The Real Trouble with Recalcitrant Emotions

Abstract

Cognitivists (sometimes called ‘Judgementalists’) about the emotions minimally hold that it is a necessary condition for being in an emotional state that one make a certain judgement or have a certain belief. For example, if I am angry with Sam, then I must believe that Sam has wronged me. Perhaps I must also elicit a certainly bodily response or undergo some relevant experience, but crucial to the view is the belief or judgement. In the face of ‘recalcitrant emotions’, this once very popular view has come under heavy criticism that has led many theorists to either abandon the view or to offer more nuanced representational views of the emotions. Against what seems to now be received wisdom, I argue that cognitivists have tools at their disposal that allow them to alleviate the apparent conflicts presented by cases of recalcitrance. But I also believe that cognitivists are still in trouble. Although cognitivists have a range of underexplored resources, their use comes at a high cost. In particular, cognitivists must adopt a widespread and thoroughgoing inaccessibility to our own thoughts and judgements that should strike one as implausible. It is mental opacity rather than mental conflict that is the real problem posed by recalcitrance.

1 Introduction

Cognitivists (sometimes called ‘Judgementalists’) about the emotions minimally¹ hold that it is a necessary condition for being in an emotional state that one make a certain judgement or have a certain belief.² For example, if Sally is angry with Sam, then Sally must believe that Sam has wronged her. Perhaps she must also elicit a certain bodily response or undergo some relevant experience in order to be in an emotional state, but, crucially for cognitivists, the subject must form an appropriate belief or judgement. In the face of the Jamesian theory of the emotions, cognitivism struck many as a serious advance. The view helps makes sense both of the intentionality of the emotions and of their rational assessability, and these are no small feats.³ But the view has also come under heavy criticism that has led many theorists to either abandon the view or to offer more nuanced representational views of the emotions.⁴ This paper concerns one especially important line of attack against cognitivists that is often taken to be devastating: Cognitivists cannot make sense of the everyday fact that one can be in an emotional state, say, fearing the dog, while believing that the dog is not dangerous; cognitivists cannot make sense of the everyday fact that one can, say, be angry with Jane while believing that she has not wronged one. More generally (following D’Arms and Jacobson’s (2003) influential paper) let us say that an emotion is recalcitrant when it exists despite the agent having a belief or judgement that is in tension with it (129). The attack with which this paper is concerned (an attack that, as I will argue, hasn’t been sufficiently drawn out) has it that the existence of recalcitrant emotions shows that cognitivism predicts unacceptable mental conflicts and hence is false.

Against what seems to now be received wisdom, I argue that cognitivists have tools at their disposal that allow them to alleviate the apparent conflicts. Cognitivists have valuable resources that have gone under-appreciated and part of my goal is to shed light on those resources. But I also believe that cognitivists are still in trouble. Although cognitivists have a range of under-explored resources, their use comes at a high cost. In particular, cognitivists must adopt a widespread and thoroughgoing inaccessibility to our own thoughts and judgements that should strike one as implausible. It is mental opacity rather than mental conflict that is the real problem posed by recalcitrance.

2 The Problem of Recalcitrance

The objection to cognitivism is due to Greenspan (1981). She asks us to imagine a situation in which Fido, a known-to-be lovable and toothless dog, elicits fear in a subject—call her ‘Pat’. Pat is well aware (and indeed affirms) that the dog poses no threat. She is afraid nevertheless. When faced with Fido, Pat trembles, flees, and so on. But, as Greenspan argues, in order to explain the subject who affirms that Fido is not dangerous but reacts fearfully, cognitivists must attribute an ‘incoherent’ set of judgements (162): The judgement that Fido is dangerous and the judgement that Fido is not dangerous. In light of cases like this one, Greenspan concludes that cognitivism is flawed.⁵

The argument will be drawn out with more care in a moment, but before doing so let us look at some recent appeals to the argument, for it is worth seeing just how influential Greenspan’s case has been. For many authors, it is given as a clear and well known reason for departing from cognitivism, a reason that allegedly needs very little discussion.

