Abstract
One of the most discussed challenges to metaethical expressivism is the embedding problem. It is widely presumed that the reason why expressivism faces this difficulty is that it claims that moral sentences express non-cognitive states, or attitudes, which constitute their meaning. In this paper, it is argued that the reason why the embedding problem constitutes a challenge to expressivism is another than what it usually is thought to be. Further, when we have seen the real reason why expressivism is vulnerable to this difficulty, it is plausible to argue that it makes up a challenge to certain other metaethical views as well.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
In what follows, I will take this type of traditional semantics to state that the semantic contents of purely descriptive sentences consist in propositions, but my arguments can be formulated more generally in terms of truth conditions.
Thus, we should distinguish between metaethical expressivism, which is a theory about the meaning of moral sentences, and expressivism as a general semantic view, which can be applied to both moral and non-moral sentences. In what follows, ‘expressivism’ refers to the first notion unless I indicate otherwise. The idea that the meaning of a sentence consists in the mental state it expresses can be explicated in different manners. One way, which has become influential in metaethics, explains this in terms of assertability or acceptability conditions. Thus understood, the meaning of a sentence consists roughly in the mental state in which a person needs to be, in order for it to be semantically appropriate to assert or accept the sentence. See Schroeder (2008): Ch. 2. Cf. Ridge (2003): 563–574, and Kalderon (2005): Ch. 2. For an alternative account, see e.g. Gibbard (2003): Ch. 4.
It might be objected that there are non-cognitive mental states which do not have these features. However, it seems uncontroversial to assume that on expressivism, moral sentences express a type of non-cognitive states that do have these features, and this is all that is needed for the arguments in subsequent sections.
Importantly, ‘approval’ or ‘disapproval’ does not refer to a non-cognitive mental state in this context, but to a mental state which is such that if a person is in it, she approves or disapproves of something.
I follow the metaethical convention of letting ‘judge’ and ‘judgment’ refer to mental states.
Geach (1965): 449–465. Cf. Searle (1962): 449–465. In the first formulations of the problem, it was directed against views that conceive of the meaning of sentences in terms of what speech acts they are used to perform. It was later argued that it also applies to expressivism, which is the modern descendant of these views. See e.g. Schueler (1988): 492–493; Dreier (1996): 30–32; van Roojen (1996): 313–314; Unwin (1999): 337–338; Horwich (2005): 78–80; Eklund (2009): 705–706; Weintraub (2011): 601–602; Kurth (2011): 161–180.
For example, it would be problematic to explain how a moral sentence can make systematic and predictable contributions to the meanings of the various complex sentences of which it is part. Cf. Schroeder (2010): 47–54.
Cf. Schroeder (2008): 21.
It may perhaps be suggested that if expressivists are able to explain how the meaning of a complex sentence consists in a mental state which is a function of the mental states which constitute the meanings of its part, it is inconsequential if it turns out that a moral sentence does not express the same mental state in all sentential contexts. However, there seems to be prima facie reason to prefer a view which evades this problem even if it succeeds in other respects.
It should be stressed that the phrase ‘disapproval of lying’ does not refer to a non-cognitive mental state in this context, but to a mental state which is such that a person who is in it disapproves of lying. This becomes clear when Sinnott-Armstrong clarifies the argument with saying that one could assert (2) whether or not one disapproves of lying. If ‘disapproves of lying’ refers to a non-cognitive state in this passage, the argument would be question begging.
On certain versions of expressivism, this sentence does not express a non-cognitive state towards lying, but a second order non-cognitive state, or some other more complex non-cognitive state, with regard to lying. As a result, the fact that a person is in this mental state would not entail that she disapproves of lying. This complication is however insignificant in the present context. What is important is that expressivism entails that a sentence of the type ‘φing is wrong’ consists in a non-cognitive state which is such that if a person is in this state, she disapproves of something.
Needless to say, the intrinsic claim is not the only conceivable view that would be vulnerable to the embedding problem, since we can imagine all types of claims according to which a given sentence has different meaning in different sentential contexts. The question before us is whether there are metaethical views other than expressivism that entail the intrinsic claim and on these grounds can be challenged by this difficulty.
