Does the Automaticity in Human Behaviour Undermine Human Moral Responsibility?

Simon Kittle
Tuesday 14 June 2022
  1. Introduction
  2. Fields of Study
  3. Discussion
  4. Conclusion

1. Introduction and Hypothesis

Christianity holds that people are morally accountable to God for at least some of their actions. This moral accountability is central to Christian teaching because God is the pinnacle of moral goodness, and the capacity for moral action is one way in which humans are thought to image God.

In both common sense thinking and philosophical theorising about moral responsibility it is widely accepted that there is a control or freedom condition on moral responsibility. Call this the Responsibility-Requires-Control view. According to this view, if something is outside someone’s control then one cannot be responsible for it.[1] The Responsibility-Requires-Control view is explanatorily powerful; it provides a straightforward account of why human beings cannot be held responsible for (e.g.) their time and place of birth, why someone who (say) acts badly while in the grip of a strong negative emotion is usually considered less responsible than someone who acts badly after careful planning, and much more.

If Responsibility-Requires-Control is true, then, since it is possible that the sciences of human behaviour and action may reveal that certain aspects of human behaviour and action lie beyond the person’s control, it is possible that these sciences might reveal that some aspects of a person’s behaviour lie beyond their moral responsibility. Some thinkers have made bold claims in this regard. Sam Harris, for instance, makes the apparently universal claim that our “[t]thoughts and intentions emerge from background causes of which we are unaware and over which we exert no conscious control” with the result that “[w]e do not have the freedom we think we have” (Harris 2012, 5; cf. Haggard 2005, 291; Banks 2006, 237). There is little evidence for such a universal scepticism about conscious control; nevertheless, human freedom and moral responsibility could potentially be undermined by empirical findings that found a significant level of automaticity in human behaviour.

Automaticity refers to those aspects of a person’s choice, behaviour, or action which are initiated or produced without the conscious subject’s active involvement. Now, not all instances of automaticity so defined threaten agential control. Indeed, an agent’s conscious control is subserved by, and necessarily dependent on, an array of automatic processes. When I pick up a mug, for example, the precise trajectory of my arm and the grip which my hand forms around the mug are not under my direct conscious control; automatic, unconscious processes tailor my movements to the mug in question with only minimal awareness on my part. This is no threat to the control I exercise when I decide to pick up the mug because my decision doesn’t pertain to such details. Sometimes, however, unconscious processes yield an explanation of human behaviour which conflicts with the explanation offered by the conscious subject. Such cases challenge accounts of moral responsibility committed to Responsibility-Requires-Control.

The hypothesis to be considered here is that certain results from psychology and neuroscience provide reason to think that some of those aspects of human behaviour which we ordinarily think of as under our control are in fact produced by automatic processes in a manner which undermines the person’s moral responsibility.

2. Fields of Study

2.1 Systematic Theology

Christianity holds that people are morally accountable to God for at least some of their behaviour. Indeed, a prominent stream in traditional Christian theology has it that “God is a king-like ruler of the universe, [and] the promise of God’s punishment or reward after death is the foundation of morality” (Gottlieb 2017, 117). The Catechism of the Catholic Church states that:

Sin is an offense against God…disobedience, a revolt against God… (Catholic Church 2019, 453)

… each will be rewarded immediately after death in accordance with his works and faith (Catholic Church 2019, 266).

John Frame summarises the teaching of the Reformed stream of Protestant thought like so:

In Scripture, human beings are clearly responsible … to God as the supreme evaluator of human conduct. (Frame 2002, 119)

Human responsibility (accountability) means that human beings are subject to God’s evaluation… (Frame 2002, 119)

This is the dominant Protestant position. Greg Boyd, a theologian opposed to Frame’s position on many points, writes:

We are commanded and equipped by our Creator to love everyone as well as to lovingly care for the whole animal and plant kingdom … for which we are to some extent morally responsible. (Boyd 2001, 165)

Richard Swinburne, prominent philosopher of religion and member of the Eastern Orthodox church, also concurs:

Man’s dependence on God is so total that he owes it to him to live a good life. (Swinburne 1989, 124)

This moral accountability before God is no incidental part of Christian teaching. First, Christianity affirms that God is perfectly good. And as Michael Murray and Michael Rea explain, theists have typically taken God’s perfect goodness to involve God’s moral perfection, meaning that God is “morally unsurpassable … maximally loving and benevolent towards every created thing” (Murray and Rea 2008, 26). Jeffrey Jordan considers it unproblematic to define theism as “the proposition that there exists an all powerful, all knowing, morally perfect being” (Jordan 2011, 120; my emphasis). For many (if not most?) Christian thinkers, then, God’s perfect goodness includes the claim that God is morally perfect (Plantinga 1974; Adams 1999, 14; Baggett and Walls 2011, 132; O’Brien 2011; for a survey, see Garcia 2011).[2] Moreover, as Laura Garcia states, “[f]or most theists, moral perfection is the attribute most essential to a divine being” (2011, 218; my emphasis).

