Perceptual symbols in language comprehension: can an empirical case be made?