Other type of views that entail the intrinsic claim are hybrid theories according to which the meaning of a sentence such as ‘φing is wrong’ is constituted by a cognitive state (belief) in combination with a non-cognitive state (desire). See e.g. Ridge (2006): 302–336. It has been argued that such views also are susceptible to the embedding problem; see Strandberg (2014). Cf. Schroeder (2009): 264–275. It might be argued that the embedding problem generalizes in other manners as well; see e.g. Eklund (2009): 705–712.
For an account of the nature of this kind of mental state, see Millikan (1995): 185–200. Among the authors that can be interpreted to embrace this view, see e.g. McDowell (1978): 13–29; Little (1997): 59–79; Bedke (2009): 189–209. Cf. Tenenbaum (2006): 235–264. Gunnar Björnsson has briefly suggested that a view according to which moral judgments consist in this kind of mental states might be vulnerable to a difficulty akin to the embedding problem: Björnsson (2001): 87.
Another objection is that the present view can avoid the embedding problem in the same manner as hybrid theorists want to avoid the Frege-Geach problem. (For a hybrid theoretical solution to this problem, see e.g. Ridge (2006): 324–330.) On this objection, the present view can explain the validity of moral arguments such as (1)–(3) by appealing to the fact that moral sentences have truth values. Thus, (3) follows from (1) and (2) because (3) cannot be false if (1) and (2) are true. First, my point is not that it would be impossible for the present view to avoid the embedding problem, but rather that this problem constitutes a challenge to the present view for the same reason, and to the same extent, as it constitutes a challenge to a view according to which moral sentences express non-cognitive states, either wholly or partly. Second, it has been argued that the indicated proposal is insufficient to account for the validity of moral arguments (Strandberg (2014)). Third, it should be observed that the mentioned proposal exclusively addresses validity. It does not address the generic problem how a moral sentence can be taken to express a mental state that involves approval or disapproval in different sentential contexts. So, even if it would succeed to account for validity, it is not obvious that it avoids the embedding problem as understood here.
Schroeder (2008): 23.
Schroeder maintains that according to an expressivist semantics, sentential connectives need to operate on mental states rather than on propositional contents (Schroeder (2008): esp. Ch. 7). This contention rests on the assumption that an expressivist semantics is combined with metaethical expressivism. Metaethical expressivists need to explain the meaning of a complex moral sentence (e.g. ‘If it is wrong to lie, then it is wrong to get one’s little brother to lie’) in terms of sentential connectives operating on the non-cognitive mental states that constitute the meanings of the sentences which constitute its parts. Now, there are also complex sentences which combine moral sentences and purely descriptive sentences (e.g. ‘If it is wrong to lie, then few will do so’). In order to explain the meaning of such ‘mixed’ complex sentences, metaethical expressivists need to maintain that sentential connectives quite generally operate on mental states (non-cognitive states and beliefs) rather than on propositional contents. However, if an expressivist semantics instead is combined with metaethical cognitivism it might be argued that it need not understand sentential connectives as operating on mental states, but should be free to argue that they operate on propositional contents. The basic reason is that on this view it would be possible to explain compositionality in terms of the propositional contents of sentences rather than in terms of the beliefs they express.
It might be wondered whether this argument can be directed against metaethical views which understand meaning in some other manner than in terms of expressed mental states. For example, I have in mind a view according to which the meaning of a moral sentence of the type ‘φing is wrong’ consists in a proposition which makes beliefs that have this proposition as their objects motivating. This issue requires an extended discussion that has to be postponed for another occasion.
References
Bedke, M. S. (2009). Moral judgment, purposivism: saving internalism from amoralism. Philosophical Studies, 144, 189–209.
Björnsson, G. (2001). Why emotivists love inconsistency. Philosophical Studies, 104, 81–108.
Blackburn, S. (1998). Ruling passions. Oxford: Clarendon Press.
