Radford (1997) vs. my handouts

Throughout the semester, there have been various points where I have diverged in my presentation on the handout from what Radford has in his textbook (for what I think are good reasons). However, it may be somewhat hard to keep track of, and so I am collecting here some comments on Radford's book and how his analyses relate to the ones I presented. This is still a work in progress. I intend to complete it in a couple of days and then keep it up to date for the rest of the semester.

Comments:

On participles: There was a bit of discussion about the features [+Participle] and [+Past]. My reasoning on this has been repeated in various places,but my basic objection to the features that Radford used is that [+Past] should be a feature of I and not a feature of the participle. Instead, I suggested [+Perfect] for participles like written and [+Progressive] for participles like writing.

Concerning the labeling of nodes, I depart from Radford's convention and label any node in the tree which is both a maximal and a minimal projections as being an XP (rather than an X). I do this because for the vast majority of the situtations we see in which something moves in the tree, it involves movement to a specifier or an XP-adjoined position, things that only maximal projections can participate in. So, for example, on p. 87, (5) would have an NP and not an N if drawn my way.

Morphology and underlying forms (week 3a: PowerPoint, PDF). The view of morphology (e.g., the morphology of be) I am putting forward here differs from Radford's view (which he discusses in his chapter 4). What I have presented to you is essentially the Distributed Morphology theory of the syntax-morphology mapping (Halle & Marantz 1993, and others). I recently found a web page devoted to DM if you're looking for more technical information.

Constituency tests. Skipped in chapter 3 was most of the discussion of sections 3.7-8 (Testing structure, Additional tests). The concept of coordination as a test has come up since, occasionally.

I augmented some of the argumentation for DP in the handout. Radford essentially just gave it to us because 'this is how people do it in recent work' (paraphrased, but from p. 96). I also borrowed some of the discussion here from chapter 4 (section 4.7 on null determiners and section 4.8 on pronouns). I'll return below to divergences.

Binding theory. Radford basically skipped binding theory altogether, except for a couple of comments on pp. 114-116. For a secondary reference, you might look at Santorini's textbook (get comments from blog).

Head Movement. The presentation of head movement on the handout is fairly loosely related to the discussion in the textbook,but most of the concepts are the same. The textbook does not discuss Irish or French, however. Standard textbooks do. Other places to look might be Carnie's textbook, maybe Santorini's textbook.

HM: 6.5 strong and weak features. Radford's use of the term "strong" and "weak" is rather nonstandard, and we are not following it. His discussion of greed and attraction in 6.5 we are not following.

Neg: 6.6 His view of negation (headed by not) we are not following, because we have a more sophisticated view based on French. Where can the spec-verison be looked up, where can the SpecNegP version be looked up? Check Santorini's textbook.

6.7 Syntax of have -- pretty much not dealt with

6.8 "Tense affix" and the introduction of TP. We'll introduce TP with AgrP in chapter 10, Radford's reasoning in chapter 6 makes no sense. We'll also not split I up into a base-generated head-adjunct. Modals are I (or T) for our purposes. Also avoids the rather ugly excoporation analyss required in 6.9.

6.9 Question affix. We assume I moves to C in its entirety (we don't excoroporate a modal from I). We'll also assume if is not adjoined to Q, it just is Q.