Presentation is loading. Please wait.

Presentation is loading. Please wait.

Episode 8a. Passives and remaining issues 6.4-6.7 CAS LX 522 Syntax I.

Similar presentations


Presentation on theme: "Episode 8a. Passives and remaining issues 6.4-6.7 CAS LX 522 Syntax I."— Presentation transcript:

1 Episode 8a. Passives and remaining issues 6.4-6.7 CAS LX 522 Syntax I

2 Some words from the blog Several issues arose in the course of HW6 that may are worth highlighting here. Several issues arose in the course of HW6 that may are worth highlighting here. One concerns the case of the object of a preposition: Computers break near me. One concerns the case of the object of a preposition: Computers break near me. Now that we’ve incorporated case into our system, we’re stuck with it. Noun phrases come with case. Computers has case (nominative) and me has case (accusative). Now that we’ve incorporated case into our system, we’re stuck with it. Noun phrases come with case. Computers has case (nominative) and me has case (accusative). The question is: How is the case of me checked? The question is: How is the case of me checked?

3 Computers break near me Computers break is unaccusative; there’s no agent, and computers is the Theme/Patient, it is the affected object. Computers break is unaccusative; there’s no agent, and computers is the Theme/Patient, it is the affected object. Thus, we have in our numeration: Thus, we have in our numeration: break [V, uN*] break [V, uN*] v unaccusative [v, uInfl:, uV*] v unaccusative [v, uInfl:, uV*] computers [N,  :3pl, case] computers [N,  :3pl, case] T [T, u  :, pres, uN*] T [T, u  :, pres, uN*] As well as near and me, which we’ll get to in a moment. As well as near and me, which we’ll get to in a moment.

4 Computer break First, let’s just do computers break. First, let’s just do computers break. We start by merging break and computers. We start by merging break and computers. NP computers [N,  :3pl, case] VP V break [V]

5 Computers break v [v, uInfl:, uV*] v [v, uInfl:, uV*] We Merge v with VP (HoP). We Merge v with VP (HoP). NP computers [N,  :3pl, case] VP V break [V] vPvPvPvP v [v, uV* uInfl:]

6 Computers break The V moves up to adjoin to v to check the [uV*] feature of v. The V moves up to adjoin to v to check the [uV*] feature of v. NP computers [N,  :3pl, case] VP vPvPvPvP v v [v, uV*, uInfl:] V break [V]

7 Computers break The T is Merged with vP (HoP). The T is Merged with vP (HoP). T has the features: T has the features: [T, pres, u  :, uN*, nom]. The [nom] feature of T can now match the [case] feature of computers. NP computers [N,  :3pl, case] VP vPvPvPvP v v [v, uV*, uInfl:] V break [V] T T [T, tense:pres, u  :, uN*, nom]

8 Computers break The [nom] feature of T matches, values, and checks the [case] feature of computers, checking itself in the process. The [u  :] feature of T can also match the [  :pl] feature of computers. NP computers [N,  :3pl, nom] VP vPvPvPvP v v [v, uV*, uInfl:] V break [V] T T [T, tense:pres, u  :, uN*, nom]

9 Computers break The [  :3pl] feature of computers matches, values, and checks the [u  :] feature of T. The [tense:pres] feature of T matches the [uInfl:] feature of v, which will be valued by both the tense and  -features of T. It’s [tense:pres] that matches the [uInfl:] feature, but the  - features “come along” when the [uInfl:] feature is valued. NP computers [N,  :3pl, nom] VP vPvPvPvP v v [v, uV*, uInfl:] V break [V] T T [T, tense:pres, u  :3pl, uN*, nom]

10 Computers break The [uN*] feature of T matches the [N] feature of computers. This is not sufficient to check the [uN*] feature because they are not local, so computers is moved up to SpecTP. NP computers [N,  :3pl, nom] VP vPvPvPvP v v [v, uV*, uInfl:pres3pl] V break [V] T T [T, tense:pres, u  :3pl, uN*, nom]