Charlow, N. (2014). The problem with the Frege–Geach problem. Philosophical Studies, 167, 635–665.
Chrisman, M. (2011). Expressivism, inferentialism and the theory of meaning. In M. Brady (Ed.), New Waves in Metaethics (pp. 103–125). New York: Palgrave.
Dreier, J. (1996). Expressivist embeddings and minimalist truth. Philosophical Studies, 83, 29–51.
Eklund, M. (2009). The Frege-Geach problem and Kalderon’s moral fictionalism. Philosophical Quarterly, 59, 705–712.
Eriksson, J. (2009). Homage to Hare: ecumenism and the Frege-Geach problem. Ethics, 120, 8–35.
Geach, P. T. (1965). Assertion. Philosophical Review, 74, 449–465.
Gibbard, A. (1992). Reply to Blackburn, Carson, Hill and Railton. Philosophy and Phenomenological Research, 72, 969–980.
Gibbard, A. (2003). Thinking how to Live. Cambridge, Mass.: Harvard University Press.
Hale, B. (1986). The compleat projectivist. Philosophical Quarterly, 36, 65–84.
Horwich, P. (2005). The Frege-Geach point. Philosophical Issues, 15, 78–93.
Kalderon, M. E. (2005). Moral fictionalism. Oxford: Clarendon.
Kurth, C. (2011). Logic for morals, morals from logic. Philosophical Studies, 155, 161–180.
Little, M. O. (1997). Virtue as knowledge: objections from the philosophy of mind. Noûs, 31, 59–79.
McDowell, J. (1978). Are moral requirements hypothetical imperatives? The Aristotelian Society, 52, 13–29.
Miller, A. (2003). An introduction to contemporary metaethics. Cambridge: Polity Press.
Millikan, R. G. (1995). Pushmi-Pullyu representations. Philosophical Perspectives, 9, 186–200.
Ridge, M. (2003). Non-cognitivist pragmatics and Stevenson’s ‘“Do so as well!”’. Canadian Journal of Philosophy, 4, 563–574.
Ridge, M. (2006). Ecumenical expressivism: Finessing Frege. Ethics, 116, 302–336.
Schroeder, M. (2008). Being For. Oxford: Oxford University Press.
Schroeder, M. (2009). Hybrid expressivism: virtues and vices. Ethics, 119, 257–309.
Schroeder, M. (2010). Noncognitivism in Ethics. New York: Routledge.
Schueler, G. F. (1988). Modus Ponens and moral realism. Ethics, 98, 492–500.
Searle, J. (1962). Meaning and speech acts. Philosophical Review, 71, 423–432.
Sinnott-Armstrong, W. (2000). Expressivism and embedding. Philosophy and Phenomenological Research, 61, 677–693.
Strandberg, C. (2014). Options for hybrid expressivism. Ethical Theory and Moral Practice. doi: 10.1007/s10677-014-9511-2.
Tenenbaum, S. (2006). Direction of fit and motivational cognitivism. In R. Shafer-Landau (Ed.), Oxford Studies in Metaethics (Vol. 1, pp. 235–264). Oxford: Oxford University Press.
Unwin, N. (1999). Quasi-realism, negation and the Frege-Geach problem. Philosophical Quarterly, 49, 337–352.
van Roojen, M. (1996). Expressivism and irrationality. Philosophical Review, 105, 311–335.
Wedgwood, R. (2007). The nature of normativity. Oxford: Oxford University Press.
Weintraub, R. (2011). Logic for expressivists. Australasian Journal of Philosophy, 89, 601–616.
Zangwill, N. (1992). Moral Modus Ponens. Ratio, 5, 177–193.
Acknowledgments
I am very grateful to the audiences in Umeå, Gothenburg, Eindhoven, Oslo and Stockholm, for valuable discussions about earlier drafts of this paper. Special thanks are due to John Eriksson and Sara Packalén for their helpful comments.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Strandberg, C. Can the Embedding Problem Be Generalized?. Acta Anal 30, 1–15 (2015). https://doi.org/10.1007/s12136-014-0232-x
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12136-014-0232-x