A long time ago, I was taught that logical positivism, though very productive in explicating what it means for one science to reduce to another (e.g. chemistry to physics), died because the verification criterion for meaning could not itself be verified.
Since that long-ago time, I have seen people defend verificationism. Was news of its death "greatly exaggerated"?
There is nothing without prefiguration in philosophy; and the dead usually rise from their graves. So, to follow Mauro's lead:
According to the standard textbook account, verificationism is a doctrine that was popular back in the days of logical positivism. The positivists promised us a 'verifiability criterion of meaningfulness' which would help to distinguish 'cognitively significant' discourse from nonsense, but despite their best efforts, no workable criterion was ever produced. This led to verificationism being thoroughly discredited and consigned to the philosophical scrap-heap, along with most of the other doctrines characteristic of logical positivism.
Now this story is only partially accurate. For it ignores the fact that verificationist ideas, in one form or another, continue to find adherents among contemporary philosophers. Quine, Putnam and Dummett, to pick the three best-known cases, are all openly sympathetic to the basic verificationist idea that the content of a statement is closely tied to what would count as evidence for its truth. None of these philosophers believes in the 'criterion of meaningfulness' that the positivists hankered after, but the spirit of verificationism lives on in their work. Due in large part to the interest generated by Putnam's and Dummett's writings, verificationist ideas have enjoyed something of a resurgence in recent years, though not always under that name. In this paper, I examine one of the factors that lies behind the continued appeal of verificationism.
In contemporary discussions, verificationist ideas usually surface by way of opposition to some form of realism. Putnam explicitly contrasts his version of verificationism - which he calls 'internal realism' - with a position called 'metaphysical realism'. Metaphysical realism holds that truth is a 'radically non-epistemic' notion, with no conceptual connection to justification or warranted assertion, while internal realism holds that truth is warranted assertibility in the ideal limit. Similarly Dummett contrasts the realist view that meaning is to explained in terms of 'evidence transcendent' truth-conditions, with his own anti-realist view that meaning should be explained in terms of verification-conditions, which cannot transcend our ability to detect them. For Putnam, Dummett and others, realism provides the foil for the development of their verificationist viewpoints.
(Samir Okasha, 'Verificationism, Realism and Scepticism', Erkenntnis (1975-), Vol. 55, No. 3 (2001), pp. 371-385: 371-2.)
The answer by Geoffrey Thomas is helpful, but a point made in the text deserves amplification. It could reasonably be said that verificationism lives on. But only because the question of "meaningfulness" has been mostly divorced from it:
None of these philosophers believes in the 'criterion of meaningfulness' that the positivists hankered after.
Science needs to be testable, and the ideas behind verificationism are useful for articulating the conditions for testability, and for distinguishing between productive discussion and unproductive speculation. Consider current debates about string theory to see why this is still important.
But the turn to "meaningfulness" was an overreach. Consider a work of fiction. Nothing in a work of fiction is verifiable or falsifiable, because fiction has no truth conditions at all. It doesn't claim to be a true description of anything, and so it can neither be true nor false. But to say that fiction is therefore "meaningless" massively distorts conventional notions of meaning. There's even a strong case to be made that this line of reasoning applies not only to fiction, but also to huge swaths of pure mathematics. The most extreme version of this view is mathematical fictionalism, but even relatively modest views about the nature of mathematical truth make it difficult to support the notion that mathematical statements meet the verifiability criterion of meaning.
It's true that sometimes philosophical terms do have narrow, technical definitions that diverge from standard usage. Verificationism failed precisely because "meaning" was not given such a narrow, technical definition, and if it had been, it would have been clear that "meaning" was not the best word to use.