11 Computers break Once the [N] feature of computers is a sister to the T that has the [uN*] feature (the feature projects from T to T—it’s the same feature), the [uN*] feature is checked. VP vPvPvPvP v v [v, uV*, uInfl:pres3pl] V break [V] T T [T, tense:pres, u  :3pl, uN*, nom] NP computers [N,  :3pl, nom] TP

12 Computers break near me Now, let’s consider Computers break near me. Me is clearly accusative. There’s nothing here that can value a case feature as accusative. All we’re adding to this is me (which has accusative case) and the P near. VP vPvPvPvP v v [v, uV*, uInfl:pres3pl] V break [V] T T [T, tense:pres, u  :3pl, uN*, nom] NP computers [N,  :3pl, nom] TP

13 Computers break near me Conclusion: It must be near that is responsible for the accusative case on me. VP vPvPvPvP v v [v, uV*, uInfl:pres3pl] V break [V] T T [T, tense:pres, u  :3pl, uN*, nom] NP computers [N,  :3pl, nom] TP P near [P, uN*, acc] NP pronoun [N,  :1sg, case]

14 Computers break near me Merge near and me (1sg pronoun). The [N] feature of me checks the [uN*] feature of near. The [acc] feature of near values and checks the [case] feature of me (checking itself in the process). VP vPvPvPvP v v [v, uV*, uInfl:pres3pl] V break [V] T T [T, tense:pres, u  :3pl, uN*, nom] NP computers [N,  :3pl, nom] TP P near [P, uN*, acc] NP pronoun [N,  :1sg, acc] PP

15 Near me computers break The last step: Adjoin the PP to the TP. To the TP? Near me can appear on either side of TP, not vP. VP vPvPvPvP v v [v, uV*, uInfl:pres3pl] V break [V] T T [T, tense:pres, u  :3pl, uN*, nom] NP computers [N,  :3pl, nom] TP P near [P, uN*, acc] NP pronoun [N,  :1sg, acc] PP TP

16 P checks accusative So, in general: A preposition P… So, in general: A preposition P… Has a [P] category feature Has a [P] category feature Has a [uN*] feature, motivating a Merge with its object. Has a [uN*] feature, motivating a Merge with its object. Has an [acc] feature, valuing and checking the [case] feature of its object. Has an [acc] feature, valuing and checking the [case] feature of its object. T has [T], [uN*] (EPP), [u  :], [nom] T has [T], [uN*] (EPP), [u  :], [nom] v has [v], [uInfl:], [uV*], and, if v assigns a  -role, it has [uN*] and [acc]. v has [v], [uInfl:], [uV*], and, if v assigns a  -role, it has [uN*] and [acc].

17 Feature matching/valuation A modal like might has a category feature [T], but we’ll say it also has a [modal] feature. A modal like might has a category feature [T], but we’ll say it also has a [modal] feature. [modal] values a [uInfl:] feature as none. [modal] values a [uInfl:] feature as none. Pat might leave. They might leave. Pat might leave. They might leave. If there is no [modal] feature, [tense] values a [uInfl:] feature with the tense features and (valued)  -features. If there is no [modal] feature, [tense] values a [uInfl:] feature with the tense features and (valued)  -features. Pat leaves. They leave. Pat leaves. They leave. If a [uInfl:] on Aux (Perf and Prog) is valued by a tense feature, it is strong, and must be local to the valuing feature: If a [uInfl:] on Aux (Perf and Prog) is valued by a tense feature, it is strong, and must be local to the valuing feature: Pat has not left. Pat has not left. Pat might not have left. Pat might not have left.

18 Double-object constructions We’ve by now covered the sentence We’ve by now covered the sentence Pat gave books to Chris. Pat gave books to Chris. Pat, books, and Chris are all noun phrases, they all need case. Pat, books, and Chris are all noun phrases, they all need case. Pat gets (nom) case from T. Pat gets (nom) case from T. books gets (acc) case from v. books gets (acc) case from v. Chris gets (acc) case from P (to). Chris gets (acc) case from P (to). What about Pat gave Chris books? What about Pat gave Chris books?

