Charniak (1993) covers statistical language learning from the perspective of computer science and computational linguistics, and is recommended.
McEnery and Wilson (1996) is strongly recommended. An electronic version is available at http://www.ling.lancs.ac.uk/monkey/ihe/linguistics/contents.htm . This text is less concerned than we are to provide details of the underlying statistical and computational ideas, but is an admirable survey of the landscape of the field.
One introduction, written for linguists, which comes very close to our approach is Abney (1996). This makes a very clear case for the use of statistical methods in linguistics. The commentary on Chomsky's arguments against Shannon is particularly acute and apposite.
Krenn and Samuelsson (1997) is an ongoing effort to provide a very comprehensive guide to statistics for computational linguists. It should suit those who want to add a clear and formal grasp of the maths and stats to a pre-existing background in computational linguistics. It is maths-heavy and application-light, so it is better for people who already know why the applications to which statistics are being applied are of interest.