Sunday, October 22, 2006

Some Considerations on Non-Cognitivism

I’m sorry that this is a day late.

I spent the day with my wife. We went to see the movie “Flyboys.” There was a scene in the movie that was nice to see. One of the pilots – a competent and brave pilot – got killed in the movie. He left a letter behind that said, “I have no religion and I do not request any service…” That was a refreshing statement to hear in a movie.

Anyway, I did not get home in time to write yesterday’s posting, so I wrote it today.

It concerns comments made on my posts last weekend on subjectivism that I ignored non-cognitivism as a viable form of subjectivism. Non-cognitivism differs from other forms of subjectivism in that it denies that moral claims are truth-bearing propositions at all. So, any objection that I have to ‘subjective truth’ does not apply to this form of subjectivism.

Interrogatories and commands are two examples of statement types that do not have a truth value.

True or false. What color is my car?

This makes no sense. The question, "What color is my car?" is not the type of statement that is capable of being true or false.

The same is true of commands.

True or false. Close the door.

Moral statements, it is argued, represent a third class of statements incapable of being true or false. A moral statement is like laughing at a joke, or crying at the end of a movie, or shouting, "Ow!" when you hit your thumb with a hammer, or screaming and jumping on the bed at the sign of a large spider.

Moral Claims Universally Treated as Truth-Bearing Propositions

The first problem with this view is that it fails the first and most direct test that we can use to categorize statements as true or false.

Hand somebody a list of interrogatories and ask them if they are true or false and they will stare at you with incomprehension.

The same will happen if you hand them a list of commands.

However, if you hand them a list of moral statements; "Abortion is always wrong," "Capital punishment is never justified," "Society should not permit marriage between two people of the same gender," they will take their pencil and start filling in answers.

If it is a mistake to assign truth value to moral claims then 99.99% of the population capable of making coherent statements in any language is making this mistake. In fact, I would put the level of error at 100%. Even those who claim to be non-cognitivists, I assert, if you look at their use of sentences when they are not thinking about their theories - the value claims they make in every day life - they find themselves constantly claiming that value statements are true or false.

This is because they ARE true or false.

'True Of Me' vs. 'True For Me'

I need to make a quick digression because a lot of people get the impression that if somebody says that a value claim is objectively true, this means that everybody experiences the same value.

This argument is no more true than saying, "If location statements are objectively true or false, then everybody should be standing at exactly the same spot," or "If height and weight statements are objectively true or false, then everybody should have exactly the same height and weight."

Some value statements are objectively true or false statements about the speaker, and make no inference that others should have the same quality.

My statement that I crave chocolate is exactly like the statement that I am 6'1" tall. I am saying something that is objectively true of me. It is also objectively true of chocolate that it is something that I crave - just as it is objectively true of 6'1" that it is the same as my height.

Many value claims are claims like this that tell the listener what is true of the speaker.

However, there is nothing in this that is true for me that is not true for everybody else. That I am 6'1" tall is true for everybody. It is not true of everybody that they are over 6' tall. It makes no sense to say that it is true of everybody that I am over 6' tall. However, it makes perfectly good sense to say, and it is objectively true for everybody that I am over 6' tall.

As recently as yesterday I criticized the idea that moral statements are statements about what is true of the speaker. In an analogy to location, I assert that moral statements are more like claims about the center of population - something that depends on the location of each individual in the group but considers the location of everybody in the group. Morality depends on the desires of each individual but considers the desires of all individuals. Thus, moral claims are claims of the form, true of us more than true of me.

Propositions, Meaning, and Reference

Richard told me last week that, "Ewwww" is not a statement at all. He is correct. It does not have a subject or a verb. It is more like a basic noise - like a hiccup or a sneeze, or the reflex one gets when the doctor pounds on that nerve just below the knee with his hammer.

However, just like hiccups and sneezes the "Ewwww" can be fully captured by a set of objectively true propositions. There is "something going on." Our explanation of that "something" will be a set of objectively true propositions - as long as we have the right explanation. There is no room for anything other than objectively true or false propositions in our explanation.

