Abstract
Theories of free will are often measured against how well they capture everyday intuitions about free will. But what are these everyday intuitions, and what theoretical commitments do they express? Empirical methods have delivered mixed messages. In response, some free will theorists have developed error theories to undermine the credentials of countervailing intuitions. These efforts are predicated on the idea that people might misunderstand determinism in any of several ways. This paper sheds light on the comprehension problem. We first discuss recent efforts to explain systematic errors in how people interpret determinism. Then, we present the alarming results of two new preregistered studies exploring three types of comprehension failure: (a) epiphenomenal bypassing, (b) fatalistic bypassing, and (c) indeterministic intrusion. Our findings suggest that misunderstanding runs deeper than others have supposed. This casts doubt on existing models of commonsense thinking about free will. Unless and until researchers properly control for the kinds of misunderstandings we identify, research on free will intuitions cannot shed light on whether ordinary thinking reflects commitments to compatibilism or incompatibilism.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
This paper presumes that folk intuitions sometimes provide relevant constraints on theory construction. This might be because folk intuitions are relevant for fixing the reference of theoretical terms (as in Vargas, 2017; see also Chihara & Fodor, 1965), or because congruence with folk intuitions might constitute a theoretical virtue that warrants (all else being equal) selecting a more congruent theory over a less congruent theory. Whatever the case may be, our research is situated within an ongoing debate about whether folk thinking about free will is best characterized as compatibilist or incompatibilist, which includes the assumption that such thinking is sometimes theoretically relevant.
Deery et al. (2015) found that participant responses to different scale items indicated context-sensitive commitments to both compatibilism and incompatibilism (see p. 791). This might appear to be thin evidence at best for natural compatibilism. However, compatibilism is the thesis, roughly, that free will and determinism are logically compatible. The results from Deery et al. suggest that many people do not think determinism per se precludes having free will, but only that determinism precludes having free will in certain circumstances. This implies that there are compatibilist elements in folk psychological thinking about free will, even if such thinking is not uniformly compatibilist.
For an early diagnosis of the problem, see Cokely and Feltz (2010). Their focus is on why so many people fail what we are calling ‘surface’ comprehension checks (see Sect. 2.1.1). We are interested instead in how many people understand determinism.
To say that people falsely believe that determinism entails epiphenomenalism is not to say that people are disposed to state this entailment when prompted; rather, to say that people falsely believe that determinism entails epiphenomenalism is to say that responses elicited under experimental conditions are best explained by positing a false belief that determinism entails epiphenomenalism.
‘□’ denotes the necessity operator, ‘ → ’ denotes entailment, and ‘&’ denotes the conjunction operator. ‘Po’ names some proposition that denotes the state of the world at time o, ‘L’ names some proposition that denotes a set of laws of nature, and ‘P’ names some proposition that denotes the state of the world at the present time. Informally, Po stands for “the past”, L stands for “the laws”, and P stands for “the present”, so that the above could be read roughly as noting the mistake of equating: “Necessarily, the past and the laws entail the present” with “The past is necessary, the laws are necessary, and together they entail the present”. Determinism does not rule out the possibility of alternative laws of nature or past sequences of events.
To determine sample size, we conducted a power analysis using G*Power. For a linear regression with two predictors (vividness and either Epiphenomenalism, Fatalism, and intrusion predicting free will or moral responsibility) to detect a minimum effect size of interest (d = 0.15) at 95% power (error rate = 5%), we needed 107 participants. Given that only 1/3 of the sample would be administered each of the dependent variables of interest, we multiplied by a factor of 3 to get 321 participants. We over-recruited by 15% to account for exclusions.
MTurk is an online survey service that enables researchers to recruit and pay for participants for completing surveys of studies. For findings concerning the benefits of using MTurk—including the quality of the data and the improved diversity of the participant pool—see Burhmester et al. (2011), Paolacci et al. (2010) and Rand (2012).
