Arnold's theory of emotion in historical perspective

Abstract

Magda B. Arnold's theory of emotion is examined from three historical viewpoints. First, I look backward from Arnold to precursors of her theory of emotion in 19th century introspectionist psychology and in classical evolutionary psychology. I try to show that Arnold can be regarded as belonging intellectually to the cognitive tradition of emotion theorising that originated in Brentano and his students, and that she was also significantly influenced by McDougall's evolutionary view of emotion. Second, I look forward from Arnold to the influence she had on Richard S. Lazarus, the theorist who deserves the most credit for popularising the appraisal approach to emotion. Here, I try to document that Lazarus’ theory of the stress emotions preserved most assumptions of Arnold's theory. Finally, I look back at Arnold from today's perspective and address points of success of the appraisal paradigm in emotion psychology, as well as some remaining problems.

Parts of this article are based on Chapters 1 and 2 of Reisenzein, Meyer, and Schützwohl (2003).

Parts of this article are based on Chapters 1 and 2 of Reisenzein, Meyer, and Schützwohl (2003).

Acknowledgments

I am grateful to James Averill, Nico Frijda, Arvid Kappas, Ira Roseman, and Stephanie Shields for their helpful comments on an earlier version of the article.

Notes

Parts of this article are based on Chapters 1 and 2 of Reisenzein, Meyer, and Schützwohl (2003).

¹Alexius Meinong (1853–1920) was professor of philosophy in Graz, Austria from 1889 to 1920. He became famous in philosophy for his theory of objects, an early form of logical semantics (cf. Simons, 1986; Smith, 1994). Meinong's achievements in psychology include the founding of the first experimental psychology laboratory of Austria-Hungary (1895) and the establishment of the “Graz” school of Gestalt psychology. His students include the Gestalt psychologist Christian von Ehrenfels and Fritz Heider, the founder of attribution theory. Heider (1983) refers to Meinong as the teacher who influenced him most. An excellent biography of Meinong was written by Dölling (1999).

²All translations from German are mine. Meinong's term for “cognitive” in this quotation is “intellectual” (“intellektuelle Leistung”).

³Within this movement, “phenomenological analysis” originally referred to a special philosophical method devised by Husserl, the intuition of essences (eidetic seeing, Wesensschau). Husserl sharply distinguished this method from introspection, and it is in fact closer to conceptual analysis (Künne, 1983). However, Husserl's special views of the nature and reach of phenomenological analysis are controversial and were not shared by all subsequent phenomenologists. To what degree Arnold endorsed Husserl's conception of phenomenological analysis remains unclear to me, as she did not comment on the issue. In any case, the present description of Arnold's method—a combination of introspection and analysis of commonsense psychology (which is often referred to as a conceptual analysis of the mentalistic terms of ordinary language; cf. Heider, 1958)—seems to capture well what she actually did.

⁴Similar to Stumpf (1907), Arnold draws a sharp distinction between emotions on the one hand and what she calls “feelings” on the other hand. Whereas emotions are “reactions to objects or situations”, feelings are “reactions to a subjective experience” (Arnold, 1960a, p. 81). Paradigmatic examples of feelings are sensory feelings, the experiences of pleasure and displeasure caused by simple sensations, such as the pleasant feeling caused by the smell of a rose or the unpleasant feeling elicited by a bitter taste. Although Arnold believed that feelings, like emotions, are mediated by a process of appraisal (evaluation), she admitted that in the case of feelings, this claim was not based on phenomenological evidence. Rather, in the case of feelings, appraisal is postulated “as a hypothetical construct … to account for the facts” (Arnold, 1960a, p. 73). The present review is restricted to Arnold's theory of proper emotions. For further discussion of Arnold's theory of sensory feelings, see Kappas (this issue).

⁵In Arnold and Gasson (1954), where Arnold's appraisal theory was first described, the term “appraisal” does not in fact yet appear. There, appraisals are still called “value judgments” (e.g., p. 305) or “evaluations” (e.g., p. 295). The introduction of the technical term “appraisal” by Arnold (1960a, 1960b) seems to have been motivated by two considerations: (1) to have available a broader term that covers the factual beliefs underlying emotions as well as the evaluations; and (2) to be able to refer to (presumably existing) forms of evaluation that do not comfortably fit the concept of an evaluative belief. These include the evaluations that, in Arnold's view, underlie sensory pleasures and displeasures (see Footnote 4), but possibly also the “intuitive” (as opposed to “reflective”) evaluations that underlie emotions (see the last part of this article).

