Abstract
This paper takes stock of Parfit’s latest defence of his meta-ethical view, non-realist cognitivism. In the third volume of On What Matters, Parfit gives his account a new—as I am going to show: conceptual—spin. Also, quite surprisingly, he takes back much of his earlier criticism of rival theories and claims instead that he and his opponents, Allan Gibbard and Peter Railton, are climbing the same meta-ethical mountain on different sides.
Mainly focusing on the new spin in Volume III, I argue for the following four claims. Firstly, non-realist cognitivism can easily be accepted by all those believing in the irreducibility of normative concepts (not: facts). Secondly, Parfit succeeds in avoiding Mackie-style queerness objections. However, thirdly, once we fully grasp the conceptual rather than metaphysical core of non-realist cognitivism, it becomes clear that the view, if successful, would accomplish much less than Parfit’s talk about irreducible, non-natural normative facts may have previously suggested. Finally, I argue that the conceptual spin generates a problem regarding one of the crucial pillars of non-realist cognitivism, namely the status of normative facts as objective or mind-independent. It remains entirely unclear how the view could account for this status. All in all, non-realist cognitivism doesn’t clear up the mists covering the heights of the meta-ethical mountain. It is still a long climb to the summit, and in order to reach it, we have to answer many questions Parfit doesn’t address.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
In recent years, many books and articles about On What Matters have been published. I cannot, here, give an overview of all valuable contributions. But let me name a few. Three particularly helpful anthologies are Kirchin (2017); Hoesch et al. (2017); Singer (2017). Schroeder (2011) is an insightful review of Volumes I and II.
Part of the development takes place in Volume III’s companion volume Does Anything Really Matter? (Singer 2017), to which I will also refer below.
In Volume III, Parfit includes two comments from Gibbard and Railton that play a crucial role regarding the convergence claim. The official «timeline» of the exchange between Parfit, Gibbard and Railton is this: Parfit presents non-realist cognitivism in Volume II (Parfit 2011b). Gibbard and Railton comment on the view in Singer (2017). Parfit replies to these comments in Volume III (Parfit 2017), and also includes Gibbard’s and Railton’s replies to his replies in the same volume.
In the following, I will use «fact» and «truth» interchangeably, as Parfit does.
Thereby providing an «ostensive definition» (Finlay 2017).
In what follows, all reasons and oughts are understood to be pro tanto.
This implies, of course, that non-normative concepts could never refer to normative facts. The kind of irreducibility Parfit has in mind here—namely a conceptual one—will become clearer further below.
For example, the fact that I should go get some milk could never turn out to be the same as its supervenience-base; whatever that base is. Different normative theories characterize the reason-giving features of the world in different ways. Subjectivists claim that a desire-related fact determines and explains that, e.g., I should go buy some milk. On the other hand, objectivists claim that reasons for obtaining milk provisions are provided by a (desire-independent) feature of having milk at one’s disposal (cf. Parfit 2011a, p. 45). It is important to understand, however, that these disagreements about the reason-giving features of the world are orthogonal to the problem we are dealing with here. Whatever the reason-giving facts turn out to be, Parfit, Railton and Gibbard agree that they could not be the same as the reason-facts they determine. This is why Parfit discusses subjectivism and objectivism as normative views in Volume I, and meta-ethical views about the nature of the relation between empirical and normative facts in Volume II.
This is why Laskowski (2018, p. 499) calls the introduction of these distinctions the new «big idea» of Volume III.
Parfit (2017, p. 150). Regarding the terminology: When Parfit introduces what I just called informational fact, he calls it «pleonastic» (Parfit 2017, p. 68). This label, however, is a bit difficult to understand, which is why Railton (2017, p. 54) proposes «linguistic» instead. (Parfit explicitly uses the notion of a linguistic property at one point in Volume II (Parfit 2011b, p. 348).) Railton’s proposal is more intuitive, I take it, for it suggests that the sense of «fact» we are after has something to do with language. For reasons that will soon become clear I go with «informational fact», which suggests, much like Railton’s label, that the sense of «fact» we are after is sensitive to the information provided by appealing to the fact. See also Gibbard (2017, pp. 74–6). For Parfit’s reaction to Gibbard, see Parfit (2017, pp. 89–90, 196). For Railton’s comments on the issue, see Railton (2017, p. 54). For Parfit’s convergence regarding «facts» and «truths», see Parfit (2017, pp. 99–102).
My depiction follows Parfit (2017, p. 90).
A reader may find this way of setting up the terminology slightly awkward. We will come back to worries of this kind further below. For now, our aim is solely to present Parfit’s considerations.
How exactly states of affairs are individuated need not concern us here. Very roughly, my answer would be: by their causal powers.
Parfit (2017, p. 229). To be precise: In the quoted passage, Parfit characterizes a sense of «property». But since it works analogically to the informational sense of «fact», this is not a problem.
