A Tale of Two Peripheries
A Tale of Two Peripheries
Evidence from Chinese Adverbials, Light Verbs, Applicatives and Object Fronting
Abstract and Keywords
Concerning the peripheral area around the edge of vP, Chinese provides a perfect testing ground for investigation in that linguistic expressions are very often constructed analytically and their “loosened” parts may spread over functional projections in a minimalist manner. As a result, Chinese proves to be an ideal testing ground for the cartographic approach due to its robust analyticity, as functional elements spread over the sentential projections in an orderly manner, revealing some “hidden” cartography obscured by morpho-syntactic displacement in other languages. In particular, there is the contrast of the vP periphery vs. the left periphery, in the inner-outer dichotomy of wh-adverbials, as well as that of reflexive adverbials. The same pattern is found again in light verbs, applicatives, and object fronting with striking conceptual connections, and so our finding hopefully will bring us closer to a full understanding of the nature of syntax-semantics mapping under the cartographic approach.
This paper sets out to join a recent inquiry into the peripheral area around the edge of vP along the line of Belletti (2004, 2005).* It is suggested that, in comparison with other languages more agglutinating in nature, Chinese provides an ideal testing ground for this venture in that its linguistic expressions are often constructed analytically (cf. Huang 2004), and their “loosened” parts may spread over functional projections in a minimalist manner (i.e., Merge instead of Move, very much in line with the notion of Lexical Courtesy).1 We show that their distribution concentrates on two specific areas—that is, the vP periphery vs. the left periphery in the sense of Rizzi (1997, 2001), which in turn are associated with comitativity and causality respectively through the syntax-semantics interface.
1. Inner Adverbials Vs. Outer Adverbials
1.1. How-Why Alternations in a Wider Context
Since Collins’s (1991) seminal work on the asymmetry between why and how come in English, there is a growing interest in studying the distribution and interpretation of wh-adverbials across languages. What we have (p.2) observed in Chinese is another line of division in terms of their syntactic distribution (i.e., inner wh-adverbials vs. outer wh-adverbials, roughly corresponding to the classic distinction between VP-adverbs and sentential adverbs (cf. Tsai 2008a)). Here we would like to argue that the division between the vP periphery and the left periphery is most easily seen in this inner-outer dichotomy: When zenme appears before a future modal, it gets interpreted as causal, as in (1a). In contrast, it receives an instrumental reading in a postmodal position, as in (1b):2
There are two forms of Chinese why as well: as shown in (2a), the outer why patterns with an adverb, can never appear after the future modal, and the interpretation is strictly a reason question. On the other hand, the inner why behaves more like a PP, and typically appears after the future modal, resulting in a purpose question, as illustrated by (2b):
In fact, as noted by Stepanov and Tsai (2008), there are also two types of why-questions in Russian: počemu, a reason why, must appear above negation, as in (3a), whereas its purpose counterpart, začem, cannot, as in (3b):
On the other hand, while it is impossible to put počemu ‘whyR’ in the scope of negation by forming an indefinite wh, the same practice is perfectly fine with začem ‘whyP’, as evidenced by the contrast of (4a,b):
This is reminiscent of the negative island effects on the Chinese outer why, as illustrated by the contrast in (5a,b). On the other hand, the inner why, by virtue of assuming a PP form, can indeed be rescued by an indefinite construal under negation (cf. Tsai 1994a,b; Stepanov and Tsai 2008), as in (6):
Furthermore, a recent fine-grained study of Japanese wh-adverbials also reveals that nande can be interpreted as either reason or instrumental, depending on its syntactic position as well as its categorial status (cf. Huang, 1982; Fujii et al. 2014), as exemplified by the ambiguity displayed by (7a,b):
The above ambiguity thus provides another fine example of how-why alternations in Tsai’s (2008a) sense, which can be dissolved by replacing nande with its uncontracted counterpart nani-de ‘what-with’, presumably a PP in syntactic terms, as in (8a), or by adding a de-D-linking marker mata on earth’, as in (8b):
Interestingly enough, only instrumental nande (call it nandeI) and its PP form can appear below certain sentential adverbials such as kanarazu ‘necessarily’ or tokidoki ‘sometimes’, as shown by (9a). In contrast, the corresponding reason question is blocked in the same configuration, as shown by (9b):
On the other hand, when nande appears above those adverbials with mata, the reading is distinctively reason (call it nandeR), and the instrumental question is blocked instead, as evidenced by (10a,b):
On the historical front, we can also find evidence for the inner-outer dichotomy in Tsou. First consider (11a): Mainenu ‘how’ starts as a verbal predicate linked to the main predicate by the conjunction ho, although in the latter stage of development ho has evolved into a complementizer introducing a control complement (cf. Tsai and Chang 2003; Tsai 2007a):
(p.5) Mainenu can also be construed as a matrix predicate taking a finite clause headed by ci ‘such that’ as its complement, as evidenced by (11b). As a result, a causal question is formed. The further development has seen a collapsing of the bi-clausal structure, with mainenu and ci contracted into a sentential adverbial mainci ‘why’, as shown in (11c). We therefore have another case of the inner-outer distinction cutting across how-questions and why-questions.
