350 is there a hidden vocabulary

Prof Beach at ees: no, there is no very useful mid-level vocabulary for causation (how to do causal mapping, even at scale)

• So I was just at the European evaluation Society meeting in helimony last week and heard an interesting presentation from Professor Beach, from our horse in Denmark. And he was calling for a or discussing the possibility of a common vocabulary for the A series of change and program theory. And I think this is reminds me of Nancy Cartwright, whom he frequently sighted with her visual vocabulary, which includes elements like blockers, and also Gary Gertz has is fascinated by the question of what different elements what One might need in a explanatory map, but I'm quite sure that the answer to all of this is that it isn't going to work, and if there was such a possible common set of elements one could use to build up many different kinds of program theory in a way that would cover most of the relevant use cases, then we would have discovered one, probably hundreds of years ago, and certainly by now. Can also think of Rick Davies work on theory of change. I we have something like an alternative vision, which says that the only elements we need or are ever really going to have, are Simply boxes for the elements of the causal map, which we call factors, although we have to point out that factors can also be the ends of causal chains, and not just explanatory factors. Is factors and lines between them, where the lines simply mean an undifferentiated influence, or in our case, when we're talking about causal maps that somebody believes or claims there is an undifferentiated influence. This means we don't have or hold out any hope for any kinds of special symbols or categories with which we might differentiate the kinds of causal influence, like necessary or sufficient conditions or packaging factors into packages, or designating functions like moderators or inhibitors. In fact, we are so radical as to say that we have to abandon the idea that the nodes are somehow naming variables which can somehow go up or down or take different values, and that the arrows, therefore, at least can or should express At least a positive or negative direction of influence and or, more ambitiously, a Strength of influence. For us, the boxes are simply a bare propositions and the arrows are simply bare causal influence. But you might say, What an extraordinary, impoverished universe you live in. But that's not the case, because what we do is what we believe actual human speakers do, which is, in the most cases, to make the labels do all the work. And make the links do none of the work. This obviously simplifies things, because you only have the labels to worry about, not the links. So for example, people say things like, but because I don't have the money. I'm not going on holiday this year. So we can see this as an undifferentiated causal link between causal factors which are relatively sophisticated, namely, I the absence of something on the left, or possibly the absence of a sufficient amount of something on the left, and its influence on a future expectation or lack of future expectation. So evidence. This makes transcribing what people say in terms of causal influences very simple, because we don't have to worry about anything like necessity or sufficiency or anything else, and all the work is done by the labels the question then, of course, becomes, then means that the usual problem of, how do we organize our labels more generally in order to be able to gain a knowledge by saying that two or more labels are somehow related, which is the fundamental problem of qualitative text analysis and it's This function to which we give the job of combining for example, this person's child is ill with that person's child is ill, and subsuming them under a Common label. And in the same way, we have the problem of somehow combining the information about this person's going on holiday this year and this person, this other person isn't going on holiday this year. And somehow being able to present them as in one way, the same idea, but with differing I somehow differing parameters, somehow we need to, on the one hand, accept them as related or even instances of the same thing, while not losing the information that they are saying something different about this thing, the fact is that although humans sometimes do actually express causal influences in the form of an expression of how one variable influences another with a certain polarity and strength. These cases are really rare, and in the vast majority of human speech and writing, we express a relatively undifferentiated causal influence, but have a hair raising uncountable universe of different kinds of causal factor labels, which are a I also be somehow related or not related to one another. And I'm saying both that this barefoot style is the way that people, in fact, do express themselves, but also it's how we should try to understand and code what's going on.