Second, Christian teaching, drawing from its Jewish roots, teaches that humans are made in the image of God. And one of the main ways that humans image God is in their capacity for moral action. Keith Ward writes:

Our distinctive human capacities, implanted in us by God, are our main clues to the sorts of lives we should seek to lead, and the sorts of ideals at which we should aim. Those capacities should themselves be images, however partial and remote, of the character of God. (Ward 2017, 144)

Christian teaching includes, of course, the claim that the image of God in humans has been marred by sin. But this does not negate the point that humans are accountable before God. Hans Schwarz (2013, 214–215) contends that the majority Christian teaching is that sin has tarnished but not obliterated the image of God in human beings (cf. Brunner 1965, 136). If this is right, then even fallen humans, with their compromised agency and associated inabilities, are capable of bearing some moral responsibility.

This highly condensed survey illustrates that it would be a serious problem for Christian theology if empirical findings showed that humans do not have the control required for moral responsibility.

2.2 Psychology

There is evidence from a range of subdisciplines in psychology which seems to show that aspects of human cognition, behaviour, and action are produced by automatic processes of which the agent is unaware. This includes work on cognitive dissonance (Festinger 1964), cognitive illusions and the systematic errors they lead to (Pohl 2017), the self-attribution of motivation and affect states (Nisbett and Wilson 1977), the bystander effect (Latané and Darley 1970), false memory attribution and the misinformation effect (Loftus (2005) and Laney and Loftus (2013) are helpful reviews), as well as work on priming, implicit bias, and similar phenomena (Brownstein and Saul 2016a; Brownstein and Saul 2016b). I focus here on two fields which demonstrate the pervasiveness of confabulation.

Confabulation is where someone makes up – confabulates – an explanation for why they behaved in a certain way (see Hirstein (2005) for an introduction). The confabulated explanation is not a genuine explanation; it is a post-hoc fabrication which the person offers apparently in order to demonstrate, to themselves or others, that their own behaviour makes sense. Crucially, the agent does not realise that the explanation they’ve offered is a confabulation—they think their explanation accurately explains why they behaved as they did.

Although originally studied in various pathologies (Hirstein 2009), it is now recognised that confabulation is pervasive in everyday life (Wheatley 2009). Timothy Wilson and Richard Nisbett’s (1978) famous study on the behaviour of shoppers exemplifies this. Wilson and Nisbett had shoppers evaluate items of clothing laid out on a table. In one condition, shoppers faced four identical pairs of nylon stockings, though they weren’t told the stockings were identical. Wilson and Nisbett asked the shoppers to pick the best pair of stockings; the shoppers were then asked why they had chosen as they did. The shoppers offered various reasons for their choice (softness, stretchiness, etc). None of them, however, mentioned the position of the stockings on the table. Yet the right-most pair of stockings were preferred to the left-most pair by a factor of four to one. When asked directly whether the position of the stockings influenced their choice, all but one denied that the position could have affected their decision (Wilson and Nisbett 1978, 124). This suggests that (a) in many cases, people were (without realising it) choosing the right-most pair of stockings simply because of their position on the table, and (b) in such cases, the reasons given by the shoppers were fabrications. In a replication of the Wilson and Nisbett study, A. W. Kurglanski et al. (2005) found that the preference for the right-most object was positively correlated with how much time pressure the participants were put under. In another replication, Anton Kühberger et al. (2006) found a preference for the right-most item emerged when the identical items were placed 70cm apart but not when they were placed 1cm apart. What’s significant for our purposes, however, is that the explanation offered by the subject is a confabulation.

Another field of research which demonstrates the presence of confabulation in everyday life is the work on false memories. One of the classic works in this field is Elizabeth Loftus and John Palmer’s (1974) investigation into the role that the wording of post-event questioning can have on memory formation. In one experiment, Loftus and Palmer had participants watch a short video of a car accident. The participants were then asked about the event, but the wording of the questions varied slightly among participants. Loftus and Palmer found that the question, “‘About how fast were the cars going when they smashed into each other?’ elicited higher estimates of speed than questions which used the verbs collided, bumped, contacted, or hit in place of smashed” (Loftus and Palmer 1974, 585). Similarly, when they tested participants for recall a week later, participants who received the verb ‘smashed’ were more likely to report that they saw broken glass in the video, although there was no broken glass in the video. The Loftus and Palmer study suggests that both initial memory formation and later recall are generative or constructive processes subject to influence by a variety of incidental factors.

Since Loftus and Palmer’s seminal study, a large amount of research has corroborated the claim that human memory is a (re-)constructive process. As Cara Laney and Loftus put it in a review article:

Studies conducted in laboratories around the world have demonstrated that human memory is susceptible to errors as a result of exposure to post-event information such as leading questions and reports of others, contact with other people, intentional suggestions, and even tiny differences in language. (Laney and Loftus 2013, 138)

The work on false memory attribution is relevant to the concern of the present article because it shows not merely that automatic cognitive processes of which the subject is unaware contribute to the formation of the person’s memories, but that the processes in question are sometimes influenced by factors which, from a normative point of view, they shouldn’t be. More recent studies have shown that people come to form preferences and alter their behaviour based on the formation of false memories, and these systematic inaccuracies produced by the mechanisms of human memory transfer over to the subject’s own explanation of their behaviour and action.