1.2. Inner Selfhood vs. Outer Selfhood
It is worthwhile to note that there is a strong correspondence between outer adverbials and eventuality construals, in particular with regard to event-level causation. Namely, when we ask a why-question, we are essentially presupposing the existence of a cause or reason for the current event, and asking the addressee to point out what it is (cf. Karttunen 1977; Bromberger 1992). At this level, outer wh-adverbials always scope over the subject. As a result, they do not observe the agentivity restriction like their inner counterparts such as the instrumental how and purpose why. The latter is strictly agent-oriented, expressing a kind of proto-comitativity associated with various usages of the English with (cf. Tsai 2008a). As evidenced by the absence of instrumental readings in the unaccusative construction (12) and the locative-existential construction (13), non-agent subjects are only compatible with outer wh-adverbials:
The same asymmetry obtains for the pair of Chinese why’s as well, as illustrated below:
As a matter of fact, a parallel to the inner-outer dichotomy can be found in Chinese reflexive adverbials as well: In (16a), the premodal ziji is interpreted as anti-causal (e.g., without others’ coercion/persuasion), while in (16b), its postmodal counterpart is construed as anti-comitative (i.e., without others’ company/help):
Again we found subject agentivity effects on inner reflexive adverbial construals: As shown by unaccusative sentences like (17) and locative-existential sentences like (18), only the anti-causal readings survive, whereas the anti-comitative readings are blocked:
These parallels lead us to the conclusion that outer wh-adverbials pattern with outer reflexive adverbials in expressing (anti-)causality, while inner wh-adverbials pattern with inner reflexive adverbials in expressing (anti-)comitativity (i.e., a with-relation in the sense of Parsons (1995)). Based on Rizzi’s split-CP analysis, we propose to associate the property of outer adverbials with the left periphery, while attributing the property of inner adverbials to the vP periphery in the spirit of Belletti (2004, 2005). The two peripheries roughly correspond to the edges of two strong phases (i.e., CP and vP, along the line of Chomsky (2000, 2001)). We may thus visualize this inner-outer dichotomy in the following diagram: (p.7)
Building upon Ramchand’s (2008) distinction between the causation projection and the process projection within vP, we implement the insight in terms of a causative phrase (CauP) in the left periphery to encode the eventuality causation mentioned above (also cf. Shlonsky and Soare 2011). The Spec of CauP thus readily provides a locus for outer reflexive and wh-adverbials, except perhaps for the causal zenme, which may well serve as the head of the interrogative phrase (IntP), patterning with perché in Italian and how come in English (see Tsai 2008a for detailed discussion). In contrast, inner reflexive and inner wh-adverbials are agent-oriented in nature. They serve as a vP-adjunct in syntax, while behaving like a focus operator in terms of semantics (cf. Tsai, forthcoming).
Argumental reflexives, on the other hand, either stay within VP as anaphors, typically subject-oriented, or situate high above as logophors in the so-called source phrase (SrcP), typically speaker-oriented (cf. Huang and Liu 2000).
All in all, the syntax-semantics correspondences of Chinese inner/outer adverbials thus provide a strong argument for the proposed two-periphery analysis. As a point of interest, it is worthwhile to note that a doubled strong pronoun in a vP-internal focus position in Italian can produce an adverbial-like “in person” reading (cf. Belletti 2005), as shown below:
(p.8) In a sense, this so-called “strong pronoun doubling” provides cross-linguistic support for our proposal to tie inner Self with the vP periphery. We will return to the connection between focus construals and the vP periphery below.
