Abstract
In this chapter, I start to argue for an articulated VP structure by investigating the position of derived objects. While syntacticians generally agree that there is a derived object position, there is less agreement on the details of the landing site. I review some past proposals for derived object positions. I look at cases where either the merged object position is filled or the derived object position is filled as well as cases where both of these positions are filled simultaneously. I then use cross–linguistic data to argue (i) that the derived position is below the merged position of the external argument and (ii) that elements that appear in this position come to be there through A–movement.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
As will become obvious, this articulated VP structure has much in common with Chomsky’s “little” vP structure (Chomsky 1995: Chapter 4), Hale and Keyser’s I-syntax structure (Hale and Keyser 1993, 2002), and the work of others such as Arad (1998), Harley (1995), Pylkkänen (2002), and Rackowski (2002). Because the translation from one system to another is not trivial, I use my own labels of V1 and V2 instead of v and V.
- 2.
In Chapter 7, I will discuss some very particular cases where a theta-role may be assigned to the Spec of this category.
- 3.
Some work (e.g., Pesetsky 1995; Hornstein 1999) posits movement that violates the Theta-Criterion but the claim that movement into a theta-position is ruled out is still widely assumed.
- 4.
Number is not marked within many Malagasy DPs. Glosses and translations will vary in terms of whether the singular or plural is given.
- 5.
In fact, there are arguments that objects raise in English based on more subtle data from word order as well as binding facts. See Lasnik (2001), for example. On the other side, there are accounts in which the Malagasy word order shown is accounted for without raising to object (e.g., Pearson 1997).
- 6.
The v in these examples is meant to indicate where the finite verb originates.
- 7.
This example is glossed as in the original. Hvers vegna means ‘why’. The translation was provided by A. Holmberg (p.c.).
- 8.
This is a simplified version of Johnson’s full account but sufficient for our needs. It is important to note here that overt movement of the object in English necessitates overt movement of the verb as well in order to arrive at the appropriate word order.
- 9.
Sportiche (1998) is the published version of Sportiche (1990). The idea of a VP-internal derived object position is also contained in Koopman and Sportiche (1991).
- 10.
Another early account using a type of split VP hypothesis is Lefebvre (1991). In an earlier version of the present work, I included her data from Haitian Creole as supporting evidence. Because of the complications introduced by agree with no movement, however, the conclusions are less clear. I thank a reviewer for pointing this out.
- 11.
Koizumi gives other arguments for his structure that involve adverb placement, the Minimal Link Condition, Participial agreement, and the Chain Condition. I refer the reader to his work for more development of his ideas. Since the early 1990s, others have used this sort of structure as well. My aim here is simply to give an early account.
- 12.
Some English speakers find this awkward.
- 13.
This argument alone might suffer if Boskovic’s (2004) proposal that floated quantifiers cannot be found in theta-positions is correct. However, as mentioned in footnote 11 of Chapter 1, Koizumi provides other evidence as well.
- 14.
Much work has been done on this sort of object movement. For more literature on this and different viewpoints, see Holmberg (1986), de Hoop (1996), Neeleman and Reinhart (1998), Fox and Pesetsky (2005), and references cited therein.
- 15.
The term “Chinese” will refer to Mandarin throughout the book.
- 16.
There is also a change in the verb from cha ‘stick’ to cha-ma n ‘stick-full’. This type of verb change will be important to the discussions in Sections 4.5.2 and 8.2.1.
- 17.
More will be said of the interaction of specificity and the aspectual class of the predicate in Chapter 5.
- 18.
Others have worked on the issue of the varying position of the object in this and related languages and accounted for it via case-related object shift (e.g., Noonan 1992b; Guilfoyle 1993; Bobaljik and Carnie 1996). An interesting point in Bobaljik and Carnie’s work (p. 229) is that one of their arguments that subjects in Irish cannot be in situ is dependent on their assumption that the only position for a derived object is above the merged position of the external argument: “… if the object has raised overtly to the specifier position of AgrOP yet the subject still precedes the object, then the subject must have raised past the object.” Making the opposite argument, Ramchand assumes that the subject is in situ and therefore the derived object position must be lower. Since McCloskey (1996) presents convincing arguments against having subject in situ in Irish, I choose not to assume that it is in situ.
- 19.
The direct case is also what is used for nominative subjects.
- 20.
I have left this example as in Baker (1988: 247), even though it seems as if the translation should be in the past tense.
- 21.
I discuss the possibility of base-generation in Section 2.5.4.
- 22.
I assume that the non-promoted logical object will be licensed (case-marked) in situ. This will be similar to non-promoted logical subjects in Western Malayo-Polynesian languages (see (43) below).
- 23.
It is important to keep in mind here, as discussed in Chapter 1, that evidence from other languages is being used to determine what is universal to language.
- 24.
Koizumi’s (1995) analysis of na constructions in Zarma, a Nilo-Saharan language, is very similar to Sybesma’s analysis of ba constructions in Chinese. Note that, in fact, in these analyses, while the object is argued to move, the movement is obligatory. The preverbal vs. postverbal status of the object is determined not by whether there has been object movement but by whether there has been verb movement across the derived object. Sybesma’s arguments for object movement are given below. Given that this obligatory movement is string-vacuous in the case of the non-ba construction, I will be assuming, contra Sybesma, that only in the ba-construction is the movement obligatory, accounting for the differences in the restricted interpretation of ba-constructions. We have seen (and will discuss again) a similar case in Scots Gaelic where verb movement obscures the position of the derived object. When the verb does not move, however, we can see the two object positions clearly.
