Psychological Research on Free Will Intuitions: A Critical Review

3. Phenomenal Abilities

Phenomenal Abilities: Incompatibilism and the Experience of Agency
Oisín Deery, Matt Bedke and Shaun Nichols
Oxford Studies in Agency and Responsibility Volume 1 2013, ed. D. Shoemaker, Oxford University Press, 126–50

Author's Abstract

Incompatibilists often claim that we experience our agency as incompatible with determinism, while compatibilists challenge this claim. This chapter reports a series of experiments that focus on whether the experience of having an ability to do otherwise is taken to be at odds with determinism. It was found that participants in the studies described their experience as incompatibilist whether the decision was (i) present-focused or retrospective, (ii) imagined or actual, (iii) morally salient or morally neutral. The only case in which participants did not give incompatibilist judgments was when the question was explicitly about whether one's ignorance of the future was compatible with determinism. This lends empirical support to claims made by incompatibilists about the experience of agency, while also showing that compatibilist accounts of ability are inadequate to the reported phenomenology. These results also inform recent debates about the presuppositions of deliberation.


Study 1: 84 US participants recruited online (participant demographics unreported)
Study 2: 155 US participants recruited online (participant demographics unreported)
Study 3: 106 US participants recruited online (participant demographics unreported)

The first part of the paper [Deery et al 2013: §1.2] gives a snapshot of phenomenological arguments for indeterministic free will put up by some important philosophers' and some arguments against. The authors then move on to correcting some methodological problems with earlier studies. On a positive note, in order to ensure understanding, the experimenters tested the subjects prior to their answering the questionnaires. This is a mark of good design and I commend this aspect of the protocol. Nonetheless, I found serious design errors with this study.

In Study 1, participants who agreed with a statement indicating that they could have steered left or right on an imaginary snow sled proceeded to a 'training' session on determinism (called 'causal completeness' in the study). In this session, participants were instructed with the statement that:

According to causal completeness, if we replayed the past right up until Obama's decision—including everything that was going through Obama's mind—then Obama would once again make exactly the same decision. That is, all the events leading up to Obama's decision (including everything that was going through Obama's mind), made it so that it had to happen that Obama would pick Biden.

[Deery et al 2013: 133]

My problem here is that this training statement is going beyond a strict explanation of determinism. It is elucidating a particular version of determinism (hard determinism) that is incompatibilist in nature. By saying Obama's picking Biden 'had to happen', this training statement is priming participants to interpret determinism as incompatibilism.

To test their comprehension of determinism, participants were then checked for agreement with the following statement:

According to causal completeness, if a week from now Barack Obama decides to have soda with dinner, all the events leading up to that decision will make it the case that he has to decide to have a soda with dinner.

[Deery et al 2013: 134]

Once again, the phrasing, 'he has to decide', is an incompatibilist rendering of an agent's voluntary choice. These two statements prime participants into interpreting determinism as taking away their ability to choose otherwise. That the authors chose these particular phrasings is baffling considering the authors discuss the real risk of priming participants in §4.1, while in §5.2 they explore the various compatibilist interpretations of the ability to 'do otherwise' that explicitly reject the notion that voluntary decisions 'had to happen'.

The experimenters [134] then asked participants whether determinism means that there is 'something mistaken' about their feeling that they could have gone left or right on the imaginary snow sled. The problem here is that this way of asking the question presumes that compatibilists think that there is 'something mistaken' in people's 'feeling' of being able to choose one way or the other. Far from denying that a snow sled rider 'could either go to the left or go to the right', compatibilists explain how this can be so.

Another serious misgiving I have about the design of the experiments is the priming fostered by the overall context of the study. The 'training' and the questions asked give the impression to subjects not trained in philosophy and language that the issue of whether agents could have done otherwise is bound up exclusively with the truth of complete event causation. The context of the study primes such novices to see the phrase, 'had to happen', as having one and only one modal interpretation.

In Free Will and Compatibilism [Allan 2016], I demonstrate how 'could have done otherwise' is open to two different modal interpretations, depending on the context. By focusing subjects exclusively on 'causal completeness', they are prone to follow the path of many first year philosophy students when initially introduced to the concept of determinism. I've seen the same situation in ethics and epistemology classes, where many novice students initially adopt radical views, such as universal skepticism, moral nihilism and relativism. It is only later, after students have honed their philosophical tools and refined their analytical thinking skills, that they adopt more nuanced approaches. For these reasons, I think the design of this study is deeply flawed. In a nutshell, the participant training and questionnaires prime the participants by explicitly presuming what the study intends to demonstrate empirically.

The results for Study 1 [135] indicated that participants tended to interpret their imagined experiential choices, past and present, as being incompatible with determinism. Interestingly, the study authors' prediction [131] that participants would be more incompatibilist about current imagined choices compared with past imagined choices was not borne out.

In Study 2, the authors sought to mitigate the limitations on the design of Study 1. They attempted this by making it also about actual moral choices and by rewording the statement referring to 'mistaken' feelings. The reworded statement given to participants is as follows:

Even though it felt like I could either choose to X or choose to Y, if causal completeness is true then I couldn't really have chosen differently than I did.

[Deery et al 2013: 136]

Book cover: The Moral Landscape by Sam Harris

Unfortunately, the rewording has not diminished the weaknesses found in Study 1. As with the prior study, the reworded training statement presumes that under determinism decisions 'had to happen' the way they did. Once again, this wording primes participants to answer in an incompatibilist fashion. Consistent with their earlier 'training', participants are led to interpret, 'I couldn't really have chosen differently than I did', as being incompatible with determinism. As before, this is precisely the question at issue.

The results for Study 2 [138–9] indicated no differences in participants' acceptance of incompatibilist feelings for imagined choices compared with actual choices. Likewise, there were no measured differences for moral choices compared with non-moral choices.

In Study 3, the authors [140] sought to eliminate a source of participant priming by removing the phrase, 'Even though it felt like I could either choose to X or choose to Y', from the statement given to participants. Once again, I don't think this rephrasing helps to eliminate the bias ingrained in the study. As I indicated above, the primary source of participant priming is in the 'training' given to participants. This 'training' remained unchanged in Study 3 compared with the previous two studies.

The results for Study 3 indicated, as per the previous two studies, a strong tendency by the participants to interpret their experience as being incompatible with determinism. However, given the participant priming that remained unabated in this final study, I don't consider it prudent to give the results any more credence than that earned by the initial two studies.

poor methodology from priming
Study results:
strongly supports incompatibilism
weakly supports incompatibilism

Copyright © 2016

You will be interested in
Book cover: Mistaken Identity: The Mind-Brain Problem Reconsidered by Leslie Brothers
The God Argument: The Case against Religion and for Humanism by A. C. Grayling
Book cover: A Brief History of the Soul by Stewart Goetz
Book cover: Consciousness Explained by Daniel C. Dennett
Book cover: Cognitive Science: An Introduction to the Science of the Mind by Jose Luis Bermudez
Book cover: Enlightenment Now: The Case for Reason, Science, Humanism, and Progress by Steven Pinker

Share This

  • twitter
  • facebook
  • linkedin
  • googleplus
  • gmail
  • delicious
  • reddit
  • digg
  • newsvine
  • posterous
  • friendfeed
  • googlebookmarks
  • yahoobookmarks
  • yahoobuzz
  • orkut
  • stumbleupon
  • diigo
  • mixx
  • technorati
  • netvibes
  • myspace
  • slashdot
  • blogger
  • tumblr
  • email
Short URL: