Abstract
It is a well-known characteristic of German syntax that “incomplete” or “partial” constituents can be fronted. Consider the following examples involving topicalization of a verbal category.1
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
Preview
Unable to display preview. Download preview PDF.
References
See, e.g., Paul (1919, §68) for some early discussion. Paul gives examples like Gemahnt hab ich ihn schon oft (lit. ‘Reminded have I him already often’), and notes: “In besonderen Fällen kann der Inf. oder das Part. wie ein anderes Satzglied als psychologisches Subj. an die Spitze des Satzes treten.”
Huang (1993) and Bayer (1993) in addition assume that ß also contains a trace of the subject, which follows under the Predicate-Internal Subject Hypothesis (cf. Kitagawa (1986), Sportiche (1988), Speas (1990), and Koopman Sportiche (1991), among many others). Here and henceforth, I will mostly ignore possible VP-internal subject traces, unless this affects an issue under discussion; see in particular chapter 5.
In (3), string-vacuous scrambling of NP2 ends up in a VP-adjoined position, and remnant VP topicalization affects only the lower VP segment. Such an analysis might be problematic under several theories of projection and adjunction; see, e.g., Kayne (1994) and Kolb (1995). Alternatively, one might postulate (under the remnant movement approach) that NP2 is adjoined to some functional XP between VP and IP. In what follows, I will abstract away from this issue, assuming adjunction to VP and topicalization of a VP-segment in (3), mainly for reasons of simplicity.
This is only partially true in the case of van Riemsdijk (1989, 129), who suggests that after V- or V’-movement, the fronted category is subject to a regeneration operation that creates a full VP.
lnterestingly, work on incomplete category fronting in German which has been carried out within the framework of HPSG reflects more or less the same fundamental dichotomy. Extraction in HPSG is handled in terms of the creation and passing on of a SLASH feature; so, Hinrichs Nakazawa’s (1994, 11ff) approach, which makes crucial use of a SLASH feature on an incomplete topicalized VP, is in many respects similar to the Principles and Parameters approach in terms of remnant movement. On the other hand, it is suggested by Nerbonne (1994, 112ff) and others that an incomplete topicalized VP does not bear a SLASH feature signalling the absence of one or more arguments of V; this approach is similar in certain respects to the Principles and Parameters approach that relies on the movement of “small” verbal categories.
For an attempt to derive both the assumptions in (5) and (7) from independently established principles of movement and projection, cf. Chomsky (1995).
Although there can be little doubt that the sentences in (10) are well formed from a purely grammatical point of view, markedness is also an issue here — sentences of the type in (10) (where the indirect object and the verb are fronted together, and the direct object remains in an IP-internal position) are generally perceived as more marked than sentences of the type in (6) (where the direct object and the verb are fronted together, and the indirect object stays behind). One might hold that these markedness considerations somehow relate to the issue of whether VP fronting retains `normal’ or ‘unmarked’ word order; but, for lack of an articulated theory of markedness as opposed to (un-) grammaticality, I will not pursue the matter any further here.
It should be noted that examples like these are considered ungrammatical by Frey Si Tappe. Similarly, Sabel (1995, 67–72) gives analogous examples that he considers ill formed, and he concludes that non-argument traces may never occur unbound. I do not follow these judgements here (or, in the case of some of Sabel’s examples, I surmise that the illformedness has an independent reason), and assume that remnant movement may create both unbound argument traces and unbound non-argument traces, as long as no independent constraints are violated. That said, the issue of markedness as opposed to (un-) grammaticality that was brought up in the last footnote arises here, too.
Of course, the question arises of why English lacks scrambling in the first place. Here, I will not have anything new to say about that issue; see, e.g., Koster (1986), Webelhuth (1988; 1992), Fanselow (1992; 1993; 1997), Haider (1993), Reuland Kosmeijer (1993), Müller Sternefeld (1993), and Müller (1995) for proposals.
This does not mean that every specific proposal of Chomsky’s minimalist program can be adopted. For instance, in a context that is unrelated to remnant movement constructions of the type I am interested in here, Chomsky (1995, 365) introduces a constraint that strictly blocks movement of any category that contains a trace (his ex. (200)). As observed by Fanselow (1997), adherence to this constraint implies giving up the remnant movement approach to incomplete category fronting. However, independently of the issue of remnant movement, it seems that such a constraint is untenable — among other things, it incorrectly rules out topicalization of a complex NP that dominates a clause in which some other movement operation (topicalization, V/2, etc.) has applied, or topicalization of a clause in which wh-movement has taken place.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 1998 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Müller, G. (1998). Incomplete Category Fronting. In: Incomplete Category Fronting. Studies in Natural Language and Linguistic Theory, vol 42. Springer, Dordrecht. https://doi.org/10.1007/978-94-017-1864-6_1
Download citation
DOI: https://doi.org/10.1007/978-94-017-1864-6_1
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-4941-4
Online ISBN: 978-94-017-1864-6
eBook Packages: Springer Book Archive