Norm-like Patterns in Joint Action: A dual-process theory
Date given: Tuesday, 6th June 2023
These are the notes for a talk with John Michael based on some work-in-progress.
Joint action in humans and other animals involves various patterns of behaviour which appear to accord with norms. To illustrate, humans will often match the effort they put in to the effort a partner contributes (Székely & Michael, 2023), and dogs will perform play bows which encourage others to drop their guard (Bekoff, 2001). These norm-like patterns facilitate successful joint action. But what is going on psychologically and normatively? The main philosophical tools available are joint commitment (Gilbert, 2013) and reliance (Scanlon, 1998; Bratman, 2006). Neither seems sufficient for understanding psychological and normative aspects of norm-like patterns, or so we will argue. Instead we offer a novel conjecture with two parts. First, norm-like patterns in joint action are a consequence of fast, cognitively undemanding processes that are liable to produce behaviours in conflict with reasoned judgements about norms. Second, these fast processes are identical to those postulated by dual-process theories of ethical cognition (Greene, 2015). This conjecture generates predictions. We should be able to construct cases in which people follow norm-like patterns while denying that any corresponding norm applies. And their tendency to follow norm-like patterns should be modulated by factors that have been shown to modulate fast ethical processes, including cognitive load, disgust and bitterness. If correct, the norm-like conjecture may contribute to characterising aspects of joint action that are shared with other species, and to illuminating the origins of aspects which are unique among humans.
Norm-Like Patterns are patterns in behaviour that conform to principles, upheld (in part) by cost-imposing responses to violations, where conformity matters for solving a coordination problem these agents or their ancestors faced.
Norm-like patterns are widespread in humans and many other animals.
Why are norm-like patterns not norms? Because the principles need not be true, nor would the agents necessarily endorse them. And they may not be currently appropriate.
The existence of norm-like patterns raises two questions:
What, if any, normative notions are needed to characterise norm-like patterns?
What are the psychological mechanisms underpinning them?
Joint Commitment and Reliance
The notion of joint commitment is nicely illustrated by a famous example:
‘If they are walking together, both Andrea herself and Heinrich will have the understandings [that]: by virtue of their walking together Andrea has a right to Heinrich’s continued walking alongside her, together with the standing to issue related rebukes and demands.’ (Gilbert, 2013, p. 25)
Could norm-like patterns reflect joint commitments in Gilbert’s sense?
The existence of joint commitments could explain how some norm-like patterns are upheld. But this cannot be the whole story because dogs and other nonhuman animals are unlikely to have the understandings required for joint commitment.
Michael & Butterfill (2022) offer a complementary problem for joint commitment.
A hint about the notion of reliance is contained in Principle M:
‘Principle M: [...] it is not permissible for one person, A, in order to get another person, B, to do some act, X [...], to lead B to expect that if he or she does X then A will do Y [...] when in fact A has no intention of doing Y if B does X, and A can reasonably foresee that B will suffer significant loss if he or she does X and A does not reciprocate by doing Y.’ (Scanlon, 1998, p. 298)
This principle could also explain how some norm-like patterns are upheld. But as there are norm-like patterns that depend on what is actually done rather than what is intended or foreseen, it cannot be the whole story.
A further problem—which applies to stealing Gilbert’s and Scanlon’s ideas alike—is that those authors are both attempting to genuine norms. These are norms that the agents concerned would, on reflection, agree with; or at least they would not, with full knowledge and understanding, reject them. This is a problem because, as mentioned earlier, the principles to which norm-like patterns conform need not be true, nor would the agents necessarily endorse them. And they may not be currently appropriate. To illustrate: effort-matching occurs even when it reduces performance; and people may uphold norm-like patterns involving food (cannibalism, for example) and and sex (incest, for example) while rejecting the principles to which the patterns conform.
Although there is broad consensus that avoiding poisons is one of the functions of sensitivity to bitterness, the correlation between bitterness and toxicity is not super strong (Nissim, Dagan-Wiener, & Niv, 2017). Bitter things can be beneficial (such as caffeine, for example). Indeed, as well as indicating toxicity, bitterness is also associated with medicinal properties. Nissim et al. (2017) suggest that animals may exploit this association by eating bitter substances when ill. To illustrate, chimpanzees suffering from diarrhoea and other symptoms of parasite infection extract and chew an extremely bitter pith, which appears to improve their health (Huffman, 2001, p. 939).