Jonas believes firmly that this spider is not dangerous, yet he is terribly frightened. Mary is convinced that she has done nothing wrong, yet she is assailed with crushing guilt. Should we then, to save the theory, say that they in fact have contradictory beliefs, one which is unacknowledged or unconscious, and in so doing, attribute to them a radical form of irrationality? (Deonna and Teroni 2012: 54–5)

The objection is that judgementalism does not permit one’s emotions to conflict with one’s considered judgement; yet such conflict is a familiar psychological phenomenon. […] Since traditional judgementalism holds that the relevant belief is a necessary constituent of an emotion, the theory seems committed to denying the possibility of emotional recalcitrance. […] If judgementalism can accommodate recalcitrant emotion, it is only through the dubious attribution of peculiarly conflicted beliefs. (D’Arms and Jacobson 2003: 129)

According to judgmentalists, for Jane to be afraid that Fido may hurt her is, at least, for her to judge that Fido may hurt her. Yet she believes that Fido can hurt no one. The judgmentalist account thus forces us to ascribe to Jane two contradictory judgments. (Benbaji 2013: 579)

This is but a sampling of the widespread reliance on Greenspan’s objection. But thus far we have only the beginnings of a problem for cognitivism, for what is so bad about ascribing contradictory beliefs or judgements? It is certainly possible to believe/judge that p and believe/judge that not-p, so the view doesn’t predict something impossible or absurd. Belief in contradictions might even be quite normal depending on how beliefs are individuated. So we need a bit more before the cognitivist is in trouble.

Notice that Greenspan herself worries that if we adopt cognitivism we will attribute to a subject ‘incoherent’ beliefs and Deonna and Teroni say that we must attribute ‘a radical form of irrationality’. They are hinting at a key assumption—in order to generate a problem for cognitivism, it must be assumed that the subjects in question are rational. In fact, in Greenspan’s original example, she builds into the the case that the subject who is afraid of the harmless dog meets our usual standards of rationality. Once that assumption is on the table, the argument against cognitivism can be made precise and quite compelling:

• 1. S is rational. (premise)

• 2. S believes that Fido is not dangerous. (premise)

• 3. S is undergoing a fear response in the face of Fido. (premise)

• 4. S believes that Fido is dangerous. (by the truth of cognitivism and 3)

• 5. S believes that Fido is dangerous and believes that Fido is not dangerous. (conjunction of 2 and 4)

• 6. Any subject who believes that p and believes that not-p is not rational. (‘Contradictory-so-not-rational’ principle)

• 7. S is not rational. (by 5 and 6)

• 8. Contradiction.

Given the contradiction, some premise has got to go, but which? There are two contentious assumptions in this argument: Cognitivism and the ‘Contradictory-so-not-rational’ principle stated in 6. The authors quoted above give up cognitivism and that’s the usual reaction. Rather than denying cognitivism, cognitivists should attend more carefully to 6. There are independent reasons for thinking it is false. We’ll return to that thought it just a moment.

Before moving on it is worth considering whether a subject who fears Fido and believes that Fido isn’t dangerous really is rational. Benbaji was quoted above. An important aim of Benbaji’s paper is to make sense of the irrationality of recalcitrant emotions—something taken for granted but in need of explanation.^{6 , 7} To hold that the subject of a recalcitrant emotion is irrational is quite plausible.⁸ After all, we might very naturally say to Pat, ‘What are you afraid of? You are being irrational.’ Or to the subject who has recalcitrant guilt, ‘You are being irrational, just let it go.’ But if the subjects in question are not rational, the objection gets no footing. The “worry” would amount to this: we must attribute conflicting beliefs to irrational subjects. This is no objection at all and so the view that subjects undergoing recalcitrant emotions are irrational will be put aside. Despite some sympathy with the irrationality position, the present paper is concerned with the argument given in 1–8. It’s the argument due to Greenspan and it’s what most people in the literature seem to have in mind when they quickly dismiss cognitivism. In any event, the argument as given in 1–8 is a compelling version of the case against cognitivism from recalcitrance and it deserves careful consideration. If cognitivists can save themselves with the simple claim that recalcitrant emotions aren’t had by rational subjects, then all the better for cognitivists, but that’s not what’s at issue here.

Return now to the thought that premise 6 should be pressured. Cognitivists can benefit from conclusions drawn from Kripke’s (1988) now famous puzzle about belief. A number of philosophers have taken Kripke’s puzzle (completely independently of anything concerning the debate over cognitivism) to call into question the ‘Contradictory-so-not-rational’ principle stated in 6. If 6 is problematic, then cognitivists can reject it rather than their own view.

