Radically Self-Deceived? Not So Fast

Fleur Jongepier and Quassim Cassam

Suppose Phineas gets into a railroad accident, and a large iron bar is driven through his head.¹ Upon regaining consciousness, Phineas— previously a kindhearted, modest fellow—now has a general contempt for humankind and a strong belief in his own superiority. If we want to know why Phineas believes he is superior to everyone else, we might ask him for his reasons. In response, he might say that he believes he is superior because his colleagues are untrustworthy or because he has a higher IQ or something along those lines. This is a rational explanation that centrally involves the reasons individuals themselves cite in favor of believing what they believe. Alternatively, we can give a non-rationalizing explanation of why Phineas believes what he believes. A non-rationalizing explanation in Phineas’s case would concentrate on the iron bar being driven through his brain and the fact that significant brain injuries can lead to significant psychological change.

Pelham et al. suggest that being ignorant of why we believe what we believe in the non-rationalizing sense gives us reason to think we don’t know why we believe what we believe. They for instance write that “people often have little or no access to their own thought processes”. But, first, knowing one’s thoughts and knowing one’s thought processes are not equivalent, and ignorance about the latter does not amount to ignorance about the former. We often find ourselves with ‘passing thoughts’ that occur to us for no apparent reason.² One might suddenly find oneself thinking about cats, say, without really knowing why. But one still knows one is thinking about cats. So ignorance about underlying thought processes doesn’t undermine the fact that we know what we are thinking.

Second, Pelham et al. are not clear about what they mean by the notion of a thought ‘process’. A thought process might be the inferential steps one takes to reach a conclusion (a personal-level explanation) or the cognitive mechanisms that were involved (a sub-personal-level explanation). The former is something a person does, whereas the latter is something in which “the person, qua person, does not figure”.³ When it comes to knowing why we believe what we believe, we expect people to know about personal-level thought processes, not sub-personal ones.

Third, acquiring more knowledge about our reasons for belief in the sub-personal non-rationalizing sense does not automatically undermine knowing what we believe in the personal, rationalizing sense. Take Phineas: the explanation of why he believes what he believes in terms of the iron bar running through his head may well leave intact the reasons Phineas himself has for believing other people are inferior. Only if Phineas himself were to become convinced that the non-rationalizing explanation is the (only) right explanation of his beliefs will his rationalizing explanation be weakened or undermined. But even if this were to happen, it’s questionable whether results from social psychology can play a similar role as the iron bar did for Phineas.

The more general point is that knowing about sub-personal thought processes does not seem necessary for most types of self-knowledge. One can know that one believes the weather is fine or that one has the desire to go for a walk without knowing anything about the neurobiological or sub-personal thought processes that underpin these mental states. To think individuals need to know about the latter to have self-knowledge would amount to requiring that individuals need have to have a degree in neurobiology or cognitive science to have self-knowledge.

Regarding knowledge of our actions, the same ambiguity of asking ‘why’ applies. When we want to know why someone performed action A, we can either ask for ‘agential reasons’ or what according to the individual in question speaks in favor of doing A, or we can ask about the ‘motivational factors’ that lead people to do A. Constantine Sandis gives the following helpful example:

I may be motivated to buy a bottle of a certain brand of rum by the advert portraying beautiful, happy, people on a yacht in some exotic location. My agential reason here will not be the thought that I could in some way be like them but, rather, the purported fact that the rum will be of good quality.⁴

Apart from advertisements, there are all kinds of motivating factors that move individuals to do the things they do, including as Sandis writes, their “upbringing, a past trauma, or a fact about human nature” as well as people’s “neu- roscientific make-up”. Things like upbringing or neuroscientific makeup help explain, in a non-rationalizing sense, why someone did what he or she did, but they are not the sort of thing the individual would take to speak in favor of doing A. To claim that individuals don’t know why they do what they do is thus ambiguous between claiming they are ignorant about their own agential reasons and being ignorant about the relevant motivational factors.⁵

This distinction is particularly relevant to the ‘implicit egotism’ studies that lie at the heart of Pelham et al.’s argument. These are studies about people’s unconscious bias to “gravitate towards things that resemble the self” (see Pelham, Harding, and Hardin, Chapter 9). It turns out for instance that people have a preference for marrying people with surnames matching their own and also that “people with the same birthday number [are] disproportionately likely to marry”. Pelham and colleagues also studied common career names that are also common surnames such as ‘baker’, ‘cook’, or ‘carpenter’ and found that there is a ‘surname-career matching effect’. The longer people named Cook had been married, for instance, “the more strongly they gravitated toward working as cooks”.

