Pedagogical Grammar ENG 240 Fall 2024 PDF

Pedagogical Grammar ENG 240 Fall Semester 2024 Mondays & Thursdays, 13.30–15.00 Chapter 6  (Im)possible Word Orders The word order in (1) is ungrammatical, but that in (2) is fine: (1) *Adrian has liked always this yellow chair. (2) This yellow chair, Adrian has a...

Pedagogical Grammar ENG 240 Fall Semester 2024 Mondays & Thursdays, 13.30–15.00 Chapter 6  (Im)possible Word Orders The word order in (1) is ungrammatical, but that in (2) is fine: (1) *Adrian has liked always this yellow chair. (2) This yellow chair, Adrian has always liked. And this goes across a clause boundary, and even more: (3) This yellow chair, I never said [ that John has always liked ]. (4) This yellow chair, I don’t think [ I ever said [ that John has always liked ]].  Asking Questions (5) Which yellow chair has Adrian always liked __? A less frequent variant of this (echo question) would be: (6) Adrian has always liked which yellow chair? And asking a question can also cross clause boundaries: (7) Which yellow chair did you say [ that Adrian has always liked __ ]? (8) Which yellow chair do you think [ that I said [ that Adrian has always liked __ ]]?  Towards Movement (9) This yellow chair, Adrian has always liked. Here we move this yellow chair from its original complement position of VP to a sentence-initial position (to be identified). (10) Which yellow chair is such that you think that I said that Adrian has always liked this yellow chair? (11) Which pictures of herself do you think that Mary posted? (12) Which pictures of herself do you think that Mary posted ? Here we move which pictures of herself from its VP-position to the front of the whole.  Merge Again (13) Before Merge: After Merge: W X WP 3 W XP (14) YP 3 Y WP This WP can now be merged with some 3 other element Y, W XP which then projects.  And Another Merge (or Two) (15) ZP 3 And this YP now can Z YP then be merged with 3 some further element Z, which subsequently Y WP projects. 3 W XP (16) YP Alternatively, Y+WP 3 projects to Y’ to which But of course our XP is XP Y' then a specifier XP is the same we had earlier—that is, we 3 merged before YP. can understand Y WP movement of XP as 3 re-merging this XP. W XP  Movement as Remerge (17) Which yellow chair has Adrian __ always liked __? (18) FinP 3 DP Fin' Adrian 3 Fin VP has % always liked this yellow chair  The Auxiliary in Wh-Questions In English wh-questions, there is so-called auxiliary inversion— where subject and auxiliary invert (‘swap position’). Plus: (19) a. Which yellow chair has John been trying to find? a'. *Which yellow chair has been John trying to find? b. Which yellow chair will John be trying to find? b'. *Which yellow chair will be John trying to find? c. Which yellow chair must John have tried to find? c'. *Which yellow chair must have John tried to find? Wh-Movement & Auxiliary Movement (20) XP 3 DP X' which chairs 3 X FinP has 6 Adrian always liked  Embedded Clauses: The CP (21) I don’t think that Adrian has always liked yellow chairs. (22) V' 3 V CP think 3 C FinP that 6 Adrian has always liked yellow chairs (23) I asked if Adrian had always liked yellow chairs.  C-Features Complementisers head CP and as such (like other categories) have an interpretable categorial feature [C]. So, complementisers like that and if introduce a CP—but of course these two complementisers are different (see Hwk. 1): (24) a. that: [C: Declarative] (for short: [C: D]) b. if: [C: Question] (for short: [C: Q]) Thus, that introduces embedded declaratives, if embedded questions—a perhaps better term for CP as Complementiser Phrase might then be CP as Clause-type Phrase. Wh-Questions and Features within CP And we have seen differences with respect to the head of CP (vis-à-vis auxiliary inversion) differences in embedded vs. non- embedded questions. Let’s capitalise on this with C having two relevant subfeatures, C: Non-embedded and C: Question. (25) CP wo DP C' which yellow chair wo C FinP [C: Non-emb., Q] 6 has Adrian always liked Has Adrian always liked this yellow chair? (26) CP wo C FinP [C: Non-emb., Q] wo DP Fin’ Adrian ei Fin VP has 6 always liked this yellow chair  Questions vs. Declaratives Now, the feature [C: Non-emb., Q] signals that this sentence should be a non-embedded matrix question like (27a) rather than a matrix declarative as in (27b): (27) a. Has