On Thu, 12 Jun 2003 12:05:41 -0700, andie nachgeborenen wrote: >[Popper] admitted that, therefore, falsification >cannot be atomic, proposition by proposition, and can >only be tentative and propvision, not conclusive. He >disputed, however, that this meant that therefore >there was no point in talking about falsifiability, >or that it could not be the demarcation criterion, or >that there was no demarcation criterion. I think he >was right both in his concession and his conclusions >about its limited effect. If you think > otherwise, explain why.
Three related reasons: 1) Because Popper honoured this self denying ordinance as often in the breach as the observance whenever he started actually saying things about science, pseudoscience and nonscience. I'm sure that sufficient diligence might dig up some example of him saying that, for example, there might be constructed some valid science of personality based on planetary movements and that the validity of the Sirius-Nile cycle and lunar cycles in things from suicide rates to stock returns needed to be explained in some manner, but it was hardly representative of the man's work and by and large was discarded completely by his wannabes. There has to be some sort of product liability for academic theories, and Popper is as culpable for the influence of the cruder version of his theory as the producers of Vicodin. 2) A lot of the explanation for 1) is that the grammer of the word "demarcation" makes it very difficult to use in contexts where you are talking about something that admits of degrees and can only be judged holistically. And language is telling us something here; it is specifically telling us that a "Criterion" which can't be relied upon to give a definitive answer, isn't much of a criterion at all. If Popper had only talked about a "falsificationist approach", a way of thinking which set score by attempting to knock down null hypotheses, then he'd have been true to this interpretation of them, but he'd have been saying something more similar to Lakatos. 3) And finally, a lot of the force behind 2) above comes from the fact that a lot of the *point* of falsificationism is that it's meant to give us a short way with all the problems of holism and tentativeness which dog theories of science based on confirmation. If falsificationism isn't going to definitively rule anything out, and if it can't be used to assess the scientificity of particular propositions, then what's the advantage over Bayesianism? Popper was originally fighting a war on two fronts, and it strikes me that in making these concessions to one side he's very much opening himself up to attack on the other. dd PS: If Ian had been quicker on the draw he might have pointed out that Quine thought that it *was* a problem for Popper and demanded that you be the one to "explain why" you disagreed. I'm not sure you realise how dismissive you're being. It must be irritating to have to revisit ground you thought had been covered and lost in the academic literature ages ago, but it's also quite annoying for us to be magisterially directed to read a 400-page doctoral thesis before we're allowed to have opinions. Everyone on this list has to explain the basic material of their own specialist fields every once in a while.
