Basic-state expressivism: thought, talk and dispositions
A common view is that communication is the transmission of thoughts from one head to another using language as a means. "Snow is white" transfers the belief that snow is white - a state that describes the world. The close connection between moral thinking and motivation, on the other hand, suggests that moral sentences, e.g., "Stealing is wrong" doesn't transfer a belief, but a more emotionally-laden state, e.g., disapproval of stealing. This has led many philosophers, so-called expressivists, to argue that moral sentences function to express (dis)approval and that this explains meaning. This view seems plausible if applied to simple sentences. When embedded, however, problems arise. "If stealing is wrong, then helping to steal is wrong" doesn't express any disapproval, but the meaning of "wrong" doesn't change. This suggests that expressivism is flawed. The point of departure for this project is that there is nothing wrong with the expressivist view per se. It is the standard explanation of the meaning of descriptive sentences in terms beliefs and moral sentences in terms (dis)approvals that is problematic. The purpose of this project is to explore the hypothesis that terms and sentences express more basic dispositional states of mind. The aim is to publish articles that develop this framework while considering standard problems. This promises to contribute both to the development of the expressivist program and understanding of thought and talk more generally.
Final report
How should we think about moral thought and talk? One idea is that moral thought and talk is not, at least not primarily, something that purports to represent the world, but something that has a non-descriptive function. To think that it is wrong to steal, for example, is to have a negative attitude (a kind of disapproval) towards stealing. To say that stealing is wrong is to express disapproval towards stealing. These are central ideas of moral expressivism. Another central idea of moral expressivism is that the meaning of a term or sentence is to be explained by the mental states they express. For example, the meaning of “stealing is wrong” is to be explained by virtue of the mental state that the sentence functions to express, viz., roughly disapproval of stealing. This takes us to a much-discussed problem for expressivist theories. Even if it is plausible to think that “stealing is wrong” expresses disapproval of stealing and thereby get its meaning, it seems that there are contexts where “stealing is wrong” can occur without expressing any negative attitude. A standard example is the following: “If stealing is wrong, then getting your colleague to steal is wrong.” In this context, “stealing is wrong” does not seem to express any negative attitude, but “stealing is wrong” nevertheless seems to have the same meaning whether embedded or not. If this is right, there seems to be something very problematic about the expressivist explanation of the meaning of moral words and sentences.
The overarching purpose of this project was to investigate whether this, and many other, problem for expressivism arises because of the standard expressivist explanation of the meaning of moral terms and sentences. The hypothesis was that there is not really a problem with expressivism as such, i.e., the idea that we should explain the meaning of (moral) terms and sentences in terms of the mental states they express, but that we should understand the mental state that words and sentences express differently. A moral expressivist should not explain the meaning of, for example, “stealing is wrong” in terms of disapproval of stealing. Similarly, we should not understand the meaning of “grass is green” in terms of the belief that grass is green. Instead, the hypothesis was that we should understand the meaning of words and sentences in terms of a more basic state of mind.