What are we to make of this? Pelham et al. write that “virtually none” of the participants in their lab experiments on implicit egotism were “able to report the true reasons for their preferences for stimuli that resembled the self”. This almost makes it sound as if we should have expected them to. Imagine the surprise if these participants explained their actions (e.g., career choices) by citing the most recent literature on implicit egotism (“I wanted to be a cook because my last name is Cook”) or if the participants in the classic Nisbett and Wilson stockings experiments explained their choice by saying they wanted the particular pair because they had a right-sided response bias. That individuals don’t come up with explanations of their actions in terms of motivational factors should come as no surprise. If it does come as a surprise, one has confused agential reasons and motivational factors.

More importantly, the authors imply that the ‘true reasons’ here are the motivational factors, but why should this be so? Why should implicit egotism be the ‘true reason’ for marrying one’s partner rather than the agential reasons individuals have? The concept of ‘true reasons’ is not particularly helpful. There are valuable and less valuable explanations of people’s actions, but what makes an explanation valuable is highly context dependent.

Pelham et al. claim that the fundamental assumption that we are capable of free choice is undermined by “growing list of unconscious influences on social judgements and decision-making” and that the “edifice of free will is being chipped away, study by study, by hundreds of modern studies on implicit social cognition”. But if those studies concern motivational factors, and motivational factors don’t automatically undermine people’s agential reasons, this dramatic conclusion doesn’t follow.

The authors (implicitly) suggest that we cannot make our own free choices because we are constantly being ‘influenced’. But the fact that one’s actions are influenced by factors we are not aware of does not mean one’s actions were unfree or that one was not responsible for performing them. It depends on the shape that influence takes, and influence comes in many forms.⁶ Persuading and coercing both count as influencing a person. Influencing someone’s attitudes and actions by engaging in rational debate is not freedom undermining, whereas influencing them by holding a gun to their head is. Coercion and manipulation are standardly assumed to undermine our capacity for free choice, so if the ‘influence’

Pelham er al. describe takes on a coercive or manipulative form, then that would give reason to think our actions aren’t free in the relevant sense.

Coercion involves drastically reducing people’s options and forcing them to perform a particular action or making it impossible for them to perform the action they wish to perform. Coercion can come in radical and easily detectible forms (e.g., life or death choices) but also comes in subtler forms that are more difficult to detect, such as offers one can’t refuse where it’s hard to assess whether the option was genuinely impossible to refuse (e.g., being ‘coerced’ to stay on Facebook, given the lack of adequate alternatives). But it’s evident that the studies discussed by Pelham et al. do not involve coercion in either a drastic or subtle sense. One clearly isn’t forced to buy stockings on the right-hand side or marry partners with the same birth date.

What about manipulation? At times, the authors refer to the ‘psychological forces at work’ as ‘nudging’ individuals to act in certain ways, suggesting that being nudged is incompatible with acting freely.⁷ The implicit thought seems to be that nudges are ways of manipulating people and being manipulated is incompatible with acting freely. But being nudged as such isn’t freedom undermining and can in fact be freedom enhancing. To the extent that biases or weakness of will, for instance, stand in the way of someone’s acting in accordance with their own goal of eating a healthier diet, certain nudges, such as placing healthy products at eye level, can allow them to act more in line with their own goals.^s Also, it isn’t clear why being nudged should lead us to think people aren’t responsible or prepared to take ownership for their own choices. Suppose we confront individuals that were nudged to pick an apple:

“you were being nudged. You picked the apple. Is that really what you wanted?” It seems unlikely that many people would object. At least, the deliberate choice architecture in supermarkets has not yielded mass protest on this account (“Honey, I don’t know what just happened. I wanted to pick up some fruit, but I only have Twinkies in my bag. Darn these folk at Walmart!”).⁹

Nudges aren’t necessarily manipulative. Being nudged to do something counts as manipulative by definition only if we understand manipulation in a descriptive sense as simply the bringing about of a certain change (one might for instance ‘manipulate’ a pencil by sharpening it). But manipulation in the descriptive sense isn’t freedom undermining. Manipulation is only potentially freedom undermining on a normative conception of manipulation. Manipulation in the normative sense typically involves intentionally “perverting” the way a person “reaches decisions, forms preferences or adopts”.¹⁰ Often, this happens through bypassing, undermining, or subverting someone’s capacity for rational deliberation (although not necessarily).¹¹ Clearly, manipulation of this sort isn’t going on in the studies discussed by Pelham et al. This is so even if we drop the requirement that manipulation necessarily involves an intention to manipulate because, even on weaker accounts of what manipulation involves, there is no going around the fact that manipulation involves a person or institution getting another person to do something. So unless we choose to anthropomorphize implicit egoistic or right-sided response biases, biases don’t manipulate in a way that affects free choice.