Furthermore, that explanation does not require any type of 'subjective reality'. We do not need to invent such an entity to capture any real-world phenomenon. Objective reality is quite sufficient.

Clearly, if a person chops somebody up with an axe for raping his daughter, this is something going on that is not fully captured by the statement, "He is expressing an opinion that rape is wrong." The event can, in fact, be described in terms much like that of the person who says, "Ow!" when he pounds on his thumb with a hammer, or laughs or cries at a show. He is expressing a physical reaction.

Yet, the event, like everything else in the physical universe, can still be entirely accounted for by a set of objective propositions. There is no need to invent other types of entities.

Explaining vs Analyzing Speech Acts

Here, I want to draw a distinction between explaining a speech act and analyzing a proposition.

Imagine a math teacher standing in front of a classroom. The teacher teaches Pythagoras' Theorem (that, on a right triangle, the square of the length of they hypotenuse is equal to sum of the squares of the two sides).

In order to explain this action (why that person uttered that claim at that time), we would look at his beliefs and desires. For example, we may discover that the teacher wants to teach geometry, wants to continue earning a paycheck, believes that this theorem is true, and believes that this is the point at which it is appropriate to teach this theorem.

Explanations of speech acts are 'subjective' in that the explanation always makes reference to the mental states of the agent.

However, we need to distinguish this from an assessment of what the statement means. The meaning of Pythagoras’ Theorem has nothing to do with the teacher’s beliefs and desires. It is a statement about triangles.

Similarly, if we look at an act of a person praising or condemning another individual, the explanation for that act must refer to the beliefs and desires of the agent. If we mistake the causes of a speech act from its meaning, given that the proximate cause must be the beliefs and desires of the agent, we are certainly going to think that some form of subjectivism is correct. The person who chops at his daughter’s alleged rapist may well believe that he deserves to be chopped up, and his desire to chop the man up may be stronger than his aversion to the consequences. Yet, this is the explanation for an action, not an assessment as to its meaning.

Moral Statements as Objects of Propositional Attitudes

We are still left with the question, “If a person believes that X is immoral, what does he believe?”

The mere fact that this is a sensible question is an issue for the noncognitivist. It makes no sense to ask the question, “If a person believes how old are you anyway?” or “If a person believes shut the door.” Interrogatories and commands cannot be used as the object of a ‘belief’ statement, because belief statements are propositional attitudes, and they require a proposition as its object.

We have the same problem using other non-cognitivist alternatives as objects for belief claims. A person cannot ‘believe that ewwww’ or ‘believe that ’ or ‘believe that ’.

Beliefs are propositional attitudes. Beliefs take propositions as their objects. The fact that a person can believe that abortion is immoral, or believes that there should be separation between church and state, suggests that ‘abortion is immoral’ and ‘there should be separation between church and state’ are propositions. That is to say, they are statements capable of being true or false. These statements are filling a role that only cognitive propositions can fill. This suggest that moral statements are cognitive (truth-bearing statements).


In the space of a posting it is difficult to write a full argument on any of these topics. Here, I hoped to raise a number of concerns that I want people to keep in mind as they assess non-cognitivism.

Keep in mind the fact that almost everybody, if not everybody, if presented with a list of moral statements, understand that it makes sense to answer that those statements are true or false.

There is a distinction between ‘true of me’ and ‘true for me’ and, while the former makes sense, there is no need for the latter.

Even non-cognitivist utterances can be fully captured by a set of objectively true or false statements that fit the utterance into a chain of cause and effect.

If we mistake the causes of a speech act (which must necessarily be the beliefs and desires of the speaker) with its meaning, we may become confused into believing that its meaning depends on the beliefs and desires of the speaker.

Moral statements can fill roles that only propositional attitudes can fill – such as being the object of a propositional attitude - a ‘belief’ or a ‘desire’.