19 give possessive vs. give locative Notice that the two constructions mean slightly different things: Notice that the two constructions mean slightly different things: Pat gave Chris books. Pat gave Chris books. Pat caused Chris to have books. Pat caused Chris to have books. *Pat sent Chicago letters. *Pat sent Chicago letters. #Pat taught preschoolers Calculus. #Pat taught preschoolers Calculus. Pat gave books to Chris Pat gave books to Chris Pat caused books (to go/be located) to Chris. Pat caused books (to go/be located) to Chris. Pat sent letters to Chicago. Pat sent letters to Chicago. Pat taught Calculus to preschoolers. Pat taught Calculus to preschoolers. So far, we’ve been dealing with give locative. So far, we’ve been dealing with give locative.

20 Chris has books If have (possessive) is somehow part of the meaning of give possessive (cause+have), let’s think a bit about have. If have (possessive) is somehow part of the meaning of give possessive (cause+have), let’s think a bit about have. In Chris has books, Chris is not an agent. Suppose instead that Chris is a theme/patient, affected by the possession. In Chris has books, Chris is not an agent. Suppose instead that Chris is a theme/patient, affected by the possession. Since there is no agent, suppose that Chris has books has the “unaccusative v”. Since there is no agent, suppose that Chris has books has the “unaccusative v”. v [v, uV*, uInfl:] v [v, uV*, uInfl:]

21 Chris has books We now have two NPs, and only one avenue for checking case (T, which will value Chris as nominative). Books needs (acc) case from somewhere. Well, what are the options? We now have two NPs, and only one avenue for checking case (T, which will value Chris as nominative). Books needs (acc) case from somewhere. Well, what are the options? NP books [N,  :3pl, case] V V have [V, uN*, uN*, …] VPVPVPVP NP Chris [N,  :3sg, case] vPvPvPvP v [v, uInfl:, uV*]

22 Chris has books It must be that books gets case from have itself. It must be that books gets case from have itself. What  -role does books have? What  -role does books have? UTAH does not yet cover this case (NP daughter of V) UTAH does not yet cover this case (NP daughter of V) NP books [N,  :3pl, case] V V have [V, uN*, uN*, acc] VPVPVPVP NP Chris [N,  :3sg, case] vPvPvPvP v [v, uInfl:, uV*]

23 Chris has books Let’s say that books has a Possessee  -role. Let’s say that books has a Possessee  -role. So we add to UTAH: So we add to UTAH: NP daughter of V: Possessee. NP daughter of V: Possessee. NP books [N,  :3pl, acc] V V have [V, uN*, uN*, acc] VPVPVPVP NP Chris [N,  :3sg, case] vPvPvPvP v [v, uInfl:, uV*]

24 Chris has books Have (possessive) is a V that takes two arguments ([uN*, uN*]) and checks accusative case ([acc]). Have (possessive) is a V that takes two arguments ([uN*, uN*]) and checks accusative case ([acc]). It’s almost like a “transitive preposition” (P HAVE ). It’s almost like a “transitive preposition” (P HAVE ). Now, suppose we add an Agent. Now, suppose we add an Agent. NP books [N,  :3pl, acc] V V have [V, uN*, uN*, acc] VPVPVPVP NP Chris [N,  :3sg, case] vPvPvPvP v [v, uInfl:, uV*]

25 Pat gave Chris books That is, Pat cause Chris to have books, more or less. That is, Pat cause Chris to have books, more or less. Now using the v agent, we have [uN*, acc] features on v as well. The [acc] feature can value and check the case on Chris. Now using the v agent, we have [uN*, acc] features on v as well. The [acc] feature can value and check the case on Chris. NP books [N,  :3pl, acc] V V have [V, uN*, uN*, acc] VPVPVPVP NP Chris [N,  :3sg, case] v v [v, uInfl:, acc, uN*, uV*]

26 Pat gave Chris books Move V to v, Merge in Pat. Assume that v agent +have = give (possessive), and we have Pat gave Chris books. Move V to v, Merge in Pat. Assume that v agent +have = give (possessive), and we have Pat gave Chris books. Pat will get its [case] feature checked by T (nom), and v will have its [uInfl:] feature valued by T (past3sg). Pat will get its [case] feature checked by T (nom), and v will have its [uInfl:] feature valued by T (past3sg). NP books [N,  :3pl, acc] V VPVPVPVP NP Chris [N,  :3sg, acc] v v vPvPvPvP NP Pat [N,  :3sg, case] V give [V, uN*, uN*, acc] v [v, uInfl:, acc, uN*, uV*]

