Haidt’s “Emotional Dog” has three legs, a broken tail, and fleas
The Broken Tail
Picking up from Part 2a, I further develop the argument that Haidt failed to properly account for moral development and, in particular the process of moral internalization, in his Social Intuitionist Model (SIM).
The development of moral reasoning
In their critique of SIM, Saltzstein and Kasachkoff (2004) say:
Haidt’s theory is limited by his taking a nondevelopmental approach to morality, confusing development with deployment. (By “developmental,” we mean minimally that the deployment of a mental operation is shaped by the process, including the order and timing, by which it is acquired.)
In Haidt’s response, he goes into quite some defence of the two sections in his paper about moral development in children. Saltzstein and Kasachkoff meant moral development in a more general sense, though, as they sought to clarify with the parenthetical comment in the quote above. Furthermore, the very first sentence in the contentious section of their critique states that “[b]ehaviors that function in a nondeliberative and automatic way are not necessarily formed or developed in that way” (p. 279). Saltzstein and Kasachkoff are making a general comment about moral internalization. The previous article addressed this at length. Here the focus is mainly on moral internalization in childhood, a process discussed in Haidt’s article, but not obvious in the model.
Haidt should have made the point that the development of morals is qualitatively different in childhood. Presumably, his response would be the same as to Saltzstein and Kasachkoff, that there are two sections devoted to moral development in children. This is true, and what’s more, they form an exceptionally good synthesis of the work on childhood moral development. However, if, as seems to be implied by the necessity of those two sections, moral development in children is qualitatively different to that of adults, where is this illustrated in the SIM?
Indeed, I would suggest that Saltzstein and Kasachkoff are saying that the model, as drawn (see Image 1, below), is more a model of iterative moral deployment than one of moral development from the cradle to the grave.
You might argue that the Social Intuitionist Model, as drawn, does allow for this qualitative difference between the adult world of development, primarily through deployment, and the childhood world of development, primarily through the externalization of evolved abilities. But questions arise: Is the qualitative difference in changes to the relative strength of the influence of links five and six? Links three and four? Both? Something else?
Pizarro and Bloom (2003) state that Haidt’s “model allows for virtually all conceivable relationships between environment[al] stimuli, deliberative reasoning, moral intuitions, and moral judgments” (p. 195), which is basically saying that the model is more a survey of the general landscape than a model of the particular processes. By extension, this means that the model, as drawn, is not the one that is defended in the text.
In the previous article, I discussed moral development over time, with a focus on the ongoing moral development of adults. In what follows, I will mainly focus on moral development (and deployment) in children, focusing mostly on the text of the two sections Haidt offers as a developmental model of moral psychology as it relates to children.
More moral internalization
Haidt says, “…the most important developmental question about intuitions is not, ‘How do they get into the child?’ but rather, ‘How do they get out?’” If evolved intuitions emerge at around the same time in all children, as Haidt goes on to say, then the question is wrong. I would argue that it is not an intuition that is suddenly available to the child, it is a type of information processing that gives rise to a particular type of intuition. Haidt seems to more-or-less concur with this when he says:
This pattern of sudden similarly timed emergence with overgeneralization suggests the maturation of an endogenous ability rather than the learning of a set of cultural norms.
One can imagine for example, that the maturation of Theory of Mind (the ability to conceptualize other people’s knowledge, thoughts and desires as distinct from one’s own and as real to them as yours are to you), would give rise to an explosion in taking other people’s mental contents into account. Haidt discusses this when talking about Western parents teaching their children to share and play fairly. So, we’re talking about an ability that gives rise to a certain class of intuitions, not the intuitions themselves and certainly not “the externalization of intuitions” (p. 826), which is what Haidt inexplicably calls the entire section in which he discusses this endogenous ability.
Young children, going through these developmental phases relating to neural development, are socialized by their families to conform to the social norms of their culture (or at least the family’s idiosyncratic take on them). Sometimes this will involve children coming to terms with their preferences not being acceptable. Haidt also talks about the importance of peer socialization in children and in particular about a sensitive period (between the ages of 9 and 15), during which time the moral norms that the child will go on to deploy throughout their lives are primarily internalized.