This study was approved by the Institutional Review Board at College of Charleston
All stimuli and measures were presented on the same page for the study. This means that participants could always look back at the scenario to see the details.
A surface comprehension check can detect whether participants are able to correctly identify basic features of how the scenarios are worded but cannot detect whether they have a deeper understanding of the nature and implications of the key elements of the scenario.
As we noted above, 6% (19/325) failed this surface comprehension check.
Measures of vividness were included to explore the relationship between the quality of mental imagery and responses to items about the vignettes, but there were no pre-registered hypotheses about the relationship between vividness and judgments of free will. The 5-pt. Likert scale is a standard tool to measure vividness of mental imagery (see, e.g., De Brigard et al., 2013). Because analyses of vividness responses were merely exploratory, we do not discuss the results here. We mention them only to be fully transparent about all measures used in this research.
In their original study, Nichols and Knobe only asked about moral responsibility, which limited their ability to speak to the debate about natural incompatibilism, so we included statements about both free will and responsibility.
All six items were presented as a group and their order was randomized.
In designing these four items, we tried to word them in a way that addresses the main objection raised by Björnsson (2014) and Björnsson and Pereboom (2014) about earlier work on epiphenomenalism—namely, that the wording of the epiphenomenalism items used by Murray and Nahmias (2014) and Nahmias and Murray (2011) allows for an incompatibilist reading of their findings. Given the way we have worded our items (and given our results), we believe we have sidestepped this possible strategy for explaining away our findings.
We included items that expressed both naïve and modally sophisticated forms of fatalism.
This information was collected in Study 1 and Study 2 to explore whether the responses from naïve participants significantly differ from experienced participants since past research suggests experience may make a difference (Crone & Levy, 2018). Because this element of our study was exploratory and we did not make any preregistered predictions, we do not discuss these findings here. The complete data set can be found on our OSF page.
Percentages for free will and responsibility judgments were calculated by taking all participants who registered 5–7 on the free will item or responsibility item as agreeing that people could have free will or be morally responsible in a deterministic universe, while participants who registered 1–3 were counted as disagreeing that people could have free will or be morally responsible in a deterministic universe. Ratings of 4 were not counted either way (hence, percentages do not add up to 100). We round percentages down (.1–.4) or up (.5–.9) accordingly throughout the paper. We dichotomized responses made on a continuous scale for purposes of comparing our results to those reported in Nichols and Knobe (2007). The data for free will and responsibility judgments was bimodally distributed, which suggested that the data could be represented ordinally.
Because our study design departed in some ways from Nichols and Knobe (2007), our results reflect a failed conceptual replication rather than a failed direct replication. A conceptual replication of some study attempts to find relationships between variables using slightly different methods than the original study to see whether empirical findings are robust to methodology. A direct replication of some study attempts to find relationships between variables using the exact same procedure as the original study. Thus, because we did not find a relationship between reading about a deterministic scenario and judgments about free will using slightly different measures, we failed to conceptually replicate Nichols and Knobe (2007).
It is surprising because most participants in that condition agreed with fatalism, so it is odd that this did not influence their intuitions about free will.
This study was approved by the Institutional Review Board at College of Charleston.
The original Supercomputer scenario was set in the future in our own universe. We thought that by having participants instead imagine a generic universe where such a supercomputer existed, this would make it easier to ask how similar participants thought the Supercomputer scenario is to our own universe.
As we noted above, 5% (16/325) failed this surface comprehension check.
We would like to thank Alfred Mele for suggesting we include this item to control for the possibility that some people don’t find Jeremy blameworthy because they didn’t view the bank robbery negatively.
96% of participants (291/302) judged that Jeremy’s robbing the bank was a bad thing.
