Person matters in impersonality

DOI: 10.1111/synt.12230 REGULAR ARTICLE Person matters in impersonality Ane Berro1Ane Odria2Beatriz Fernández2 1University of Deusto 2University of the Basque Country Correspondence: Ane Berro, University of Deusto, Unibertsitate Etorb. 24, 48007, Bilbao, the Basque Country Email: [email protected] Abstract TheBasqueimpersonalisadetransitivizedconstruction where the internal argument is the only overt argument and the external argument, although semantically present, does not have any morphological reflex. This article argues that, despite its intransitive shape, the impersonalinvolvesaparticularkindofVoiceprojection that we term defective. For case and agreement, being defective means having no uninterpretable ϕfeatures and no Case to assign. However, a silent person pronoun is introduced in its specifier position, and thus, there are two arguments within VoiceP. The two arguments compete to value the ϕfeatures of the next functionalhead, namelyT.With thisanalysis weaccountfor the main properties of the impersonal, such as the syntactic activeness of the implicit external argument and thepersonconstraintontheinternalargument. 1INTRODUCTION Basque has a detransitivized construction called impersonal (Ortiz de Urbina 1989,1991, Albizu1997a,2001a, Ortiz deUrbina2003,2011–2019,Fernández&Berro2021)or,alternatively, mediopassive(Brettschneider1979,OrtizdeUrbina2006,DeRijk2008,Rezac2009b).Inthisconstruction,whichwewillrefertoasimpersonal,theinternalargumentisthe onlyovertargument, since the external argument, although semantically present, does not have any morphological reflex. Thus, the Basque impersonal has an intransitive shape; even though the lexical verb is This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. © 2022 The Authors. Syntax published by John Wiley & Sons Ltd. Syntax. 2022;25:147–187. wileyonlinelibrary.com/journal/synt 147 148 BERRO et al. transitive in this construction, the auxiliary selected is the intransitive auxiliary izan ‘be’. This is illustrated in (1).1 (1) Liburu horiek erraz sal-tzen dira. book those[abs] easily sell-ipfv be;3pl.abs ‘Those books are easily sold.’ As can be seen in (1), the internal argument liburuhoriek ‘those books’ is the only argument pronounced and also the only argument crossreferenced on the intransitive auxiliary, which shows third-person plural absolutive agreement. On the basis of the intransitive morphology, the agentive interpretation, and other properties that we will discuss, the Basque impersonal could be considered to be in fact passive (see section 2 for various previous approaches) and to be built on top of a fully constructed verb, like passives that have been analyzed in other languages, such as English and German: see Doron 2003, Collins2005,Bruening2012,Alexiadouet al.2015.In those approaches,it is arguedthatpassives are formed when a Passive head (or a Voice head, in Collins 2005) selects for a Voice projection (or, alternatively, a vP, when v is the head introducing the external argument) consisting of activeVoice.2Inthisway,thoseapproachesexplainthesystematicityandhighproductivityof the construction—in contrast, for example, to middles—and the syntactic activeness of the external argument. Like in passives, the external argument of Basque impersonals is syntactically active. Thesyntactic presenceof theimplicitexternalargumentissupportedby,amongother things,the fact that it can license agent-oriented adverbs and instrumental phrases and the fact that it can control into purpose clauses. All of this makes Basque impersonals different from English-like middles (such as Books read easily), which similarly are intransitive and have an implicit agent argument. In turn, these tests situate the Basque impersonal closer to the passive construction (Fernández & Berro 2021). However,theBasqueimpersonaldivergesfromthepassiveinsomeimportantrespects:unlike in passives, in impersonals the internal argument—which is marked absolutive—must necessarily be third person, as (2) illustrates, and the external argument cannot be overtly realized in an adpositional phrase, as (3) illustrates. (2) Ni/zu hondora-tu nintzen/zinen. I[abs]/you[abs] sink-pfv be;1sg.abs;pst/be;2.abs;pst Inchoative: ‘I sank.’/‘You sank.’ Intended impersonal (ungrammatical): ‘I was sunk.’/‘You were sunk.’ (3) *Itsasontzi-a hondora-tu zen Billez. ship-det[abs] sink-pfv be[3.abs].pst Bill.ins Intended: ‘The ship was sunk by Bill.’ 1Abbreviations used in the glosses in this article: 1 first person, 2 second person, 3 third person, abs absolutive, abl ablative, all allative, cl clitic, com comitative, comp complementizer, dat dative, det determiner, dom differential object marking, erg ergative, f feminine, ine inessive, ins instrumental, ipfv imperfective, loc locative, m masculine, neg negation, pfv perfective, pl plural, pros prospective, psr possessor, pst past, reln relational, res resultative, sg singular, term terminative. 2The contribution of the passive head (or the Voice head, in Collins 2005) is different depending on the author. Bruening and Alexiadou et al. argue that a Pass head saturates the external argument of Voice by existentially binding it. On the other hand, Doron claims that the passive head in Hebrew assigns to the external argument the thematic role of actor (but does not existentially bind it). Finally, Collins suggests that Voice (similar to the passive head in the analyses just mentioned) requires the participle (PartP) to move to spec,VoiceP in English, introduces the preposition by as its head, and checks the accusative case of the DP in spec,vP. Thus, in Collins 2005 the passive head does not prevent the projection of the external argument. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 149 Thus,in this articleweanalyze the impersonalwith a structuredifferentfromthat ofthe passive. Particularly,theanalysisputforwardinthisarticledoesnotassumethattheimpersonalisbuilton top of a fully constructed verb. Instead, it is argued that the type of Voice involved in impersonals is different from that of common personal transitive and intransitive clauses. The present analysis has two important aspects that account for the main properties of the impersonal. On the one hand, we argue that the implicit subject is a silent person pronoun, interpreted as [+human], that has an unspecified person feature (i.e., a person feature with no specificvalue)andnonumberfeatures.BecauseitisapronounofcategoryDP,wecanaccountfor the fact that the implicit subject can license secondary predicates and bind reciprocal anaphors and certain kinds of reflexive anaphors (Landau 2010). Additionally, being a person pronoun but unspecified for person, the implicit external argument must be interpreted as an animate agent but cannot license adpositional phrases with agentive interpretation that have a specific person feature (first, second, or third person). In all these respects, Basque impersonals are similar to reflexive-marked impersonals/passives in Spanish (Mendikoetxea 1999, Sánchez 2002, Ormazabal & Romero 2019,2020). On the other hand, the person restriction on the internal argument—also attested in Spanish reflexive-marked passives/impersonals (Mendikoetxea 2008, MacDonald 2017, Ormazabal & Romero 2019,2020)—is a consequence of the defectiveness of the version of the Voice projection thatisinvolved.AssumingthetypologyofVoiceproposedinSchäfer2008bandlaterdevelopedin Alexiadouetal.2015,weproposethatthedefectiveVoiceinimpersonalsissemanticallytransitive but lacks a D feature. Thus, although this Voice introduces a semantic argument, it is different fromtheVoiceofcommonpersonaltransitiveclauses.Asregardscaseandagreement,wepropose thatbeingdefectivemeanshavingnouninterpretableϕfeaturesandnoCasetoassign.Sincethere are nevertheless two arguments within VoiceP, the two arguments compete to value and delete theuninterpretableϕfeaturesof thenextfunctional head, namelyT. Giventhat Thasasingleset of features (one uPers and one uNum), the unvalued ϕfeatures are shared by the two arguments, and this leads to the person restriction on the internal argument. As we will show, this split in the ϕ-feature valuation of a given functional head is reminiscent of the split ϕchecking attested by the Person–Case Constraint, seen in ditransitive constructions (Anagnostopoulou 2003,Béjar &Rezac2003,Rezac2007,2008a,2009a,2011). As suggested in Rezac 2009b: 314–316, the silent externalargumentinthespecifierofVoice(v,inRezac’sterms)playsthesameroleastheapplicativedativein ditransitivesandthus preventsthe internal argument fromAgreeingin both person and number features with T. However,thepersonrestrictionontheinternalargumentisovercomeifanadditionalϕ-probe head is introduced, as a last-resort repair strategy, at the syntax–LF interface (in order to assure Full Interpretation; Rezac 2011,Kalin2018,T.Levin2019). As we will show, in the Basque impersonal the internal argument can be first or second person if it is marked dative: (4) Zer duzu, baina, gaur triste ikus-ten zaizu.3 what[3.abs] have;2.erg[3.abs] but today sad see-ipfv be;2.dat[3.abs] ‘But what’s going on with you? You look sad today.’ Lit. ‘…You are seen sad today.’ We will argue that dative Case is assigned by activating uninterpretable ϕfeatures in a KP projected on top of the internal-argument DP. With this extra Case licenser, the internal-argument 3Lyrics from the song “Oihana” (a name meaning ‘jungle’ or ‘forest’), written and performed by Mikel Urdangarin; track 1 of the album Heldu artean (Soraluze–Placencia de las Armas, the Basque Country: Gaztelupeko Hotsak, 2002): https:// www.badok.eus/euskal-musika/mikel-urdangarin/heldu-artean. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 150 BERRO et al. DP does not need to share the unvalued ϕfeatures of T, and as a consequence there is no person constraint. This article is structured as follows. In section 2 we consider the various previous approaches to the construction under analysis. In section 3 we present an overview of Basque impersonals and provide evidence for the projection of a semantically active Voice. In section 4 we compare the impersonal with the middle and the passive and show that it cannot be paralleled with either ofthoseconstructions.Particularly,wearguethattheimpersonalcannotberegardedasapassive, given that it has a person restriction on the internal argument and does not allow the presence of an overt agent argument in an adpositional phrase. In section 5, we offer our analysis of impersonals, and in section 6 we show that the person restriction is overcome if an additional ϕprobe is projected and the internal argument is marked with differential case (differential object marking). Finally, in section 7, we conclude by claiming that Basque impersonals involve a defective Voice head but two DP arguments. With this analysis, we account for the intransitive morphologyof impersonals,thesyntacticactivityandanimatenatureofthesilent externalargument, and the person restriction on the internal argument. 2PASSIVE, MEDIOPASSIVE, OR IMPERSONAL? Instancessuchas(1)and(5a)havebeendescribedandanalyzedasimpersonalsinBasquelinguistics (Ortiz de Urbina 1989,1991,Albizu1997a,2001a, Ortiz de Urbina 2003,2011–2019, Fernández & Berro 2021). Nevertheless, the impersonal label obscures the similarity of this construction toOldBasque(medio)passives,alabelusedbyBrettschneider1979,OrtizdeUrbina2006,DeRijk 2008, and Rezac 2009b. The Old Basque (medio)passive is exemplified in (5b), a sentence from the 16th-century writer Joan Perez de Lazarraga. (5) a. Liburu-ak sal-tzen dira. book-det.pl sell-ipfv be;3pl.abs ‘Books are sold.’ b. Dirurren sal-du ninçan esclabea.4 money.for sell-pfv be;1sg.abs;pst slave ‘I was sold for money as a slave.’ (Mounole 2014: 149) There is an important difference between the examples in (5a) and (5b): the internal argument is third person in the former and nonthird (specifically, first) person in the latter. Thus, the third-person restriction on the internal object (see (2)) holds for impersonals in Contemporary Basque but not for Old Basque (medio)passives, as shown by Lazarraga’s sentence in (5b) (see Altuna 1980, Zulaika 1998, Ortiz de Urbina 2006,Mounole2014). Moreover, although, as exemplified in (3), in Contemporary Basque the external argument cannotbeexpressedasanadjunctinimpersonals(Brettschneider1979:381,OrtizdeUrbina2003: 773–776, Berro & Fernández 2019b,Urrestarazu2019, Fernández & Berro 2021), this was not the 4Gidor Bilbao, Ricardo Gómez, Joseba A. Lakarra, Julen Manterola, Céline Mounole, and Blanca Urgell, editors, Lazarraga eskuizkribuaren edizioa eta azterketa (Bilbao, the Basque Country: University of the Basque Country, 2010): B17.60. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 151 caseinOldBasque(medio)passives(OrtizdeUrbina2006:773).Thiscanbeseenin(6),asentence from Joanes Leizarraga’s translation of the New Testament. (6) Gauza guzti-ak ni-ri neure Aita-z eman çaizquit. Thing all-det.pl I-dat my father-ins give.pfv be;3pl.abs;1sg.dat ‘All things have been given to me by my Father.’ (Mt 11:27; Mounole 2014: 150) This example is particularly interesting in that it would be acceptable nowadays—note that the internal argument is third person—except for the explicit mention of the agent argument by means of an instrumental phrase. (See section 5.3 for more on both of the contrasts just mentioned.) Hence, we are dealing with a construction, known both as impersonal and (medio)passive, that is strikingly similar to a passive, as can be seen by comparison with the English translation of (6). In fact, the label passive is used in Altuna 1980: 232, 234, Zulaika 1988: 779–784, and Mounole 2014: 148–155 when referring to instances such as (5b) and (6).5 Leavingasidetheterminologicalchaos,itcouldbepositedthatthereisashiftinBasquefroma truepassivetoaconstructionthatsharesmanybutnotallofthepropertiesofapassive(seesection 5.3). The label passive by itself does not seem appropriate for referring to the constructions in (1) and(5a)availableinContemporaryBasque(although,aswewillseebelow—seespecificallyfootnote 5—the label impersonal passive deserves some consideration); hence, the term mediopassive mightbe moresuitable here. However, when comparedto Englishand Romancemiddles—typeI and type II respectively, in Ackema & Schoorlemmer 2005’s classification—the Basque construction does not seem to be as restricted as the English type, exemplified in (7a), and behaves more in line with the Romance type, exemplified in (7b) by Spanish. (7) a. Booksreadeasily. b. Los libros se leen fácilmente. det.pl book.pl cl read.3pl easily ‘Books read easily.’ Interestingly enough, the properties of the Basque construction differ considerably from those attributed to English-like middles (see section 4.1), but the construction shows striking similarities with the Romance-type middle, in particular the Spanish reflexive-marked impersonal/passive (Mendikoetxea 1999; see section 5.1). Circling back to the beginning, (1) and (5a) might also qualify as impersonals. In fact, this is the term generally used to refer to this construction, as we noted above, and it is tentatively used by us here. The term impersonal actually covers a very broad area, given that it has been used in Basque linguistics to refer not only to the detransitivized construction under discussion but also to constructions with nonreferential or impersonal interpretation of personal verb markers (Ortiz de Urbina 2003: 572–576) and to weather expressions (Ortiz de Urbina 2003: 577–579, Arteatx & Artiagoitia 2018), among other things. The term has also been used to refer to subject-suppressing impersonals, attested in Balto-Slavic, Balto-Finnic, and Celtic languages (Blevins 2003) and discussed by Fernández & Berro 2021 in comparison to Basque. As with the 5The very same passive label is, however, also used by the same authors to refer to another construction that does not seem to fit the definition of a true passive (see section 4.3). 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 152 BERRO et al. Basqueconstruction,thereis no consensusonthepropercharacterization of subject-suppressing impersonals. For instance, Polish -no/-to constructions (Siewierska 1988,Kibort2004), which exemplifythem,havebeenconsideredimpersonalpassivessincetheylackasubject(Comrie1977) and also nonpromotional passives (Zúñiga & Kittilä 2019: 84–85) since the internal argument remains in the object position. As shown by Fernández & Berro 2021, there are many similarities between subject-suppressing impersonals and the Basque construction. However, unlike them, Basque impersonals are not morphologically transitive but rather show a detransitived morphology.6 In a nutshell, the Basque construction under discussion shares several properties with passives, middles, and impersonals, and as a consequence it has received various labels in the previousliterature.Herewehave optedforthe termimpersonal,eventhough weareawareof the fact that it may not fully describe the nature of this construction. 3BASQUE IMPERSONALS AND VOICE In this section, we argue that Basque impersonals involve a Voice layer that is responsible for introducing a thematic subject. Even though the impersonal shows intransitive morphology, namely the auxiliary izan ‘be’ and a lack of ergative agreement (section 3.1), we show that the external argument is semantically present in impersonals, unlike in inchoatives (section 3.2). 3.1 Morphologically intransitive Basque is an ergative language, in that the subject of a transitive verb is marked by ergative case (-k) whereas the subject of an intransitive verb and the transitive object are assigned absolutive case, that is, the unmarked case (Levin 1983, Ortiz de Urbina 1989). Also, Basque shows a three-way agreement system: the arguments marked by ergative, absolutive, and dative case trigger agreement on the auxiliary verb (Hualde 2003). Additionally, Basque is a language with two auxiliaries that alternate, izan ‘be’ and *edun ‘have’.7Transitive verbs require *edun,whereas intransitives can choose between the two. Broadly speaking, agentive or unergative verbs require *edun, while patientive or unaccusative verbs select izan (Levin 1983; see Pineda & Berro 2020 for a recent approach). Albizu 2001b and Arregi 2004 argue that *edun shows up when ergative agreement marking is present in the auxiliary. As for impersonals, these show an intransitive shape and look exactly like inchoatives. In particular,thesoleargumentshowsabsolutivecase,theauxiliary selectedisizan,andnoergative agreement is present (Fernández & Berro 2021). If we compare the examples in (8), it can be 6We leave for further research the study of these subject-suppressing impersonals or impersonal passives in relation to the analysis pursued here and also that of other impersonal/passive configurations that seem to be similar to Basque impersonals (such as Spanish reflexive-marked impersonals/passives). It is worth noting that the Basque construction under discussion has also exceptionally been called impersonal passive (Holmer 1999: 191—but see fn. 7) and been compared to constructions with the same designation (Eguzkitza 1981: 242). Although we are not using this designation here, it might well be the most accurate one, given the properties that the construction shares with both impersonals and passives. 7The citation form of the transitive auxiliary, *edun, is preceded by an asterisk because it is a historically reconstructed participial form and thus unattested as a nonfinite verbal form. See Hualde 2003 for further details on this reconstructed auxiliary and others, which are omitted here for the sake of brevity. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 153 observedthataverblikehil‘kill/die’thatallowsthecausative–inchoativealternationcanoccurin threeconstructions:inthecausative,asin(8a),intheinchoative,asin(8b),andintheimpersonal, as in (8c). (8) a. Unai-k armiarma-k hil ditu. Causative Unai-erg spider.det-pl[abs] kill.pfv have;3pl.abs[3.erg] ‘Unai has killed the spiders.’ b. Armiarma-k (berez) Inchoative spider.det-pl[abs] by_themselves hil/hil-tzen dira. die.pfv/die-ipfv be;3pl.abs ‘The spiders have died/die (by themselves).’ c. Armiarma-k (zapata honekin) Impersonal spider.det-pl[abs] shoe this.com hil/hil-tzen dira. kill.pfv/kill-ipfv be;3pl.abs ‘The spiders have been killed/are killed (with this shoe).’ Inthe causativeexample(8a), the auxiliary selected is*edun, whichagreeswith the ergativesubject Unaik ‘Unai’ and the absolutive plural object armiarmak ‘the spiders’. By contrast, both in the inchoative example (8b) and in the impersonal example (8c), there is only one argument, armiarmak; the auxiliary is izan; and the sentence shows third-person plural absolutive agreement.As shown inFernández&Berro2021,theonlyway todisambiguateexampleslike (8b) and (8c) is to introduce modifiers like berez ‘by themselves’ or zapata honekin ‘with this shoe’. The former forces an inchoative reading, whereas the latter favors an agentive reading in which the external argument is not formally present but receives an arbitrary interpretation. Without those modifiers, though, the inchoative and impersonal look exactly the same. As for the auxiliary alternation, in this article we will simply assume, in accord with Berro et al. 2018,that*edun is selected when the Voice head has a D feature and izan is selected when Voice lacks such a feature. Assuming that auxiliaries are inserted in T, this can be stated with the following spellout rules. (9) a. izan ‘be’ auxiliary →T/_____Voice{} b. *edun ‘have’ auxiliary →T/_____Voice{D} Thus, we propose that in both unaccusatives and impersonals there is a Voice head projected that has no D feature. However, the Voice head selected in impersonals is different from that in unaccusatives in that in impersonals Voice is semantically active. In section 3.2, we will offer evidence for the presence of a semantically active Voice in impersonals. Another question that must be considered here is that of promotion of the internal argument of the impersonal. One of the formal explanations for the promotion to subject that is found in middlesandpassivesinaccusativelanguagesinvolvestheraising-to-subjecttype ofAmovement, asdiscussedbyAckema&Schoorlemmer2005.However,thisAmovementisnotsoobviouswhen dealing with an ergative language such as Basque. To begin with, in an accusative language with overt case or agreement marking, the subject of a transitive verb is marked nominative whereas the object is marked accusative. This being the case, promotion to subject is morphologically visible, since the theme is marked by accusative case in object position and nominative in subject 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 154 BERRO et al. position. By contrast, in an ergative language such as Basque, the theme is marked absolutive in both transitive-object and intransitive-subject positions. As a consequence, the posited A movement,if available, cannotbe morphologicallyidentified. Wordorderdoesnot clarifythe question either. In an SVO language, such as English or a Romance language, the preverbal position is associatedwiththesubjectandthepostverbalpositionwiththeobject. In contrasttothis,Basque being an SOV language, both the transitive-subject and transitive-object positions are preverbal, which blurs the differences between them with respect to the verbal position. Be this as it may, wewill consider the absolutive argument to bemerged in object position and to be assigned Case inthatpositionbyT.Wewillnotspeculateaboutwhetherit raisestothe specifierofTafterwards, since this is not strictly necessary for our analysis. 3.2 Voice in impersonals In this section, we will provide evidence for the presence of a semantically active Voice in impersonals. As mentioned in section 3.1, we consider Voice to be present in both impersonals and unaccusatives, although with important differences. In both, the Voice head projected does not haveaDfeature;butinimpersonals,unlikeinunaccusatives,theVoiceheadintroducesasemanticallyactiveexternalargument.Additionally,inthisarticlewewillargue(seeparticularlysection 5) that the specifier of Voice in the impersonal is also syntactically active. As shown in Fernández & Berro 2021, the implicit external argument of impersonals behaves like a syntactic argument according to several tests: it can license agent-oriented modifiers, as in (10a),instrumentalandcomitativephrases,asin(10b),andsecondarypredicates,asin(10c);and it can control the PRO argument of an adjunct clause (Rodet 1992), as in (10d). (10) a. Itsasontzi-a nahita / kontu handiz hondora-tu zen. ship-det[abs] deliberately carefully sink-pfv be[3.abs].pst Impersonal: ‘The ship was sunk deliberately/carefully.’ Inchoative (pragmatically deviant): ‘The ship sank deliberately/carefully.’ b. Karramarro-a-k erraz harrapa-tzen dira eskuekin/ crab-det-pl[abs] easily catch-ipfv be[3pl.abs] hand.det.pl.com/ salabardoarekin. trap.det.com Impersonal: ‘Crabs are easily caught by hand/with a trap.’ Lit. ‘Crabs are easily caught with hands/a trap.’ Inchoative (ungrammatical) c. Ibai-a erraz zeharka-tzen da oinutsik. river-det[abs] easily cross-ipfv be[3.abs] barefoot Impersonal: ‘The river is easily forded barefoot.’ Inchoative (ungrammatical) d. Itsasontzi-a hondora-tu zen aseguru-a kobratzeko. ship-det[abs] sink-pfv be[3.abs].pst insurance-det[abs] collect.to Impersonal: ‘The ship was sunk to collect the insurance.’ Inchoative (pragmatically deviant): ‘The ship sank to collect the insurance.’ Inchoatives are at best pragmatically deviant according to the same tests for a syntactic external argument. Hence, while the examples in (10) are licit as impersonals, they are ungrammatical or pragmatically deviant when interpreted as inchoatives. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 155 Thus, even if Voice may be present in both inchoatives and impersonals, the semantic and syntactic contribution of this head is clearly different in the two instances. As stated above, we assume that in both inchoatives and impersonals Voice lacks a D feature; this is the reason why they select the auxiliary izan ‘be’. Nevertheless, assuming the typology of Voice proposed in Schäfer2008banddevelopedinAlexiadouetal.2015,wewillclaimthat,unlikeininchoatives,in impersonals Voice introduces a semantically active subject. Apart from that, we will argue that, even though Voice does not have a D feature in impersonals, a silent person pronoun—in other words, a full DP—is merged in its specifier and that this explains the ability of the subject to license secondary predicates and to bind reciprocal anaphors and certain reflexive anaphors (see section 5.1). 4IMPERSONALS ARE NEITHER MIDDLES NOR PASSIVES In section 3, we showed that even though impersonals have intransitive morphology (i.e., they have just one overt argument, marked absolutive, and select the auxiliary izan ‘be’), they involve a Voice layer that introduces a semantically active external argument. In this section, we will discard two possible analyses that posit a nonovert but semantically present external argument. In particular, we will show that impersonals cannot be regarded as either middles or passives (Fernández & Berro 2021). Additionally, we will discuss briefly another construction that could potentially be considered passive in Basque, and we will conclude that Contemporary Basque lacks any construction that matches the properties of passive configurations crosslinguistically. 4.1 Impersonals are not middles AreasonablemovewouldbetoconsiderBasqueimpersonalsmiddles,sincetheyhavesomeofthe properties that are characteristic of middle constructions. According to Schäfer 2008a and 2008b, in middles (i) the grammatical subject corresponds to the thematic or internal object; (ii) the agent is demoted and has an arbitrary interpretation; and (iii) the aspectual interpretation of the sentenceis nonepisodic.Apartfromthosefeatures,ithas also beenarguedthat middles (iv) need to co-occur with a certain kind of modal adverb, namely the kind that generates potential modal interpretations;(v)areformedonlywithasubsetofverbs,namelythosethatentailachange,such asactivitiesandaccomplishments;(vi)donotoccurintheimperativeortheprogressive;and(vii) select the have auxiliary in languages with auxiliary selection (Keyser & Roeper 1984, Ackema & Schoorlemmer 2005). Basque impersonals have some but not all of these properties (Fernández & Berro 2021).8 For instance, as shown in section 3.1, the auxiliary selected in impersonals is izan ‘be’, not *edun ‘have’. Furthermore, they can be episodic; when the verb combines with the perfective 8It must be noted that crosslinguistically middles do not show uniform behavior (Lekakou 2005). For instance, Ackema & Schoorlemmer 2005 makes a distinction between middles that have most or all the properties listed, termed type I middles, and those that do not, termed type II middles. The Basque impersonal patterns with type II middles, in that it seems to be a broader and more permissive construction than the semantically similar but much more restrictive type I middles. In accord with Ackema & Schoorlemmer 2005,Lekakou2005, Fábregas & Putnam 2014,andFernández& Berro 2021, we consider middles to be a particular semantic interpretation obtained in a syntactic construction that is more broadly used in the language, namely the impersonal construction. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 162 BERRO et al. this section, we will examine more closely the nature of the implicit external argument and will argue that it is a silent pronoun of category DP. Thus, we agree with previous proposals such as Albizu 1997a and 2001a that argue that the subject in impersonals is an arbitrary pro.Wefurther specify that the silent pronoun is a person pronoun that is consequently necessarily interpreted as [+human] and that has an unspecified value for the person feature and no number features. Firstly, we will show that the implicit subject must be interpreted as an animate entity with no number features (Rodet 1992). The sentences in (20) involving the verb eraman ‘take’ illustrate this restriction. Eraman does not allow the inchoative–causative alternation. In the active example, either an animate subject (langilea ‘the worker’) or an inanimate subject (haizea ‘the wind’) can be the initiator performing the action of carrying the tiles to the beach. By contrast, in the impersonal example, only the agentive-subject interpretation can remain. With regard to this animateagent,thenumberinterpretationisarbitrary;itcouldbeinterpretedaseitheroneperson or a group of people. (20) a. Langile-a-k/haize-a-k teilatueta-ko Active/transitive worker-det-erg/wind-det-erg roof.pl-reln teila-ak eraman ditu tile-det.pl[abs] take.pfv have;3pl.abs[3.erg] hondartza-raino. beach.det-term ‘The worker/storm has carried the roofing tiles to the beach.’ b. Teilatueta-ko teila-k eraman dira Impersonal roof.pl-reln tile.det-pl[abs] take.pfv be;3pl.abs hondartza-raino. beach.det-term ‘The roofing tiles have been carried to the beach’ (by someone/a group of people/*the wind). Thus,eventhoughtheverbisinprinciplecompatiblewithaninanimatecauser(andconsequently with a Voice head that introduces that), the impersonal is more restrictive in that it only allows an animate subject. A similar restriction has been attested for the silent causee of Icelandic indirect causatives, as shown by E. F. Sigurðsson & Wood 2020. The causee, if it remains implicit as in (21b), cannot be a nonvolitional causer such as a storm. The causee must be animate, even though this leads to a semantic anomaly in the example given. Compare this with the explicit causee in (21a).13 (21) a. Guðlét storminn kasta skipinu God.nom let.pst storm.the.acc toss.inf ship.the.dat til. about ‘God made the storm toss the ship about.’ b. #Guðlét kasta skipinu til. God.nom let.pst toss.inf ship.the.dat about ‘God made (someone/*the storm) toss the ship about.’ 13Similar restrictions regarding the implicit argument have been observed in the Icelandic new impersonal passive (Maling & Sigurjónsdóttir 2002,Maling2006, Ingason et al. 2013) and in the Icelandic impersonal modal construction (H. Á. Sigurðsson & Egerland 2009). 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 163 Sigurdsson & Wood link the animacy condition on the null causee to its ϕP nature. A ϕPisan impoverishedargument,syntactically“smaller”thanafullandreferentialDP(Landau2010,E.F. Sigurðsson & Wood 2020). They suggest that the animacy restriction may be related to ϕP status, given the existing relation between animacy and ϕfeatures. However, the implicit subject of Basque impersonals should not be considered a ϕP, given that, as we will show shortly, it behaves like a DP pronoun in the contexts analyzed by Landau2010. Thus,we proposethat theanimacy conditionillustratedin (20) is a consequence of the implicit subject being a person pronoun, therefore [+human], not of it having an impoverished status as a ϕP. In any case, this is yet one more restriction that differentiates Basque impersonals from passivesinotherlanguages,suchasEnglishandSpanish(Ormazabal&Romero2019,2020).InSpanish analytic passives, the event can be understood to be initiated by a nonvolitional causer (e.g., a fire), as can be seen in (22a). In fact, Spanish analytic passives contrast with reflexive-marked passives in this respect. The implicit subject of reflexive-marked passives, such as (22b), must be an animate agent, like in Basque impersonals (Mendikoetxea 1999, Sánchez 2002, Ormazabal & Romero 2019,2020). (22) a. Muchos libros fueron quemados/destruidos many books be.3pl.pst burn.pfv.m.pl/destroy.pfv.m.pl (por el fuego) aquella noche. by the fire that night ‘Many books were burned/destroyed (by the fire) that night.’ (Ormazabal & Romero 2019: 62) b. Se quemaron muchos libros aquella noche. SE burn.3pl.pst many books that night ‘Many books were burned that night’ (animate generic agent only). (Ormazabal & Romero 2019: 62) In Spanish analytic passives, a silent person argument is not projected in the specifier position ofVoice,andtheVoiceheadprojectedisnotdefective.Therefore,thereisnorestrictionregarding the interpretation of the implicit argument as animate or inanimate. Regarding the categorial nature of the silent argument, we claim that the person argument of Basque impersonals must be regarded as a DP, not as a ϕP. Landau 2010 argues that only DPs, not ϕPs, can license secondary depictive predicates and bind anaphors. In fact, the null causee of Icelandic indirect causatives is argued to be a ϕP, which explains why this construction does not accept secondary predicates (E. F. Sigurðsson & Wood 2020). By contrast, the null subject of Basque impersonals can license secondary predicates naturally (Fernández & Berro 2021): (23) a. Ibai-a erraz zeharka-tzen da oinutsik. river-det[abs] easily cross-ipfv be[3.abs] barefoot ‘The river is easily forded barefoot.’ b. Herri-ko ereserki-a zutunik kanta-tzen da. town-reln hymn-det[abs] standing_up sing-ipfv be[3.abs] ‘The town’s anthem is sung standing up.’ c. Kanta hori mozkortuta kanta-tzen da. song that[abs] drunk sing-ipfv be[3.abs] ‘Thatsongissungdrunk.’ Again, the availability of depictive predicates predicated of the implicit subject is another property that differentiates impersonals from passives. The implicit subject of passives does not 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 164 BERRO et al. generally license secondary predicates in languages like English and Spanish (although see footnote 13). According to Landau 2010 secondary predicates must be predicated of strong implicit arguments, namely DPs. Landau argues that this is the reason for the implicit argument of passives not allowing such predicates in English:14 (24) a. It is impossible for me [to be visited (*together)]. b. The room was left (*angry). c. The issue was decided (*unassisted). b. The game was played (*shoeless). In contrast to the implicit argument of passives, he shows that PRO arguments can license secondary predicates: (25) a. It is possible [PRO to visit me together]. b. They expected [PRO to leave the room angry]. As for Spanish, as shown by Ormazabal & Romero 2019, analytic passives do not allow adjectival phrases predicated of the implicit subject: (26) a. *María no es besada borracho. Mary no be.3sg kissed.f.sg drunk.m.sg Intended: ‘You (generic) cannot kiss Mary drunk.’ b. *Allí, todos los años las fiestas son there all det year.pl det.f.pl festivity.pl be.3pl celebradas vestidos con trajes típicos. celebrated.f.pl dressed.m.pl with dress.pl typical.m.pl Intended: ‘There, every year they (generic) celebrate the festivities dressed in regional costumes.’ (Ormazabal & Romero 2019: 76; glosses and translations ours) 14Whether English passives allow such secondary predicates is under debate. Collins 2005 argues that they do, based on examples like those in (i), and takes that as evidence of the syntactic presence of the external argument (see also Baker et al. 1989, Landau 2000, and the references in those works for further discussion). (i) a. The book was written drunk. (Collins 2005: 101) b. At the commune, breakfast is usually eaten nude. (Collins 2005: 101) c. The song must not be sung drunk. (Baker 1988: 318) Also, the situation across languages seems to be complex. For instance, in German, passives can license secondary predicates, as Alexiadou et al. 2015: 132 shows: (ii) Der Raum wurde wütend verlassen. the room became angry left ‘The room was left in an angry mood.’ 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 165 By contrast, reflexive-marked passives can license secondary predicates once tense, modal, and aspectual factors are controlled, as (27) shows. Note the contrast between the reflexive-marked passives in (28) and the analytic passives in (26). (27) a. El río se atraviesa descalzo. det.m river cl cross.3sg barefoot.m ‘The river is forded barefoot.’ b. En esta casa, se ve la tele sentado. in this house cl watch.3sg det.f TV seated.m ‘In this house TV is watched seated.’ (28) a. ¡No se besa a María borracho! no cl kiss.3sg dom Mary drunk.m ‘One cannot kiss Mary drunk.’ b. Allí, todos los años se celebran las fiestas there all det year.pl cl celebrate.3pl det.f.pl festivity.pl vestidos con trajes típicos. dressed.m.pl with dress.pl typical.m.pl ‘There, they (generic) celebrate their festivities every year dressed in regional costumes.’ (Ormazabal & Romero 2019: 75–76; glosses and translations ours) Landau2010alsoarguesthatonlyDPscanactasantecedentsforConditionAbinding.Landau states that a Condition A binder must be a strong implicit argument, that is, it requires a D layer. With its D feature, the binder enters into an Agree relation with T and in this way checks the D feature of the anaphor. As can be seen in (29) and (30), reciprocal anaphors can be bound by the implicit subject (Albizu 1997a, Ortiz de Urbina 2003), especially when the verb has imperfective aspect and the interpretation is nonepisodic. (29) elkar engaina-tzen denean each_other[abs] deceive-ipfv be[3.abs].when ‘when everyone deceives each other’ Lit. ‘when each other is deceived’ (Ortiz de Urbina 2003: 588) (30) a. Hona heltzean, elkar agur-tzen/besarka-tzen da. here.all arrive.when each_other[abs] greet-ipfv/hug-ipfv be.[3.abs] ‘When one arrives here, everyone must greet/hug each other.’ Lit. ‘When one arrives here, each other is greeted/hugged.’ b. Hemen-dik erraz ikus-ten da elkar. here-abl easily see-ipfv be[3.abs] each_other[abs] ‘Everyone can see each other from here.’ Lit. ‘Each other can be seen from here.’ As noted by Ortiz de Urbina 2003, third-person reflexive anaphors are not acceptable in this construction: (31) *bere buru-a engaina-tzen denean 3.psr head-det[abs] deceive-ipfv be[3.abs].when Intended: ‘when oneself is deceived’ (Ortiz de Urbina 2003: 588) 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 166 BERRO et al. However, we believe that the unacceptability of reflexive anaphors in impersonals is a consequence not of the categorial nature of the implicit subject but of the person specification of the anaphor, which is incompatible with the unspecified value of the person feature of the implicit subject (Albizu 1997a). A similar restriction has been observed for Spanish reflexive-marked passives. Ormazabal & Romero 2020 argues that se in reflexive-marked passives cannot license reflexive pronouns like su ‘his/her’ because su is specified for person features whereas se is not. Similarly, the Basque reflexive anaphor bereburua =his/her head ‘himself/herself’, which varies dependingonthepersonfeaturesoftheantecedent(nireburua=myhead‘myself’forfirst-person antecedents, zureburua =your head ‘yourself’ for second-person antecedents), cannot be bound by the implicit subject of impersonals because it requires a binder with the same person specification, whereas the implicit subject has an unspecified person feature: it can be interpreted as first person, second person, indefinite, or generic, like the person pronoun se in Spanish (Mendikoetxea 1999). Interestingly, Basque has another reflexive anaphor, one that requires an unspecified animate antecedent, namely norbere burua =oneself’s head ‘oneself’. This anaphor can occupy the thematic position of either the direct object, as in (32a), or the indirect object, as in (32b). (32) a. norbere buru-a engaina-tzen denean oneself.psr head-det[abs] deceive-ipfv be[3.abs].when ‘when oneself is deceived’ b. Ikastaroko lehen egunean, norbere buruari gutun bat course.reln first day.ine oneself.gen head.det.dat letter det[abs] idatziko zaio. write.pros be[3.abs];3.dat ‘On the first day of the course, we will write a letter to ourselves.’ Lit. ‘…a letter will be written to oneself.’ Sincenorbereburuarequiresananimateantecedentthatisunspecifiedforperson,itcannotoccur insentenceswithanovertorsilentDPantecedentspecifiedforfirst,second,orthirdperson,such as (33b), contrasting with (33a). (33) a. Beste inorena baino lehen, norbere other anybody.gen.det[abs] than before oneself.gen burua maita-tzen da hemen. head.det[abs] love-ipfv be[3.abs] here ‘Before anybody else, we love ourselves here.’ Lit. ‘…oneself is loved here.’ b. *Beste inorena baino lehen, (ni-k/zu-k/ other anybody.gen.det[abs] than before I-erg/you-erg/ horrek) norbere burua maita-tzen that one.erg oneself.gen head.det[abs] love-ipfv dut/duzu/du. have[3.abs];1sg.erg/have[3.abs];2.erg/ have[3.abs;3.erg] Intended: ‘Before anybody else, I/you/that one love(s) myself/yourself/himself/herself.’ Intheactivetransitiveexample(33b),the overt subject nik ‘I’,zuk ‘you’,orhorrek ‘thatone’orits silent counterpart is crossreferenced by the corresponding agreement markers on the auxiliary. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 167 If, instead of norbere burua, the example in (33b) contained a reflexive anaphor with a specific person value, it would be grammatical, as expected: (34) Beste inorena baino lehen, (ni-k/zu-k/horrek) other anybody.gen.det[abs] than before I-erg/you-erg/that one.erg nire/zure/bere burua maita-tzen my/your/her head.det[abs] love-ipfv dut/duzu/du. have[3.abs];1sg.erg/have[3.abs];2sg.erg/have[3.abs;3.erg] ‘Before anybody else, I/you/that one love(s) myself/yourself/himself/herself.’ In (34), the reflexive anaphors agree in person with their corresponding subject, and as a consequence, the sentence is grammatical. Thus, the only problem in (33b) is that norbere burua is incompatible with a person-specific antecedent. In contrast, since in an impersonal the implicit subject is animate and has an unspecified value for person, it is a suitable antecedent for norbere burua,asin(33a). Due to the unspecified nature of the silent person argument and to the kind of Voice head involved in the impersonal, its subject differs considerably from a common DP argument, in that it cannot be overtly pronounced,does not bear ergative case, does not trigger ergative agreement ontheauxiliary,anddoesnotaffectauxiliarychoice.Inapersonalfiniteclause,whentheexternal argumentisaDP(overtor silent),theauxiliarymustshowthecorrespondingergativeagreement marker, and the auxiliary root must be *edun ‘have’ instead of izan ‘be’: (35) (Gazte-ek) kanta hori mozkortuta kanta-tzen young_people-pl.erg song that[abs] drunk sing-ipfv dute/*da. have[3.abs];3pl.erg/be[3.abs] ‘Young people/they sing that song drunk.’ In contrast, as we have shown, the implicit external argument of impersonals does not have any morphological reflex on the auxiliary and does not change the auxiliary root into *edun.Aswe will explain in the next section, this is a consequence of its having a defective kind of Voice, whichlacks a D feature and doesnot provideuninterpretableϕfeaturestobe valuedby a DP and deleted. 5.2 Defective Voice As mentioned at the outset of section 5.1, we claim that the Voice head in impersonals is thematic but has no D feature (like in Schäfer 2008b’s passives or Alexiadou et al. 2015’s Greek passives). We label this kind of Voice head defective for convenience, in order to compare it with the Voice head that occurs in common transitive sentences, which is both thematic and has a D feature. As we noted, in our account the fact that Voice does not have a D feature does not mean that a certain kind of external argument cannot be merged in its specifier position. We argued in section 5.1 that the external argument of impersonalsis a silent person pronoun with an unspecified person feature, similar to the se pronoun in Spanish reflexive-marked passives (Ormazabal & Romero 2019,2020). If so, in impersonals there are two active arguments: a silent 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 168 BERRO et al. personpronounwithunspecifiedpersonfeaturesandaninternalargument,whichcanbeeither overt or silent. The fact that there are two arguments within a defective VoiceP causes the person restriction described in section 4.2. We assume that Voice is the head that introduces the external argument (Kratzer 1994,1996) and assigns Case to the internal object of transitive predicates (like little v in Chomsky 1986). In order to explain the derivation of impersonals, we will first consider that of personal transitive clauses. In such clauses, Voice has a D feature and the ability to assign absolutive Case to the internal argument. Thus, it can be regarded as a strong phase, equivalent to the v*P of Chomsky 2000 and Chomsky 2001. We will remain agnostic about the Case assignment of the external argument: this account is compatible with an analysis where the external argument receives inherent ergative case from Voice (in line with proposals such as Legate 2002, Massam 2002, Aldridge 2004, Woolford 2006,Legate2008, Coon 2010)orwithonewhereitisCaseassigned structuralergativecasebyT(forarecentapproach,seeRezacetal.2014).Inanycase,wepropose that, as shown in (36), in personal transitive sentences, the external argument values and deletes the set of uninterpretable ϕfeatures on T (uϕ, consisting of uPers and uNum) and the internal argumentvaluesanddeletestheuninterpretableϕfeaturesonVoice (uϕ, also consistingofuPers and uNum). This gives rise to the agreement markers (ergative and absolutive respectively) that occur in personal transitive sentences. (36) Personal transitive sentences; nondefective Voice By contrast, in impersonals, Voice is defective in that it lacks a D feature. For case and agreement, we consider that being defective means having no uninterpretable ϕfeatures and thus no structural Case to assign. Defective VoiceP can also be characterized as a weak phase, similar to theonethatispresentinunaccusativepredicates(Chomsky2000,2001).Hence,contrarytowhat weseeinpersonaltransitives,VoicePinimpersonalsdoesnotinvolveacompletedomainforCase licensing, and thus the internal argument introduced within it enters into Agree with the next functionalheadthathasuninterpretableϕfeatures,namely,T.InaccordwithChomsky2000and Chomsky 2001, we assume that the uninterpretable ϕfeatures in T are in fact inherited from the phasal head C. