GGS ● Freie Universität Berlin ● 7.-9. Mai 2010

Zum Passiv reflexiver Verben

und den Theorien der Reflexivierung

Florian Schäfer Universität Stuttgart

florian@ifla.uni-stuttgart.de

1. The phenomenon • Passives of reflexive and reciprocal verbs (PoRs):1 (1) Zuerst wird sich geküsst, später dann geheiratet first becomes REFL kissed, later then married ‘First people kiss each other, then they marry’ • PoRs seem to be rare, crosslinguistically. To my knowledge, they are reported for three

languages: German, Icelandic (see below) and Lithuanian (Geniušien÷ 1987, Wiemer 2006).2

1.1 A first characterisation of PoRs (2) a. Hier haben die Römer sich gewaschen (active) Here have the.NOM Romans REFL.ACC washed ‘Here, the Romans washed’ b. Hier wurde sich (von den Römern) gewaschen (passive) Here was REFL.CASE? (by the Romans) washed (i) The external argument disappears. (ii) The above examples lack a (overt) nominative DP; most (though not all) PoRs are, therefore, instances of ‘impersonal passives’.3 (iii) The reflexive pronoun does not change its shape. What case does it have? Clearly, it

does not shift to nominative as referential objects would.4 It seems to keep accusative. � Two theoretical questions pop up immediately:

� How is Principle A of the Binding Theory satisfied?

� What about Case/Burzio’s Generalization?

1 I concentrate on passives involving reflexive pronouns such as German ‘sich’ which can have a reflexive and a reciprocal reading. PoRs can also be found with reciprocal pronouns such as German ‘einander’ (each other). I have not investigated the latter cases in any detail. 2 For English, the example in (i) is mentioned in Baker, Johnson & Roberts (1989). Clearly, the phenomenon in English is much more restricted than in the languages to be discussed. (But note that the predicate ‘keep to X’ is inherently reflexive; see the discussion below).

(i) Such privileges should be kept to oneself. 3 German, Icelandic and Lithuanian have PoRs which are not impersonal (cf. (i)) and/or which are unproblematic for case theory but not for binding theory (cf. (ii)). See section 6.3.

(i) Da wurde wieder viel Unsinn von sich gegeben. there was again many nonsense.NOM from oneself given (People uttered lot’s of nonsense)

(ii) Anschließend wurde sich ein Hamburger gekauft Afterwards was REFL.DAT a.NOM hamburger bought 4 As in most languages, the paradigms of reflexives lack a nominative form in German, Icelandic and Lithuanian (cf. Everaert 1990, Woolforth 1999).

2

• Note that, in the active, (German) reflexives clearly carry ordinary Case. (3) a. Ich wasche mich b. Du wäschst dich c. Er wäscht sich

I wash me.ACC you wash you.ACC he washes REFL.ACC

(4) a. Hans wäscht sich wie kein anderer/wie keinen anderen John.NOM washes REFL like no-one.nom other/ like no-one.ACC other

b. Hans hilft sich wie kein anderer/wie keinem anderen John.NOM helps REFL like no-one.NOM other/ like no-one.DAT other

� The reflexives in the passive should have Case, too; there is no reason whatsoever to think that they are exempt from Case Theory just iff they are in the passive.5 • But the above tests do not really work in PoRs. Claim: (5) is out for formal reasons (i.e., (default) agreement; see below). Claim: (7b) - (10b) are marked for semantic reasons (While the nominative comparison opposes two reflexive uses of the predicates involved, the object comparison opposes a reflexive use and a disjoint use. The latter is incompatible with the observation discussed below that the events in PoRs are interpreted as naturally reflexive). (5) *Es wurde mich/dich gewaschen it was me.ACC/you.ACC washed (6) a. ?Dann wurde wie ein Profi getanzt then was like a.NOM professional danced (7) a. ?Es wurde sich wie ein Gentleman benommen

b. *Es wurde sich wie einen Gentleman benommen it was REFL like a.NOM/ACC gentleman behaved (8) a. ?Es wurde sich wie ein Gentlemen gekleidet

b. ?(*)Es wurde sich (genau so) wie einen Gentlemen gekleidet it was REFL (exactly) like a.NOM/a.ACC gentlemen dressed (9) a. ??Dann wurde sich wie ein Gentleman rasiert

b. ?(*)Dann wurde sich (genau so) wie einen Gentleman rasiert it was REFL like no-oneNOM/ACC washed (10) a. ??Hier wird sich wie Freunde geholfen

b. ?(*)Hier wird sich (genau so) wie Freunden geholfen here was REFL like no-one.NOM/DAT helped • German PoRs were ignored for a long time in the theoretical literature. It seems that the

theoretical challenges provided by this construction concerning Case Theory and Binding Theory lead many theoretically oriented scholars to a categorial rejection of PoRs (e.g. Reis 1982:20f, Haider 1985, Kiss 2003:fn 15, Bierwisch 2006).

5 German ‘sich’ can occur only in accusative or dative positions; in genitive positions, a 3rd person pronoun is used. Once again, this suggests that ‘sich’ has case.

3

• Other authors recognized that this view is not generally correct, and more concretely, they acknowledged that the formation of PoRs is a productive option provided by the

grammar of German. First examples and preliminary discussions of PoRs can be found e.g. in Wunderlich (1985:222), Abraham (1986), Fanselow (1987), Sells, Zaenen & Zec (1987), Grewendorf 1988, or Frey (1993). Plank (1993) and Vater (1995) are the first to investigate PoRs in more depth; they show that PoRs are, in principle, accepted among speakers of German; (see also Ágel 1997, Hundt 2007; see Müller 1999 for a HPSG-analysis).

• PoRs are not substandard: I heard the following examples in the evening news (20

o’clock Tagesschau) of the public-law television (ARD): (11) Während sich heute über Sonne gefreut werden konnte, while REFL today about sun pleased become could, muss morgen mit Regen gerechnet werden must tomorrow with rain calculated become ‘While, today, the sun could be enjoyed, for tomorrow, rain must be expected.’ (12) Bei der Kieler Koalitionskrise wird sich schon gar nicht mehr bemüht, at the Kieler.adj coalition-crisis becomes REFL already really no longer bestirred die taktischen Mätzchen zu verstecken the tactical tricks to hide ‘Politicians involved in the coalition crisis in the regional capital Kiel do not even try (bestir themselves) any longer to hide the tactical tricks.’ • PoRs are very frequent in spoken language, but they are also found in written texts,

i.e. newspapers. The following example is from “Spiegel online”, 16.8.2009. (13) Bei der ARD wurde sich eiligst für den Fauxpas entschuldigt At the ARD was REFL hastily for the faux-pas apologized ‘The people responsible at the ARD hastily apologized for the faux pas’ • Contrary to what has occasionally been claimed PoRs are not just used to convey

requests. They occur frequently in declaratives and interrogatives (Plank 1993).