2.3 Neuroscience

In neuroscience, the studies most frequently discussed in relation to the topic of automaticity are Benjamin Libet’s well-known experiments, though there are other important bodies of work which may point in a similar direction, such as David Milner and Melvyn Goodale’s (2006) work on the visual system. In Libet et al.’s (1983) study, participants were connected to an EEG machine to measure brain activity and an ECG machine to measure muscle movement of the right arm. Participants were asked to flex their right wrist at a moment of their choosing while watching a specially designed clock which allowed them to report when they felt the urge or made the decision to flex. Libet aligned the EEG measurements, the ECG measurements, and the participant’s verbal report of the time of decision. He found that the verbal report of the time of urge or decision preceded the onset of muscle movement by around 200ms—no real surprise. But Libet also found that a distinctive readiness potential preceded the time at which the participant reported making the decision by around 350ms. Libet interpreted this to mean that the decision about when to flex had “already been initiated cerebrally” before the person became aware of the decision (Libet et al. 1983, 623). This implied that the person’s conscious choice did not, contrary to what we ordinarily think, initiate their flexing. In a meta-analysis of 37 Libet-style studies covering 43 experiments, Moritz Braun et al. (2021) found a temporal pattern of brain activity “largely consistent” with the one Libet reported; they also note, however, that there is “a high degree of uncertainty associated with the onset of [the unconscious] brain activity” and so conclude that aspects of Libet’s findings remain “fragile” (Braun et al. 2021, 196).

The experimental paradigm that Libet pioneered has been extended by numerous neuroscientists. Chun Siong Soon et al. (2008) adapted the basic setup to use fMRI instead of EEG. They were able to predict with 59.5% accuracy what decision a person would make around seven seconds before the person reported making the decision. Soon et al.’s study is significant because the timeframe involved looks much more problematic given a focus on free will. In Libet’s study, where the timeframes are on the order of hundreds of milliseconds, one might suggest that these timeframes are within the margin of error of a person’s ability to accurately report when they made the decision to flex, and in this way dismiss Libet’s results.[3] Clearly, however, this response is unavailable when the timeframe is around seven seconds.

Uri Maoz et al. (2012) conducted a Libet-inspired study where patients played a head-to-head betting game against the experimenter. The patients involved were suffering from severe epilepsy and as part of their treatment had electrodes implanted deep in the brain. This presented an opportunity to study decision-making using electrical readings taken directly from the brain rather than via EEG or fMRI. In Maoz et al.’s study, the brain readings were fed into a computer in real-time which displayed the predicted decision on a display behind the patient’s head. The system produced a prediction of what the patient would do 500 milliseconds before the “bet now” signal was shown to both players. Maoz et al. conducted the study on two patients. With the first patient, they were able to predict with 72±2% accuracy which bet the patient would make; with the second patient, they achieved 63±2% accuracy. These predictions concerned decisions in a competitive game in which small amounts of money were at stake and so suggest that Libet-style predictions can be made about decisions that the subject considers meaningful.

3. Discussion

3.1 Systematic Theology

Whether the findings outlined above undermine human morally accountability before God depends on (a) whether Responsibility-Requires-Control is true and (b) whether the empirical results surveyed provide good evidence for the absence of agential control. In this section I give two reasons for thinking Responsibility-Requires-Control is true.

First, Responsibility-Requires-Control is a highly intuitive principle supported by pre-theoretical intuitions and everyday moral practices. When I step on your foot, you blame me; when I step on your foot because someone pushes me over, you do not blame me. Why? Because the latter is not an instance of my agency, not something I control.[4] The refusal to praise or blame people for things outside of their control is pervasive. When someone is born into a family with plenty of material resources, we consider the person fortunate, not morally praiseworthy; when someone is born into severe poverty, we consider the person unlucky, not morally blameworthy. We neither praise nor blame people for their genetics, for their first language, for the colour of their eyes or shape of their nose, and so on. Nor do we praise or blame people for biological processes internal to their person – such as a person’s digestive processes, or processes which lead the person to get a certain disease – which cannot be consciously controlled. Apparent exceptions to this idea only prove the underlying point. If someone knows that smoking causes cancer but nevertheless smokes forty cigarettes a day,[5] we might be inclined to say that the person bears some responsibility for getting lung cancer. But even if correct, blame would be appropriate here only if the person could have decided not to smoke—a decision which bestows indirect control over whether they got cancer. The same basic point holds when uncontrolled neurological processes affect a person’s conscious thoughts or experiences. We do not blame Capgras’ syndrome patients for coming to believe that their closest relatives are imposters; we recognise that these beliefs are generated by neurological processes which cannot be consciously modulated—and for precisely this reason we deem blame inappropriate.

