Your complimentary articles
You’ve read one of your four complimentary articles for this month.
You can read four articles free per month. To have complete access to the thousands of philosophy articles on this site, please
The Cognitive Gap
Justin Bartlett explores a basic distinction between understandings of ethics.
We are all concerned, to a greater or lesser degree, with ethical issues. Whether it be concerns over crime and punishment, humanitarian aid, ecological destruction, or simply the fact that your friend has broken a promise, ethical considerations seem to creep into almost all areas of life. But what is it that makes humans prone to thinking in terms of right and wrong or good and bad? Why are we psychologically predisposed to make judgements of the moral variety, and what does it actually mean to make a moral judgement? These are what philosophers might call ‘second-order’ moral questions. They are among the chief concerns of the philosophical sub-field known as ‘metaethics’.
© Susan Platow 2023 Please visit smplatow.com
The Lay of the Land
Metaethics as a distinct discipline is little over a century old. It is generally considered to have begun with the book Principia Ethica written by G.E. Moore in 1903. In it Moore stressed the difference between thinking about which things are good (ethics), and thinking about what ‘good’ means (metaethics). Since then, many moral judgements have been metaethically sliced, diced and analysed. Indeed, metaethics has been a battlefield. Metaphysical muddles, semantic silliness, and logical log-jams have made this terrain rather muddy but it is far from being a wasteland. In my estimation the battle has been dominated by two opposing sides, and, as we will see, there is a vast conceptual chasm between them. On the one side, we have the cognitivists, and on the other the non-cognitivists. These two factions of metaethicists are sitting on either side of what we could call the cognitive gap. But what sets them apart?
Let us start with the cognitivists. Cognitivists come in many shades, but they all share some common ideas about moral talk. They all agree that moral judgments express cognitive (that is, intellectual) states which track mind-independent moral properties such as rightness and wrongness. These moral properties are thought by the speakers to be part of objective reality. Cognitivists, then, claim that moral sentences are truth-apt; they express truths (or falsehoods) about the world. Hence, when someone says ‘Killing is wrong’, then according to cognitivists the speaker is describing the act of killing as really having the objective, opinion-independent property of wrongness – just as when someone says ‘Grass is green’ they are describing the grass as really being green. The speaker is simply describing the truth of the situation as they see it.
Non-cognitivists argue, conversely, that emotional and expressive (that is, non-intellectual) mental states must be underlying moral language. Non-cognitivists also come in many flavours, but all of them reject the cognitivist claim that moral language attempts to report objective moral facts. Instead, non-cognitivists interpret moral sentences such as ‘Killing is wrong’ in a totally different way to physical factual statements such as ‘Grass is green’. By the non-cognitivist interpretation, we are not picking up on any moral facts of the situation when we make moral assertions, because, many non-cognitivists say, there aren’t any objective moral facts at all. Instead, we are simply expressing our personal emotions or convictions about moral issues. Non-cognitivists claim that by saying ‘Killing is wrong’, we in fact mean something like ‘Boo to killing!’ or ‘Don’t kill!’
Strengths & Weaknesses
With these two polarised positions in mind, it may seem that it should be easy to spot which one of them is wrong when contrasted with the other view. Unfortunately, this is not the case. Even today the situation is one of stalemate; neither side has really conquered the other. I will endeavour to briefly outline why this is so, by looking at the main obstacles which stand in the way of each side.
Cognitivism makes claims that might seem a little far-fetched to some. Claims of objective morality might seem intuitively wrong to those people, or possibly even distasteful, if they’re seen as asserting an absolute morality that must be applied to all. But such feelings are not themselves strong philosophical arguments against cognitivism. Any actual arguments behind them would still need to be articulated.
One of the biggest problems to haunt cognitivist theories has been that if moral language attempts to express objective moral facts, then we need strong metaphysical and epistemological accounts of such facts. In other words, cognitivists need to answer such questions as, ‘In what do moral facts consist?’ (or ‘What is the nature of moral truths?’) and ‘How do we come to know these facts?’.
