The Left Edge Ban and the Architecture of the Grammar

Abstract This book proposes and motivates a ban on words and sentences of a particular form. It shows that there are differences between the strategies by which this ban is satisfied at the word level, and the way this ban is satisfied at the sentence level. One striking finding is that syntactic operations may be motivated as a means to avoid violating the ban, which we might not expect on a strictly feed-forward grammatical architecture. A theoretically interesting consequence of this is that the syntax must be able to see at least some information that the phonology must be able to have access to, motivating an architecture of the grammar that is different from that commonly assumed. Empirically, this ban allows an understanding of a variety of apparently disparate linguistic facts as having the same underlying cause, including, but not limited to: cross-linguistic skews in attested stress patterns, morphophological differences between prefixes and suffixes, restrictions on certain disharmonic word orders and cases where this restriction is apparently lifted, and a requirement that heavy elements undergo obligatory extraposition from a variety of fronted constituents. Each of these cases can be understood as the consequence of the grammar of a particular language employing a limited set of strategies to ensure the proposed ban is satisfied.