Just keep these facts in mind and be ready to use them when analyzing the claims of somebody who defends non-cognitivism in ethics.


Staircaseghost said...

I see you're still using the same old examples and exhibiting the same misunderstandings of the position you want to argue against. Noncognitivism is a semantic hypothesis, not a syntactic one -- "but people say they're true or false" is irrelevant. To say that because whenever S is engaged in Θ, the process of Θing can be described in terms of a set P of truth-functional propositions is not at all the same thing as saying that whenever S Θ's, S is asserting that everything in P is true. (When my Bible falls off my shelf, that does not mean that my Bible is "asserting that" Newton's theory of gravity is true).

Eventually you're going to have to engage with some good introductory texts to get a handle on the actual position instead of this tilting at windmills (what on earth does the "true for me/true of me" distinction have to do with noncognitivism?). Here's a good bibliography to get you started (going pseudochronologically):

Stevenson, C.L. 'The Emotive Meaning of Ethical Terms'

-- 'The Emotive Conception of Ethics and its Cognitive Implications'

Hare, R. M. The Language of Morals

Blackburn, S. Spreading the Word

-- Essays in Quasi-realism

[see especially in the above 'Errors and the Phenomenology of Value', 'How to be an Ethical Anti-realist', and 'Attitudes and Contents']

Gibbard, A. Wise Choices, Apt Feelings

-- Thinking How to Live

And going back to your sort of general confusion of the ways in which one might deny various aspects of objectivity about morals, an indispensable review of the contemporary metaethical landscape (available online as a pdf) is Darwall, Gibbard, & Railton's 'Towards Fin de Siecle Ethics: Some Trends'. This was reprinted in the great textbook edited by the same three authors, Moral Discourse and Practice, along with dozens of representative essays from the varying positions currently active in the literature; definitely worth picking up from your library or Amazon.

Good luck, and happy reading!

Anonymous said...

Good reading, as always (I almost feel like I should comment from time to time just because the blog is such a deep read and yet there are so few comments...).

Of course, I s'pose it would be great if I could contribute something and not leave simply comments like these. For now, my forays in computer science and arts share little in common with your writing and I have precious little to contribute.

Alonzo Fyfe said...


Thanks for the note.

As a college student, I was forced to take classes that I though had nothing to do with what I was interested in - so I hated those classes.

Yet, twice, these 'unrelated classes' gave me pieces of information that significantly changed the core ideas that are the foundation of my writing.

One of those was the computational theory of mind - the way that minds are like computers. This, combined with some studies in the philosophy of psychology (which I also thought would be a waste of time) brought me to accept the idea that the brain was like a computer and beliefs and desires the programming language for the human brain.

There is, indeed, a lot of study going on in this area.

Alonzo Fyfe said...


Since I saw your comment to law week's postings, I used these examples specifically to help you feel more at home.

Anyway, I have, of course, already read Hare and Blackburn and Stevenson. In fact, I talked with Hare a bit. He came to my college to give a philosophy lecture. The department also arranged for a private reading of a philosophy paper - by invitation only. I was one of eight people invited to hear a paper on prescriptivism.

There is a lost in common between Hare's two-level theory of ethics (the 'archangel' and the 'prole') that gets folded into the idea that morality is ultimately concerned with the evaluation of desires. Maybe I'll write a post on it in the near future.

However, it was in studying Hare that I came up with the objection that it makes no sense to write fundamental 'ought' premises into an argument in order to get an 'ought' conclusion. These 'ought' premises are nothing more than pulling a rabbit out of a hat in order to get a desired conclusion. There's no justification for it.

So, if we can't derive 'ought' from 'is' - then we have no reason to keep 'ought' around at all. We have no reason not to discard it and toss it in the same bin with God, devils, souls, free will, and other mythical entities that more primative cultures once accepted.

As for Stephenson's emotivism, the points I made above (particularly the first and last point) sufficiently refute that thesis.