to have, it is a fact that I do not have them.” Then mother of course should not insist (you ought to have them because, if you haven’t, you are a bad girl …) but she should say something like: “This fact does not really matter, because even if it is so, your aunt can expect the said feelings on your side, and this expectation is appropriate. Therefore you should behave as if you had them and pay Aunt Tillie a visit.”
To sum up what I have argued for in this section: Raw feelings cannot be arguments, but elaborated feelings can. If an elaborated emotion claims to take up the function of an argument, we have to look out whether there is a theoretical basis available about the ”appropriateness” of the respective emotion. And if it is, then a judgement about the (Non-) Appropriateness is justifiable. This has an interesting consequence: Whenever an elaborated emotion is appropriate and can work as an argument it needs not even to be felt. Not the emotion itself but its appropriateness makes the argument.
2. Emotion in argument
The first section was somehow preliminary. Its message was that emotion, when it shall function as a premise in an argumentation, plays a somewhat neglectable role. In the second section I will muse about the more important questions concerning the role of emotion in argument. I hope that I can at least clarify the territory a little bit.
Let me look back to the question whether argumentation is a purely intellectual process. As I have stated above the answer depends on what “argumentation” is. If it is e.g. a sequence of propositions whose truth-values are determined and provide for the truth value of the conclusion, then this would certainly be something in the purely intellectual sphere. We have, however, meanwhile reached a kind of consensus, that this is not the only and maybe not the best view on argumentation. A lot of scholars have looked out for better approaches and they have been more or less successful. In Hamburg we have been developing a validity-oriented approach during the past 20 years and I shall here very shortly sketch some of its elements.
I take argumentation as a dialogical procedure to construct a conclusion as a “New Orientation” out of proven orientations. New Orientation is not mere opinion and it is not yet knowledge. It is something in between the two. It is a piece of theory which shall cover some lack of orientation. To fulfil this function it may not only be interesting and original, but it must be reliable. The reliability of a piece of theory, taken as orientation, proves in practice. It can, however, be risky to simply introduce a thesis, viz. a new piece of theory, into a practice. Let me mention a very great example: Can a European sailor in the late medieval age afford to put to sea to India behind the Atlantic ocean if there is no India behind the Atlantic ocean? A caravel at that age was able to stay on the sea for only four weeks. Columbus’ thesis about India transoceana therefore contained the utmost risks for all those who participated in its fulfilment. What could be done to reduce those risks? Yes the best knowledge has to be collected and as it is not sufficient the thesis has to be constructed as an extrapolation of that knowledge. And this is the genuine function of argumentative dialogue: to check and test as far as possible in advance whether or not the thesis is reliable. Argumentation in this view is the intellectual testing procedure for a thesis to find out about its reliability. The procedure consists of constructive steps (reasons, grounds, evidences) and refutative steps (objections, counter-evidences). Sometimes the procedure leads to a conclusion; that means: the thesis can be introduced into practice as a New orientation. This kind of result is reached when the constructive steps have been productive and the refutative steps have all been defeated or integrated. I call this quality “unrefutedness” (or “freedom of objections”). It is a purely formal quality, even if it concerns more than the logical consistency of the thesis and argumentation. Yet to give a conclusion the splendour of reliability as a New orientation we need more. We expect that the thesis offers us an insight into the material matters in which the initial lack of orientation was realized and motivated the rise of argumentation. If an argumentation is successful the opponent not only accepts the conclusion but he/she would experience something like: “Now I understand the point. I can see it with the eyes of my mind.”
In the perspective of the Hamburg Approach the quality of a satisfying conclusion is therefore twofold: it contains both, “unrefutedness” and “insight”. Unrefutedness is, as I said a formal quality, it is a purely intellectual affair. What about insight?
I believe that it is here, where the emotions enter the field of argument and argumentation theory in a far more important way than they could do through the treatment of the emotional fallacies. Insight reflects the whole orientation-system of the participants. An addressee who engages in argumentative dialogue introduces a system of all kinds of beliefs and wishes and fears etc. and in this orientation-system a large emotional layer. Its emotions are partly spontaneous and instable but the more important ones are emotional bonds and deep attachments which have been acquired in the course of life over the years and decades. These attachments are now able to strongly structure our insight and even to block an insight if it would feel bad or unsafe or dull or horrible etc.
