| 5862 | A single counterexample is enough to prove that a truth is not necessary [Aristotle] |
| Full Idea: If we have a single counter-instance, the argument is refuted as not necessary, even if more cases are otherwise or more often otherwise. | |
| From: Aristotle (The Art of Rhetoric [c.350 BCE], 1403a07) | |
| A reaction: This is Aristotle (pioneering hero) pointing out what we now tend to think of as Karl Popper's falsification, the certain way to demonstrate the falseness of a supposed law of nature, by finding one anomaly from it. |
| 5629 | If a proposition implies any false consequences, then it is false [Kant] |
| Full Idea: If only a single false consequence can be derived from a proposition, then this proposition is false. | |
| From: Immanuel Kant (Critique of Pure Reason [1781], B819/A791) | |
| A reaction: Seems right. Of course, it might imply entirely true consequences, and still be false. This idea has to be one of the foundations (sic) of coherentism about truth and justification. |
| 22658 | Experimenters assume the theory is true, and stick to it as long as result don't disappoint [James] |
| Full Idea: Each tester of the truth of a theory …acts as if it were true, and expects the result to disappoint him if his assumption is false. The longer disappointment is delayed, the stronger grows his faith in his theory. | |
| From: William James (The Sentiment of Rationality [1882], p.42) | |
| A reaction: This is almost exactly Popper's falsificationist proposal for science, which interestingly shows the close relationship of his view to pragmatism. Believe it as long as it is still working. |
| 22137 | Observation can force rejection of some part of the initial set of claims [Duhem, by Boulter] |
| Full Idea: Logic and observation alone do not force a scientist to reject a scientific claim if experimental observations so not turn out as expected. The scientist must reject something of the initial set of claims, but that is a matter of choice. | |
| From: report of Pierre Duhem (The Aim and Structure of Physical Theory [1906]) by Stephen Boulter - Why Medieval Philosophy Matters 2 | |
| A reaction: This is a key point against any simplified Popperian notion of falsification. Tiny observations can't kill huge well supported theories. |
| 19464 | We only discard a hypothesis after one failure if it appears likely to keep on failing [Ayer] |
| Full Idea: Why should a hypothesis which has failed the test be discarded unless this shows it to be unreliable; that is, having failed once it is likely to fail again? There is no contradiction in a hypothesis that was falsified being more likely to pass in future. | |
| From: A.J. Ayer (The Problem of Knowledge [1956], 2.viii) | |
| A reaction: People may become more likely to pass a test after they have failed at the first attempt. Birds which fail to fly at the first attempt usually achieve total mastery of it. There are different types of hypothesis here. |
| 22191 | Kuhn's scientists don't aim to falsifying their paradigm, because that is what they rely on [Kuhn, by Gorham] |
| Full Idea: In Kuhn's view scientists are decidedly not interested in falsifying their paradigm, because without a paradigm there is no systematic inquiry at all. | |
| From: report of Thomas S. Kuhn (Structure of Scientific Revolutions (2nd ed) [1962]) by Geoffrey Gorham - Philosophy of Science 3 | |
| A reaction: This seems to be one of the stronger aspects of Kuhn's account. You'd be leaving the big house, to go out on the road with a tent. |
| 18076 | Most theories are continually falsified [Kuhn, by Kitcher] |
| Full Idea: Kuhn contends that almost all theories are falsified at almost all times. | |
| From: report of Thomas S. Kuhn (Structure of Scientific Revolutions (2nd ed) [1962]) by Philip Kitcher - The Nature of Mathematical Knowledge 07.1 | |
| A reaction: This is obviously meant to demolish Karl Popper. |
| 22188 | Give Nobel Prizes for really good refutations? [Gorham on Popper] |
| Full Idea: Popper implies that we should be giving Nobel Prizes to scientists who use severe tests to show us what the world is not like! | |
| From: comment on Karl Popper (The Logic of Scientific Discovery [1934]) by Geoffrey Gorham - Philosophy of Science 2 | |
| A reaction: A lovely simple point. The refuters are important members of the scientific team, but not the leaders. |
| 7780 | Falsification is the criterion of demarcation between science and non-science [Popper, by Magee] |
| Full Idea: According to Popper, falsification is the criterion of demarcation between science and non-science. | |
| From: report of Karl Popper (The Logic of Scientific Discovery [1934]) by Bryan Magee - Popper Ch.3 | |
| A reaction: If I propose something which might be falsified in a hundred years, is it science NOW? Suppose my theory appeared to be falsifiable, but (after much effort) it turned out not to be? Suppose I just see a pattern (like quark theory) in a set of facts? |
| 16830 | We don't only reject hypotheses because we have falsified them [Lipton on Popper] |
| Full Idea: Popper's mistake is to hold that disconfirmation and elimination work exclusively through refutation. | |
| From: comment on Karl Popper (The Logic of Scientific Discovery [1934]) by Peter Lipton - Inference to the Best Explanation (2nd) 05 'Explanation' | |
| A reaction: The point is that we reject hypotheses even if they have not actually been refuted, on the grounds that they don't give a good explanation. I agree entirely with Lipton. |
| 6794 | If falsification requires logical inconsistency, then probabilistic statements can't be falsified [Bird on Popper] |
| Full Idea: In Popper's sense of the word 'falsify', whereby an observation statement falsifies a hypothesis only by being logically inconsistent with it, nothing can ever falsify a probabilistic or statistical hypothesis, which is therefore unscientific. | |