2. Inner Light Verb Vs. Outer Light Verb
Another case in favor of our claim has to do with the fact that there are also two classes of Chinese light verbs which, rather surprisingly, display the now familiar inner-outer asymmetry (cf. Tsai 2007b). The first class includes eventuality predicates such as CAUSE, as first discussed in the seminal work of Huang (1994, 1997). It is argued that the syntax-semantics mismatch of (21) can be resolved by analyzing its underlying structure parallel to (22), where a lexical causative verb rang is present, and the resultative verbal complex chi-de is returned to its normal position:4
Specifically, we entertain the possibility that in (21) there is an implicit causative predicate CAUSE, situated in exactly the same position as its lexical counterpart in (22), as illustrated in (23a):
This implicit predicate takes a Cause as its external argument, and an effect event as its complement. (21) is thus derived by raising the resultative verb to CAUSE, which is phonologically defective and requires verb-raising to fulfill the PF requirement.5
The second class of light verbs involves various construals associated with dynamic properties and contrastive foci (e.g., instrumental, locative, and benefactive— see Lin 2001, and Feng 2003, 2005, among others). Take an instrumental construal like (24) for example: Here we have another case (p.9) of a syntax-semantics mismatch, which can once again be paraphrased as a sentence with a lexical light verb yong ‘use’, as in (25):
On the assumption that (24) has an implicit light verb USE corresponding to yong in terms of both syntax and semantics, as in (26a), we may derive the sentence by raising the verb qie ‘cut’ to USE, as sketched in (26b):
In terms of syntax, the first class of light verbs is most likely to head a functional projection on the CP layer, which we may call “outer light verbs.” The second class, in contrast, presumably heads the vP projection. Hence the notion of “inner light verbs.” To really tease them apart, we may employ two syntactic criteria: The first one involves Chinese verb-copying, which is essentially a vP-internal phenomenon. Just as our theory predicts, raising- to-CAUSE, where the landing site is beyond the vP periphery, is not subject to verb-copying, as evidenced by (27a). In contrast, raising-to-USE is fully compatible with verb-copying, since it only involves an inner light verb position within the vP periphery, as evidenced by (27b):
The second test has to do with light verbs’ ability to take various types of postverbal complements; as we have seen in (23), an outer light verb like CAUSE typically involves a resultative construal, presumably encoded by a result projection in the sense of Ramchand (2008). In addition, there is (p.10) another type of implicit outer light verb which may be dubbed as COST (Gu 2002). Take (28) for instance: here the main verb kan-le raises to COST in the now familiar fashion, which typically introduces a duration expression such as yi xiawu “one afternoon” in the postverbal position:6
Inner light verbs, however, fail the second test. On the one hand, raising-to-USE is not allowed in presence of a resultative complement such as hen lei ‘very tired’ in (29a), whereas the resultative construal is perfectly compatible with the corresponding lexical light verb, as evidenced by (29b):7
On the other hand, the same restriction is duly observed by the duration construal of (30a), where the main verb qie is prevented from adjoining to the implicit light verb USE. In contrast, (30b) shows that raising-to-USE does not pose any problem for its lexical counterpart yong, just as we might expect from the above pattern of contrasts:
In terms of semantics, it is not difficult to see the interpretive similarity between outer adverbials and outer light verbs (i.e., they both involve causality), as well as that between inner adverbials and inner light verbs (i.e., they both involve a with-relation between an Instrument/Comitant and its corresponding event (Parsons 1995; Tsai forthcoming)). All the above observations clearly point to the conclusion that it is imperative to separate outer and inner (p.11) light verbs in terms of their syntactic topography under the cartographic approach (see Rizzi 1997 and Cinque 1999, among many others), as sketched below (RC: resultative complement):
From the cross-linguistic perspective, it has been reported in the literature that there are two tiers of causative projections, where the inner-outer distinction is based on various morpho-syntactic criteria (see, for instance, Svenonius 2005). Furthermore, our theory predicts that inner and outer light verbs should be able to appear together in the sentential projection in question. This is indeed borne out. As shown by (32), we may combine the causative and instrumental usages in one sentence, where the external argument na-ba dao ‘that knife’ appears to play both roles as Cause and Instrument:
Here we propose that na-ba dao is actually thematically related to CAUSE in the left periphery, and gets identified with the instrument argument in Spec-VP through null operator movement in the classic sense (cf. Chomsky 1977, 1986). On the other hand, the main verb qie-de first combines with the inner light verb USE through head movement. The resulting verbal complex (p.12) then adjoins to the outer light verb CAUSE.8 The whole derivation can be visualized in the following tree diagram:
3. Inner Affective Vs. Outer Affective
It is instructive to note that there is also an inner-outer distinction between Mandarin affective construals (cf. Tsai 2012). (34a) presents a typical case of outer affective construals, which are marked by an applicative head gei, and arguably located in the left periphery:
The inner affective of (34b), on the other hand, appears in the form of a pseudo-double object construction (pseudo-DOC), in that there is no directional possession between the Affectee wo and the Theme san-ping jiu.
In terms of semantics, there is a truth-conditional distinction between the two types of affective construals: In the scenario that a doctor asked a patient not to drink wine at home, but the patient did not follow the instruction, the doctor may utter (34a), but not (34b). As a matter of act, Mandarin inner affective construals require static possession between the two objects (i.e., the Affectee wo ‘me’ and the Theme san-ping jiu ‘three bottles of wine’) in direct contrast with the ‘ “to-the-possession-of” interpretation typically associated with English low applicatives such as Mary baked Bill a cake.
