The argument for weak relativism
But although there is an argument against strong relativism, the argument against weak relativism is harder to make. We could come to learn that the Azande had a concept of the soul, or mbisimo, of a person, which operated in certain ways, and that they took the behavior of conscious people to be evidence for the existence of that mbisimo.
There seems to be, at least prima facie, no difficulty in understanding this claim. Nor does it seem difficult to understand that in their way of thinking—their conceptual scheme—what we took to be evidence that someone wanted meat was evidence that their mbisimo wanted meat. These seem to be different claims, and we might eventually feel that we understood what each of them meant. After learning English (and ENGLISH, with it) we could learn Azande (and AZANDE) in the way Zande children learn the language—not by translation, but directly. But we might still be able to think of no way of marshaling evidence that discriminated between these two ways of thinking about human mentality and behavior.We might also agree that you could only use one of these conceptual schemes at a time, but that nothing in the evidence forced you to use one or the other. As Evans-Pritchard found, it is possible to get used to using extremely alien forms of thought.
I want now to argue that this sort of weak cognitive relativism is possible. I shall argue, more precisely, that it is possible, as Kant thought, that the way we think about the world—our conceptual scheme—helps to determine what it is reasonable for us to believe. I shall also argue, however, that this is not too surprising.
To see why weak relativism is less puzzling than it might at first appear, all we need to do is to begin with a simple case.
In Middle German, the language spoken in Germany in the Middle Ages, there was no word that translated our word “brown.” The only word Middle German speakers had that covered brown things covered purple things also. They called things that were brown-or-purple “braun.” These people could certainly tell brown and purple things apart by looking at them. But if you had asked them to put marbles together into natural groupings, they would have put all the brown and purple marbles—all the braun ones—together.This difference is connected systematically with other differences between Middle German and modern English, for it follows that they did not have a word that accurately translated “color,” for example. They had the word “Farbe” instead. If “Farbe” translated “color,” then every truth about color would correspond to a truth about Farbe. But they did not think that brown and purple marbles were of two Farben; they thought they were of one Farbe.
Still, it is not too hard to see how we would translate this language. “Braun” translates as “brown or purple”; “Farbe” refers to colors, excluding brown and purple, but including brown-or-purple.
There would be a difference between operating these two conceptual schemes. Middle German speakers might have remembered the Farbe of many things but not—or not so easily—their color. We would continue to remember colors. Each of us could work out what the other would remember and take to be important about the looks of things, but different things would continue to strike each of us as important. Now there might be reasons for preferring one scheme to another: perhaps all the brown mushrooms in our country are edible and all the purple ones poisonous. Sensitivity to color would help here, and Farbosensitivity might be lethal. But the problem would not be that one scheme said that something was true that the other said was false.
These would be different ways of looking at the world; and evidence would lead them to say that brown things were “braun” and us to say that they were “brown.” And it would not be a matter of evidence which way of looking at the world was right.This simple case leads naturally into the more complex case of Zande belief in the mbisimo. Remember what I said in Chapter 1 about a functionalist theory of the mental. If there can be a functionalist theory of the mind, why could there not be a functionalist theory of the mbisimo? Indeed, if you remember what I said about functionalism in Chapter 1, you can argue that there must be such a theory. In Chapter 1 I said that, at the most general level, a functionalist theory explains the internal states of a system by fixing how they interact with input, and with other internal states, to produce output. But the only things we know about directly are the inputs and outputs. That is all the evidence there is. There seem, therefore, to be the same reasons for thinking that there must be a functionalist theory of the mbisimo as there are for believing there must be a functionalist theory of the mind.
If the Zande theory of the mbisimo and our theory of the mind made exactly the same predictions about what inputs would lead to what outputs, no amount of evidence would distinguish them. You might argue that this just showed that mbisimo meant the same as mind. But I think this would be wrong. For the internal states that the two theories proposed could operate in different ways. To put it in the terms of Chapter 1, the Ramsey-sentences of the two theories could have different structures, even if their consequences for input and output were the same.
The two theories might then differ, in the ways that Middle German and English differ. Classifications of states of the mbisimo that struck the Azande as natural might correspond to no natural classifications of ours.
Perhaps, over time, the Azande would find that our theory suited them better; perhaps we could take a cue from theirs. Most likely, however, as our understanding of the world developed, both of us would change our theories. And there would be nothing to guarantee that we would end up with the same theory, at least so long as we continued to speak different languages.If I am right, evidence and reason cannot, by themselves, lead us to one truth. There may be different ways of conceptualizing the one reality. To say this is to say more than that our knowledge of the world is fallible. We do, indeed, know that our own theories are not perfect. Many of the things that happen in our world we cannot explain; many others are actually inconsistent with our best current theories. But we also usually suppose that with time and effort we could make our theories better—explaining what could not be explained before, and modifying the theories to avoid their false consequences. Even those who believe that, because fallibilism is true, we are always at risk of being wrong think that it is possible to use evidence to get reasonable evidence that one theory—say, our everyday theory of belief and desire—is less adequate to the facts than another—say, neurophysiological theories of the mind.
But if I am right, this is not so. Relative to one conceptual scheme, it might be natural to say, “Jane believes that it's raining”; relative to another, it might be better to say, “Jane is in neural state X”; or even “Jane's mbisimo is in state Y.” And it might be impossible for one person to make all of these equally their natural way of reacting to the evidence, so that, in that sense, these conceptual schemes were incompatible. The choice between the three “realities” would be settled not by evidence but by asking: “Which conceptual scheme is it easier to live with?” There is no reason to suppose that two people in the same culture, let alone in different ones, would be bound to agree on the answer to this question. Nevertheless, of course, reasons and evidence are essential tools of thought in every conceptual scheme.
9.8