The older children, those in the sensitive period (I’ll call them adolescents for ease), seem likely to be dealing with many instances where their intuitions clash. Presumably, adolescents will need to figure out whether to assimilate this new information into their existing intuitions or make an accommodation for it, effectively carving out a new intuition. This would especially be the case where their intuitions are at odds with those of their families and their peer groups, and where expressing their misgivings to either group might expose them to ridicule (or worse).
Motivations
The fear of ridicule or other social consequences motivates individuals (and, one might suggest, especially adolescents) to reason in certain ways. The individual might be motivated to take into account their appearance, accountability, accuracy, or some combination of these. With these, I am referring to Haidt’s citing of ‘impression motivation’ (appearance) and ‘defence motivation’ (accountability) as the bases for post hoc reasoning. Cordelia Fine (2006) notes a third category from the same research, saying, “we are also importantly motivated by the desire to hold objectively true beliefs and attitudes (‘accuracy motivation’)” (p. 95). It’s odd that Haidt makes such strong cases for appearance and accountability as motivations but rarely does so for accuracy given that he, like Fine, cites Chaiken, Giner-Sorolla and Chen (1996), whose paper explicitly talks about appearance and accountability as motives “beyond [not instead of] accuracy” (p. 553).
If we accept that adolescents are not moral philosophers and that “reasoning,” especially in the heat of the moment, is defensive self-justification, it seems likely that many children in the sensitive period will be confronted with events for which they do not have prior experience. They might be motivated to ‘go along to get along,’ in the first instance, regardless of any explicit beliefs. If an interaction, whether with a parent or a peer, is something that the other group might consider immoral, then the adolescent’s two primary sources of moral information are at odds. It seems reasonable to suggest that they might do quite a lot more reflective contemplation (link 6) than other people. The question then becomes, what skills do they bring to this process? What have they learned, through socialization, from their parents?
The Nurturant Parent and the Strict Father
George Lakoff (1995) described two distinct underlying metaphors for parenting styles. The Nurturant Parent, appeals to a child’s empathy and directs it with consequentialist reasoning (e.g. how would doing Z make X/Y feel? How would you feel if X/Y did Z to you?). The intention is that the child reflects on the situation and internalizes the link between the physical act and the relevant moral emotion. This is to say, the parent provides a foundation for reasoning and also a path to internalization of that reasoning. The Strict Father, by contrast, places an emphasis on explicit rules, possibly highlighting the authority behind the rules, but seldom giving explanations as to why the rules exist or what they are meant to achieve.
In the situation where an adolescent encounters a behaviour or a justification of behaviour that contravenes their expectations, what are they to do? It may be similar to how children come to understand what words mean; surprising exceptions drive greater learning (Fazekas, et al., 2020). But what is learned when the surprise contravenes an inflexible rule? What happens when there are no explicit rules to refer to?
Several of the critiques point out that this situation happens to adults, too. However, it seems probable that the qualitative difference between children and adults has to do with a greater frequency of such instances occurring in adolescence, mostly as a product of socialization outside of the home. But the frequency for adults is not inconsiderable, as Pizarro and Bloom (2003) say:
…although many of our views might be shaped by our cultural context, there are innumerable instances in which people — not necessarily professional philosophers — take moral stands that put themselves very much at odds with members of their community. Examples include “righteous Gentiles” in Nazi Germany … college professors who defend the abolition of tenure, and many pacifists during wartime. Also, modern humans often face moral issues that have not been anticipated by evolution or by the culture in which they were raised: Should research on stem cells be permitted? Should graduate students be permitted to unionize? Should the American government pay reparations to the descendants of slaves? These are not questions about empirical fact; they are questions of right and wrong. (p. 195) [examples involving children edited out to keep to the point being made]
It seems clear that a child provided with only a set of rules delivered by a moral authority (aka truisms) is at a disadvantage compared to a child that is encouraged to self-generate, understand and internalize the reasons behind those rules. The former is likely to be more motivated by appearance, the latter by accuracy.