Second, most theorists attempting a systematic account of moral responsibility and who thus attend to a wide variety of cases and considerations have concluded that the pre-theoretical intuitions are correct on this point. Traditionally, responsibility was tied by both compatibilists (Moore 1912; Ayer 1977; Vihvelin 2013) and incompatibilists (Reid 1788; O’Connor 2000) to the presence of alternative possibilities. But control remains the focus even of those accounts which reject the need for alternatives. Harry Frankfurt’s (1971) hierarchical account requires that the agent’s first-order desire be causally efficacious—i.e. a source of control. John Martin Fischer and Mark Ravizza’s (1998) reasons-responsive account of moral responsibility focuses on the notion of guidance control; Robert Kane (1996) develops the notion of plural voluntary control; and Dana Nelkin articulates a type of control associated with being able to act for good reasons. And as Nelkin writes, “[i]n general, when we find out that someone lacks a capacity for recognizing reasons or cannot control her behavior in light of them, we are tempted to excuse that person for her actions” (Nelkin 2011, 8; my emphasis). So there are both pre-theoretical intuitive reasons and theoretical reasons to accept Responsibility-Requires-Control.

3.2 Psychology

Confabulation undermines the explanation of human action given by the subject because it renders the reasons explanation offered by the subject causally irrelevant to the action performed, therefore implying that the agent’s reasons-involving mental states played no role in controlling the action. If, unbeknownst to myself, I pick the right-most pair of nylon stockings because they are nearest to my hand, then the reason I cite for my choice – the softness, stretchiness, or whatever – does not play the causal role needed to secure my agential control. And if Responsibility-Requires-Control is true, then confabulation undermines moral responsibility as well.

The seriousness of the threat to human responsibility posed by confabulation depends on (a) whether the evidence for confabulation does indeed support its existence, (b) the nature of the mechanisms underlying confabulation, and (c) how widespread confabulation is.

The existence of confabulation is not in serious doubt; as outlined above, evidence for confabulation comes from a range of subdisciplines within psychology and the other sciences of behaviour. It’s true that some celebrated studies implying the existence of confabulation have (in)famously failed to replicate. For instance, Stéphane Doyen et al. (2012) were unable to replicate John Bargh et al.’s (1996) highly publicised study purporting to show the effects of priming for old-age on goal-setting behaviour. But critics should proceed with caution: Doyen et al. did find evidence of behavioural priming, just not the type which Bargh had claimed to find. Moreover, as David Meyer (2014) has noted, other types of priming, such as semantic priming, which may under certain circumstances contribute to confabulated explanations, are supported by a robust body of evidence, and the neurological processes underlying them are increasingly well-understood. There is also a robust evidence base for false memory attributions and the self-(mis)attribution of certain emotional states. So the prospects for removing the threat to responsibility by maintaining that confabulation is an unsupported phenomenon are dim; certainly, one cannot simply dissolve the worry by pointing to the replication crisis which came to light in 2012.[6]

So the precise nature of the threat that confabulation poses to responsibility hinges on (b) and (c): the nature of the mechanisms which produce the confabulation; and how widespread confabulation is. There is an active research programme investigating the mechanisms underlying confabulation and there is ongoing theorising drawing from this expanding evidence base (Hirstein (2005, Ch 1) discusses the major contours of the debate).

More important for our concerns is the question of scope: how widespread is confabulation? Some authors impressed by the evidence on unconscious cognition suggest, or come close to suggesting, that all explanations of human behaviour in terms of subjectively accessible propositional attitudes are confabulations (Doris 2015; Bargh 2017).[7] John Doris accepts that people deliberate a great deal, that choosing for reasons seems to bestow control, and that this experience is compelling from a first-person point of view (Doris 2015, 78–80). He explains (away) the apparent efficacy of these first-person accounts by noting that for confabulations to be useful, the mechanisms which generate them would have to be highly constrained such that they usually produce plausible explanations (Doris 2015, 97): if they didn’t, the resulting confabulations wouldn’t allow the person to make sense of their behaviour. But given this requirement, we should, Doris says, consider it no surprise that first-person explanations appear to the subject to be so compelling.

Now, when any given piece of behaviour is taken in isolation, it may be possible, as Doris suggests, to make plausible the idea that the subject’s own reasons-involving explanation of that behaviour is a confabulation. However, when a person’s behaviour is considered as a whole, the idea of universal confabulation should strike us as incredible. Doris’s theory fails to account for the many stretches of complex activity which depend for their existence – not to mention their success – on the accurate sequencing and successful performance of component actions. Plausibly, the only way to explain such complex activity is by appeal to the agent’s subjectively accessible reasons-involving propositional attitudes. For instance, even the modest activity of writing a coherent paragraph of text – writing a sentence, re-reading it back, asking oneself if it accurately captures what one intends to say (i.e. introspecting), perhaps asking oneself if the text matches one’s grammatical aspirations (more introspection), re-reading again, writing a sentence which follows on from the previous – surely falsifies any theory of global confabulation. And this is to say nothing of the power of beliefs, desires, and intentions to explain an agent’s long-term projects like writing an article, building a garden shed, or directing a company. The explanation for why I have spent many mornings working on this article is that, since I desire to learn about human agency and believe that writing about it is a great way to do that, I formed an intention to write an article. That intention has for weeks tethered and guided my behaviour around that goal in myriad ways. It’s scarcely plausible that such an explanation is a confabulation (cf. Arpaly 2018, 754–755).