In Kripke’s famous puzzle about belief we are introduced to a subject who seems to believe that p and who seems to believe that not-p. Appealing to the plausibility of the Contradictory-so-not-rational principle, Kripke argues that our concept of belief is, at the very least, puzzling. We needn’t rehearse Kripke’s puzzle now. It’s sufficient to highlight a very plausible response to puzzles of that sort, namely that subjects who believe contradictory propositions may still be rational as long as they entertain those propositions using different concepts or by entertaining them under different modes of presentation.⁹ When different modes of presentation or concepts are at issue, a thinker may be unaware that he or she is thinking a pair thoughts that in fact have contradictory contents and so his or her rationality isn’t threatened. If this is correct, then the Contradictory-so-not-rational principle (i.e. 6) is too hasty. To establish irrationality, we need a subject who thinks that p and that not-p in the same way/under the same modes of presentation (save for negation). In light of the well known and well received independent considerations against 6 in the literature on belief, cognitivists can and should help themselves to the denial of 6. If 6 goes, the case against cognitivists as offered above goes as well.

3 The Repaired Puzzle

Isn’t it easy enough to repair the worry from recalcitrance and defeat the cognitivist once again? On first pass, it seems implausible to maintain that Pat is using distinct modes of presentation or is thinking of Fido in distinct ways in her belief and her fear. Benbaji (2013) is a recent representative of this common thought:

Now we may uphold contradictory judgments, provided they are kept distinct in our mind. However, [Pat’s] alleged contradictory judgments cannot be so readily consigned to separate areas of her belief box, since [Pat’s] belief that Fido is innocuous and her fear of Fido (the alleged contradictory belief) are both alive in her mind, as is the conflict between them. (579)

To defeat cognitivists, anti-cognitivists must simply trade 6 for something that takes account of the lessons learned from Kripke’s puzzle and run the objection again, carefully pointing out that our subject is not using distinct concept or modes of presentation. Accordingly, let’s exchange 6 for 6′:

6′. Any subject who believes that p and believes that not-p and does so by deploying the same (save for the negation) modes of presentation/concepts is not rational.

(‘Contradictory-with-modes-so-not-rational’ principle)

Even if 6 is false, since 6′ is plausible and the recalcitrant subject is using the same modes across both fear and explicit belief, the argument against cognitivists appears to go through anew. This is a much improved version of the problem facing cognitivists.

4 Slow Switching and Pat’s Distinct Concepts

Anti-cognitivists took themselves to have a devastating objection to cognitivism—if rationality is assumed, the view yields a contradiction. But we’ve seen that the argument requires a repair. Once repaired, anti-cognitivists can argue once again that cognitivism yields a contradiction and so must be abandoned.

The strengthened puzzle against cognitivists relies on 6′ rather than 6 along with the assumption that a subject like Pat undergoing a recalcitrant emotion uses the same modes or concepts across the relevant representational states. But this is not an assumption that cognitivists must accept. In this section I will offer some reasons cognitivists can and should give for rejecting the prima facie plausible claim that Pat is deploying the same concepts or entertaining under the same modes of presentation.

There are at least two reasons—the second more interesting than the first—that cognitivists might offer for being wary about the assumption that Pat deploys the same concepts or modes across the attitudes in question. First, as modes of presentation are usually introduced, she couldn’t be entertaining under the same modes. In footnote 9 I quoted Salmon, but it is worth repeating the quote here:

The important thing is that, by definition, [modes of presentation] are such that if a fully rational believer adopts conflicting attitudes (such as belief and disbelief, or belief and suspension of judgement) toward propositions p and q, then the believer must take p and q in different ways, by means of different guises, in harboring the conflicting attitudes toward them—even if p and q are in fact the same proposition (1989: 246).¹⁰

If Pat is indeed rational, then by definition she must be deploying distinct modes or concepts. I don’t want to rely on a stipulation at this juncture, but it is worth pointing out that those who might want to utilize the principle that subjects like Pat are thinking of things under the same modes of presentation are diverging from standard usage. In a careful attack on cognitivism, how anti-cognitivists are understanding modes of presentation should be addressed.

The second and more serious reason that cognitivists should give for being wary of the assumption that Pat uses the same modes of presentation across the attitudes in question is that introspection is not a perfect faculty. As far as I can tell, the only reason we have for supposing that Pat deploys the same modes across her fear and belief is that it seems to her, via introspection, that she is thinking about the very same dog and in the very same way. But there are good reasons to question the outputs of introspection. For example, we know that there are some mental states that go ‘unseen’ by introspection, such as repressed beliefs, and we know that people often harbor attitudes implicitly that those very subjects deny harboring (e.g. implicit sexist or racist beliefs). Just because Pat, via introspection, takes herself to be deploying the same modes or concepts, it doesn’t follow that it is so. Of course, introspection often gets things right, so perhaps Pat is getting things right (a thought I’ll return to in Sect. 5). But it certainly isn’t guaranteed that Pat is getting things right. The crucial point is that cognitivists are not obliged to accept that Pat is deploying the same concepts or modes of presentation and indeed they have some reasons for thinking she very well may not be. If correct, their view hasn’t been shown to yield a contradiction.

