3.1 General information about the project <B6> new 3.1.1 Title: Underspecification in Voice systems and the syntax-morphology interface Kurztitel: Voice

3.1.2 Research areas: Syntactic theory, comparative syntax, syntax-morphology, syntax-lexicon interface, lexical se-mantics 3.1.3 Principal investigator: Prof. Dr. Alexiadou, Artemis, 13.02.1969 Institut für Linguistik: Anglistik Universität Stuttgart Keplerstr. 17 70174 Stuttgart Telefon: 0711-6858-3121 Telefax: 0711-6858-3122 Email: artemis@ifla.uni-stuttgart.de Does the principal investigator hold a permanent position?

Yes No, limited until _________________ Further employment is planned until ____________

3.1.4 The project includes:

• research on human subjects or material Yes No A copy of the required approval of the responsible ethics committee is included with the project proposal Yes No

• clinical studies Yes No

• experiments involving animals Yes No

• experiments involving recombinant DNA Yes No

• research with human embryonic stem cells Yes No

Legal authorisation has been obtained Yes No

• research concerning Convention on Biological Diversity Yes No

3.1.5 Requested funding

Financial Year

Funding for staff Funding for direct costs

Funding for in-strumentation

Total

201X/2 201X 201X 201X 201X/1 (Figures in thousands of euros)

3.2 Summary Short summary: The main research topic of this project is the relation between syntactic Voice alternations (and syntactic configurations more generally) and Voice morphology, and its impli-cation for theories of Voice and theories of morphology/syntax interactions. The main goal of this project is to offer an integrative theory of the interaction between the morphological realiza-tion of Voice and argument structural Voice within the framework of Distributed Morphology. Deutsche Zusammenfassung: Der hauptsächliche Untersuchungsgegenstand dieses Projektes ist ei-nerseits die Relation zwischen syntaktischer Voice-Alternation (und syntaktischen Konfiguratio-nen im allgemeinen) und Voice-Morphologie, andererseits deren Auswirkung auf Voice-Theorien und Theorien der Wechselbeziehung zwischen Morphologie und Syntax. Das Hauptziel dieses Projektes besteht in der Entwicklung einer integrativen Theorie der Interaktion von morphologi-scher Voice-Realisation und Argument-Struktur-Voice im Rahmen der Distributed Morphology. Extended Summary: The main research topic of this project is the manner in which syntactic voice alternations relate to voice morphology, and the implications of this for theories of Voice and theories of morphology/syntax interactions. A central observation is that, across languages, these two dimensions of Voice do not stand in a one-to-one relation. Specifically, different operations on a verb’s argument structure are often encoded by the same morphological device, i.e. Voice systems show massive syncre-tisms. While we find such syncretisms in all languages - there is no language that uses a specific morphological realization for each argument structure Voice - their specific shape is subject to considerable variation which, however, is not totally random. Rather they are ordered in subset relations. The main goal of this project is to offer an integrative theory of the interaction between the morphological realization of Voice and argument structural Voice, within the framework of Distributed Morphology. Specifically, we will pursue an analysis that treats Voice morphology as being sensitive to the output of the syntax. This clarifies our position with respect to alternatives which attempt to capture Voice syncretisms by making morphology sensitive to the workings of Argument Struc-ture, with the latter conceptualized as a module of the grammar distinct from the syntax (in the sense of e.g. Grimshaw 1990). Our approach advocates a direct interaction between interfaces: the morphology realizes the features of the syntax directly, without any intermediate levels or features added after the syntactic derivation. Furthermore, we would like to see whether the claim that the middle Voice is distinct from the passive Voice can indeed hold, and whether a proper characterization of the middle voice can be formulated in our theoretical terms. In our investigation, we look at the behavior of Voice in Germanic, Romance and Greek, and we investigate a range of analytic and synthetic constructions which participate in the reali-zation of Voice.

Our results are expected to be significant theoretically as well as empirically. On the theo-retical side, we hope to show how different interpretations of Voice come about from the various structural combinations of the different heads within the building blocks of verbal meaning. On the empirical side, our analysis will seek to identify possible and impossible combinations across individual languages, and to explain why certain generalizations seem to hold cross-linguistically. 3.3 Starting point of the project 3.3.1 State of the art 3.3.1.1 AS- and M-Voice Among the different theoretical frameworks of contemporary linguistics, one finds considerable variation in the way the term Voice is applied. In our project we will use the term Voice in the following two senses: a) Voice denotes a particular alternation in a verb’s argument structure (AS-Voice here) and b) as Voice alternations are typically marked on the verb's morphology, Voice is considered a morpho-syntactic category of the verb (M-Voice here). In B6 we are interested in the interaction between AS and M-Voice, and in particular the manner in which semantico-syntactic Voice alternations (and syntactic configurations more gen-erally) relate to Voice morphology, and the implications of this for theories of Voice and theories of morphology/syntax interactions. Main types of AS-Voices: A central alternation in the literature on AS-Voice is the one between the active Voice and the eventive passive Voice, exemplified in (1) for English: (1) a. John read the book (active) b. The book was read (by John) (passive) Besides the active and the passive Voice, three further AS-Voices are frequently identi-fied across languages, illustrated below with English examples (but see Haspelmath 1990 for a list of further, less frequent AS-Voices and Siemund & Hole 2000 for a broader use of the term ‘Voice’ or ‘Voice alternation’). (i) anticausative Voice refers to spontaneous events like break, open, or melt which can also be construed as transitive/causative verbs: (2) a. John broke the vase (causative) b. The vase broke (anticausative) (ii) generic or dispositional middle Voice: (3) Such books always sell well (iii) reflexive (and reciprocal) Voice involves inherently and naturally reflexive (or reciprocal) verbs, e.g. ‘body care verbs’ (wash, comb), or ‘verbs of assuming position’ (sit down, turn). (4) John washed and combed every morning Terminological clarification: Some studies subsume the three AS-Voices just discussed under the term middle Voice and thus establish a contrast between the basic active Voice with two different non-active Voices: the

passive Voice and the middle Voice (Lyons 1969, Givón 1979, Shibatani 1985, Geniušien÷ 1987, Kemmer 1993, Klaiman 1991, Siewierska 1984). However, the term middle Voice has a range of interpretations in contemporary linguistics. Sometimes it denotes a formal category, i.e. it refers to an inflectional category of the Classical Greek verb (see Kaufmann 2001 for a recent discus-sion). In other cases, it receives a purely semantic characterization, as e.g. in Lyons (1969: 373), where the middle Voice indicates that "the action or state affects the subject of the verb or his interests". Kemmer (1993) suggests that the appropriate semantic characterization of middle Voice makes reference to two properties: i) initiator as affected (endpoint) and ii) low degree of elaboration of events. We note here that the phenomena that these studies subsume under middle Voice are rather heterogeneous both from a functional and a formal perspective, as we will im-mediately see below. In most formal studies, on the other hand, the term middle is used in a more restricted way, namely to refer to dispositional constructions as in (3). Here, following Doron (2003), and Alexiadou & Doron (2007), we will use the term dispositional middle to refer to constructions of the type in (3). We will use the term Middle Voice to refer to a Voice different from the passive Voice which potentially sub-sumes the other three AS-Voices described here. In doing so, we are interested in the question of whether this Voice can indeed be established in its own right, and whether Affectedness is the proper characterization for it. Properties of the main types of AS-Voices: Descriptively, in the eventive passive (1b) the external argument of the verb is not overtly ex-pressed. However, it is generally agreed upon that this argument is implicitly present, as it is se-mantically and syntactically active. This is suggested by a number of well-known tests such as the (non-)licensing of (i) purpose clauses, (ii) agentive by-phrases, (iii) agentive adverbs or (iv) instrumental phrases. Anticausatives differ from actives and passives in that they do not make reference to an (implicit) external argument (e.g. Levin & Rappaport Hovav 1995 or Reinhart 2000; but see sec-tion 3.3.2). It has been widely acknowledged that passives differ from anticausatives in all the tests mentioned in (i-iv) above (Burzio 1981, 1986, Manzini 1983, Williams 1985, Jaeggli 1986, Baker, Johnson and Roberts 1989 and others). Furthermore, unlike all the other AS-Voices, anti-causatives are compatible with phrases like English ‘by itself’ (see e.g. Chierchia 1989/2004, Levin and Rappaport Hovav 1995, Reinhart 2000, Härtl 2003 among others.) Dispositional middles, on the other hand, are closer to passives in that they make refer-ence to an implicit external argument. However, the implicit agent is syntactically less active in middles. While middles, like passives, license instrumental phrases, control of implicit PRO-subject of adjunct clauses and do not allow ‘by itself’, they disallow the realization of the implicit external argument via a prepositional by-phrase and do not combine with agent-oriented adverbs or purpose clauses. Furthermore, middles are available only with a restricted set of verb classes and they are necessarily generic-dispositional, i.e. stative predications (Keyser & Roeper 1984, Fellbaum 1986, Hale & Keyser 1986, 1987, Fellbaum & Zribi-Hertz 1989, Condoravdi 1989, Ackema & Schoorlemmer 1995, 2005, Stroik 1992, 1999, Zribi-Hertz 1993, Rapoport 1999, Reinhart 2000, Steinbach 2002, Lekakou 2005, Marelj 2004 among many others). Finally, reflexives (and reciprocals) make reference to two theta roles (agent and pa-tient) which are both assigned to the same entity. Many languages thereby make a morphological difference between naturally reflexive/reciprocal/non-other directed verbs (as well as inherent reflexives) on the one hand and ordinary transitive verbs in a reflexive construal/other-directed verbs on the other hand (e.g. Haimann 1983, Kemmer 1993, Reinhart & Reuland 1993, König & Siemund 2000a, b, König & Vezzosi 2004, Smith 2004). For the latter class of verbs, English uses the anaphor himself in object position (5a) while the former class of verbs is typically in-transitive under a reflexive interpretation. The example in (5b) clearly involves two thematic

roles (an agent and a theme) which are both assigned to the subject referent; (5b) differs from other “derived” intransitives involving implicit objects such as ‘John ate’ where the theme is also not overtly expressed but understood as disjoint from the subject referent. (5) a. John hates himself b. John washed and shaved 3.3.1.2 The morphological realization of AS Voice across languages The different AS-Voices mentioned above can be semantically identified crosslinguistically. However, there is large variation concerning the morphological realization of these AS-Voices across languages as we will show in this section based on a small selection of languages. Beginning with the eventive passive, two main morphological strategies of passive for-mation can be found in the languages of the world (Siewirska 1984, Shibatani 1985, 1988 (ed.) Haspelmath 1990, Klaiman 1992, Andersen 1994, Kazenin 2001, Abraham 2006, to mention a few): (i) Analytic/periphrastic passives are formed with the help of a passive auxiliary and a passive participle as in the English example in (1b), repeated below in (6b). (ii) In synthetic passives, exemplified with Greek in (7b), a verbal suffix (tik) is attached directly to the verbal stem and is followed by other inflectional categories like aspect, tense and agreement marking. Alternatively, the suffix is attached outside of other inflectional categories as in the reflexive passive found in some Scandinavian, Slavic or Romance languages (see below) (See Haspelmath 1990 for other rare strategies to form synthetic passives). (6) a. John reads the book b. The book is read (by John) (analytic passive) (7) a. O Jianis anikse tin porta John opened the door b. I porta anixtik e (apo ton Jiani) (synthetic passive) the door open.NACT (by the John) I. AS-Voice/M-Voice interaction in languages with analytic passives The active/passive alternation is morphologically marked in all languages of the world that have a passive in the first place (Haspelmath 1990; but see Comrie 2008). This is clearly not the case with other AS-Voices which sometimes surface in the same M-Voice as ordinary active verbs. A case in point is English. The other three AS-Voices discussed are morphologically unmarked in English and involve plain (active) verb morphology (but see Brame 1983 for morphologically marked inherent reflexive predicates in English such as John prided himself). However, in lan-guages other than English, the morphological realization of anticausatives, reflexives and mid-dles shows certain interesting properties. German, for instance, differs from English in that it has a light reflexive pronoun ‘sich’ (a SE-Anaphor in the terminology of Reinhart & Reuland 1993) which is used to mark reflexive verbs, generic middles as well as (the majority of its) anticausative verbs:1 (8) a. Der Mann wäscht sich (reflexive) the man washes REFL b. Diese Art von Büchern verkauft sich immer gut (middle) this sort of books sells REFL always well 1 See e.g. Reis 1976, 1982, Höhle 1978, Haider 1985, Abraham 1986, 1987a, b, 1995, Haider & Bierwisch 1989, Grewendorf 1989, Primus 1989, Fagan 1992, Frey 1993, von Stechow 1996, Kaufmann 2001, 2003, Steinbach 2002, Gast & Haas 2008, among others.

c. Die Tür öffnet sich (anticausative) the door opens REFL Dutch is located between English and German. On the one hand, it forms a periphrastic passive akin to the one in German and English. Like German, it marks inherent reflexives and a (minority of its) anticausatives with the reflexive element ‘zich’. But standard Dutch, like Eng-lish, leaves its generic middles unmarked (e.g. Everaert 1986, Reinhart & Reuland 1993, Hoek-stra & Roberts 1993, Ackema & Schoorlemmer 1995, 2005, Geurts 2004, Marelj 2004, Lekakou 2005). (9) Dit sort boeken heft (*zich) altijd goed verkocht (middle) this sort books has (*REFL) always well sold To conclude, all three Germanic languages set their passive morphologically apart from the other three AS-Voices. English leaves all the other AS-Voices unmarked, Dutch marks two of them and German all three of them with a reflexive pronoun. The situation in the Romance languages as well as the Scandinavian and Slavic languages is quite similar to the situation in German.2 It seems that the languages that use reflexive morphology as an M-Voice marker share one important property. Specifically, this marker is morphologically light or simple. Languages differ, however, in the exact morpho-syntactic, i.e. phrase-structural status of their light/simplex reflexive element (Faltz 1985, Haspelmath 1987, 1990). This can be a light reflexive pronoun (Dutch, German), a reflexive clitic (Romance) or a verbal reflexive (Russian). While both in German and in Dutch the reflexive is a (light) pronoun, and hence an XP, (Everaert 1986, Koster 1987, Steinbach 2002, Steinbach & Gärtner 2000), the two languages differ in that the German reflexive can be focused without the addition of an intensifier, while the Dutch reflexive can only be focused with the help of an intensifier (e.g. Geurts 2004). In Romance (and many Slavic languages), the reflexive is a clitic, and hence a head. The fact that only light reflexive elements act as M-Voice markers becomes especially clear in languages with a so-called two-form reflexive system like Dutch (10) (see Kemmer 1993). These languages have in addition to SE-anaphors a so-called SELF-anaphor (in the termi-nology of Reinhart & Reuland 1993) which is morphologically heavy or complex (typically in-volving an intensifier). As in English, the heavy reflexive is used to form the reflexive use of other-directed verbs. However, non-other directed verbs are expressed with the light reflexive marker (10b,c): (10) a. Zij haat ??zich/zichself She hates REFL/REFL-SELF b. Jan schaamt zich/*zichzelf/*Marie John shames REFL/REFL-SELF/Mary ‘John is ashamed’ c. Jan waste zich/??zichself/Marie John washed REFL/ REFL-SELF/Mary ‘John washed (Mary)’

2 For Romance see e.g. Ruwet 1972, Napoli 1976, Grimshaw 1981, Burzio 1981, 1986, Cennamo 1993, Manzini 1986, Rosen 1988, Dobrovie-Sorin 1998, 2003, 2005, Zribi-Hertz 1982, 1987, Labelle 1990, 1992, Fellbaum & Zribi-Hertz 1989, Zubizarreta 1982, Centineo 1995, Cuervo 2003 among many others. For Slavic, see Babby & Brecht 1975, Babby 1975, Brecht & Levine 1984, Siewierska 1988, Rivero 2004, Rivero & Sheppard 2001, 2003, Geniušien÷ 1987. For Scandinavian, see e.g. Sigurðsson 1989, Åfarli 1992, Anderson 1990, Thráinsson 2007, Hellan 1988, Laanemets to appear.