• PoRs are not a recent invention. Behagel (1924, II:214) (one of the earliest sources of examples) provides some middle-high German examples.

2. A semantic/conceptual restriction on the formation of PoRs • Languages with a two-form reflexive system (e.g. Dutch) make a morphosyntactic

distinction between three classes of reflexive verbs. The morphological distinction reflects a semantic difference (see Kemmer (1993) and the references there).

A: Inherently reflexive verbs: The reflexive pronoun cannot be replaced by a referential DP. Only the simple reflexive is allowed:

(14) Jan schaamt zich/*zichself/*Marie John shames REFL/REFL-SELF/Mary ‘John is ashamed’

4

B: Naturally reflexive verbs (and naturally reciprocal verbs): The reflexive pronoun can be replaced by a referential DP. Out of the blue, the simple reflexive is strongly preferred. The complex reflexive is possible in the case of strong focus.

(15) a. Jan waste zich/??zichself/Marie b. Jan scheerde zich/??zichself/Peter John washed REFL/ REFL.SELF/Mary John shaved REFL/REFL.SELF/Peter ‘John washed (Mary)’ ‘John shaves (Peter)’

Naturally reflexive verbs come from a number of semantic subclasses which all represent events which carry

“inherent in their meaning [..] the lack of expectation that the two semantic roles they make reference to will refer to distinct entities” (Kemmer 1993:58).

So called “grooming verbs” such as ‘shave’, ‘wash’ and ‘dress’ form one main subgroup of naturally reflexive verbs; other subgroups will be mentioned below.6

C: “Naturally disjoint verbs” (also called “other directed verbs” (e.g. König & Vezzosi

2004): The reflexive pronoun can be replaced by a referential DP. The complex reflexive is strongly preferred. These verbs express events which carry the expectation that the two semantic roles they make reference to will refer to distinct entities. (e.g. ‘hate’, ‘accuse’, ‘kill’, ….).

(16) Zij haat ??zich/zichself/Peter She hates REFL/REFL-SELF/Peter ‘John hates himself/Peter’ • In German the simple reflexive pronoun ‘sich’ can be used in all three contexts. The

addition of the intensifier ‘selbst’ (self) is never obligatory. This means that German does

not make a (obligatory) morphological difference between inherently/naturally reflexive verbs and naturally disjoint verbs.

(17) a. Hans hasst sich/Maria (naturally disjoint) John hates REFL/Mary b. Hans wäscht sich/Maria (naturally reflexive) John washes REFL/Mary c. Hans schämt sich/*Maria (inherently reflexive) John shames REFL/Mary

Claim: The same lexical-semantic restrictions that determine the distribution of simple and complex reflexives in Dutch determine the formation of PoRs: PoRs are mainly formed with inherent reflexive and naturally reflexive verbs/predicates; naturally

disjoint verbs are very rarely found.

• Such a semantic restriction has already been suggested in Abraham 1986, Sells & al. 1987, Kaufmann 2001 and especially Ágel 1997, but it was never empirically corroborated.

6 I will argue that inherent/natural reflexivity is not confined to the verbal co-argument domain; I will therefore talk about inherently/naturally reflexive predicates, vPs or events.

5

• Plank (1993) and Vater (1995) provide the first and only questionnaire studies about PoRs. These authors do not mention the above generalization, but they observe that different examples of PoRs receive quite different acceptance.

• Google query: In order to test the above claim, I carried out a Google query: Kemmer (1993: chapter 3, 4) divides the class of naturally reflexive/reciprocal verbs into a number of semantic subclasses. For each of these subclasses, I selected one verb and checked whether it forms a PoR. The result was positive; i.e. for each subclass, already the first verb that I selected lead to a number of matches. As a comparison, I checked for a number of naturally disjoint verbs whether they form reflexive passives. It turned out that these verbs hardly ever form such passives. I. grooming verbs: e.g. ‘sich waschen’ (to wash)

II. verbs of change in body posture: e.g. ‘sich hinsetzen’ (to sit down) III. verbs of non-translational motion: e.g. ‘sich verbeugen’(to bow) IV. verbs of translational motion: e.g. ‘sich bewegen’ (to move)

V. verbs of antagonistic events: e.g. ‘sich prügeln’ (to trash)

VI. verbs of affectionate actions: e.g. ‘sich küssen’ (to kiss)

VII. social/encountering actions: e.g. ‘sich treffen’ (to meet)

VIII. verbs of interlocution: e.g. ‘sich unterhalten’ (to talk to s.o.)

IX. verbs of emotion: e.g. ‘sich wundern’ (to wonder)

X. emotive speech actions: e.g. ‘sich beschweren’ (to complain (about))

XI. simple cognitive events: e.g. ‘sich überlegen’ (to consider)

XII. verbs of perception: e.g. ‘sich anschauen’ (look at) XIII. intentive verbs: e.g. ‘sich wünschen’ (to desire) XIV. ordinary transitive and ditransitive verbs (naturally disjoint verbs): sehen (to see): e.g. “sich (selbst) gesehen wurde” - 0 matches.

loben (to praise): e.g. “wird einander gelobt” - 0 matches.

beschuldigen (to accuse): e.g. “wurde sich (selbst) beschuldigt” - 0 matches. angreifen (to attack): e.g. “einander angegriffen wird” - 0 matches.

beladen (to load): e.g. “sich beladen wurde” - 0 matches

geben (to give): e.g. “wurde sich (selbst) gegeben” - 2 matches. • Conclusion: The claim by Ágel (1997) that only (mainly) inherently reflexive and

naturally reflexive verbs form PoRs could be further supported.7

• Potential problem: The above findings could be an artefact: (i) By definition, naturally disjoint verbs are rarely used in a reflexive construal. (ii) Actives are much more frequent than passives. (iii) As a result, the PoRs of naturally reflexive verbs could simply be very rare in the corpus.

7 Hundt (2007) provides a list of real-world examples of PoRs containing 95 modern and 26 historical examples. 7 of his examples involve verbs which probably are naturally disjoint verbs. Interestingly, counterexamples to the above generalization very often contain the adverb gegenseitig (mutually) as in (i). Other counterexamples turn out to be from specialized communities or to involve irony (“as if the event were naturally reflexive”). (i) Hier wird sich nicht gegenseitig umgebracht. Here is REFL mutually killed

6

Questionnaire study: 48 Target sentences: judged on a scale from 1-7 (acceptable – not acceptable). 4 inherently reflexive verbs + 4 inherently reciprocal verbs (active and passive) (18) a. Auf dem Spielplatz wird sich dann richtig ausgetobt. b. Auf dem Spielplatz toben sie sich dann richtig aus. ‘At the playground they let off steam.’ 4 naturally reflexive verbs + 4 naturally reciprocal verbs (active and passive) (19) a. Auf dem Volksfest wurde sich wieder heftigst geprügelt. b. Auf dem Volksfest prügelten sie sich wieder heftigst. ‘At the fair they beat each other heavily.’ 4 naturally disjoint verbs in a reflexive use + 4 naturally disjoint verbs in a reciprocal use (a,p) (20) a. Oftmals wird sich durch die enge Startaufstellung behindert. b. Oftmals behindern sie sich durch die enge Startaufstellung. ‘Often they impede each other due to the close start-positions.’

