TYPE Original Research PUBLISHED 23 February 2023 DOI 10.3389/fpsyg.2023.1068058 On the distribution and OPEN ACCESS interpretation of voice in Greek anticausatives EDITED BY Artemis Alexiadou, Leibniz-Centre General Linguistics (ZAS), Germany REVIEWED BY Evripidis Tsiakmakis 1*, Joan Borràs-Comes 2 and Anna Roussou, M. Teresa Espinal 1* University of Patras, Greece 1 Department of Filologia Catalana/Center for Theoretical Linguistics, Autonomous University of Despina Oikonomou, Barcelona, Barcelona, Spain, 2 University of Barcelona, Barcelona, Catalonia, Spain Humboldt University of Berlin, Germany *CORRESPONDENCE This paper provides experimental evidence in support of the view that Greek Evripidis Tsiakmakis does not have three productive morphological classes of anticausative verbs, but Evripidis.Tsiakmakis@uab.cat M. Teresa Espinal only two: the class of verbs that bear non-active voice morphology and the class teresa.espinal@uab.cat of verbs that are morphologically active. Across two experiments, native Greek SPECIALTY SECTION speakers are found to prefer for each anticausative verb either non-active or active This article was submitted to voice morphological marking, in the presence or absence of explicit contextual Language Sciences, information. It is also shown experimentally that native speakers prefer an a section of the journal Frontiers in Psychology interpretation that involves a specific cause for all anticausatives, especially when the existence of such a cause is favored by the contextual setting. Our empirical RECEIVED 12October 2022 ACCEPTED 31 January 2023 findings are consistent with the view that the Voice Phrase that is realized as non- PUBLISHED 23 February 2023 active voice morphology in Greek anticausatives is expletive. From a theoretical CITATION perspective, we analyze the expletiveness of this Voice projection as the result of Tsiakmakis E, Borràs-Comes J and semantic redundancy: the Voice head of Greek anticausatives combines with a Espinal MT (2023) On the distribution and interpretation of voice in Greek anticausatives. v head that encodes a redundant cause meaning component and is, therefore, Front. Psychol. 14:1068058. interpreted merely as introducing an identity function. doi: 10.3389/fpsyg.2023.1068058 COPYRIGHT KEYWORDS © 2023 Tsiakmakis, Borràs-Comes and Espinal. This is an open-access article distributed under expletive voice, anticausatives, semantic redundancy, Greek, experimental results the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the 1. Introduction copyright owner(s) are credited and that the original publication in this journal is cited, in Across languages, there are groups of verbs that exhibit what is known as the causative accordance with accepted academic practice. No use, distribution or reproduction is alternation (Levin, 1993; Levin and Rappaport Hovav, 1995; Schäfer, 2008; Rappaport Hovav, permitted which does not comply with these 2014, a.o.): they display an intransitive variant describing a change of state and a transitive terms. variant describing a situation where somebody causes this change of state. The former, exemplified by (1a), is dubbed the anticausative variant, while the latter, exemplified by (1b), is labeled as the causative variant. (1) a. The door opened. Anticausative b. Jane opened the door. Causative In the present paper we will be focusing on the anticausative part of the alternation and on one language only, namely Greek. Greek verbs are morphologically marked for tense, aspect, voice, and agreement with the subject (Triantafyllidis, 1941; Holton et al., 1997, among many others; see also Ralli, 2005). Especially regarding voice, a dual distinction is made between active voice morphology (2a) and non-active voice morphology (2b).1 1 In the rest of the paper, the word marked refers exclusively to the presence of either active voice morphological marking or non-active voice morphological marking. Frontiers in Psychology 01 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 (2) a. ziyi-se Active voice morphology (7) I porta anikse. Class B weigh-perf.act.past.3.sg the door opened.act. ‘he weighed’ ‘The door opened.’ b. ziyi-stike Non-active voice morphology (8) To kotetsi gremistike/ gremise. Class C weigh-perf.nact.past.3.sg the hencoop crumbled.nact crumbled.act. ‘he was weighed’ ‘The hencoop crumbled.’ In the framework of Alexiadou et al. (2015), Under the morphology-syntax mapping adopted by Alexiadou Greek voice morphology, as exemplified in (2), is mapped onto voice et al. (2015), the three morphological classes of anticausative verbs syntax in the following way: Non-active voice marking above can be associated with corresponding syntactic classes: Class A (2b) is uniformly considered as the realization of a syntactic verbs are characterized by the projection of (non-active) VoiceP, Class non-active Voice Phrase (VoiceP), that does not project a specifier B verbs are characterized by the absence of (active) Voice, and Class (Embick, 1997) and appears in intransitive construals like C verbs are identified via the optionality of the projection of a (3) below. (non-active) VoiceP. The question is readily raised whether the three morphological and/or syntactic classes of anticausative verbs are (3) To pedhi vaftistike xtes. also mapped onto different meanings. Concretely put, do the kid baptize.perf.nact.past.3.sg yesterday anticausatives of one class receive a systematically different ‘The kid was baptized yesterday.’ interpretation from anticausatives of the other classes? And in the same vein, what is the interpretation of voice in Greek On the other hand, active voice marking (2a) may correspond to anticausatives? There are two ways to address these questions: two different structures: (i) the projection of an active VoiceP that either by comparing the meaning of Class A and Class B verbs does project a specifier and appears in transitive construals (4), or (ii) (projection of a VoiceP in a verb X vs. non-projection of VoiceP the absence of a VoiceP altogether, in the case of unaccusatives for in a verb Y), or by comparing Class C anticausative minimal pairs instance (5).2 [projection of a VoiceP in a verb X vs. non-projection of VoiceP in the same verb X; see example (8)]. Researchers interested in the topic have mostly followed the latter (4) O Kostas elise ton ghrifo. alternative. Alexiadou and Anagnostopoulou (1999, 2004), for the Kostas solve.perf.act.past.3.sg the riddle. example, report that Class C anticausatives with active voice ‘Kostas solved the riddle.’ morphology are incompatible with adverbs such as endelos (5) To luludhi anthise. ‘completely’, while their non-actively marked counterparts are the flower blossom.perf.act.past.3.sg not: ‘The flower blossomed.’ (9) a. To kotetsi gremise (#endelos). the hencoop crumbled.act completely Regarding specifically anticausative verbs, Alexiadou et al. (2015); see also Theophanopoulou-Kontou, 2000; Tsimpli, 2006) b. To kotetsi gremistike endelos claim that in Greek they can be divided into three morphologically the hencoop crumbled.nact completely distinct classes: verbs of Class A bear non-active voice morphological ‘The hencoop crumbled completely.’ marking (6), verbs of class B are morphologically active (7), and verbs of Class C can appear with either active or non-active voice morphology (8).3 The authors take the contrast to suggest that morphologically non-active Class C verbs [such as gremistike in (9b)] can convey either partial or complete change of state, while the ones bearing active voice (6) To karavi vithistike. Class A morphology receive only partial-change interpretation [hence the the boat sank.nact 4 ill-formedness of (9a)]. ‘The boat sank.’ It is important to note that Alexiadou and Anagnostopoulou (1999, 2004) report their own judgments. But even if we accept that their generalization holds, it is hard to link a partial vs. complete 2 We use lowercase voice to refer to the corresponding morphological change of state interpretative distinction to the presence (non-active category and uppercase Voice to refer to the corresponding syntactic category. voice morphology) vs. the absence of a VoiceP (active voice 3 The co-existence of morphologically distinct classes of anticausative verbs morphology), the semantics of which have been related mostly to the within the same language is a typologically frequent situation; see Haspelmath introduction (or existential binding) of the external argument since (1993, 2016). Kratzer (1996). 4 In the rest of the paper, we gloss only the morphological distinctions that Oikonomou (2014) focuses on a different contrast attributed to are relevant to our discussion, namely active vs. non-active voice. Class C anticausative verbs, as exemplified in (10): Frontiers in Psychology 02 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 (10) a. To ftero tu aftokinitu mu # tsalakose/ tsalakothike.5 the fender of.the car mine crumpled. act crumpled. nact. ‘The fender of my car crumpled.’ b. I fusta mu tsalakose/ tsalakothike. the skirt mine crumpled. act crumpled. nact ‘My skirt crumpled.’ [Oikonomou, 2014: 45, exs. (84a, b)] According to the author’s intuitions, Class C anticausatives with non-active voice morphology, and (ii) on the interpretation of these active voice morphology are infelicitous when the described change verbs. The study presented here pursues two goals: (i) to fill the gap of state is violently caused by an external initiator –as in the crumpling in the empirical argumentation by offering novel experimental of a car fender (10a). In the case of not externally (nor violently) evidence on the distribution and meaning of voice in Greek caused changes, as is the crumpling of a skirt (10b), both active and anticausatives, and (ii) to explore an analysis that explains the non-active Class C verbs are appropriate. compatibility of an expletiveness approach to Voice with an (external) Ιt appears that Oikonomou (2014) offers a feasible answer to what causation approach to anticausatives. the interpretation of voice in Greek anticausatives may be: it allows the The paper is structured as follows: Section 2 consists in a detailed external cause that brings about the described change of state to enter description of our experimental study. In Section 3 we discuss the the derivation as a semantic argument.6 However, this approach empirical and theoretical consequences of our findings regarding the should be taken with caution for two reasons: First, this author also distribution and interpretation of anticausative voice in Greek. Section builds on judgments not proven to be shared by other Greek speakers. 