There are three categories -- "meaningful," "meaningless," and "tautological" statements -- at least in Ayer's categorization. "Statements which are not testable are meaningless or tautological" would be an example of a tautology: just a definition of terms.

Because if you /could/ test the statement to see if it were true (not absolutely true, but, per Ayer, "probable"), you'd conduct an experiment where you took a sample of statements, tried to come up with tests (ways in which they refer to sense experiences that would serve to verify or disprove them), and then saw which ones were or were not meaningful. But in Ayer's framework, meaningfulness is defined as referring to sense experiences that would serve to verify or disprove, so it's circular, thus tautological, which isn't a term of abuse in Ayer's categorization the way meaninglessness is. He thinks that philosophers deal in tautologies all the time -- constructively! -- and that meaningful statements are more in the domain of science anyway.