All of this goes back to the distinction between agential reasons and motivating reasons. For the claim to be plausible that there is no such thing as our capacity to make free choices in the supermarket, at work, and in life in general, what needs to be shown is that there’s something suspicious about our agential reasons, but it’s not clear that the studies discussed by Pelham et al. give reason for suspicion in that regard.

One might worry that we have not really been taking seriously the challenge put forward by Pelham et al. If it’s true that nonrational explanations of our actions are available, and that in a significant class of cases these explanations do not match the reasons we ourselves give, then our personal-level reasons are not as decisive as we take them to be; indeed they seem to be way off the mark. Even if motivational factors, and thus insights from (social) psychology, do not automatically undermine people’s personal-level reasons and their knowledge of them because these are two distinct domains; large-scale mismatches between the two should still be unsettling, and we have reason to be suspicious of our own agential reasons all the same.

The question, though, is whether they really are mismatches and whether they are really large scale. As Pelham et al. themselves point out, a longer-standing worry about experiments in social psychology is that the results usually only emerge in “artificial confines of the lab where people make unimportant decisions”. For dramatic claims about self-knowledge and free will to be genuinely unsettling and thus relevant to our self-understanding, the results should be replicable in the wild, with people who have intact brains and who make nontrivial decisions. Hence Pelham et al.’s proposal to focus on implicit egotism studies that are “robust outside the lab” and which concern “major life decisions”¹² is admirable. The result is, though, that Pelham et al.’s dramatic claims hinge primarily on a handful of studies of which the effect is, as the authors acknowledge, rather modest (there was a 6.5 percent bias for people to marry others who shared their birthday numbers). Just how ‘large scale’ the effects really are is not entirely clear.

More importantly, the effects don’t bring about any mismatches per se. The fact that for a (small) number of people their last names or birth dates turn out to be a factor in their decisions remains just that: a factor. Sometimes arbitrary facts about one’s last name or birth date со-determine one’s actions. But the ‘со’ is important here: nonrational factors play a role, but not the only role, and it seems in most cases not even a particularly dominant role. This at least helps explain why Adam Smith didn’t become a smith, Lynne Rudder Baker didn’t become a baker, and Judith Butler didn’t become a butler. Clearly also people whose occupations and last names do match don’t need to go into existential crises for also in their cases many other factors led them to pursue the careers they chose to pursue. The presence of nonrational factors explaining our actions do not as such point to there being any mismatches because the two aren’t necessarily competing explanations. Having Cook as one’s last name might indeed influence one’s decision to become a cook, but this doesn’t mean one has become a cook because one is named Cook or that being named Cook is the ‘real’ reason one became a cook, again unless one had no personal-level reasons whatsoever to become a cook.

Perhaps there will be many more implicit egotism studies. It may for instance will turn out that people who live in places like Catford (London) or Dogtown (California) are more likely to get themselves a pet; that people living in Drogen (Germany) are more likely to use drugs, or that people living on Chopin lane or Mozart street are more inclined to listen to classical music. None of this would show that our choices aren’t free in the relevant sense or that we are massively self-deceived about why we do what we do. There’s more that moves one to buy a pet, do drugs, or listen to classical music than facts about where one lives. It does follow that many of our choices also have nonrational co-determinants, but that should not be news to anyone.

This is not to say that agential reasons are somehow immune to alternative non-rationalizing explanations or that finding out about the motivational factors of our actions can have no impact on one’s agential reasons. The question is what is required for this to happen. Helpful in this context is John Christman’s proposed condition of ‘hypothetical resistance’: a person is not free or autonomous relative to his or her attitude or action if he or she would have resisted the development leading to that action or attitude.¹³ What matters is, according to Christman, “what the agent thinks about the process of coming to have the desire, and whether she resists that process when (or if) given the chance” (ibid.). If one came to realize that one’s decision to marry one’s partner was the result of coercion or manipulation, or indeed a massive brain lesion as in Phineas’s case, then that would presumably lead to resistance in the relevant sense and thus to the conclusion that one’s choice was not (sufficiently) free.¹⁴

The question then becomes: would the implicit egotism studies discussed by Pelham et al. lead to hypothetical resistance? This seems unlikely. Suppose the couples in the birthday number studies who married partners with the same birthday number were given a call and were informed about the existence of implicit egotism and were told that their preference of marrying their partner was influenced by an unconscious preference for things that resemble the self. This is unlikely to be the sort of‘process’ rhar would lead to resistance and is not something that would undermine the agential reasons for marrying their partners or for going on in their jobs as cooks and bakers. Perhaps the news might make one uncomfortable for a moment, but the discomfort will fade so long as one also has robust personal-level reasons for marrying one’s partner or pursuing one’s career.