- 25.
‘Cry’ in this construction behaves like a manner adjunct. A paraphrase might be ‘This thing made Zhang-san tired through a crying event’ where Zhang-san has been doing the crying. See Tomioka (2006) and Vinka (1999) for similar analyses of Swedish, Japanese, and English resultatives.
- 26.
27 EXTP is Extent Phrase, which is similar to Hoekstra’s (1992) small clause. See Sybesma (1992: 74ff) for details.
- 27.
See Fox and Pesetsky (2005) for a way of accounting for these facts.
- 28.
Vinka is reporting on a dialect of Swedish spoken in Northern Sweden and in Finland. I am grateful for his input on these data.
- 29.
The purpose of Vinka’s paper is to argue for a phrase structural distinction between two different types of particles in Swedish. My interest is only in the predicative type of particle since this is the one that allows both low object shift (across the particle) and high object shift.
- 30.
In fact, the phrase structure is probably much more complicated than what I present here. Most likely there is iterative predicate movement in many of these languages, as proposed by Pearson (1998), Rackowski (1998), and Rackowski and Travis (2000). I believe that none of the claims that I make in this book are affected by these proposals.
- 31.
Because an Experiencer is also found in this position, the label Agent has been avoided in the literature, replaced by Actor. I will often just use the term Agent, however, especially when discussing particular constructions where the external argument is in fact an Agent.
- 32.
This, in fact, would be the subject counterpart to the applicative constructions discussed above where both the merged and the derived object position are filled simultaneously.
- 33.
I use the term traditional term “topic” here mainly to be consistent with much of the literature in this area and to distinguish this position from the VP-internal “subject” position. See Kroeger (1993), however, for arguments that this “topic” bears little similarity to the discourse notion of Topic. Others, however, such as Sells (2000), Richards (2000), and Pearson (2001), have maintained that this position is a topic, or at least an A’-position.
- 34.
Obviously, some explanation for why a DP is licensed in this position is required. See Section 3.3.1.2 for an account of Case in Malagasy.
- 35.
The apostrophe in the Malagasy examples and glosses indicates a process labeled N-bonding by Keenan (2000).
- 36.
There are a variety of ways of promoting the object to subject position. I give only what is called the “suffix passive” (see Paul 2000 and Pearson 2001 for more on this in a Chomskian framework).
- 37.
I will return to a possible analysis of this in Section 3.4. I have benefited from discussions with Mark Campana many years ago concerning Kalagan and the problems that it raises for phrase structure.
- 38.
As an anonymous reviewer pointed out, it could be that the derived DP is in a high derived position outside of the VP and that both the verb and the external argument (Agent DP) have moved even higher. I resist this type of account since I see no need for the extra movements. Further, if the external argument is in situ, we can explain why the thematic content of this position is restricted to the external argument. In other words, if this DP were in a high derived position, we might expect other arguments (benefactives, instrumentals) to appear there.
- 39.
As shown by, for example, Bobaljik (1995), there is an effect on the information structure, however.
- 40.
A reviewer points out that this is similar to the English I hit Mary’s ar m and I hit Mary on the arm vs. I saw Mary’s ar m and *I saw Mary on the arm. However, this type of structure is much more productive in languages with true possessor raising. See example (58) where an idiom chunk can raise. Landau (1999) presents arguments from Hebrew supporting a movement analysis of Possessor Raising. He also calls into question whether affectedness is the appropriate restriction on the predicate or whether the restriction has to do with the nature of the external argument. I believe that his observations could be accommodated with minor changes to my account.
- 41.
This was not the first time this sort of structure was proposed in an articulated VP format. See, for example, Hoekstra (1995), Harley (1996).
- 42.
A reviewer points out that this distinction may come from the fact that through can be analyzed as a transitive preposition in (55b). It is true that other examples (where the preposition cannot be analyzed as being transitive) may not show the same distinction (I burned up the papers vs. I burned the papers up). This weakens the claim that the English particle constructions also show a shift in meaning. However, the point remains that the lower derived object position may host non-arguments, as shown in (56).
- 43.
See the discussion in Sybesma (1992: 146ff) concerning Huang’s disagreement with Goodall’s conclusion. Part of the issue is what counts as an idiom and thereby an idiom chunk, an investigation which would take us too far afield and must be left for future research. Recall that Cheng (1986) has shown that the ba NP must be an affected argument. More likely, the distinction has to do with stative vs. dynamic predicates.
- 44.
As we have seen above, Chomsky (1995) proposes that the object moves to a second Spec, vP position (“little v”). Movement, therefore, does not provide evidence for Spec of an additional nonlexical category. Further, he considers v to be a functional category. I differ from Chomsky in assuming that case-checking in a Spec, head configuration only happens with nonlexical categories, and that V1 therefore is a lexical category. My view of lexical vs. functional categories is introduced briefly in Chapter 1 and fleshed out in Section 6.5.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2010 Springer Science+Business Media B.V.
About this chapter
Cite this chapter
Travis, L.d. (2010). Inner Derived Objects. In: Inner Aspect. Studies in Natural Language and Linguistic Theory, vol 80. Springer, Dordrecht. https://doi.org/10.1007/978-90-481-8550-4_2
Download citation
DOI: https://doi.org/10.1007/978-90-481-8550-4_2
Published:
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-8549-8
Online ISBN: 978-90-481-8550-4
eBook Packages: Humanities, Social Sciences and LawSocial Sciences (R0)