(p.13) In terms of pragmatics, Mandarin outer affectives are strictly speaker-oriented: the ungrammaticality of (35a) indicates that it can only be a first-person singular pronoun, hence speaker-oriented. By contrast, the same restriction is not observed for inner affectives in (34b), as evidenced by the well-formedness of (35b):
Furthermore, while it is very awkward to employ an outer affective in a declarative sentence such as (36a), its inner affective counterpart in pseudo-DOCs is quite compatible with the same environment, as evidenced by the grammaticality of (36b):
From a cross-linguistic point of view, the so-called ethical datives in modern Greek (as well as Romance languages in general) present an interesting comparison with Mandarin outer affectives (cf. Perlmutter 1971; Jaeggli 1982; Cuervo 2003; Michelioudakis and Sitaridou 2008, among others). As shown in (37), there is also a speaker/addressee-oriented restriction on the high applicative usage of these dative clitics, which typically express some sort of affectedness between an individual and an event:
Furthermore, modern Greek ethical datives also require licensing from imperative, optative, subjunctive, or negative moods, as shown by the contrast between (38a) and (38b):
This is very much in line with the outer affective construals in Mandarin, which are disallowed in declarative sentences, as we have already seen in (36a).
Furthermore, Kikushima (2013) has reported that in Japanese, the transference verb kureru can appear in a similar environment to that of the CP-related high applicative in Chinese. As shown below, the applicative morpheme is attached to the main verb okuru send’, and roughly means ‘give (me)’. The resulting interpretation can be either benefactive, as in (39a), or adversative, as in (39b):
To sharpen the above intuition further, we may spell out the first-person dative Affectee, and add an evaluative adverb odoroitakotoni ‘surprisingly’ which carries the exclamative force. As shown by (40b), the reading is very much in line with the outer affective in Mandarin:
In terms of structural distribution, Kikushima (2013) also points out that kureru always scopes over other transference verbs such as yaru ‘give’ and morau ‘receive’, as illustrated by (41) and (42) respectively:
Given the mirror principle effects, this indicates that kureru locates much higher than other verbal elements, presumably within the same range of Mandarin outer affectives.
It is therefore reasonable to suggest that outer affectives differ from their inner counterparts in situating on the CP layer rather than the VP layer. We propose to take the affective marker gei in (36a) to head a high applicative project in the left periphery, which hosts the first-person Affectee in question. The high applicative head then raises to the evaluative phrase (EvaP), and the subject DP topicalizes to the sentence-initial position. We may visualize the relevant derivations in the following tree diagram:
Inner affectives, on the other hand, involve an implicit applicative head Aff. In a pseudo-DOC such as (36b), after V adjoins to Aff, the [V-Aff] complex mediates a static possessive relation between the Affectee and the Theme. We thus place the Affectee argument in the Spec of a “middle” applicative projection in-between vP and VP, as shown below:
Our last case has to do with the curious specificity effects on object fronting in Chinese, which provide further evidence for our distinction between the two peripheries: It is possible to use the future adverbial mingtian ‘tomorrow’ as a delimitator to distinguish two preverbal focus positions in Mandarin (cf. Tsai 2008b). As shown in the following contrastive focus construction, if an object is preposed in between mingtian and the main verb, it is interpreted as either definite, as in (46a), or nonspecific, as in (46b):
Given that the future adverbial is associated with the IP layer, it is reasonable to assume that the object in question is situated in the peripheral area of vP, presumably at the Spec of an inner focus phrase (FocPinner). On the other hand, if the object raises to a place higher than mingtian, the nonspecific reading is blocked, as in (47b), and only the definite reading survives, as evidenced by (47a):
This is a good indication that the fronted object in question has acquired topicality of some sort in front of the future adverbial, presumably at the Spec of an outer focus phrase in the left periphery.
Our observation is strengthened by the fact that a numeral NP can undergo object fronting only when it is specific or definite. This restriction is illustrated by the following contrast: In (48a), the numeral NP liang-ben shu ‘two books’ is raised to a position below the future adverbial, and receives a nonspecific reading. By contrast, it is prevented from landing above the future adverbial, as in (48b), since unlike bare NPs, it is impossible for a numeral NP to receive an interpretation other than nonspecific:
On the other hand, one may rescue the outer focus construal by adding a demonstrative such as zhe ‘this’, as in (49a), or an existential modal such as you ‘have’. which produces some sort of partitive interpretation, as in (49b):
In addition to the semantic distinction mentioned above, there is also a syntactic test for our distinction of Chinese foci: It has been noted in the literature that only a base-generated topic allows left dislocation through the resumptive pronoun strategy (cf. Cinque 1990, among others). Chinese displays exactly the same pattern: As evidenced by the contrast between (50a,b), a typical topic can be associated with a bound pronoun in-situ, whereas a fronted object cannot:
When applying the resumptive pronoun test to our inner-outer dichotomy, it becomes clear that the strategy is available only for an object raised above the future adverbial mingtian ‘tomorrow’, as in (51a); (p.18) the same construal is blocked for object-fronting below mingtian, evidenced by (51b):
All these facts point to the conclusion that the fronted object in an outer focus position is actually a contrastive topic (also known as a focus topic), which must be definite or specific by nature. In comparison, an inner focus is found in between FinP (which hosts the future adverbial) and VP (which is headed by the main verb), presumably situated in the Spec of FocPinner in the vP periphery:
(52) Topography of inner and outer focus (irrelevant details omitted):
(p.19) In this picture, there are two ways to derive the semantic distinction of the two types of Chinese foci. One is to assume that there is an existential operator associated with the implicit irrealis mood (MoodIrr), which may license a nonspecific indefinite in the inner focus position. By contrast, an outer focus is simply too high for this construal, hence the definiteness/specificity effects in question.