� The target sentences were arranged in two questionnaires a 24 sentences so that no verb was presented twice. Both questionnaires were filled with the same 48 distractors.

� Each questionnaire was presented in two randomized orders. � Each sentence was read by 24 speakers.

Reflexive and reciprocal verbs

1234567

active 1,56 1,87 1,68 1,66 2,26 1,79

passive 2,67 3,21 3,1 3,19 5,21 4,66

i-ref i-rec n-ref n-rec nd-ref nd-rec

Reflexives and reciprocals combined

1

234

56

7

active 1,71 1,67 2,03

passive 2,94 3,14 4,93

inherent refl. / reciproc. naturally refl. /reciproc. naturally disjoint

Conclusion: Naturally disjoint verbs are clearly and strongly dispreferred in PoRs.8

8 The individual subjects showed very variable behavior. Some rejected most PoRs; some accepted most PoRs.

7

3. PoRs in Icelandic9

• PoRs are first mentioned by Sigurðsson (1989:355, fn. 60). Eythórsson (2008) judges the

examples below as totally acceptable (no question mark) and adds that most speakers he consulted agreed with him (the first verb is inherently reflexive, the second naturally reflexive).

(21) a. Börnin leika sér allan daginn the.children play REFL.DAT all the.day ‘The children are playing all day’ b. (?)Það var leikið sér allan daginn expl. was played REFL.DAT all the.day (22) a. Fólkið baðaði sig á laugardögum the.people bathed REFL.ACC on saturdays ‘The people took a bath on saturdays.’ b. (??)Það var baðað sig á laugardögum expl was bathed REFL.ACC on saturdays Note: Icelandic reflexive pronouns have an overtly case-inflected paradigm (sig-ACC, ser-DAT, sin-GEN) (the same holds e.g. for Romanian). • The questionnaire study Maling & Sigurjónsdóttir (2002) shows that not all PoRs receive

the same acceptance. • Claim: The lexical semantic restriction identified for PoRs in German holds in Icelandic,

too; Icelandic PoRs are best with inherently and naturally reflexive verbs. • Examples with simple reflexives: (23) a. Svo var bara drifið sig á ball Then was just hurried REFL to the.dance (Elsewhere 78% | Inner Rvík 67% | Adults 40%) b. Það var haldið sig innan dyra út af óveðrinu it was kept REFL in doors due to bad.weather (Elsewhere 82% | Inner Rvík 65% | Adults 37%) c. Það var skoðað sig um á svæðinu it was looked REFL around in the.area (Elsewhere 72% | Inner Rvík 43% | Adults 31%) � (23a) - (23c) involve inherently reflexive verbs. 9 Side remark on the relation between Icelandic PoRs and the ‘new passive’: Icelandic PoRs are typically

discussed in the context of the ‘new passive’. The recent literature concluded that the new passive is a real passive in that the external argument is absorbed although the (even definite) internal argument remains in situ and keeps accusative (Eyþórsson 2008, Jónsson 2009, Sigurðsson (to appear)), not an active impersonal (cf. Maling & Sigurjónsdóttir 2002). While there is arguably a relation between the new passive and PoRs (in that the latter might have grounded the way for the development of the former), the recent literature concluded that PoRs are NOT genuine instances of the new passive. Eythórsson (2008:189) mentions that the PoR “seems to be an innovation of Modern Icelandic which is increasingly gaining ground and is accepted by many speakers who do not accept the NC [New Construction, i.e. the new passive] with non-reflexive verbs”. The last conclusion can already be drawn from the questionnaire study by Maling & Sigurjónsdóttir (2002); PoRs receive much more acceptance than new passives without a reflexive pronoun.

8

• Examples with complex anaphor: (24) a. Það var horft á sjálfan ig í speglinum it was looked at SELF REFL in the.mirror (Elsewhere 58% | Inner Rvík 48% | Adults 34%) b. Það var bent á sjálfan sig á myndinni it was pointed to SELF REFL in the.picture (Elsewhere 19% | Inner Rvík 11% | Adults 13%) � The two examples are syntactically identical. There is a clear difference in acceptance. The predicate (vP) in (24a) is naturally reflexive; this is clearly not the case in (24b). • Examples with possessive reflexives: (25) a. Það var haldið með sínu liði It was held with SELF's team People supported their own team (Elsewhere 63% | Inner Rvík 49% | Adults 36%) b. Í morgun var hrint systur sinni af hjólinu this morning was pushed sister SELF's off the.bike (Elsewhere 13% | Inner Rvík 7% | Adults 2%) c. Það var klippt hárið á dúkkunni sinni it was cut the.hair on doll SELF's (Elsewhere 5% | Inner Rvík 2% | Adults 2%) ���� (25a) involves a naturally reflexive predicate. (25b-c) are clearly NOT naturally reflexive. NOTE: (25a) makes it very clear that natural reflexivity cannot be computed within the

co-argument domain of a verb. � Conclusion 1: The same semantic/conceptual restriction holds in German and Icelandic. � Conclusion 2: Icelandic gives us a clear clue for the case-question. Direct-object reflexives keep the object case under passivization. 4. PoRs in other languages: Dutch and Norwegian • In other languages, PoRs are claimed to be totally ungrammatical. Since most passives of

reflexive verbs are instances of “impersonal passives”, we only look at two languages that have an impersonal passive in the first place, Dutch and Norwegian.

4.1 Dutch

• PoRs are typically judged as totally ungrammatical in Dutch (p.c. Martin Everaert).

(26) *Er werd zich gewassen (Reinhard & Siloni 2004) There was REFL washed10 10 Jan-Wouter Zwart (p.c.) agrees that this example is totally out. But he judges (i) as “not so bad” and (ii) as “not strikingly bad, but somewhat uneasy.” Note that (i) involves an inherently and idiomatic reflexive

9

4.2 Norwegian

• The examples in (27) are from Maling & Sigurjónsdóttir (2002). My informants (Terje Lohndal, Torgrim Solstad, Kristi Kroch Christensen) find these and other examples from Åfarli (1992) very bad. Most of the examples involve inherently or naturally reflexive predicates; we see no improvement.11

(27) a. *Det ble hygget seg It was amused REFL b. *det ble låst seg (selv) inne i fabrikken It was locked REFL (self) inside in the.fatory c. *Det ble lyttet til hverandre på møtet It was listened to one.another at the.meeting � Conclusion: Dutch and Norwegian differ from German and Icelandic in that they do

not form (impersonal) PoRs (For some complications, see fn. 10 and 11). 5. Theoretical evaluation

