Editing Minimalist program (section)

== Derivation by phase ==

A '''phase''' is a syntactic domain first hypothesized by Noam Chomsky in 1998.<ref>Chomsky, Noam (1998). "Minimalist Inquiries: The Framework" MIT Occasional Papers in Linguistics 15. Republished in 2000 in R. Martin, D. Michaels, & J. Uriagereka (eds.). ''Step By Step: Essays In Syntax in Honor of [[Howard Lasnik]]''. 89–155. MIT Press.</ref> It is a domain where all derivational processes operate and where all features are checked.<ref name=":2">{{Citation|title=Derivation by Phase|date=2001|url=http://dx.doi.org/10.7551/mitpress/4056.003.0004|work=Ken Hale|publisher=The MIT Press|doi=10.7551/mitpress/4056.003.0004|isbn=978-0-262-31612-5|access-date=2020-12-04 |last1=Chomsky |first1=Noam |pages=1–52 |url-access=subscription}}</ref> A phase consists of a phase head and a phase domain. Once any derivation reaches a phase and all the features are checked, the phase domain is sent to transfer and becomes invisible to further computations.<ref name=":2" /> The literature shows  three trends relative to what is generally considered to be a phase:

# All CPs and some vPs are phases: Chomsky originally proposed that CP and vP in transitive and [[unergative verb]]s constitute phases. This was proposed based on the phrases showing strong phase effects discussed  above.<ref name=":1" />
# A specified set of phrases are phases: CP, DP (based on parallels between DP and CP<ref>{{Citation|last=Svenonius|first=Peter|title=On the Edge|date=2004|url=http://link.springer.com/10.1007/1-4020-1910-6_11|work=Peripheries|volume=59|pages=259–287|editor-last=Adger|editor-first=David|series=Studies in Natural Language and Linguistic Theory|place=Dordrecht|publisher=Springer Netherlands|doi=10.1007/1-4020-1910-6_11|isbn=978-1-4020-1908-1|access-date=2020-12-05|editor2-last=De Cat|editor2-first=Cécile|editor3-last=Tsoulas|editor3-first=George|url-access=subscription}}</ref>), all ''v''Ps,<ref name=":6" /> TP (in some languages<ref name=":3">See Assmann et al. (2015) Ergatives Move Too Early: On an Instance of Opacity in Syntax. Syntax 18:4 pp. 343–387</ref>)
# Every phrase is a phase, with moved constituents cycling through all intermediate phrase edges.<ref name=":3" />

=== Strong phases: CP and ''v''P  ===
A simple sentence can be decomposed into two phases, CP and ''v''P. Chomsky considers CP and ''v''P to be strong phases because of their  propositional content, as well as their interaction with movement and reconstruction.<ref name=":1">{{Cite book|last=Chomsky|first=Noam|title=Minimalist Inquiries: The Framework. In Roger Martin, David Michaels, and Juan Uriagereka, eds., Step by Step: Essays on Minimalist Syntax in Honor of Howard Lasnik|publisher=MIT Press|year=2000|isbn=026213361X|location=Cambridge, Mass.|pages=89–156}}</ref>

'''Propositional content''': CP and vP are both propositional units, but for different reasons.<ref name=":0">{{Cite journal|last=Obata|first=Miki|date=2006-01-01|title=Phase and Convergence|url=https://repository.upenn.edu/pwpl/vol12/iss1/23|journal=University of Pennsylvania Working Papers in Linguistics|volume=12|issue=1}}</ref> CP is considered a propositional unit because it is a full clause that has tense and force: example (1) shows that the complementizer ''that'' in the CP phase conditions finiteness (here past tense) and force (here, affirmative) of the subordinate clause. ''v''P is considered a propositional unit because all the [[theta role]]s are assigned in ''v''P: in (2) the verb ''ate'' in the ''v''P phase assigns the Theme theta role to the DP ''the cake'' and the Agent theta-role to the DP ''Mary''.<ref name="sportiche2014" />
 (1) ''John said'' [<sub>CP</sub> '''''that Mary will eat the cake''''' ].

 (2) [<sub>CP</sub> ''Mary'' [<sub>''v''P</sub> '''<Mary> ''ate the cake''''' ].

 {| class="wikitable"
 |<!--col1-->[[File:Syntax_Tree_of_simple_sentence.png|300x300px]]
 |<!--col2-->[[File:Propositional Evidence of CP phase.png|300x300px]]
 |<!--col3-->[[File:Propositional Evidence of vp phase.png|300x300px]]
 |-
 |<!--col1-->structure of simple sentence
 |<!--col2-->CP phase: ''that Mary will eat the cake''
 |<!--col3-->''v''P phase: ''Mary ate the cake''
 |}

'''Movement''': CP and vP can be the focus of [[Cleft sentence|pseudo-cleft]] movement, showing that CP and ''v''P form syntactic units: this is shown in (3) for the CP constituent ''that John is bringing'' ''the dessert'', and in (4) for the ''v''P constituent ''arrive tomorrow''.<ref name=":0" />
 (3) a.  ''Mary said'' [<sub>CP</sub> '''''that John is bringing the dessert'''''].
     b.  ''What Mary said was'' [<sub>CP</sub> '''''that John is bringing the dessert'''''].