The other way is to side with Tsai (2001) in claiming that the domain of existential closure (i.e., nuclear scope) is not associated with VP (cf. Diesing 1992), but defined by the notion of the syntactic predicate. In the case of (47) and (48b), the syntactic predicate is formed by raising the verb to MoodIrr at LF. Consequently, an indefinite in the outer focus position is well beyond the scope of existential closure, and must be licensed in a “marked” manner. On the other hand, its inner counterparts in (46) and (48a) are subject to the default existential quantification, and receive the nonspecific interpretations, just as predicted.
5. Ins and outs in Perspective
5.1. Adverbials vs. Focus
To complete the whole picture, it is imperative to put all classes of ‘ins’ and ‘outs’ against one another for a final showdown. The mapping result would look like the following topography:
Let’s deal with each dueling challenge one by one. First we locate the Mandarin outer how in between topics and foci, as it is higher than both inner and outer foci, as in (54a,b), but lower than a (discourse) topic, as in (54c). On the other hand, inner how appears lower than either an inner focus, as in (54d), or an outer focus, as in (54e):
Furthermore, parallel to outer how, outer self appears higher than both inner and outer foci, as in (55a,b), but below a (discourse) topic, as in (55c):
Likewise, inner self is on a par with inner how in occurring below either an inner focus, as in (56a), or an outer focus, as in (56b):
As for Mandarin high applicatives, the outer affectives stand firmly above both inner and outer foci, just as one might expect from their speaker-oriented construals in the left periphery:
As shown in the contrastive focus construals of (58a,b), it is impossible to raise the objects further to the right of the outer affective. On the other hand, it is perfectly fine to place a topic in front of the outer affective, as evidenced by (58c):
Finally, when we apply the same test to the pseudo-DOCs, it becomes clear both inner and outer foci are situated above the inner affective: As in (59a), the fronted objects zhurou ‘pork’ and niurou ‘beef’ occur higher than the affected object wo ‘me’. When raised further above the temporal adverbial (p.22) zuotian ‘yesterday’, these objects occupy the even higher outer focus position, as shown in (59b):
5.3. Adverbials vs. Light Verbs
How about the locus of light verbs? It turns out that Mandarin outer how appears above the outer light verb CAUSE (which attracts the main verb); as we can tell from the contrast between (60a,b), zenme ‘how’ here is interpreted as causal rather than instrumental, a telltale sign of outer wh-adverbials. Furthermore, it is impossible to place zenme below the outer light verb, as evidenced by (61):
By contrast, when we put zenme in front of the inner light verb USE (which again attracts the main verb), it can be interpreted as either causal or instrumental, as illustrated by the ambiguity of (62a,b):
(p.23) Likewise, outer self is found in a position higher than the outer light verb in the left periphery: As indicated by the absence of the anti-comitative interpretation in (63b), inner self is not allowed in front of CAUSE. In comparison, the anti-causal reading of (63a), though marginal for pragmatic reasons, is still intelligible:
It is also instructive to note that inner self cannot follow CAUSE either, as evidenced by (64), presumably due to the lack of a vP periphery in this particular construction:
Also, parallel to their interrogative counterparts, inner self and outer self precede the inner light verb: As illustrated below, both anti-causal and anti-comitative readings are available for the reflexive adverbial construal in front of USE:
It can then be concluded that outer adverbials can only occur above outer light verbs, while inner adverbials are situated in between inner and outer light verbs. We thus move a step closer to mapping a comprehensive topography of all the ins and outs involved.
5.4. Affectives vs. Light Verbs
Now we may proceed to the question of how Mandarin affectives fare with light verbs: It seems that outer affectives, just like outer adverbials, precede both inner and outer light verbs, as evidenced by (61a,b):
On the other hand, although they’re a bit marginal, both inner and outer light verbs can be found higher than inner affectives in pseudo-DOCs such as (62a,b):
5.5. Focus vs. Light Verbs
Now how about pitching object fronting and inner light verb construals against each other? We have no difficulty in placing an inner focus in front of an inner light verb, as seen in (63a) and (64a); the same can be said about the outer focus in (63b) and (64b) (recall that we can tell inner and outer foci apart by their positions relative to temporal adverbials such as mingtian ‘tomorrow,’ and zuotian ‘yesterday’):
The situation with outer light verbs proves a bit harder to resolve. Let’s first consider the following case, where CAUSE co-occurs with the information focus in the typical object positions:
Next we place the object in between a temporal adverbial such as mingnian ‘next.year’ and the main predicate, as in (66a). Here a contrastive focus interpretation is required to license the construal, and the outer light verb CAUSE clearly has the structural advantage over the fronted object Xila ‘Greece’:
Although it’s a bit awkward, it is also possible to raise the object further to the left of CAUSE and the temporal adverbial—that is, to the outer focus position in (66b).
