Overall: Excellent work -- Very impressive.  You have presented a thoroughly thought
out, quite (appropriately) sophisticated analysis of a rather difficult problem, and, as
usual, your concerns and side issues are exactly those that should concern and attract
you. The balance struck between the particulars of the language and concerns of
theoretical elegance is just right throughout.

Thank you for seeing to it that the pages were numbered this time.

Some specific observations:

p.2	Tree (1a)

Do you really think "beag compordach" is a constituent, as your tree indicates?  (The
brackets seem to say something different.)  Remember what adjunction is for.

" ... a right adjunct in English is unusual ..."

How about PP adjuncts to NP, VP?

p.3, (1e): the PP `as an rialtas' is surely a complement rather than an adjunct

p.3, last paragraph: excellent discussion, ...

p.3, bottom: "it seems propitious to adopt the hypothesis that AP/DegP adjunction occurs
at the NP level."

Then why don't you assume that in the tree for (1a)?

p.4, top: "Finite clauses in Irish ... more closely resemble the embedded clauses we've
seen in German."

??  Surely not in terms of word order?  The discussion here is quite muddled.  What is
the purpose of the tree?  And why does the tree show the DP "an cat" as a *right*
specifier of VP, when there is no evidence for such an ordering anywhere?  (It looks
like you are systematically saying "left" when you mean "right", but that's just a
guess.) It's also not clear to us what the phrase `due to the ancestral structure of the
trees themselves' means.

p.5: Now this makes sense and the discussion of filled C-positions is particularly good.

pp.7-8: the discussion of the complement structures of perception verbs is excellent --
perceptive and accurate.

pp.8-9: Excellent discussion; but why can't 'ag' be a T?

p.10, around (12z): the printed copy is messed up at this point.

p.10, Part IV:  Looks like you nailed it.

p.11: Concerning the idea that Neg raises to C: I wonder if negative clauses can be
embedded (with an overt C).

P.11: Why assume that the entire NegP gets deleted when the Neg head raises to C?

p.12, end of Part IV and beginning of Part V: *Excellent* discussion.

p.13, "zombie movement" -- interesting concept.

Very clear and sophisticated discussion of the theoretical conjecture.

p.13, last line: `equivilant' --> `equivalent' (think of it as `equi-valent')

p.14: "... the idea of strong head control over the operations of its phrase ..."

What does this very impressive-sounding idea mean?

p.15: Can't *one* of you learn the rules of punctuation around 'however'?

p.16: " either constituents may raise from deleted phrases, or head movement is not
final."  -- An intriguing place to get to.

pp. 17-18: Beautiful treatment of the antepenultimate problem. It is very perceptive and
asks just the right questions.

pp.19-20, on Gapping: You have opened yourselves up a nice can of worms.  Just a couple
of comments:

p.19: But V-T raising is *not* possible for main Vs in English.

p.20 "John does/did play the violin and Susan the clarinet."

These are perfectly grammatical.

p.21: "If Gapping is not ABM then why do they appear to be so similar?"

That's a very good question.  We'll try to find ways to address it in the coming weeks.
Here we will just note that some contemporary syntacticians have been trying to analyze
Gapping as involving ABM rather than ellipsis.

p.22: Very interesting line of thought.  This idea too is echoed in certain contemporary
views about the overall operation of grammatical processes.  You are just about ready
for graduate school.