Haidt’s “Emotional Dog” has three legs, a broken tail, and fleas
The Broken Tail
In his paper, The Emotional Dog and Its Rational Tail: A Social Intuitionist Approach to Moral Judgment, Jonathan Haidt suggests that “moral emotions and intuitions drive moral reasoning, just as surely as a dog wags its tail.”
Of lesser anatomical importance to most dogs, but of huge significance to Haidt’s Emotional one, is that (rational) tail. The four legs and tail of the Emotional Dog form part of an imperfect metaphor; the removal of the fourth leg (see part 1, The Three-Legged Dog) also puts the first break in its tail. The breaks are mostly in the definition(s) of reasoning that Haidt is using.
(Links 5 and 6 in the model below) curls back over its (emotional) body of intuition and judgment.
(From left to right, photos by Olga Shenderova and Juan García on Pexels.)
These are the breaks
The explicit definition of moral reasoning that Haidt used in his article was that it is:
conscious mental activity that consists of transforming given information about people in order to reach a moral judgment. To say that moral reasoning is a conscious process means that the process is intentional, effortful, and controllable and that the reasoner is aware that it is going on (p. 818).
In my previous article, I pointed out that Haidt redefined moral behaviour to exclude the overriding of immoral impulses because that is a process involving moral reasoning (per the above definition). I argued that this probably excludes the majority of moral judgments and therefore leaves the Social Intuitionist Model describing moral judgment only where it leads to outwardly visible moral behaviour.
It is difficult to observe someone not doing something you didn’t know they had an impulse to do but overrode (see also Behaviorism).
It is possible that Haidt is instead taking it to be the case that the overriding of impulses is also intuitive in nature and therefore not a particular problem for his model. However, it seems that most people are conscious of having immoral thoughts and actively choosing not to act on them. Again, this seems to come under Haidt’s definition of moral reasoning. As such, it is appropriate to ask whether there is another definition of reasoning under and beneath the explicit one adopted by Haidt. If there is, Haidt may have conflated the two in the Social Intuitionist Model (see my Glossary for the difference between conflation and equivocation).
In what follows, I will address several key instances of reasoning being more directly causal in moral judgment than Haidt allows, to see if the underlying definition of reasoning changes.
Before I get underway, I (along with the authors of all three critiques I will mention) agree with Haidt’s proposal to reduce the influence of reasoning on models of human morality and to recognise the influences of intuitions and social interaction. However, I (and those authors) think his overall proposal swings too far in the opposing direction.
Moral dumbfounding
Among Haidt’s first references in defence of the Social Intuitionist Model (SIM), is his own work with Fredrik Björklund and Scott Murphy (Haidt, et al., 2000), discussing “moral dumbfounding” (PDF). That paper defines this phenomenon as “the stubborn and puzzled maintenance of a judgment without supporting reasons” (p. 1). This paper remains unpublished, has not undergone peer review and, in fact, not all of the video footage of participants was coded prior to the write-up (as per a note appended to the paper in 2010). For a full, up-to-date review of the concerns relating to moral dumbfounding, see the work of Cillian McHugh.
Haidt takes moral dumbfounding to support the idea that we have intuitions that aren’t underwritten by reasons. However, there are (at least) two ways that dumbfounding might come about that don’t rely on intuition.
1. Truisms
When morally dumbfounded, people may be referencing moral truisms (Maio & Olson, 1998): “beliefs that are widely shared and rarely questioned” (p. 294). These would be part of the “set of virtues held to be obligatory by a culture or subculture” in Haidt’s definition of moral judgment (p. 817). Another way to look at such truisms is through a lens that Haidt provides. He says, “solitary moral reasoning may be common among philosophers” (p. 820), and “such an ability may be common only among philosophers” (p. 829). As such, it’s likely that non-philosophers will rely on axioms that are further up the reasoning chain than the bedrock axioms that philosophers are searching for or arguing about.
Pizarro and Bloom (2003) refer to these non-philosopher axioms as first principles, suggesting that they “serve as a starting point for deliberative reasoning” (p. 194). Asking an individual why (or why not) to engage in a certain action and asking for reasons — as Haidt and colleagues do — is therefore pointless; there are no reasons, but there aren’t necessarily intuitions, either, just (idiosyncratic) first principles or truisms.
2. Internalization
When morally dumbfounded, an individual may have actively reasoned to a particular response on numerous similar occasions in the past. However, over time, it has become so practised as to be automatic (this is one type of what is called moral internalization in the literature; see Hoffman, 2000, for a book-length overview). This being the case, the individual’s reasoning is directly causal, but it is so far in their past that they no longer have accurate access to it.