The generalization is that only SE (and not SELF-anaphors) are used to mark AS-Voices (see also König & Siemund 2000a, b). Besides reflexive middles and reflexive anticausatives, some Romance (Slavic and Scandinavian) languages have a second passive Voice formed with reflexive morphology.3 This reflexive passive must be kept apart from the dispositional middle, because it has an episodic reading which is absent from dispositional middles which are derived statives. An example of the reflexive passive in French is given below. (11) is clearly eventive and hence cannot be analysed as a dispositional middle. Note that while the reflexive passive in French does not license the by-phrase, reflexive passives in Slavic languages do license by-phrases. Haspelmath (1990) suggests that the availability of a reflexive passive is correlated with a very light nature of the reflexive element as a clitic or an affix. At least the better studied languages that have a reflexive passive all have an analytic passive in addition. (11) Ce poème s’est lu hier pendant la fête (*par Marie) This poem REFL’is read yesterday during the party by Mary ‘The poem was read during the party yesterday’ (Dobrovie-Sorin 2005) II. AS-Voice/M-Voice interaction in languages with synthetic passives Turning to languages that mark the eventive passive with synthetic morphology, we ob-serve that they are also not uniform. At least two groups must be distinguished. In one group of languages, the passive is not morphologically set apart from the other AS-Voices. In Greek (12), for example, (a subset of) anticausative verbs (12c), reflexive verbs (12d) and dispositional mid-dles (12e) are formed with the same non-active (NACT) morphology used for the passive (see Tsimpli 1989, Alexiadou & Anagnostopoulou 2004, Zombolou 2004). Albanian is relatively similar to Greek (Rivero 1990, Massey 1992, Kallulli 2006, Manzini & al. 2009), but see section 3.4. (12) a. O Janis ekapse ti supa (active) the John-nom burnt-ACT the soup-acc ‘John burnt the soup’ b. To vivlio diavastike apo to Jani (passive) the book read-NACT by the John The book was read by John c. I supa kaike (anticausative) the soup-nom burnt-NACT-3sg ‘The soup burnt’ d. I Maria htenistike (reflexive) the Mary-nom combed-NACT-3sg ‘Mary combs (herself)’ e. Afto to vivlio diavazete efkola (middle) this the book-nom reads-NACT-3sg easily ‘This book reads easily’

3 In addition, some Romance and Slavic languages have impersonal constructions marked with a reflexive marker (Cinque 1988). These have to be kept apart from reflexive passives (D'Alessandro 2007). We do not include these constructions in this presentation for reasons of space, although they are obviously relevant for our investigation.

Note that non-inherently reflexive/other-directed verbs in Greek can appear in reflexive constructions when the verb is prefixed with the marker afto 'self' in addition to bearing NACT morphology (Rivero 1992, Embick 1998):4 (13) i Maria afto-katastrefete (Self-Reflexives)

the Mary-nom self-destroys-NACT ‘Mary destroys herself’ Anticausatives formed with NACT can be distinguished from passives (also formed with NACT) in Greek on the basis of different tests (AAS 2006a, b, and subsequent work). For example, like in the other languages only anticausatives are compatible with the phrase ton eaftos tu ‘by itself’.5 Middles can also be distinguished from passives by their generic/dispositional semantics. Reflexives are also set apart in that they license agentive adverbs referring to the sole argument DP. But importantly, such forms are in principle ambiguous between a passive and a non-passive interpretation, with the exception of examples like (13), which are clearly reflexives and the verbs which lack a passive reading altogether (12c) (see Alexiadou & Anagnostopoulou 2004, 2009 for details). In the second group of languages, the synthetic eventive passive is set apart from the other AS-Voices. For example, Hebrew distinguishes between morphological passive and mor-phological middle Voice (both are synthetic). But only the latter is used for anticausatives, re-flexives and dispositional middles (Doron 2003). III. Summary: Voice syncretisms Summarizing, the relation between AS-Voices and M-Voices is blurred by syncretisms in all languages; i.e. there is no language that uses a special M-Voice for each AS-Voice. This seems to be a universal property. A consequence of this is that languages typically use only one or two M-Voices in addition to the active Voice. Table I graphically illustrates an interesting property of Voice systems. While languages differ in whether they single out the passive morphologically or not, the other AS-Voices are ordered in a subset relation with respect to their morphological marking. To our knowledge, this synchronic subset dimension has not been discussed, let alone analyzed, in the syntactic discus-sion of Voice systems (but see Alexiadou & Doron 2007 for a first attempt). It has, however, been mentioned within the typological literature. For example, both Haspelmath (1987, 1990, 1993) and Geniušien÷ (1987) observe that there seems to be no language where reflexive and passive under exclusion of anticausatives form a syncretism. (Haspelmath explicitly excludes middles from this subset relation, a point which we do not subscribe to; see also below). The subset relation has also been acknowledged within diachronic analyses of Voice; especially for the well studied Romance (but also for Slavic) languages it was observed that reflexive clitics were used first to form reflexive verbs; in a next step, this morphological device shifted to anti-causatives and to generic middles and finally to passive uses (Lehmann 1982, Croft, Shyldkrot & Kemmer 1987, Haspelmath 1987, 1990, 1993, Cennamo 1993, Oesterreicher 1992, Michaelis 1997, Selig 1998, König & Siemund 2000a, b).6

4 There is a fourth case involving the Non-active form, namely deponent verbs which are not part of an alternation (Mackridge 1987, Embick 1998). 5 Most anticausatives formed with NACT morphology have no passive counterpart, while those anticausatives formed with ACT morphology do have a passive counterpart, a fact that can be diagnosed by the thematic role of by-phrase they allow (see AAS 2006, Alexiadou (to appear) for discussion). 6 In table 1 we put together the reflexive passive and the synthetic passive mainly for reasons of space and illustration. It is an open question whether we should analyze the reflexive element as a part of the verbal morphology in the same way as the NACT morphology in Greek.

active analytic

passive synthetic passive

dispositional middle

anticausative inherent reflexive

Greek/Albanian Act ∅ Nact Nact Nact Nact Hebrew Act ∅ Pass Middle Middle Middle Romance Act + Refl Refl Refl Refl German Act + ∅ Refl Refl Refl Dutch Act + ∅ Act Refl Refl English Act + ∅ Act Act Act Table 1: Voice-Syncretisms forming subsets across languages: A further question that, to our knowledge, has not been investigated in the syntactic lit-erature is whether it is a general property of languages with synthetic passive that they use the same NACT-device to mark further AS-voices (besides the active). The situation in Hebrew sug-gests that the subset relation holds for synthetic passives in the same way as it holds for reflexive passives. The historical development of Greek supports this idea as well (see e.g. Kaufmann 2001). 3.3.1.3 Non-canonical analytic passives, Affectedness and AS(Middle)-Voice As already mentioned, periphrastic passives are built on the basis of the combination copula + participle. Across languages, other constructions involve what from a morphological point of view seems to be the same participle found in passives, i.e. the participle is involved in syncre-tisms, too. These are referred to as non-canonical passives in the literature and typically involve auxiliaries other than be. For instance, in English, formally the same participle used in the even-tive passive also shows up in the so-called get-passive (Haegeman 1985). Other environments where the same form is found are adjectival passives (Wasow 1977, Levin & Rappaport 1986, Embick 2004a), and the present perfect: (14) a. The door was slowly opened (passive) b. The door got opened (get-passive) c. The door is already opened (adjectival passive) d. He has opened the door (perfect) A similar state of affairs holds in German. The very same participle II found in the even-tive passive shows up in the so-called bekommen-passive (or kriegen-passive) (15). Like in En-glish, the participle appears in the adjectival passive and the perfect (Rapp 1996, 1997, Kratzer 1996, 2000; for an exhaustive list, see Höhle 1978). For a recent discussion of the morpho-phonological properties of the German participle II, see Rathert (2009). (15) a. Er bekam das Obst (klein) geschnitten (bekommen-passive) He got thw fruits small cut b. Er bekam den Text ab-ge-schrieben He got the text copied

c. Er bekam das Buch ver-kauft He got the book sold Leaving the perfect and the adjectival participle aside, as these have been extensively discussed in the literature,7 we are mainly concerned here with the participial syncretism between the eventive verbal passive, and the non-canonical passives i.e. the English get-passive and the German bekommen-passive. The reason for this is twofold: i) such constructions raise the ques-tion of alternative realization of analytic passive Voice, and ii) a number of researchers view non-canonical passives as a form of middle Voice, since their subjects are interpreted as ‘af-fected’ (in the sense of Klaiman 1992, Kemmer 1993 described above, see e.g. Hatcher 1949, Givon & Yang 1994, Arrese 1999, McIntyre 2005)). For this reason, we call the get constructions and their cognates across languages affected participial constructions here. Let us consider the properties of such constructions. It has been observed that the bekom-men-passive which advances a dative to nominative and preserves the structural accusative is not possible with all dative verbs (Reis 1976, Haider 1984, 1986, 2001 Reis 1985, Wegener 1985, 1991, Fanselow 1987, Leirbukt 1997 Vogel & Steinbach 1998). Leaving aside monotransitive dative verbs, which form the bekommen-passive only for some speakers, the underlying generali-zation is that only verbs with the basic/unmarked word order dat >> acc can form bekommen-passives, while verbs with the basic/unmarked word order acc >> dat cannot (Czepluch 1988, Haider 1993, Molnárfi 1998, Fanselow 2000, McFadden 2004, Cook 2006),8 see (16-17); for detailed discussion, see Cook (2006): (16) a. Die Mutter schickt dem Jungen das Paket (basic order: dat >> acc) the.nom mother send the.dat boy the.acc parcel b. Der Junge kriegt/bekommt das Paket geschickt. the.nom boy gets the.acc parcel sent (17) a. Der Professor unterzog den Studenten der Prüfung (basic order: acc >> dat) the.nom professor subjected the.acc student the.dat test b. *Die Prüfung kriegte/bekam den Studenten unterzogen the.nom test gets the.acc student subjected As explicitly demonstrated in Cook (2006), the two basic word orders (dat >> acc and acc >> dat) do not go along with the animateness of the two object DPs (cf. Fanselow 2003, Heck 2000, Vogel & Steinbach 1998, Müller 1999), but they go along with a difference in conceptualization, which means that the high and low datives have different thematic roles. Low datives express locations or goals and irrespectively of their animateness they cannot be advanced in the bekom-men-passive. High datives, on the other hand, are affected objects, associated with the θ-roles Beneficiary or Maleficiary, and these can, again irrespectively of their animateness, be advanced in the bekommen-passive (see Leirbukt 1997 for a similar conclusion). Turning to English, it has been noted that the get-passive is not permitted with stative verbs and verbs that do not allow for the subject of the construction to be interpreted as affected, see Siewierska (1984). Similarly to German, the English get-passives describe events that are perceived to have a fortunate or unfortunate consequence on the subject. The same holds for the French construction se faire (Labelle 2002), as in (18). All these constructions are often com-

7 See e.g. Klein 1992, Kratzer 2000, Musan 2002, Rapp 1996, 1997, von Stechow 2002, Rathert 2006, Rothstein 2008, Alexiadou et al. 2003, Embick 2004a, McFadden & Alexiadou (to appear), Alexiadou & Anagnostopoulou 2008. 8 For the two classes of ditransitive verbs in German, see Czepluch 1988, Haider 1992, Haider 1993, Vogel & Stein-bach 1998, Müller 1999, Heck 2000, Fanselow 2000, 2003, Haider & Rosengren 2003, and the contributions in Hole et al. 2006, among others.

pared with the so-called adversity passive in Japanese, which is named after its affectedness se-mantics (e.g. McCawley 1972, Kuroda 1979, Miyagawa 1989, Shibatani 1994, Pylkkänen 2008). (18) Jean s’est fait écraser (par une voiture). Jean refl aux made run-over (by a car) ‘Jean was run over by a car.’ 3.3.1.4 Previous analyses Since the early days of the investigation of Voice, the different sub-phenomena of AS-Voice have always been taken up separately and studied as independent ones. Thus, some authors excusively deal with passive formation, while others are primarily concerned with a particular AS-Voice, e.g. reflexive, anticausative or dispositional middle or a particular M-Voice (reflexive morphology, NACT-morphology) and yet others are concerned with affected participial constructions, which are contrasted to eventive passives mainly. As our overview suggests, there are several contradicting analyses for each sub-phenomenon, but none that unifies the constructions we are interested in. In this are, we observe the familiar tension betwen syntactic and lexicalist analyses (but see Steinbach 2002 for a postsyntactic account of reflexive verbs, middles and anticausatives). Some studies (e.g. Kemmer 1993, Manney 2000) utilize cognitive grammar terminology to de-scribe the Voice relations at hand (Givón 1979, Talmy 1988, Langacker 1987), since this ap-proach has developed a system of diagrams which are useful in depicting the internal relations between different participants of an event. Within the Government and Binding (GB) framework, the standard analysis of the eventive passive is based around two principles (Baker 1988, Baker, Johnson & Roberts 1989 among many others):9 i) the passive morphology absorbs accusative Case and ii) the passive morphology absorbs the external theta-role. In the standard treatment, these two properties are related if we assume that the passive morphology is itself an argument that is assigned Case and receives the external theta-role. Recently, the specifics of this analysis have been challenged, see e.g. Embick (1997) and most prominently Collins (2005). Importantly, Kratzer (1996) develops a neo-Davidsonian association of external arguments in the syntax, according to which external arguments are not arguments of the verb, but are introduced by a VoiceP. On this view, the implicit argument of the passive has not been absorbed but it is introduced by a Voice-projection, similarly to the external argument in actives, albeit in an existentially bound fashion. Most lexical analyses assume that SE-reflexive elements, besides a basic anaphoric role, have a second use as valency reducing morphemes; they either block the projection of an argument into syntax or absorb a θ-role in the lexicon (Williams 1981, Grimshaw 1981, Wehrli 1986, Cinque 1988, Fellbaum & Zribi Hertz 1989, Fagan 1992, Ackema & Schoorlemmer 1995; cf. Dobrovrie-Sorin 1998). In order to derive the different AS-Voices, different types of reflexives are often assumed. Zubizarreta (1982, 1987) assumes three different versions of the French se: one is the ordinary reflexive anaphor, one is a verbal suffix deleting the external theta role (anticausatives) and a third one is the medium/middle se that blocks the assignment of the external theta role. Cinque (1988) assumes even five different si-morphemes for Italian. Wehrli (1986), on the other hand, assumes only one reflexive element which always absorbs either an internal or an external argument: in reflexives, the internal argument is absorbed, in anticausatives, the external argument is absorbed. The theta system developed by Reinhart (1996, 2000, 2002) and Reinhart & Siloni (2004) assumes different lexical operations that can apply to a verb’s underlying theta structure.

