Skip to main content

Ethical evidence

Abstract

This paper argues that ethical propositions can legitimately be used as evidence for and against empirical conclusions. Specifically, I argue that this thesis is entailed by several uncontroversial assumptions about ethical metaphysics and epistemology. I also outline several examples of ethical-to-empirical inferences where it is extremely plausible that one can rationally rely upon their ethical evidence in order to gain a justified belief in an empirical conclusion. The main upshot is that ethical propositions can, under perfectly standard conditions, play both direct and indirect evidential roles in (social) scientific inquiry.

This is a preview of subscription content, access via your institution.

Notes

  1. By ‘ethical propositions’, I mean propositions which feature evaluative or normative words like, ‘good’, ‘bad’, right’, ‘wrong’, ‘just’, ‘fair’ and so on. More stringently (and in order to avoid complications involving disjunctive propositions), we can define an ethical proposition as one whose truth essentially depends on whether an ethical property is instantiated (or some anti-realist-friendly equivalent of this). ‘Empirical propositions’ are propositions whose truth essentially depends on the instantiation of the kind of property which would be described by our best scientific theories (although in this paper I will only be concerned with empirical propositions which purport to describe the world at the level of description of the social sciences). Propositions which feature “thick” predicates (e.g., ‘selfish’, ‘admirable’) plausibly fall somewhere between ethical and empirical propositions as I have defined them. If the argument in this paper succeeds in showing that ethical propositions can be used as evidence for and against empirical propositions, then it shows a fortiori that “thick” propositions can also be used as evidence in this way.

  2. Throughout this paper, I assume that bodies of evidence are composed of propositions, so it would be the proposition that Spencer cut the cake unfairly which is my evidence for the conclusion that Spencer cut the cake unevenly, rather than my mere belief about this proposition, or experience of the unfairness, etc. Not much turns on this assumption. The discussion below could be easily adapted to show that ethical beliefs (rather than ethical propositions) can be legitimately used as evidence for empirical conclusions. The matter is a bit trickier if we think of evidence as constituted only by experiences or immediate sense-data, since it is not obvious that we have distinctively ethical experiences or sense-data. In this case, however, it would be sufficient for the discussion in this paper to establish that ethical propositions can play the same pseudo-evidential role which empirical propositions standardly play, e.g., when these empirical propositions are learned via testimony. To state the overall point of this paper as abstractly as possible, the idea is that—whatever account of evidence one prefers—there are no grounds for treating ethical propositions or beliefs as failing to play whatever evidential or quasi-evidential role which one ascribes to empirical propositions or beliefs within this account of evidence.

  3. The main location where there has been discussion (or rather, a conspicuous lack of discussion) about the idea that ethical propositions can count as evidence for empirical conclusions is the large philosophical literature on the role of value-judgements in scientific inquiry. In this debate, the distinction is commonly drawn between ‘direct’ and ‘indirect’ roles (Longino, 1990; Douglas, 2000, p. 564) which value-judgements—i.e., ethical beliefs—can play in scientific research, and it is usually assumed without argument (or at most a gesture towards a worry about wishful thinking, see Sect. 1 below) that value-judgements cannot play a direct evidential role in influencing scientific research. However, there have been some notable exceptions to this trend, especially from feminist and pragmatist philosophers of science who have argued that value-judgements can play an evidential or quasi-evidential role in scientific inquiry (e.g., Anderson, 2004; Brown, 2020; Campbell, 1998; Clough, 2003, 2011; Nelson, 1990; White, 1981). The aim of this paper is to build on their work by giving a clearer and more systematic treatment of the metaphysics and epistemology underlying this position.

    Outside of the values-in-science debate, the question of whether ethical propositions can be used as evidence for empirical conclusions has also recently cropped up in metaethics, with Barber (2013) and Basl and Coons (2017) each trying to use a negative answer to this question in order to motivate a controversial result about ethical metaphysics or epistemology (see Sect. 2 below). However, it is worth noting that in other areas of this literature, it is assumed to be true without much argument that ethical propositions can legitimately be used as evidence for empirical conclusions. For instance, this features as a key premise in Dorr’s (2002) influential argument against certain versions of metaethical non-cognitivism, and in Podgorski’s (2020) recent argument about the ethical significance of moral uncertainty.