But perhaps anti-cognitivists can strengthen their case. There is an attractive principle that, if true, would vindicate the claim that Pat deploys the same concepts or modes in both belief and fear. That principle is the ‘Introspective Knowledge of Comparative Contents’ principle (IKCC). It has been explored recently by Falvey and Owens (1994) and Boghossian (1992), but it has perhaps been put most plausibly and succinctly by Sainsbury and Tye (2012):

IKCC: When our faculty of introspection is working normally, we can know a priori via introspection with respect to any two present, occurrent thoughts whether they exercise the same or different concepts.¹¹

Anti-cognitivists can grant that introspection is an imperfect faculty but they might still maintain that (when properly qualified) our recalcitrant subject will be able to reflect on her fear and her judgement and will be able to tell, on that basis, whether or not the same concepts or modes are at issue. By appealing to IKCC, cognitivism is finally shown to yield a contradiction.

Interestingly, there are good reasons for thinking that IKCC is false. If it is, then it can’t serve as support for the view that insists that Pat deploys the same mode or concept of Fido across the two states—the entailment is still left wanting.

‘Slow Switching’ cases apply serious pressure to IKCC. Suppose that one afternoon Oscar sees Pavarotti, the famous singer, floating at the hotel pool. On this basis, Oscar thinks to himself that Pavarotti swims. That night, unbeknownst to Oscar, Oscar is transported to Twin-Earth which is superficially just like Earth except all the watery-stuff is made of XYZ rather than H2O and there is no Pavarotti on Twin-Earth, only a microphysical duplicate (call him ‘Twin-Pavarotti’). On Twin-Earth (as is standard in the Twin-Earth examples philosophers are familiar with) everyone speaks a language phonetically just like English, but of course on Twin-Earth when someone says ‘water’, she refers to XYZ and when someone says ‘Pavarotti’, she refers to Twin-Pavarotti. Oscar is now on Twin-Earth (and let’s suppose he has been there for a few years now and is embedded in the Twin-English speaking community). He sees Twin-Pavarotti on stage. Upon seeing the man on stage, Oscar sincerely says to himself, ‘Pavarotti sings’. Recalling the day back at the pool, Oscar then thinks to himself, ‘Pavarotti swims and sings!’ But of course Oscar has made a bad inference, an inference that is no better than a common equivocation. But Oscar is unaware of his conflation. Although he has performed a bad inference, it isn’t something he is able to detect by introspection. Recall that unbeknownst to Oscar he was transplanted, and by being transplanted the things he thinks about are switched. But now IKCC is in trouble. It seems implausible that by merely moving someone unharmed to Twin-Earth we destroy his faculty of introspection or memory, so the antecedent of IKCC (‘When our faculty of introspection is working normally’) is in the clear. But it is false of Oscar that he can know a priori via introspection with respect to any two present, occurrent thoughts whether they exercise the same or different concepts. He thinks that he is making a perfectly good inference because he thinks that he is deploying a single concept that refers to Pavarotti throughout his inference. He is incorrect. He is thinking a Pavarotti-thought followed by a Twin-Pavarotti thought. IKCC isn’t obviously such a good principle after all.¹²

Without IKCC it looks very hard to require of cognitivists that Pat deploys the same modes of presentation of Fido across her belief and her fear. If IKCC fails, cognitivists can demand that their opponents offer some other reason for requiring of them that Pat is deploying the very same concepts or modes. Prima facie, the reason was that Pat, upon reflection, tells us that the modes are the same—she insists that she is thinking about the very same dog and in the very same way! But without IKCC, Pat’s own introspective reports are no guarantee. So, a perfectly acceptable cognitivist line might run like this: ‘Pat represents (due to her belief) that the dog is not dangerous but she represents (due to her fear) that the dog is dangerous and she does so under different modes of presentation. Nevertheless, subjects like Pat very often mistake such distinct modes for the same modes.’

Return now to our argument against cognitivism. We saw above that 6 should be replaced with 6′, yielding the following:

• 1. S is rational. (premise)

• 2. S believes that Fido is not dangerous. (premise)

• 3. S is undergoing a fear response in the face of Fido. (premise)

• 4. S believes that Fido is dangerous. (by the truth of cognitivism and 3)

• 5. S believes that Fido is dangerous and believes that Fido is not dangerous. (conjunction of 2 and 4)

• 6′. Any subject who believes that p and believes that not-p and does so by deploying the same (save for the negation) modes of presentation/concepts is not rational.