On the positive side, cognitivism can answer some questions in a seemingly simple and easy way. Thus the cognitivist account gives a straightforward answer to why we speak in terms of right and wrong: we are simply assessing and reporting on the facts of the matter as we see them – we observe moral reality and report it. In this way, cognitivism also avoids a certain logical problem that has plagued non-cognitivist theories for the past sixty years, as we will see.
Non-cognitivist theories of ethics may seem at first glance to have something intuitively appealing about them. For example, we might feel angry at someone and express that anger by telling them that their action was wrong – this certainly happens – and then think that ethics in general works this way. Moreover, a real strength of non-cognitivism is that it doesn’t require us to explain, find, or believe in objective moral facts, and hence does not require answers to those questions about how we come to know such facts. Non-cognitivists can just say, “There are no moral facts, so we don’t need to know about them.”
This shedding of the metaphysical and epistemological burdens of cognitivism is advantageous for non-cognitivism, but the expressivist view then runs straight into one of the biggest problems in metaethics: the Frege-Geach Problem. This problem was first explained by Peter Geach in 1963. It concerns the semantics of moral sentences when interpreted in the non-cognitivist or expressivist way.
Let’s go back to our example where a non-cognitivist asserts that ‘Killing is wrong’ really means something like ‘Boo to killing!’. This interpretation apparently works for simple moral assertions such as this; but when we start to construct more sophisticated sentences and try to use the same substitution, we run into some serious logical problems.
Let’s look at a simple logical inference first using a cognitivist interpretation:
P1: Killing is wrong.
P2: If killing is wrong, then getting your little brother to kill is wrong.
C: Therefore, getting your little brother to kill is wrong.
This argument gives us a valid inference, meaning that if we accept both P1 and P2, we are justified in drawing the conclusion C. The argument works because the sentence ‘Killing is wrong’ holds the same meaning throughout it. This is just fine for the cognitivists, as they believe that moral language functions in the same way as other language about reality. But let’s look at this from an expressivist perspective. Let us suppose that the non-cognitivist says that the meaning of the assertion ‘Killing is wrong’ is in fact ‘Boo to killing!’. So we can put this phrase as P1, and so on. Doing so, we get:
P1: Boo to killing!
P2: If ‘Boo to killing!’, then getting your little brother to kill is wrong.
C: Therefore, getting your little brother to kill is wrong.
We can see that in P2, in the first case, ‘Killing is wrong’ is part of a conditional “if… then…” phrase, and makes sense there. But in the second argument the original moral judgement ‘killing is wrong’ has been translated as ‘Boo to killing!’ and so P2 of the second argument makes no sense. In fact, the rule turns out to be that non-cognitivist, or expressivist, translations of the meanings of moral claims only work when uttered as assertions, as in P1, and not when unasserted or embedded in larger sentence structures such as conditionals as in P2. And, thus, the validity of the argument on such an interpretation appears suspect, as it commits the fallacy of equivocation.
The Frege-Geach Problem is a formidable issue for the expressivist, and has generated vast amounts of literature since it was first formulated. It is the non-cognitivist’s responsibility to give an account of the semantics of expressivist moral language that would overcome this problem.
So where do we go from here? As the two sides have drawn down upon each other, have we reached an unavoidable stalemate, or are we beginning to see the possibility of a victory for one side, or perhaps a reconciliation between the two? I claim the latter, which will appear in the shape of a hybrid theory – a theory that takes the strongest claims from both sides of the cognitivist gap and reconciles them into one all-encompassing account of moral language. But, although much work is being done in this area, we are yet, I believe, to find the hybrid model that effectively bridges the gap. We have our work cut out for us.
© Dr Justin J. Bartlett 2023
Justin J. Bartlett is a philosopher at Srinakharinwirot University, Bangkok. He is currently doing research in metaethics and philosophy of language.