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 169 In spite of its defectiveness, we nevertheless claim that the VoiceP in impersonals actually contains two arguments: the external argument, in its specifier position, and the internal argument, in its complement position—that is, within vP. Given the nonphasal nature of Voice, both the external and internal arguments coincide in the same phasal domain, namely, CP. Therefore, the two of them will try to Agree with T, the only functional head bearing uninterpretable ϕ features within CP. The external argument being structurally higher than the internal argument, T first probes the ϕfeatures of the external argument. Given that this is a silent person pronoun not specified for number, it bears person but not number features and is thus defective in its ϕ features. Consequently, T will only Agree in person features with it, and as a consequence, the internal argument will have the option to Agree with T in number features only. Thus, we propose that, as shown in (37), the feature set present in T is divided: the external argument values the uPers feature, whereas the internal argument values uNum. (37) Impersonal sentences; defective Voice Since the internal argument cannot value uPers, firstand second-person absolutive objects are banned in this context, given that these bear both person and number features. By contrast, third-person arguments are allowed, because—as argued in the literature on the Person–Case Constraint, which applies in ditransitive constructions—these bear number but not person features (see, among others, Anagnostopoulou 2003, Béjar & Rezac 2003,Rezac2007,2008a,2009a, 2011, Adger & Harbour 2007). We thus claim that the nature of the person restriction in impersonal constructions is identical to the Person–Case Constraint as attested in ditransitives. See Rezac 2009b: 314–316 for this same claim. In ditransitives, the Person–Case Constraint captures the incompatibility of firstand second-person objects with an agreeing (indirect-object) dative (Bonet 1991,1994). This is a syntactic constraint that arises when the two internal arguments—the direct and indirect objects—co-occur in the same Agree/Case locus, namely, v (equivalent to transitive Voice). For Basque, the Person–Case Constraint is illustrated by Laka1993’s classic examplesin(38); see also Albizu 1997band1997cforathoroughexamination of the Person–Case Constraint in Basque and crosslinguistically. (38) a. *Zu-k harakina-ri ni sal-du naiozu. you-erg butcher-dat I[abs] sell-pfv have;1sg.abs;3sg.dat;2.erg ‘You have sold me to the butcher.’ 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 170 BERRO et al. b. Zu-k harakina-ri liburua sal-du diozu. you-erg butcher-dat book[abs] sell-pfv have[3.abs];3sg.dat;2.erg ‘You have sold me the book.’ (Laka 1993: 27) As has been argued with respect to ditransitive constructions, we assume that in firstand second-person arguments it is not sufficient to check their [number] feature in order for them to be Case licensed (Anagnostopoulou 2003). The Agree relations in impersonal constructions are then identical to those that hold in ditransitives: due to the intervention of a higher argument—the dative in ditransitives, the implicit external argument in impersonals—firstand second-person internal arguments are banned from undergoing regular Agree with their actual Case licenser, namely, v in ditransitives and T in impersonals. This violates the Person-Licensing Condition, which states that “an interpretable first/second person feature must be licensed by entering into an Agree relation with a functional category” (Béjar & Rezac 2003: 53). Another factor in favor of the claim that the internal argument only values uNum on T is that the internal argument, apart from being third person, is usually nonhuman (Rodet 1992) or inanimate (Berro & Fernández 2019b, Fernández & Berro 2021). As mentioned by Berro & Fernández 2019b, the impersonal construction is much more frequent in the corpora when the internal argument is inanimate, although inanimacy is not a necessary condition to form an impersonal (see section 6.2 for the derivation of third-person animate internal arguments). A similar tendency has been observed in the spread of the reflexive-marked passive in Spanish (Cabañas 2006). In sum, we argue that in impersonals the only feature set, provided by T, is shared by two arguments and that this is the source of the person restriction found. Instead of functioning as a homogeneousblock,theϕfeaturesonThappentobevaluedseparatelybytheexternalargument on the one hand and the internal argument on the other: the former Agrees with T in person features and the latter in number features. In ditransitives, as argued in Béjar & Rezac 2003: 54, given that dative Case is licensed by Appl, the indirect object can only Agree defectively with v, andthusthepersonfeaturesinvgetadefaultvalue,thatis,thirdperson.Similarly,inimpersonals, since the implicit external argument bears unspecified person features, uPers on T gets a default value: again, third person.15 Nevertheless, the implicit external argument does not trigger ergative agreement on the auxiliary, as external arguments usually do in Basque (leaving aside ergative-displacement phenomena).Whyabsolutiveinsteadofergative?Wesuggestthatthechoicebetweenabsolutiveandergative agreement markers might be governed by a PF-interface operation, a vocabulary-insertion rule that varies depending on the nature of the Voice head adjacent to T (in a similar vein as the vocabulary-insertion rule for auxiliaries given in (9)). When T has an adjacent Voice head that has a D feature, the valued ϕfeatures of T will be realized as ergative agreement markers; otherwise, if the adjacent Voice head does not have a D feature, they will be realized as absolutive agreement markers. Aswillbeshowninsection6,thepersonrestrictionontheinternalargumentcanbeovercome if it is marked dative instead of absolutive. In that scenario, we will claim that an extra Case licenserhasbeen introducedintothe derivation(Rezac2011,Kalin2018,T.Levin2019). The fact 15The claim that the external argument values the person probe on T as third person by default is supported by the fact that, with first and second person, the person features do not make interpretive sense without the number features (Taraldsen 1995, Anagnostopoulou 2003). 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 171 that Case assignment affects the person restriction—that is, the fact that the source of the person restriction is the inability of defective Voice to license two different arguments—is evidence in favor of our analysis. 5.3 A brief look at Old Basque (medio)passives Beforemovingontosection6,wewouldliketostopbrieflytolookatOldBasque(medio)passives, aconstructionsimilarto(contemporary)impersonalsbutwithcertainpropertiesthatcontrastsignificantly (see section 2). In keeping with the literature, we will refer to the construction attested in Old Basque as (medio)passives, but the reader should note that we are basically dealing with an (apparently) very similar construction. As presented in section 4.2, Basque impersonals have some properties that make them different from well-behaved passives in languages like English and Spanish, especially the person restriction on their internal argument and their incompatibility with agent postpositional phrases. Interestingly, this does not seem to be the case for (medio)passivesinOldBasque,asanalyzedbyIrigoyen1992,Zulaika1998,OrtizdeUrbina2006, and others. Firstly, Old Basque (medio)passives have no person restriction on the subject: no ban on firstand second-person internal arguments, as can be seen in (39) and (40a) below. Secondly, they can express an agent by means of an instrumental postpositional phrase, as in (40a) and (40b). Thefirst exampleshowingthe lack of person restriction, (39), is from Bernard Etxepare’s Linguae Vasconum primitiae.16 (On this property of Old Basque (medio)passives, see particularly Ortiz de Urbina 2006: sects. 2, 3, Zulaika 1998: sect. 16.2.) (39) Oray, othoy, enadila oguen gabe difama. now please neg;be;1sg.abs;comp fault without defame ‘Now, please let me not be defamed without blame.’ (9:14; Ortiz de Urbina 2006: 768) As can be seen, the internal argument of the (medio)passive in (39) is first person, being crossreferenced by the first-person absolutive marker on the auxiliary. The second property, overt expression of the agent, can be seen in the Old Basque (medio)passives in (40), both from Joanes Leizarraga’s 1571 translation of the New Testament.17 (40) a. Parti nadinean Espaigniarát, ethorriren naiz leave be;1sg.abs;comp Spain.all come.pros be;1sg.abs çuetara: ecen sperança dut iragaitean you.all for hope have[3.abs];1sg.erg pass.ine ikussiren çaituztedala eta çuueça-z hara see.pros have;2pl.abs;1sg.erg;comp and you-ins there guidaturen naicela. guide.pros be;1sg.abs;comp ‘Whensoever I take my journey into Spain, I will come to you: for I trust to see you in my journey, and to be brought on my way thitherward by you.’ (Rom 15:24; Zulaika 1998: 782) 16The first book published in Basque, in 1545. See Patxi Altuna’s critical edition: Altuna 1980. 17See the edition of Theodor Linschmann and Hugo Schuchardt, I. Leiçarraga’s baskische Bücher von 1571 (Neues Testament, Kalender und Abc) im genauen Abdruck (Strasbourg, France: K. J. Trübner, 1900). 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 178 BERRO et al. In Basque, dative Case is commonly considered to be assigned either by (High/Low) Appl or by a P head (see, among many others, Albizu 2001a,Rezac2008a,2008b,Albizu2009,Etxepare& Oyharçabal 2009,Rezac2009a,2009b, Fernández 2010a,2010b, Fernández & Ortiz de Urbina 2010,Rezac2011, Fernández & Ortiz de Urbina 2012, Etxepare & Oyharçabal 2013, Odria 2014, Fernández 2015, Odria 2017, Ormazabal & Romero 2017, Odria 2019). However, in order to distinguish the dative object of impersonals from other inherent datives, we propose that the dative Case assigner in impersonals is instead K, a more general functional head that stands for (structural) Case—this is in fact the head that assigns dialectally attested differential object marking in transitive constructions(Odria2017).Hence,wealsoassume that the dativemarkersin the finite verbinvolvecliticdoubling,thatis,themovementofaheadthat,afterAgreeingwiththelicensed nominal, brings its interpretable ϕ-feature bundle to Voice/T, as is argued by Etxepare 2006, Rezac 2006,2007,Arregi&Nevins2008,Rezac2008a,2008b,Preminger2009,Rezac2011,Arregi &Nevins2012, Etxepare 2014, Preminger 2014, Rezac et al. 2014, Odria 2017, and Odria 2019, among others. Ascanbeseenin(49),whentheobjectAgreeswithK,thenumberfeaturesinTremainunvalued; they are thus interpreted as singular by default, since this is the morphologically unmarked valuefornumberfeatures(seeKramer2014,Preminger2014,andT.Levin2019,amongothers).24 This contrasts with third-person objects that receive absolutive Case by Agreeing in number features with T, because the absolutive agreement marker depends on the number specification of the object. The latter is seen in (50), whereas, when the object is marked dative as in (51), the number features in T remain unvalued/singular, regardless of the number features of the object, since these are actually crossreferenced by dative agreement. (50) a. Liburu hori erraz sal-tzen da. book that[abs] easily sell-ipfv be[3.abs] ‘That book is easily sold.’ b. Liburu horiek erraz sal-tzen dira. book those[abs] easily sell-ipfv be;3pl.abs ‘Those books are easily sold.’ (51) a. Pozik ikus-ten zaizu. happy see-ipfv be;2.dat[3.abs] ‘You (singular) look happy.’ 24As a reviewer points out, the proposal that the uNum feature of T can remain unvalued makes the prediction that intransitive verbs are also possible in the impersonal configuration in Basque, with the sole argument being the silent external argument in the specifier of Voice. As (i) illustrates, this prediction is in fact borne out: unergative verbs like dantzatu ‘dance’, eskiatu ‘ski’, and so on can easily be found in this construction (Ortiz de Urbina 2003), with the sole argument remaining implicit. (i) Hemen ondo dantza-tzen/eskia-tzen da. here well dance-ipfv/ski-ipfv be;3.abs ‘People dance/ski well here.’ Certain unaccusative verbs can also occur in the impersonal construction, but their distribution is much more limited. We would suggest that in intransitive impersonals the implicit external argument checks uPers on T and uNum remains unvalued. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 179 b. Pozik ikus-ten zaizue. happy see-ipfv be;2pl.dat[3.abs] ‘You (plural) look happy.’ Again, the contrast between (50) and (51) corroborates the idea that the intervention of the implicit subject exclusively affects uPers in T. Otherwise, we would not expect a number contrast when the object Agrees with T and appears absolutive marked, as in (50). 