Readers familiar with Frege’s distinction between Sinn and Bedeutung (Frege 1980) will probably experience a déjà-vu at this point. The idea that we can refer to the same entities by means of different thoughts is, of course, nothing new. Sometimes, Parfit would be a bit easier to read if he provided readers with references helping them to recognize some of the «classic» ideas underlying his thoughts.
Railton (2017, p. 53) emphasizes that Parfit’s definitions «generate new properties wherever there is new information». New wording, new fact.
«[I]t’s one thing to be the positive square root of 4, and quite a different thing to be the only even prime number» (Parfit 2017, p. 69). One problem of this account of property-identity is that it seemingly cannot live up to what Bader (2017, p. 113) calls the «worldliness constraint». It is unclear why the differences tracked by Parfit’s account should be taken to be worldly differences, and not merely nomological or conceptual ones. We will come back to this point in our assessment below.
Parfit uses a more complicated formulation for this natural fact. I choose a simpler one. Nothing substantial is lost.
Parfit’s focus on informational facts in Volume III renders this part of his view a lot more lucid for the reader. See also Laskowski (2018, pp. 499–500).
J.L. Mackie first formulated the argument from queerness against the idea that there are objectively prescriptive moral or evaluative facts. Such facts, he claimed, would be utterly different from all other (referential) facts we know of; which is why we should not believe in their existence (Mackie 1977, ch. 1).
In multiple passages throughout Volume III, Parfit stresses that only normative thoughts or beliefs could help us «to make good decisions and to act well» (cf. Parfit 2017, p. 85). Natural or empirical thoughts could not. He seems to think that normative thoughts have a specific deliberative role that non-normative thoughts lack. This, however, can easily be accepted by naturalists (Parfit 2017, p. 101).
See also Railton’s comments on Methodological Naturalism and how it differs from what Parfit calls «Metaphysical Naturalism» (Railton 2017, pp. 44–5).
Railton (2017, p. 46). In Railton’s own terminological framework we encounter «job descriptions of phenomenological concepts» and «conceptual roles» instead of Parfit’s informational facts and linguistic properties—without losing anything Parfit wants to say (ibid., pp. 49–50).
See also Laskowski (2018, pp. 501–3).
At least, it seems plausible if you aren’t an analytical naturalist.
In Volume III, Parfit considerably moves towards Gibbards expressivism: «Like Gibbard, I have changed my view. I now believe that, as Gibbard claims, his oblique expressivist account helps to explain much of our normative thinking» (Parfit 2017, p. 231). He also admits that his previous criticism of Gibbard’s and Blackburn’s positions was a «bad mistake» (ibid., p. 236).
For Railton’s approach, see Seligman et al. (2016).
One way to capture (1) and (2) might be a subjectivist account, according to which the truth-makers of normative claims include facts about desire-like states. But such an account probably would have problems capturing (3) (in the sense Parfit wants to capture it). Moreover, Parfit goes through lengths to reject subjectivism in Volume II (Parfit 2011b, pp. 263–356). For similar worries regarding Parfit’s balancing act between truth-minimalism and robust non-naturalism, see Muders and Rüther (2017), who ultimately conclude that Parfit would be best off by becoming a robust non-naturalist.
In response, Parfit might point to the domain of mathematics or logic, saying: «Whatever sense of truth they use, I use it as well». However, as (Copp 2018, p. 577), among others, points out, this is not a helpful reply because the ontology and semantics of mathematical and logical claims are at least as contested as the ontology and semantics of normative claims.
As an anonymous referee helpfully points out: Another possible route to objective moral truths without assuming ontologically robust moral facts is proposed in Hare (1981). Hare’s basic idea is that the objective truth of a moral judgment consists in the fact that all rational persons must agree to it. Such a view might also be labeled ‹non-realist cognitivism›. It is ‹non-realist› because there are no ontologically robust moral facts. And it is ‹cognitivist› in so far it includes the truth-assessability of moral claims.
Furthermore, it would seem a bit strange to say that a normative claim «refers to»—rather than, say, «expresses»—a true thought. And even if we say that it expresses a true thought, we still don’t know why the expressed thought should be taken to be objectively true.
Parfit states at one point that, if there are no objective normative truths of the kind he envisages, then he, Ross and Sidgwick «would have wasted much of our lives» (Parfit 2011b, p. 367).
Again, I wouldn’t go as far as claiming that Parfit’s conceptual turn in Volume III was entirely new. It was, however, a lot more lucid than before, and this is mainly due to Parfit’s new distinction between informational and referential facts.
References
Bader, Ralf. (2017). The grounding argument against non-reductive moral realism. In Oxford Studies in Metaethics, Vol.12, ed. Russ Shafer-Landau. Oxford, Cambridge (MA): Oxford University Press.
Brandhorst, Mario. (2015). “Der neue normative Realismus: einige kritische Fragen.” Zeitschrift für Philosophische Forschung 69 (3): 275–305.
Copp, David. (2018). “A semantic challenge to non-realist cognitivism.” Canadian Journal of Philosophy 48 (3–4).
Finlay, Stephen. (2014). Confusion of Tongues. Oxford, New York: Oxford University Press.