9 We deal with the issue of the representation of the implicit argument in section 3.4.

Saturation has the effect of existentially binding an argument; in passives, the external argument is saturated. Reduction eliminates theta-roles. Reduction of the external theta-role leads to anticausatives; reduction of the internal theta-role leads to reflexive verbs. Later work on the theta system (Reinhart & Siloni 2005) assumes that reflexives are derived by a process called theta-bundling. This process applies in some languages in the lexicon and in other languages in the syntax; its semantic effect is, however, always the same: it collapses two arguments (one necessarily the external argument) and, thereby, transforms a relation into a one place predicate which projects as an unergative verb. The different M-Voices either indicate that a lexical operation saturation or reduction/bundling has applied to a verb or are needed for formal reasons to avoid that the derived Voices crash at the interfaces after the syntactic derivation. In German, for example, the reflexive pronoun is inserted in reflexives and anticausatives to absorb the structural accusative case feature of the (basically transitive) verb. Kallulli (2006), adopting some key ingredients of Reinhart's (2002) theta system, namely the assumption that θ-roles are features, analyzes M-Voice as the result of the suppression of the first element in a thematic feature cluster. Such an account shares some of the problems of the theta-system. Recently, Doron & Rappaport Hovav (2007) criticized the proposal that θ-bundling can take place either in the lexicon or in the syntax (see also Labelle 2008). Specifically, they question the claim that bundling can take place in the syntax and they argue that most examples analyzed as syntactic bundling by Reinhart & Siloni (2005) are actually instances of ordinary anaphoric binding. Obviously, the claim that reflexive verbs are derived by anaphoric binding asks for an alternative explanation for the well-known facts that suggest that reflexive verbs are intransitive in French (Kayne 1975, Grimshaw 1981). The operation of Reduction has also been criticized in the literature for the main reason it has never been semantically defined; crucially, Doron (2003) argues that an operation eliminating a predicate’s semantic argument cannot be semantically formalized and Koontz-Garboden (2007) argues that natural languages in general do not have operations eliminating semantic arguments. Syntactic analyses of middles typically assume an unaccusative syntax similar to the one in passives. Since these analyses assume that the external argument is realized as a zero element (Stroik 1992, 1999, Hoekstra & Roberts 1993), they have problems to explain why the implicit argument of passives differs from the one of middles in being less active (see Ackema & Schoorlemmer (1995, 2005) for detailed discussion). A relatively well known syntactic theory of reflexive verbs is the unaccusative analysis, originally proposed in Marantz (1984), and taken up in Embick (2004b) and especially in McGinnis (1998, 2000). On this view, the reflexive element is merged in the external argument position, Spec,v and the full DP is merged in the internal argument position. That is, reflexive verbs are ordinary transitive verbs that involve two thematic arguments. The unaccusative aspect comes in with the following derivation: since the reflexive element needs an antecedent (Principle A of the Binding Theory (Chomsky 1981, 1986)), the internal argument must raise to a position c-commanding the external argument position. This syntactic movement is similar to the movement found with one-place unaccusatives and it is responsible for the unaccusative properties shown by reflexive verbs in Romance languages (participle agreement, auxiliary selection). The proposal by McGinnis faces a number of severe theoretical problems, most specifically, it is not in accordance with phase theory (Chomsky 2001, seq., Marantz 2006). Syntactic accounts of reflexive anticausatives and dispositional middles (Haider 1985, Everaert 1986) often assume that the reflexive is an A’-element. For a detailed discussion and criticism of Haider’s account see Fagan (1992) and especially Steinbach (2002). One central observation is that the German reflexive element in anticausatives, dispositional middles as well as reflexive verbs has the same free distribution as ordinary NPs, i.e. it does not have to be

adjacent to C0 nor to the lexical verb. Instead, it freely scrambles. Furthermore, it is clearly case marked. This behavior is hardly expected if the reflexive in anticausatives and middles occupies an A’-bar position. Instead, most authors discussing German reflexive anticausatives and middles come to the conclusion that they are syntactically transitive even though they only involve one theta-role (Fagan 1992, Steinbach 2002, Haider & Bierwisch 1989, von Stechow 1996, Wunder-lich 1997, Haider 2000). As is clear thus far, researchers working on reflexives as M-Voice markers have nothing to say about languages that use non-active morphology. The reverse also holds: researchers working on non-active morphology have nothing to say about reflexive morphology. For instance, Embick (1998, 2004b) proposes that the morphological realization of Voice is condi-tioned by the following requirement: the non-projection of the external argument is sufficient to give a NACT-form. In this sense the morphological realization does not coincide with the ex-pression of agentivity. Turning to the affected participial constructions, these do not figure prominently in the most recent syntactic literature. (For some discussion on the similarity between these constructions and eventive passives, their proper analysis in terms of raising or control and the status of the auxiliary verb see Haegeman 1985, Fox & Grodzinsky 1998, Huang 1999 for Eng-lish, Labelle 2002 for French and the Reis vs. Haider debate for German). 3.3.3 Preliminary work by this research group Our previous work focused on the causative alternation. We developed criteria to distinguish the properties of anticausatives as opposed to passives and we identified the semantic and syntactic properties of the two morphological classes of anticausatives in German and Greek. Furthermore, we offered an explanation of the gaps in the distribution of M-Voice in the case of anticausatives in Greek and German. These results as well as our results from our work in nominalizations and participles will be presented in a book for Oxford University Press (External arguments in tran-sitivity alternations: a layering approach authored by Artemis Alexiadou, Elena Anagnostopou-lou and Florian Schäfer). Alexiadou & Anagnostopoulou (2004) investigated anticausatives in Greek and con-cluded that the two morphological types of anticausatives (active vs. non-active) are structurally different. Alexiadou, Anagnostopoulou & Schäfer (2006a, b) and Alexiadou & Schäfer (2006)develop a syntactic theory of the causative alternation. The central claim is that both cau-satives and anticausatives are built from a common root plus a verbalizer; they differ only in the presence vs. absence of a Voice projection. This means that causatives and anticausatives do not differ in their event decomposition but only in the presence vs. absence of a canonical external argument. However, against common assumptions in the literature, anticausatives crosslinguisti-cally can make reference to an external argument which is necessarily a causer (a causative event or a natural force, (19a-b). (19) a. The vase broke from the pressure b. Die Tür öffnete sich durch den Windstoß the door opened REFL through the gust-of-wind As argued by Schäfer (2008c, 2009d), partly on the basis of data like the one in (20a,b), this im-plicit causer argument cannot be lexically encoded, but is licensed in the context of a bi-eventive, resultative event structure. (20) a. Der Ball rollte (*durch den Wind) The ball rolled throughcauser the wind b. Der Ball rollte (durch den Wind) über die Torlinie

The ball rolled throughcauser the wind across the goal-line Building on AAS (2006), the investigation of the causative alternation was extended in the domain of aphasia (Stavrakaki, Alexiadou, Kambanaros, Bostantjopoulou & Katsarou sub-mitted) and in the domain of language acquisition (Zombolou, Varlokosta, Alexiadou & Anag-nostopoulou 2009). Alexiadou (to appear b) is concerned with morphological patterns in the causative alter-nation, and argues that the morphology we see in the alternation helps us explain why anticausa-tive and causative formation is more productive in some languages as opposed to others. The crosslinguistic variation relates to the pieces languages have at hand in order to realize Voice and v heads, the two heads that are involved in the formation of causative verbs. Alexiadou (to ap-pear a) is concerned with postverbal nominatives as a diagnostic for unaccusativity. Alexiadou (2005) offers a syntactic analysis of the participles in the get-passive arguing that this is actually a resultative participle. Alexiadou & Anagnostopoulou (2008) discuss crosslinguistic patterns in the formation of stative and resultative participles. On the basis of Greek and Hebrew, Alexiadou & Doron (2007) tried to offer a first ap-proximation that there are indeed two non-active voices, passive and middle, both preventing the insertion of an external argument, but with different properties. Finally, Alexiadou (submitted) offers a survey of the general relation between syntax and the lexicon. Schäfer (2002) gives an overview of different syntactic accounts to the binding theory with an emphasis on theories that abandon the independent Principles B (and C) and derive their effects from general economy principles on bound elements. Schäfer (2003) is specifically con-cerned with the implementation of locality constraints in such a framework. Building on the theory of the causative alternation developed by AAS (2006), Schäfer (2007, 2008a, to appear) develops a (post-)syntactic theory of anticausative Voice-morphology, with special emphasis on the reflexive morphology found in Indo-European languages. The over-all claim is that anticausative morphology reflects a way to avoid the merging of a thematic ex-ternal argument. Schäfer's work is concerned with the question why dispositional middles are crosslinguis-tically marked with the same device as anticausatives in reflexive languages. Schäfer (2007, 2008a, b) argues that reflexive anticausatives and middles have exactly the same morphosyntax; they differ only in the presence vs. absence of a generic, dispositional operator in the sense of Lekakou (2005). Furthermore, this work explicitly states that if a language has marked anticausa-tives, then the same morphological device is used to form middles in this language; the rationale is that middles are newly coined anticausatives in a generic context. Dutch, which has reflexive anticausative but unmarked middles, is not an exception to this claim because the reflexive mark-ing of anticausatives is no longer a productive strategy in this language: for example, it is not used for newly coined anticausatives. Heidinger & Schäfer (2008) investigate how the distribution and the semantic properties of French par-phrases introducing external arguments in the context of the reflexive passive and anticausative developed from the 16th century to present-day French. They argue that the the-matic role of the element introduced by the preposition provides a strong clue for the categoriza-tion of the construction as a passive or an anticausative. Schäfer (2007, 2008a, 2009b) develops a theory of free datives in the context of transitive and intransitive change-of-state verbs. These datives denote either the beneficiary/maleficiary of the change-of-state event or, in the case of anticausatives, the oblique causer of the event; i.e. thematically; these are exactly the kind of datives that can be advanced in the bekommen/get-passive. Schäfer (2009a) provides a survey on the theoretical questions that the causative alterna-tion poses as well as on the different types of analyses that have been proposed. It makes specific

reference to recent work and explicitly discusses the question of morphological syncretisms found in this area. We have recently been invited to contribute to the following projects: i) Alexiadou: a paper for a volume on the Art of Getting: GET verbs in European languages from a synchronic and diachronic point of view, edited by A. Lenz and G. Rawoens, ii) Alexiadou: a paper for a special volume of Linguistics on the diachrony of labile verbs edited by N. Lavidas and L. Kuli-kov and iii) Alexiadou, Anagnostopoulou and Schäfer an overview paper for the project men-tioned in (ii). 3.3.4 List of relevant publications

• Peer-reviewed publications

Alexiadou, A. To appear a. Post-verbal nominatives: an unaccusativity diagnostic under scrutiny. In R. Folli, C. Ulbrich (eds.), Interfaces in Linguistics: New Research Perspectives. Oxford University Press, Oxford.

Alexiadou, A. To appear b. On the morpho-syntax of (anti-)causative verbs. In M. Rappaport Hovav, E. Doron, I. Sichel (eds.), Syntax, Lexical Semantics and Event Structure. Oxford University Press, Oxford.

Alexiadou, A. Anagnostopoulou, E. 2004. Voice morphology in the causative-inchoative Alter-nation: evidence for a non unified structural analysis of unaccusatives. In: A. Alexiadou, E. Anagnostopoulou, M. Everaert (eds.), The Unaccusativity puzzle. Oxford University Press, Oxford, 114-136.

Alexiadou A., Anagnostopoulou, E., Schäfer, F. 2006a. The properties of anticausatives crosslinguistically. In: M. Frascarelli (ed.), Phases of Interpretation. Berlin: Mouton, 187-211.

Alexiadou A., Anagnostopoulou, E., Schäfer, F. 2006b. The fine structure of (anti-)causatives. In: C. Davis, A.-R. Deal, Y. Zabbal (eds.), Proceedings of NELS 36, UMASS, Amherst.

Heidinger S., Schäfer, F. 2008. On the French reflexive passive and anticausative. A diachronic view from the par-phrase. In: B. Fagard, S. Prevost, B. Combettes, O. Bertrand (eds.), Etudes de linguistique diachronique. Peter Lang, Bern.

Schäfer F. to appear. Local Case, Cyclic Agree and the Syntax of truly Ergative Verbs. In Local Modelling of Non-Local Dependencies in Syntax, A. Alexiadou, T. Kiss, G. Müller (eds.), Linguistische Arbeiten. Niemeyer, Tübingen.

Schäfer F. 2009a. The Causative Alternation. In Language and Linguistics Compass, Vol. 3. Schäfer F. 2008a. The Syntax of (Anti-)Causatives. External arguments in change-of-state con-

texts. John Benjamins, Amsterdam. Schäfer, F. 2003. Relativierte Bindungsminimalität in OT: einfache vs. serielle Optimierung. In:

L. Gunkel, G. Müller,G. Zifonun (eds.), Arbeiten zur Reflexivierung. Niemeyer, Tübingen.

• Submitted manuscripts Alexiadou, A. 2009. Syntax and the Lexicon. In Syntax, 2nd Edition: an international Handbook.

Mouton de Gruyter, Berlin. Schäfer, F. 2009d. Two types of external argument licensing - the case of causers. Submitted to

Studia Linguistica. Stavrakaki, S., A. Alexiadou, M. Kambanaros, S. Bostantjopoulou, Z. Katsarou. The production

and comprehension of verbs with alternating transitivity by patients with nonfluent. Submitted to Journal of Neurolinguistics.

• Non peer-reviewed publications

Alexiadou, A. 2005. A note on non-canonical passives: the case of the get passive. In: H. Broek-huis & al. (eds.), Festschrift für Henk van Riemsdijk. Mouton de Gruyter, Berlin.

Alexiadou, A., Anagnostopoulou, E. 2009. Agent, causer and instrument PPs in Greek: implica-tions for verbal structure. MIT Working Papers in Linguistics 57: 1-16.

Alexiadou, A., Anagnostopoulou, E. 2008. Structuring participles. Proceedings of WCCFL 26, 33-41.

Alexiadou A, Anagnostopoulou, E., Schäfer, F. 2006b. The fine structure of (anti-)causatives. In C. Davis, A.-R. Deal, Y. Zabbal (eds.), Proceedings of NELS 36, UMASS, Amherst.

Alexiadou A., Schäfer, F. 2006. Instrument Subjects Are Agents or Causers. In: D. Baumer, D. Montero, M. Scanlon (eds.), Proceedings of the 25th West Coast Conference on Formal Lin-guistics, 40-48. Cascadilla Proceedings Project, Somerville, MA.

Schäfer F. 2009b. The Oblique Causer Construction across Languages. In: A. Schardl, M. Walkow, M. Abdurrahman (eds.), Proceedings of NELS 38. GLSA, Amherst, MA.

Schäfer F. 2008b. Middles as Voiced Anticausatives. In: E. Efner, M. Walkow (eds.), Proceedings of NELS 37. GLSA, Amherst, MA.

Schäfer F. 2007. On the nature of anticausative morphology: External arguments in change-of-state contexts. Doctoral Dissertation, University of Stuttgart.

Schäfer, F. 2002. Aspekte einer wettbewerbsgetriebenen Bindungstheorie. Unpublished Diplom-Thesis, University of Potsdam.

Patents: does not apply

• Presenations Alexiadou, A., Doron, E. 2007. The syntactic construction of two non-active voices: passive and

middle. GLOW XXX Workshop on Global selective comparison. Tromsø. Schäfer, F. 2008c. Two types of external argument licensing. The case of causers. GLOW 31,

Main Session. Newcastle. Schäfer, F 2009c. The Passive of Reflexive Verbs. Talk at the Grimm seminar, Lunds Universitet Zombolou, K., S. Varlokosta, A. Alexiadou, E. Anagnostopoulou. Acquiring passives vs. anti-

causatives in Greek. 2009. BU Conference on Language Development 34.

