Morality is like vision: you can just see when something is wrong. And when you see that something is wrong, it feels objective, it feels like you’re observing something true about the world. The apple is red. Killing babies is wrong.
Analogies are good. Metaphors, comparisons, contrasts – they all help us understand things that we might otherwise not grasp, to organise and put words to things that otherwise might slip around in our minds making no sense. Often, it’s about translating an abstract concept into something more concrete – the main metaphor I always think of here is “ideas are food” (see Lakoff & Johnson, 1980, for an excellent introduction to “the metaphors we live by”). But analogies also have their limitations, and that’s what I’ll try to demonstrate in this post.
In psychology, naturally, analogies and metaphors are used to understand psychological concepts – but they are often contested. For example, the brain is (not) a computer. Moral disgust is (or is not) bodily disgust. The five moral foundations are like the five taste buds (or not).
In their paper The Visual Guide to Morality (which inspired this post) Schein, Hester, and Gray (2016)* argue that “vision provides an intuitive way to summarize and communicate moral psychology research and also offers perspectives on current debates in the field”. The current debates that they are referring to are intuition vs. reason, pluralism vs. universalism, and constructionism vs. modularity. Overall I was convinced. Yes, the visual analogy does indeed seem intuitive and useful for thinking about morality. Yes, it helps organise our discussions and provides an interesting perspective on various debates. The one disagreement I had with it, was when I read in the conclusion that “vision has already settled many of the issues that moral psychology currently debates” – at that point I thought “and so what?” But more on that, later in this post.
As I was reading the paper, I also had another analogy lurking at the back of my mind: The Linguistic Analogy (LA). According to this analogy, morality is like language. Although it shared some aspects with the morality-vision analogy (you can see that an apple is red, you can hear that a sentence is ungrammatical, you can know that killing is wrong), proponents of this analogy (e.g. Mikhail, 2007; Dwyer, 2009) went further than Schein and colleagues – further both in the broader sense, and in the deeper sense. Basically, this analogy was going to revolutionize the study of morality, just like Chomsky’s program of research revolutionized linguistics in the 80s.
But… I’m writing in the past tense, because although the LA was pretty popular for a while, it has since sort of fizzled out. Why might that be?
When I first came across the LA, it was pretty exciting to me. I studied linguistics (in addition to psych) in undergrad, but for my PhD I wanted to focus on morality – could I really combine my two interests by studying something like moral grammar (a particular aspect of the LA)?! That sounds awesome!
Well, yes and no. While I eventually left behind moral grammar in favour of a focus on soldiers, an important study did come out of my initial interest in LA (Simon and I had it published in Thinking & Reasoning last year, check it out!). In this paper, we make the point that although the LA sounds all well and good as an analogy, the theorists have drawn the bow too far when it comes to method. Remember how I said that the LA goes pretty deep? I meant deep as in “potentially permeating all aspects of the study of morality” – in particular, Mikhail and others propose that because of the analogy between linguistics and morality, we can also use the techniques the linguists use to study language, to study morality. Two things work the same way –> use the same tools to study them. It made a certain kind of sense to me, but when we tried it out… no. At the moment, the best way to study morality does not involve using the tools suggested by the LA.
It’s a sort of negative message, but I think it helps explain why the LA did not take off; why the study of morality now does not look like the study of linguistics in the 80s. You can evaluate an idea according to whether it is true, abstract, or useful (thanks TAPAS), and the LA falls at the last hurdle because you can’t use its methods for producing any kind of empirical research. It’s a neat idea, with a good level of abstraction, so perhaps it can offer a helpful perspective on some debates in morality (e.g. competence vs. performance, universalism vs. pluralism) – but until we can operationalise the key concepts adequately and pin them down in our studies, it simply remains too theoretical (at least for mainstream moral psych). We don’t even know if it’s true, in any practical sense.
Which brings us back to the visual analogy. Schein, Hester, and Gray don’t go as far as the LA. They are more clearly sticking to their limits, by suggesting that we use the analogy as an aid to understanding, rather than as something from which we can derive innovative tools and future ideas for study and revolution etc etc. In other words, there is less danger of some poor naïve PhD student taking their idea too literally, and embarking on an attempt to study morality like vision. Which is definitely a good thing. What might be less of a good thing, is what I was hinting at earlier – that by saying “vision has already settled many of the issues moral psychology currently debates”, the reader is left with the impression that (the authors think that) “we should just accept the conclusions that the vision researchers have come to, regarding intuition vs. reason, pluralism vs. universalism, and constructionism vs. modularity”.
But that seems to be going too far.* Just like we (well, Simon and I) had to test the tools of the linguistic analogy to know that they didn’t work for studying morality, we have to also test the different hypotheses about morality suggested by the visual analogy. In other words, vision researchers might know where they stand on various debates, but so what? We can use their framework to help structure investigations in moral psychology, but to find out whether morality really is intuitive, universal, modular (for example) there is still research left to be done.
So that’s the limitation of analogies. They are great tools for thinking and theorizing. But to know whether they “work”, you have to actually do the work.
*Two things about these authors:
1) I know them! Last year I spent 2 months in Kurt Gray’s lab, and hung out with Chelsea and Neil as well. It was great, and so I’m probably positively biased towards their writing now. 🙂
2) They don’t actually say that “we should accept the conclusions the vision researchers have come to”. I just feel like it’s a plausible (but problematic) interpretation of their article overall, and I’m cautioning future readers more than the authors themselves.