Finally, we should put adverbials and affectives together, and see how they fare with each other: (67) represents a realis sentence where only the instrumental reading of inner how is disallowed:
This restriction may be due to the tense-anchoring effect of the inchoative aspect, which blocks the LF movement of inner how. Outer wh-adverbials, by contrast, always scope over TP, and the causal reading of (67a) is therefore not affected. Now when we add an outer affective in the construction, it becomes clear that outer how may c-command the Affectee, as in (68a), but not vice versa, as in (68b).
Furthermore, there is no way for an inner affective to scope over outer how, since the Affectee involved is deeply embedded in the pseudo-DOC such as (69):
The next step is to examine inner wh-adverbials against Mandarin affectives. (70) is an irrealis construction where inner how can only follow the volitional modal xiang ‘would like’. This is indicated by the fact that the causal reading of outer how is blocked here, as illustrated by the contrast of (70a,b):
On the other hand, when we substitute an inner affective for its outer counterpart, as in (72), the inner how naturally precedes the Affectee in the indirect object position, since it is situated in the vP periphery:
After exhausting all the possible combinations of the ins and outs in Mandarin, we are now in a position to map out their relative heights in both peripheries, as visualized below:
All in all, we have drawn evidence from adverbials, applicatives, light verbs, object fronting, and their interactions to verify the existence of the vP periphery, and to explore the uncharted territory in the vP periphery, which presents a sharp contrast with the left periphery in expressing (proto-)comitativity rather than causality in terms of the range of its interpretative potentials. Hopefully, this study will bring us a step closer to the full understanding of syntax-semantics mapping under the cartographic approach.
Belletti, Adriana. 2004. “Aspects of the Low IP Area.” In The Structure of CP and IP, edited by L. Rizzi, 16–51. Oxford: Oxford University Press.
Belletti, Adriana, 2005. “Extended Doubling and the VP Periphery.” Probus 17 (1): 1–35.
Bromberger, Sylvain. 1992. On What We Know We Don’t Know. Chicago: Chicago University Press.
Chomsky, Noam 1977. “On Wh-Movement.” In Formal Syntax, edited by P. Culicover, T. Wasow and A. Akmajan, [date t/k]. New York: Academic Press.
Chomsky, Noam. 1986. Knowledge of Language: Its Nature, Origin, and Use. New York: Praeger.
(p.31) Chomsky, Noam. 2000. “Minimalist Inquiries: The Framework.” In Step by Step: Essays on Minimalist Syntax in Honor of Howard Lasnik, edited by R. Martin, D. Michaels, and J. Uriagereka, 89–155. Cambridge, MA: MIT Press.
Chomsky, Noam. 2001. “Derivation by Phase.” In Ken Hale: a Life in Language, edited by Michael Kenstowicz, 1–52. Cambridge, MA: MIT Press.
Cinque, Guglielmo. 1990. Types of A’ Dependencies. Cambridge Mass: MIT Press.
Cinque, Guglielmo. 1999. Adverbs and Functional Heads: A Cross-Linguistic Perspective. New York: Oxford University Press.
Collins, Chris. 1991. “Why and How Come.” MIT Working Papers in Linguistics 15: 31–45. Cambridge: MIT.
Cuervo, M. 2003. Datives at Large. PhD diss., MIT.
Diesing, Molly. 1992. Indefinites. Cambridge, MA: MIT Press.
Feng, Shengli. 2003. “Light Verb Syntax in Classical Chinese.” Paper presented at the Conference on Research and Pedagogy in Classical Chinese and Chinese Language History. Columbia University, New York, March 2003.
Feng, Shengli. 2005. Studies on Chinese Prosodic Grammar. Beijing: Peking University Press.
Fujii, Tomohiro, Kensuke Takita, Barry Chung-Yu Yang and Wei-Tien Dylan Tsai. Forthcoming. 2014. “Comparative Remarks on Wh-adverbials in Situ in Japanese and Chinese.” In Japanese Syntax in Comparative Perspective, edited by Mamoru Saito. Oxford University Press.
Gu, Yang. 2002. “On the Event Structure of the Causative Predicate Shi in Mandarin Chinese.” Paper presented at The 11th Annual Conference of the International Association of Chinese Linguistics, Aichi Prefectural University, Nagoya, August 2002.
Huang, C.-T. James. 1982. Logical Relations in Chinese and the Theory of Grammar. Ph.D. diss., Massachusetts Institute of Technology.
Huang, C.-T. James. 1994. “Verb Movement and Some Syntax-Semantics Mismatches in Chinese.” Chinese Languages and Linguistics 2(39): 587–613.
Huang, C.-T. James. 1997. “On Lexical Structure and Syntactic Projection.” Chinese Languages and Linguistics 3: 45–89.