A more reasonable hypothesis is that the majority of our reasons-based explanations are exactly what they seem – explanations which cite mental states that cause and facilitate control over our behaviour – but that confabulation systematically occurs under various edge-case conditions (as when, for instance, someone is unknowingly presented with four identical items and pressed to distinguish between them). If this is right, then since confabulation indicates lack of control, we should modulate attributions of responsibility when it occurs. It will thus be important for all scholars writing on responsibility – including theologians – to consult the literature to determine just when confabulation does and doesn’t occur.

3.3 Neuroscience

One frequently voiced objection to Libet’s studies is that they concern inconsequential, non-moral decisions with limited ecological validity (Roskies 2011, 18–19; Messer 2017, 77). A similar complaint is that subjects in the Libet task have no reason to favour one moment to flex than another (Mele 2009, 83). The idea here is that since the mechanisms which produce inconsequential decisions are likely to differ from those involved in meaningful (especially moral) decisions, it is illegitimate to apply insights from Libet’s results to meaningful, moral decisions. But while it’s correct that different types of decision may turn out to be subserved by or correlated with different neurological processes, this objection is nevertheless unconvincing.

First, many of our everyday decisions are morally neutral and inconsequential, yet we still think we control the unfolding of these parts of our lives by deciding what to do. It would be surprising, and troubling, if, despite how things appear to us, these decisions were not freely made.

Second, the Maoz et al. study discussed above demonstrates that Libet-style findings can be achieved for consequential decisions which the person cares about. It seems reasonable to conclude that, as technology improves, the ecological validity of Libet-style studies will continue to improve. And if the Maoz et al. study is any indication, such improvements will preclude appeal to this objection.

Another objection is that the readiness potential might be a necessary but not sufficient condition for the subsequent decision. One version of this thought suggests that the readiness potential might be something like an inclination or a “slight bias” to one of the options (Mele 2014, 30). Most theorists contend that any viable account of free will should be consistent with decisions being causally influenced by prior factors.[8] However, even if correct, the following point is worth noting: if the readiness potential is necessary but not sufficient for the decision, and if the readiness potential is itself uncontrolled, then this still imposes a limitation on the control which the agent can exercise. Suppose, for instance, that when choosing between A and B, in order to be able to choose A, the readiness potential RP1 must first occur. If so, then one’s being able to choose A will be limited to contexts where RP1 has occurred—but such RP1 contexts are themselves beyond one’s control. That seems problematic inasmuch as we ordinarily think that we are free to direct our thoughts in almost any direction.

Some have also challenged the usefulness of asking the subject to report on the timing of their felt urge to move. There are several versions of this objection. One is that, given that all conscious experiences themselves take time to generate, subjects will be incapable of achieving the required accuracy. Another makes the conceptual point that an urge and an experience of an urge are different entities (Holton 2004, 220; Mele 2009, 32), such that the subjects’ report of the timing of the latter doesn’t identify the timing of the former. Some recent studies have attempted to address this reliance on the subjects’ own reporting of the timing (Matsuhashi and Hallett 2008), but the conceptual issues have been harder to resolve. Moreover, there are deeper conceptual worries with the Libet-paradigm. In Libet’s original study, he referred to the subject’s urge, desire, will, and decision to flex interchangeably (e.g. Libet 1985, 530). But these items of mental furniture are not equivalent. An urge is something that one experiences (and may be entirely passive), whereas a decision is something one does (an action). Given these conceptual points, it cannot be ruled out that the Libet-style studies establish, at most, that an urge necessarily precedes a decision (something which is not obviously a threat to free will).

Another serious methodological problem concerns Libet’s use of back-averaging. In short, the problem is this: to get usable data, Libet-style studies have to run many trials for each participant, the results of which are then averaged. This requires a marker which can be used to temporally align data from the different trials. For this, Libet used the onset of the muscle movement. As a result, the prior brain activity is recorded from the EEG only if wrist muscle movement is detected by the ECG. This means that “no one knows” whether similar readiness potentials occur in those cases when the person chooses not to move, or when the person is instructed to wait for the urge to flex and then veto the movement—in such cases, the brain activity simply isn’t recorded (Mele 2009, 74). It is therefore not possible to claim that Libet’s studies provide conclusive evidence that the readiness potential is sufficient for the occurrence of the muscle movement. But it is precisely the sufficiency claim which undergirds the conclusion that the conscious urge or decision bestows no agential control.

As will be clear from this survey, the Libet-paradigm has been subject to a wide array of criticisms.[9] The conceptual point about urges being distinct from decisions and the methodological point concerning the use of back-averaging seem especially damaging to the conclusions Libet and others have sought to draw from these results. As a result, it appears safe to agree with Joshua Shepherd that “nothing in the Libet tradition suggests that consciousness is irrelevant to decision-making” (Shepherd 2015, 202). Nevertheless, caution is sensible here. Although extant Libet-style studies do not yet support a sceptical conclusion about free will, it may be that, suitably extended, the paradigm will in the future reveal limits to agential control.