 (4) a. ''Alice will'' [<sub>vP</sub> '''''arrive tomorrow'''''].
     b. ''What Alice will do is'' [<sub>vP</sub> '''''arrive tomorrow'''''].
'''Reconstruction.''' When a moved constituent is interpreted in its original position to satisfy binding principles, this is called reconstruction.<ref name=":5" /> Evidence from reconstruction is consistent with the claim that the moved phrase stops at the left edge of CP and ''v''P phases.<ref name=":6">See, among others, Legate, Julie Anne. 2003. Some Interface Properties of the Phase. Linguistic Inquiry 34: 506–516 and Chomsky, Noam. 2008. On Phases. In ''Foundational Issues in Linguistic Theory. Essays in Honor of Jean-Roger Vergnaud''. eds. Robert Freidin, Carlos Peregrín Otero and [[Maria Luisa Zubizarreta]], 133–166. Cambridge, Massachusetts: MIT Press</ref>

* '''Reconstruction at left edge of CP phase''': In (5), the reflexive ''himself''  can be understood as being co-referential with either ''John'' or ''Fred'', where co-reference is indicated by co-indexation. However, the constituent that contains ''himself'', namely the sentence-initial phrase [''which picture of himself''], is not c-commanded by either ''John'' or ''Fred'', as is required by Principle A of the Binding Theory. The fact that co-indexation of ''himself'' with either one of ''John'' or ''Fred'' is possible is taken as evidence that the constituent containing the reflexive, namely [''which picture of himself''] has moved through a reconstruction site—here the left edge of the lower CP phrase—from where it can satisfy Principle A of the Binding Theory relative to the DP ''John''.<ref name=":0" />
*'''Reconstruction at left edge of vP phase''': In (6), bound variable anaphora requires that the pronoun ''he'' must be c-commanded by ''every student'', but Condition C of the Binding Theory requires that the R-expression ''Mary'' be free. However, these requirements cannot be satisfied by the sentence-initial constituent that contains both ''he'' and ''Mary'', namely the phrase [''which of the papers that he gave Mary'']. The fact that the sentence is nevertheless well-formed is taken to indicate that this phrase must have moved through a reconstruction site first, from where it is interpreted. The left edge of the ''v''P phase is the only position where these binding requirements could be satisfied: (i) ''every student'' c-commands the pronoun ''he''; (ii) ''Mary'' is free from any c-commanding DP.<ref name=":6" />
 (5) a. [Which picture of '''himself'''<sub>k</sub>] did '''John'''<sub>k</sub> think ___ '''Fred'''<sub>j</sub> liked __?
     b. [Which picture of '''himself'''<sub>j</sub>] did '''John'''<sub>k</sub> think ___ '''Fred'''<sub>j</sub> liked __?

 (6) [Which of the papers that '''he'''<sub>k</sub> gave '''Mary'''<sub>j</sub>] did every student<sub>k  __</sub> ask her<sub>j</sub> to read __ carefully?

 {| class="wikitable"
 |<!--col1-->[[File:Syntax Tree- Reconstruction effects of vP phase.png|400x400px]]
 |<!--col2-->[[File:Syntax tree- Reconstruction effect of vP phase.png|400x400px]]
 |-
 |<!--col1-->reconstruction at left edge of CP phase
 |<!--col2-->reconstruction at left edge of vP phase
 |}

=== Phase edge ===
Chomsky theorized that syntactic operations must obey the phase impenetrability condition (PIC) which essentially requires that movement be from the left-edge of a phase. The PIC has been variously formulated in the literature. The [[extended projection principle]] feature that is on the heads of phases triggers the intermediate movement steps to phase edges.<ref name=":0" />

====Phase impenetrability condition (PIC)====

Movement of a constituent out of a phase is (in the general case) only permitted if the constituent has first moved to the left edge of the phase (XP).

The edge of a head X is defined as the residue outside of X', in either specifier of X and adjuncts to XP.<ref>{{Cite book|last=Chomsky|first=Noam|url=https://books.google.com/books?id=E-IbAQAAIAAJ|title=Derivation by Phase|date=1999|publisher=MIT, Department of Linguistics|language=en}}</ref>

English successive cyclic wh-movement obeys the PIC.<ref name=":0" /> Sentence (7) has two phases: ''v''P and CP. Relative to the application of movement, ''who'' moves from the (lower) ''v''P phase to the (higher) CP phase in two steps:

*Step 1: '''''who''''' moves from the complement position of VP to the left edge of ''v''P, and the EPP feature of the verb forces movement of '''''who''''' to the edge of vP.
* Step 2: '''''who'''''  moves from the left edge of the lower ''v''P phase to the specifier of the (higher) CP phase.