5.1 Four questions raised by PoRs Q1: How can the reflexive pronoun stay in the accusative case? (Burzio’s Generalization) Q2: What is the antecedent of the reflexive pronoun? (Principle A of the Binding Theory) Q3: Why are PoRs acceptable only with inherent/naturally reflexive verbs but not with naturally disjoint verbs? (In the active, the three verb classes behave the same!) Q4: Why don't we find PoRs in other (Germanic) languages (especially, if they have a similar anaphoric system involving SE-anaphors).

construction; (ii) involves an inherently reflexive predicate. Note furthermore that (i) and (ii) are not impersonal passives as they have a nominative DP. The significance of this difference should be checked with more examples and more speakers. (i) Overdag werd er gewerkt, maar 's avonds werd er zich tegoed gedaan daytimes became there worked but evenings became there REFL well-done aan de heerlijkste spijzen en dranken. on the most.delicious foods and drinks ‘By day, people work; but in the evenings, people enjoy delicious foods and drinks’ (ii) Op de workshop moest er weer een hoop geklets tot zich genomen worden on the workshop must.PAST there again a heap nonsense to REFL taken become ‘On the workshop one had to take in a lot of rubbish again.’ 11 Tor Åfarli (p.c.) insists that “the following is perfectly o.k. in colloquial Norwegian”. Terje Lohndal (p.c.) agrees that this example is better. Note that (i) is not necessarily an impersonal passive as it could involve a nominative (beer). But the same holds for other examples in Åfarli (1992) which received bad judgements. (i) Vi var på puben i går kveld, og det ble kjøpt seg øl over en lav sko We went to the.pub in last night and it was bought REFL beer all the time “We went to the pub last night, and it was bought SELF beer all the time.”

10

5.2 Evaluation of theoretical assumptions and frameworks

♦ add Q2:

• One could suggest that PoRs are hidden transitives with a covert external argument (Maling & Sigurjónsdóttir 2002 for Icelandic). This would solve Q1 and Q2.

• One could suggest that the implicit argument of passives can antecede the reflexive

(Barðdal & Molnar 2003, Sternefeld 2006, Sigurðsson to appear). This would solve Q2. Both hypotheses cannot answer Q3, the second one cannot answer Q4. German provides a further counter-argument: • German impersonal passives easily combine with agentive by-phrases which can even

introduce 1st/2nd person pronouns. (28) a. Von mir wurde nicht gelacht by me was not laughed b. weil von dir zu viel gelabert wird because by you too much babbled becomes • 1st and 2nd person by-phrases are possible with PoRs, too. In the active, subject and

anaphor agree in Person/Number (29a). Crucially and unequivocally, no agreement

between by-phrase and anaphor is possible in PoRs (29b); instead, the 3rd person reflexive is obligatory (Plank 1993).

(29) a. Nur wir waschen uns / *sich hier täglich (active) only we wash us.ACC / REFL here daily b. Nur von uns wird sich / *uns hier täglich gewaschen (passive) only by us is REFL/ us.ACC here daily washed � I conclude that the reflexive in PoRs does not have an antecedent in the syntax at

all.12 ♦ The “the unaccusativity analysis” of reflexive verbs: • Lexical version: The external theta role is absorbed/reduced/bundled in the lexicon; only

the internal argument is projected to the syntax. (The reflexive morphology is a sign that the lexical operation has taken place).

• Syntactic version (Marantz 1983, McGinnis 1998, 2000, 2004, Embick 2004): The reflexive pronoun is merged in the external argument position; the full DP is merged in the internal argument position. In order to fulfil Principle A of the Binding Theory, the internal argument A-moves to a position c-commanding the anaphor. (The correct case-pattern emerges through specific assumptions).

Both analyses are obviously untenable for languages with PoRs.13

12 An alternative, not implausible, idea is that the existentially bound implicit argument of passives is always 3rd person (or impersonal) and that by-phrases do not have to match the implicit argument in phi-features. But this would leave it unexplained why naturally disjoint verbs do not form PoRs.

11

♦ The “Bundling” approach to reflexive verbs (and other intransitivity approaches14

):

• Reinhard & Siloni (2005) assume that all types of reflexive verbs (inherently reflexive

verbs, naturally reflexive verbs and naturally disjoint verbs in a reflexive use) are derived by a process bundling an internal with an external theta role.

• As a result, only one argument with a complex theta-role is merged in the syntax. The reflexive element is NOT an argument of the verb. It is either merged as a sign that bundling has taken place or as a case-reducer absorbing object-case. Motivation for this theory comes from the well known tests (Kayne 1975, Grimshaw 1981) that suggest that reflexive verbs are intransitive in French.

(30) Bundling (Reinhard & Siloni 2005: 400): [θi] [θj] � [θi - θj], where θi is an external theta role. • Bundling takes place in the lexicon (e.g. English) or in the syntax (e.g. French, German). (31) Bundling in the lexicon: The case feature is reduced before syntax. a. Max washed b. Verb entry: washacc[Agent] [Theme] c. Reflexivization output: wash[Agent-Theme] d. Syntactic output: Max[Agent-Theme] washed. e. Interpretation: ∃e [wash (e) & Agent(e, Max) & Theme(e, Max) (32) Bundling in the syntax: The case feature is absorbed by the reflexive element. a. Jean se lave Jean REFL washes b. VP: [se laveθi-Agent, θk-Theme] c. IP: [Jean<θi, θk> [se lavej [VP tj ]]] d. Interpretation: ∃e [wash (e) & Agent(e, Max) & Theme(e, Max) � Their account seems to answer Q2; the reflexive is not an anaphor and, therefore, not subject to the Binding Theory. The account cannot answer Q3 and Q4. Furthermore, it cannot account for Q1; the accusative should be absorbed by the passive morphology so that there should be no need to insert a case-absorber. • Doron & Rappaport Hovav (2007) (see also Labelle 2008) provide a collection of

arguments against the Bundling approach for so-called syntax-languages, and especially, against the claim that reflexive verbs are intransitive in these languages.

• I will give here only one of their arguments which is originally from Labelle (2008): If the

external and the internal theta role are bundled and assigned to one DP, it should be impossible to focus only one of them; especially, it should be impossible to focus just the

13 Reinhard & Siloni (2004) provide a number of empirical arguments against the lexical version. The syntactic version is untenable for theoretical reasons; there is no way (besides stipulation) to avoid that the binding of the reflexive by the moved internal arguments leads to a situation of Lethal Ambiguity (the latter also discussed and well motivated by McGinnis 1998, 2004). 14 E.g. Sells et al. 1987, Ágel 1997, Hundt 2006, Eythórsson 2008.

12

internal argument of a reflexive verb just as it is impossible to focus the unrealized object in ‘John ate’. However, in languages which form reflexive verbs with reflexive morphology, both arguments can be focused independently.