4. Conclusion

The hypothesis under consideration is that empirical results provide reason to think that some aspects of human behaviour which we ordinarily think of as under our conscious control are in fact produced by automatic processes in a manner which undermines the person’s moral responsibility. The evidence outlined suggests that this hypothesis is confirmed with respect to behaviour and action where the subject engages in confabulation. Since it is not plausible that the majority of first-person explanations are confabulations, the scope of this worry is limited, though theologians and others who theorise about moral responsibility should modulate responsibility attributions in the relevant cases. Moreover, theologians should note that there is potential for a more wide-ranging indirect challenge to responsibility deriving from the evidence that supports a view of human memory according to which it is an essentially constructive process. The challenge to responsibility from the Libet-tradition is not currently borne out by the evidence, though continued advances in technology may change this situation.[10]


Adams, Robert M. 1999. Finite and Infinite Goods: A Framework for Ethics. USA: Oxford University Press.

Arpaly, Nomy. 2018. “Comments on Talking to Our Selves by John Doris”. Philosophy and Phenomenological Research 97: 753–757. doi:10.1111/phpr.12547.

Ayer, A. J. 1977. “Freedom and Necessity”. In The Philosophical Imagination, edited by Raziel Abelson, Marie-Louise Friquegnon and Michael Lockwood. New York: St. Martin’s Press.

Baggett, David, and Jerry L. Walls. 2011. Good God: The Theistic Foundations of Morality. New York: Oxford University Press.

Banks, William P. 2006. “Does Consciousness Cause Misbehaviour?”. In Does Consciousness Cause Behavior?, edited by Susan Pockett, William P. Banks and Shaun Gallagher, 235–256. Cambridge, MA: MIT Press.

Bargh, John A. 2017. Before You Know It: The unconscious reasons we do what we do. New York: Touchstone.

Bargh, John A., M. Chen, and L. Burrows. 1996. “Automaticity of Social Behavior: Direct Effects of Trait Construct and Stereotype-Activation on Action”. Journal of Personality and Social Psychology 71: 230–244. doi:10.1037//0022-3514.71.2.230.

Boyd, Gregory A. 2001. Satan and the Problem of Evil: Constructing a Trinitarian Warfare Theodicy. Downers Grove Ill.: InterVarsity Press.

Braun, Moritz N., Janet Wessler, and Malte Friese. 2021. “A meta-analysis of Libet-style experiments”. Neuroscience and biobehavioral reviews 128: 182–198. doi:10.1016/j.neubiorev.2021.06.018.

Brownstein, Michael, and Jennifer M. Saul, eds. 2016a. Implicit Bias and Philosophy: Volume 1. Oxford United Kingdom: Oxford University Press.

—, eds. 2016b. Implicit Bias and Philosophy: Volume 2. Oxford United Kingdom: Oxford University Press.

Brunner, Emil. 1965. Man in Revolut: A Christian Anthropology. Philadelphia: The Westminster Press.

Catholic Church. 2019. Catechism of the Catholic Church. Vatican City, Washington, D.C.: Libreria Editrice Vaticana; [Distributed by] United States Conference of Catholic Bishops.

Doris, John M. 2015. Talking to Our Selves: Reflection, Ignorance, and Agency. Oxford: Oxford University Press.

Dougherty, Trent A. 2014. The Problem of Animal Pain. London: Palgrave Macmillian.

Doyen, Stéphane, Olivier Klein, Cora-Lise Pichon, and Axel Cleeremans. 2012. “Behavioral Priming: It’s All in the Mind, but Whose Mind?”. PloS one 7: e29081. doi:10.1371/journal.pone.0029081.

Festinger, Leon. 1964. Conflict, Decision and Dissonance. Stanford: Stanford University Press.

Fischer, John M., and Mark Ravizza. 1998. Responsibility and Control: A Theory of Moral Responsibility. Cambridge: Cambridge University Press.

Frame, John M. 2002. The Doctrine of God. Phillipsburg, N.J.: P & R Pub.

Frankfurt, Harry G. 1971. “Freedom of the Will and the Concept of a Person”. The Journal of Philosophy 68: 5–20.

Garcia, Laura. 2011. “Moral Perfection”. In The Oxford Handbook of Philosophical Theology, edited by Thomas P. Flint and Michael C. Rea, 217–238. Oxford, New York: Oxford University Press.

Gottlieb, Anthony. 2017. The Dream of Enlightenment: The Rise of Modern Philosophy. New York and London: Liveright Publishing Corporation.

Haggard, Patrick. 2005. “Conscious Intention and Motor Cognition”. Trends in Cognitive Sciences 9: 290–295.

Harris, Sam. 2012. Free Will. New York: The Free Press.

Hirstein, William. 2005. Brain Fiction: Self-deception and the Riddle of Confabulation. Cambridge, MA: MIT Press.

—. 2009. “Confabulation”. In The Oxford Companion to Consciousness, edited by Tim Bayne, Axel Cleeremans and Patrick Wilken, 174–177. Oxford, New York: Oxford University Press.