27 Adverbs Before today, we’d always drawn adjuncts as adjoined to vP. This explains why sloppily can be either to the left or to the right of vP: Before today, we’d always drawn adjuncts as adjoined to vP. This explains why sloppily can be either to the left or to the right of vP: Pat sloppily ate lunch. Pat sloppily ate lunch. Pat ate lunch sloppily. Pat ate lunch sloppily. Pat has sloppily eaten lunch. Pat has sloppily eaten lunch. Pat has eaten lunch sloppily. Pat has eaten lunch sloppily. Quickly also seems to be able to adjoin to PerfP or ProgP, at least marginally. Quickly also seems to be able to adjoin to PerfP or ProgP, at least marginally. ?Pat might sloppily have eaten lunch. ?Pat might sloppily have eaten lunch. ?Pat should sloppily be eating lunch. ?Pat should sloppily be eating lunch. But it can’t be between a subject and T: But it can’t be between a subject and T: *Pat sloppily might eat lunch. *Pat sloppily might eat lunch.

28 Manner vs. propositional adverbs Sloppily, slowly, quickly— all describe the manner in which an action takes place. These are manner adverbs. They adjoin to vP. Sloppily, slowly, quickly— all describe the manner in which an action takes place. These are manner adverbs. They adjoin to vP. There are other kinds of adverbs as well, however. One such kind are propositional adverbs: perhaps, fortunately, interestingly. These express a kind of attitude on the part of the speaker toward the content of the sentence. There are other kinds of adverbs as well, however. One such kind are propositional adverbs: perhaps, fortunately, interestingly. These express a kind of attitude on the part of the speaker toward the content of the sentence.

29 Propositional adverbs Propositional adverbs seem to adjoin to TP. Propositional adverbs seem to adjoin to TP. Fortunately, Pat ate lunch. Fortunately, Pat ate lunch. Pat ate lunch, fortunately. Pat ate lunch, fortunately. ?Pat fortunately ate lunch. ?Pat fortunately ate lunch. ?Pat might have fortunately eaten lunch. ?Pat might have fortunately eaten lunch. Temporal adverbs also seem to adjoin high. Temporal adverbs also seem to adjoin high. Today Pat ate lunch. Today Pat ate lunch. Pat ate lunch today. Pat ate lunch today. *Pat today ate lunch. *Pat today ate lunch.

30 Adverb positions Generally speaking, where an adverb attaches depends on its meaning. Generally speaking, where an adverb attaches depends on its meaning. vP for manner adverbs, TP for temporal adverbs, … vP for manner adverbs, TP for temporal adverbs, … Notice that we predict this now: Notice that we predict this now: Yesterday Pat completely finished lunch. Yesterday Pat completely finished lunch. Pat ate lunch completely yesterday. Pat ate lunch completely yesterday. *Pat ate lunch yesterday completely. *Pat ate lunch yesterday completely. Later, perhaps, we’ll consider additional complexity in adverb placement. Later, perhaps, we’ll consider additional complexity in adverb placement.

31 Passives The passive construction is one where: The passive construction is one where: The original subject disappears (or is moved to a by- phrase) The original subject disappears (or is moved to a by- phrase) The original object becomes the subject. The original object becomes the subject. The verb appears as be+passive participle. The verb appears as be+passive participle. The passive participle in English sounds just like the perfective participle. The passive participle in English sounds just like the perfective participle. Pat took pretzels.active Pat took pretzels.active Pretzels were taken (by Pat).passive Pretzels were taken (by Pat).passive

32 Passives Pat stole books. Pat stole books. Books were stolen (by Pat). Books were stolen (by Pat). In both cases, books is getting the Theme/Patient  -role. By UTAH, it must be originally Merged as NP daughter of VP, in both the active and the passive. In both cases, books is getting the Theme/Patient  -role. By UTAH, it must be originally Merged as NP daughter of VP, in both the active and the passive. In fact, the passive is a lot like the unaccusative. An “underlying object” becomes the subject. In fact, the passive is a lot like the unaccusative. An “underlying object” becomes the subject.

