Tuesday, July 30, 2024

Phasal versus Non-Phasal Movement

Abstract: In this short speculative blog post, I introduce the distinction between phasal and non-phasal movement. I show how the distinction fits in with the distinction between A and A’-movement. I discuss the implications of the distinction in resolving the freezing dilemma for smuggling.

1. Introduction

With the exception of head movement, movement operations are taken to fall into one of two categories: A-movement and A’-movement. A-movement is movement to an A-position (e.g., Spec TP), and A’-movement is movement to an A’-position (e.g., Spec CP). To a first approximation, A-movement is triggered by phi-features, and is related to Case checking. A’-movement is triggered by operator features, and is related to focus and other discourse properties. 

But some kinds of movement do not fit naturally into this dichotomy. For example, Collins 2005 (see also Collins 2024) argues extensively that the passive involves movement of a PartP to Spec VoiceP, illustrated below:

1.

[ PartP Voice [EA v <PartP>]]

In this diagram, PartP starts out as the complement of v, and moves over EA (external argument), landing in Spec VoiceP. In Collins 2005, this kind of derivation is called smuggling, because the movement of the PartP smuggles the direct object over the intervening EA (on smuggling, see Belletti and Collins 2020).

Is the movement of PartP in (1) A-movement or A’-movement? 

It does not fit naturally into either category of movement types. It is definitely not operator movement, because the participle is not focused in anyway. Nor does it have any other kind of operator feature. It also does not seem to involve A-movement, which is usually movement of a DP. 

Other types of movement that raise similar issues for the A versus A’ distinction are found in Kayne 1994 (e.g., TP movement internal to the CP) and Cinque 2005 (NP movement internal to the DP) and VP movement in quotative inversion (see Collins 2003, Storment 2024, forthcoming).

2. Phasal versus Non-Phasal Movement

I propose that there is a three-way distinction in movement types (once again, putting aside head movement, which clearly counts as non-phasal movement). On the one hand, there is phasal movement, which is movement of a phase (e.g., DP, CP, vP, etc.). All A and A’-movements are phasal movements. For example, movement of vP to Spec CP is a classic case of phasal movement:

2.

I told John to do his homework, and [do his homework], he did.

In this example, the vP [do his homework] moves to Spec CP, in a kind of A’-movement.

On the other hand, there is non-phasal movement, which is movement of a non-phase (any category other than the phasal categories of DP, CP, vP, etc.). For example, movement of the PartP in (1) is non-phasal movement.

For brevity sake, I will call phasal movement P-movement, and non-phasal movement P’-movement.  On the basis of this dichotomy, we can also define P-positions (positions occupied by phases) and P’-positions (positions occupied by non-phases). Just like A-positions do not block A’-movement (see Rizzi 1997 and much later work), we expect that P-positions will never block P’-movement. Consider again the example in (1), the EA (external argument) is in Spec vP, a classical A-position, but it does not block PartP movement. The reason is that A-positions do not block P’-movement. A similar lack of blocking (for VP movement) can be shown for quotative inversion, see Collins 2003, Storment 2024, forthcoming.

Since movement to Spec CP is always P-movement, it follows that P’-movement can never leave a CP, and so is always local to a single CP. In this, P’-movement is similar to A-movement, which also cannot land in Spec CP (because of the constraint on improper movement). Both P’-movement and A-movement are contained within a single finite CP.

3. Smuggling and Freezing

An often-remarked property of smuggling derivations of the kind shown in (1) is that they violate the freezing constraint (see Corver 2017 for a comprehensive overview). In particular, since PartP has moved to Spec VoiceP, it should obey the freezing constraint. That means that any extraction from PartP (in Spec VoiceP) will be blocked, since PartP is frozen.

But it is essential to the smuggling analysis of Collins 2005 that the direct object be able to move out of PartP into Spec TP:

3.

[ DP T [was [ [Part [write <DP>]] Voice [EA v <PartP>]]

“The book was written by John.”

In this derivation, PartP raises to Spec VoiceP. Then the DP internal to PartP moves to Spec TP. Such movement should be impossible by the freezing constraint, since PartP has undergone movement. See Collins 2005 and 2024 for details.

I propose that only P-movement obeys the freezing constraint. Once a phase A has undergone movement, it is completely frozen so that no subconstituent can undergo movement out of A. But if a non-phasal constituent undergoes movement (internal to a phase), it is not frozen, and subextraction is possible. This is what allows the movement of DP from PartP in 3 above.

Following Collins 2005, in (1) and (3), the phase is VoiceP. PartP moves to Spec Voice, an instance of P’-movement. Since PartP is in the edge of VoiceP, and since PartP is not frozen, nothing blocks A-movement of the direct object to Spec TP.

Why is movement of a non-phase not frozen? I speculate that freezing is the result of spelling out a moved constituent. But in syntax, only phases and parts of phases are spelled out. First, when a phase is formed, its complement domain is spelled out. Then after the phase is moved, I propose that the whole phase is spelled out, effectively freezing its contents and so no further movement out of the moved phase is possible. Non-phases are not spelled out in this way. Rather, they only get spelled out when a larger containing phase is spelled out.

If these speculations are correct, the difference between P and P’ movement with respect to freezing is related to the way that phases are spelled out in a derivation.

4. Conclusion

In summary, I have argued for a very general distinction between two types of movement: phasal and non-phasal movement, with the following properties:

4. P-movement

a. Movement of a phase.

b. Examples: movement of vP, DP, CP.

c. Either A or A’-movement.

d. Obeys freezing constraint.

5. P’-movement

a. Movement of a non-phase.

b. Examples: movement of PartP, VP, NP, TP

c. Includes head movement

d. Does not obey freezing constraint.

This blog post is meant as speculative. I leave it to further research to flesh out the details of this distinction.

References:

Belletti, Adriana and Chris Collins. 2020. Smuggling in Syntax. Oxford: Oxford University Press.

Cinque, Guglielmo. 2005. Deriving Greenberg’s Universal 20 and Its Exceptions. Linguistic Inquiry 36, 315-332.

Collins, Chris. 2003. The Distribution of Particle Verbs in Quotative Inversion. Manuscript, Cornell University.

Collins, Chris. 2024. Principles of Argument Structure: A Merge-Based Approach. MIT Press, Cambridge.

Corver, Norbert. 2017. Freezing Effects. In Edited by Martin Everaert and Henk C. van Riemsdijk (eds.), The Wiley Blackwell Companion to Syntax, Second Edition. John Wiley and Sons.

Kayne, Richard. 1994. The Antisymmetry of Syntax. MIT Press, Cambridge.

Rizzi, Luigi. 1997. Relativized Minimality. MIT Press, Cambridge.

Storment, John David. 2024. Quotative Inversion as Smuggling: Evidence from Setswana and English. Presented May 3, 2024 at the 55th Annual Conference on African Linguistics, Montreal, Québec.

Storment, John David. Forthcoming. Projecting (Your) Voice: A Theory of Inversion and Circumvention. Doctoral Dissertation, Stony Brook, New York.


No comments:

Post a Comment

Note: Only a member of this blog may post a comment.