3.4.1 General considerations
Recall that, descriptively, Comparative Deletion (CD) is a process which elimi-nates the QP or the quantified DP from the subclause, if it is logically identical to its antecedent in the matrix clause (Bacskai-Atkari 2010b, Bacskai-Atkari 2012a) in examples such as (1), repeated here for the sake of convenience as (37):
(37) a. Ralph is more qualified than Jason isx-qualified.
b. Ralph has more qualifications than Jason hasx-many qualifications.
c. Ralph has better qualifications than Jason hasx-good qualifications.
Comparative subclauses exhibit regular operator movement (see Chomsky 1977, Kennedy 2002) to a [Spec,CP] position. This is illustrated in (38):
(38) CP
C′ C than
CP Op. C′
C IP
As for the structure of the left periphery, I assume that the CP can be split (cf.
Rizzi 1997: 297, Rizzi 1999: 1, Rizzi 2004: 237–238), even though I do not adopt a cartographic approach and the various CPs should rather be regarded as in-stances of an iterated CP. Rizzi’s split CP is illustrated in (39):
(39) [CP [TopP* [FocP [TopP* [CP]]]]]
Rizzi assumes that multiple TopPs and a designated FocP may appear between the two CP projections; however, this is irrelevant for the present analysis, and therefore I will neither include them in the representations, nor will I discuss possible arguments against a strict cartographic approach. Further, Rizzi (1997;
1999; 2004) attributes different functions to the two CPs: he assumes that the higher C head is responsible for the “illocutionary” Force of the clause, while the lower is responsible for Finiteness.
The term “illocutionary Force” is fundamentally used to cover clause types, that is, categories such as declarative, interrogative, relative, comparative, etc.; it is terminologically unfortunate to involve the concept of illocution since the kind of illocution discussed by Rizzi has little to do with how Austin (1962) and Searle (1969) introduced the term, the sentence types in question not being performative.
In addition, the distinction between Force and Finiteness is problematic as well because, though the relative position of a given C head in a combination (that is, whether it is a lower or a higher one) is straightforward, it is hard to disentangle the various functions in cases where a single C head marks both. Due to these reasons, I will henceforth not mark the Force/Finiteness distinction.
3.4.2 Predicative versus attributive and nominal structures
Turning back to the representation given in (38), the complementiser head of the comparative subclause (than) occupies the higher C position, while the compar-ative operator (Op.) moves to the specifier of the lower CP.
In predicative structures, such as (37a), the QP containing the AP is headed by a phonologically empty operator (x), and the entire QP moves up to the specifier of the CP, where it is deleted. By contrast, in nominal and attributive structures, such as (37b) and (37c), respectively, the QP is an adjunct within the DP (Ken-nedy & Merchant 2000, Kántor 2008a) and thus the entire DP moves up and is deleted. This is because the QP cannot be extracted from the DP due to the DP-island constraint (cf. Kayne 1983, Ross 1986, Izvorski 1995: 217, Grebenyova 2004, Bošković 2005).
Movement in predicative structures is represented in (40), based on (37a):
(40) CP
C′ C than
CP QPi x-qualified
C′ C
∅
IP Jason is ti
Movement in attributive and nominal structures is represented in (41), based on the examples given in (37b) and (37c):
(41) CP
C′ C
than
CP DPi
x-many qualifications x-good qualifications
C′ C
∅
IP Jason has ti
All this can be derived from more general rules and is hence not specific for comparative subclauses, as similar phenomena can be observed in other construc-tions containing operators (cf. Kennedy & Merchant 1997: 7). Consider the exam-ples given in (42):
(42) a. *Howis Ralphqualified?
b. How qualifiedis Ralph?
c. *How bigdid Ralph seecats?
d. How big catsdid Ralph see?
e. *How manydid Ralph seecats?
f. How many catsdid Ralph see?
As can be seen, the QPhow qualifiedand the DPhow big catsorhow many cats can be moved only as a whole: neither the Q head may be extracted from the QP nor the QP from the DP. I will return to the issue later, also casting light upon how it varies cross-linguistically. At this point, suffice it to say that in cases such as (42a) and (42b) above the Q head cannot be extracted because then it would have to occupy a phrase position in the lower [Spec,CP] as a head. On the other hand, as I will show later, in some languages the quantifier may also be realised as a QP modifier within the QP heading the adjective in question, and it can in such cases be extracted, cf. Kántor (2008a). Similarly, the extraction of the QP out of the DP is highly dependent on the parametric settings of a given language:
while English, Bulgarian and Greek prohibit it, it is allowed in Polish and Czech (Kennedy & Merchant 2000); these questions will be addressed inChapter 4 in detail.