One interesting difference between Studies 1 and 2 concerns the reliability of the error measures. While Intrusion items exhibited high reliability across studies, the reliability of Bypassing and Fatalism items improved markedly. In Study 1, the reliability of Fatalism items was driven down by Item 4. Dropping Item 4 raises the internal reliability of Fatalism items to .804 (Cronbach’s alpha). However, this difference in Fatalism items did not emerge in Study 2. We do not know why these differences emerged and whether they resulted from our sample or from some feature of the vignette.
Someone might view our method as stacking the deck against comprehension, where one mistake counts as misunderstanding. However, computing comprehension failures as a function of average responses to the four items does not significantly change results. We still observe high failure rates for epiphenomenalism (91% in S1; 43% in S2), fatalism (90% in S1; 67% in S2), and intrusion (57% in S1; 71% in S2) across both studies. The total failure rate using averages is 70% (409 participants). Hence, even on more conservative estimations of failure, we still observe high rates of misunderstanding.
Someone might still object to this interpretation of participant responses. After all, it might seem that we should utilize a principle that meaning is fixed or settled by use, such that ‘determinism’ refers to whatever best explains the balance of participant responses in adequately powered, representative studies. On that view, we might wonder whether ‘determinism’ just means ‘fatalism’ or ‘epiphenomenalism’ or ‘bypassing’ given how participants behaved in our studies (more precisely, we might treat these terms as names of propositions, where ‘determinism’ and either of ‘fatalism’, ‘epiphenomenalism’, or ‘bypassing’ corefer to the same proposition). However, in [omitted], we found that when participants are presented with measures from all three error scales, nearly 50% of participants made all three errors. Given that fatalism, epiphenomenalism, and bypassing are jointly inconsistent with each other, it seems more plausible to interpret participants as making erroneous judgments about determinism rather than fixing the reference of ‘determinism’ in a way that is consistent with the usage of the term.
It might seem the obvious response to invoke principles of reflective equilibrium. Empirical evidence (especially judgments taken to indicate folk conceptual commitments) must always be balanced against other evidence when drawing conclusions about concepts. However, this response misses the point. There is no work for reflective equilibrium to do here because there is nothing to bring into balance. Our results call into question the inferential value of extant evidence entirely.
References
Andow, J., & Cova, F. (2016). Why compatibilist intuitions are not mistaken: A reply to Feltz and Millan. Philosophical Psychology, 29(4), 550–566. https://doi.org/10.1080/09515089.2015.1082542
Audi, R. (1991). Responsible action and virtuous character. Ethics, 101(2), 304–321.
Björnsson, G. (2014). Incompatibilism and “bypassed agency.” In A. Mele (Ed.), Surrounding free will (pp. 95–112). Oxford University Press.
Björnsson, G., & Pereboom, D. (2014). Free will skepticism and bypassing. In W. Sinnott-Armstrong (Ed.), Moral psychology: Free will and moral responsibility (Volume 4) (pp. 27–36). M.I.T. Press.
Burhmester, M., Kwang, T., & Gosling, S. (2011). Amazon’s Mechanical Turk: A new source of inexpensive, yet high-quality data? Perspectives on Psychological Science, 6, 3–5. https://doi.org/10.1177/1745691610393980
Chihara, C. S., & Fodor, J. (1965). Operationalism and ordinary language: A critique of Wittgenstein. American Philosophical Quarterly, 2(4), 281–295.
Cokely, E., & Feltz, A. (2010). Questioning the free will comprehension question. Proceedings of the Annual Meeting of the Cognitive Science Society, 32(32), 2440–2445.
Cover, J. A., & O’Leary-Hawthorne, J. (1996). Free agency and materialism. In J. Jordan & D. Howard-Snyder (Eds.), Faith, freedom and rationality (pp. 47–71). Lanham: Roman and Littlefield.
Crone, D., & Levy, N. (2018). Are free will believers nicer people? Four studies suggest not. Social Psychological and Personality Science. https://doi.org/10.1177/1948550618780732
De Brigard, F., Addis, D. R., Ford, J. H., Schacter, D. L., & Giovanello, K. S. (2013). Remembering what could have happened: Neural correlates of episodic counterfactual thinking. Neuropsychologia, 51(12), 2401–2414.