4 concludes the paper. Second, her proposal is developed around something very close to the internal vs. external causation distinction (Levin and Rappaport Hovav, 1995), the grammatical relevance of which has been cast under 2. The experimental study doubt (Rappaport Hovav, 2020). A radically different proposal is put forth by Alexiadou et al. An experimental study was designed in an attempt to gather (2015); see also Oikonomou and Alexiadou, 2022), who claim that no evidence concerning the speaker’s choice of non-active vs. active voice interpretative distinction can be systematically mapped to the morphological marking in Greek anticausatives, and contribute to presence vs. absence of a VoiceP in Greek anticausatives. The authors settling the debate laid out in the Introduction. Precisely, we aimed to argue that, instead, actively and non-actively marked Class C examine the hypothesis put forth in Alexiadou et al. (2015) that anticausatives are semantically equivalent as regards their event postulates that the Voice Phrase of Greek non-actively marked structure, and postulate that the VoiceP projected in Greek anticausatives is semantically expletive. anticausatives with non-active morphology is expletive (Schäfer, 2008; In order to test this hypothesis, we carried out our experimental Wood, 2014, 2015); it is interpreted as introducing an identity function study in two steps. First, we conducted a very simple acceptability (Winter, 2016) over predicates of events.7 judgment task with intransitive sentences whose main verb was Putting together the above, the interpretation of voice in Greek anticausative. Second, we investigated whether the speaker’s choice anticausatives has given rise to an unresolved debate. Setting the and interpretation of actively vs. non-actively marked anticausative partial vs. complete change discussion (Alexiadou and verbs was sensitive to contextual information. Concretely, the presence Anagnostopoulou, 2004) at the side, as its relevance was refuted by of brief contextual settings would allow us to test whether these the very authors that opened it (Alexiadou et al., 2015), we are left settings might condition the speakers’ preferences for (i) some specific with the external causation hypothesis (Oikonomou, 2014) on the morphological setup and (ii) a cause vs. no-cause interpretation of the one hand, and the expletiveness hypothesis (Schäfer, 2008), on the anticausative event. other. Still, all these previous studies do not support their respective In what follows, we present the two experiments that formed the analyses with broad experimental data that report (i) on the basis of our study. acceptability and naturalness judgments that Greek native speakers attribute to Class C anticausative verbs with either active or 2.1. Experiment 1 5 We have replaced Oikonomou’s (2014) question mark with the # symbol. 6 See Lavidas et al. (2012) for a view similar in essence, according to which As anticipated, the linguists interested in the interpretative non-active voice morphology in Greek alternating verbs correlates with the contribution of voice in Greek anticausatives focused mostly on Class existence of an implicit external argument. C verbs, that is the set of verbs that allegedly form both actively and 7 Non-actively marked anticausatives have been related also to reflexivity, non-actively marked anticausatives. We therefore thought that our especially with reference to Romance languages (Kallulli, 2006; Labelle, 2008; focus should be also on the nature of Class C verbs, in order to find Koontz-Garboden, 2009, among many others). This alternative is left aside out whether this class has the same status in the grammar of native because it has not been proposed, to our knowledge, in relation to Greek. See speakers as Classes A (only non-actively marked anticausatives) and Horvath and Siloni (2011), Alexiadou et al. (2015), Schäfer and Vivanco (2016), B (only actively marked anticausatives). Our research question in the for arguments against the reflexivity analysis of anticausatives. very first step of our study was whether native speakers have a clear Frontiers in Psychology 03 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 preference for associating a subset of Class C verbs with active By contrast, if speakers have a clear preference for one of the two morphology and a different subset with non-active morphology or forms of each verb and assign significantly different acceptability they do not. Two patterns seemed likely: Participants might not ratings to the members of each pair in our experiment, this can differentiate between the two morphological options; this would be regarded as evidence in support of the view that Class C is suggest that Class C, including verbs that randomly display active or not productive. non-active voice morphology, is productive in the grammar of Greek. The set of materials for Experiment 1 was complemented with 20 Alternatively, participants might systematically choose active voice control items: 10 sentences featured Class A anticausative verbs, i.e., morphology for some verbs and non-active voice morphology for a anticausatives that obligatorily bore non-active voice morphology different group of verbs; this would indicate that Class C is (metavalome ‘change’, vithizome ‘sink’, anatrepome ‘turn over’, not productive. peristrefome ‘rotate’, mionome ‘diminish’, epidhinonome ‘deteriorate’, To the above end, actively marked Class C anticausatives were veltionome ‘improve’, anaptisome ‘grow’, trelenome ‘go crazy’, ekrighnime contrasted directly to their counterparts with non-active voice ‘explode’); additionally, 10 sentences featured Class B anticausative marking. Participants had to rate the acceptability of sentences that verbs, i.e., anticausatives that were necessarily marked as active (alazo featured either one form or the other.8 The survey was administered ‘change’, vuliazo ‘sink’, anapodhoyirizo ‘turn over’, yirizo ‘rotate’, via Alchemer. lighostevo ‘diminish’, xiroterevo ‘deteriorate’, kaliterevo ‘improve’, meghalono ‘grow’, salevo ‘go crazy’, skao ‘explode’). The control items 2.1.1. Participants were included in order to confirm the participants’ competence to Experiment 1 was voluntarily completed by 90 native evaluate voice morphology independently of the specific questions that speakers of Greek (44 male, 44 female, 2 other; mean age 28.91 years, our experiment addressed. We further made sure that each Class A SD = 3.99), recruited via Facebook and other social media verb used had a synonym in the group of Class B verbs chosen for the platforms. experiment, which allowed us to maintain a control item design similar to the one of the criticals. Two minimal item pairs pertaining to the 2.1.2. Materials controls are given below, translated into English (see Appendix for the For the materials of this experiment, 10 anticausative verbs that set of items). allegedly belong to Class C (Alexiadou et al., 2015) were used: rayizo ‘crack’, madhao ‘pluck’, zarono ‘wrinkle’, erimono ‘desert’, zesteno ‘heat’, (13) a. I orasi tu veltiothike. Class A. lerono ‘sully’, dhialio ‘disperse’, gremizo ‘crumble’, tsalakono ‘crumple’, the vision his improved.nact. skizo ‘tear’. Each verb appeared in its active and its non-active voice morphological guise in affirmative sentences and as part of the same b. I orasi tu kaliterepse. Class B. sentential environment. This led to a total of 20 critical experimental the vision his improved.act. items. We provide two of the minimal item pairs featured in ‘His vision improved.’ Experiment 1 below, along with the English translation. See Appendix (14) a. To politiko skiniko metavlithike. Class A. for the full list of critical items. the political scenery changed.nact. b. To politiko skiniko alakse. Class B. (11) a. To plithos dhielise. the political scenery changed.act. the crowd dispersed.act. ‘The political scenery changed.’ b. To plithos dhialithike. the crowd dispersed.nact. ‘The crowd dispersed.’ We gave participants the following instructions: “In what follows, you will be presented with a set of sentences. Each sentence is followed (12) a. I tixi rayisan. by a scale. We ask you to use this scale to rate how good each sentence the walls cracked.act. is in your opinion (left extreme = bad, right extreme = good).” b. I tixi rayistikan. Our participants rated the total of items, producing 40 ratings the walls cracked.nact. each (20 critical items +20 control items). A sum of 3,600 responses (90 participants × 40 ratings) were statistically analyzed. ‘The walls cracked.’ 2.1.3. Procedure Recall that, if Class C is productive in the grammar of Greek Participants completed Experiment 1 using their personal native speakers, participants are not expected to distinguish between computer or smart device. First, they had to read the instructions and the active (11a, 12a) and the non-active voice morphology (11b, 12b). answer a brief questionnaire concerning their sociolinguistic background (see Appendix). Then the main task started, in which they 8 We acknowledge that the speakers’ ratings of the test sentences may were asked to read sentences in isolation (i.e., in total absence of be affected by factors other than their preference for active or non-active context) and use a scale to report whether they found each sentence morphological marking on a given verb, such as the type of the internal good or bad. argument. However, upon designing the experimental study, we chose a limited The items were randomized. An example follows of what set of experimental conditions. participants saw on their screens, translated into English. Frontiers in Psychology 04 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 (15) To pandeloni skistike. the pants teared.nact ‘The pants teared.’ kaki kali ‘bad’ ‘good’ The median duration of the experiment was 6′ 93″. Concerning the second part of our research question, we again Before we present the results of this acceptability judgment task, considered two alternatives: Participants might interpret or we introduce in Section 2.2 a modified variant of Experiment 1, one accommodate a specific cause for the anticausative event when the that required the evaluation of sentences against specific contextual context makes such a cause salient, but deny the existence of a specific settings and aimed to investigate not only the speaker’s preference for cause when the contextual setting is not sufficiently informative or active vs. non-active voice morphology but also the meaning ascribed restrictive in this respect. This would suggest that speakers base their to anticausative verbs bearing this morphology. interpretation on contextual information. Alternatively, participants might interpret that there is a specific cause when the verb bears non-active voice morphology but prefer an interpretation where no 2.2. Experiment 2 concrete cause for the described change of state is specified for verbs with active voice marking. This would suggest that speakers are more Our second experiment aimed at getting additional evidence sensitive to morphological than contextual information when concerning the original hypothesis we investigated, according to interpreting anticausatives. which non-active voice in Greek anticausatives is semantically Given this twofold research question, we tested the active vs. expletive. Recall that while Experiment 1 meant to investigate whether non-active morphology distinction in Class C anticausatives against speakers have a clear preference for one of the two anticausative two types of contexts: (i) a context that introduced via linguistic