In 'Logical consequence, a defense of Tarski' (Journal of Philosophical Log
ic, vol. 25, 1996, pp. 617-677), Greg Ray defends Tarski's account of logic
al consequence against the criticisms of John Etchemendy. While Ray's defen
se of Tarski is largely successful, his attempt to give a general proof tha
t Tarskian consequence preserves truth fails. Analysis of this failure show
s that de facto truth preservation is a very weak criterion of adequacy for
a theory of logical consequence and should be replaced by a stronger absen
ce-of-counterexamples criterion. It is argued that the latter criterion ref
lects the modal character of our intuitive concept of logical consequence,
and it is shown that Tarskian consequence can be proved to satisfy this cri
terion for certain choices of logical constants. Finally, an apparent incon
sistency in Ray's interpretation of Tarski's position on the modal status o
f the consequence relation is noted.