3.4 Project description (goals, methods, work program) While Voice phenomena have been studied from a variety of perspectives, we observe that this has been done in a very fragmentary manner. Specifically, previous researchers focus on certain sub-aspects of the phenomenon or on the description of individual languages or phenomena. In addition, researchers either focus on reflexive morphology or non-active morphology, but rarely consider both in parallel. We think that the lack of a unified treatment of reflexive and non-active morphology is a considerable drawback of the theory of grammar. Furthermore, most previous research has not attempted to unify the treatment of what we call here affected participial constructions with their canonical cognates. But importantly, for an accurate characterization of Voice phenomena such constructions are of importance in order: a) to understand the difference between canonical and non-canonical passives; b) to understand the syntactico-emantic properties the affected constructions are subject to; c) to formally define Affectedness in a principled manner and ultimately d) to answer the question whether middle Voice has its own right in the grammar. A second important drawback is the fact that the question of interaction between AS and M-Voice is largely ignored in most recent syntactic literature assuming a decompositional approach to verbal structure, see e.g. Borer (2005), Ramchand (2008). Especially in a system such as Ramchand's the only way to treat M-Voice would be to assume multiple lexical enrties

for verbs, thus leading to an explosion of the lexicon, which should be avoided (see Reinhart 2000). In addition to these general issues, further specific problems for previous accounts can be identified. For instance, in lexicalist accounts it is not clear why different lexical operations should always be marked by the same morphological device across languages. Moreover, such accounts give no answer to the question why the same lexical operations are marked across languages and offer no explanation concerning the difference and/or similarity between the reflexive device and the non-active/middle device. Thereby different accounts are often not compatible with each other. Another problematic issue concerns the assumption that different morphemes exist with exactly the same morpho-syntax. This seems to us to be a weakness that should be avoided if possible. As mentioned, there are clear typological and historical indications that the formation of reflexive verbs is the “basic” AS-Voice in languages with SE-anaphor while all other AS-Voice marked with SE-anaphors come in later. We think that all other uses of SE-anaphors must be related to this basic use. Either there exists only one element SE that brings about different AS-Voices because it combines with different syntactic structures or there are indeed different SE-elements. However, in the latter case there must be similarities in the featural make-up between the different SE-elements that motivate why they are spelled-out in an identical way (see Burzio 1991 for an account along these lines explaining the syncretism between reflexive si and impersonal si in Italian). We note also that while it is quite uncontroversial that a heavy reflexive realizes an inter-nal argument of a (di-)transitive predicate and shares its referent with a different (typically the external) argument of this predicate, concerning the class of reflexive verbs involving light mor-phological marking quite different theoretical analyses have been put forward and there is no agreement as to whether the reflexive element is an argument of the verb or not, as well as whether these predicates, if intransitive, are better analyzed as unergative or unaccusative predi-cates. Finally, and most importantly, previous accounts (lexicalist as well as syntactic ones) have nothing to say about the sub-set relationships summarized in table 1.

3.4.1 Questions The main research topic of this project is the manner in which syntactic Voice alternations relate to Voice morphology, and the implications of this for theories of Voice and theories of morphol-ogy/syntax interactions. In this project we assume that a broader typological perspective can guide our deeper theoretical understanding of individual Voices and the relations among different Voices and lead us to a formal characterization of the phenomenon of Voice. In this project, we take the phenomena of Voice to be a good testing ground for the gen-eral discussion on underspecification vs. ambiguity in the grammar. In particular, the phenomena and constructions we are interested in will shed light on the following questions: i) does the grammar produce an identical morpho-syntax that is subject to different readings at the interpret-ative component or ii) does it produce different morpho-syntaxes and the elements used in a par-ticular language to realize them are underspecified, i.e. are they sensitive to one (or a limited amount) of features that are shared by these constructions? The task is then to find out which features underlie the distribution. The questions of our project can be identified as follows: (i) How can we offer a formal characterization of Voice that captures its morphological, syntactic and semantic properties as well as its cross-linguistic variability? (ii) In connection to (i), we would like to investigate the Number of Voices available across lan-guages and whether the view that considers the middle voice as distinct from the passive Voice can indeed be established, and, if so, to find the proper characterization of the middle Voice.

(iii) Syncretisms: which is the appropriate level of grammar where syncretism should be dealt with: the core grammar or the interface? (iv) What is the nature of reflexive and non-active marking and can they be seen as comparable to one another? (v) What is the relationship between semantics and morphology, given our view on the place of morphology within grammar? (vi) Why do gaps in the distribution of a particular M-Voice exist? (vii) Formation of subsets: Why do they exist and why do languages select different cut-off points? (viii) Which of our constructions contain implicit arguments in the syntax? If so, how are these represented in the grammar? (ix) What is the status of light verbs in non-canonical passives? (x) What is the role of Affectedness in grammar and in the characterization of Voice apart? Can a formal characterization be offered? (xi) How do the participles in non-canonical and canonical passives differ from another within a language and across languages?

3.4.2 Goals The central aim of this project is to offer an integrated theory of Voice, by accounting for the syntactic, semantic and morphological properties of different Voice systems. We will do this by first formulating generalizations about M- and AS-Voice in the formal framework we are com-mitted to (Theory of Principles and Parameters and Distributed Morphology), as it offers a rigor-ous formalism within which explanations can be given. Specifically, we will advance an analysis that treats voice morphology as being sensitive to the output of the syntax. Our approach advocates a direct interaction between interfaces: the morphology realizes the features of the syntax directly without any intermediate levels or features added after the syntactic derivation. Our results are expected to be significant theoretically as well as empirically. On the theo-retical side, we hope to show how different interpretations of voice come about from the various structural combinations of the different heads that constitute the building blocks of verbal mean-ing. On the empirical side, our analysis will seek to identify possible and impossible combina-tions within individual languages, and to explain why we find certain generalizations cross-linguistically. Taking as a starting point the observation that languages vary with respect to the cutoff points for assigning M-morphology to the different derivations that form the core of our research proposal, we want to understand why specific options are chosen, others are blocked, and how morphology interacts with the constructions which appear in the languages under dis-cussion. A second theoretical aim is to contribute to the theory of Underspecification in Morphol-ogy and Syntax. While in most work within Distributed Morphology (DM), underspecification is made use of to systematically express syncretisms, in the recent syntactic literature, we observe a conceptual move against underspecification, see Kayne (2008), Caha (2008) and Manzini & al. (2009); especially Manzini & al. favor an ambiguity view. Since, however, different morpho-syntaxes can be systematically teased apart for the different AS-Voices, we cannot see how such an ambiguity account can be maintained. Thus, we assume that in order to understand the nature of syncretisms, morphological elements may be underspecified with respect to the syntactic envi-ronment in which they appear. ‘Underspecification’ in this sense refers to the properties of pho-nological exponents (M-Voice markers here) with respect to the syntactico-semantic environ-ments in which these exponents will be inserted. The underspecification of phonology with re-spect to morphosyntax or syntax/semantics in this manner allows for the pervasive patterns of

syncretism found in natural language to be captured systematically, a point which has figured prominently in the critique of Lexicalist approaches to morphology. 3.4.3 Methods and work program Methods In this project, we will build on work carried out in Stuttgart and in co-operation with our col-leagues in Greece, Israel and the USA within the framework of DM, a piece-based, realizational theory of morphology. Specifically, we will adopt a syntactic approach to word structure (Hale and Keyser 1993, Halle & Marantz 1993, Kratzer 1994, 1996, Marantz 1997, Embick 2000, Alexiadou 2001, Embick 2003, Doron 2003, Harley & Noyer 2003, Alexiadou, Anagnostopou-lou & Schäfer 2006, and subsequent work). The architecture of the model of grammar that we adopt here is illustrated in (21). The syntax consists of a set of rules that generate syntactic structures, which are then subjected to further operations in the course of the derivation of the PF and LF interface levels: (21) The Grammar Syntactic Derivation (Spell Out) 3

PF LF Morphological adjustment Every word is formed by syntactic operations (Merge, Move). The principles of morphology are therefore to a large extent the principles of syntax. In the default case, the morphological struc-ture at PF simply is the syntactic structure. In more complex cases, which are in no way excep-tional, some further operations apply at PF to modify the syntactic structure (morphological ad-justment). The units that are subject to the syntactic operations Move and Merge are the mor-phemes. On the widely-held view according to which syntactic structures are hierarchical tree structures, the morphemes are the terminals of such trees. There are two types of morphemes: i) the Roots, items like √CAT or √SIT, make up the members of the “open class” and ii) the abstract morphemes– such as [pl] or [past]– are the (con-tents of the) familiar functional categories of syntactic theory. In this model, all word formation is syntactic, i.e. involves syntactic operations on Roots and abstract morphemes. Word (and thus verbal too) meaning comes about via the combination of roots and func-tional elements. In particular, we assume that roots become verbal via v (which introduces an eventuality) and external arguments are introduced by VoiceP, Kratzer (1996). On this view, the same head that introduces a DP in the active licenses an implicit argument (or a PP) in the pas-sive Voice. Unlike Kratzer, however, we assume that no (not even the internal one) argument is specified in a lexical entry, as these do not exist as primitives. That is, we assume a configura-tional theta-theory (as in Hale & Keyser 1993) and together with project B1 we will develop a model for the introduction of internal arguments (but cf. B4/B5). (22) [VoiceP [vP [Root ]]

For the particular concerns of this project, a piece-based approach allows us to investigate M-Voice from the perspective of available units to encode it. If we compare English, German, Ro-mance and Greek, we see that the languages have different units in their Vocabulary to encode M-Voice. Only German (and Romance) have a SE-anaphor and use it productively as an M-Voice marker. Greek and English both lack a SE-anaphor, but they fill this gap in different strategies. Greek has non-active morphology available which it uses extensively for Voice mark-ing. English lacking both SE-anaphors and non-active morphology has to fill the gap by active morphology. The question is whether this is the right direction to look at the differences among languages and of course how the pieces in (22) are distributed in the case of the participial con-structions. In view of the complexity of the enterprise, our approach is by definition comparative in nature. We will deal with languages we are familiar with and for which detailed descriptions exist in order to be able to make sense of the subset relationships. In view of our expertise and the general expertise present in Stuttgart, and that of our colleagues abroad, we will consider and compare Germanic languages (English, German and Dutch), with Romance (Italian, French, Spanish and Romanian), and Greek and Hebrew primarily. These languages provide us with the core morphological properties we are interested in in the domain of realization of AS-Voices (unmarked, reflexively marked (both XP and clitic) and non-active) and further provide an inter-esting split when it comes to the affected participial constructions, namely Greek lacks them al-together in spite of having both light verbs and participles of the right type. Scandinavian and Slavic languages might eventually be used as control cases. As we show below, Albanian might also turn out to be of special interest.

Work program I Characterization of Voices As already stated, we aim at providing a formal characterization of the different Voices that we described in section 3.3. To this end, we need to answer the following questions: what are the relevant categories to fully describe Voice? Are passive and middle the right terms/enough? What about the term medio-passive, a label often used in formal and typological studies alike? Importantly, how can we define the different Voices semantically and syntactically? Doron (2003) and Alexiadou & Doron (2007), following Kemmer (1993), suggested that passive and middle are two instantiations of Nonactive Voice. Thus the main opposition is be-tween active and non-active. These two non-active Voices are alike in that they both prevent the insertion of the external argument. They both derive intransitive verbs, but the crucial difference between the two voices is that passive applies to verbs and is found only if the active exists, whereas middle applies to roots, so that the existence of a middle verb does not depend on an active verb. But is this the correct characterization, in view of the diversity in the interpretation and the syntactic behavior of the constructions sub-sumed under Middle? And how does it gen-eralize to reflexive verbs? Can they uniformly be sub-sumed under Middle? Kratzer's Voice hypothesis suggests that the Voice projection reflects the active vs. non-active contrast. But what about the Middle Voice? Does this also involve the presence of a Voice head? But then this Voice head must be clearly different from the passive Voice head, and should be sensitive to different features across the other instantiations of Middle Voice. And is this Voice head also present in the affected participial constructions? II Number of Voices Do all languages have access to all Voices or are there languages that have more semantic Voices than others? Reflexive languages have reflexives, anticausatives and middles. But they differ as to the presence vs. absence of reflexive passives, e.g. Spanish vs. German. German does not have

a reflexive passive. Is the Spanish reflexive passive identical in use to the periphrastic passive? If not, does this mean that German has a gap or are the functions of the reflexive passive taken over by the periphrastic passive? Non-active languages have non-active morphology in the above environments. But do they have other AS-Voices in addition to passive? For instance, Lekakou (2005) argued that the dispositional middle in Greek is not a discrete Voice-category but a dispositional passive, as it licenses by-phrases and is available with a much bigger class of verbs than in other languages. This brings about the general productivity question as well: why are passives more pro-ductive in some languages (English/German/Romance) but not in others (Greek)? III Explaining Voice syncretisms How can we explain the syncretism between semantically different Voices? Are they syntactically encoded (DM) or do they emerge at the conceptual interface (Manzini & al. 2009). Let us briefly consider the two alternatives. Beginning with the latter, in their discussion of Voice syncretisms in Albanian and Greek, Manzini & al. state that single lexical items may be open to a number of interpretations (syncretism), in the same way that the same morphosyntactic structure may be open to them (passive, reflexive, anticausative or even impersonal in the case of middle-passive) and vice versa: the same interpretation may map to different morphosyntaxes, as is the case in Albanian that shows different realizations across tenses. On this view, syncretism does not reflect a mismatch between syntax and morphology, but a lack of non-isomorphism between morphosyntactic structure and interpretation. In a DM approach, a single lexical item can be realized in a number of syntactico-semantic environments. This is so because the item in question is 'underspecified' for the domain of insertion and because there is a common property that all these environments share, see e.g. Embick's work on Greek Voice. We think that Manzini & al. overlook an important insight: namely, the AS-voices never share all syntactic-semantic properties, suggesting that they differ syntactically. In other words, they minimally differ in a number of respects described for some of the AS-voices most forceful-ly in AAS (2006) and Doron (2003). If this is so, the interface approach cannot be maintained and pursued. Rather the interfaces 'receive' a transparent syntactic structure and what remains under-determined is the morphological realization. In this project, we want to systematically describe the complete set of AS-Voices and propose a syntactic theory of morphological encoding. We will spell this out in more detail be-low. IV Reflexive marking vs. non-active marking. As already mentioned, reflexive marking and non-active marking are dealt with separately in the literature, with few notable exceptions (e.g. the work by Reinhart 2000, 2002, Reinhart & Siloni 2005). This naturally raises the question whether they are identical or totally different? If they are identical, why do they have distinct realizations? If they are different, how do they differ and why are they used for the same task? Perhaps the behavior of Albanian is illuminating: in this lan-guage, the morphological realization of different AS-Voices can take the form of a specialized inflection, a clitic (u) associated with the ordinary 3rd person active forms of the verb, or a spe-cialized auxiliary (jam, i.e. 'I am', as opposed to kam 'I have' in the active followed by a parti-ciple). The fact that e.g. anticausatives, reflexives and passives can alternate between the three forms depending on tense (and mood) enables the formulation of the hypothesis that the three realizations serve the same purpose. As we already mentioned, reflexive languages differ as to the phrasal status of the reflex-ives, e.g. XP (Germanic) vs. clitic (Romance), which influences the auxiliary selection patterns (have in Germanic vs. be in Romance). But even within Romance, we find considerable differ-