variants of so-called Class C verbs in the absence of contextual means an external cause or initiator interpreted as causing the information, the purpose of Experiment 2, which was also based on described change of state (e.g., a fire, a storm, an earthquake) –let us an acceptability judgment task, was to explore whether the call this the overt cause context, and (ii) a context where no such acceptability of active vs. non-active forms and their interpretation external initiator bringing about the change of state was explicitly differ when participants are forced to directly contrast the actively and introduced in the situation description –let us dub this the non-overt the non-actively marked anticausative of a Class C verb against cause context.9 In Experiment 2, participants were asked to read some different contextual settings. texts consisting of brief contexts followed by short sentences, rate the Given that there are infinite aspects of the contextual setting that naturalness of the sentences, and provide judgments on their possible could possibly affect the preference for and interpretation of active or interpretations: a cause interpretation that identified a specific external non-active voice marking on anticausative Class C verbs, and taking cause in the described event, and an interpretation that negated the into account the previous literature (namely Oikonomou, 2014), existence of a specific cause, labeled as no-cause interpretation merely we identified the saliency of an external cause of the described change for ease of reference. This survey was also administered via Alchemer. of state as the most relevant factor. Concretely, our research question at this second stage of our study was twofold; we meant to test (i) 2.2.1. Participants whether the speakers’ preference for active or non-active voice A total of 76 participants (20 male, 55 female, 1 other; mean age morphology on a Class C anticausative verb is sensitive to the (non-) 29.91 years, SD = 5.69), all native speakers of Greek, completed saliency of an external cause linguistically phrased in the description Experiment 2. In this case, too, we used Facebook and other social of the situational environment, and (ii) whether the speakers’ media platforms to recruit participants. interpretation of an anticausative as involving a specific cause or not is sensitive to the voice morphology of the verb or the description of 2.2.2. Materials a (non-)salient external cause linguistically phrased in the For the critical items of Experiment 2 we used the same 10 verbs as contextual setting. in Experiment 1; in fact, we used exactly the same sentences. However, As regards the first part of our research question, there were two this time we further introduced a contextual setting. Accordingly, each possibilities: Speakers might not systematically associate the verb appeared in its active and its non-active variant, and each variant existence of a salient external cause with one voice morphology or was preceded by an overt cause context and a non-overt cause context, the other; this would suggest that no connection can be established bringing our critical items to a total of 40 (see Appendix for the full list between voice morphology and the presence or absence of cause- of critical items). Below we show how our example (12) from Experiment related information. (Note that, in that case, we would expect to 1 was adapted for the purposes of Experiment 2. The contexts are here obtain a morphological preference pattern close to that obtained translated into English for the reader’s convenience. from Experiment 1). Alternatively, speakers might relate systematically the non-active vs. active voice morphology distinction to the (non-)salient external cause distinction, regardless of the 9 In the case of non-overt cause contexts, there was still the possibility that specific verb; this would imply that the choice of one morphological participants accommodated some cause for the described change of state, variant over the other reflects information regarding the cause of the although one of a different type than the external initiator introduced explicitly described change. by overt cause contexts. Frontiers in Psychology 05 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 (16) a. Overt cause context – Active voice. ‘His vision improved.’ [The earthquake was rather strong. The seven-storey b. Non-overt cause context – Active voice. building was deeply shaken.] [The puppy was so young he could not even see. Weeks passed.] I tixi rayisan. I orasi tu kaliterepse. the walls cracked.act. the vision his improved.act. ‘The walls cracked.’ ‘His vision improved.’ b. Non-overt cause context – Active voice. c. Overt cause context – Non-active voice. [After seventy years, the building was now uninhabitable. [Grandpa was complaining that he could not see from one eye. I got The signs were obvious.] him a basic collyrium from the pharmacy.] I tixi rayisan. I orasi tu veltiothike. the walls cracked.act. the vision his improved.nact. ‘The walls cracked.’ ‘His vision improved.’ c. Overt cause context – Non-active voice. d. Non-overt cause context – Non-active voice. [The earthquake was rather strong. The seven-storey [The puppy was so young he could not even see. Weeks passed.] building was deeply shaken.] I orasi tu veltiothike. I tixi rayistikan. the vision his improved.nact. the walls cracked.nact. ‘His vision improved.’ ‘The walls cracked.’ d. Non-overt cause context – Non-active voice. Recall that the research question that we addressed via Experiment [After seventy years, the building was now uninhabitable. 