Some comments below on analytic-synthetic, with thanks to JLS for the useful excerpts from Popper's writings on evolution. In these, Popper acknowledges a change in his position but in a way that is perhaps overgenerous and might provide blank ammunition for the less scrupulous of his critics. His change of view is more a clarification of what-is-and-what-is-not testable under the guise of 'D' than any radical change in view: where at earlier times he emphasised important ways in which D is not falsifiable, he later wished to clearly acknowledge that many D-type explanations are well-tested. This is not a fundamental change of view as to the testability of D but amplification based on the fact that different D-type claims having different degrees of testability. But D as an explanatory framework (to be contrasted with, say, a Creationist or Lamarckist framework) remains (almost) as 'metaphysical' as its rivals: for what observation could ever falsify it? One of Popper's key arguments, for the 'metaphysical' status of D as a framework, remains untouched by his later clarifications - if we found a planet with traces of primitive life that corresponded to those from which life on earth evolved, but on that planet no life further evolved, would that falsify D? And if we found a planet with conditions for life that corresponded to those from which life on earth evolved, but where no life evolved, would we take that as falsifying D? Popper thinks not (we would instead probably explain the absence of evolution of life in such circumstances by some contingent facts - such that the primitive life was unable to survive or that the conditions for life, though seemingly favourable, simply did not trigger the same specific contingencies that led to the evolution of life on earth). And so it is very hard, in this vein, to conceive what kind of observable fact would ever falsify D as a framework:- for it would appear that any pattern in evolution - including even the failure of life to evolve - could be explained within a D-type framework. What has evolved is that Popper came more clearly to see that his own theory of knowledge is an 'evolutionary epistempology' in the sense put forward by D.T. Campbell. However, this evolutionary aspect is evident even from Popper's first book, The Two Fundamental Problems of The Theory of Knowledge. Also young Popper was hostile to the appropriation of D and ideas of "fittest to survive" by Social Darwinists, Nazis etc. and wary of D as a pseudo-scientific platform for vile ethics, and this influenced perhaps his initially circumspect and even "contemptuous" treatment of evolutionary philosophies. [That said, it might be noted that Popper thinks Bergson's elan vital is onto something important]. More below. ________________________________ From: "Jlsperanza@xxxxxxx" <Jlsperanza@xxxxxxx> To: lit-ideas@xxxxxxxxxxxxx Sent: Saturday, 25 February 2012, 23:50 Subject: [lit-ideas] Re: Darwiniana >It seems that, at first Popper was concerned that some claims, e.g. The fittest survives. would be _analytic_, rather than 'metaphysical'. (Only later, when he granted scientific status to Darwinism, did he clarify that he was against some _metaphysical_ ideas about evolution he would find in Whitehead, or Bergson). So it would be good to compare how Popper (who, like Grice/Strawson 'In defense of a dogma', accepts the analytic-synthetic distinction) relates the "analytic" (Popper's 'almost tautological', as per the quotes below) with the metaphysical. > Here I pause. For to say Popper accepts the analytic-synthetic distinction is not to say he accepts Grice/Strawson's way of drawing the distinction, and I suspect Popper's views are closer to Quine's in "Two Dogmas". What might be noted is that, for Popper, the analytic-synthetic distinction is not to be confused with the distinction between non-testable/testable by observation. An analytic statement will be true by definition as it were, and as such its truth is never testable by observation. But the class of synthetic statements may be sub-divided into those that are testable by observation and those that are not. When are we given a statement, like "All swans are white", we cannot tell whether it is analytic, synthetic-untestable or synthetic-testable simply by looking at it - we can only tell by looking at the methods used to defend the truth of the statement. So if we produce what appears to be black swan and someone says, "But that does not falsify 'All swans are white', because it is not white and so cannot be a swan", their method of defending the truth of "All swans are white" indicates that, for them, whiteness is a defining characteristic of a swan and so, for them, "All swans are white" is an analytic truth. But if they defend the the truth of "All swans are white" by saying that there is some actual-albeit-unobservable difference between the black-seeming-swan and an actual swan, and so the black-seeming-swan does not falsify "All swans are white", then their defence may indicate that they take "All swans are white" to be synthetic but in a way that is not testable by observation. If they accept what is observationally the same as a white swan (apart from its colour) as a swan (albeit black), and so accept that it falsifies "All swans are white", they show that for them "All swans are white" is synthetic and testable by observation. This point is of some importance because Popper, the "tottering, old metaphysician" as he once described himself, sees metaphysics as seeking synthetic, substantive truths, albeit not truths that are testable by observation - and is opposed to views that decry that there can be synthetic, substantive but empirically untestable truths. He opposes, therefore, the conception of philosophy as limited to the pursuit of so-called conceptual or analytic truths, and indeed decries philosophising that engages in this pursuit (- except to the extent that what is framed as an argument pursuing a conceptual/analytic truth may be reframed as one for a synthetic, substantive metaphysical truth). So to get back to the question JLS raises: We might say a tautology like "All table are tables" is a kind of analytic truth, but perhaps not all analytic truths are tautological in this sense - is "All swans are white" a tautology like "All tables are tables" if whiteness is a defining property of swans? If tautological, there is no 'almost' about it in these cases. For analytic truths or tautologies in this sense have no explanatory content as to how the world is, and this is because they have no synthetic content: they explain only how words are being used. But in this they are simply conventions that are being stipulated. It is for this reason that treating D as analytic or tautologous must be a mistake if we think D is at the same time some kind of explanation of how things are in the world. So, for Popper, the interesting content is always synthetic content - whether it is testable by observation or not. [And 'analytic content' is only interesting when it is synthetic-content-in-disguise (hence the importance of the question "Are there interesting analytic truths?")] Donal Who here issues an ongoing apology for spelling and other grammatical errors in his posts London