Is there evidence for automatic processing of spatial and color attributes present in pictures and words?