ences. For instance, Kayne's (1975) tests which suggest that reflexive verbs are intransitive do not apply the same way in Italian as they do in French. What is the reason for this difference, and how does the status of the reflexive influence auxiliary selection and participial agreement? Embick (1998, 2004b) argued that reflexive and non-active M-Voices have in common aspects of an unaccusative syntax. This view is challenged by the following observation. In Ger-man and Icelandic reflexive verbs can build a passive under preservation of the reflexive pronoun (see Abraham 1986, Sells, Zeanen & Zec 1987, Plank 1993, Vater 1995, Ágel 1997, Maling & Sigurjónsdóttir 2002, Eythórsson 2008, Jónsson 2009, Sigurðsson to appear, Schäfer 2009c). (23) a. Heute kann sich über Sonne gefreut werden Today can REFL about sun rejoiced be b. Zuerst wird sich geküsst, dann wird geheiratet Fist is REFL kissed, afterwards is married The observation that reflexive verbs passivize in some languages makes the unaccusative analysis for at least these languages untenable; if we think that AS-Voices are syntactically uniform across languages (the null hypothesis, as Reinhart & Siloni (2005), in our view correctly, point out), the unaccusative analysis has to be refuted for other languages, too. Furthermore, under the proposal that the reflexive pronoun in German acts as a case absorber in the derivation of reflexive verbs, we would expect that the reflexive pronoun disappears in passives because structural accusative case should already be absorbed by passive formation thereby making the reflexive superfluous. Data such as the ones in (23) raise new questions. First of all, which reflexive verbs can form a passive? Some preliminary research seems to suggest that the construction is limited to inherent and naturally reflexive verbs both in Icelanding and German. Second, why does this not appear in other languages such as Dutch or Norwegian even though both languages form their reflexive verbs with a SE-anaphor? Finally, consider once again the Greek data in (13). Embick took such examples as a strong argument that non-active morphology is incapable of reflexivizing any type of verb. Verbs that are reflexive with non-active only are verbs which are capable of being ‘inherently reflex-ive’, which never appear with afto-. This is supported by the observation that nominals, which clearly lack non-active morphology prefixed with afto- receive a reflexive interpretation ('afto-travmatismos' self-hitting). But practically no research exists on the nature of this element. We can safely conclude that this is not an intensifier, as Greek uses another element in this function (o idios 'the same'). But how exactly does afto- interact with non-active marking, and in what relationship does it stand to the syntactic projection Voice? V Relationship between semantics and morphology A central question of this project concerns the relationship between semantics and morphology. The fact that we find syncretisms suggests that the constructions in question have something in common. Otherwise we would not find these systematic patterns. Standardly, morphological marking is a reflex of an operation on AS (e.g. Williams 1981 and much subsequent work). But why should different AS operations be realized with the same marking? Why should we find the subset relation? Lexical treatments have no answer to these questions. Lexicalist accounts in fact say nothing about the specific type of morphology. For instance, what is it that qualifies a reflex-ive pronoun as an M-Voice marker? Other approaches such as e.g. Manzini & al. deal with this problem as an issue of ambiguity. Working within a piece-based approach, such as DM, we are forced to answer this question. Specifically, within DM a particular exponent has to be ranked in terms of 'specification' for its context of insertion, as it competes with other items. Thus the fact that distinct syntactic environments opt for the same exponent suggests that first there is a link

between them, and second this particular item is flexible enough to be inserted. But what are the features that this marking is sensitive to? These will emerge once we are able to provide formal characterizations of the different Voices. Related to that is the following issue: Can we define the semantic and/or syntactic proper-ties shared by these constructions in a systematic way? Within a realizational framework such as DM, two options are in principle possible. One option is to assume that different Voices share syntactic substructures, while they differ in other syntactic properties, leading to a common mor-phological marking (as in Embick 1998, 2004b). Alternatively, the syncretisms could be seen as the spell-out of a syntactic item, a head or an XP that takes part in the syntax of the different Voices. Which of the two can account for the phenomena at hand in a satisfactory manner? We note here that Embick's approach faces problems with the analysis of passives if we assume that these do contain an implicit argument. If so, this must be projected in the syntax (see below) and cannot be reduced to features on a functional head. VI Gaps in the paradigm Why do we find gaps in the distribution of morphology only with anticausatives but not with middles and reflexives? In other words, anticausatives in languages other than English are di-vided into two main categories, those that carry morphological marking and those that surface with active M-Voice. Such gaps are not found in the other AS-Voices. Does this suggest that the marking of middles and reflexives is indeed a process on AS, while the marking on anticausa-tives is simply expletive and, therefore, more idiosyncratic? This has consequences for the cor-rect characterization of the features the morphological realization is sensitive to and the number of Voice patterns to be assumed. VII Subset relationships A puzzling fact about the syncretisms is the subset relation. Why do we find this? The morpho-logical subset property suggests that we have a semantic and /or syntactic subset. How can we account for this? To our knowledge, no theoretical account of the phenomenon exists. Are the cut-off points related to the type of marking? As mentioned above, Alexiadou & Doron (2007) suggested that anticausatives, reflexives and dispositional middles instantiate middle Voice which is seen in opposition to the passive Voice. By comparing English to Greek and Hebrew, they pointed out that in Greek, there is no morphological distinction, hence a unique NACT morphology is used for all AS derivations, though on the basis of tests such as agentive by-phrase, causer PP and by-itself modification, they behave differently from one another (contra e.g. Lekakou 2005). In Hebrew, middle morphology marks 1-3, and passive marks 4. In English, 1-3 is unmarked, 4 is marked as passive. This sug-gests that if a language has two distinct markings for middle and passive they will distribute across these lines. 1. Anticausative 2. Reflexive 3. Disp. Middle 4. Passive Greek Nact Nact Nact Nact Hebrew Middle Middle Middle Passive English Active Active Active Passive But the case of German and Dutch suggests that the picture is far more complicated than this. In this project, adding other languages to this picture, we want to achieve a deeper understanding of the subset relation. In addition, we want to understand why i) some languages develop a second passive, namely a reflexive passive; ii) why in some languages (German) reflexives can also form a passive, and iii) why participial constructions are not uniformely passive (non-canonical-passives in English, German and Romance and Greek stative participial constructions).

VIII Implicit arguments Any study of Voice would be incomplete without raising the issue of implicit argument represen-tation. The literature on implicit arguments has defined them as syntactically active elements that nevertheless do not occupy a syntactically projected position, though the definitions vary in their details (cf. e.g. Roeper 1987, Roberts 1987, and section 3.3.). In agreement with Bhatt & Pan-cheva (2005), we find the existence of elements that are syntactically active but not syntactically projected conceptually problematic. Thus, unlike Williams (1985, 1987), we are unable to con-clude that implicit arguments are not projected syntactically. While we can safely conclude that anticausatives lack an implicit agent, work by Kallulli (2006) and AAS (2006a, b) pointed out that anticausatives can involve a causer argument. The question is whether this should be represented as an implicit argument similar to the one found in passives (Kallulli 2006, 2007) or whether its licensing can be reduced to the presence of a causa-tive event in anticausatives as suggested by AAS (2006a, b) and Schäfer (2008a, c, 2009b) (cf. also Kratzer 2005). Things do not become easier in the case of the implicit arguments of passives and mid-dles. We have already discussed and criticized above accounts that represent the implicit agent of middles in the syntax. On the other hand, the implicit argument of passives is syntactically much more active suggesting that it is also syntactically represented. But how exactly is this argument projected? As a kind of pronoun in Spec,Voice (e.g. Sternefeld 1995, but see e.g. Jaeggli 1986 for critcism), by the passive morphology (Jaeggli 1986, Baker, Johnson & Roberts 1989, (see Embick 1997, and Collins 2005 for criticism), a potentially silent by-phrase (Collins 2005, Kal-lulli 2007) or as a thematic feature on Voice (Embick 1998, 2004b, Kallulli 2006)? Arguably, a decision should be taken on the basis of the semantic and syntactic properties associated with the implicit argument. One such property is that the implicit argument cannot be controlled (Chomsky 1986, Jaeggli 1986, Williams 1987, Bhatt & Pancheva 2005); in (24a), but not in (24b) can the giver(s) of the testimony be they, an observation that is not compatible with-out further ado with the proposal that the implicit argument is a covert pronoun (pro, PRO). Si-milarly, in example (24c) the implicit argument is disjoint from John, or at least is vague on that point in the same way as John wants Mary to be seen by someone is. (24) a. They expected [PRO to give damaging testimony] b. They expected [damaging testimony to be given] c. John wants Mary to be seen The same holds for German (25a). Note, however, that the addition of a by-phrase allows the implicit argument to co-refer with the matrix subject, suggesting that there is no general disjoint reference effect on the implicit argument. This is not compatible with the view in Williams (1987) that the existentially bound implicit argument acts like an R-expression. (25) a. Der Komissar hofft, dass der Verdächtige bald überführt wird b. Der Komissari hofft, dass der Verdächtige bald von ihmi überführt wird An important, related question that has not received much attention in the literature is how the properties of the implicit argument of passives can differ across languages. As discussed in AAS (2006), Alexiadou & Schäfer (2006) and Doron (2003), there are languages such as Greek (Hebrew) or Icelandic that allow only agent subjects but not causer subjects as implicit arguments of passives (27b vs. 26c); this is so even if the corresponding transitive sentence al-lows causer subjects (26a).

(26) a. I komotria/O ilios stegnose ta mallia mu The hairdresser/The sun dried-3sg the hair my The hairdresser/The sun dried my hair b. Ta mallia mu stegnothikan apo tin komotria The hair my dried-Nact by the hairdresser ‘My hair was dried by the hairdresser

c. ?*Ta mallia mu stegnothikan apo ton ilio The hair my dried-Nact by the sun ‘My hair were dried by the sun’ Interestingly, impersonal passives across languages show a general human restriction on their implicit argument (e.g. Maling & Sigurjónsdóttir 2002). (27) a. Es wurde (von dem Mann/?*vom Zug/*von dem Vogel) laut gepfiffen b. Er werd door de jongen/*de trein gefloten it was by the boys/*the train whistled A further potential difference concerns secondary predication. Chomsky (1986) claims that the implicit argument of passives differs from PRO in that it cannot be modified by a depictive. (28) a. They expected [PRO to leave the room sad] b. *The room was left sad Müller (2004) gives examples that show that this does not hold for German. (29) dass das Buch nackt gelesen wurde that the book naked read was ‘that the book was read naked’ However, it is not clear that we face a real difference between English and German. Baker (1988) and Collins (2005) both provide examples which suggest that the implicit argument of English passives behaves in fact like its German counterpart. (30) a. At the commune, breakfast is usually eaten nude b. This song must not be sung drunk But in Greek, the modification of the implicit argument is definitely ungrammatical as (31) shows: (31) a. O Janis tragudise to tragudi methismenos The John-nom sang-3sg the song-acc drunk-msc-sg b. *To tragudi tragudithike *methismeno/*methismenos The song-nom sang-Nact-3sg drunk-neut.sg/drunk msc.sg While this could indicate a peculiarity of the implicit argument of Greek passives, an alternative explanation might be more promising. As discussed by Jónsson (2009), the implicit argument of Icelandic passives can be modified by uninflected present participles but not by adjectives which are inflected for person. This suggests that the problem of the above examples relates to the li-censing of the adjectival inflection: if the implicit argument of passives makes no φ-features

available and if adjectives do not allow default agreement, the differences between the examples is explained. (32) a. Minningargreinin var lesin grátandi the.obituary was read crying b. *Morgunmatur er alltaf borðaður nakinn breakfast is always eaten naked.sg.masc The same set of questions emerges also in the area of nominalizations in B1, and we will address them jointly. An important issue here is whether nominalizations constitute a form of passive Voice, or whether at least some of them (e.g. examples of the type the city's destruction) are dispositional middle constructions, as suggested by Longobardi (2001). Longobardi points out that a difference between verbal passives and nominal passives is that unaffected objects can passivize in the former, but not in the latter. This raises the question once again, what kind of constructions are subject to Affectedness, and whether the same form of Affectedness is relevant in all cases (see below). IX Get-verbs We use this label as a cover term for all non-canonical passive auxiliaries. What is the status of these verbs? Are they semi-lexical (Haider 2001, van Riemsdijk 1998) or fully lexical? Or do they have the status of auxiliaries (Haegeman 2005, McIntyre 2005)? And if so, how do they differ from their be counterparts? We note that while English, German and French all have such verbs, they differ as to the type of light verb they use. English/German use a possessive light verb, Romance uses a causa-tive light verb. In addition, only Romance languages contain an overt reflexive (se faire, 18), although German also uses a reflexive with its dispositional middles and anticausatives. Why do the Romance languages require a reflexive to achieve an affected interpretation? How are pos-session structures (arguably involved in English and German get/bekommen constructions, see McIntyre 2005), light causative verbs and reflexive marking linked to affectedness? (see below) We further observe that Italian se faire constructions differ from their French counter-parts, as they allow both causer and agent par-PPs (French allows only Agentive phrases). What is the source for this difference and how similar are these then to real passive constructions? X Affectedness Typological studies all agree that the single argument present in Middle Voice is 'affected'. In the syntactic literature, the notion of affectedness was used in Anderson (1979) to describe the rela-tion between a preposable object and its head in noun phrases (e.g. the city's destruction). Since then, others have extended this notion to dispositional middles and to dative/possessive construc-tions (e.g. Landau 1999) which constitute areas of our investigation. Authors, however, dis-agree as to the correct description of the phenomenon and the factors that are responsible for it (see Anderson's 2006 overview for detailed discussion and references). Specifically, for Tenny (1994) and Krifka (1989) affectedness is primary a term for argu-ments that are taken to be affected by the eventuality denoted by the predicate. Often this term is seen as comparable to the term quantized arguments. The other notion Affectedness is unders-tood as relating to possession. It is not clear how the quantized view on Affectedness extends to the possessive one. Do we have different types of affectedness, i.e. does affectedness make refer-ence to scales, as suggested in C2? If so, how are these structurally encoded? If possessors do give rise to affectedness, how is this represented in the structure? Is it represented in terms of a possessive structure which relies on the presence of an applP (Marantz 1993, Cuervo 2003, Pylkkänen 2008), or via an Affect head, as recently argued for by Bosse et al. (2009)? Why do

some languages use reflexives for applicative structures? How is this different from/similar to what we find in dispositional middles? What are the general constructions that languages use to encode Affectedness? XI Participial constructions We know by now that participles fall into a number of distinct categories (stative, resultative, eventive), see Kratzer (2000) and Embick (2004a) for German and English respectively. In some languages these are distinct morphologically (Greek, Alexiadou & Anagnostopoulou 2008) but in most languages they are not (e.g. English, German). Which of the participles is involved in the periphrastic get-passive and bekommen-passive? How are the auxiliary and the participle compositionally interpreted in these periphras-tic constructions? Can such constructions be alternative realizations of middle Voice? For in-stance, a closer look at the syntactic behavior of get-passives reveals that these are indeed similar to dispositional middles and anticausatives. Get-passives, unlike be-passives, but like disposi-tional middles and anticausatives, do not license agentive adverbs or control into purpose clauses and (33-34) (McIntyre 2005, Toyota 2007 for a recent discussion and references; cf. Bowers 2002): (33) a. The ship was sunk [PRO to collect insurance money] (passive) b. The ship was sunk deliberately c. *The ship got sunk [PRO to collect insurance money] (get-passive) d. *The ship got sunk deliberately (34) a. *This bureaucrat bribes easily to avoid the draft (dispositional middle) b. *This bureacrat bribes deliberately c. *The ship sunk to collect the insurance (anticausative) d. *The ship sunk deliberately What explains the difference between the eventive verbal passive and the above constructions? How do the participles in the canonical and non-canocnical constructions differ? Are they glo-bally different or do they share some but not all syntactic/semantic properties (subset-relation)? Why are the differences between the different participles not morphologically reflected, i.e. con-trary to what isomorphism/iconicity suggests? What explains the cross-linguistic distribution of affected consrtrucions (possible in Ger-manic and Romance but not in Greek)? We note that e.g. Greek has participles quite similar to the participles used for in other periphrastic consturctions in languages like German or English (Alexiadou & Anagnostopoulou 2008, Anagnostopoulou 2003). Specifically, -menos- participles involve stativization of a phrase that contains an external argument. Nevertheless, Greek has not developed a second periphrastic passive, besides its synthetic passive. Thus, the question is: since all building blocks of this alternative passive are available in Greek, why is this construc-tion not passive? In addition, the Greek participial constructions cannot be used in manner simi-lar to the get-passives either. Rather, the Greek participial constructions seem to convey the Per-fect of Result reading associated with the Present Perfect in languages like German. In Greek this is systematically expressed via the be + participle construction (and not via the have +participle one, which is used for the other readings of the Present Perfect). The exact ways this comes about need to be identified, and this can only be done if Greek is contrasted to its Germanic and Rom-ance counterparts. Especially, we would like to investigate Italian and French participles in the light of Kratzer's and Embick's results in order to better understand the restrictions observed. In fact the problem is not limited to the status of participial constructions. Greek has light verbs of the appropriate type but still is not able to form such affected constructions.