Holton, R. 2004. “Review of Daniel Wegner’s The Illusion of Conscious Will”. Mind 113: 218.

Jordan, Jeffrey J. 2011. “Pascal and the Wager”. In Philosophy of Religion: The Key Thinkers, edited by Jeffrey J. Jordan, 118–136. London, New York: Continuum.

Kane, Robert H. 1996. The Significance of Free Will. New York: Oxford University Press, USA.

Kittle, Simon. In press. “Free Will and Cognitive Biases”. In Free Will and Theology, edited by Olli-Pekka Vainio and Aku Visala. Routledge (TBC).

—. ms, “The Incompatibility of Free Will and Prior Objective Probabilities”.

Kruglanski, A. W., Chun, W. Y., Sleeth‐Keppler, D., & Friedman, R. S. 2005. “On the psychology of quasi‐rational decisions: The multifinality principle in choice without awareness”. Advances in Consumer Research 32: 331–332.

Kühberger, Anton, Christoph Kogler, Angelika Hug, and Evelyne Mösl. 2006. “The Role of the Position Effect in Theory and Simulation”. Mind & Language 21: 610–25. doi:10.1111/J.1468-0017.2006.00290.X.

Laney, Cara, and Elizabeth F. Loftus. 2013. “Recent advances in false memory research”. South African Journal of Psychology 43: 137–146. doi:10.1177/0081246313484236.

Latané, Bibb, and John M. Darley. 1970. The Unresponsive Bystander: Why doesn’t he help? New York: Meredith Corporation.

Libet, Benjamin. 1985. “Unconscious cerebral initiative and the role of conscious will in voluntary action”. Behavioral and Brain Sciences 8: 529. doi:10.1017/S0140525X00044903.

Libet, Benjamin, Curtis A. Gleason, Elwood W. Wright, and Dennis K. Pearl. 1983. “Time of Conscious Intention to Act in Relation to Onset of Cerebral Activity (Readiness-Potential)”. Brain 106.

Loftus, Elizabeth F. 2005. “Planting misinformation in the human mind: A 30-year investigation of the malleability of memory”. Learning & memory 12: 361–366. doi:10.1101/lm.94705.

Loftus, Elizabeth F., and John C. Palmer. 1974. “Reconstruction of automobile destruction: An example of the interaction between language and memory”. Journal of Verbal Learning and Verbal Behavior 13: 585–589. doi:10.1016/S0022-5371(74)80011-3.

Maoz, Uri, S. Ye, I. Ross, A. Mamelak, and C. Koch. 2012. “Predicting Action Content On-Line and in Real Time before Action Onset – an Intracranial Human Study”. In Advances in Neural Information Processing Systems, edited by P. Pereira et al. Cambridge, MA: MIT Press.

Matsuhashi, Masao, and Mark Hallett. 2008. “The timing of the conscious intention to move”. The European journal of neuroscience 28: 2344–2351. doi:10.1111/j.1460-9568.2008.06525.x.

Mele, Alfred R. 2009. Effective Intentions. Oxford: Oxford University Press.

—. 2014. Free: Why Science hasn’t Disproved Free Will. Oxford: Oxford University Press.

Messer, Neil. 2017. Theological Neuroethics: Christian Ethics Meets the Science of the Human Brain. London and New York: Bloomsbury T&T Clark an imprint of Bloomsbury Publishing PIc.

Meyer, David E. 2014. “Semantic Priming Well Established”. Science 345, no. 6196: 523. doi:10.1126/science.345.6196.523-b.

Mill, John S. 1878. An Examination of Sir William Hamilton’s Philosophy. London: Longmans, Green, Reader, and Dyer.

Milner, A.D, and M.A Goodale. 2006. The Visual Brain in Action. Oxford University Press, USA.

Moore, G. E. 1912. Ethics. London: Humphrey Milford; Oxford University Press.

Murray, Michael J., and Michael C. Rea. 2008. An Introduction to the Philosophy of Religion. Cambridge and New York: Cambridge University Press.

Nelkin, Dana K. 2011. Making Sense of Freedom and Responsibility. Oxford University Press.

Nisbett, R.E, and Timothy Wilson. 1977. “Telling more than we can know: Verbal reports on mental processes”. Psychological review 84: 231.

O’Brien, Dan. 2011. “God, Omniscience and Understanding Evil”. In God, Goodness and Philosophy, edited by Harriet A. Harris, 167–177. Farnham Surrey England, Burlington VT: Ashgate.

O’Connor, Timothy. 2000. Persons and Causes: The Metaphysics of Free Will. New York: Oxford University Press.

Plantinga, Alvin. 1974. God, Freedom and Evil. Grand Rapids, MI: William B. Eerdmans.

Pohl, Rüdiger, ed. 2017. Cognitive Illusions: Intriguing Phenomena in Thinking, Judgment and Memory. London, New York: Routledge Taylor & Francis Group.

Reid, Thomas. 1788. Essays on the Active Powers of Man. Edinburgh: John Bell and G. G. J. & J. Robinson.