(33) Jean-Pierre s’est dénoncé lui-même Jean-Pierre REFL is denounced himself (i) ‘Jean-Pierre denounced himself, it was not others who denounced him (ii) ‘Jean-Pierre denounced himself, he did not denounce others (34) Jean-Pierre ne se rase pas lui- même Jean-Pierre NEG REFL shaves not himself (both readings available) (35) Am Morgen wäscht sie sich immer SELBER at morning washes she REFL always self (i) agent focus: She washes herself, no-one else washes her. (She is a disabled patient) (ii) theme focus: She washes herself, she washes no-one else. (She is a nurse) � Doron & Rappaport-Hovav (2007) conclude that French ‘se’ is an anaphoric object

clitic bearing the internal θ-role (see also Alencar & Kelling 2005). � Reflexive verbs are not intransitive but transitive. The well-known data by Kayne

(1975)/Grimshaw (1981) need a different explanation (see Doron & Rappaport Hovav (2007), Labelle (2008) or Alencar & Kelling (2005) for different proposals).

♦ Lexical Reflexivization: • Doron & Rappaport Hovav (2007) argue that reflexive verbs involving SE-anaphors have a

syntactic derivation. Specifically, the reflexive use of naturally disjoint verbs is derived in the syntax by anaphoric binding of the reflexive object.

• But these authors assume that there exist, in addition, the lexical operations “Reflexivization” and “Reciprocalization”.

• These operations identify an internal θ-role with an external θ-role. The result is an intransitive lexical entry which projects one complex θ-role. When this intransitive entry is projected to syntax, a lexical marker of reflexivization is added.

• Inherently reflexive verbs are always derived in the lexicon. Naturally reflexive verbs have a syntactic and a lexical derivation.

• These authors explicitly assume that in all languages with SE-anaphor there exists a syncretism between the anaphor and the marker of reflexivization.

� While Q1 and Q4 remain open, their proposal seems to answer Q2 and Q3: PoRs involve a marker of Reflexivization, not an anaphor. Problem 1: If PoRs could only be derived on top of the lexical derivation, it should be

impossible to focus the two θ-roles independently. This prediction is NOT borne out. Both, the implicit external argument as well as the internal argument can be focused independently via the addition of an intensifier. (Adverbs and the form of the intensifier lead

13

to different preferences; see Hole (2006, 2008) on modification of the implicit argument of passives via intensifiers (e.g. selber/selbst (self)). (36) a. Morgens wäscht sie sich immer/erst mal SELBER at.morning washes she REFL always/first-of-all self (� agent focus, � theme focus) b. Morgens wird sich immer/erst mal SELBER gewaschen at.morning becomes REFL always/ first-of-all self washed (� agent focus, � theme focus) (37) a. Morgens werden sich die Zähne selbst geputzt, At.morning become REFL.DAT the teeth.NOM self brushed am Abend macht das die Mama at.evening makes this the mama b. Erst wird sich selbst gewaschen, dann die Kinder First becomes REFL washed, then the.ACC children (!accusative on ‘kinder’!) Problem 2: Doron & Rappaport-Hovav (2007) take Reflexivization to semantically identify the external with an internal θ-role of a verb. But both, the domain of inherent reflexivity

and the domain of natural reflexivity exceed the verbal coargument domain.

• Recall that the Icelandic PoR with a possessive reflexive in (38) received very high acceptance. I suggested this to be so because ‘to support X's team’ is naturally reflexive.

(38) Það var haldið með sínu liði it was held with SELF's team ‘People supported their own team’ No lexical process can relate the external theta role of the verb with the possessor of the object DP. • Similarly, German (as other Germanic languages) has many inherently reflexive predicates

where the reflexive is embedded in and θ-marked by a non-subcategorized PP. An example is the semi-idiomatic string “etwas von sich geben” (sth. from oneself give). Such strings passivize very productively.

(39) Auf der Konferenz wurde viel Unsinn von sich gegeben At the conference was many nonsense from REFL given ‘People uttered lots of nonsense at the conference’ � The phenomenon of inherent reflexivity is not (cannot be) restricted to the coargument

domain of verbs. If one wants to argue that inherently reflexive structures are derived by a lexical process, one has to assume that this lexical process can apply to [verb + PP] complexes, certainly not the standard assumption.15

15 Similar data led Reinhart & Siloni (2004) to the assumption of a syntactic version of bundling.

14

6. Towards an analysis of PoRs

• General assumption:

� Whenever we see a SE-anaphor, there is a SE-anaphor. The reflexive element in all

types of reflexive predicates (inherently reflexive, naturally reflexive and naturally disjoint predicates) is an anaphor; it’s either the (in-)direct object of a verb, the object of a preposition or the possessor of a DP.

� Natural reflexivity and natural disjointness (and at least some aspects of Inherent reflexivity such as in (39)) are a post-syntactic phenomenon determined at the C-I interface. It is computed on the basis of information within vP (phase-level).

• Recall: Above I claimed that the anaphor in PoRs has no syntactic antecedent. If the implicit agent of the passive would act as an antecedent, we could not explain why naturally disjoint verbs do not form PoRs. Furthermore, the following agreement facts would be problematic.

(40) a. Nur wir waschen uns / *sich hier täglich (active) only we wash us.ACC / REFL here daily b. Nur von uns wird sich / *uns hier täglich gewaschen (passive) only by us is REFL/ us.ACC here daily washed � The traditional version of Principle A of the Binding Theory is not fulfilled in PoRs. Claim: Some languages can repair a violation of Principle A, but the formally repaired structure is acceptable (easily interpretable) only iff the underlying predicate/event is conceptualized as inherently or naturally reflexive. The distribution of PoRs (≈ The availability of Repair for Principle A violations)

(formally) ungrammatical (formally) grammatical (semantically) acceptable

PoRs in all languages (all types of verbs)

PoRs in German/Icelandic

(all types of verbs)

PoRs with

inherently and

naturally reciprocal verbs

Dutch

Norwegian

15

6.1 Basic assumptions on binding, case and agreement (Schäfer 2008, to appear): Note: I leave Phase Theory aside for reasons of space.

• Anaphoric binding is grounded in a syntactic Agree-relation between a DP-antecedent and a variable (e.g. Burzio 1998, Reuland 2001, Fischer 2004, Heinat 2006, Chomsky 2008, Kratzer 2009).

� A variable is totally underspecified with respect to φ-features; it is a set of a

categorical D-feature and unvalued φ-features {D, uφ}. Therefore, it is

referentially defective. Furthermore, a variable needs an antecedent to value its

features under Agree; Agree is local (-> This replaces Principle A). � A variable is active due to its unvalued φ-features; it acts as a probe, the antecedent

as a goal.

� Proposal: Variables probe upwards (cf. Baker 2008 for the option of upward-probing). � At LF, the syntactic Agree-relation is evaluated semantically as a binding relation (as

expressed via coindexation in earlier stages of the theory, or λ-abstraction; see Reuland 2001 for a discussion that Violation of Inclusiveness is inevitable).

� At PF, the Agree relation is evaluated morpho-phonologically and the variable is spelt

out in a language specific way (Halle & Marantz 1993, Heinat 2006). The specific spell out (as a reflexive or as a (locally bound) pronoun) depends on the φ-features of the antecedent and language-dependent morpho-phonological economy conditions (Burzio 1998).