Turning back to comparative subclauses in English, it is important to investi-gate the issue of copies. In our case, there are only two copies to consider: the lower one in the base position of the QP or the DP and the higher one in the lower [Spec,CP] as a result of movement. The higher copy, as has already been seen, is deleted by Comparative Deletion; note that this is independent from whether the AP or NP is identical to the one in the matrix clause. The lower copy is regularly deleted by PF (cf. Bobaljik 2002, Chomsky 2008, Bošković & Nunes 2007: 44–
48), which is possible because the QP or DP in question is e-given. The deletion processes taking place in (37) are shown in (43):
(43) a. Ralph is more qualified [CPthan [CP[QPx-qualified] Jason is [QP x-qualified]]].
b. Ralph has more qualifications [CPthan [CP[DPx-many qualifications]
Jason has [DPx-many qualifications]]].
c. Ralph has better qualifications [CPthan [CP[DPx-good qualifications]
Jason has [DPx-good qualifications]]].
As should be obvious, Comparative Deletion takes place in all structures, hence there is no difference between predicative and attributive/nominal constructions:
the fact that the entire DP has to be eliminated in the latter is due to different, independent constraints.
One obvious advantage of this approach is that it accounts for the deletion of QPs and DPs without having to resort to extra mechanisms: Comparative Dele-tion takes place in the lower [Spec,CP] posiDele-tion and it deletes any material that is there. In turn, differences in terms of what phrases are found there arise sim-ply out of movement constraints. I will return to the issue of why Comparative Deletion has to take place at all later, also accounting for the differences found between languages and varieties. At this point, suffice it to say that a movement analysis claiming that the entire QP or DP moves (and not only the operator) can successfully account for the elimination of both copies by assuming that Compar-ative Deletion obligatorily takes place in the lower [Spec,CP], eliding the higher copy, and that lower copies are regularly deleted at PF. Though it is a prerequi-site that deleted material has to be e-given, the fact that obligatory deletion takes place is not directly linked to these elements being recoverable, contrary to Ken-nedy (2002): rather, it is associated with a syntactic position where it happens independently of whether the material there is e-given or not.
3.4.3 Comparative Subdeletion
The case of Comparative Subdeletion, as found in subcomparatives, may at first sight seem to be a counterexample for what has been established for Comparative Deletion. In these (predicative) structures, as was mentioned at the beginning of this chapter, the QP in the subclause remains overt:
(44) The table is longer than the desk iswide.
However, even in such cases Comparative Deletion takes place regularly in the [Spec,CP] position: if Comparative Deletion did not occur, then the higher copy should remain (cf. Bacskai-Atkari 2010b). On the other hand, the lower copy cannot be eliminated since it is F-marked: it contrasts with the AP (long) in the matrix clause. As pointed out by Bošković & Nunes (2007: 48), lower copies may remain overt if the pronunciation of the higher copy would make the derivation crash at PF. Thus, the following happens in (44):
(45) The table is longer [CPthan [CP[QPx-wide]Fthe desk is [QPx-wide]F]].
As can be seen, the higher copy of the QP is deleted by Comparative Deletion exactly the same way as in (43a) and the two clauses differ in fact only with re-spect to whether the lower copy remains; however, this difference can be derived from recoverability. This all indicates that subcomparatives are not exceptional in terms of Comparative Deletion, and thus there is no separate Comparative Subdeletion process.
In this way, the relation between Comparative Deletion and Comparative Sub-deletion can be easily handled, without having to resort to distinguishing the two on the basis of whether they include overt or covert movement, as was seen in connection with Kennedy (2002). Again, the role of information structure is not directly related to Comparative Deletion itself: Comparative Deletion is treated as a mechanical process eliminating material from the lower [Spec,CP] position and the fact that the lower copy of the QP can remain overt is due to F-marking.
Note that being F-marked is not identical to not being e-given; it is rather intended to express some kind of contrast. For instance, the QPx-widein (45) is in contrast with the QPlonger of the matrix clause. Also, this QP appears in a clause-final position, which is the canonical position for foci and/or contrasted elements in English: see Selkirk (1984; 1986), Nespor & Vogel (1986), McCarthy &
Prince (1993). This QP expresses the main contrast involved in comparison and it follows logically that it appears in a position where it can bear main sentential stress.
As far as the overt lower copy of an e-given AP is concerned, it is usually ungrammatical because it should regularly be eliminated as a lower copy and it should not appear in a contrastive position. However, if there is a context in which it can be interpreted as a contrasted element even though it is given, it may remain overt: see also (25). The difference is illustrated in (46) below:
(46) a. ⁇/*The table is longer than the desk islong.
b. A: The table is longer than the desk is wide.
B: No, the table is longer than the desk isLONG.
In both cases the subclause contains an overt lower copy of the QP that is identical to the one in the matrix clause. However, in (46a) it should have been eliminated as there is no additional instruction for PF to preserve the lower copy.
As opposed to this, (46b) is grammatical because the QP in question is contrasted:
this contrast holds not with the QP in the matrix clause but with the one in the preceding sentence.
It can be concluded that subdeletion constructions also include Comparative Deletion in the regular way, and the fact that the lower copy remains overt stems from constraints independent from the mechanism of Comparative Deletion.