| From: comment on Karl Popper (The Logic of Scientific Discovery [1934]) by Alexander Bird - Philosophy of Science Ch.5 | |
| A reaction: In general, no prediction can be falsified until the events occur. This seems to be Aristotle's 'sea fight' problem (Idea 1703). |
| 6795 | When Popper gets in difficulties, he quietly uses induction to help out [Bird on Popper] |
| Full Idea: It is a feature of Popper's philosophy that when the going gets tough, induction is quietly called upon to help out. | |
| From: comment on Karl Popper (The Logic of Scientific Discovery [1934]) by Alexander Bird - Philosophy of Science Ch.5 | |
| A reaction: This appears to be the central reason for the decline in Popper's reputation as the saviour of science. It would certainly seem absurd to say that you know nothing when you have lots of verification but not a glimmer of falsification. |
| 18284 | Particulars can be verified or falsified, but general statements can only be falsified (conclusively) [Popper] |
| Full Idea: Whereas particular reality statements are in principle completely verifiable or falsifiable, things are different for general reality statements: they can indeed be conclusively falsified, they can acquire a negative truth value, but not a positive one. | |
| From: Karl Popper (Two Problems of Epistemology [1932], p.256), quoted by J. Alberto Coffa - The Semantic Tradition from Kant to Carnap 18 'Laws' | |
| A reaction: This sounds like a logician's approach to science, but I prefer to look at coherence, where very little is actually conclusive, and one tinkers with the theory instead. |
| 4223 | Unfalsifiability may be a failure in an empirical theory, but it is a virtue in metaphysics [Lowe] |
| Full Idea: Although unfalsifiability is probably a defect in scientific hypothesis, because it is deprived of empirical content, it seems rather to be a virtue in a metaphysical hypothesis. | |
| From: E.J. Lowe (A Survey of Metaphysics [2002], p.241) | |
| A reaction: Presumably nothing could ever be found to count against a necessary truth. A nice point. 'Find me an instance where 2+2 is not 4'. |
| 4644 | A proposition such as 'some swans are purple' cannot be falsified, only verified [Baggini /Fosl] |
| Full Idea: The problem with falsification is that it fails to work with logically particular claims such as 'some swans are purple'. Examining a million swans and finding no purple ones does not falsify the claim, as there might still be a purple swan out there. | |
| From: J Baggini / PS Fosl (The Philosopher's Toolkit [2003], §3.29) | |
| A reaction: Isn't it beautiful how unease about a theory (Popper's) slowly crystallises into an incredibly simple and devastating point? Maybe 'some swans are purple' isn't science unless there is a good reason to propose it? |
| 22174 | The discoverers of Neptune didn't change their theory because of an anomaly [Okasha] |
| Full Idea: Adams and Leverrier began with Newton's theory of gravity, which made an incorrect prediction about the orbit of Uranus. They explained away the conflicting observations by postulating a new planet, Neptune. | |
| From: Samir Okasha (Philosophy of Science: Very Short Intro (2nd ed) [2016], 1) | |
| A reaction: The falsificationists can say that the anomalous observation did not falsify the theory, because they didn't know quite what they were observing. It was not in fact an anomaly for Newtonian theory at all. |
| 22175 | Science mostly aims at confirming theories, rather than falsifying them [Okasha] |
| Full Idea: The goal of science is not solely to refute theories, but also to determine which theories are true (or probably true). When a scientist collects data …they are trying to show that their own theory is true. | |
| From: Samir Okasha (Philosophy of Science: Very Short Intro (2nd ed) [2016], 2) | |
| A reaction: This is the aim of 'accommodation' to a wide set of data, rather than prediction or refutation. |
| 22189 | Why abandon a theory if you don't have a better one? [Gorham] |
| Full Idea: There is no sense in abandoning a successful theory if you have nothing to replace it with. | |
| From: Geoffrey Gorham (Philosophy of Science [2009], 2) | |
| A reaction: This is also a problem for infererence to the best explanation. What to do if your best explanation is not very good? The simple message is do not rush to dump a theory when faced with an anomaly. |
| 22190 | If a theory is more informative it is less probable [Gorham] |
| Full Idea: Popper's theory implies that more informative theories seem to be less probable. | |
| From: Geoffrey Gorham (Philosophy of Science [2009], 3) | |
| A reaction: [On p.75 Gorham replies to this objection] The point is that to be more testable they must be more detailed. He's not wrong. Theories are meant to be general, so they sweep up the details. But they need precise generalities and specifics. |
| 19685 | Falsificationism would be naive if even a slight discrepancy in evidence killed a theory [McGrew] |
| Full Idea: Data do not quite speak for themselves, which speaks against a naive form of falsificationism according to which even the slightest mismatch between theory and evidence suffices to overturn a theory. | |
| From: Timothy McGrew (Evidence [2011], 'Interp..') | |
| A reaction: [He cites Robert Boyle wisely ignoring some data to get a good fit for his graph] |
| 14576 | Smoking disposes towards cancer; smokers without cancer do not falsify this claim [Mumford/Anjum] |
| Full Idea: Smoking disposes towards cancer, and has its way in many instances. The existence of some smokers without cancer, however, does nothing to falsify this dispositional claim. | |
| From: S.Mumford/R.Lill Anjum (Getting Causes from Powers [2011], 7.5) | |
| A reaction: Indeed, falsification by one instance will only work against absolute and universal claims, and nature contains hardly any of those. |