2 also had a second part, related to whether participants interpret a The signs were obvious.] specific cause when encountered with anticausative event descriptions I tixi rayistikan. or not. Aiming at exactly this, every test sentence was followed by two the walls cracked.nact. possible interpretations: one that attributed the change of state ‘The walls cracked.’ described by the sentence to a specific cause accessible from the contextual setting and one that excluded the existence of a concrete cause for the same change of state. The prediction derived from the The context in (16a,c), in contrast with the context in (16b,d), stated general hypothesis regarding the expletiveness of anticausative voice explicitly an external initiator for the change of state described in the test (Alexiadou et al., 2015) was that voice morphology would not affect the sentence; it was the strong earthquake that is meant to have caused the interpretation. With context being the main source of relevant walls to crack. It is in this sense that we considered the former an overt information, the 40 items that were introduced by an overt cause cause context, contrasting with the latter that included no overt initiator; context would elicit high saliency ratings for the cause interpretation, the cracking of the walls is attributable to the effect of time or the while the 40 remaining items, introduced by a non-overt cause context, tendency of all material things to perish. If anticausative voice is indeed would lead to high saliency ratings for the no-cause interpretation. expletive (Alexiadou et al., 2015), then the (non-)overtness of the cause The participants were given the following instructions: “In what is not expected to interact with the speakers’ preference for active or follows you will read a set of small texts. Each text consists of the non-active voice morphology. In other words, no significant differences description of a situation followed by an utterance. First, we ask you to between (16a) and (16b) on the one hand, and (16c) and (16d) on the rate how natural each utterance is with respect to the situation using other, are to be found. It is further expected that the difference between a scale (left edge = totally unnatural, right edge = absolutely natural). (16a,b) and (16c,d) will mirror the findings of Experiment 1. Second, we ask you to use a similar scale to rate how salient each of Having in mind a potential comparison of the results from the two provided interpretations of the utterance is, always in relation Experiment 1 and Experiment 2, we decided to keep in the latter the to the situation (left edge = impossible, right edge = extremely set of control items used for the former (Class A and B anticausatives). possible).”10 The controls were also adapted by including contextual information, so as to have a structure parallel to the criticals. Consider the 10 The reader may notice at this point that the instructions to the two reformulation of example (13) in (17), with the contexts translated experiments differed. The sentences in Experiment 1 appeared in isolation and into English again here for expository purposes. See Appendix for the participants had to rate them on a “bad-to-good” scale. The sentences in set of items used in Experiment 2 in their original form. Experiment 2, on the other hand, were preceded by specific contextual information and participants had to judge them on a “totally unnatural-to- absolutely natural” scale with respect to the preceding context. Given that the (17) a. Overt cause context – Active voice. results obtained from the two experiments showed mostly a similar behavior [Grandpa was complaining that he could not see from one eye. I got when looking at the item-specific variation, especially in the case of critical him a basic collyrium from the pharmacy.] items, and in order to simplify the exposition, the results of the two experiments I orasi tu kaliterepse. are presented jointly. Concretely, in the description of the results, we use the vision his improved.act. acceptability as a superordinate term that includes both the good vs. bad ratings and the natural vs. unnatural ratings elicited from the participants. Frontiers in Psychology 06 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 This second experiment included the same test sentences as of the same 10 alleged Class C verbs in the absence and in the presence Experiment 1 (20 criticals +20 control items), each appearing in 2 of context–, we report our results in a single section, divided into two different contextual settings. This led to a total of 80 experimental parts. First, we analyze the acceptability judgments obtained via items. In order to reduce the expected duration of the task, we decided Experiments 1 and 2, which have been combined into a single to split the items in half and create two different versions of database. The possible effect of Voice (active, non-active) is analyzed Experiment 2. Each version included 20 critical items (5 critical verbs across every specific item as well as the occurrence of each item in the × 4 conditions) and 20 control items (5 control verb meanings × 4 different contextual conditions. This analysis is run separately on conditions). Our participants rated all the items, producing 3 control and critical items. Second, we analyze the saliency of the two different ratings for each: one related to the naturalness of the test possible interpretations that have been provided in Experiment 2, as sentence