⁶Nonetheless, it should be acknowledged that Arnold's reference to desires or motives as (partial) causes of emotion is much less explicit than Lazarus’ (e.g., 1966, 1991). One reason for this is probably that Arnold: (a) did not make a sharp distinction between desires to act (action tendencies) and desires that something be the case (colloquially, wishes; these are more typically the desires that underlie emotions), and (b) identified emotions with desires to act. As a consequence, the claim that joy about S presupposes the desire for S translates, for Arnold, into the claim that joy about S presupposes (another) emotion directed at S, which may seem threateningly close to circular. Interestingly, however, Arnold did make a parallel claim for negative emotions—she asserted that they presuppose positive emotions (to which she counts liking and wanting): “We must like something, must want or possess it before our aim can be frustrated or our possession disturbed so that we feel anger or fear” (Arnold, 1960a, p. 194). But if negative emotions presuppose desires, how can positive emotions make do without them?

⁷Therefore, the factual and the evaluative belief about S are both partial, direct causes of the emotion directed at S. This point has been occasionally overlooked in the subsequent appraisal literature, where the factual belief is sometimes depicted as being only an indirect cause of the emotion (factual belief → evaluation → emotion). To be sure, this casual model is not entirely wrong: The factual belief is typically also an indirect cause of the emotion, in that it instigates the process of evaluation (cf. Figure 2). Nonetheless, it needs to be stressed that evaluative beliefs are alone just as insufficient for an emotion as are factual beliefs alone and that the quality of the experienced emotion depends on both. For example, to experience joy about having won in the lottery, it is not sufficient that one evaluates winning positively; one must also believe that one has won. Furthermore, one must believe this firmly: If one is uncertain whether or not one has won (in Arnold's words, as long as the object is still “absent”, rather than “present” and “resting in possession”), one will not experience joy, but hope, even though the evaluative belief is unchanged (Arnold, 1960a; chapter 11).

⁸Alternatively, or in addition, Arnold may have been influenced by the evolutionary emotion theory proposed by Shand (1914), who was himself significantly influenced by McDougall and in turn influenced him. Shand's emotion theory is actually more congenial to Arnold's than is McDougall's, because it is more “cognitive”. Although Arnold's references to Shand's theory of emotion are sparse, the following comment suggests that she found herself in essential agreement with this theory: “Shand (1914) and others [assumed] that there are inherent systems in the mind that are connected with bodily systems … In this way, the importance of physical changes in emotion was preserved while the mental part of the system took care of the fact that the situation has to be interpreted by the individual before an emotion can be aroused” (Arnold, 1970, p. 170).

⁹Actually, appraisal plays a larger role in McDougall's theory than Arnold suggests (cf. Meyer, Schützwohl, & Reisenzein, 1999). This is true at least if one takes into account McDougall's later theory of the “derived emotions” (which include joy, sorrow, hope, disappointment, and despair; see supplementary chapter 3 in McDougall, 1960; and McDougall, 1928). According to McDougall (who was in this case influenced by Shand, 1914), these emotions are forms of pleasure or displeasure, or mixtures of both, that occur when one cognises an actual or possible fulfilment or frustration of an instinctive action impulse. Hence, McDougall's theory of the derived emotions is close to Meinong's theory of the judgment-based emotions. Furthermore, a close look at the “natural” eliciting conditions of McDougall's basic emotions suggests that even some of these comprise appraisals or something very much like them. For example, anger is elicited by the perceived obstruction or blocking of other instinctive actions; and the instincts of dominance and submission are elicited by the presence of other people to whom one perceives oneself to be, respectively, superior or inferior.

¹⁰However, Arnold frequently also expresses scepticism about the utility of emotional action impulses in humans, at least in today's society. The motives of “the normal person”, she writes, “are rational rather than emotional” (1960a, p. 237). “Clearly, emotion interferes and disturbs if it urges us in a direction different from that indicated by deliberate judgment” (Arnold, 1970, p. 177). Emotional impulses therefore need to be controlled and guided, to be in concert with the dictates of reason.

¹¹Interestingly, a “syndrome” definition of emotions had also been proposed by McDougall. He (McDougall, 1928) distinguished between two senses of the term “emotion”, a narrow sense (emotion = emotional experience) and a wide sense (emotion = the totality of the instinctive mental and bodily processes). According to McDougall, both definitions have their use and justification. Arnold and Gasson (1954) likewise seem to allude to a syndrome definition of emotions when they propose that “an emotion is complete when there is the whole sequence described above, including the practical estimate of the situation, the reaction of wanting or dislike, the somatic expression and organic changes, and the awareness of these changes” (p. 295). If the emotional expression or the organic changes are missing, the emotion is said to be “incomplete” (p. 295) although, because emotion is defined as a felt action tendency, “it is [still] possible in these cases to speak of emotion” (p. 296).