This book expresses the core theory of Logical Positivism in its more developed form. In short: a factually significant proposition (i.e. a proposition that actually says something about the observable world) must be verifiable or falsifiable at least in principle, by some possible observations which would increase or decrease the probability of the propostion being true. The only other meaningful statements are tautologies, they say the same thing twice: a tautology is true of false in virtue of the defintions of terms used. Or put another way, in analytic language one may say anything they want and make up any defintions with the one condition that one may not contradict oneself. This doctrine (of logical positivism in general) derives from classic empiricism, which asserts that all knowledge of the world must derive solely from sense-data (using our 5 senses) and our human ability to conceptualise and organise such data.
From these claims, Ayer developes the emotivist theory of values and argues that literal assertions about God, of any supernatural entity, or of literally 'good' or 'evil' behaviour are literally senseless (they say nothing at all about the empirical world). Religious and moral language do, of course, have aethetic value in so much as they express how we feel about the world.
He also explains how Mathematical reasoning is possible and how Mathematically theories can be described as 'true' - Maths is a form of analytic reasoning so that a Mathematician may say anything he wants and define any symbol however he wants (i.e. lays down axioms and definitions) so long as he does not contradict himself. Theorums can be derived by carefully investigating what is implied by these axioms and defintions.
Language, Truth and Logic is not too long (it can be read in about 6 hours or so) but clearly and systematically developes a coherant account of human logic. Ayer answers many classic problems in Philosophy while at it - problems such as God, monism vs pluralism etc can often be reduced to meaninglessness. He attempts to answer the problem of whether there is any reason to believe in other minds, through analogy of observable body to unobservable mind, but he himself (in his introduction) admits that he had not resolved the dialema altogether (e.g. it may still be questioned why other material bodies may behave as humans without experiencing sensations like you yourself do [assuming you yourself have a mind!].)
This work ranks alongside Russell's 'Problems of Philosophy' and Russell and Whitehead's 'Principle's of Mathematics' as one of the key works of early 20th Century analytic philosophy; crucial to anyone with an interest in Philosophy (or to some extent any Science).