6.3 Differential object marking in impersonals and nonimpersonal transitives In line with Rezac 2011,Kalin2018 argues that finite clauses have one obligatory Case licenser that is always merged whereas other licensers are merged secondarily—that is, when needed for convergence. Thus, as in Rezac 2011, the addition of a secondary Case licenser is seen as a last-resort operation, given that it is exclusively applied in cases where the derivation would otherwise crash. In this way, Kalin accounts for the crosslinguistically widely attested phenomenon ofdifferentialobject marking, by arguing that the differentialmarking arises as a consequence of the activation of such a secondary Case licenser, which licenses otherwise unlicensed objects. Differential object marking refers to the phenomenon where objects high in animacy and/or specificity bear a given marking not attested in other objects (Bossong 1985,1991, Aissen 2003). DifferentialobjectmarkingisalsoavailableinsomeBasquesouthwesternvarieties.Anexampleis in (52b), whose dative-marked object contrasts with the absolutive-marked object in (52a). In the varieties in question, objects high in animacy and specificity, especially firstand second-person objects, tend to bear this dative marking instead of the canonical absolutive expected in an ergative language like Basque (Fernández & Rezac 2010,Mounole2012, Odria 2014, Fernández & Rezac 2016, Rodríguez-Ordóñez 2016, Odria 2017, Rodríguez-Ordóñez 2017, Odria 2019). (52) a. Gaur triste ikus-ten zaitut (zu). today sad see-ipfv have;2.abs;1sg.erg you[abs] ‘You look sad to me today.’ Lit. ‘I see you sad today.’ b. Gaur triste ikus-ten dizut (zu-ri). today sad see-ipfv have[3.abs];2.dat;1sg.erg you-dat ‘You look sad to me today.’ Lit. ‘I see you sad today.’ This differential object marking is similar to that found in Spanish and Hindi, where animate and specific objects show the same marking as dative indirect objects: amarking in Spanish and -ko marking in Hindi. Kalin 2018 argues that in languages like Spanish and Hindi transitive v is an obligatory licenser and Appl an optional one, which merges when the object needs licensing and this requirement is not satisfied by v. Thus, according to Kalin, in these languages Appl behaves as an obligatory licenser for indirect objects and as an optional one for direct ones. Similarly, T. Levin 2019 regards differential object marking in Palauan as an exceptional-licensing phenomenon since it requires the addition of a Case licenser not otherwise present in the derivation. As we have already pointed out, the same mechanism applies in Basque impersonals: firstandsecond-personobjectsendupbeingdativemarked,asin(41c),byAgreeingwithapreviously 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 180 BERRO et al. inactive ϕprobe (Rezac 2007,2009a,2011) or, to put it otherwise, by Agreeing with a secondary or optional licenser (Kalin 2018,T.Levin2019). Therefore, it seems reasonable to think that the dativeCaseassignedbyKinBasqueimpersonalsis in factaninstanceofdifferentialobjectmarking, since it is also attested with human but not nonhuman objects and is a pattern available for differential-object-marking speakers.25 Nevertheless,observethat in transitivesdifferentialobjectmarking co-occurs with thetransitiveauxiliary *edun‘have’(52b)andin impersonals with the intransitiveizan ‘be’ (41c);thelatter fact reduces to the defectiveness of the Voice head in impersonals. Furthermore, in the differential object marking found in transitive constructions, firstand second-person datives alternate withtheabsolutive;thephenomenonisoptionalformanyspeakers,andthosewhohaveitobligatorily also use the absolutive in nonfinite contexts (Fernández & Rezac 2016, Odria 2017). This is not what we see with dative objects in impersonals: although third-person objects may alternate between the dative and the absolutive, firstand second-person objects can only appear with the dative(however, see footnote 11).As argued in Odria 2014,Fernández & Rezac2016, Odria2017, and Odria 2019, the alternation found between differential object marking and absolutive markingintransitiveconstructionsmaybeduetothefactthat,unlikeinimpersonals,theobjectAgrees with v/Voice before receiving the differential marking. In sum, both Kalin 2018 and T. Levin 2019 argue that differentially marked objects, due to their ϕ-feature specification, always require Case licensing and that the absence of it leads to ungrammaticality. This is, again, analogous to what happens with firstand second-person objects under the Person–Case Constraint as manifested in both ditransitives and the impersonals under study. In fact, as Levin notes (p. 193), “we could intuitively maintain that all [differential-object-marking] patterns are triggered by ‘extended’ [Person-Licensing Condition]–like statements.” Kalin 2017 takes a further step in unifying the Person–Case Constraint in ditransitives with differential object marking, stating that the similarities between the two include (i) the number of arguments involved in each (two: the indirect and direct objects with the Person–Case Constraint and the subject and object in differential object marking); (ii) the fact that in both cases the phenomenon affects the lower argument, namely the object; (iii) the fact that both the Person–Case Constraint and differential object marking are triggered by the intervention of a higher argument and neither of them persists when the higher argument is removed; and (iv) the fact that both are repaired by adding a secondary licenser to the construction. In addition, it is important to recall that the relationship between the dative marking of the object and the Person–Case Constraint appears even closer in the case of Basque impersonals than in transitive contexts, because the differential-object-marking pattern is actually caused by a person restriction that could simply be an instance of the Person–Case Constraint. 7CONCLUSIONS In this article we have analyzed the impersonal in Basque. We have proposed that in this construction Voice introduces a semantically active argument but has no D feature (Schäfer 2008b, Alexiadou et al. 2015), and we have called this type of Voice defective. In this analysis, Voice 25However, as pointed out by Ortiz de Urbina 2011–2019, the dative in impersonals is independent from the dialectal differential object marking discussed in the literature. In other words, speakers that do not show dialectal differential object marking do have the dative-marking pattern observed in impersonals, although we presume that this strategy is restricted to southwestern dialects. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 181 beingdefectivealsoimplieshavingnouninterpretableϕfeatures,leavingTastheonlyfunctional head with ϕfeatures to be valued by a DP and deleted. However, even though Voice is defective in impersonals, we have suggested that a silent person pronoun is introduced in its specifier position and thus that there are two arguments within VoiceP. As a result, the uninterpretable ϕfeatures of T are split between the two arguments, and this causes a person constraint on the internal argument: it can only be third person. In line with Rezac 2009b, this article has shown thattherearesignificantsimilaritiesinnaturebetweenthepersonconstraintsoccurringinimpersonalsandinditransitiveconstructions.Theircommonalitiesalsoincludetherepairstrategythat is used to overcome them, which consists of activating a ϕprobe that was previously inactive in the derivation (Rezac 2007,2009a,2011). As a consequence of the repair, in impersonals, firstand second-person internal arguments are embedded under KP and morphologically marked with dative case and agreement, circumventing in this way the person restriction (Kalin 2018,T. Levin 2019). With this analysis, we have accounted for the main properties of the impersonal: the fact that the implicit subject is syntactically active as well as able to license secondary predicates and bind reciprocal anaphors and certain reflexive anaphors; the construction’s intransitive morphology; and the person constraint on the internal argument. Additionally, we have provided an analysis of impersonals with dative-marked internal arguments and proposed that this is an instance of differentialobjectmarking.The impersonalconstructionwithdifferentialobjectmarkingisbasically similar to that without it, the difference being that in the former, an extra ϕprobe has been introduced into the derivation in order to license a firstor second-person internal argument. ACKNOWLEDGMENTS Theauthorsaregratefultothereviewersandtheeditorfortheircarefulreadingandvaluablecomments. They also acknowledge financial support from research project PGC2018-096380-B-I00 of the Spanish Ministry of Science, Innovation, and Universities, from research project ANR-17-CE27-0011-01 of the Agence Nationale de la Recherche, and from research project IT1169 of the Basque government. DATA-AVAILABILITY STATEMENT The English, Spanish, and Icelandic data discussed in this article, as well as some of the Basque data, have been previously reported in the literature cited. The Basque data that were generated by the present study have been given explicitly in this article. REFERENCES Ackema, Peter, and Maaike Schoorlemmer. 2005. Middles. In: Martin Everaert and Henk van Riemsdijk, editors. The Blackwell companion to syntax. Malden, MA: Blackwell Publishing. 3.131–203. Adger, David, and Daniel Harbour. 2007. Syntax and syncretisms of the Person Case Constraint. Syntax 10.1.2–37. Aissen, Judith. 2003. Differential object marking: Iconicity vs. economy. Natural Language and Linguistic Theory 21.3.435–483. Alberdi, Xabier, and Ibon Sarasola. 2001. Euskal estilo libururantz [Towards a Basque style book]. Bilbao, the Basque Country: Euskal Herriko Unibertsitateko Argitalpen Zerbitzua. Albizu, Pablo. 1997a. En torno a la inacusatividad sintáctica de las construcciones impersonales del euskera [Regarding the syntactic unaccusativity of Basque impersonal constructions]. Interlingüística 8.13–18. Albizu, Pablo. 1997b. Generalized person-case constraint: A case for a syntax-driven inflectional morphology. In: AmayaMendikoetxeaandMyriamUribe-Etxebarria,editors.Theoreticalissuesatthemorphology-syntaxinterface.(Anejos del Anuario del Seminario de Filología Vasca “Julio de Urquijo” 40.) Bilbao, the Basque Country: University of the Basque Country. 1–33. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 182 BERRO et al. Albizu, Pablo. 1997c. The syntax of person agreement. Unpublished paper. Work conducted at: University of Southern California. Albizu, Pablo. 2001a. Datibo sintagmaren izaera sintaktikoaren inguruan: Eztabaidarako oinarrizko zenbait datu [On the syntactic nature of the dative phrase: Some data for discussion]. In: Beatriz Fernández and Pablo Albizu, editors. Kasu eta komunztaduraren gainean/On case and agreement. Bilbao, the Basque Country: Euskal Herriko Unibertsitateko Argitalpen Zerbitzua. 49–69. Albizu, Pablo. 2001b. Sobre la distribución sintáctica de las formas finitas del verbo vasco: Condicionamiento léxico y sintáctico [On the syntactic distribution of the finite forms of the Basque verb: Lexical and syntactic conditioning]. Anuario del Seminario de Filología Vasca “Julio de Urquijo” 35.1.65–106. Albizu, Pablo. 2009. Construcciones inacusativas con dativos posesivos y dativos de interés en vasco: Un análisis derivacionalista [Unaccusative constructives with possessor datives and datives of interest in Basque: A derivationalist analysis]. Handout. Presented at: Seminario de Lingüística Teórica, Spanish National Research Council, May 11. Aldridge, Edith. 2004. Ergativity and word order in Austronesian languages. Doctoral thesis. Ithaca, NY: Cornell University. Alexiadou, Artemis, Elena Anagnostopoulou, and Florian Schäfer. 2015. External arguments in transitive alternations. Oxford, the United Kingdom: Oxford University Press. Altuna, Patxi, editor. 