and two related to the saliency of the two possible reported by the participants. interpretations. A total of 9,120 responses (76 participants × (40 × 3) Concerning the statistical analyses, a series of beta mixed-effects ratings) were statistically analyzed. ANOVAs were performed using the glmmTMB package in R. To fulfill the requirements of a beta distribution, the 0–100 response values 2.2.3. Procedure obtained were first divided by 100 (to obtain a 0–1 distribution), and The procedure followed for Experiment 2 was similar to the then the two ends were replaced by very close values (0.0000001 for 0, one described for Experiment 1. Participants carried out the and 0.9999999 for 1). The omnibus test results are reported, which are required tasks using their personal computers or smart devices. complemented with Sequential Bonferroni-corrected pairwise After reading the instructions, they filled in the same sociolinguistic contrasts (obtained using the emmeans package) and Cohen’s d as a questionnaire as in the first experiment (see Appendix). Then, they measure of effect size. In each reported model, the chosen random moved to the main part of Experiment 2. For each item, effects’ structure was the most complex structure providing no model participants were presented with a context, a test sentence and two convergence problems. possible interpretations of the test sentence. First, they were asked to rate the naturalness of the sentence with respect to the context. 2.3.1. Acceptability results Then, they were asked to decide how salient each of the interpretations was, in relation to the context-sentence pair they 2.3.1.1. Acceptability results for control items had to evaluate. Figure 1 displays the results of the perceived acceptability ratings The items were randomized. Below we demonstrate what among the control items of Experiments 1 and 2. As can be seen, there participants saw on their screens, using the example from (16a). Both is a preference for non-active voice items, a sort of preference for items the instructions and the example are here translated into English for presented without a context, and, among the items that did involve a the reader’s convenience. context, a preference for those with an overt cause one. In general, our (18) [The earthquake was very strong. The seven-storey building was deeply shaken.] I tixi rayisan. the walls cracked.act. ‘The walls cracked.’ Rate how natural the utterance is in relation to the situation. katholou fisiko apolita fisiko ‘totally unnatural’ ‘absolutely natural’ Rate how salient each interpretation of the utterance is in relation to the situation. a. An earthquake was the cause for what happened. katholu pithani ekseretika pithani. ‘impossible’ ‘extremely possible’. b. There was no specific cause for what happened. katholu pithani ekseretika pithani. ‘impossible’ ‘extremely possible’. The median duration of the experiment was 18′84″. control items received mean acceptability ratings higher than 70%, indicating that our participants were indeed capable of providing judgments on Greek voice morphology. 2.3. Results A beta mixed-effects model was performed for the acceptability responses for control items of Experiments 1 and 2. The fixed factors Due to the comparable design of the two experiments –testing the were Voice (i.e., active, non-active), Context (i.e., no context, overt same sentences containing active and non-active anticausative pairs cause context, non-overt cause context) and their paired interaction. Frontiers in Psychology 07 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 The random effects’ structure included a random intercept for Subject experiments into a single database and compared the ratings attributed plus a random slope for Context by Item. to the two anticausative verbs within each pair of synonyms. The omnibus test results showed a significant main effect for both A model was run including Voice (active, non-active), Context Voice, χ2(1) = 11.676, p = 0.001, and Context, χ2(2) = 10.152, p = 0.006, (no context, overt cause context, non-overt cause context), the specific but no significant interaction. The main effect of Voice relates to an Item (pair of synonyms), and all their possible interactions as fixed overall greater acceptability of items presented with non-active factors, with a random intercept for Subject. morphology (d = 0.177, p = 0.001). The main effect of Context indicates The three main effects and two paired interactions were found to lower acceptability rates for items presented with a non-overt cause be significant, including the double interaction Voice × Item, context, compared to those presented with no context at all (d = 0.696, χ2(9) = 30.610, p < 0.001. This was not the case for the triple interaction, p = 0.016), and those presented with an overt cause context (d = 0.381, χ2(18) = 16.050, p = 0.589. The results of the pairwise contrasts for the p = 0.039), with no significant difference between the latter (d = 0.315, Voice × Item interaction in which Voice is taken as the contrast field p = 0.561). can be summarized in Table 1. An analysis of the potential effect of Voice within every specific As for the non-significant interaction Voice × Context × Item, the control item was also performed. Recall that the control items used in results of the pairwise contrasts show a scarce statistical relevance Experiments 1 and 2 were built around 20 anticausative verbs, for Voice, i.e., only for three out of the thirty combinations of Item organized in 10 pairs of synonymous lexical items. Each pair consisted and Context. In this respect, alazo vs. metavalome ‘change’ shows a of one verb that obligatorily displayed active voice morphology (Class preference for active voice when no context is provided (d = 0.419, B) and one verb that was obligatorily marked as non-active (Class A). p = 0.049), kaliterevo vs. veltionome ‘improve’ shows a preference for With this in mind, we combined the participants’ responses to both non-active voice when no context is provided (d = −1.115, p < 0.001), and vuliazo vs. vithizome ‘sink’ shows a preference for non-active voice when a non-overt cause context is provided (d = −0.665, p = 0.047). 