  4. To clarify: what I call the ‘metaphysical’ objection to Ethical Evidence involves both a worry about the metaphysics of ethical properties and about the content of ethical beliefs. This latter worry is not strictly about ethical metaphysics, but the metaphysics of ethical properties and the content of ethical beliefs are typically discussed together in the metaethical literature, so I lump them together here.

  5. See Bedke (2017) for an introduction and overview of the debate about the conative content of ethical beliefs.

  6. Technically, this objection would render the Ethical Evidence thesis trivially true rather than false.

  7. It is worth emphasising that even metaethical ‘non-cognitivists’ almost universally accept this kind of minimal cognitivism and truth-aptness. See, e.g., Bedke (2017, pp. 297–298).

  8. McPherson (2012, p. 205) notes that ‘the supervenience of the ethical has been a rare locus of near-consensus in metaethics’.

  9. For explanatory connections in realist theories, see Rosen (2017). Anti-realist philosophers are usually less explicit about the (indirect) explanatory connection between the empirical and ethical facts in abstract discussion, but it is easy to extract a commitment to this assumption from their specific theories. For example, on Gibbard’s (2003) influential expressivist metaethical theory, the explanation for why a particular contingent ethical statement is true is (roughly) that (i) a relevant bunch of empirical facts P are true and (ii) the agent (relative to whom the ethical fact can be treated as a true) has the relevant set of ‘planning attitudes’. Even though the explanatory connection is indirect, the truth of a particular empirical proposition P can nevertheless feature as part of the explanation for why a particular ethical proposition E is true.

  10. On McCain’s account of evidential support, evidence is constituted by appearances rather than propositions. As I discussed in fn.2, this does not substantively change the discussion in this paper. Additionally, McCain thinks that the explanandum of this Best Explanation is not just why P is true, but rather why P is possessed as evidence (i.e., why it seems to one as if P, or why one has a justified belief that P, etc.). This does not matter for the current discussion: if an empirical proposition can feature in the Best Explanation of why a particular ethical proposition is true, then it is uncontroversial that this empirical proposition can also feature as part of the Best Explanation of why it seems to someone as if that ethical proposition is true.

  11. Basl and Coons (2017, pp. 162–167) argue for a similar conclusion from the opposite direction. That is, while I argue from first principles (i.e., assumptions which are widely shared in the current metaethical debate), they argue via elimination from existing categories of metaethical theories. See also Brown (2020, pp. 89–100, 163–165), who argues that on any ‘modest’ cognitivist metaethical theory, our beliefs about ethical propositions (or value-judgments more broadly) will be the right sort of thing to play an evidential role in scientific inquiry.

  12. It is worth noting that Basl and Coons (and to some degree Barber) are only concerned with the Ethical Evidence thesis insofar as it bears on the question, ‘why doesn’t moral reasoning serve as a tool for learning about the physical world?’ (Basl and Coons 2017:161, emphasis added). In other words, these authors are more concerned with why it is impermissible to go about researching or investigating empirical reality by means of investigating ethical truths. However, (i) their arguments against “moral science” still bear directly on the Ethical Evidence thesis, so I focus on this aspect of their discussion in the present section; (ii) on a wider, more social interpretation of what ‘investigation’ amounts to, it is plausible that seeking out reliable testimony can be a means of investigation—and if so, then this section of the paper should also be interpreted as arguing that we can investigate the empirical world by means of ethical inquiry and investigation; and finally, (iii) although it is beyond the scope of this paper to defend this stronger thesis, Brown (2020) argues that we can (and should) do scientific investigation and inquiry which is epistemically influenced by our moral reasoning and imagination—which if correct might also entail that a form of “moral science” is epistemically permissible. Thanks to a reviewer for pressing me to say more about this.

  13. Barber (2013, p. 634) and Basl and Coons (2017, pp. 161, 168) give several other examples of ethical-empirical inferences which also appear to be epistemically irrational. The discussion below does not specifically address all of these examples, and one might think that some of them fail to transmit justification for reasons which are more complicated than those I discuss here. This is probably true, but the basic point remains that even if ethical propositions can generally be relied upon as empirical evidence, there would still be a wide range of cases where inferring from an ethical proposition to an empirical conclusion fails to give an agent epistemic justification for believing the conclusion, but the explanation for this has to do with (perhaps quite complicated) general epistemological principles governing the transmission of justification rather than anything to do with the ethical nature of the premises.