1980. Linguae Vasconum primitiae [First fruits of the language of the Basques]. By Bernard Etxepare. Bilbao, the Basque Country: Euskaltzaindia. Altuna,Patxi. 1981.Ziburuko Etxeberrirenaditzadela eta[Regardingtheverbused byEtxeberrifromZiburu]. In: Euskalarien nazioarteko jardunaldiak.(Iker 1.) Bilbao, the Basque Country: Euskaltzaindia. 323–329. Anagnostopoulou, Elena. 2003. The syntax of ditransitives: Evidence from clitics. Berlin, Germany: Mouton de Gruyter. Arregi, Karlos. 2004. The have/be alternation in Basque. Unpublished paper. Work conducted at: University of Illinois at Urbana-Champaign. Arregi, Karlos, and Andrew Nevins. 2008. A principled order to postsyntactic operations. Unpublished paper. Work conducted at: University of Illinois at Urbana-Champaign and Harvard University. Arregi,Karlos,andAndrewNevins.2012.Morphotactics:Basqueauxiliariesandthestructureofspellout.Dordrecht, the Netherlands: Springer. Arteatx, Iñigo, and Xabier Artiagoitia. 2018. Eguraldiaren gramatika euskaraz eta Gramatikaren Teoria (Meteorological expressions in Basque grammar and the Theory of Grammar). Anuario del Seminario de Filología Vasca “Julio de Urquijo” 52.1–2.3–60. Baker, Mark. 1988. Incorporation: A theory of grammatical function changing. Chicago, IL: University of Chicago Press. Baker, Mark, Kyle Johnson, and Ian Roberts. 1989. Passive arguments raised. Linguistic Inquiry 20.2.219–251. Béjar, Susana, and Milan Rezac. 2003. Person licensing and the derivation of PCC effects. In: Ana Teresa Pérez-LerouxandYvesRoberge,editors.Romancelinguistics:Theoryandacquisition.Amsterdam,theNetherlands: John Benjamins Publishing Company. 49–62. Berro, Ane. 2019. Basque adjectival participles: Stative, resultative or experiential. Glossa 4.66. Berro, Ane, and Beatriz Fernández. 2019a. Applicatives without verbs. Natural Language and Linguistic Theory 37.4.1273–1317. Berro, Ane, and Beatriz Fernández. 2019b. (Euskal) bozak: Erdikoak, inkoatiboak eta arazleak [(Basque) voices: Middles,inchoativesandcausatives].In:ItziarLaka,editor.Hitzaksarean:PelloSalabururieskeronez.Bilbao, the Basque Country: Euskal Herriko Unibertsitateko Argitalpen Zerbitzua. 39–50. Berro, Ane, Bernard Oyharçabal, and Beatriz Fernández. 2018. Inkoatibo/arazle alternantzia euskaraz eratortzen [Deriving the inchoative/causative alternation in Basque]. In: Beatriz Fernández and Ane Berro, editors. Euskaratik giza hizkuntzara: Jon Ortiz de Urbinari gorazarre. Special issue of Fontes Linguae Vasconum 126.99–122. Blevins, James P. 2003. Passives and impersonals. Journal of Linguistics 39.3.473–520. Bollebacher, John. 1977. The Basque passive. In:William A. Douglass, Richard W. Etulain, and William H. Jacobsen Jr., editors. Anglo-American contributions to Basque Studies: Essays in honor of Jon Bilbao.Reno,NV: Desert Research Institute. 181–192. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 183 Bonet, Eulalia. 1991. Morphology after syntax: Pronominal clitics in Romance. Doctoral thesis. Cambridge, MA: Massachusetts Institute of Technology. Bonet, Eulalia. 1994. The Person-Case constraint: A morphological approach. In:Heidi Harley and Colin Phillips, editors.The morphology-syntaxconnection.(MITWorkingPapersinLinguistics22.) Cambridge, MA:The MIT Press. 33–52. Bossong, Georg. 1985. Empirische Universalienforschung: Differentielle Objektmarkierung in den neuiranischen Sprachen. Tübingen, Germany: Gunter Narr Verlag. Bossong, Georg. 1991. Differential object marking in Romance and beyond. In: Dieter Wanner and Douglas A. Kibee,editors.NewanalysesinRomancelinguistics.Amsterdam,theNetherlands:JohnBenjaminsPublishing Company. 143–170. Brettschneider, Gunter. 1979. Typological characteristics of Basque. In: Frans Plank, editor. Ergativity: Toward a theory of grammatical relations. New York, NY: Academic Press. 371–384. Bruening, Benjamin. 2012. By phrases in passives and nominals. Syntax 16.1.1–41. Cabañas, Rebeca. 2006. La animacidad como factor de variación entre la pasiva perifrástica y la pasiva refleja [Animacyasthetriggeringfactorforvariationbetweentheanalyticpassiveandthereflexive-markedpassive]. Anuario de Letras: Lingüística y Filología 44.125–146. Chomsky, Noam. 1986. Knowledge of language: Its nature, origin and use. New York, NY: Praeger Publishers. Chomsky, Noam. 2000. Minimalist inquiries: The framework. In: Roger Martin, David Michaels, and Juan Uriagereka, editors. Step by step: Essays on Minimalist syntax in honor of Howard Lasnik.Cambridge,MA: MIT Press. 89–115. Chomsky, Noam. 2001. Derivation by phase. In: Michael Kenstowicz, editor. Ken Hale: A life in language. Cambridge, MA: MIT Press. 1–52. Collins, Chris. 2005. A smuggling approach to the passive in English. Syntax 8.2.81–120. Comrie, Bernard. 1977. In defense of spontaneous demotion: The impersonal passive. In: Peter Cole and Jerrold M. Sadock, editors. Grammatical relations. (Syntax and Semantics 8.) New York, NY: Academic Press. 47–58. Coon,Jessica.2010.ComplementationinChol(Mayan):Atheoryofsplitergativity.Doctoralthesis.Cambridge,MA: Massachusetts Institute of Technology. de Rijk, Rudolf. 2008. Standard Basque: A progressive grammar. Cambridge, MA: MIT Press. Doron, Edit. 2003. Agency and voice: The semantics of the Semitic templates. Natural Language Semantics 11.1.1–67. Eguzkitza, Andolin. 1981. On the so-called passive in Basque. Fontes Linguae Vasconum 37.233–254. Etxepare,Ricardo.2006.Numberlongdistanceagreementin(substandard)Basque.In:JosebaA.LakarraandJosé Ignacio Hualde, editors. Studies in Basque and historical linguistics in memory of R. L. Trask. Special issue of Anuario del Seminario de Filología Vasca “Julio de Urquijo” 40.1–2.303–350. Etxepare, Ricardo. 2014. Contact and change in a restrictive theory of parameters. In: Carme Picallo, editor. Linguistic variation in the Minimalist framework. Oxford, the United Kingdom: Oxford University Press. 108–139. Etxepare,Ricardo,andBernardOyharçabal.2009. Bi datiboegitura ipar-ekialdeko zenbait hizkeratan [Twodative structures in some northeastern varieties]. Lapurdum 13.145–158. Etxepare, Ricardo, and Bernard Oyharçabal. 2013. Datives and adpositions in northeastern Basque. In: Beatriz Fernández and Ricardo Etxepare, editors. Variation in datives: A micro-comparative perspective. Oxford, the United Kingdom: Oxford University Press. 50–95. Euskaltzaindia. 1987. Euskal gramatika: Lehen urratsak II. Bilbao, the Basque Country: Euskaltzaindia. Euskaltzaindia. 1991. Euskal gramatika: Lehen urratsak I. Bilbao, the Basque Country: Euskaltzaindia. Euskaltzaindia. 1993. Euskal gramatika laburra: Perpaus bakuna. Bilbao, the Basque Country: Euskaltzaindia. Fábregas,Antonio,andMichaelT.Putnam.2014.Theemergenceofmiddlevoicereadingswithandwithoutagents. Linguistic Review 31.2.193–240. Fernández,Beatriz.2010a.Goidatiboaketaaregoragokoakeuskaraz:Jabedatiboaketadatibohunkituakvs.datibo etikoak [High datives and even higher ones in Basque: Possessive datives and affected datives versus ethic datives]. Gogoa 10.1–2.1–20. Fernández, Beatriz. 2010b. Goieta behe-datiboak eta Datibo-komunztaduraren murriztapena: Ipar-ekialdeko hizkerak eta euskara estandarra [High and low datives and dative-agreement constraint: Northeastern varieties and standard Basque]. In: Irantzu Epelde, editor. Euskal dialektologia: Lehena eta oraina.(Anejos del 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 184 BERRO et al. AnuariodelSeminariode FilologíaVasca“JuliodeUrquijo” 69.)Bilbao,the BasqueCountry:Universityofthe Basque Country. 37–69. Fernández, Beatriz. 2015. The syntax of -(k)i, a Basque preposition. In: Xulio Sousa, Carlota de Benito, and Víctor Lara, editors. Syntactic variation in Western European Languages: From the noun phrase to clause structure. Special issue of Dialectologia 5.81–107. Fernández, Beatriz, and Berro Ane. 2021. Basque impersonals in comparison.Linguistics. https://doi.org/10.1515/ ling-2020-0057. Fernández, Beatriz, and Jon Ortiz de Urbina. 2010. Datiboa hiztegian [The dative in the dictionary]. Bilbao, the Basque Country: University of the Basque Country. Fernández, Beatriz, and Jon Ortiz de Urbina. 2012. Dative (first) complements in Basque. In: Ernestina Carrilho and Beatriz Fernández, editors. Syntactic microvariation in westmost European languages. Special issue of Journal of Portuguese Linguistics 11.1.83–98. Fernández, Beatriz, and Milan Rezac. 2010. Datibo osagarri bitxiak eta Datiboaren Lekualdatzea: Ari nai diyot eta kanta egin nazu bidegurutzean[Quirkydative objectsand dativedisplacement]. In:Beatriz Fernández, Pablo Albizu, and Ricardo Etxepare, editors. Euskara eta euskarak: Aldakortasun sintaktikoa aztergai.(Anejos del AnuariodelSeminariodeFilología Vasca“JuliodeUrquijo” 52.)Bilbao, theBasque Country:Universityofthe Basque Country. 113–149. Fernández, Beatriz, and Milan Rezac. 2016. Differential object marking in Basque varieties. In:Beatriz Fernández and Jon Ortiz de Urbina, editors. Microparameters in the grammar of Basque. Amsterdam, the Netherlands: John Benjamins Publishing Company. 93–139. Holmer, Arthur. 1999. An active analysis of Basque ergativity. Fontes Linguae Vasconum 31.189–225. Hualde,JoséIgnacio.1988.CaseassignmentinBasque.AnuariodelSeminariodeFilologíaVasca“JuliodeUrquijo” 22.1.313–330. Hualde, José Ignacio. 2003. Verbs: Finite forms. In: José Ignacio Hualde and Jon Ortiz de Urbina, editors. A grammar of Basque. Berlin, Germany: Mouton de Gruyter. 205–246. Ingason, Anton Karl, Julie Anne Legate, and Charles Yang. 2013. The evolutionary trajectory of the Icelandic new passive. University of Pennsylvania Working Papers in Linguistics 19.2.91–100. Irigoyen, Alfontso. 1992. Euskal litteratura zaharrari dagokiola verbo-systema periphrastikoan boz aktivoari oposatzen zaion boz pasivoaz ohar gisako elea berriz ere [Again a word about the passive voice opposing the activevoiceintheverbalsystemofoldBasqueliterature].In:Volume4ofDerephilologicalinguaeVasconicae. Bilbao, the Basque Country: Editorial Wilsen. 287–290. Kalin, Laura. 2017. Dropping the F-bomb: An argument for valued features as derivational time-bombs. In: Andrew Lamont and Katerina Tetzloff, editors. NELS 47: Proceedings of the forty-seventh annual meeting of the North East Linguistic Society. Amherst, MA: Graduate Linguistics Student Association, University of Massachusetts Amherst. 2.119–132. Kalin, Laura. 2018. Licensing and differential object marking: The view from Neo-Aramaic. Syntax 21.2.112–159. Kazenin, Konstantin I. 2001. The passive voice. In: Martin Haspelmath, Ekkehard König, Wulf Oesterreicher, and Wolfgang Raible, editors. Languagetypology andlanguage universals:Aninternational handbook.Berlin, Germany: Water de Gruyter. 2.899–916. Keenan, Edward L. 1985. Passive in the world’s languages. In: Timothy Shopen, editor. Clause structure. Volume 1ofLanguage typology and syntactic description, first edition. Cambridge, the United Kingdom: Cambridge University Press. 243–282. Keenan, Edward L., and Matthew S. Dryer. 2007. Passive in the world’s languages. In: Timothy Shopen, editor. Clause structure. Volume 1 of Language typology and syntactic description, second edition. Cambridge, the United Kingdom: Cambridge University Press. 1.325–361. Keyser, Samuel Jay, and Thomas Roeper. 1984. On the middle and ergative constructions in English. Linguistic Inquiry 15.3.381–416. Kibort, Anna. 2004. Passive and passive-like constructions in English and Polish. Doctoral thesis. Cambridge, the United Kingdom: University of Cambridge. Kramer,Ruth.2014.Cliticdoublingorobjectagreement:TheviewfromAmharic.NaturalLanguageandLinguistic Theory 32.2.593–634. Kratzer, Angelika. 1994. The event argument and the semantics of voice. Unpublished paper. Work conducted at: University of Massachusetts at Amherst. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 185 Kratzer, Angelika. 1996. Severing the external argument from its verb. In: Johan Rooryck and Laurie Zaring, editors. Phrase structure and the lexicon. Dordrecht, the Netherlands: Kluwer Academic Publishers. 109–137. Laka, Itziar. 1993. The structure of Inflection: A case study in X0syntax. In: José Ignacio Hualde and Jon Ortiz de Urbina, editors. Generative studies in Basque linguistics. Amsterdam, the Netherlands: John Benjamins Publishing Company. 