Situación contable de la C.O.S.E.M.A

The two analyses in 5.1 and 5.2 approach the phenomenon of distance-distributivity from a purely syntactic perspective. Neither of them is a fully adequate account for English

each, nor can they be applied to German jeweils or other distance-distributive elements

that are licensed in subject position. Given this state of affairs, it is unclear what a purely syntactic account of distance-distributive elements should look like if the two basic tools of syntactic analysis – an analysis in terms of binding and an analysis in terms of extraction – do not work.

One may wonder if an integrated syntactic and semantic account would not produce better results, especially in light of the peculiar semantic behaviour of distance- distributive elements.

The special semantic behaviour of distance-distributive elements in various languages has been analysed in Choe (1987), Junker (1995), Link (1998), and Moltmann (1991, 1997). These analyses will be discussed in chapter IV.3. Unfortunately, the above studies (with the exception of Moltmann 1997 perhaps) do not pay much attention to surface compositionality. Nor do they aim at providing a unified cross-linguistic analysis of distance-distributive elements that would shed light on their different syntactic distribution in different languages.

In chapter III to V, I attempt to fill this gap. I will provide a unified analysis for distance-distributive constructions in a variety of languages. All constructions are argued to share one underlying structure, and all constructions are argued to be interpretable from

surface structure.36 _{The observable cross-linguistic differences in distribution are derived}

from three independent factors: (i.) differences in the grammatical properties of the distance-distributive element itself (this is reminiscent of Burzio’s anaphor account); (ii.) general syntactic differences between languages; and (iii.) the interaction of semantic operations involved in interpreting distance-distributive constructions with syntactic factors, which may cause the semantic derivation to crash. It will be argued that a complete and accurate account of the behaviour of distance-distributive elements, both within one language and cross-linguistically, is possible only by taking into account all three factors.

6 Conclusion

This introductory chapter on jeweils in particular, and on distance-distributivity in general

has brought to light the following facts:

(80) i. Jeweils is ambiguous between an adverbial and an adnominal reading.

ii. The two readings of jeweils correspond to two different syntactic positions:

Jeweils is interpreted adverbially when it can be construed as being adjoined to

VP. Jeweils is interpreted adnominally when it (or its trace) is adjacent to a

numeral or indefinite expression, the DistShare (and when the semantic requirements are satisfied).

iii. Some of the properties of jeweils are found back with a number of formally

related expressions, the so-called s-expressions, arguing for a unified analysis.

iv. Adnominal jeweils is not a floated quantifier, but a distance-distributive

element.

v. Distributive elements with similar properties are attested in a variety of languages, including Italian, French, Dutch, Icelandic, Norwegian, Russian, Czech, Bulgarian, Korean and Japanese.

vi. Observable cross-linguistic differences in the expression of distance-

distributivity arguably follow from the values of two parameters: [+/- D- features] and [+/- DP-internal movement].

vii. The discussion of Burzio (1986) and Safir & Stowell (1988) on distance-

distributive (= binominal) each has shown that a purely syntactic analysis does

not give a satisfactory account of the phenomenon. This argues for an integrated syntactic and semantic account.

Chapter III presents the syntactic analysis of jeweils in particular, and of distance-

distributivity in general. The two parameters in (80vi) will be motivated and their application will be discussed in detail. Chapter IV presents the semantic analysis of distance-distributivity. It will be shown that a surface compositional interpretation of distance-distributive constructions is possible.

The main results of chapter II are the following. There are two instances of the distributive

element jeweils: Jeweils can occur both in adverbial and in adnominal position. The

difference in syntactic position is accompanied by a difference in interpretation.

Furthermore, adnominal jeweils was shown to behave like a distance-distributive

quantifier. As such, it poses a problem for compositionality. Finally, it was shown that distance-distributive quantifiers occur in a variety of languages in systematic fashion. This fact calls for a unified explanation.

The objectives of this chapter are threefold. The first objective is to present a complete

syntactic analysis of both instances of jeweils. The syntactic analysis identifies the

syntactic structures that will be interpreted in chapter IV.

The second objective is to shed more light on the syntactic nature of the phenomenon of distance-distributivity in German and other languages. Of course, this enterprise goes

hand in hand with providing a general syntactic account of adnominal jeweils. It will

emerge that jeweils-DPs exhibit characteristic properties of so-called ‘Inverse Linking

Constructions’ (henceforth ILCs), illustrated in (1): (1) [DP One apple [PP in [QP every basket]]] is rotten.

‘In every basket, there is rotten apple.’

The syntactic analysis of ILCs will then form the basis for the syntactic analysis of

adnominal jeweils. The upshot of the discussion will be that adnominal jeweils involves a

regular QP that is embedded inside a complex DP. The quantifier Q syntactically combines with its restriction (in line with compositionality). This restriction is realised as an overt or covert proform which is co-indexed (hence co-referent) with the DistKey antecedent.

The third objective of chapter III is to provide an explanation for the cross-linguistic variation observed with distance-distributive elements (DDs). The discussion, which is based on relatively few, mostly Indoeuropean languages, is intended merely as laying the basis for a more comprehensive typology of distance-distributivity. The languages under discussion will be shown to differ regarding two parameters. This leads to a cross- classification into four potential language types.

The first parameter concerns the possibility of DP-internal (predicate) fronting for information structure reasons. It turns out that the DD can occur in prenominal position in languages that allow for such DP-internal fronting (e.g. German, Dutch, French, Slavic, Japanese, Romanian). The DD follows the DistShare in languages that disallow discourse- driven DP-internal movement (e.g. English). This way, the relative order of DD and DistShare is reduced to a general syntactic property of the languages under discussion. The second parameter concerns the morphological shape of the DD, which I take to reflect

its feature contentovertly. It turns out that those DDs that contain D-features, and that for

this reason are formally identical with regular D-quantifiers, cannot distribute over non- DP antecedents (e.g. pluralities of events as denoted by verb conjunction), nor can they occur with underlying subjects. This is the case in English, Dutch, Norwegian, Icelandic,

Russian, and the Romance languages. The impossibility to occur with underlying subjects will be attributed to the obligatory checking of D-features (between the DistKey and the DD) under agreement, which is subject to c-command in line with Chomsky (1998). DDs in underlying subject position are not c-commanded by the DistKey, and are consequently excluded from this position in languages where D-feature checking is obligatory. In contrast, DDs that do not contain D-features can distribute over any pluralic entity expressed in the clause. Furthermore, they can appear in subject position for they do not have D-features to check under agreement.

The structure of chapter III is as follows. In section 1, I present the syntactic analysis

of adverbial jeweils. In section 2, we will take the first steps towards a syntactic analysis

of adnominal jeweils, the more interesting case because of the compositionality problem.

More arguments for the constituency of jeweils and the DistShare expression are

provided. This constituent is argued to form a complex DP syntactically. Section 3 discusses the syntax of ILCs in form of an interlude. In 4, we turn back to the syntactic

analysis of adnominal jeweils, which will be based on the syntactic analysis of ILCs. In

section 5, the analysis is extended to other languages. There, d-distributivity is approached from a cross-linguistic perspective along the lines sketched in the preceding paragraph. The chapter concludes in section 6.