Huang, C.-T. James. 2004. “Syntactic Analyticity and Parametric Theory.” Paper presented at The 12th Annual Conference of the International Association of Chinese Linguistics, Nankai University, Tianjin, June 2004.
Huang, C.-T. James, and Luther C.-S. Liu. 2000. “Logophoricity, Attitudes and Ziji at the Interface.” In Syntax and Semantics, vol. 33: Long-Distance Reflexives, edited by Peter Cole, Gabriella Hermon, and C.-T. James Huang, 141–195. New York: Academic Press.
Jaeggli, Osvaldo 1982. Topics in Romance Syntax. Dordrecht: Foris.
Karttunen, Lauri. 1977. “The syntax and semantics of questions.” Linguistics and Philosophy 1: 3–44.
Kikushima, Kazunori. 2013. “On the Architecture of Modals and Applicatives in Japanese: A Cartographic Analysis.” PhD. diss., National Tsing Hua University, Taiwan.
Lin, T.-H. Jonah. 2001. “Light Verb Syntax and the Theory of Phrase Structure.” Ph.D. diss., University of California, Irvine.
Michelioudakis, Dimitris and Ioanna Sitaridou. 2008. “The Ethic Dative in Modern Greek and Romance.” Paper presented at the 29th Annual Meeting of the Department of Linguistics, Aristotelian University of Thessaloniki, May 2008.
Parsons, Terence. 1995. “Thematic Relations and Arguments.” Linguistic Inquiry 26(4): 635–662.
Ramchand, Gillian. 2008. Verb Meaning and the Lexicon: A First Phase Syntax. Cambridge: Cambridge University Press.
Rizzi, Luigi. 1997. “The Fine Structure of the Left Periphery.” In Elements of Grammar: Handbook of Generative Syntax, edited by L. Haegeman, 281–338. Dordrecht: Kluwer.
Rizzi, Luigi. 2001. “On the Position of ‘Int(errogative)’ in the Left Periphery of the Clause.” In Cinque, G. and G. Salvi, eds., Current Studies in Italian Syntax: Essays Offered to L. Renzi. New York: Elsevier, 287–296.
Shlonsky, Ur, and Gabriela Soare. 2011. “Where’s ‘Why’?” Linguistic Inquiry 42(4): 651–669.
Stepanov, Arthur, and W.-T. Dylan Tsai. 2008. “Cartography and Licensing of WH-Adjuncts: A Cross-Linguistic Perspective.” Natural Language & Linguistic Theory 26(3):589–638.
Svenonius, Peter. 2005. “Two Domains of Causatives.” Ms., CASTL, University of Tromsø.
Tsai, W.-T. Dylan. 1994a. “On Nominal Islands and LF Extraction in Chinese.” Natural Language & Linguistic Theory 12(1): 121–175.
Tsai, W.-T. Dylan. 1994b. “On Economizing the Theory of A-bar Dependencies.” PhD diss., Massachusetts Institute of Technology.
Tsai, W.-T. Dylan. 1999. “On Lexical Courtesy.” Journal of East Asian Linguistics 8(1): 39–73.
Tsai, W.-T. Dylan. 2001. “On Subject Specificity and Theory of Syntax-Semantics Interface.” Journal of East Asian Linguistics 10(2): 129–168.
Tsai, W.-T. Dylan. 2007a. “Conjunctive Reduction and its Origin: A Comparative Study of Tsou, Amis, and Squliq Atayal.” Oceanic Linguistics 46(2): 585–602.
Tsai, W.-T. Dylan. 2007b. “Two Types of Light Verbs in Chinese.” Paper presented at the Joint Meeting of the 15th Annual Conference of the International Association of Chinese Linguistics and the 19th North American Conference on Chinese Linguistics, Columbia University, New York City, May 2007.
Tsai, W.-T. Dylan. 2008a. “Left Periphery and How-Why Alternations.” Journal of East Asian Linguistics 17(2): 83–115.
Tsai, W.-T. Dylan. 2008b. “Object Specificity in Chinese: A View from the vP periphery,” The Linguistic Review 25(3–4): 479–502.
Tsai, W.-T. Dylan. 2012. “High Applicatives Are Not High Enough: A Cartographic Solution.” Paper presented at the 14th China International Conference on Contemporary Linguistics, Xi’an Foreign Language University, Xi’an, May 2012.
Tsai, W.-T. Dylan. Forthcoming. “Causality, Comitativity, Contrastivity, and Selfhood.” In Interfaces in Grammar, edited by Jianhua Hu and Haihua Pan. Philadelphia: John Benjamins.
Tsai, W.-T. Dylan and Melody Y.-Y. Chang 2003. “Two Types of Wh-Adverbials: A Typological Study of How and Why in Tsou.” In The Linguistic Variation Yearbook, vol. 3, edited by Pierre Pica, 213–236. Philadelphia: John Benjamins.