The case of bitterness illustrates the role of helpful principles in characterising fast process. In this case, the helpful principle is: do not eat poisons. Relating the fast process to the helpful principle enables us to understand why the fast process exists. Importantly, the helpful principle need not match what is good or the useful; nor need it be a principle that agents would endorse, or follow, in full knowledge and understanding. Note also that the fast process need not involve any representation of the helpful principle (nor, as far as we can see, of any principles).
If there were a social equivalent of bitterness, namely a sensational response to socially toxic events, then that would likely explain norm-like patterns. This is probably too much to ask for. But perhaps we can exploit research in moral psychology that comes close to this idea.
A Dual-process Theory
The helpful principles may be false, and those whose actions manifest the norm-like patterns may know this.
The fast processes need not involve any representation of the helpful principles (nor any representations at all).
We are not committed to the idea that fast moral processes are the only source of norm-like patterns, just that they are one source.
- Norm-like patterns will dissociate from normative judgements.
- Increasing cognitive load will enhance norm-like patterns.
- Disgust and bitterness will modulate judgements following observation of violation of norm-like patterns.
Since automaticity and cognitive efficiency are matters of degree, it is only strictly correct to identify some processes as faster than others.
The fast-slow distinction has been variously characterised in ways that do not entirely overlap (even individual author have offered differing characterisations at different times; e.g. Kahneman, 2013; Morewedge & Kahneman, 2010; Kahneman & Klein, 2009; Kahneman, 2002): as its advocates stress, it is a rough-and-ready tool rather than an element in a rigorous theory.
On some views norm-like patterns might actually involve norms after all. For instance, Westra & Andrews (2022) state that a normative regularity is ‘A socially maintained pattern of behavioral conformity within a community.’ On this definition, norm-like patterns are normative regularities. We are neutral on whether they are right. ↩︎
Bitterness appears innately aversive. In rats, ‘[t]he modal elicitor of aversive behaviours is a bitter, normally avoided substance like quinine, which evokes chin rubs, gapes, face washes, forelimb flails, and paw treads’ (Forestell & LoLordo, 2003, p. 141; Grill & Norgren, 1978). ↩︎
A range of animals including sea anemones become averse to a food type after a single bitter encounter (Garcia & Hankins, 1975). And in humans mixing a neutral flavour (vanilla, for example) with a bitter substance can reduce liking for that flavour (Baeyens, Crombez, De Houwer, & Eelen, 1996; Dickinson & Brown, 2007). ↩︎
What is the evidence that avoiding poisons is one of the functions of sensitivity to bitterness? Animals who encounter a greater proportion of poisonous foods in their normal diet (herbivores) show both higher sensitivity to bitterness (Li & Zhang, 2014) and a higher tolerance for it (Ji, 1994). This makes sense because herbivores need to take more risks: they could not get enough to eat if they rejected everything bitter. Further, two changes in diet which reduce exposure to toxins, namely eating more animals or cooking with fire, may have gradually reduced sensitivity to bitterness in humans (Wang, Thomas, & Zhang, 2004). ↩︎
Although unlikely, it is possible that regular sensations of bitterness also play a social role. Chapman, Kim, Susskind, & Anderson (2009) establish that (1) responses to bitterness are marked by activation of the levator labii muscle ‘which raises the upper lip and wrinkles the nose’; (2) bitter responses are made not just to bitter tastes but also to ‘photographs of uncleanliness and contamination-related disgust stimuli, including feces, injuries, insects, etc.’; and (3) in a dictator game, ‘objective (facial motor) signs of disgust [are] proportional to the degree of unfairness they experienced.’ ↩︎
Heyes (2023) offers a dual-process theory, and our conjecture is close to her view. Where we focus on norm-like patterns, Heyes introduces a notion of ‘implicit normativity’. On her view, implicit normativity is a consequence of implicit processes. These implicit processes are habitual processes (‘Acting on implicit processes alone, a person will comply with and enforce any behaviour that is common and has yielded positive outcomes in their recent experience. It does not matter whether, in the long term, the behaviour promotes or interferes with their own well-being or the thriving of their narrow social group – the people with whom they interact – or of the wider society in which they live. Implicit processes also allow an individual’s compliance and enforcement behaviour to change relatively rapidly with experience.’ p. 52). As Heyes stresses, this implies that implicit normativity is limited to what has recently been experienced as rewarding. So it cannot explain norm-like patterns which were rewarding to conform to a long time ago only; it may also have difficulty identifying rewards for norm-like patterns such as effort-matching, where conformity produces lower rewards than nonconformity would. ↩︎