Link Search Menu Expand Document

Appendix: Dual Process Theory and Auxiliary Hypotheses

The stripped-down dual process theory (see A Dual Process Theory of Ethical Judgement) requires at least one auxiliary hypothesis in order to relate it to available evidence. At least six auxiliary hypotheses have been proposed. Which should we accept?

This recording is also available on stream (no ads; search enabled). Or you can view just the slides (no audio or video). You should not watch the recording this year, it’s all happening live (advice).

If the video isn’t working you could also watch it on youtube. Or you can view just the slides (no audio or video). You should not watch the recording this year, it’s all happening live (advice).

If the slides are not working, or you prefer them full screen, please try this link.

The recording is available on stream and youtube.


This is covered in the lecture in 2023—4. You do not need this material, but you may wish to consider it if you are writing about dual-process theories of ethical judgement.

Proponents of dual process theories of ethical judgments tend to offer complex sets of hypotheses which go far beyond the core idea of a dual process theory and are not all equally well supported by evidence (for two examples, see Green and Kumar in A Dual Process Theory of Ethical Judgement).

Since the hypotheses that are associated with different proposals can be isolated and tested separately, combining them yields something more like a parlay bet than a theory.[1] Better to keep them separate.

What Is the Core Idea

Just this and nothing more.

Two (or more) ethical processes are distinct in this sense: the conditions which influence whether they occur, and which outputs they generate, do not completely overlap.

One process is faster than another: it makes fewer demands on scarce cognitive resources such as attention, inhibitory control and working memory.

Because much of the available evidence involves observations of just one response to a moral scenario (typically a verbal judgement[2]), additional, auxiliary hypotheses are needed to generate predictions linked to the currently available evidence. But which auxiliary hypothesis should we adopt?

What Are the Candidate Auxiliary Hypotheses?

At least six candidates for auxiliary hypotheses can be found in the existing literature (and there are probably many more):

id fast slow source
1. never consequentialist[3] sometimes consequentialist Greene (2014)
2. always deontological[4] sometimes not deontological Greene (2014)[5]
3. always only act-types sometimes not only act-types Cushman (2013)
4. never distal outcomes sometimes distal outcomes Cushman (2013)[6]
5. model-free model-based Cushman (2013)
6. affective cognitive Greene (2014)

UPDATE A further candidate auxiliary hypothesis might concern the distinction between personal and impersonal. Since this introduces even more problems than we will eventually have to face (McGuire, Langdon, Coltheart, & Mackenzie, 2009), it does not appear to be a good bet as things stand.

Which Auxiliary Hypothesis to Choose?

I propose that we use just one auxiliary hypothesis, which is a version of (4) above:

Only the slow process ever flexibly takes into account the more distal outcomes of an action.

This auxiliary hypothesis has two virtues. Together with the stripped-down dual process theory, it generates nearly all the predictions for which we currently have good evidence. And it cannot be significantly weakened while still doing this.

Ask a Question

Your question will normally be answered in the question session of the next lecture.

More information about asking questions.


automatic : As we use the term, a process is automatic just if whether or not it occurs is to a significant extent independent of your current task, motivations and intentions. To say that mindreading is automatic is to say that it involves only automatic processes. The term `automatic' has been used in a variety of ways by other authors: see Moors (2014, p. 22) for a one-page overview, Moors & De Houwer (2006) for a detailed theoretical review, or Bargh (1992) for a classic and very readable introduction
characteristically consequentialist : According to Greene, a judgement is characteristically consequentialist (or characteristically utilitarian) if it is one in ‘favor of characteristically consequentialist conclusions (eg, “Better to save more lives”)’ (Greene, 2007, p. 39). According to Gawronski, Armstrong, Conway, Friesdorf, & Hütter (2017, p. 365), ‘a given judgment cannot be categorized as [consequentialist] without confirming its property of being sensitive to consequences.’
characteristically deontological : According to Greene, a judgement is characteristically deontological if it is one in ‘favor of characteristically deontological conclusions (eg, “It’s wrong despite the benefits”)’ (Greene, 2007, p. 39). According to Gawronski et al. (2017, p. 365), ‘a given judgment cannot be categorized as deontological without confirming its property of being sensitive to moral norms.’
cognitively efficient : A process is cognitively efficient to the degree that it does not consume working memory and other scarce cognitive resources.
distal outcome : The outcomes of an action can be partially ordered by the cause-effect relation. For one outcome to be more distal than another is for it to be lower with respect to that partial ordering. To illustrate, if you kick a ball through a window, the window’s breaking is a more distal outcome than the kicking.
fast : A fast process is one that is to to some interesting degree cognitively efficient (and therefore likely also some interesting degree automatic). These processes are also sometimes characterised as able to yield rapid responses.
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.
outcome : An outcome of an action is a possible or actual state of affairs.
proximal outcome : The outcomes of an action can be partially ordered by the cause-effect relation. For one outcome to be more proximal than another is for it to be higher with respect to that partial ordering. To illustrate, if you kick a ball through a window, the kicking is a more proximal outcome than the window’s breaking.