 (7)  [<sub>CP</sub> '''''Who''' did you'' [<sub>vP</sub> ''see'' <s>who</s>]]?

 {| class="wikitable"
 |<!--col1-->[[File:Phasal_Movement_Syntax_tree-_step_1.png|300x300px]]
 |<!--col2-->[[File:Phasal_movement_tree-_step_1.png|300x300px]]
 |-
 |<!--col1-->Step 1: wh-phrase moves to edge of vP phase
 |<!--col2-->Step 2: wh-phrase moves to edge of CP phase
 |}

====Medumba wh-movement====
Another example of PIC can be observed when analyzing A'-agreement in [[Medumba language|Medumba]]. A'-agreement is a term used for the morphological reflex of A'-movement of an XP.<ref name=":5" /> In Medumba, when the moved phrase reaches a phase edge, a high low tonal melody is added to the head of the complement of the phase head. Since A'-agreement in Medumba requires movement, the presence of agreement on the complements of phase heads shows that the wh-word  moves to the edges of phases and obeys PIC.<ref name=":5">{{Cite thesis|last=Keupdjio|first=H. S.|date=2020|title=The syntax of A' -dependencies in Bamileke Medumba (T).|url=https://open.library.ubc.ca/collections/ubctheses/24/items/1.0389713|journal=University of British Columbia|doi=10.14288/1.0389713}}</ref>

Example:<ref name=":5" />

The sentence (2a) has a high low tone on the verb '''''nɔ́ʔ''''' and tense '''''ʤʉ̀n''''', therefore is grammatical.

(2a) [<sub>CP</sub> á wʉ́ Wàtɛ̀t '''nɔ́ɔ̀ʔ''' [<sub>vP</sub> '''ⁿ-ʤʉ́ʉ̀n''' á?]]

'Who did Watat see?'

The sentence (2b) does not have a high low tone on the verb '''''nɔ́ʔ''''' and tense '''''ʤʉ̀n''''', therefore is not grammatical.

(2b) *[<sub>CP</sub> á wʉ́ Wàtɛ̀t '''nɔ́ʔ''' [<sub>vP</sub> '''ⁿ-ʤʉ́n''' á?]]

<nowiki>*</nowiki>'Who did Watat see?'

To generate the grammatical sentence (2a), the ''wh''-phrase ''á wʉ́'' moves from the vP phase to the CP phase. To obey PIC, this movement must take two steps since the wh-phrase needs to move to the edge of the vP phase in order to move out of the lower phase.

* Step 1: First, the ''wh''-phrase moves from the complement of VP to the edge of the vP phase to avoid violating PIC. In this position, the agreement is expressed on the verb ''ʤʉ̀n'' and surfaces as a high low (HL) tone melody (''ⁿ-ʤʉ́ʉ̀n''). The agreement is expressed on the verb which is the head of the complement of the v phase head.
[[File:Syntax_tree_wh-movement_in_Medumba.png|none|thumb|297x297px|Step 1: wh-phrase moves to edge of vP]]
* Step 2: Now that it is at the edge of the vP phase, the ''wh''-phrase is able to leave the vP phase and move to the Spec-C position of the CP phase. Agreement is expressed on the tense ''nɔ́ʔ'' as a high low tone melody (nɔ́ɔ̀ʔ). The tense which agreement is expressed on is the head of the complement of the C phase head
[[File:Syntax_tree_Wh-movement_in_Medumba_(part_2).png|none|thumb|293x293px|Step 2: wh-phrase moves from vP phase to CP phase]]

One can confirm that A' agreement only occurs with movement by examining sentences where the wh-phrase does not move. In sentence (2c) below, one can observe that there is no high low tone melody on the verb '''''nɔ́ʔ''''' and tense '''''fá''''' since the wh-word does not move to the edge of the vP and CP phase.<ref name=":5" />

(2c) [m-ɛ́n '''nɔ́ʔ''' '''fá''' bɔ̀ á wʉ́ á]

'The child gave the bag to who?'

=== Cycle ===
The spell-out of a string is assumed to be cyclic, but there is no consensus about how to implement this. Some analyses adopt an iterative spell-out algorithm, with spell-out applying after each application of Merge. Other analyses adopt an opportunistic algorithm, where spell-out applies only if it must. And yet others adopt a wait-til-the-end algorithm, with spell-out occurring only at the end of the derivation.

There is no consensus about the cyclicality of the Agree relation: it is sometimes treated as cyclic, sometimes as a-cyclic, and sometimes as counter-cyclic.