“…Studies of intelligibility have consistently demonstrated reliable differences in performance between natural and synthetic speech. In contrast, comprehension studies using passages of connected speech have produced conflicting evidence, some reporting sizable differences between natural and synthetic speech (Hersch and Tartarglia, 1983;Luce, 1981;Moody and Joost, 1987) and others reporting no differences at all (Jenkins and Franklin, 1981;McHugh, 1976;Pisoni and Hunnicutt, 1980;Schwab, Nusbaum, and Pisoni, 1985). How can these findings be accounted for in a principled way?…”