Three of the project’s most important results are the following. One result has to do with the project’s most basic hypothesis, viz., that moral sentences, e.g., “stealing is wrong,” express disapproval and thereby acquires its meaning. The first hypothesis of the project was that moral sentences do not express disapproval, but classificatory dispositions that functions to regulate moral attitudes. While working on this project, this hypothesis has in many respects changed. One result of the project is that we should think about the meaning of “wrong” as functioning to emotionally mark whatever “wrong” is predicated about. To think that N is wrong is, very briefly, to emotionally mark N in a particular negative way. To emotionally mark N thus functions to present N for us in a particular way. It is important to emphasize that an emotion is not the same as having a feeling. An emotion is rather a kind of bodily change with accompanying action-tendencies. This is not the same thing as disapproving of something, but something more basic. The hypothesis is that we should think of the meaning of moral sentences in terms of expressing these more basic mental states. Regardless of whether “wrong” occurs in “stealing is wrong” or ““If stealing is wrong, then getting your colleague to steal is wrong” it functions to emotionally mark whatever it is predicated about. A consequence of this is that the project had an emotivist turn. Moral thought and talk is intimately connected to emotions. A different result has to do with the nature of moral judgments. For an expressivist, moral judgments are not beliefs, they do not aim at truth or to represent some kind of moral reality. By contrast, cognitivists think that moral judgments are beliefs, that they aim at truth and to represent some kind of moral reality (although there is much disagreement about how to understand “moral reality” and indeed whether what moral judgments purport to represent exists). If moral judgments are beliefs, then it seems plausible to think that we would change our judgments in response to changes in our evidence. For example, if a person judges that an act is wrong, e.g., cannibalism, because it causes harm, but learns that the act is harmless, chances are that the person nevertheless maintains the moral judgment. In other words, people sometimes are morally dogmatic. This may provide clues regarding the nature of moral judgments. One idea is that the conceivability of moral dogmatists gives us reason to think that moral judgments are not beliefs – many philosophers think that beliefs are a kind of mental state that changes in response to changes in evidence. A final result concerns the nature of normativity. Questions regarding the nature of normativity are many and controversial, but a common view, that fits well with expressivism, is that the internal connection between moral judgments and motivation is the key. An alternative, and, I believe, a more plausible explanation connects to the hypothesis above. “Wrong” functions to emotionally mark whatever “wrong” is predicated about in such a way that this is presented to us in a particular way: it acquires a particular role in our deliberation – a role that it would have had if it was not for our emotions.
Two examples of new research questions are the following. 1. What is it to have a moral belief? Cognitivists usually do not say much about this beyond saying that beliefs aim at truth or have a particular direction of fit. But in what sense are beliefs evidence responsive? This is a central in relation to the challenge from moral dogmatism, but also interesting more generally in relation to the nature of belief. 2. How is it possible to negotiate about the meaning (metalinguistic negotiation) and thereby change the concept we live by? This connects to debated regarding so-called “conceptual engineering.” However, in the debate regarding metalinguistic negotiation and conceptual engineering, the emotional dimension is remains largely unaddressed. Given the hypothesis that “wrong” functions to emotionally mark whatever “wrong” is predicated about, an interesting question is how we use language, gestures, pictures and more to emotionally mark different things (both intentionally and unintentionally) and how this can result in associations between object and events and emotions (again, to emotionally mark something is not for it to feel in a particular way). This may explain why many purportedly descriptive words are actually evaluative and the kind of mechanisms that we use to make certain terms and concepts evaluative.
Many central thoughts and ideas of the project have been presented at various conferences nationally and internationally. I have presented material relevant for the project at conferences in Stockholm, Lisbon, Kent, Bratislava, Paris and Madison (online). Part of the result of the project have also been published in peer reviewed journals. There are also a number of drafts which I hope to publish in the near future.
One possible development of the project is to more thoroughly develop the emotivist idea outlined above, and examine how we in different ways negotiate emotional components of thought and talk, and how this may function to create terms and concepts that are, at least in part, evaluative, i.e., on the present hypothesis, connected to emotions.
The overarching purpose of this project was to investigate whether this, and many other, problem for expressivism arises because of the standard expressivist explanation of the meaning of moral terms and sentences. The hypothesis was that there is not really a problem with expressivism as such, i.e., the idea that we should explain the meaning of (moral) terms and sentences in terms of the mental states they express, but that we should understand the mental state that words and sentences express differently. A moral expressivist should not explain the meaning of, for example, “stealing is wrong” in terms of disapproval of stealing. Similarly, we should not understand the meaning of “grass is green” in terms of the belief that grass is green. Instead, the hypothesis was that we should understand the meaning of words and sentences in terms of a more basic state of mind.