� Full fledged referential pronouns are the combination of a D-feature and a set of

interpretable φ-features {D, φ} and always spell-out their inherent φ-features.

� Morphological case and verbal agreement are mainly PF phenomena (Marantz (2000), McFadden (2004), Sigurðsson (2003, 2005, 2006, 2009). Yet, PF evaluates

syntactic Agree-relations to determine case and agreement.

� Structural case is established on the basis of information computed inside TP (or

alternatively, vP/VoiceP (Sigurðsson 1989, 2000, 2003, Schäfer 2008, to appear)). T is equipped with unvalued phi-features and searches its c-command domain for the closest valued DP.

� Dependent case (ACC): A DP is realized at PF with dependent case if a different DP

has valued local T via Agree. � Default case (NOM): A DP which is not realized with dependent case appears with

default case. (Inherent/lexical case takes precedence over default and dependent case.)

• A derivation involving ordinary subject-object binding: (41) a. daß Hans sich mag

that John.NOM REFL.ACC likes ‘that John likes himself’

16

b. TP 3

T vP {uP, uN, uG} 3 Hans{P, N, G} v’ 3 v VP 3 sich{uP, uN, uG} V

• T enters the derivation with unvalued φ-features. The variable enters the derivation with unvalued ϕ-features. The subject is merged with a full set of valued ϕ-features.

• T probes its c-command domain. The closest element with valued φ-features is the external

argument in Spec,vP which therefore agrees with and values T. • The variable probes the tree upwards for an antecedent. The subject values the variable. • These Agree-relations will be evaluated at the interfaces.

� At PF, the internal argument is marked with dependent ACC because there is a different DP (the subject) within the same clause/phase which has valued the features on T via Agree. The subject gets default case.

� The variable is spelled out according to the phi-features of its antecedent.

� At LF, the Agree-relation between the subject and the variable is interpreted as

semantic binding.

6.2 Deriving PoRs: a proposal

(42) als sich gewaschen wurde when REFL washed became Q1: How can the reflexive pronoun stay in the accusative case? (Burzio’s Generalization) Q2: What is the antecedent of the reflexive pronoun? (Principle A of the Binding Theory) Q3: Why are PoRs acceptable only with inherent/naturally reflexive verbs but not with naturally disjoint verbs? Q4: Why don't we find PoRs in other (Germanic) languages (especially, if they have a similar anaphoric system involving SE-anaphors)

• Recall: I claimed that the anaphor in PoRs has no syntactic antecedent. Furthermore,

PoRs (often) lack any DP with valued φ-features because they are impersonal constructions.

17

• The derivation of PoRs - I:

(43) TP 3 T vPpassive

{uP, uN, uG} ! v’ 3 v VP 3 DP{uP, uN, uG} V • Both, the variable and T enter the derivation with unvalued φ-features. T searches its c-

command domain. It agrees with the variable (forming the chain < T – variable >, but no valuation can take place.

• Expectation: A crash at the interfaces. This looks like the correct result for Dutch or Norwegian; but it isn’t for German and Icelandic.

• Note that we have a similar situation/problem with ordinary impersonal passives (or

quirky intransitives in Icelandic). (44) a. weil (*es) hier gestern lange getanzt wurde (German) because expl. here yesterday long danced was b. ĺ gœr hefur (*það) verið dansað (Icelandic) yesterday has expl. been danced c. TP 3 T{uP, uN, uG} vPpassive ! v’ 3 v VP ! ? V • There is no DP available to value Tense. We have an unvalued one-member chain < T >

which should lead to a crash at the PF-interface, contrary to fact. • Impersonal passive strategies:

� Expletive insertion: Det (Norwegian 2) is a fully specified nominal category (neuter, singular). It checks EPP on Tense and triggers participle agreement. Der (Norwegian 3) is not specified for number and gender. It simply checks EPP on Tense and does not trigger agreement (Holmberg 2002).

18

(45) a. Nokre gjester er komne/*kome Some guests is come.PL/come-N.SG b. Det er kome/ *komne nokre gjester (Norwegian 2; Holmberg 2002) EX is come-N.SG/ come.PL some guests c. Der er komne / *kome nokre gjester (Norwegian 3; Holmberg 2002) EX is come.PL / come.SG some guests

� Default agreement: In the absence of a nominal category (Norwegian 3, German Icelandic), impersonal constructions can be rescued by Default Agreement (DA).

Proposal: DA is a last resort operation taking place before Transfer. It prohibits a crash by inserting default φ-features into unvalued objects (i.e., a functional head; a chain between functional heads).

The derivation of PoRs – II (46) a. als sich gewaschen wurde when REFL washed became b. TP 3 T{uP, uN, uG} vP ! v’ 3 v VP 3 DP{uP, uN, uG} V • (46) involves an unvalued two-member agreement chain <T0 - DP>. Ad Q2: The anaphor has no antecedent; it gets it’s φ-features valued by DA. Ad Q4: DA is a last resort repair mechanism rescuing unvalued agreement chains: � There are different types of unvalued agreement chains: (i) <head{uφ}> (One member chain) (ii) <head{uφ} – head{uφ}> (Two member chain involving heads) (iii) <head{uφ} – DP{D, uφ}> (Two member chain involving head and DP) Proposal: Rescuing the latter is more complex than rescuing the former; not all languages allow those repair mechanisms. Some languages can realize DA only on (chains of) heads of the extended verbal domain (T, v, V) but not on non-homogeneous agreement-chains involving verbal heads and D-elements like anaphors. Ad Q3: DA avoids a crash at the PF-interface. But the output of syntactic derivations must

be comprehensible at the CI-interface, too. If the syntax does not provide enough information for interpretation, conceptual knowledge might help. � With inherently reflexive predicates, it is conceptually clear that no further referent besides

the external argument can be involved. Naturally reflexive events carry the conceptual

19

expectation that the argument position realized by the unbound reflexive denotes the same entity that acts as the external argument. That is, the semantic coindexation that usually is computed from syntactic binding/agree can be reconstructed on conceptual grounds beyond any doubt.

Ad Q1: Case is a post-syntactic, i.e. morphological phenomenon. We need to update our assumptions about dependent case: � Dependent case (ACC): A DP is realized at PF with dependent case if a different DP

something else (e.g. DA) has valued T via (default) Agree. 6.3 Extensions, Predictions and Problems ♦ PoRs with nominative DPs:

(47) a. Auf der Konferenz wurde viel Unsinn von sich gegeben At the conference was many nonsense.NOM from REFL given ‘People uttered lots of nonsense at the conference’

b. Anschließend wurde sich ein Hamburger gekauft Afterwards was REFL.DAT a.NOM hamburger bought • Upward probing must be allowed to pass an internal argument; otherwise we would get

obligatory binding of the variable by the internal argument. We know from double object verbs that internal arguments are never obligatory antecedents (48).