Adrian always liked the yellow chair? b. Adrian has always liked the yellow chair. So, what needs to happen to (26) is movement of Fin to C. The same movement is also required for matrix wh-questions, as the following tree illustrates. Thus for (26) to proceed to give us (27a), one movement is required—and in wh-questions a second movement step is needed, that of the wh-phrase DP. Which yellow chair has Adrian always liked? (28) CP 3 DP C’ which yellow chair 3 C FinP has 3 DP Fin' Adrian 3 VP has 3 AdvP V' always 3 V liked which yellow chair  Embedded vs. Non-Embedded So, a filled C can derive a yes/no question from a declarative: (29) a. Adrian has always liked the yellow chair. b. Has Adrian __ always liked the yellow chair? But whatever sits in (un)expressed C, there can be only one C: [C: Non-emb., Q] vs. [C: Emb., Q] for (non-)embedded questions or [C: Non-emb., Decl.] vs. [C: Emb., Decl.] for declaratives. (30) a. *I wonder has Adrian always liked this yellow chair? b. *I wonder which yellow chair has Adrian always liked? (NB: We won’t go through the entire paradigm here.) Seem- vs. Hope-Constructions Again (31) a. John seems to always win the race. b. John hopes to always win the race. (32) a. It seems that John always wins the race. b. *It hopes that John always wins the race. (33) THEME (full clause) John seems [ to win the race ]. AGENT THEME (34) AGENT THEME (full clause) John hopes [ PRO to win the race ]. AGENT THEME  John seems to win the race. (35) FinP 3 DP Fin' John 3 Fin VP –s 3 V FinP seem 3 Fin VP to 3 V DP win the race  John hopes to win the race. (36) FinP 3 DP Fin' John 3 Fin VP –s 3 V FinP hope 3 DP Fin' PRO 3 Fin VP to 3 V DP win the race  Another Example of Movement But how does John get its θ-role in the seem-construction? (37) Mary said that John kissed Bill. ≠ “John said that Mary kissed Bill.” So (see also which yellow chair from above, for example): (38) All θ-roles must be assigned clause-internally. Hence, John gets its θ-role in the embedded clause and moves—we say, the DP John raises to the matrix clause.  John seems to win the race. (39) FinP 3 DP Fin' John 3 Fin VP –s 3 V FinP seem 3 Fin' John 3 Fin VP to 3 V DP win the race  Another Example of Movement Hence, John gets its θ-role in the embedded clause and moves —we say, John raises to the matrix (= raising construction). (40) a. John seems [ to always win the race ]. b. John hopes [ PRO to always win the race ]. In (40a) John raises from embedded to matrix subject position and in (40b) matrix subject John controls the embedded PRO. Thus, a seem-construction involves subject-to-subject raising and is now called subject-raising construction, whereas the hope-construction involves subject control and is thus known as a control construction.  Taking Stock We encountered a problem earlier trying to unify all syntactic dependencies (see chapter 5 on Case Theory): “If we adopt the condition on uninterpretable features requiring that all uninterpretable features must be c- commanded by their matching interpretable counterparts for feature checking to take place, nominative case agreement could not be accounted for.” (41) An uninterpretable feature [uF] must be c-commanded by a matching interpretable feature [F] in the same clause; otherwise the sentence is ungrammatical.  Mary loves herself. (42) FinP wo DP Fin' wo Fin VP wo V DP [D][N] [Fin] [V] [D][N] [φ: 3, SG] [uφ: 3, SG] [uV] [uFin] [uφ: 3, SG] Mary –s love herself  Mary loves herself. (43) FinP wo DP Fin' ei Fin VP ei V' ei V DP [D][N] [Fin] [D][N] [V] [D][N] [φ: 3, SG] [uφ: 3, SG] [φ: 3, SG] [uV] [uFin] [uFin] [uφ: 3, SG] Mary –s love herself  VP-Internal Subject Hypothesis Thus, if we assume the VP-Internal Subject Hypothesis (VISH), then the interpretable [Fin] on Fin does indeed c-command the uninterpretable [uFin] on the subject DP—when it’s in SpecVP. (44) There is a man walking in the street. What’s the subject here? There does not refer to anything or anyone—but a man does. We can express (44) differently: (45) A man is walking in the street. So, a man is the logical (or associated) subject in both sentences, while there is called an expletive subject. VISH Argument 1: Expletive Subjects (46) FinP ei DP Fin' there ei Fin VP is ei DP V' a man ei V PP walking in the street VISH Argument 2: Floating Quantifiers (47) a. All the teachers are dancing on the table. b. The teachers are all dancing on the table. c. *The teachers are dancing all on the table. (48) a. Both Harold and Maude are sitting in the garden. b. Harold and Maude are both sitting in the garden. c. *Harold and Maude are sitting both in the garden. (49) a. ?Wildly the teachers are dancing on the table. b. The teachers are wildly dancing on the table. c. The teachers are dancing wildly on the table. (50) a. Who was dancing on the table? All the teachers. b. Who was sitting in the garden? Both Harold and Maude. c. Who was dancing on the table? *Wildly the teachers.  