Deery, O., Bedke, M., & Nichols, S. (2013). Phenomenal abilities: Incompatibilism and the experience of agency. In D. Shoemaker (Ed.), Oxford studies in agency and responsibility (pp. 126–150). Oxford University Press.
Deery, O., Davis, T., & Carey, J. (2015). The free-will intuitions scale and the question of natural compatibilism. Philosophical Psychology, 28(6), 776–801. https://doi.org/10.1080/09515089.2014.893868
Dennett, D. (1984). I could not have done otherwise: So what? Journal of Philosophy, 81(10), 553–565.
Doris, J. M., & Knobe, J. (2010). Responsibility. In J. M. Doris (Ed.), The moral psychology handbook (pp. 321–354). Oxford University Press.
Ekstrom, L. W. (2002). Libertarianism and Frankfurt-style cases. In R. Kane (Ed.), The Oxford handbook of free will (pp. 309–322). Oxford University Press.
Feltz, A., Cokely, E., & Nadelhoffer, T. (2009). Natural compatibilism versus natural incompatibilism: Back to the drawing board. Mind & Language, 24(1), 1–23. https://doi.org/10.1111/j.1468-0017.2008.01351.x
Feltz, A., & Millan, M. (2013). An error theory for compatibilist intuitions. Philosophical Psychology, 28, 529–555. https://doi.org/10.1080/09515089.2013.865513
Ismael, J. (2019). Determinism, coounterpredictive devices, and the impossibility of Laplacean intelligences. The Monist, 1–2(4), 478–498.
Jackson, F. (1998). From metaphysics to ethics: A defense of conceptual analysis. Oxford University Press.
Kane, R. (1999). Responsibility, luck, and chance: Reflections on free will and indeterminism. Journal of Philosophy, 96, 217–240.
Kane, R. (2005). A contemporary introduction to free will. Oxford University Press.
Knobe, J. (2014). Free will and the scientific vision. In E. Machery & E. O’Neill (Eds.), Current controversies in experimental philosophy (pp. 69–85). Routledge.
Lycan, W. (2003). Free will and the burden of proof. In A. Hear (Ed.), the Proceedings of the Royal Institute of Philosophy for 2001–02 (pp. 107–122). Cambridge: Cambridge University Press.
Monroe, A. E., Dillon, K. D., & Malle, B. F. (2014). Bringing free will down to Earth: People’s psychological concept of free will and its role in moral judgment. Consciousness and Cognition, 27, 100–108. https://doi.org/10.1016/j.concog.2014.04.011
Murray, D., & Nahmias, E. (2014). Explaining away incompatibilist intuitions. Philosophy and Phenomenological Research, 88, 434–467. https://doi.org/10.1111/j.1933-1592.2012.00609.x
Nadelhoffer, T., Rose, D., Buckwalter, W., & Nichols, S. (2020a). Natural compatibilism, indeterminism, and intrusive metaphysics. Cognitive Science, 44(8), e12873. https://doi.org/10.1111/cogs.12873
Nadelhoffer, T., Yin, S., & Graves, R. (2020b). Folk intuitions and the conditional ability to do otherwise. Philosophical Psychology. https://doi.org/10.1080/09515089.2020.1817884
Nahmias, E. (2011). Intuitions about free will, determinism, and bypassing. In R. Kane (Ed.), The Oxford Handbook on Free Will (2nd ed., pp. 555–576). Oxford University Press.
Nahmias, E., Morris, S., Nadelhoffer, T., & Turner, J. (2004). The phenomenology of free will. The Journal of Consciousness Studies, 11, 162–179.