Time schedule 2010/2: Work on reflexive passives and participial constructions and get verbs. Implicit argu-ment in passive constructions. Work on afto-. 2011: Unification of reflexive marking and non-active morphology. Identification of reasons for the subset relation. Affectedness and the scales it is subject to. Beginning to develop a theory of Voice syncretisms. 2012: Development of a theory of Voice syncretisms. Number of Voices available across lan-guages. Gaps in the paradigm. 2013: Enlarge the typological picture. Work on monograph. 2014/1: Completion of monograph on Voice. 3.5 Role within the Collaborative Research Center 3.5.1 Position with respect to the overall concept of the Research Center The project addresses questions of mismatch of form and interpretation that are the core concerns of the research concerns of this SFB. It is concerned with the theoretical question of underspeci-fication vs. ambiguity and it aims at advancing our understanding of the theory of underspecifica-tion at syntax-morphology interface. 3.5.2 Interaction with other projects The SFB presents an ideal environment for this particular research to be carried out as a number of sub-issues that are important for this proposal figure prominently in other projects as well. To begin with, B1 and B6 study phenomena that have been argued to be comparable, nominalization and verbal argument alternations. We believe that a joint study of these phenom-ena will lead us to better understand the syntactic constraints on a) implicit arguments and pas-sivization and b) questions of argument structure. What is its source? How can arguments be reduced, if at all, and what are the different mechanisms available cross-linguistically? Within area B, projects B4 and B5 address issues of argument alternations (verb particles and causative alternation respectively), in areas which neighbour our own. We expect fruitful discussions with both these projects concerning the projection of arguments and the role root classification plays in determining verb meaning. We expect to interact actively with project C2 not only on general issues of argument alternations but also on the role of Affectedness in grammar, its semantic ingredients and its syn-tactic representation. Thus we hope to profit from each other's results. Project B6 is interested in active-passive Voice alternations and in light verbs of a partic-ular type, phenomena that are being looked at from a different perspective within D2. We expect to profit from each other's results in pursue of an appropriate characterization of the construc-tions at hand.

3.6 Demarcation from other funded projects of the principal investigator Does not apply.

3.7 Auxiliary support for the project

201X/2 201X 201X 201X 201X/1

Funding for staff Salary scale.

No. Sum €

Salary

scale.

No. Sum €

Salary scale.

No. Sum €

Salary scale.

No. Sum €

Salary scale.

No. Sum €

Total: Total: Total: Total: Total:

Funding for direct costs

Sum €

Sum €

Sum €

Sum €

Sum €

Small equipment (up to € 10,000)

Consumables Laboratory animals Travel Other

Funding for instrumentation (instru-ments exceeding € 10,000 gross)

Funding for instrumentation

Total

Funding for instrumentation

Total

Funding for instrumentation

Total

Funding for instrumentation

Total

Funding for instrumentation

Total

<Tabelle wird von der Geschäftsleitung anhand der Angaben des Teilprojekts erstellt>

3.7.1 Staffing of the project

Name, acad. title, position

Field of research Department of the uni-versity

or non-university institution

Requested funding period:

Work performed in the project in hours/week (consultancy: C)

Salary scale

Core support 3.7.1.1 Research assis-tants (incl. student assistants)

3.7.1.2 Non-scientific assistants

Auxiliary support 3.7.1.3 Research assis-tants (incl. student assistants)

3.7.1.4 Non-scientific assistants

1) Bitte durchnummerieren und Aufgabenbeschreibung nachfolgend erläutern 2) Bitte Verfahrensgrundsätze der DFG zur Bezahlung wissenschaftlichen Personals beachten. <Tabelle wird von der Geschäftsführung anhand der Angaben aus den Teilprojekten erstellt.>

Job description of staff paid from core support of the funding period requested zu 1: <Artemis Alexiadou>

Responsible for the co-ordination of research. Work on the theory of implicit arguments, argument structure, subset-relations, get-verbs in English; responsible for English and Greek. zu n: <Name>

<Aufgabenbeschreibung> Job description of staff paid from auxiliary support for the funding period requested zu 1: <Florian Schäfer>

Florian Schäfer is an expert in issues concerning the causative alternation, dispositional middles, the morpho-syntax of sich, and the syntax of nominalization, as can be seen by his record of publication. In this project, he will be concerned with the following tasks:

i) Auxiliary selection and participial agreement with reflexive verbs, ii) the passive of ref lexive verbs, iii) the structure of participles in Germanic, iii) the morpho-syntax of unac cusativity diagnostics iv) bekommen-constructions in German and v) subset relations. zu 2: <Cinzia Campanini>

Cinzia Campanin finished her BA in Venice, and is currently an MA student at the University of Stuttgart. She will complete her MA thesis on optional pseudo reflexives in Italian and their relationship with applicative and dative constructions:

In this project, she will be responsible for the investigation of the inner-Romance varia tion in the following areas: i) reflexives and their relation to applicatives, ii) the se faire construction, iii) the structures of participles and iv) the application of Kayne's diagnos tics. It is expected that she will complete a Ph.D. thesis on issues of inner-Romance varia tion and Romance vs. Germanic comparison.

Both core and auxiliary members will work on the theoretical question of argument struc

ture, syncretisms, reflexive and non-active marking and the characterization of Voice. 3.7.1 Distribution and justification of funding for direc t costs (excl. salaries and instru-

mentation) (by financial year)

201X/2 201X 201X 201X 201X/1 Estimated core support available for direct costs

Requested funding for direct costs

(All figures in euros)

<Text> <Tabelle wird von der Geschäftsführung anhand der Angaben des Teilprojekts erstellt.> Justification for auxiliary support for direct costs (as above) <Cost element> for <financial year> <Designation and justification of application for this item> € <Sum applied for> etc. <Hinweise zu "Sachmitteln": Vor der Begründung der Einzelpositionen sind die Mittel anzugeben, die aus der Grundausstattung für das Teilpro-jekt für Sächliche Aufwendungen zur Verfügung stehen werden. Als Ergänzungsausstattung können Mittel unter den im folgenden erläuterten Kostenarten bei den einzelnen wissen-schaftlichen Teilprojekten beantragt werden. Zu Mitteln, die zentral für den gesamten Sonderforschungsbereich beantragt werden können, vgl. Hinweise zum Antragsmuster für das Verwaltungsprojekt. • Kleingeräte bis 10.000,- EUR (brutto) ohne Fahrzeuge:

Bei der Begründung ist ggf. auch kurz darauf einzugehen, warum vergleichbare Geräte, die an den am Teilpro-jekt beteiligten Instituten eventuell vorhanden sind, nicht herangezogen werden können. Es wird darauf hingewiesen, dass die Ausstattung mit Arbeitsplatzrechnern grundsätzlich aus Mitteln der Grund-ausstattung bereitzustellen ist und nur bei ausdrücklicher projektspezifischer Begründung beantragt werden kann.

• Verbrauchsmaterial: Bitte den Bedarf für einzelne Verbrauchsmaterialien (z.B. Chemikalien, Glasgeräte) spezifizieren und begründen. Mittel für Bürobedarf sollen im zentralen Verwaltungsprojekt für den gesamten Sonderforschungsbereich beant-ragt werden, nicht bei den einzelnen Teilprojekten.

• Reisen: Mittel für projektspezifische Reisen (Forschungsreisen, Expeditionen etc.) sind bei den einzelnen Teilprojekten zu beantragen und aufgeschlüsselt nach Fahrt- und Aufenthaltskosten zu begründen. Allgemeine Reisemittel ei-nes Sonderforschungsbereichs (Tagungsreisen etc.) sind im Rahmen des zentralen Verwaltungsprojekts zu beant-ragen. Soweit Veranstaltung, Teilnehmer und Art der aktiven Teilnahme zum Zeitpunkt der Antragstellung be-kannt sind, sind diese Angaben im Finanzierungsantrag aufzuführen, im übrigen der Größenordnung nach plausi-bel zu machen.

• Sonstiges: Sollten die für das Projekt notwendigen sächlichen Aufwendungen nicht unter eine der angegebenen Kategorien fallen, können Mittel unter „Sonstiges“ beantragt werden. Unter Sonstiges beantragte Positionen sind einzeln aufzuführen und zu begründen. Bitte achten Sie dabei, falls mehrere Positionen aufgeführt werden, auf eine gute Übersichtlichkeit.

3.7.3 Instrumentation (instruments exceeding € 10,000 gross and vehicles)

Requested for financial year

201X/2 201X 201X 201X 201X/1 Total:

(All prices in thousands of euros incl. VAT and transportation costs, etc.)

Literature: Abraham, W., 1986. Unaccusatives in German. Groninger Arbeiten zur germanistischen Linguistik

(GAGL) 28, 1-72. Abraham, W., 1987a. Was hat sich in ”Damit hat sich”s”? In: Centre de Recherche en Linguistique Ge-

rmanique (Nice) (ed.), Das Passiv im Deutschen. Niemeyer, Tübingen, 51-71. Abraham, W., 1987b. Zur Typologie des Mediums in der Westgermania.. In: W. Abraham, R. Århammer

(eds.), Linguistik in Deutschland. Niemeyer, Tübingen, 3-23. Abraham, W., 1994. Diathesis: the middle, particularly in West-Germanic. What does reflexivization

have to do with valency reduction?. In: W. Abraham, T. Givón, S. Thompson (eds.), Discourse grammar and typology. John Benjamins, Amsterdam, 3-47.

Abraham, W., 2006. Introduction: Passivization and typology. Form vs. Function – a confined survey into the research status quo. In: W. Abraham, L. Leisiö (eds.), Passivization and typology. Form and function. John Benjamins, Amsterdam, 1-28.

Ackema, P., Schoorlemmer, M., 1995. Middles and nonmovement. Linguistic Inquiry 26, 173-197. Ackema, P., Schoorlemmer, M., 2005. Middles. In: M. Everaert, H. van Riemsdijk, R. Goedemans, B.

Hollebrandse (eds.), The Blackwell companion to syntax, vol. III. Blackwell, Oxford, 131-203. Åfarli, T. A., 1992. The syntax of Norwegian passive constructions. John Benjamins, Amsterdam. Ágel, V., 1997. Reflexiv-Passiv, das (im Deutschen) keines ist. Überlegungen zu Reflexivität, Medialität,

Passiv und Subjekt. In: Ch. v. Dürscheid, K. H. Ramers, M. Schwarz (eds.), Sprache im Fokus. Fest-schrift für Heinz Vater zum 65. Geburtstag. Niemeyer, Tübingen, 147-187.

Alexiadou, A., Rathert, M., von Stechow, A., 2003. Introduction: The modules of the Perfect construc-tions. In: A. Alexiadou, M. Rathert, A. von Stechow (eds.), Perfect Explorations. Mouton de Gruyter, Berlin, v-xxxvii.

Alexiadou, A., 2001. Functional structure in nominals: nominalization and ergativity. John Benjamins, Amsterdam.

Anagnostopoulou, E., 2003. The syntax of ditransitives: evidence from clitics. Mouton de Gruyter, Ber-lin.

Andersen, P. K., 1994. Empirical studies in diathesis. Nodus, Münster. Anderson, M. 1979. Noun phrase structure. Ph.D. Dissertation University of Connecticut. Anderson, M., 2006. Affectedness. In: M. Everaert, H. van Riemsdijk, R. Goedemans, B. Hollebrandse

(eds.), The Blackwell Companion to Syntax. Blackwell, Oxford, 121-140. Anderson, S. R., 1990. The grammar of Icelandic verbs in -st. In: J. Maling, A. Zaenen (eds.), Modern

Icelandic Syntax. Academic Press, San Diego, Ca, 235-273. Arrese, J., 1999. Conceputalization of events, semantic relations between constructions and topology: a

catastrophe theoretic study of get and be. Journal of English Studies I, 97-117. Babby, L. H., 1975. A transformational analysis of transitive -sja verbs in Russian. Lingua 35, 297-332. Babby, L. H., Brecht, R. D., 1975. The syntax of Voice in Russian. Language 52 (2), 342-367. Baker, M., 1988. Incorporation: a theory of grammatical function changing. University of Chicago

Press, Chicago. Baker, M., Johnson, K., Roberts, I., 1989. Passive arguments raised. Linguistic Inquiry 20, 219-252. Bhatt, R., Pancheva, R., 2005. Implicit Arguments. In: M. Everaert, H. van Riemsdijk, R. Goedemans, B.

Hollebrandse (eds.), The Blackwell Companion to Syntax, vol. II, Blackwell, Oxford, 554-584. Borer, H., 2005. Structuring Sense. An Exo-Skeletal Trilogy. Vols. 1 and 2. Oxford University Press,

Oxford. Bosse, S., Bruening, B., Cathcart, M.E., Peng, A. E., Yamada, M., 2008. Affect Arguments Cross-

linguistically. In: M. Tadic, M. Dimitrova-Vulchanova, S. Koeva, S. (eds.), Proceedings of the Sixth International Conference on Formal Approaches to South Slavic and Balkan Languages. 41-47.

Bowers, J., 2002. Transitivity. Linguistic Inquiry 33, 183-224. Brame, M., 1983. Bound Anaphora is not a Relation between NPs: Evidence for Local Word Grammar

(Without Trees). Linguistic Analysis 11, 139-166. Brecht, R., Levine, J., 1984. Conditions on Voice Marking in Russian. In: M. Flier, R. Brecht Issues in

Russian Morphosyntax. Slavica Publishers, Columbus, Ohio, 118-37. Burzio, L., 1981. Intransitive Verbs and Italian Auxiliaries. Dissertation, MIT. Burzio, L., 1986. Italian Syntax. A Government-Binding Approac., Kluwer, Dordrecht.

Burzio, L., 1991. The morphological basis of anaphora. Journal of Linguistics 27, 81-105. Caha, P., 2008. The Superset principle. Ms., University of Tromsø. Cennamo, M., 1993. The Reanalysis of Reflexives: a diachronic perspective. Naples: Liguori. Centineo, G., 1995. The Distribution of si in Italian Transitive/Inchoative Pairs. In: Proceedings from

Semantics and Linguistic Theory. Austin: University of Texas, 54-71. Chierchia, G., 1989/2004. A semantics for unaccusatives and its syntactic consequences. In: A. Alex-

iadou, E. Anagnostopoulou, M. Everaert (eds.), The unaccusativity puzzle: explorations of the syn-tax-lexicon interface. Oxford University Press, Oxford, 22-59.

