Popper formulates and explains his non-justificationist theory of knowledge. Science--empirical science--aims at true explanatory theories, yet it can never prove, finally establish, or justify any of its theories as true, not even if it is in fact a true theory. Science must continue to question and criticize all its theories, even those which happen to be true.

The controversy surrounding Popper's proposed solution to the problem of induction is beginning to display many of the symptoms of being interminable. For decades the discussion has continued, apparently without any progress being made. Again and again, Popperians and their critics have accused each other of ‘missing the point’. The essay attempts to explain what exactly is ‘the point’ of the problem of induction, and asks whether Popper does indeed miss it. An answer is proposed, and on this basis an (...) explanation for the puzzling interminability and emptiness of the above dialogue is put forward. (shrink)

The Comprehensibility of the Universe puts forward a radically new conception of science. According to the orthodox conception, scientific theories are accepted and rejected impartially with respect to evidence, no permanent assumption being made about the world independently of the evidence. Nicholas Maxwell argues that this orthodox view is untenable. He urges that in its place a new orthodoxy is needed, which sees science as making a hierarchy of metaphysical assumptions about the comprehensibility and knowability of the universe, these assumptions (...) asserting less and less as one ascends the hierarchy. This view has significant implications: that it is part of scientific knowledge that the universe is physically comprehensible; that metaphysics and philosophy are central to scientific knowledge; that science possesses a rational, if fallible, method of discovery; that a new understanding of scientific method and rationality is required. Maxwell argues that this new conception makes possible a natural resolution of long-standing philosophical problems about science, regarding simplicity, induction, and progress. His goal is the reform not just of the philosophy of science but of science itself, and the healing of the rift between the two. (shrink)

Charles Babbage was an English mathematician, philosopher and mechanical engineer who invented the concept of a programmable computer. From 1828 to 1839 he was Lucasian Professor of Mathematics at Cambridge, a position whose holders have included Isaac Newton and Stephen Hawking. A proponent of natural religion, he published The Ninth Bridgewater Treatise in 1837 as his personal response to The Bridgewater Treatises, a series of books on theology and science that had recently appeared. Disputing the claim that science disfavours religion, (...) Babbage wrote 'that there exists no such fatal collision between the words of Scripture and the facts of nature'. He argues on the basis of reason and experience alone, drawing a parallel between his work on the calculating engine and God as the divine programmer of the universe. Eloquently written, and underpinned by mathematical arguments, The Ninth Bridgewater Treatise is a landmark work of natural theology. (shrink)

Originally published in 1969. This book explains what is wrong with the traditional methodology of "inductive" reasoning and shows that the alternative scheme of reasoning associated with Whewell, Pierce and Popper can give the scientist a useful insight into the way he thinks.

The main text of Popper’s Logik der Forschung consists of seven chapters outlining the main features of his falsificationist philosophy of science, followed by two involving probability theory and quantum mechanics, and finally one on corroboration, which is the main concern here., noting its newer material when appropriate.) The chapter begins with a section on non-verifiability of theories, which would have been better placed in chapter 6 on testability, and then two sections on the probability of events and of hypotheses (...) and on probability logic, which belong rather in chapter 8. Only then come four short sections on corroboration, the property of a theory that it passes its tests. The chapter is placed last because of a link between corroboration and probability; he even claimed that there was no difference between the two in that both are appraisals. In a later appendix *ix, and in more detail in other work, he offered quantitative measures of corroboration involving probabilities, but at the end of the latter text he almost dismissed the enterprise; and in similar vein, I shall not consider it further, since the measures depend upon the tests which happen to have been carried out and so are not very significant. Otherwise, he did not discuss corroboration very much in his writings, and it is not well handled by commentators;1 but this rather low status contrasts markedly with the actual practice of science, where scientists and engineers regularly work with highly corroborated theories, even to the extent that ‘reliability science’ is a subject in its own right. (shrink)

The main text of Popper’s Logik der Forschung consists of seven chapters outlining the main features of his falsificationist philosophy of science, followed by two involving probability theory and quantum mechanics, and finally one on corroboration, which is the main concern here., noting its newer material when appropriate.) The chapter begins with a section on non-verifiability of theories, which would have been better placed in chapter 6 on testability, and then two sections on the probability of events and of hypotheses (...) and on probability logic, which belong rather in chapter 8. Only then come four short sections on corroboration, the property of a theory that it passes its tests. The chapter is placed last because of a link between corroboration and probability; he even claimed that there was no difference between the two in that both are appraisals. In a later appendix *ix, and in more detail in other work, he offered quantitative measures of corroboration involving probabilities, but at the end of the latter text he almost dismissed the enterprise; and in similar vein, I shall not consider it further, since the measures depend upon the tests which happen to have been carried out and so are not very significant. Otherwise, he did not discuss corroboration very much in his writings, and it is not well handled by commentators;1 but this rather low status contrasts markedly with the actual practice of science, where scientists and engineers regularly work with highly corroborated theories, even to the extent that ‘reliability science’ is a subject in its own right. (shrink)