21–70. Landau, Idan. 2010. The explicit syntax of implicit arguments. Linguistic Inquiry 41.3.357–388. Legate, Julie Anne. 2002. Walrpiri: Theoretical implications. Doctoral thesis. Cambridge, MA: Massachusetts Institute of Technology. Legate, Julie Anne. 2008. Morphological and abstract case. Linguistic Inquiry 39.1.55–101. Lekakou, Maria. 2005. In the middle, somehow elevated. Doctoral thesis. London, the United Kingdom: University College London. Levin,Beth. 1983.Onthenatureofergativity. Doctoralthesis.Cambridge,MA: MassachusettsInstituteofTechnology. Levin, Theodore. 2019. On the nature of differential object marking. Natural Language and Linguistic Theory 37.1.167–213. MacDonald, Jonathan E. 2017. An implicit projected argument in Spanish impersonaland passive-se constructions. Syntax 20.4.353–383. Maling, Joan. 2006. From passive to active. In: Benjamin Lyngfelt and Torgrim Solstad, editors. Demoting the agent: Passive, middle, and other voice phenomena. Amsterdam, the Netherlands: John Benjamins Publishing Company. 197–223. Maling, Joan, and Sigríður Sigurjónsdóttir. 2002. The new impersonal construction in Icelandic. Journal of Comparative Germanic Linguistics 5.1–3.97–142. Massam, Diane. 2002. Fully internal case: Surface ergativity can be profound. In: Andrea Rackowski and Norvin Richards,editors.ProceedingsofAFLA8:TheeighthmeetingoftheAustronesianFormalLinguisticsAssociation. (MIT Working Papers in Linguistics 44.) 185–196. Cambridge, MA: MIT Working Papers in Linguistics. Mendikoetxea, Amaya. 1999. Construcciones con se: Medias, pasivas e impersonales [Constructions with se:Middle,passive,andimpersonal].In:IgnacioBosqueandVioletaDemonte,editors.Lasconstruccionessintácticas fundamentales: Relaciones temporales, aspectuales y modales. Volume 2 of Gramática descriptiva de la lengua española. Madrid, Spain: Espasa-Calpe. 1631–1722. Mendikoetxea, Amaya. 2008. Clitic impersonal constructions in Romance: Syntactic features and semantic interpretation. Transactions of the Philological Society 106.2.290–336. Mounole, Céline. 2012. Evolution of the transitive verbs in Basque and apparition of datively marked patients. In: Gilles Authier and Katharina Haude, editors. Ergativity, valency, and voice. Berlin, Germany: Mouton de Gruyter. 335–379. Mounole, Céline. 2014. Le verbe basque ancien: Étude philologique et diachronique [The old Basque verb: A philological and diachronic study]. Anuario del Seminario de Filología Vasca “Julio de Urquijo” 48. Odria,Ane.2014. DifferentialobjectmarkingandthenatureofdativecaseinBasquevarieties.LinguisticVariation 14.2.289–317. Odria, Ane. 2017. Differential object marking and datives in Basque syntax. Doctoral thesis. Vitoria-Gasteiz, the Basque Country: University of the Basque Country. Odria, Ane. 2019. DOM and datives in Basque: Not as homogeneous as they might seem. In: Anna Pineda and Monica A. Irimia, editors. Differential objects and datives: A homogeneous class? Special issue of Lingvisticæ Investigationes 42.1.8–31. Ormazabal, Javier, and Juan Romero. 1998. On the syntactic nature of the me-lui and the Person Case Constraint. Anuario del Seminario de Filología Vasca “Julio de Urquijo” 32.2.415–433. Ormazabal, Javier, and Juan Romero. 2007. The Object Agreement Constraint. Natural Language and Linguistic Theory 25.2.315–347. Ormazabal, Javier, and Juan Romero. 2013. Object clitics, agreement and dialectal variation. Probus 25.2.301–344. Ormazabal, Javier, and Juan Romero. 2017. Historical changes in Basque dative alternations: Evidence for a P-based (neo)derivational analysis. Glossa 2.78. Ormazabal, Javier, and Juan Romero. 2019. The formal properties of non paradigmatic SE.Borealis 8.55–84. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 186 BERRO et al. Ormazabal, Javier, and Juan Romero. 2020. Deconstructing SE constructions: Number agreement and post-syntactic variation. Unpublished paper. Work conducted at: University of the Basque Country and University of Extremadura. OrtizdeUrbina,Jon.1988.Interferentziasintaktikoberezibatez[Onaspecialsyntacticinterference].In:Hizkuntz deskribapena. Volume 1 of II. Euskal Mundu-Biltzarra: Euskara Biltzarra. Vitoria-Gasteiz, the Basque Country: Eusko Jaurlaritzaren Argitalpen Zerbitzu Nagusia. 411–417. Ortiz de Urbina, Jon. 1989. Parameters in the grammar of Basque. Dordrecht, the Netherlands: Foris Publications. Ortiz de Urbina, Jon. 1991. Deskribapenetik azalpenetara euskal gramatikan [From description to explanation in Basquegrammar].In:AdolfoArejita,editor.Euskalsintaxiaz.Bilbao,theBasqueCountry:LabayruIkastegia. 11–27. Ortiz de Urbina, Jon. 2003. Impersonal clauses. In: José Ignacio Hualde and Jon Ortiz de Urbina, editors. A grammar of Basque. Berlin, Germany: Mouton de Gruyter. 572–591. Ortiz de Urbina, Jon. 2006. Old Basque (medio)passives. In: Joseba A. Lakarra and José Ignacio Hualde, editors. Studies in Basque and historical linguistics in memory of R. L. Trask. Special issue of Anuario del Seminario de Filología Vasca “Julio de Urquijo” 40.1–2.763–786. Ortiz de Urbina, Jon. 2011–2019. Aditz inpertsonalak [Impersonal verbs]. In: Pello Salaburu, Ibon Sarasola, and Patxi Goenaga, editors. Sareko euskal gramatika. Bilbao, the Basque Country: Basque Language Institute, University of the Basque Country. http://www.ehu.eus/seg/morf/5/12/2/2. Ortiz de Urbina, Jon, and Beatriz Fernández. 2016. Datives in Basque bivalent unergatives. In: Beatriz Fernández and Jon Ortiz de Urbina, editors. Microparameters in the grammar of Basque. Amsterdam, the Netherlands: John Benjamins Publishing Company. 67–93. Ortiz de Urbina, Jon, and Myriam Uribe-Etxebarria. 1991. Participial predication in Basque. In:Joseba A Lakarra, editor. Memoriae L. Mitxelena magistri sacrum. (Anejos del Anuario del Seminario de Filología Vasca “Julio de Urquijo” 14.) Bilbao, the Basque Country: University of the Basque Country. 2.993–1012. Pineda, Anna, and Ane Berro. 2020. Hybrid intransitives. Glossa 5.22. Preminger, Omer. 2009. Breaking agreements: Distinguishing agreement and clitic doubling by their failures. Linguistic Inquiry 40.4.619–666. Preminger, Omer. 2014. Agreement and its failures. Cambridge, MA: MIT Press. Pylkkänen, Liina. 2008. Introducing arguments. Cambridge, MA: MIT Press. Rebuschi, George. 1984. Structure de l’énoncé en basque [Structure of the utterance in Basque]. Paris, France: Société d’Études Linguistiques et Anthropologiques de France. Rezac, Milan. 2006. Agreement displacement in Basque: Derivational principles and lexical parameters. Unpublished paper. Work conducted at: University of the Basque Country. Rezac, Milan. 2007. Escaping the Person Case Constraint: Reference-set computation in the ϕ-system. Linguistic Variation Yearbook 6.97–138. Rezac,Milan.2008a.Phi-Agreeandtheta-relatedCase.In:DavidHarbour,DanielAdger,andSusanaBéjar,editors. Phi theory: Phi features across interfaces and modules. Oxford, the United Kingdom: Oxford University Press. 83–130. Rezac, Milan. 2008b. The syntax of eccentric agreement: The Person Case Constraint and absolutive displacement in Basque. Natural Language and Linguistic Theory 26.1.61–106. Rezac, Milan. 2009a. On the unifiability of repairs of the Person Case Constraint: French, Basque, Georgian, and Chinook.In:RicardoEtxepare,Ricardo Gómez,andJosebaA.Lakarra, editors.BeñatOihartzabaligorazarre. Special issue of Anuario del Seminario de Filología Vasca “Julio de Urquijo” 43.1–2.769–790. Rezac, Milan. 2009b. Person restrictions in Basque intransitives. Lapurdum 13.305–322. Rezac,Milan. 2011.Phifeaturesandthemodular architectureof language.London, the UnitedKingdom:Springer. Rezac, Milan. 2016. Gaps and stopgaps in Basque verb finite agreement. In: Beatriz Fernández and Jon Ortiz de Urbina, editors. Microparameters in the grammar of Basque. Amsterdam, the Netherlands: John Benjamins Publishing Company. 139–192. Rezac, Milan, Pablo Albizu, and Ricardo Etxepare. 2014. The structural ergative of Basque and the theory of Case. Natural Language and Linguistic Theory 32.4.1273–1330. 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PERSON MATTERS IN IMPERSONALITY 187 Richards, Marc. 2008. Defective Agree, case alternations, and the prominence of person. In: Marc Richards and Andrej L. Malchukov, editors. Scales.(Linguistische Arbeits Berichte 86.) Leipzig, Germany: Institute of Linguistics, University of Leipzig. 137–162. Rodet, Ana. 1992. Construccions amb lectura arbitrària en basc [Constructions with arbitrary interpretation in Basque]. Anuario del Seminario de Filología Vasca “Julio de Urquijo” 26.1.145–179. Rodríguez-Ordóñez,Itxaso.2016.DifferentialobjectmarkinginBasque:Grammaticalization,attitudesandideological representations. Doctoral thesis. Urbana, IL: University of Illinois at Urbana-Champaign. Rodríguez-Ordóñez, Itxaso. 2017. Reexamining Differential Object Marking as a linguistic-contact-phenomenon in Gernika Basque. Journal of Language Contact 10.2.318–352. Saltarelli, Mario, Miren Azkarate, David Farwell, Jon Ortiz de Urbina, and Lourdes Oñederra. 1988. Basque. London, the United Kingdom: Croom Helm. Sánchez, Cristina. 2002. Las construcciones con se [Constructions with se]. Madrid, Spain: Visor Libros. Sarasola,Ibon,PelloSalaburu,andJosuLanda.2001–2007.EreduzkoProsaGaur (ContemporaryReferenceProse). Database.Bilbao, the BasqueCountry:Basque Language Institute,Universityof theBasqueCountry.https:// www.ehu.eus/en/web/eins/ereduzko-prosa-gaur-epg-. Sarasola,Ibon,PelloSalaburu,andJosuLanda.2016a.EgungoTestuenCorpusa(CorpusofContemporaryBasque). Database. Bilbao, the Basque Country: Basque Language Institute, University of the Basque Country. http:// www.ehu.eus/etc. Sarasola,Ibon,PelloSalaburu,andJosuLanda.2016b.“Goenkale”Corpusa(“Goenkale”Corpus).Database.Bilbao, theBasque Country:BasqueLanguage Institute,University ofthe BasqueCountry.https://www.ehu.eus/en/ web/eins/goenkale-corpusa. Schäfer, Florian. 2008a. Middles as voiced anticausatives. In: Emily Elfner and Martin Walkow, editors. NELS 37: Proceedings of the thirty-seventh annual meeting of the North East Linguistic Society. Amherst, MA: Graduate Linguistics Student Association, University of Massachusetts Amherst. 2.183–196. Schäfer,Florian.2008b.Thesyntaxof(anti-)causatives:Externalargumentsinchange-of-statecontexts.Amsterdam, the Netherlands: John Benjamins Publishing Company. Siewierska,Anna.1984.Thepassive:Acomparativelinguisticanalysis.London,theUnitedKingdom:CroomHelm. Siewierska, Anna. 1988. The passive in Slavic. In: Masayoshi Shibatani, editor. Passive and voice.Amsterdam,the Netherlands: John Benjamins Publishing Company. 243–289. Siewierska,Anna.2013.Passiveconstructions.In:MatthewDryerandMartinHaspelmath,editors.Theworldatlas of language structures online. Leipzig, Germany: Max Planck Institute for Evolutionary Anthropology. http:// wals.info/chapter/107. Sigurðsson,EinarFreyr,and JimWood. 2020. Onthe implicit argumentofIcelandic indirectcausatives. Linguistic Inquiry 52.3.579–626. Sigurðsson, Halldór Ármann, and Verner Egerland. 2009. Impersonal null-subjects in Icelandic and elsewhere. Studia Linguistica 63.1.158–185. Taraldsen, Knut Tarald. 1995. On agreement and nominative objects in Icelandic. In: Hubert Haider, Susan Olsen,andStenVikner, editors.StudiesincomparativeGermanicsyntax.Dordrecht, theNetherlands:Kluwer Academic Publishers. 207–237. Urrestarazu, Iñigo. 2019. Nire barruko Setatsua: (Euskal) boza eta erdiko egiturak [(Basque) voice and middle constructions].Undergraduatethesis.Vitoria-Gasteiz,theBasqueCountry:UniversityoftheBasqueCountry. Woolford, Ellen. 2006. Lexical Case, inherent Case, and argument structure. Linguistic Inquiry 37.1.111–130. Zulaika, Esther. 1998. Ioanes Leizarragren aditza [Joanes Leizarraga’s verb]. Donostia–San Sebastián, the Basque Country: University of Deusto. Zúñiga, Fernando, and Seppo Kittilä. 2019. Grammatical voice. Cambridge, the United Kingdom: Cambridge University Press. How to cite this article: Berro, Ane, Ane Odria, and Beatriz Fernández. 2022. Person matters in impersonality. Syntax 25.147–187. https://doi.org/10.1111/synt.12230 14679612, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/synt.12230 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License