(*) I have benefited greatly from discussions with Adriana Belletti, Lisa Cheng, Candice Cheung, Yang Gu, James Huang, K. A. Jayaseelan, Kazunori Kikushima, Audrey Li, Luther Liu, Keiko Murasugi, Paul Portner, Mamoru Saito, Arthur Stepanov, Rint Sybesma, Sze-Wing Tang, Ting-Chi Wei, Iris Wu, and Barry Yang. Special thanks to Memo Cinque, Richie Kayne, and Luigi Rizzi for their inspiration and encouragement over the years. The research leading to this article is sponsored by the Chiang Ching-kuo Foundation for International Scholarly Exchange, by the Research Center for Humanities and Social Sciences of National Tsing Hua University, and by the National Science Council of Taiwan (NSC 89-2411-H007-046; NSC 94-2411-H-007-020; NSC 96-2411-H-007-026)
((i)) If a language has the option of Merge, it will always use it.
(2) The abbreviations used in this paper are glossed as follows: 1S: first person singular; 3S: third person singular; Acc: accusative case; Aff: affective marker; ApplP: applicative phrase; AV: actor voice; Cl: classifier; Dat: dative case; Eth.dat: ethic dative; EvaP: evaluative phrase; Foc: focus marker; FP: focus phrase; Inc: inchoative aspect; IntP: interrogative phrase; Lnk: linker; Nom: nominative case; Obl: oblique case; Past: past tense; Prf: perfective aspect; Q: question particle; Rea: realis mood; Res: resultative aspect; Top: topic marker.
(4) From our point of view, the raising-to-CAUSE construal discussed here should not be taken to be an instance of incorporation, in that the light verb is implicit and syntactically separate from the main verb. Huang (2004), in particular, points out that Chinese differs from English in forming analytic syntactic constructions. One fine example comes from light verbs such as da ‘hit’: While English uses the denominal verb telephone, Chinese employs its “decomposed” counterpart da dianhua ‘hit telephone/do telephoning’. It is in this sense that we take the above causative construals as an instance of verb movement rather than incorporation. Also, as noted by a reviewer, there are some differences between the implicit (p.29) causative light verb and its lexical counterpart. It may well be the case that CAUSE is actually a grammaticalized version of rang ‘cause’, with the latter still keeping most of its verb characteristics.
(5) Here an interesting issue raised by a reviewer has to do with the contrast between the following two examples:
It is pointed out that a nonagentive subject is compatible with the implicit outer light verb CAUSE, while an agentive subject is not. This is unexpected given the following legitimate usages of their lexical counterpart rang ‘make’ in both cases:
As noted in the review, the ungrammaticality of (ib) results from some form of thematic conflict. On the other hand, it should be equally pointed out that there are two types of causation that are often encoded morpho-syntactically (i.e., agentive causation vs. eventual causation, roughly corresponding to the agentive-nonagentive distinction of Causers in Pesetsky’s works on psych-verbs). Along this line, it is not unreasonable to suggest that the inner-outer dichotomy also applies to the lexical causative light verb rang in (iia,b), while its implicit counterpart CAUSE always occupies the outer light verb position in Chinese. As a result, CAUSE can never host an agentive subject laoshi ‘teacher’ at the edge of vP, hence the ungrammticality of (ib). In addition, this move fares very well with the general pattern of wh- and reflexive adverbials in that only inner adverbials observe subject agentivity.
(6) As noted by a reviewer, the extensive inventory of Chinese light verbs could be an issue. In fact, the theoretical status of Chinese light verbs is still quite debatable: For instance, we may take AFFECT to be an applicative morpheme on a par with African languages. It is equally possible to reduce COST to CAUSE, in that the former can be further decomposed into ‘cause … to spend. . .’. The exact treatment, however, is beyond the scope of this paper, where we will focus on the core cases discussed in Huang (1994,1997), Feng (2003, 2005), and Lin (2001).
(7) The exact cause of the deviance of (29a) is not entirely clear. One way to think of this issue is to suggest that raising-to-USE may create an obligatory object control configuration. As a result, the deviance of (29a) is attributed to the implausible pragmatic construal ‘until the knife became very tired.’
(8) A reviewer raises a couple of potential problems with this null operator analysis: On the conceptual side, it is unclear whether a causative predicate may take (p.30) a null operator construction as its complement. This point of technicality is well-taken although we see no a priori reason why the outer light verb cannot work like tough-predicates or Chinese long passives. On the empirical side, this analysis wrongly predicts that (32) should allow a resumptive pronoun for the null operator construal, as shown by the ungrammaticality of (i):
This restriction on the resumptive strategy, however, may have a very simple explanation: Namely, here the subject of CAUSE is inanimate, and an inanimate argument typically does not license a resumptive pronoun in Chinese, as evidenced by the deviance of (iib):
The same pattern obtains for Chinese long passives with an inanimate subject, as illustrated by the following contrast:
As a matter of fact, the unavailability of the resumptive strategy is most likely due to the [+human] origin of Chinese third person pronoun, whose distribution is still highly restricted when construed as inanimate.