  14. Barber (2013, pp. 637–639) also entertains the possibility that this inference fails to transmit justification because some extreme form of metaethical anti-realism is true. However, Basl and Coons (2017, pp. 162–167) convincingly argue that Barber’s discussion here is largely misguided.

  15. Neither Barber nor Basl and Coons explicitly reference this extensive literature in their discussion, although it is clear that this is precisely the epistemological problem with which they are concerned. In what follows, I aim to clearly reconstruct the best version of their arguments, rather than sticking to strict exegesis.

  16. A prominent exception is Tucker (2010), who argues that one actually can gain (doxastic) justification for believing that the animal is not a cleverly-disguised mule by believing this on the basis of one’s belief that the animal is a zebra. More generally, it is worth stressing that even those epistemologists who think that an agent cannot usually gain epistemic justification for believing the conclusion by making this inference should think that there are certain special cases where justification does transmit. In particular, if one received reliable testimony from a zookeeper that this animal is a zebra, then it seems that one could infer from the fact that it is a zebra and gain epistemic justification in the conclusion that it is not a cleverly-disguised mule.

  17. At least, the proposition that this animal looks like a zebra is no better evidence for one conclusion over the other once the sceptical possibility is raised (i.e., once it is relevant to consider the possibility that the animal is a cleverly-disguised mule). See Brown (2016).

  18. Moretti and Piazza (2013) give a sophisticated counterexample to this principle involving probabilistic outcomes. However, a more obvious problem is that this principle appears to entail the highly counterintuitive result that, whenever one infers from P to Q and then from Q to R, this second inference always fails to transmit justification because one’s (propositional) epistemic justification for believing both Q and R wholly depend on one’s antecedent justification for believing P. See Tucker (2010, pp. 512–514) for further discussion.

  19. Or, at least, for most cases. For instance, Basl and Coons (2017, pp. 180–184) argue that there is a different reason why inferences from ethical premises to empirical conclusions fail to transmit justification when these ethical and empirical propositions purport to describe necessary truths or law-like statements. We can put this class of inferences aside for the purpose of the present discussion.

  20. I follow Basl and Coons in using the phrase, ‘moral perception’, slightly out of line with the general philosophical literature here. Strictly speaking, what I have said above amounts to the rejection of the thesis that we gain ethical knowledge through a distinctively moral perception, and instead, this kind of direct ethical knowledge comes via making an inference from one’s justified empirical beliefs. This position it not uncontroversial; see Werner (2020) for a recent overview of the debate around moral perception.

  21. At this stage, it is instructive to consider the following objection. If one accepts a strong form of reductionism about testimonial justification, then one thinks (roughly) that one can gain epistemic justification for a testimonial belief that X only if they make a quasi-inference from the fact that such-and-such said X (Lackey, 2006, pp. 160–163). So, it seems that if one gained a belief in an ethical proposition via testimony, one’s epistemic justification for believing this proposition would wholly depend on one’s prior epistemic justification for believing an empirical proposition about the testifier’s assertion. So, if one embraces this controversial reductionist view about testimonial justification, then one’s justification for believing ethical propositions will still depend on one’s justification for believing some empirical proposition. Is this a problem for my argument? No. First, it is not clear that, just because one’s justification for believing a particular ethical proposition E depends on one’s prior justification for believing the particular empirical proposition that S testified that E, one is therefore not in a position to use E as evidence for a different empirical conclusion (i.e., one for which the fact that S testified that E would not be good evidence anyway). Second, and more importantly, even if this objection showed that ethical propositions learned via testimony are not sufficiently independently justified in order to be relied upon as evidence, the argument would also extend much further. That is, this objection would show that one can never use any proposition P which one learns via testimony as evidence, since its evidential relevance is always ‘screened off’ by the dependence on one’s prior justification for believing that so-and-so asserted that P. Even if one were to embrace this implausible conclusion, the comparative conclusion would still remain that there is no relevant difference between the evidential role played by ethical propositions and non-ethical propositions which are learned via testimony (c.f., n.2 above).