33 Passives We don’t need anything really new for the passive, just the passive auxiliary Pass. We don’t need anything really new for the passive, just the passive auxiliary Pass. be [Pass, uInfl:] selects a v unaccusative. be [Pass, uInfl:] selects a v unaccusative. By selecting for v unaccusative, the passive auxiliary “removes” an Agent. By selecting for v unaccusative, the passive auxiliary “removes” an Agent. Not allowed for intransitives, an open mystery. Not allowed for intransitives, an open mystery. The passive auxiliary works like other auxiliaries: Pass can value a lower [uInfl:] feature, if Pass’ own [uInfl:] feature is valued by a [tense] feature, it is strong. The passive auxiliary works like other auxiliaries: Pass can value a lower [uInfl:] feature, if Pass’ own [uInfl:] feature is valued by a [tense] feature, it is strong. Lunch was not eaten. Lunch was not eaten. Pass is the last auxiliary in the HoP: Pass is the last auxiliary in the HoP: Lunch may not have been being eaten. Lunch may not have been being eaten. T > (Neg) > (Perf) > (Prog) > (Pass) > v > V T > (Neg) > (Perf) > (Prog) > (Pass) > v > V

34 Lunch was eaten For Lunch was eaten, we Merge eat and lunch to build the VP, then Merge an unaccusative v… For Lunch was eaten, we Merge eat and lunch to build the VP, then Merge an unaccusative v… NP lunch [N,  :3sg, case] VP vPvPvPvP v [v, uV*, uInfl:] V eat [V, uN*]

35 Lunch was eaten The V moves up to adjoin to v to check the [uV*] feature of v. The V moves up to adjoin to v to check the [uV*] feature of v. The Pass auxiliary is Merged (HoP). The Pass auxiliary is Merged (HoP). [Pass] matches, values, checks [uInfl:] on v. [Pass] matches, values, checks [uInfl:] on v. VP vPvPvPvP v V eat [V, uN*] PassP Pass be [Pass, uInfl:] v [v, uV*, uInfl:] NP lunch [N,  :3sg, case]

36 Lunch was eaten T is Merged (HoP). T is Merged (HoP). [nom] on T matches, values, checks [case] on lunch. [nom] on T matches, values, checks [case] on lunch. [[u [  :3sg] on lunch matches, values, checks [u  :] on T. [past] on T matches, values [uInfl:] on Pass. VP vPvPvPvP v V eat [V, uN*] PassP Pass be [Pass, uInfl:] v [v, uV*, uInfl:pass] NP lunch [N,  :3sg, case] T T [T, tense:past, u  :, uN*, nom]

37 Lunch was eaten Pass moves to T (checks [uInfl:past*] on Pass). Pass moves to T (checks [uInfl:past*] on Pass). VP vPvPvPvP v V eat [V, uN*] PassP v [v, uV*, uInfl:pass] NP lunch [N,  :3sg, nom] T T Pass be [Pass, uInfl:past*] T [T, tense:past, u  :3sg, uN*, nom]

38 Lunch was eaten Lunch moves to SpecTP (checks [uN*] on T). Lunch moves to SpecTP (checks [uN*] on T). VP vPvPvPvP v V eat [V, uN*] PassP v [v, uV*, uInfl:pass] T T Pass be [Pass, uInfl:past*] T [T, tense:past, u  :3sg, uN*, nom] NP lunch [N,  :3sg, nom] TP

39 Ditransitive passives Consider again Pat gave Chris books. Consider again Pat gave Chris books. Chris was given books. Chris was given books. *Books were given Chris. *Books were given Chris. (Exercise 6.5): Pat gave books to Chris. (Exercise 6.5): Pat gave books to Chris. Books were given to Chris. Books were given to Chris. *Chris was given books to. *Chris was given books to.

40                       


Download ppt "Episode 8a. Passives and remaining issues 6.4-6.7 CAS LX 522 Syntax I."

Similar presentations


Ads by Google