Top critical review
7 people found this helpful
Academic and over-rated
on 1 August 2013
You get little from skimming this book because you miss all the fine points that contribute to following text. Instead you have to work hard to get it all, and doubly so because Popper did not use the new edition to streamline the text, preferring to add copious long footnotes instead, a few of which are highly relevant so you have to check all of them. I do not recommend that young research scientists waste their time with it; you'd be better off keeping up with the literature of your subject and watching closely those scientists who successfully carry out innovative, original research.
Popper divides science into 'inductive' (many swans are white -> all swans are white (BAD)) and 'deductive' (tentative hypothesis: all swans are white -> there should be no black swans -> look hard for a black swan and retain hypothesis tentatively until you find one (GOOD)). (My simplification of course.) I'm surprised that his binary classification of science has been given so much credence. In my experience, a range of hypotheses exists from highly controversial to well-established. Observations are collected and fitted to these hypotheses as well as can be done, and modifications are made or new hypotheses put forward when necessary for debate and further trials. This seems to fall somewhere between induction and deduction because hypotheses are 'inducted' from observations and become well-established through new observations that can be either supportive or non-falsifying. The main impediment to over-turning poor hypotheses are the important scientists who worked on them most of their lives (see Kuhn) or a lack of research funds, not philosophical mistakes. Very often in applied science, theories must be used to make political or economic decisions, and decision makers are unlikely to be impressed by a new theory simply because it has not been falsified.
Popper's other big thing is 'the problem of demarcation'. This is finding the division between "empirical sciences on the one hand, and mathematics and logic as well as 'metaphysical' systems on the other" (p 11). Nowadays, mathematics and logic are interwoven throughout science and empiricism so why do we need a division? The case is not clear to me.
Much of the book is given over to refutation of previous philosophers, making for dry academic reading. Popper makes liberal use of 'I think', 'I believe' and other opinions which do not enhance a philosophy book. He takes it upon himself to redesign a frequentist theory of probability (pp 133 - 208) for the singular circumstance of a binary sequence. We are left to wonder how he considers this to be generally applicable unless inductively. I have not noticed from my statistical studies that the theory has caught on, and Popper himself suggests in a footnote on p148 that most could be skipped on a first reading. I struggled on but was grateful eventually to take his advice. OK, so probability statements are unfalsifiable (p133) but why then was a boring and poorly referenced rewrite needed? In chapter 9, he then makes an "audacious attempt" (p209) to enquire into the foundations of quantum theory (p210). A case of misplaced overconfidence perhaps? I skipped that chapter. I would have been interested if Popper had struggled to apply his ideas to my field of marine science, where everything must be done underwater.
Chapter 10 brings all together or, perhaps, reveals the house of cards. We get an unquantifiable 'degree of corroboration' for theories that cannot be falsified. Also, various doubtful universal statements such as "the scientific method presupposes the immutability of natural processes" (did he ever think of biology?) or, worse by his own ideas: "this seems to me to exhaust the possibilities of basing the probability of a hypothesis . . ." (translate to: therefore all swans are white!). On p260, we glimpse the ivory tower: "I should decline to join in any dispute about how physicists 'in fact' proceed, since this must remain largely a matter of interpretation." Yet on p22 he says that the objectivity of scientific statements lies in "intersubjective testing" - the "mutual rational control by critical discussion".
Sorry, I am going on a bit. The book was probably a good stirrer in its day and may yet give you some useful philosophical ideas but, as an aid to current research, it's a time waster.