Whenever this happens the argumentative conclusion’s two sides can fall apart. The thesis may come out as unrefuted but the opponent may not gain any insight and would therefore not take it up as a New orientation. What can be done about this kind of dilemma? I see two possibilities.
(A) We could reinforce the dualistic view and separate the twofold quality of argumentative validity. The side of the unrefutedness would again become the expression of objective reason and insight would be something subjective which does not count very much because it is emotionally contaminated.
(B) We could adopt a more comprehensive concept of reason (where the emotional layer is also part of the consciousness) and look out for ways and methods of handling emotional attachments. These methods would show how emotional attachment can become conscious and understood and thus open to revisions and changes. Here we would again meet questions about the appropriateness of certain emotions in relationship to the context of the thesis.
Does this lead to transforming argumentative dialogue into psychotherapy? I think it is indeed worthwhile to consider that in some cases of blocked argumentation psychotherapy could be an option. We all may have experiences with dialogues in which emotional attachments were so powerful that all effort to gain some distance, viz. establish the above mentioned “gap” was in vain. Here even the skilled arguer can be seduced to present an Ad Hominem argument of the following shape: “I am afraid that because of your specific emotional attachments you are not able to understand the merits of my argumentation.” The trap here is not to break off the dialogue but to declare victory. Certainly only the opponent him/herself can state to be emotionally blocked.
Outside those delicate cases it should however be possible to proceed in the normal way: work out the respective attachments and realize that they are connected with “frames” (Wohlrapp 1998). This means that they result from (unknown) selections which now function as a restricting filter and which have to be made explicit in order to overcome the inherent limits of insight. After the frame borders have been cleared up the emotional attachments may be found secondary viz. the chances to openly discuss about their appropriateness may be enlarged.
The general idea at this point is a philosophical principle: Human beings are gifted to “overcome their subjectivities”. This means that the process of developing feelings by opening up a gap between self and emotion can be generalized and performed as a process of developing the whole personality by deepening a gap between self and subjectivity. Obviously this principle implies a strong belief in human reason: Humans need not stay in the uncivilized state of early mankind when emotions not only determine actions (like attack and flight) but also restrict and limit insight and understanding. Maybe my reader is sceptical about this idea and takes it as a particular optimism on my side. This would, however, be an over-hasty conclusion. I think that human self-understanding simply relies on this kind of trust in freedom and reason. Without it, all serious arguing would be pragmatically blind or even inconsistent. Taking up this belief in human reason as a New orientation (in fact it is a very old orientation but it has lately been buried under all kinds of cultural and postmodern thinking) is to follow the “principle of transsubjectivity” (Lorenzen (1984), 82)
(1) Emotion cannot be kept outside argumentation. Even when working with purely structural approaches it cannot be avoided that, as soon as emotions are triggered, argumentation can catch a bias.
(2) Crucial cases are those in which a conclusion is not accepted even if there are no open objections against it (any more). Such a conclusion then may feel right or feel wrong but it is not; or it maybe right or wrong but it feels not so.
(3) In these cases the emotional bonds and attachments should be investigated whether they are appropriate or not. If they are, then argumentation has to be extended until it includes the respective attachments. If they are not, then the conclusion has to be extended with demands of transsubjective distanciation from problematic attachments or even from acquired parts of the whole orientation system.
Hastedt, H. (2005): Gefühle. Philosophische Bemerkungen. Stuttgart: Reclam
Lorenzen, P. (1984): Normative Logic and Ethics. Mannheim: B.I.
Gilbert, M. (2004): Emotion, Argumentation and Informal Logic; In: Informal Logic, Vol 24, No. 3, 245-264
Goleman, D. (1995): Emotional Intelligence, New York: Bantam Books
Walton, D. (1992). The Place of Emotion in Argument. Pennsylvania: The Pennsylvania State University Press
Wohlrapp, H. (1998): A new light on non-deductive argumentation schemes; In: Argumentation 12, 341-350
Zimbardo, P.G. (1983): Psychologie. Berlin/New York: Springer