Section 3.4 of Guess Who has what Ross calls "Themost interesting consequence of Sluicing": Namely, the apparent amelioration of island violations. Ross argues that we need 'global' (i.e., non-Markovian) statements of the island constraints. This triggered some very interesting debate and discussion. The early chapters of that debate are, in the following order:

1. Lakoff (1970) Section 4, pp. 632-633
2. Chomsky (1972) pp. 72-73
3. Baker and Brame (1972) Section 4, pp. 60-63
4. Lakoff (1972) Section 1.3, pp. 81-82

Later in the semester, we will return to the question of repair of island violations, that has once again become a major topic of research in the last two decades, inspired especially by Merchant (1999). We will look carefully at several chapers of that thesis later.