Chomsky, N., 1981. Lectures on Government and Binding. Foris, Dordrecht. Chomsky, N., 1986. Knowledge of language: Its nature, origin, and use. Praeger, New York. Chomsky, N., 2001. Derivation by phase. In: M. Kenstowicz (ed.), Ken Hale: A Life in Language. MIT

Press, Cambridge, Mass, 1-52. Cinque, G., 1988. On Si constructions and the theory of Arb. Linguistic Inquiry 19, 521-581. Collins, C., 2005. A Smuggling Approach to the Passive in English. Syntax, 8 (2), 81-120. Comrie, B., 2008. What is a passive? In: Z. E. Fernández, S. Wichmann, C. Chamoreau, A. Á. González

(eds.), Studies in voice and transitivity. Lincom Europa, Munich, 1-18. Condoravdi, C., 1989. The middle: where semantics and morphology meet. MIT Working Papers in Lin-

guistics 11. MIT, 18-30. Cook, P., 2006. The Datives that aren't born equal: Beneficiaries and the Dative Passive. In: D. Hole, A.

Meinunger, W. Abraham (eds.), Datives and Other Cases. John Benjamins, Amsterdam, 141-184. Croft, W., Shyldkrot, H. B.-Z., Kemmer, S., 1987. Diachronic semantic processes in the middle voice. In:

A. Ramat, O. Carruba, G. Bernini. (eds.), Papers from the 7th International Conference on Histori-cal Linguistics, Bernini. John Benjamins, Amsterdam, 179-192.

Cuervo, M. C., 2003. Datives at Large. Ph.D. thesis, MIT. Czepluch, H., 1988. Kasusmorphologie und Kasusrelationen. Linguistische Berichte 116, 275-310. D'Alessandro, R., 2007. Impersonal si constructions. Mouton de Gruyter, Berlin. Dobrovie-Sorin, C., 1998. Impersonal se constructions in Romance and the passivization of unergatives.

Linguistic Inquiry 29, 399-437. Dobrovie-Sorin, C., 2003. Implicit agentivity without agents in the syntax: a crosslinguistic analysis of

SE-verbes. In: Actes du Colloque Représentation du sens. Montréal. Dobrovie-Sorin, C., 2005. SE-anaphors. In: M. Everaert, H. van Riemsdijk, R. Goedemans, B. Holle-

brandse (eds.), The Blackwell Companion to Syntax (Syncom). Blackwell, Oxford, 118-179. Doron, E., 2003. Agency and Voice: the semantics of the Semitic templates. Natural Language Seman-

tics, 11 (1), 1-67. Doron, E., Rappaport Hovav, M., 2007. Towards a Uniform Theory of Valence-changing Operations.

Proceedings of IATL 23. Embick, D., 1997. Voice and the Interfaces of Syntax. Ph.D. thesis, University of Pennsylvania. Embick, D., 1998. Voice Systems and the Syntax/Morphology Interface. In: H. Harley (ed.), MITWPL

32: Proceedings of the Penn/MIT Workshop on Aspect, Argument Structure, and Events. MIT, 41-72.

Embick, D., 2000. Features, Syntax, and Categories in the Latin Perfect. Linguistic Inquiry 31 (2), 185-230.

Embick, D., 2003. Locality, listedness and morphological identity. Studia Linguistica 57, 143-169. Embick, D., 2004a. On the structure of resultative participles in English. Linguistic Inquiry 35 (3), 355-

392. Embick, D., 2004b. Unaccusative Syntax and Verbal Alternations. In: A. Alexiadou, E. Anagnostopou-

lou, M. Everaert (eds.), The unaccusativity puzzle: explorations of the syntax-lexicon interface. Ox-ford University Press, Oxford, 137-158.

Everaert, M., 1986. The Syntax of Reflexivization. Foris, Dordrecht. Eythórsson, T., 2008. The New Passive in Icelandic really is a passive. In: T. Eythórsson (ed.), Gram-

matical Change and Linguistic Theory: The Rosendal Papers. John Benjamins, Amsterdam, 173-219.

Fagan, S. M., 1992. The syntax and semantics of middle constructions: A study with special reference to German. Cambridge University Press, Cambridge.

Faltz, L., 1985. Reflexivisation: A study in universal syntax. Garland, New York.

Fanselow, G., 1987, Konfigurationalität, Narr, Tübingen. Fanselow, G., 2000. Optimal Exceptions In: B. Stiebels, D. Wunderlich (eds.), The Lexicon in Focus.

Berlin, Ackademie Verlag, 173-209. Fanselow, G., 2003. Free constituent order, A Minimalist Interface Account. Folia Linguistica 37 (1-2),

191-231. Fellbaum, C., 1986. On the middle construction in English. Indiana University Linguistics Club, Bloom-

ington. Fellbaum, C., Zribi-Hertz, A., 1989. The middle construction in French and English. Indiana University

Linguistics Club, Bloomington. Fox, D., Grodzinsky, Y., 1998. Children's passive: a view from the by-phrase. Linguistic Inquiry, 29 (2),

311-332. Frey, W., 1993. Syntaktische Bedingungen für die semantische Interpretation. Akademie-Verlag, Berlin. Gast, V., Haas, F., 2008. On reflexive and reciprocal readings of anaphors in German and other European

languages. In: E. König, V. Gast (eds.), Reflexives and Reciprocals -Theoretical and Typological Explorations. Mouton de Gruyter, Berlin, 307-346.

Geniušien÷, E., 1987. The typology of reflexives. Mouton de Gruyter, Berlin. Geurts, B., 2004. Weak and strong reflexives in Dutch. In: P. Schlenker, E. Keenan (eds.), Proceedings

of the ESSLLI workshop on semantic approaches to binding theory. Nancy, France. Givón, T., 1979. On Understanding Grammar. Academic Press, New York. Givón, T., Yang, L. 1994. The rise of the English get-passive. In: B. Fox, P. Hopper (eds.) Voice: Form

and Function. John Benjamins, Amsterdam, 119-150. Grewendorf, G., 1989. Ergativity in German. Foris, Dordrecht. Grimshaw, J., 1981. On the lexical representation of Romance reflexive clitics. In: J. Bresnan (ed.), The

Mental Representation of Grammatical Relations. MIT Press, Cambridge Mass., 87- 148. Grimshaw, J., 1990. Argument Structure. MIT Press, Cambridge, Mass. Haegeman, L., 1985. The get-passive and Burzio’s generalization. Lingua 66, 53-77. Haider, H., 1984. Mona Lisa lächelt stumm: über das sogenannte Deutsche Rezipientenpassiv. Linguisti-

sche Berichte 89, 32-42. Haider, H., 1985. Von sein oder nicht sein: Zur Grammatik des Pronomens „sich“. In: W. Abraham (ed.),

Erklärende Syntax des Deutschen. Narr, Tübingen, 223-254. Haider, H., 1986. Fehlende Argument: vom Passiv zu kohärenten Infinitiven. Linguistische Berichte 101

3-33. Haider, H., 1992. Branching and Discharge. In: Working Papers of the special research group (SFB)

340, Report Nr. 23. University of Tübingen/University of Stuttgart. Haider, H., 1993. Deutsche Syntax - generativ. Vorstudien zur Theorie einer projektiven Grammatik.

Narr, Tübingen. Haider, H., 2000. The license to license. In: E. Reuland (ed.), Arguments and Case. Explaining Burzio’s

Generalization. John Benjamins, Amsterdam, 31-54. Haider, H., 2001. Heads and seletion. In: N. Corver, H. van Riemsdijk (eds.), Semi-lexical categories:

the function of content words and the content of function words. Mouton de Gruyter, Berlin, 67-96. Haider, H., Bierwisch, M., 1989. Ergebnisbericht des Projekts A1, Steuerung kompositionaler Strukturen

durch thematische Information des SFB 340, Universität Stuttgart/Universität Tübingen. Haider, H., Rosengren, I., 2003. Scrambling: Nontriggered Chain Formation in OV Languages. In: Jour-

nal of Germanic Linguistics 15, 203-267. Haimann, J., 1983. Iconic and Economic Motivation. Language 59, 781-819. Hale, K., Keyser, J., 1986. Some Transitivity Alternations in English. Lexicon Project Working Papers 7.

Center for Cognitive Science, MIT, Cambridge, Mass. Hale, K., Keyser, J., 1987. A view from the Middle. Lexicon Project Working Papers 10. Center for

Cognitive Science, MIT, Cambridge, Mass. Hale, K., Keyser, S. J., 1993. On argument structure and the lexical expression of syntactic relations. In:

K. Hale, S. J. Keyser (eds.), The View from Building 20. MIT Press, Cambridge, Mass., 53-109. Halle, M., Marantz, A., 1993. Distributed Morphology and the Pieces of Inflection. In: K. Hale, S. J.

Keyser (eds.), A View From Building 20. MIT Press, Cambridge, Mass., 111-176. Harley, H., Noyer, R., 2003. Distributed Morphology. In: L. Cheng, R. Sybesma (eds.), The Second

GLOT International State-of-the-Article Book. Mouton de Gruyter, Berlin, 463-496.

Härtl, H., 2003. Conceptual and grammatical characteristics of argument alternations: The case of decau-sative verbs. Linguistics 41 (5), 883-916.

Haspelmath, M., 1987. Transitivity Alternations of the Anticausative Type. Köln: Institut für Sprachwis-senschaft. Arbeitspapier 5.

Haspelmath, M., 1990. The grammaticization of passive morphology, Studies in Language 14, 25-72. Haspelmath, M., 1993. More on the typology of inchoative/causative verb alternations. In: B. Comrie, M.

Polinsky (eds.), Causatives and Transitivity. John Benjamins, Amsterdam, 87-120. Hatcher, A. G., 1949. To get/be invited. Modern Language Notes 64, 433-446. Heck, F., 2000. Tiefenoptimierung - Deutsche Wortstellung als wettbewerbsgesteuerte Basisgenerierung.

Linguistische Berichte 184, 441-468. Hellan, L., 1988. Anaphora in Norwegian and the Theory of Grammar. Foris Publications, Dordrecht. Hoekstra, T., Roberts, I., 1993. Middle constructions in Dutch and English. In: E. Reuland, W. Abraham

(eds.), Knowledge and Language II. Lexical and Conceptual Structure. Kluwer, Dordrecht, 183-220. Höhle, T., 1978. Lexikalistische Sxntax. Die Aktiv-Passiv-Relation und andere Infinitkonstruktionen im

Deutschen. Niemeyer, Tübingen. Hole, D, Meinunger, A., Abraham, W., (eds.) 2006. Datives and Other Cases. John Benjamins, Amster-

dam. Huang, C.-T. J., 1999. Chinese passives in comparative perspective. Tsing Hua Journal of Chinese Stud-

ies 29, 423-509. Jaeggli, O. A., 1986. Passive. Linguistic Inquiry 17, 587-622. Jónsson, J. G., 2009. The new impersonal as a true passive. In: Alexiadou et al. (eds.), Advances in Com-

parative Germanic Syntax. John Benjamins, Amsterdam, 281-306. Kallulli, D., 2006. A unified analysis of passives, anticausatives and reflexives. In: O. Bonami, P.

Cabredo-Hofherr (eds.), Empirical Issues in Formal Syntax and Semantics 6, 201-225. Kallulli, D., 2007. Rethinking the passive/anticausative distinction. Linguistic Inquiry 38 (4), 770-780. Kaufmann, I., 2001. Medium: Eine Studie zur Verbsemantik. Habilitationsschrift, University of Düssel-

dorf. Kaufmann, I., 2003. Reflexive Verben im Deutschen In: L. Gunkel, G. Müller, G. Zifonun G (eds.), Ar-

beiten zur Reflexivierung. Niemeyer, Tübingen, 135-155. Kayne, R., 1975. French Syntax. The Transformational Cycle. MIT Press, Cambridge, Mass. Kayne R., 2008. Expletives, datives and the tension between morphology and syntax. In: M.T. Biberauer

(ed.), The limits of syntactic variation. John Benjamins, Amsterdam, 175-217. Kazenin, K. I., 2001. The passive voice. In: M. Haspelmath, E. Konig, W. Oesterreicher, W. Raible

(eds.), Language Typology and Language Universals, Vol. 2.. Mouton de Gruyter, Berlin, 899-916. Kemmer, S., 1993. The middle voice. John Benjamins, Amsterdam. Keyser, S. J., Roeper, T., 1984. On the Middle and Ergative Constructions in English. Linguistic Inquiry

15, 381-416. Klaiman, M. H., 1992. Middle verbs, reflexive middle constructions, and middle voice. Studies in Lan-

guage 16 (1), 35-61. Klein, W., 1992. The present perfect puzzle. Language 68 (3), 525-552. König, E., Siemund, P., 2000a. Intensifiers and reflexives: A typological perspective. In: Z. Frajzinger,

T.S. Curl (eds.) Reflexives: Forms and Functions. John Benjamins, Amsterdam, 41-74. König, E., Siemund, P., 2000b. The development of complex reflexives and intensifiers in English, Dia-

chronica XVII, 39-84. König, E., Vezzosi, L., 2004. The role of predicate meaning in the development of reflexivity. In: B.

Wiemer, W. Bisang, N. Himmelmann (eds.), What makes Grammaticalization? A Look from its Fringes and its Components. Mouton de Gruyter, Berlin, 213-244.

Koontz-Garboden, A., 2007. States, changes of state, and the Monotonicity Hypothesis. Ph.D. thesis, Stanford University.

Koster, J., 1987. Domains and Dynasties. Dordrecht: Foris. Kratzer, A., 1994. The Event Argument and the Semantics of Voice. Ms., UMass. Kratzer, A., 1996. Severing the External Argument from its Verb. In: J. Rooryck, L. Zaring (eds.),

Phrase Structure and the Lexicon. Kluwer, Dordrecht, 109-137. Kratzer, A., 2000. Building statives. In: Berkeley Linguistic Society 26.

Kratzer, A., 2005. Building Resultatives. In: C. Maienborn, A. Wöllstein-Leisten (eds.), Event Arguments in Syntax, Semantics, and Discourse. Niemeyer, Tübingen, 178-212.

Krifka, M., 1989. Nominal reference, temporal constitution and quantification in event semantics. In: R. Bartsch, J. van Benthem, P. van Emde Boas (eds.), Semantics and Contextual Expressions. Foris, Dordrecht, 75-115.

Kuroda, S. -Y., 1979. On Japanese Passives. In: G. Bedell, E. Kobayashi, M. Muraki (eds.), Explorations in Linguistics: Papers in honor of Kazuko Inoue. Kenkyusha, Tokyo, 305-347.

Laanemets, A., to appear. The passive voice in written and spoken Scandinavian. In: M. Fryd (ed.), Passive in Germanic Languages (Groninger Arbeiten zur germanistischen Linguistik.).

Labelle, M., 1990. Unaccusatives and pseudo-unaccusatives in French. In: J. Carter, R.-M. Dechaine, B. Philip, T. Sherer (eds.), Proceedings of Nels 20. GLSA, Amherst, 303-317.

Labelle, M., 1992. Change of state and valency. Journal of Linguistics 28, 375-414. Labelle, M., 2002. The French non canonical passive in se faire. In: H. Shosuke, B. Palek, O. Fujimura

(eds.), Proceedings of Linguistics and Phonetics 2002. Charles University Press and Meikai Univer-sity, Tokyo.

Labelle, M., 2008. The French Reflexive and Reciprocal se. Natural Language and Linguistic Theory 26 (4), 833-876.