Bargh, J. A. (1992). The Ecology of Automaticity: Toward Establishing the Conditions Needed to Produce Automatic Processing Effects. The American Journal of Psychology, 105(2), 181–199.
Cushman, F. (2013). Action, Outcome, and Value: A Dual-System Framework for Morality. Personality and Social Psychology Review, 17(3), 273–292.
Davidson, D. (1971). Agency. In R. Binkley, R. Bronaugh, & A. Marras (Eds.), Agent, action, and reason, (pp. 3–25). Toronto: University of Toronto Press.
Edwards, K., & Low, J. (2017). Reaction time profiles of adults’ action prediction reveal two mindreading systems. Cognition, 160, 1–16.
Gawronski, B., Armstrong, J., Conway, P., Friesdorf, R., & Hütter, M. (2017). Consequences, norms, and generalized inaction in moral dilemmas: The CNI model of moral decision-making. Journal of Personality and Social Psychology, 113(3), 343–376.
Greene, J. D. (2007). The Secret Joke of Kant’s Soul. In W. Sinnott-Armstrong (Ed.), Moral Psychology, Vol. 3 (pp. 35–79). MIT Press.
Greene, J. D. (2014). Beyond Point-and-Shoot Morality: Why Cognitive (Neuro)Science Matters for Ethics. Ethics, 124(4), 695–726.
Kahneman, D. (2002). Maps of bounded rationality: A perspective on intuitive judgment and choice. In T. Frangsmyr (Ed.), Le prix nobel, ed. T. Frangsmyr, 416–499. (Vol. 8, pp. 351–401). Stockholm, Sweden: Nobel Foundation.
Kahneman, D. (2013). Thinking, fast and slow. New York: Farrar, Straus; Giroux.
Kahneman, D., & Klein, G. (2009). Conditions for intuitive expertise: A failure to disagree. American Psychologist, 64(6), 515–526.
McGuire, J., Langdon, R., Coltheart, M., & Mackenzie, C. (2009). A reanalysis of the personal/impersonal distinction in moral psychology research. Journal of Experimental Social Psychology, 45(3), 577–580.
Moors, A. (2014). Examining the mapping problem in dual process models. In Dual process theories of the social mind (pp. 20–34). Guilford.
Moors, A., & De Houwer, J. (2006). Automaticity: A Theoretical and Conceptual Analysis. Psychological Bulletin, 132(2), 297–326.
Morewedge, C. K., & Kahneman, D. (2010). Associative processes in intuitive judgment. Trends in Cognitive Sciences, 14(10), 435–440.
Wel, R. P. R. D. van der, Sebanz, N., & Knoblich, G. (2014). Do people automatically track others’ beliefs? Evidence from a continuous measure. Cognition, 130(1), 128–133.


  1. This is too quick: all theories are bets. And I don’t have anything against parlay bets—if you are given good enough odds, they can be spectacular (and fun, of course). But making progress in understanding is like gambling in that it requires managing risk. Just as you have to allocate cash to bets carefully in order to minimize the risk of ruin, so in your research you have to allocate time carefully in order to avoid being left with nothing. (There is some research on how this is done that starts with the Kelly Criterion.) You should not put everything into a small number of especially risky bets. This is the true problem with Parlay bets. ↩︎

  2. Measures like response time, movement trajectory, proactive eye gaze or pupil dilation do not seem to have caught on in moral psychology despite being used to good effect in supporting dual process theories in other domains including social cognition (see Wel, Sebanz, & Knoblich, 2014 or Edwards & Low, 2017 for some examples). ↩︎

  3. In this context, the term ‘consequentialist’ is short for characteristically consequentialist. ↩︎

  4. In this context, the term ‘deontological’ is short for characteristically deontological. ↩︎

  5. Greene (2014) does not explicitly separate this and the first candidate, (1). But they are clearly distinct. ↩︎

  6. Cushman (2013) phrases the distinction in terms of actions and outcomes. This is potentially confusing if, like Davidson (1971), you think that actions are individuated by their outcomes. But it is clear from the discussion that Cushman’s distinction concerns a contrast between more proximal outcomes and more distal outcomes. Note also that Cushman does not explicitly distinguish this and the third candidate, (3); perhaps the distinction is too minor to consider. ↩︎