Analysis: Floating Quantifiers (51) a. [VP all the teachers dancing] b. [VP both Harold and Maude sitting] (52) a. [Fin' are [VP all the teachers dancing]] b. [Fin' are [VP both Harold and Maude sitting]] (53) a. [FinP all the teachers [Fin' are [VP dancing]]] b. [FinP the teachers [Fin' are [VP all dancing]]] (54) a. [FinP both Harold and Maude [Fin' are [VP sitting]]] b. [FinP Harold and Maude [Fin' are [VP both sitting]]]  Two Possible Trees (55) QP (56) QP 3 3 Q DP Q DP all/both 3 all/both 3 D NP D NP the teachers the teachers Option 1: Move the full QP Option 2: Move only the DP All the teachers are dancing. The teachers are all dancing. Floating Quantifiers & Subject-Raising A clear prediction of our analysis of subject-raising (i.e. moving a subject from an embedded non-finite FinP subject position to a matrix finite FinP subject position) and the VISH analysis: If in subject-raising constructions subjects start out in the embedded clause and are remerged in the higher clause, a floating quantifier should be possible in any earlier position of the moved subject, and nowhere else. (57) a. [FinP {All} the teachers [VP {all} seem [FinP {all} to [VP {all} dance {*all} on the table]]]] b. [FinP {Both} Harold and Maude [VP {both} seem [FinP {both} to [VP {both} sit {*both} in the garden]]]]  VISH Argument 3: Idioms (58) a. John / The baker / The queen / Three of his uncles [ kicked the bucket ]. b. John / The baker / The queen / Three of his uncles [ hit the nail on the head ]. c. John / The baker / The queen / Three of his uncles [ had cold feet ]. (59) NP/DP + [VP kicked the bucket] → ‘NP/DP died.’ But there are no idioms with only subject + transitive verb: (60) [The shadow hit] NP/DP. → ‘NP/DP died.’  More on Idioms (61) THE IDIOM GENERALIZATION Only fixed constituents can receive an idiomatic interpretation. (62) a. The shit hit the fan. (= ‘Trouble started.’) b. All hell broke loose. (= ‘Everything went wrong.’) (63) a. The shit has hit the fan. b. The shit will hit the fan. If the subject originates in VP, it can qualify as a ‘fixed c. The shit must hit the fan. constituent’ (idiom here: VP) and the structure can then be (64) a. All hell has broken loose. modified further (e.g. add Fin b. All hell will break loose. and then move the subject). c. All hell must break loose.  Analysis: Idioms (65) a. Curiosity killed/*has killed/*will kill/*must kill the cat. (= ‘Being too curious can be dangerous.’) b. Elvis has left/*will leave/*is leaving the building. (= ‘It’s all over now.’) These are sentential idioms, that is, the ‘fixed constituent’ that receives the idiomatic interpretation is the full clause (here FinP) —so, if the entire FinP (= [FinP Elvis has [VP left the building]]) is the idiom, then no feature can be changed (Fin: has, *à will). (Here it doesn’t matter whether Elvis has moved from VP, what matters is that the idiomatic interpretation is assigned to the entire FinP-structure.)  A Conceptual Argument for VISH (66) FinP ei DP Fin' John ei Fin VP The subject checks nominative case and –s ei agreement in FinP. V' John ei Now the VP is well formed and Fin can V DP be merged with it. love Mary All θ-roles can be assigned within VP.  Summary Certain constituents appear in a syntactic position in which they could not have originated: They must have moved. Movement can be analysed as Remerge (merge again). The idea that constituents can move solves a range of issues: keep case and θ-relations very close or local straightforward analysis for raising constructions (θ-role) fully unified account of feature checking (nominative and accusative case, subject–verb agreement, binding) There are also some open questions, such as what constrains the different types of movements. An answer will come next.