Landau, I., 1999. Possessor raising and the structure of VP. Lingua 107, 1-37. Langacker, R. W. 1987. Foundations of cognitive grammar, vol. 1, 2. Stanford University Press, Stan-

ford. Lehmann, C., 1982. Thoughts on Grammaticalization. A programmatic Sketch. Vol. I. Arbeiten des

Kölner Universalien-Projekts, Nr. 48. Köln. Leirbukt, O., 1997. Untersuchungen zum 'bekommen'-Passiv im heutigen Deutsch. Niemeyer, Tübingen. Lekakou, M., 2005. In the Middle, Somewhat Elevated. The semantics of middles and its crosslinguistic

realization. Ph.D. thesis, University of London. Levin, B., Rappaport, M., 1986. The Formation of Adjectival Passives, Linguistic Inquiry 17, 623-661. Levin, B., Rappaport Hovav, M., 1995. Unaccusativity. At the Syntax-Lexical Semantics Interface. MIT

Press, Cambridge, Mass. Longobardi, G., 2001. The Structure of DPs: some Principles, Parameters and Problems. In: C. Collins,

M. Baltin (eds.), Handbook of Syntactic Theory. Blackwell, Oxford, 562-601. Lyons, J., 1969. Introduction to Theoretical Linguistics. Cambridge University Press, Cambridge. Mackridge, P., 1987. Modern Greek. Oxford University Press, Oxford. Maling, J., Sigurjónsdóttir, S., 2002. The new impersonal construction in Icelandic. The Journal of

Comparative Germanic Linguistics 5, 97–142. Manney, L. J., 2000. Middle Voice in Modern Greek: Meaning and Function of an Inflectional Category.

John Benjamins, Amsterdam. Manzini, M. R., 1983. On Control and Control Theory. Linguistic Inquiry 14 (3), 421-446. Manzini, M. R., 1986. On Italian si. In: H. Borer (ed.), The Syntax of Pronominal Clitics (Syntax and

Semantics vol. 18). Academic Press, New York, 241-262. Manzini, R., Roussou, A., Savoia, L., 2009. Voice morphology in Albanian and Greek: implications for

the morpholexical and LF interfaces. Ms., University of Firenze and University of Patras. Marantz, A., 1984. On the Nature of Grammatical Relations. MIT Press, Cambridge, Mass. Marantz, A., 1993. Implications of asymmetries in double object constructions. In: Sam Mchombo (ed.),

Theoretical Aspects of Bantu Grammar. Stanford: CSLI, 113-150. Marantz, A., 1997. No escape from syntax: Don't try morphological analysis in the privacy of your own

Lexicon. In: A. Dimitriadis, L. Siegel, C. Surek-Clark, A. Williams (eds.), Proceedings of the 21st Annual Penn Linguistics Colloquium. Penn Working Papers in Linguistics 4 (2), 201-225.

Marantz, A., 2006. Phases and Words. Ms., New York University. Marelj, M., 2004. Middles and Argument Structure across Languages. Ph.D. thesis, Utrecht Institute of

Linguistics OTS. Massey, V. W., 1992. Compositionality and constituency in Albanian. Ph.D. dissertation, University of

North Carolina, Chapel Hill. McCawley, N. A., 1972. On the treatment of Japanese passives. In: P. M. Perenteau, J. N. Levi, G. C.

Phares (eds.), Papers from Eighth Regional Meeting of the Chicago Linguistic Society, University of Chicago, Chicago, 256-270.

McFadden, T., 2004. The Position of Morphological Case in the Derivation: a study on the syntax-morphology interface. Ph.D. thesis. University of Pennsylvania.

McFadden, T., Alexiadou, A., to appear. Perfects, resultatives and auxiliaries in Early English. To appear in Linguistic Inquiry.

McGinnis, M., 1998. Reflexive external arguments and lethal ambiguity. In: E. Curtis, J. Lyle, G. Web-ster (eds.), Proceedings of WCCFL 16. CSLI, Stanford, 303–317.

McGinnis, M., 2000. Reflexive clitics and the specifiers of vP. In: L. Pylkkänen, H. Harley, A. van Hout (eds.), Papers from the UPenn/MIT Round Table on the Lexicon,. MITWPL 35. MIT Working Pa-pers in Linguistics, 137–160.

McIntyre, A., 2005. The Semantic and Syntactic Decomposition of get: An interaction between verb meaning and particle placement. Journal of Semantics 22, 401-438.

Michaelis, S., 1997. Antikausativ als Brücke zum Passiv: fieri, venire und se im Vulgärlateinischen und Altitalienischen. In: O. Winkelmann, M. Braisch (eds.), Neuere Beschreibungsmethoden der roma-nischen Syntax. Narr, Tübingen, 69-98.

Miyagawa, S., 1998. (s)ase as an Elsewhere Causative and the Syntactic Nature of Words. Journal of Japanese Linguistics 16, 67-110.

Molnárfi, L., 1998. Kasusstrukturalität und struktureller Kasus - zur Lage des Dativs im heutigen Deutsch. Linguistische Berichte 176, 535-580.

Müller, G., 1999. Optimality, Markedness and word order in German. Linguistics 37 (5), 777-818. Müller, S., 2004. An Analysis of Depictive Secondary Predicates in German without Discontinuous Con-

stituents. In: S. Müller (ed.): Proceedings of the HPSG-2004 Conference. CSLI, Stanford, 202–222. Musan, R., 2002. The German perfect. Its semantic composition and its interactions with temporal ad-

verbials. Kluwer, Dordrecht. Napoli, D. J., 1976. The two si's of Italian: an analysis of reflexive, inchoative, and indefinite subject

sentences in modern standard Italian. Linguistic Club, Indiana Univ., Bloomington. Oesterreicher, W., 1992. SE im Spanischen: Pseudoreflexivität, Diathese und Prototypikalität von seman-

tischen Rollen. Romanistisches Jahrbuch 43, 237-260. Plank, F., 1993. Peculiarities of Passives of Reflexives in German. Studies in Language 17 (1), 135-167. Primus, B., 1989. Parameter der Herrschaft. Reflexivpronomina im Deutschen. In: Zeitschrift für

Sprachwissenschaft 8, 53-88. Pylkkänen, L., 2008. Introducing arguments. MIT Press, Cambridge, Mass. Ramchand, G., 2008. Verb Meaning and the Lexicon: A First Phase Syntax. Cambridge University Press,

Cambridge. Rapoport, T. R., 1999. The English middle and agentivity. Linguistic Inquiry 30, 147-155. Rapp, I., 1996. Zustand? Passiv? – Überlegungen zum sogenannten „Zustandspassiv“. In: Zeitschrift für

Sprachwissenschaft. 15 (2), 231-265. Rapp, I., 1997. Partizipien und semantische Struktur. Zu passivischen Konstruktionen mit dem 3. Status.

Stauffenburg, Tübingen. Rathert, M., 2006. Simple preterit and composite perfect tense: The role of the adjectival passive. In:

Abraham, Werner, Larisa Leisio (eds.), Passivisation and typology: Form and function. John Ben-jamins, Amsterdam, 518-43.

Rathert, M., 2009. Zur Morphophonologie des Partizips II im Deutschen. Linguistische Berichte 218, 157-190.

Reinhart, T., 1996. Syntactic effects of lexical operations: reflexives and unaccusatives. OTS working papers in Linguistics. Utrecht institute of Linguistics, Utrecht University.

Reinhart, T., 2000. The Theta System: syntactic realization of Verbal concepts. OTS working papers in Linguistics. Utrecht institute of Linguistics, Utrecht University.

Reinhart, T., 2002. The Theta System – An Overview. Theoretical Linguistics 28 (3), 229-290. Reinhart, T., Reuland, E., 1993. Reflexivity. Linguistic Inquiry 24, 657-720. Reinhart, T., Siloni, T., 2004. Against the Unaccusative Analysis of Reflexives. In: A. Alexiadou, E.

Anagnostopoulou, M. Everaert (eds.), The Unaccusativity Puzzle. Oxford University Press, Oxford, 288-331.

Reinhart, T., Siloni, T., 2005. The Lexicon-Syntax Parameter: Reflexivization and Other Arity Opera-tions. Linguistic Inquiry 36, 389-436.

Reis, M., 1976. Reflexivierung in deutschen A.c.I.-Konstruktionen. Ein transformationsgrammatisches Dilemma, Papiere zur Linguistik 9, 5.82.

Reis, M., 1982. Reflexivierung im Deutschen. In: E. Faucher (ed.), Actes du Colloque du Centre de Re-cherches Germaniques de l'Université de Nancy & Journée Annuelle des Linguistes de l'Association des Germanistes de l'Enseignement Supérieur 12. Nancy, 1-40.

Reis, M., 1985. Mona Lisa kriegt zu viel: vom sogenannten Rezipientenpassiv im Deutschen. Linguis-tische Berichte 96, 140-155.

van Riemsdijk, H., 1998. Categorial feature magnetism: the endocentricity and distribution of projec-tions. The Journal of Comparative Germanic Linguistics 2, 1-48.

Rivero, M. L., 1990. The Location of Non-Active Voice in Albanian and Modern Greek. Linguistic In-quiry 21, 135-146.

Rivero, M. L., 1992. Adverb Incorporation and the Syntax of Adverbs in Modern Greek. Linguistics and Philosophy 15, 289-331.

Rivero, M. L., 2004. Datives and the Non-active Voice/Reflexive Clitic in Balkan Languages In: O. Mis-eska-Tomic (ed.), Balkan Syntax and Semantics. John Benjamins, Amsterdam, 237-267.

Rivero, M. L., Sheppard, M., 2001. On Impersonal se /się in Slavic. In: G. Zybatow et al. (eds.), Current Issues in Formal Slavic Linguistics. Peter Lang, Frankfurt/Main, 137-147.

Rivero, M. L., Sheppard, M., 2003. Indefinite Reflexive Clitics in Slavic: Polish and Slovenian. In: Natu-ral Language and Linguistic Theory, 21, 89-155.

Roberts, I., 1987. The Representation of Implicit and dethematized Subjects. Foris, Dordrecht. Roeper, T., 1987. Implicit arguments and the head-complement relation. Linguistic Inquiry 18 (2), 267-

310. Rosen, C., 1988. The Relational Structure of Reflexive Clauses. Garland, New York. Rothstein, B., 2008. The perfect time span. On the present perfect in Swedish, German and English. John

Benjamins, Amsterdam. Ruwet, N., 1972. Théorie syntaxique et syntaxe du français. Ed. du Seuil, Paris. Sells, P., Zaenen, A., Zec, D., 1987. Reflexivization variation: Relations between syntax, semantics, and

lexical structure. In: M. Iida, S. Wechsler, D. Zec (eds.), Working Papers in Grammatical Theory and Discourse Structure: Interactions of Morphology, Syntax, and Discourse. CSLI, Stanford, 169-238.

Selig, M., 1998. Pseudoreflexivität im Altitalienischen - Voraussetzungen und Richtungen eines Gram-matikalisierungsprozesses. In: H. Geisler, D. Jacob (eds.), Transitivität und Diathese in romanischen Sprachen. Niemeyer, Tübingen, 21-42.

Shibatani, M., 1985. Passives and related constructions: a prototype analysis. Language 61, 821-848. Shibatani, M., 1988 (ed.). Passive and Voice. John Benjamins, Amsterdam. Shibatani, M., 1994. An Integrational Approach to Possessor Raising, Ethical Datives and Adversative

Passives. In: S. Gahl, A. Dolbey, C. Johnson (eds.), Proceedings of the Twentieth Annual Meeting of the Berkeley Linguistics Society. Berkeley Linguistics Society, Berkeley.

Siemund, P., Hole, D., 2000. Introduction: Operations on argument structure. Special issue of Linguistics 44 (2), 203-216.

Siewierska, A., 1984. The passive: a comparative linguistic analysis. Croom Helm, London. Siewierska, A., 1988. The Passive in Slavic. In: M. Shibatani (ed.), Passive and Voice. John Benjamins,

Amsterdam, 243-289. Sigurðsson, H. A., 1989. Verbal Syntax and Case in Icelandic. Doctoral dissertation, University of Lund. Sigurðsson, H. A., to appear. On the new passive. To appear in Syntax. Smith, M., 2004. Light and heavy reflexives. Linguistics 42 (3), 573-615. von Stechow, A., 1996. The Different Readings of Wieder 'Again': A Structural Account. Journal of Se-

mantics 13, 87-138. von Stechow, A., 2002. Perfect of Result. Ms., University of Tübingen. Steinbach, M., 2002. Middle Voice. John Benjamins, Amsterdam. Steinbach, M., Gärtner, H.-M., 2000. What do Reduced Pronominals Reveal about the Syntax of Dutch

and German? In: Linguistics in Potsdam 9, 7-62. Sternefeld, W., 1995. Voice Phrases and their Specifiers. FAS Papers in Linguistics 3, 48-85. Stroik, T., 1992. Middles and Movement. Linguistic Inquiry 23, 127-137. Stroik, T., 1999. Middles and Reflexivity. Linguistic Inquiry 30, 119-131.

Talmy, L., 1988. Force dynamics in language and cognition. Cognitive Science 12, 49-100. Tenny, C. L., 1994. Aspectual Roles and the Syntax-Semantic Interface. Studies in Linguistics and Phi-

losophy 52. Kluwer, Dordrecht. Thráinsson, H., 2007. The Syntax of Icelandic. Cambridge University Press, Cambridge. Toyota, J., 2007. An adversative passive in English: in search of origins. In: N. Delbecque, B. Cornille

(eds.), Interpreting construction schemas. Mouton de Gruyter, Berlin, 143-170. Tsimpli, I. M., 1989. On the Properties of the Passive Affix in Modern Greek. UCL Working Papers in

Linguistics 1. 235-260. Vater, H., 1995. Zum Reflexiv-Passiv im Deutschen. In: H. Popp (ed.), Deutsch als Fremdsprache: an

den Quellen eines Faches. Festschrift für Gerhard Helbig zum 65. Geburtstag. München: Iudicium, 85-192.

Vogel, R., Steinbach, M., 1998. The Dative - An Oblique Case. Linguistische Berichte 173, 65-90. Wasow, T., 1977. Transformations and the Lexicon. In: P. Culicover, T. Wasow, A. Akmajian (eds.),

Formal Syntax. Academic Press, New York, 327-360. Wegener, H., 1985. Der Dativ im heutigen Deutsch. Narr, Tübingen. Wegener, H., 1991. Der Dativ - ein struktureller Kasus?. In: G. Fanselow, S. Felix (eds.), Strukturen und

Merkmale syntaktischer Kategorien. Narr, Tübingen, 70-103. Wehrli, E., 1986. On Some Properties of French Clitic se. In: H. Borer (ed.) Syntax and Semantics 19:

The Syntax of Pronominal Clitic., Academic Press, New York, 263-283. Williams, E., 1981. Argument structure and morphology. The Linguistic Review 1 (1), 81-114. Williams, E., 1985. PRO and the Subject of NP. Natural Language and Linguistic Theory 3, 297-315. Williams, E., 1987. Implicit Arguments, the Binding Theory and Control. Natural Language and Lin-

guistic Theory 5, 151-180. Wunderlich, D., 1997. Cause and the structure of verbs. Linguistic Inquiry 28, 27-68. Zombolou, K., 2004. Verbal alternations in Greek: a semantic approach. Ph.D. thesis, University of

Reading. Zribi-Hertz, A., 1982. La construction “se-moyen” du francais et son statut dans le triangle: moyen -

passif - réfléchi. Lingvisticae Investigationes 6, 345-401. Zribri-Hertz, A., 1987. La Réflexivité Ergative an Français Moderne. Le Français Moderne, 55 (1), 23-

52. Zribi-Hertz, A., 1993. On Stroik’s analysis of English middle constructions. Linguistic Inquiry 24, 583-

589. Zubizarreta, MA., 1982. On the Relationship of the Lexicon to Syntax, Dissertation, MIT, Cambridge,

Mass. Zubizarreta, MA., 1987. Levels of Representation in the Lexicon and in the Syntax. Foris, Dordrecht.