Three of the project’s most important results are the following. One result has to do with the project’s most basic hypothesis, viz., that moral sentences, e.g., “stealing is wrong,” express disapproval and thereby acquires its meaning. The first hypothesis of the project was that moral sentences do not express disapproval, but classificatory dispositions that functions to regulate moral attitudes. While working on this project, this hypothesis has in many respects changed. One result of the project is that we should think about the meaning of “wrong” as functioning to emotionally mark whatever “wrong” is predicated about. To think that N is wrong is, very briefly, to emotionally mark N in a particular negative way. To emotionally mark N thus functions to present N for us in a particular way. It is important to emphasize that an emotion is not the same as having a feeling. An emotion is rather a kind of bodily change with accompanying action-tendencies. This is not the same thing as disapproving of something, but something more basic. The hypothesis is that we should think of the meaning of moral sentences in terms of expressing these more basic mental states. Regardless of whether “wrong” occurs in “stealing is wrong” or ““If stealing is wrong, then getting your colleague to steal is wrong” it functions to emotionally mark whatever it is predicated about. A consequence of this is that the project had an emotivist turn. Moral thought and talk is intimately connected to emotions. A different result has to do with the nature of moral judgments. For an expressivist, moral judgments are not beliefs, they do not aim at truth or to represent some kind of moral reality. By contrast, cognitivists think that moral judgments are beliefs, that they aim at truth and to represent some kind of moral reality (although there is much disagreement about how to understand “moral reality” and indeed whether what moral judgments purport to represent exists). If moral judgments are beliefs, then it seems plausible to think that we would change our judgments in response to changes in our evidence. For example, if a person judges that an act is wrong, e.g., cannibalism, because it causes harm, but learns that the act is harmless, chances are that the person nevertheless maintains the moral judgment. In other words, people sometimes are morally dogmatic. This may provide clues regarding the nature of moral judgments. One idea is that the conceivability of moral dogmatists gives us reason to think that moral judgments are not beliefs – many philosophers think that beliefs are a kind of mental state that changes in response to changes in evidence. A final result concerns the nature of normativity. Questions regarding the nature of normativity are many and controversial, but a common view, that fits well with expressivism, is that the internal connection between moral judgments and motivation is the key. An alternative, and, I believe, a more plausible explanation connects to the hypothesis above. “Wrong” functions to emotionally mark whatever “wrong” is predicated about in such a way that this is presented to us in a particular way: it acquires a particular role in our deliberation – a role that it would have had if it was not for our emotions.
Two examples of new research questions are the following. 1. What is it to have a moral belief? Cognitivists usually do not say much about this beyond saying that beliefs aim at truth or have a particular direction of fit. But in what sense are beliefs evidence responsive? This is a central in relation to the challenge from moral dogmatism, but also interesting more generally in relation to the nature of belief. 2. How is it possible to negotiate about the meaning (metalinguistic negotiation) and thereby change the concept we live by? This connects to debated regarding so-called “conceptual engineering.” However, in the debate regarding metalinguistic negotiation and conceptual engineering, the emotional dimension is remains largely unaddressed. Given the hypothesis that “wrong” functions to emotionally mark whatever “wrong” is predicated about, an interesting question is how we use language, gestures, pictures and more to emotionally mark different things (both intentionally and unintentionally) and how this can result in associations between object and events and emotions (again, to emotionally mark something is not for it to feel in a particular way). This may explain why many purportedly descriptive words are actually evaluative and the kind of mechanisms that we use to make certain terms and concepts evaluative.
Many central thoughts and ideas of the project have been presented at various conferences nationally and internationally. I have presented material relevant for the project at conferences in Stockholm, Lisbon, Kent, Bratislava, Paris and Madison (online). Part of the result of the project have also been published in peer reviewed journals. There are also a number of drafts which I hope to publish in the near future.
One possible development of the project is to more thoroughly develop the emotivist idea outlined above, and examine how we in different ways negotiate emotional components of thought and talk, and how this may function to create terms and concepts that are, at least in part, evaluative, i.e., on the present hypothesis, connected to emotions.