Roskies, Adina L. 2011. “Why Libet’s Studies Don’t Pose a Threat to Free Will”. In Conscious Will and Responsibility, edited by Walter Sinnott-Armstrong and Lynn Nadel, 11–22. Oxford, New York: Oxford University Press.

Schwarz, Hans. 2013. The Human Being: A Theological Anthropology. Grand Rapids Michigan: William B. Eerdmans Publishing Company.

Shepherd, Joshua. 2015. “Scientific Challenges to Free Will and Moral Responsibility”. Philosophy Compass 10: 197–207. doi:10.1111/phc3.12200.

Sinnott-Armstrong, Walter, and Lynn Nadel, eds. 2011. Conscious Will and Responsibility. Oxford, New York: Oxford University Press.

Soon, Chun S., Marcel Brass, Hans-Jochen Heinze, and John-Dylan Haynes. 2008. “Unconscious determinants of free decisions in the human brain”. Nature Neuroscience 11: 543–545. doi:10.1038/nn.2112.

Swinburne, Richard. 1989. Responsibility and Atonement. USA: Oxford University Press.

Vicens, Leigh C. 2016. “Objective Probabilities of Free Choice”. Res Philosophica 93: 125–135. doi:10.11612/resphil.2016.93.1.8.

Vihvelin, Kadri. 2013. Causes, Laws, and Free Will: Why Determinism Doesn’t Matter. New York: Oxford University Press.

Ward, Keith. 2017. The Christian Idea of God. Cambridge: Cambridge University Press.

Wheatley, Thalia. 2009. “Everyday Confabulation”. In Confabulation, edited by William Hirstein, 205–223. New York: Oxford University Press.

Wilson, Timothy, and Richard E. Nisbett. 1978. “The Accuracy of Verbal Reports About the Effects of Stimuli on Evaluations and Behavior”. Social Psychology 41: 118. doi:10.2307/3033572.


[1] Things are more complicated than this because it is open to proponents of Responsibility-Requires-Control to say that a person may accept responsibility for something which they do not control. This need not be contrary to the principle underlying the view as long as it is held that a person’s act of acceptance is itself controlled in the requisite manner. Further consideration of this point is beyond the scope of this article.

[2] Some theologians reject this, claiming that God is good but not morally good. In brief, they argue that divine transcendence and unknowability make any idea of a moral community between God and humans incomprehensible. About this sort of view, John Stuart Mill asked “If in ascribing goodness to God … I do not mean the goodness of which I have some knowledge, but an incomprehensible attribute of an incomprehensible substance, which for aught I know may be a totally different quality from that which I love and venerate … what do I mean by calling it goodness? … Unless I believe God to possess the same moral attributes which I find, in however inferior a degree, in a good man, what ground of assurance have I of God’s veracity?” (Mill 1878, 128). I find attempts to answer this question wholly unconvincing and find myself in agreement with Trent Dougherty who has remarked that, to the extent to which he understands those views which deny that God is part of the moral order, he finds them “monstrous and something like blasphemous (libelous?) accounts of God” (Dougherty 2014, 17 fn. 2). As such, I will not consider these views any further here.

[3] Research on reaction times shows that this is not a strong objection to the original Libet-studies, but I leave this point aside.

[4] One can manufacture fanciful scenarios where someone’s pushing me over is an instance of my (indirect) agency in virtue of having been arranged by me—clearly such cases aren’t relevant to the point being made here.

[5] For completeness, we would need to add that the person has not been subject to circumstances (such as pervasive family influence, social pressure, or deeply traumatic life-experiences) that might make smoking reasonable in spite of the risks.

[6] For more on this point, see my Kittle (In press).

[7] Doris tends to limit the term ‘confabulation’ to cases of clinical pathology (Doris 2015, 78); but his view is that healthy individuals are subject to “pervasive self-ignorance” such that their explanations are best understood as attempts at rationalisation and sense-making rather than as reports as to how the agent controlled their behaviour (Doris 2015, Ch 6).

[8] There are some dissenters (see, e.g. Vicens 2016 and Kittle ms.).

[9] Mele (2009) is a sophisticated book-length assessment of Libet-style results and Sinnott-Armstrong et al. (2011) is an edited collection devoted Libet’s results. The latter includes contributions from both proponents and critics of the Libet-studies.

[10] This publication was made possible through the support of a grant from the John Templeton Foundation, God and the Book of Nature [Grant ID.  61507].  The opinions expressed in this publication are those of the author(s) and do not necessarily reflect the view of the John Templeton Foundation.

Cite this article

Kittle, Simon. 2022. “Does the Automaticity in Human Behaviour Undermine Human Moral Responsibility?” Theological Puzzles (Issue 8).

Reply to this Theological Puzzle

Disagree with the conclusions of this puzzle? Did the Author miss something? We encourage readers to reply via a ‘Note’ of up to 2000 words. Notes do not need to follow the puzzle structure. See Contribute for more information. An honorarium may be payable.

Contact the author

Simon Kittle 
Email: [email protected]

Share this story