We are far from omniscient when it comes to knowing our own mental states, and we probably credit ourselves with too much agency in making decisions on numerous occasions. But it’s a little early to conclude that because we don’t know about all of the non-rationalizing processes underlying our mental states and actions, we’re deluding ourselves into thinking we are capable of knowing our own mental states and making our own choices.

Notes

• 1. This example is a loose adaptation from the actual case of Phineas Gage, who underwent profound personality changes due to a brain lesion (Dama- sio et al. 1994).
• 2. Cassam (2011).
• 3. Elton (2000, 2).
• 4. Sandis (2015, 268).
• 5. As Ryan Cox (2018) observes, there’s also some evidence that psychologists agree that people have privileged knowledge about their reasons. Nisbett and Ross (1980), for instance, write that people are “often right in their accounts of the reasons for their behavior” and that “[a] person who answers a telephone and asserts that he did so ‘because it was ringing’ is surely right” and “[a] person who asserts that he opened the refrigerator door because he was hungry is usually right”.
• 6. See also Coons and Weber (2014).
• 7. Nudging is typically understood as changing people’s behavior by making use of people’s psychological weaknesses, such as their biases and the sub- optimal decisions they make on autopilot and under time pressure, without “forbidding any options or significantly changing their economic incentives” (Thaler and Sunstein 2009, 6). Biases themselves do not, in this sense, nudge individuals at all. Rather, biases can be used to nudge individuals. When Pelham et al. talk about ‘nudging’, they thus seem to have in mind something much broader, such as being motivated to act on the basis of nonrational means.
• 8. Thaler and Sunstein (2009).
• 9. Nys and Engelen (2017, 204).
• 10. Raz (1986).
• 11. See Gorin (2014).
• 12. Pelham and Mauricio (2015, 718).
• 13. Christman (1991).
• 14. A lot can be said about whether this criterion is a plausible one. Our main aim here though is not so much to defend this particular criterion as to sketch a possible way in which the two explanatory domains might meet, that is, how motivational factors and evidence from empirical psychology could become relevant for, or potentially undermine, agential reasons.

References

Cassam, Quassim. 2011. “Knowing What You Believe.” Proceedings of the Aristotelian Society 111: 1-23.

Christman, John. 1991. “Autonomy and Personal History.” Canadian Journal of Philosophy 21: 1-24.

Coons, Christian, and Michael Weber, eds. 2014. Manipulation: Theory and Practice. Oxford: Oxford University Press.

Damasio, Hanna, Thomas Grabowski, Randall Frank, Albert M. Galaburda, and Antonio R. Damasio. 1994. “The Return of Phineas Gage: Clues about the Brain from the Skull of a Famous Patient.” Science 264: 1102-1105.

Elton, Matthew. 2000. “The Personal/Sub-Personal Distinction: An Introduction.” Philosophical Explorations 3: 2-5.

Gorin, Moti. 2014. “Do Manipulators Always Threaten Rationality?” American Philosophical Quarterly 51 (1): 51-61.

Nisbett, Richard E., and Timothy DeCamp Wilson. 1977. “Telling More than We Can Know: Verbal Reports on Mental Processes.” Psychological Review 84 (3).

Nys, Thomas, and Bart Engelen. 2017. “Judging Nudging: Answering the Manipulation Objection.” Political Studies 65: 199-214.

Pelham, Brett, and Carvallo Mauricio. 2015. “When Tex and Tess Carpenter Build Houses in Texas: Moderators of Implicit Egotism.” Self and Identity 14: 692-723.

Raz, Joseph. 1986. The Morality of Freedom. Oxford: Oxford University Press.

Sandis, Constantine. 2015. “Verbal Reports and ‘Real’ Reasons: Confabulation and Conflation.” Ethical Theory and Moral Practice 18: 267-280.

Thaler, Richard H., and Cass R. Sunstein. 2009. Nudge: Improving Decisions about Health, Wealth, and Happiness. London: Penguin.