BookReview: The Logic of Scientific Discovery
by Karl Popper, Routledge, March 29, 2002, 978-0415278447
This book introduces Karl Popper's theory of how science advances. Published in the original German in 1934, with the English edition arriving in 1958. Popper's writing style is probably the most clear and interesting of any philosopher I have read. I understand his books almost completely -- except his mathematical proofs, which are beyond me no matter who writes them. The book is thick (544 pages), but I read it through. While the concepts are extremely simple, the argumentation is quite intense.
The detail was overwhelming at times. It's not a casual philosphy book.
Popper is a philosopher of grand theories. At the start of the preface, he claims that "all science is cosmology". I can relate, because I think it is all software, or as John Wheeler says, "it's a bit". While I do not pretend to understand the first thing about the cosmology of the Universe, I do think there's a cosmology of Software (all software). Therefore, I think there are many lessons to be learned from Popper. What disappoints me is that my computer science curriculum did not include a discussion of Popper's Logic of Scientific Discovery. I guess I would have been to young and impatient to understand such a course, anyway.
The problem with any epistemology is that it cannot prove itself. Some people think that Popper claims that his proposals are provable, and he argues with such forcefulness that he seems certain they are true, and may have even fooled himself into their being true. However, I think that's true of many of the philosopher's of science, and this doesn't mean his theory is not useful on its own. Popper was a lightning rod for criticism, and I think unfairly so. He was so certain, I think, that it was easy to mock him with relativistic arguments ("we never know anything for certain"), which might also be true, but did not devalue his theories themselves as well as their usefulness as an approach to scientific discovery.
Testable software is better, because we can define how it will behave. That doesn't mean we have written all the possible tests for the program. Rather, we can write tests, and the more we write, the more reliable the software will be. If we can't even think of how to test software, how would we know if it begins to satisify our customer's requirements.
Via Rob 10/31/2008
The Psychology of Software TestingBookReview: Art of Software TestingBookReview: Handbook of Usability Testing: How to Plan, Design, and Conduct Effective TestsBookReview: Testing Computer Software, 2nd Ed.BookReview: The Mythical Man-MonthMajor Release Syndrome: A Case for Chronological VersioningBookReview: Text, ConText, and HyperTextBookReview: Extreme Programming ExplainedBookReview: Prioritizing Web UsabilityQuantifiably SimpleBookReview: Hackers and Painters: Big Ideas from the Computer AgeBookReview: The Salmon of DoubtBookReview: Quality Software Management: Vol. 1 Systems ThinkingBookReview: Jurassic ParkBookReview: The Oxford Book of Modern Science WritingBookReview: The Emotion Machine: Commonsense Thinking, Artificial Intelligence, and the Future of the Human MindBookReview: The Logic of Scientific DiscoveryBookReview: Don't Make Me ThinkBookReview: Kuhn vs. Popper: The Struggle for the Soul of ScienceBookReview: Against Method (Third Edition)Programming is EasyBookReview: Psychology of Computer ProgrammingBookReview: The Blank Slate: The Modern Denial of Human NatureEvolutionary EconomicsBookReview: The Black Swan: The Impact of the Highly Improbable▶ More▲ Most Recent
|back to top||© 2018 Rob Nagler||Software by bivio|