Nahmias, E., Morris, S., Nadelhoffer, T., & Turner, J. (2005). Surveying free will: Folk intuitions about free will and moral responsibility. Philosophical Psychology, 18, 561–584. https://doi.org/10.1080/09515080500264180
Nahmias, E., Morris, S., Nadelhoffer, T., & Turner, J. (2006). Is incompatibilism intuitive? Philosophy and Phenomenological Research, 73, 28–53. https://doi.org/10.1111/j.1933-1592.2006.tb00603.x
Nahmias, E., & Murray, D. (2011). Experimental philosophy on free will: An error theory for incompatibilist intuitions. In J. Aguilar, A. Buckareff, & K. Frankish (Eds.), New waves in philosophy of action (pp. 189–216). Palgrave-Macmillan.
Nichols, S. (2006a). Folk intuitions on free will. Journal of Cognition and Culture, 6, 57–86. https://doi.org/10.1163/156853706776931385
Nichols, S. (2006b). Free will and the folk: Response to commentators. Journal of Cognition and Culture, 6, 305–320. https://doi.org/10.1163/156853706776931376
Nichols, S. (2012). The indeterminist intuition. The Monist, 95, 290–307. https://doi.org/10.5840/monist201295216
Nichols, S. (2015). Bound. Oxford University Press.
Nichols, S., & Knobe, J. (2007). Moral responsibility and determinism: The cognitive science of folk intuition. Noûs, 41, 663–685. https://doi.org/10.1111/j.1468-0068.2007.00666.x
Nowell-Smith, P. H. (1949). Free will and moral responsibility. Mind, 57, 45–65.
O’Connor, T. (2000). Persons and causes: The metaphysics of free will. Oxford University Press.
Paolacci, G., Chandler, J., & Ipeirotis, P. (2010). Running experiments on Amazon Mechanical Turk. Judgment and Decision Making, 5, 411–419.
Pereboom, D. (2001). Living without free will. Cambridge University Press.
Rand, G. (2012). The promise of mechanical turk: How online labor markets can help theorists run behavioral experiments. Journal of Theoretical Biology, 299, 172–179. https://doi.org/10.1016/j.jtbi.2011.03.004
Rose, D., Buckwalter, W., & Nichols, S. (2017). Neuroscientific prediction and the intrusion of intuitive metaphysics. Cognitive Science, 41, 482–502. https://doi.org/10.1111/cogs.12310
Rose, D., & Nichols, S. (2013). The lesson of bypassing. Review of Philosophy and Psychology, 4, 599–619. https://doi.org/10.1007/s13164-013-0154-3
Roskies, A., & Nichols, S. (2008). Bringing moral responsibility down to Earth. The Journal of Philosophy, 105, 371–388. https://doi.org/10.5840/jphil2008105737
Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2012). A 21 word solution. Dialogue: the Official Newsletter of the Society for Personality and Social Psychology, 26, 4–7. https://doi.org/10.2139/ssrn.2160588
Strawson, G. (1986). Freedom and belief. Oxford University Press.
Turri, J. (2017). Exceptionalist naturalism: Human agency and the causal order. The Quarterly Journal of Experimental Psychology. https://doi.org/10.1080/17470218.2016.1251472
van Inwagen, P. (1983). An essay on free will. Oxford University Press.
Vargas, M. (2013). Building better beings. Oxford University Press.
Vargas, M. (2017). Contested terms and philosophical debates. Philosophical Studies, 174(10), 2499–2510.
Woolfolk, R. L., Doris, J. M., & Darley, J. M. (2006). Identification, situational constraint, and social cognition: Studies in the attribution of moral responsibility. Cognition, 100, 283–301. https://doi.org/10.1016/j.cognition.2005.05.002
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Nadelhoffer, T., Murray, S. & Murry, E. Intuitions About Free Will and the Failure to Comprehend Determinism. Erkenn 88, 2515–2536 (2023). https://doi.org/10.1007/s10670-021-00465-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10670-021-00465-y