  (‘Contradictory-with-modes-so-not-rational’ principle)

• 7. S is not rational. (by 5 and 6)

• 8. Contradiction.

The reply presently being offered on behalf of the cognitivist runs thus: 6′ is relevant only if the beliefs formed by the subject in question have contents that are entertained under the same modes or same concepts. The prima facie reason for thinking the contents are entertained under the same modes is that it seems to subjects like Pat that they are. But we know that things are not always as they seem, so cognitivists are well within their dialectical rights to deny that Pat is using the same modes of presentation or concepts.

It is important to reiterate that the key moves thus far are not motivated by a desire to save cognitivism. The verdict that 6 is false was motived by Kripke’s puzzle. The verdict that IKCC is false was motivated by the slow-switching case. Of course those verdicts could both be called into question, but they receive wide support and are certainly live options to which cognitivists can help themselves. Anti-cognitivists may offer a new reason for thinking that 6′ applies to subjects like Pat, but until they do, cognitivists can and should deny that their view leads to a contradiction.

5 The Real Problem for Cognitivists

Up to this point, I’ve wanted to show that cognitivists are in a better position than usually supposed. Anti-cognitivists often appeal (as we saw in the quotes offered earlier in the paper) to the idea that recalcitrant pairs of mental states are ‘conflicting’, ‘irrational’, or ‘contradictory’. Such pairs are not deeply problematic for the cognitivist since there are independently plausible ways to accommodate them. But cognitivists are not in the clear. The real problem for cognitivists is that when they accommodate the conflict offered in the strengthened puzzle, they will quickly have to rely on a degree of mental opacity that one should find implausible. Let us pick the dialectic back up at IKCC.

Cognitivists can, as we saw above, deny IKCC in order to make use of distinct modes of presentation. But notice that the falsity of IKCC guarantees only that there are some cases, with respect to any two present and occurrent thoughts, where a subject is not capable of telling whether they exercise the same or different modes or concepts. Let’s be explicit. IKCC makes a universal claim—all occurrent, present thought pairs are open to introspection. The denial of this claim, supported by the slow switching case, is that some are not so open. Now, Pat’s case might be just such a case. Indeed, any case might be just such a case. For any case of recalcitrance, cognitivists can deny that the same modes or concepts are at issue even when someone like Pat says otherwise. But, ultimately, cognitivists will have to maintain something stronger. They must hold an ‘all-not’ position: all cases of recalcitrance are ones in which a subject is not in a position to tell whether he or she is deploying the same concepts or modes. The reason is that since we know there are indeed some cases of recalcitrance, it had better be that they all fail to be ones where the same modes or concepts are being deployed, else there will be a case we can plug into the revised argument above (the strengthened argument) to yield a contradiction. But this stronger demand shouldn’t be a comfortable one. Although IKCC is false, surely we sometimes (probably very often, in fact) get things right with respect to our own thoughts and there is no special reason for excluding recalcitrant cases from this set. This is the real problem for cognitivists. It’s not that their view logically leads to a contradiction nor is it that their view demands that we hold subjects who are intuitively rational to be irrational. Cognitivists can avoid contradiction, and for any case of recalcitrance presented to them, they can cast doubt on it as a case of irrationality. But ultimately cognitivists need something stronger. Cognitivists must maintain the very implausible principle that in every case of recalcitrance, two distinct modes or concepts are being deployed despite the way things seem to the subject. Cognitivism hence requires an implausibly widespread lack of access to our own mental states. This is the real problem posed by recalcitrance.

6 Conclusion

Recalcitrant emotions are typically taken, often with little explication, to present cognitivists with a serious problem. The alleged problem is that cognitivism predicts mental conflict that allows one to derive a contradiction. I’ve argued that cognitivists can help themselves to various independently plausible resources when when presented with cases of recalcitrance. Cognitivists needn’t maintain that subjects in recalcitrant cases are in mental conflict and so they can avoid the charge that their view leads to contradiction. But I’ve also argued that as the case from recalcitrance is refined and developed, the real problem for cognitivists is exposed. Cognitivists must maintain that all cases of recalcitrance are cases where subjects are, often unbeknownst to themselves, thinking of things in distinct ways. On some of these occasions, this is a reasonable position, but when sufficiently widespread, it requires an unacceptable level of mental opacity.