“…However, lip-reading alone does not guarantee a good understanding of speech, because visual information is far more ambiguous than auditory information: Spoken words often occur phonetically underspecified on the lips of the speaker (Dodd, 1987;Miller, 2006). Furthermore, as pointed out by a series of studies (see, e.g., Alegria & Lechat, 2005;Gregory & Hindley, 1996), the articulation configuration of a certain phoneme may vary depending on the articulation configuration of the previous and the following phoneme (this phenomenon is called coarticulation), and some letters (such as /B/P/ and /T/ D) are generated by a visual identical lip pattern (homologia). In particular, grammatical morphemes such as articles, prepositions, conjunctions, pronouns, and verbal auxiliaries are difficult to recognise as they are short items that are produced rapidly and with loss stress (Pizzuto, Caselli, & Volterra, 2000;Volterra, Capirci, & Caselli, 2001).…”