(48) Hek showed Maryi herselfi/himselfk in the mirror Prediction 1: The above example does not involve default agreement. The derivation, and especially the valuation of the variable is similar to middles and reflexive anticausatives (Schäfer 2008, to appear) that we find in German but also in Dutch or Norwegian; since the variable is valued in the syntax, no repair strategy is necessary. The derivation should be available even in languages that lack Default Agreement for complex agreement chains. Comment: There is at least preliminary indication that such examples involving nominative DPs are judged better in Dutch and Norwegian (fn 10, 11). Prediction 2: The variable should be spelt out according to the phi-features of the object/nominative DP. • This is in principle correct; but see the following data from Plank (1993): (49) Firma Ai will sichi Firma B einverleiben company.NOM A wants REFL.DAT company.ACC B incorporated ‘A incorporated/absorbed B’ (50) The workers of company B say: So einfach verleibt man sich uns nicht ein So easily (in)corporates one.NOM REFL.DAT us not in ‘One does not absorb us so easily’

20

(51) a. *So einfach werden sich wir nicht einverleibt So easily become.PL REFL.DAT we.NOM not incorporated b. *So einfach werden uns wir nicht einverleibt So easily become.PL us.DAT we.NOM not incorporated c. ?So einfach wird sich uns nicht einverleibt So easily become.SG REFL.DAT us.ACC not incorporated d. *So einfach wird uns uns nicht einverleibt So easily become.SG us.DAT us.NOM not incorporated

• Formally, we predict b to be the grammatical version. But the result suggests that there is a semantic binding relation between dative (recipient) and accusative (theme) (We incorporate ourself into ourself). The result is not nonsense and, therefore, a real competitor.

• What we see is a further repair strategy; default agreement for the chain <T – variable>.

The internal argument gets accusative. (See Eyþórsson (2007) on a related effect in Icelandic: In PoRs of double object verbs, the non-reflexive theme (often) keeps accusative).

♦ Languages having expletives with phi-features (Norwegian 2): If the expletive could value

the variable, no default agreement would be necessary and the PoR should be grammatical.

� Option 1: Passive v/Voice is a phase and the expletive is merged outside of the phase.

� Option 2: Since the expletive is fully nominal and c-commands the variable, the valued chain <nominalexpletive – variable> is the input to semantic interpretation; since the expletive has no theta role, no interpretation can be computed.

♦ Overgeneration: • Hopefully, the theory developed above does not predict variables/anaphors to occur in

unwanted positions.

� In general, we must be able to prohibit nominative anaphors; (Everaert 1990, Woolforth 1999). Since variables are unvalued, they can never trigger Agree; therefore, they will not qualify for NOM. The system predicts unbound variables to survive syntactically under specific circumstances. The hope is that syntactically well-formed derivations can be filtered out on conceptual grounds.

� Examples like: “suddenly REFL murdered the president” are filtered out as

conceptually deviant (Alexiadou et al 2006, Schäfer 2008b). � Examples like ‘REFL sleeps’ suffer from empty predication iff an unbound reflexive

pronoun lacks denotation and, in turn, cannot realize a thematic role at the CI-interface (cf. Schäfer 2008, to appear for such a proposal in the context of reflexive anticausatives/middles). (The unbound reflexive in PoRs receives its denotation, and, in turn, its thematic role via the conceptual backup.)

21

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

Linguistik (GAGL) 28, 1-72. Åfarli, T. A. 1992. The syntax of Norwegian passive constructions. Amsterdam: John

Benjamins. 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. Festschrift für Heinz Vater zum 65. Geburtstag (pp. 147-187). Tübingen: Niemeyer.

Alencar, L. F. & C. Kelling. 2005. Are reflexive constructions transitive or intransitive? Evidence from German and Romance. In: Miriam Butt & Tracy Holloway King (Eds.) Proceedings of the LFG05 Conference.

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

Baker 2008. The Syntax of Agreement and Concord. Cambridge University Press Baker, M., K. Johnson and I. Roberts 1989. Passive arguments raised. Linguistic Inquiry 20,

219-252. Barðdal, J. & Molnár, V. 2003. Passive in Icelandic – compared to Mainland Scandinavian. In

Structures of Focus and Grammatical Relations, J. Hetland & V. Molnár (eds), 231-260. Tübingen: Niemeyer.

Behagel, O. 1924. Deutsche Syntax. Eine geschichtliche Darstellung. Volume II. Heidelberg: Winter.

Bierwisch, M. 2006. German Reflexives as Proper and Improper Arguments. In P. Brandt & E. Fuß (eds.), Form, Structure, and Grammar : A Festschrift Presented to Günther Grewendorf on Occasion of his 60th Birthday, 15-36. Berlin: Akademie-Verlag.

Burzio, L. 1998. Anaphora and Soft Constraints. In P. Barbosa, D. Fox, P. Hagstrom, M. McGinnis & D. Pesetsky (eds.), Is the Best Good Enough? Optimality and Competition in Syntax (pp. 93-114). Cambridge, Mass.: MIT Press.

Chomsky, N. 2008. On Phases. In Foundational Issues in Linguistic Theory. Essays in Honor of Jean-Roger Vergnaud. Ed. Robert Freidin, Carlos P. Otero & Maria Luisa Zubizarreta, 133–166. Cambridge, MA: MIT Press.

Doron, E. & M. Rappaport Hovav 2007. Towards a Uniform Theory of Valence-changing Operations. Proceedings of IATL 23.

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 (pp. 41-72). MIT.

Embick, D. 2004. Unaccusative Syntax and Verbal Alternations. In A. Alexiadou, E. Anagnostopoulou & M. Everaert (eds.), The unaccusativity puzzle: explorations of the syntax-lexicon interface (pp. 137-158). Oxford: Oxford University Press.

Everaert, M. 1990. Nominative Anaphors in Icelandic: Morphology or Syntax? In: W. Abraham & E. Reuland, Germanic Syntax Workshop, J. Benjamins, Amsterdam, 277-307.

Eyþórsson, T. 2007. Reflexive passives in Icelandic. Handout of a talk given at NORMS Workshop on Pronouns, Binding and Anaphors, University of Iceland.

Eythórsson, T. 2008. The New Passive in Icelandic really is a passive. In Grammatical Change and Linguistic Theory: The Rosendal Papers, ed. T. Eythórsson, 173–219. Amsterdam & Philadelphia: Benjamins.

Fanselow, G. 1987, Konfigurationalität, Tübingen, Narr. Fischer, S. 2004. Towards an Optimal Theory of Reflexivization. PhD thesis, University of

Tübingen.

22

Fischer, S. 2006. Matrix Unloaded: Binding in a Local Derivational Approach. Linguistics 44, 913-935.

Frampton, J. and S. Gutmann 2000. Agreement is Feature Sharing. Unpublished manuscript, Northeastern University.

