A NOT "MERELY EMPIRICAL" ARGUMENT FOR A LANGUAGE OF THOUGHT