Finlay, Stephen. (2017). Defining Normativity. In Legal Norms, Ethical Norms: New Essays on Metaethics and Jurisprudence, ed. D. Plunkett, S. Shapiro and K. Toh. Oxford University Press.
Fischer, Stefan. (2018). The Origin of Oughtness. A Case for Metaethical Conativism. Berlin: De Gruyter.
Frege, Gottlob. (1980). On Sense and Reference. In Translations from the Philosophical Writings of Gottlob Frege, ed. Peter Geach and M. Black. third ed. Oxford: Blackwell.
Gibbard, Allan. (2003). Thinking How to Live. Harvard University Press.
Gibbard, Allan. (2017). Parfit on Normative Concepts and Disagreement. In Does Anything Really Matter? Essays on Parfit on Objectivity, ed. Peter Singer. Oxford, New York: Oxford University Press.
Hare, Richard Mervyn. (1981). Moral Thinking: Its Levels, Method, and Point. Oxford: Clarendon Press.
Hoesch, Matthias, Sebastian Muders and Markus Rüther. (2017). Worauf es ankommt. Hamburg: Meiner.
Jackson, Frank. (1998). From Metaphysics to Ethics. Oxford University Press.
Kirchin, Simon. (2017). Reading Parfit: On What Matters. London: Routledge.
Laskowski, Nicholas. (2018). “Parfit, Derek. On What Matters. Vol. 3. Oxford: Oxford University Press, 2017.” Ethics 128 (2): 496–505. URL: https://www.journals.uchicago.edu/doi/10.1086/694285
Leary, Stephanie. (2017). Non-naturalism and Normative Necessities. In Oxford Studies in Metaethics, Vol.12, ed. Russ Shafer-Landau. Oxford: Oxford University Press.
Mackie, John. (1977). Ethics, Inventing Right and Wrong. London: Penguin Books.
Moore, G.E. (1903). Principia Ethica. London: Cambridge University Press.
Muders, Sebastian and Markus Rüther. (2017). Was ist falsch an ontologisch robusten Tatsachen? Einige Bemerkungen zu Derek Parfits nicht-realistischem Kognitivismus. In Worauf es ankommt. Derek Parfits praktische Philosophie in der Diskussion, ed. Sebastian Muders and Markus Rüther. Hamburg: Meiner.
Parfit, Derek. (2011a). On What Matters, Vol.1. Oxford: Oxford University Press.
Parfit, Derek. (2011b). On What Matters, Vol.2. Oxford: Oxford University Press.
Parfit, Derek. (2017). On What Matters, Vol.3. Oxford: Oxford University Press.
Railton, Peter. (2017). Two Sides of the Meta-Ethical Mountain? In Does Anything Really Matter? Essays on Parfit on Objectivity, ed. Peter Singer. Oxford, New York: Oxford University Press.
Scanlon, Thomas M. (1998). What We Owe to Each Other. London, Cambridge (MA): Harvard University Press.
Scanlon, Thomas M. (2014). Being Realistic About Reasons. Oxford, New York: Oxford University Press.
Schroeder, Mark. (2007). Slaves of the Passions. New York: Oxford University Press.
Schroeder, Mark. (2011). “On What Matters, Volumes I and II.” URL: https://ndpr.nd.edu/news/on-what-matters-volumes-1-and-2/
Seligman, Martin E. P., Peter Railton, Roy F. Baumeister and Chandra Sekhar Sripada. (2016). Homo Prospectus. New York, NY: Oxford University Press.
Singer, Peter. (2017). Does Anything Really Matter? Essays on Parfit on Objectivity. Oxford, New York: Oxford University Press.
Smith, Michael. (1995). The Moral Problem. Oxford: Blackwell.
Smith, Michael. (2017). Parfit’s Mistaken Meta-Ethics. In Does Anything Really Matter? Essays on Parfit on Objectivity, ed. Peter Singer. Oxford, New York: Oxford University Press.
Street, Sharon. (2017). Nothing ”Really” Matters, but That’s Not What Matters. In Does Anything Really Matter? Essays on Parfit on Objectivity. Oxford, New York: Oxford University Press.
Suikkanen, Jussi. (2017). “Non-naturalism and Reference.” Journal of Ethics and Social Philosophy 11.
Acknowledgements
For helpful comments and conversations, I would like to thank Jacob Rosenthal, Peter Stemmer, Felix Timmermann, Christian Wendelborn, the participants of the Konstanz-Zurich Colloquium, and two anonymous referees from Zemo. Also, I am very grateful for a six month fellowship at the Zukunftskolleg Konstanz in 2019. Once more, the Zuko opened doors which, otherwise, I wouldn’t have known existed.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Fischer, S. Still a Misty Mountain: Assessing Parfit’s Non-Realist Cognitivism. ZEMO 2, 213–230 (2019). https://doi.org/10.1007/s42048-019-00044-5
Published:
Issue Date:
DOI: https://doi.org/10.1007/s42048-019-00044-5