To apply a metaphor to the situation: we wouldn’t deny that a tennis pro’s formidable abilities are the product of anything other than long hours of practice and competition. (If you’re tempted to say ‘talent,’ I would recommend reading Matthew Syed’s book, Bounce, to challenge that notion.) As Malcolm Gladwell noted, in his book, Blink (2005), those tennis pros will not be able to accurately explain how they do what they do. He quotes Vic Braden, a professional tennis coach:
Out of all the research that we’ve done with top players, we haven’t found a single player who is consistent in knowing and explaining exactly what he does … They give different answers at different times, or they have answers that simply are not meaningful. (p. 67)
That’s remarkably similar to Haidt’s definition of dumbfounding — “the stubborn and puzzled maintenance of a judgment without supporting reasons” — but we know that professional tennis abilities are the product of a long internalization process; years and years of practice and playing at the highest level. So, we can refer to the process by which the skill has arisen and doing so is not begging the question, but the specifics are nevertheless out of (conscious) reach.
All three critiques of Haidt’s paper raise the point of reasons becoming intuitions over time, though not all use the word internalization. Cordelia Fine (2006), refers to the “automatization of judgment goals based on prior moral reasoning” (p. 85) and Pizarro and Bloom (2003) state that “[p]rior reasoning can determine the sorts of output that emerge from these intuitive systems” (p. 194).
In fairness to Haidt, he does mention internalization, but only to suggest that intuitions that humans evolved to develop through childhood (“sympathy, reciprocity, and loyalty,” p. 826) are externalized “rather” than internalized. (See my note in the first article about rathering.) In all likelihood, both top-down (internalization) and bottom-up (externalization) processes are in play.
3. All points in between
A third possibility arises from Haidt’s claim, addresses the social aspect of the Social Intuitionist Model, and spans the gap between the two just mentioned. An individual might be supplied a reason (someone else’s ex post facto self-justification) and adopt it as their own. This presumably splits into further subtypes depending on whether the individual making that self-justification is an authority, a peer, or someone else, whether the recipient of the reason values authority or in-group and if (and to what degree) fairness or harm alters the valuation or deployment of those elements. These features could reasonably be seen to interact, providing a typology that causes a reason (the ex post facto self-justification) to become anything from a truism (if delivered by an authority and accepted without question) to a reason (which is internalized into intuition through repeated deployment in subsequent reasoning).
A truism could also be deployed so often as to become intuitive via internalization, the difference being that even if the individual had access to the event that caused them to adopt it, all they could say was which authority they accepted it from.
An effort in (re)construction
Moral dumbfounding underpins Haidt’s claim that:
moral reasoning is an effortful process, engaged in after a moral judgment is made, in which a person searches for arguments that will support an already-made judgment (p. 818).
If a truism underpins the judgment then providing a reason would be an effort in constructing a self-justification. Most people recognize that ‘because my parent/pastor/preferred politician said so’ is not a reason (or at least not a reason that has been reasoned to).
Trying to recall an autobiographical memory that relates to the original instance of adopting a particular line of reasoning would also be effortful. However, I would argue that it qualitatively differs from the defence of a truism.
The effort of recalling the first time you adopted a particular reason for a particular judgment of a particular act may provide an incorrect recollection of the genesis of that line of reasoning. It may provide a more recent instance — memory is, after all, reconstructive in nature — but this is nevertheless indicative of a history of adopting that line of reasoning and not indicative of reliance on a truism or effort in the construction of a rationalization.
Philosopher, Mary Salvaggio (2017) might disagree with me. It seems clear that a truism is necessarily a semantic memory. It is a “fact” (a belief, really), that can be recalled as something an individual believes in the same way that one remembers one’s name. A highly rehearsed judgment, which is to say an intuition through internalization, may also be a semantic memory but, unlike a truism, it has a history that was previously an episodic memory. However, time passes and the need to remember the specific reasons fades, either through lack of use or a lack of challenge to its use. All one can say is, it is a belief that one holds because it appeared to be a justified belief at the time it was adopted.
Saltzstein and Kasachskoff (2004) make a slightly more generalized version of this same point.
This is not justification in the strong philosophical sense for, as Salvaggio points out: if you could justify it all the way down to philosophical axioms, the memory is “beside the point” (p. 5). In everyday life, though, rigorous philosophical reasoning, as Haidt implies, is not regularly deployed. So, when called upon to justify a belief in the social (and thus Social Intuitionist) sense, one might be rationalizing (constructing a reason), but one may also be attempting to remember (reconstructing a reason). Haidt claims it is the former, but offers no way to distinguish it from the latter.
In summary
Truisms suggest that we can have reasons that we didn’t reason to, and internalization suggests that we can have reasons that we did reason to but for which the specifics of that reasoning are difficult to extract from memory. The difficulty in reconstructing a memory should not be confused with the effort of constructing a justification. Haidt’s definition of reasoning seems to see these as identical because they both give rise to dumbfounding. However, Haidt’s theories indicate that these two sources of reasons exist and might just be the polar ends of a spectrum and, at some point along that spectrum, reasons are underpinned by reasoning.
This section on the broken tail has turned out to be a bit longer than expected and is necessarily broken in two. (I know this metaphor has now gotten completely out of hand.) In part 2b, I extend the argument about moral internalization, looking at its development through childhood and the contexts in which it occurs.