  22. One way to see this is to consider the case of a testimonially justified ethical belief being used as evidence for other ethical propositions. For instance, suppose someone tells me that it’s wrong to eat meat, and I accept this testimony, thereby forming the relevant ethical belief in this proposition. Since I already (independently) know some relevant empirical propositions (e.g., the meat-production industry causes animals a lot of pain), I can plausibly use the ethical proposition which I have just learned to (justifiably) infer the ethical conclusion that this pain morally outweighs the pleasure that humans get from eating meat. In other words, when I learn an ethical proposition via testimony, I am plausibly in a position to use this ethical proposition as evidence for other ethical conclusions, in order to gain ethical understanding (c.f., Wiland, 2021, pp. 56–57). Likewise, when I learn an ethical proposition via testimony and I know the relevant ethical bridge-principles but not the relevant empirical facts, I can plausibly legitimately use this ethical proposition as evidence in favour of the relevant empirical conclusions.

  23. For a recent overview of the philosophical literature on standpoint epistemology, see Toole (2021). Standpoint theorists tend not to explicitly state that some of the truths which can be known from an marginalised standpoint have ethical content. However, this follows readily from some other claims of standpoint epistemology, in particular, that oppressed people are in a better position to understand the features of their oppression (which plausibly includes the normative features of this oppression).

  24. It is worth adding to this that, as a sociological fact, the production of ethical knowledge often operates largely independently of the production of empirical knowledge by the scientific community. For instance, Anderson (2016) documents how the ethical knowledge that the actual, historical practice of slavery was morally impermissible was generated primarily by people under risk of being enslaved themselves, while the European scientific and intellectual community continued to believe both that slavery was permissible and that Black slaves were intellectually and emotionally inferior to free Whites. When attitudes to the moral permissibility of slavery changed widely in the nineteenth century, this change was prompted by engagement with the lived experience of the oppressed and not by revised scientific beliefs about racial intellectual equality amongst European scientists.

  25. And these norms of scientific inquiry might be pragmatically justified, e.g., by the role of science in a democracy (Bright, 2018).

  26. The Culture of Poverty hypothesis is usually taken to have been first developed by Oscar Lewis (1959). It is not entirely clear whether Lewis and other early exponents used the theoretical term, ‘culture’, to refer to a broad social structural phenomenon, or to individual members of a community’s habits and values (Steinberg, 2011). The latter became the standard understanding of the hypothesis, especially after the publication of the notorious Moynihan Report (1965), which argued that one important reason why African-American communities remained poor was the fact that their ‘family values’ differed from those of White Americans, in such a way that prevented African-Americans from escaping poverty. This met with immediate criticism and controversy but shaped popular discourse about the ‘culture of poverty’ with long-lasting effects. The theory re-emerged in the 1980’s under the term, ‘underclass’ (Auletta, 1982), and it has recently sprung up again (e.g., Small et al., 2010); though in a much more nuanced form than I do justice to in the following discussion. Additionally, individualistic versions of the Culture of Poverty hypothesis appear to remain prevalent in certain circles of popular right-wing discourse, especially in the USA—often with clearly racist undertones.

  27. This is probably an exaggeration. However, the relevant point is just that even if our current sociological evidence could not definitively reject the Culture of Poverty hypothesis, it would still be plausible that we could use our justified belief in an ethical proposition as further evidence against this hypothesis.

  28. Another way of making this point is in terms of a Quinean ‘web of belief’, where the total body of possessed evidence against which we epistemically evaluate our total body of theories includes our justified ethical beliefs (see especially White, 1981, Nelson, 1990, Solomon, 2012).

References

Download references

Acknowledgements

Many thanks to Al Prescott-Couch for extensive discussion around this project and suggestions for some of the examples. Thanks also to Jonathan Ichikawa, Emily Tilton, Phyllis Pearson, Alex Bryant, Kelsey Vicars, Cam Gilbert, Mira Kuroyedov and two reviewers for this journal.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Steven Diggin.

Ethics declarations

Conflict of interest

The author declares that there is no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Diggin, S. Ethical evidence. Synthese 200, 329 (2022). https://doi.org/10.1007/s11229-022-03808-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1007/s11229-022-03808-6

Keywords

  • Values in science
  • Metaethics
  • Moral epistemology
  • Evidence
  • Philosophy of social science