# V to T movement in German

Consider the the embedded clause "du Schach gespielt hast" in this sentence

Ich glaube dass du Schach gespielt hast.

I think that you chess played have

‘I think that you have played chess.’

How do I draw the underlying structure ("deep structure") and transform it into the "surface structure" and see whether there is V to T movement?

Assuming the VP internal subject hypothesis, one option is that the embedded TP has two branches: the empty DP and the T'. The T' node branches to VP (on the left) and T (on the right). The T has "hast/have" in it already in the level of D-structure. And the VP has two branchs: the DP "du/you" and the V' which branches further to the DP "chess" and V "played".

If "hast" is not present in T in the underlying structure but moves to T from somewhere, I don't see how to draw the underlying tree. And how to see if it moves or not?

(I'm working in the framework of the first 10 chapters of Carnie)

Addition: it is also known that the following sentences are ungrammatical:

* Ich glaube dass du hast Schach gespielt.

* Ich glaube dass du hast gespielt Schach.

* Ich glaube dass Schach hast du gespielt.

• Those examples marked as wrong are orthographically wrong, but "Ich glaube das: Du hast Schach gespielt" is not only correct, though nowhere common, but also my best guess at the origin of the "dass ..." construction following DWDS, though I'm rather skeptical; It would still be natural if "das" were a relative pronoun meaning ca "what was said" instead of a determiner. – vectory Nov 1 '19 at 20:19
• The second one is completely unnatural, and I think that's the clue, in that "schach spielen" may be thought of as one word (as indeed "Schachspiel" is a fixed though transparent compound). Just today, given the thread speculating about Proto-Indo-European "and", that was bumped to the front page, I wondered how--or whether at all--German perfective ge-, akin to Lat. co- "with", was an obligatory infix similar to and; compare "go and see", "I tried and nailed it", etc, perhaps in combination with the prefix "be-" ~ "by". – vectory Nov 1 '19 at 20:23
• Third, since dass-relative clauses aren't obligatory, and kinda ugly, they can almost always be reordered to make sense: "Ich glaube, du gehst jetzt besser", "Ich glaube, dass du nicht verstanden hast", "Ich glaube, das wird nichts"; Consequently, "* Ich glaube, das Schach hast du gespielt" would be odd only because Schach is uncountable and doesn't take an article; other nouns would disagree in gender of the article ("Ich glaube, die Frage habe ich nicht richtig verstanden", "Ich gebe zu, dass ich die Frage falsch verstanden habe"). – vectory Nov 1 '19 at 20:36
• Take a look at work by Hubert Haider and Josef Bayer: they both argue that there is no V-to-T in German, and that there is no T evidence to assume T in German to begin with. Also, there is no evidence to assume obligatory raising of subjects in German. Furthermore, German does only have semi-auxiliaries (aux that behave just like lexical verbs in almost every syntactic regard). They form clusters (word-like structures) with the verbs they embed ('coherent verb constructions' in Bech's terms). As a result, the assumed phrase structure would simply be [CP dass [VP du [Schach [[gespielt] hast]]]] – aslakr Dec 5 '19 at 17:53

`du Schach gespielt hast` as you say is an embedded clause, and string-identical to the underlying form Carnie is referring to. (To answer your first question.)

As for the question on how to detect V->T movement, see @aslakr's comment on whether we should even assume that for German. My own work currently suggests for Alemannic (which is not Standard German, but related to it) that V->T movement does exist, so I'm biased here myself.

A way to find out about T, V (and little v etc.) levels in a phrase when by looking at the string is not obvious, is inserting T or V level adverbials:

``````1.a. du heute Schach gespielt hast

b. du heute einhändig Schach gespielt hast

c. ?du einhändig heute Schach gespielt hast
``````

The ? in c means it's not perfect (still not ruled out though in German, where almost every word order has some meaning). German is a bad example language to study this kind of stuff, since word order is a mess (see research on "Scrambling"). I thought that the Carnie textbook you're referring to has also French and English example pairs on this topic.