Frey, W. 1993. Syntaktische Bedingungen für die semantische Interpretation. Über Bindung, implizite Argumente und Skopus. Berlin: Akademie-Verlag.

Geniušien÷, E. 1987. The typology of reflexives. Berlin/New York/Amsterda, Mouton de Gruyter.

Grewendorf, G. 1988. Aspekte der deutschen Syntax. Narr. Grimshaw, J. 1981. On the lexical representation of Romance reflexivie clitics. In The Mental

Representation of Grammatical Relations. Ed. by Joan Bresnan. Cambridge: MIT Press. 87- 148.

Haider, H. 1985. Von sein oder nicht sein: Zur Grammatik des Pronomens „sich“. In W. Abraham (ed.), Erklärende Syntax des Deutschen (pp. 223-254). Tübingen: Narr.

Haider, H. 2000. The license to license. In E. Reuland (ed.), Arguments and Case. Explaining Burzio’s Generalization (pp. 31-54). Amsterdam: John Benjamins.

Hale, K. and S. J. Keyser 1993. On argument structure and the lexical expression of syntactic relations. In K. Hale & S. J. Keyser (eds.), The View from Building 20 (pp. 53-109). Cambridge, Mass.: MIT Press.

Heinat, F. 2006. Probes, pronouns, and binding in the Minimalist Program. PhD thesis, Lund University.

Hole, D. 2006. Agentive selbst and other instantiations of the identity function in German. Ms., University of München.

Hole, D. 2008. Focus on identity – The dark side of zìjĭ’. The Linguistic Review 25(3/4), 267-295.

Holmberg, A. 2002. Expletives and Agreement in Scandinavian Passives. Journal of Comparative Germanic Linguistics 4, 85–128.

Jónsson, J. G. 2009. The new impersonal as a true passive. In Advances in Comparative Germanic Syntax, Alexiadou et al. (eds.), Amsterdam & Philadelphia: John Benjamins.

Kaufmann, I. 2001. Medium: Eine Studie zur Verbsemantik. Habilitationsschrift, University of Düsseldorf.

Kayne, R. 1975. French Syntax. The Transformational Cycle. Cambridge, Mass.: MIT Press. Kemmer, S. 1993. The middle voice. Amsterdam: John Benjamins. Kiss, T. 2003. Die Genese der Ausnahmeanapher. In Gunkel, L., Müller, G., & Zifonun, G.

(eds.). Arbeiten zur Reflexivierung. (Studien zur Deutschen Sprache), 157 - 188. Tübingen: Narr.

Koenig, E., Vezzosi, L., 2004. The role of predicate meaning in the developement of reflexivity. In: Wiemer, B., Bisang, W., Himmelmann, N. (Eds.), What makes Grammaticalization? A Look from its Fringes and its Components. Mouton de Gruyter, Berlin, pp. 213–244.

Kratzer, A. 2009. Making a Pronoun: Fake Indexicals as Windows into the Properties of Pronouns. Linguistic Inquiry, 187-237.

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

Maling, J. & S. Sigurjónsdóttir 2002. The new impersonal construction in Icelandic. The Journal of Comparative Germanic Linguistics 5:97–142.

Marantz, A. 1984. On the Nature of Grammatical Relations. Cambridge, Mass.: MIT Press. Marantz, A. 2000. Case and Licensing. In E. Reuland (ed.), Arguments and case: explaining

Burzio's Generalization (pp. 11-30). Philadelphia: John Benjamins. McFadden, T. 2004. The Position of Morphological Case in the Derivation: a study on the

syntax-morphology interface. PhD thesis. University of Pennsylvania.

23

McGinnis, M. 1998. Reflexive external arguments and lethal ambiguity. In Proceedings of WCCFL 16, ed. Emily Curtis, James Lyle, and Gabriel Webster. Stanford: CSLI, 303–317.

McGinnis, M. 2000. Reflexive clitics and the specifiers of vP. In Papers from the UPenn/MIT Round Table on the Lexicon, ed. Liina Pylkkänen, Heidi Harley, and Angeliek van Hout. MITWPL 35. MIT Working Papers in Linguistics, 137–160.

McGinnis, M. 2004. Lethal Ambiguity. Linguistic Inquiry 35 (1), 47-95. Müller, S. 1999. Deutsche Syntax deklarativ. Head-Driven Phrase Structure Grammar für das

Deutsche. Niemeyer, Tübingen. Plank, F. 1993. Peculiarities of Passives of Reflexives in German. Studies in Language 17 (1),

135-167. Reinhart, T. and E. Reuland 1993. Reflexivity. Linguistic Inquiry 24, 657-720. Reinhart, T. and T. Siloni 2004. Against the Unaccusative Analysis of Reflexives. In A.

Alexiadou, E. Anagnostopoulou & M. Everaert (eds.), The Unaccusativity Puzzle (pp. 288-331). Oxford: Oxford University Press.

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

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

Reuland, E. 2001. Primitives of Binding. Linguistic Inquiry 32, 439-492. Schäfer, F. 2008. The Syntax of (Anti-)Causatives. External argument in change-of-state

contexts. Amsterdam: John Benjamins. 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.

Sells, P., A. Zaenen and D. Zec 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 (pp. 169-238). Stanford: CSLI Publications.

Sigurðsson, H. A. 1989. Verbal Syntax and Case in Icelandic. Doctoral dissertation, University of Lund.

Sigurðsson, H. A. 2000. The locus of case and agreement. The locus of case and agreement. Working Papers in Scandinavian Syntax 65, 65-108.

Sigurðsson, H. A. 2003. Case: abstract vs. morphological. In E. Brandner & H. Zinzmeister (eds.), New Perspectives on Case Theory (pp. 223-268). Stanford: CSLI.

Sigurðsson, H. A. 2005. Accusative and the Nom/Acc alternation in Germanic. Working Papers in Scandinavian Syntax 76, 93-133.

Sigurðsson, H. A. 2006. The nominative puzzle and the low nominative hypothesis. Linguistic Inquiry 37, 289-308.

Sigurðsson, H. A. 2009. The No Case generalization. In Advances in Comparative Germanic Syntax, A. Alexiadou, J. Hankamer, T. McFadden, J. Nuger & F. Schäfer. Amsterdam & Philadelphia: John Benjamins.

Sigurðsson, H. A. to appear, On the new passive. Sternefeld, W. 2006. Syntax. Eine morphologisch motivierte generative Beschreibung des

Deutschen. Stauffenberg. 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 (pp. 185-192). München: Iudicium.

24

Wiemer, B. 2006. Relations between Actor-demoting devices in Lithuanian. In Abraham, W., Leisiö, L. (eds.), Passivization and Typology (Form and Function), 274-309. Amsterdam, Philadelphia: Benjamins.

Woolford, E. 1999. More on the Anaphor-Agreement. Effect. Linguistic Inquiry 30: 257-287. Wunderlich, D. 1985. Über die Argumente des Verbs. Linguistische Berichte. 97, 183-227.