A.J. Ayer, answered from the texts and cited to the page.
A statement is literally meaningful if and only if it is either analytic (true in virtue of meaning, like the propositions of logic and mathematics) or empirically verifiable in principle (in which case the conditions for its verification specify its meaning).123 Statements that satisfy neither condition are not false but literally meaningless: they are pseudo-propositions that grammatically resemble assertions but have no cognitive content.
Most of traditional metaphysics — claims about the Absolute, about transcendent reality, about God's existence, about the noumenal world, about the Real beyond appearances — falls under this verdict. The elimination of metaphysics is not itself a competing metaphysical doctrine; it is the recognition that the disputes traditional metaphysicians have conducted lack the cognitive content that would make them genuine disputes at all.
The strict 1936 verifiability principle had problems acknowledged in the 1946 Introduction and refined across the long career; the considered position preserves the verification criterion while admitting weak verifiability (in principle, by some possible observation) as the relevant standard. The programme is empiricist and anti-Kantian: there is no synthetic a priori knowledge, only analytic propositions and empirical ones.4
In putting forward the principle of verification as a criterion of meaning, I do not overlook the fact that the word "meaning" is commonly used in variety of and do wish a senses, I not to deny that in some of these senses a statement may properly be said to be meaningful even though it is neither analytic nor empirically verifiable.Language, Truth and Logic, pp. 14–15
For while the statements that contain these terms may not appear to describe anything that anyone could ever observe, a "dictionary" may be provided by means of which they can be transformed into statements that are verifiable; and the statements which constitute the dictionary can be regarded as analytic.Language, Truth and Logic, p. 13
I can quite well believe that you have experiences different from mine. Let us assume (which after all is an empirical assumption) that you have even asense different from mine. I should be in the position of the blind man, and then I should admit that statements which are unintelligible to me might be meaningful for you.The Meaning of Life, p. 38
For what empirical test could possibly go to decide whether the world is one Editor's Introduction [ 15 ] or many, or whether the things that we perceive do or do not exist outside someone's mind? It is characteristic of such rival philosophical theses as realism and idealism that each is consistent with all the appearances, whatever their content may happen to be.Logical Positivism, pp. 15–16
Language, Truth and Logic (1936) is one of the most influential single books of twentieth-century philosophy and the central text of the British reception of Vienna Circle logical positivism. The verification principle has been challenged repeatedly: Popper proposed falsification as alternative; Quine's "Two Dogmas of Empiricism" (1951) attacked the analytic/synthetic distinction; Putnam and Kripke challenged the empiricist account of meaning through natural-kind arguments; Dummett's "Metaphysics of Verificationism" (Hahn volume) argued the principle has metaphysical commitments of its own. Logical positivism is widely held to have been superseded; Ayer himself acknowledged that "logical positivism died long ago" while continuing to defend the empiricist framework. The verification principle remains a live position in the tradition that runs from Carnap through contemporary neo-empiricism.