Most children experience some combination of the extremes presented by Lakoff’s two parental models, so they have some internalized reasons and some truisms. It seems clear that, in noticing this distinction, as Saltzstein and Kasachkoff suggest, there is a difference between development and deployment. Additionally, the environments in which development and deployment occur are (must be) important.
Wicked environments
Haidt (2001) supports his contention that reasoning is usually post hoc but may appear objective to the reasoner with a set of findings where participants in the cited papers were: [1] deceived as to the cause of their physiological state (Nisbett & Schachter, 1966); [2] hypnotized (Zimbardo, et al., 1993); [3] subliminally exposed to relevant stimuli (Kunst-Wilson & Zajonc, 1980); or were [4] split-brain patients (Gazzaniga, et al., 1962).
These studies, I would argue, give rise to confabulation, or what Harry Frankfurt (1986) might have called bullshit. Kopelman (1987) distinguished between provoked momentary confabulation, where real memories are displaced in context and brought about by asking questions, and fantastic confabulation which is chronic, grandiose, and due to frontal lobe damage. It seems fair to suggest that split-brain patients are engaged in fantastic confabulation. The other three are provoked. Or, as Frankfurt (1986) went on to say, “bullshit is unavoidable whenever circumstances require someone to talk without knowing what he is talking about” (p. 18).
This is not to say that individuals will never be mistaken about the cause of their physiological state or always aware of the stimuli impacting upon their thoughts. Instead, it is to say that, much of the time, people are called upon to apply their knowledge in kind environments (Hogarth, 2001) or at least environments that are sufficiently similar to the environment in which they learned the relevant fact (or within the individual’s tolerance for ambiguity), such that they will not be provoked into bullshitting. The citations that Haidt has assembled say more about the environmental constraints that we evolved in, environments where we were not being barraged by advertising, which is to say, not being deceived, hypnotized or subliminally influenced.
Cordelia Fine agrees that our reasoning can be biased by irrelevant factors, such as our mood (e.g. Forgas, 1994). However, where individuals feel they may be judged on accountability, appearance or accuracy, they may also be able to override the influence of automatic processes such as mood or recently experienced intense emotions. Fine notes that Lerner, Goldberg and Tetlock (1998) found that individuals that had been angered by viewing a video that was unrelated to the decision they had to make about the next video were better able to override the emotion if they believed that they would have to explain their judgment afterwards. This seems to be a mix of accountability, appearance and accuracy motivation, with accountability and accuracy predominating. The individual is aware of the biasing nature of feeling angry and intentionally overrides that to make a reasoned decision. It seems to be the case, therefore, that an environment where accuracy is valued would have the effect of encouraging individuals to lean into accuracy instead of (or as part of) appearance and accountability.
Some individuals, however, may seek out certain emotions in order to come to the judgment that they want to come to (the social media age has given rise to phrases like hate-watching and doomscrolling, for example). Consciously using unconscious processes seriously blurs the distinction that Haidt wants to make between intuition and reasoning. Utilizing emotions in this way relates to Pizarro and Bloom’s suggestion that individuals create their own decision-making environments.
In Summary
The implication of Haidt’s section on the development of morality in children is that there are three qualitatively different periods in a child’s life: childhood, adolescence and adulthood. This is not in the illustration of the Social Intuitionist Model. Haidt also discusses two types of motivation related to moral reasoning; appearance and accountability. These motivations more readily give rise to the constructed rationalizations of the Social Intuitionist Model. Haidt ignores a third motivation from the same literature, which would more obviously underpin moral reasoning of the type that he claims tends not to occur as often. There are two models of parenting style in the literature, with one relying more on moral reasoning and motivated by accuracy, with the other relying more on moral truisms and motivated by appearance.
In the final part of this set of articles, I pick off a few other concerns with Haidt’s paper.