2.3.1.2. Acceptability results for critical items Figure 2 displays the results of the perceived acceptability ratings among the critical items of Experiments 1 and 2. As can be seen, the acceptability obtained does not vary much across the different context conditions, even though a generalizable preference for non-active voice forms over active ones seems to occur. Nevertheless, the statistical results below indicate that this preference for non-active voice items is just an artifact caused by the specific verbs selected for the experimental tasks. Specifically, six out of the 10 verbs tested display a preference for non-active voice forms, three of the tested verbs display a preference for active voice forms, and one can take either the active or the non-active FIGURE 1 morphology (see Table 2 below). Acceptability results of control items in Experiments 1 and 2. A beta mixed-effects model was performed for the acceptability responses for critical items of Experiments 1 and 2. Again, the fixed TABLE 1 Mean (and standard deviation) values for the reported acceptability of each specific control item across active and non-active voice in Experiments 1 and 2. Mean (SD) acceptability reported Pairwise contrasts Item: Class B/Class A Active voice Non-active voice Cohen’s d Value of p alazo/metavalome ‘change’ 85.08 (20.47) 82.29 (20.98) 0.188 0.269 salevo/trelenome ‘go crazy’ 77.83 (28.21) 86.25 (19.71) −0.366 0.037 xiroterevo/epidhinonome ‘deteriorate’ 83.93 (22.87) 89.07 (18.00) −0.071 0.680 meghalono/anaptisome ‘grow’ 79.41 (23.11) 81.80 (23.93) −0.192 0.272 lighostevo/mionome ‘diminish’ 66.41 (31.30) 67.45 (32.29) −0.189 0.280 skao/ekrighnime ‘explode’ 75.96 (31.57) 82.42 (26.84) −0.174 0.347 kaliterevo/veltionome ‘improve’ 74.78 (29.83) 90.28 (17.21) −0.596 0.001 anapodhoyirizo/anatrepome ‘overturn’ 81.97 (25.33) 87.12 (19.27) −0.273 0.105 yirizo/peristrefome ‘rotate’ 60.74 (34.84) 58.16 (36.16) 0.266 0.134 vuliazo/vithizome ‘sink’ 82.89 (23.85) 87.41 (20.54) −0.341 0.044 The Active voice column corresponds to Class B verbs, while the Non-active voice column corresponds to the verbs of Class A. The last two columns indicate the results of the pairwise contrasts associated with the significant interaction Voice × Item. Statistically significant results appear in bold. Frontiers in Psychology 08 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 factors were Voice (i.e., active, non-active), Context (i.e., no context, no significant difference between the latter two (d = 0.224, p = 0.620). overt cause context, non-overt cause context) and their paired On the other hand, when non-active morphology is used, items interaction. The random effects’ structure included a random intercept accompanied with a non-overt cause context receive lower acceptability for Subject plus a random slope for Voice by Item. ratings, compared to those with overt cause contexts (d = 0.572, The omnibus test results showed significant results for Context, p = 0.003) and those presented without a context (d = 0.493, p = 0.021), χ2(2) = 14.501, p = 0.001, and for the paired interaction Voice × with no significant difference between the latter (d = 0.079, p = 1.000). Context, χ2(1) = 40.473, p < 0.001. However, no significant main effect An additional statistical model was run over the acceptability of was found for Voice, χ2(1) = 1.447, p = 0.229 (which is in line with the critical items, including Voice (active, non-active), Context (no hypothesis that the effect of Voice is verb-specific). The main effect of context, overt cause context, non-overt cause context), the specific Context indicates higher acceptability rates for items presented with Item, and all their possible interactions as fixed factors. The model an overt cause context, compared to those presented with no context included a random slope for Voice by Subject. In this analysis, we were at all (d = 0.605, p = 0.015), and those presented with a non-overt cause interested in the potential effect of Voice within every specific item context (d = 0.564, p = 0.014), with no significant difference between and every contextual condition in which each item has been presented. the latter (d = 0.041, p = 1.000). All main effects and interactions were found to be significant. The The interaction Voice × Context can be interpreted as such that ones of interest to us are, first, the paired interaction Voice × Item, different preferences for Context conditions are found when exploring χ2(9) = 123.220, p < 0.001, and, second, the triple interaction Voice × active or non-active morphology. On the one hand, when active Context × Item, χ2(18) = 75.480, p < 0.001. The pairwise contrasts morphology is used, items without a context receive lower acceptability associated with the paired interaction Voice × Item are summarized in ratings, compared to those with overt cause contexts (d = 0.939, Table 2: while active voice morphology is preferred for rayizo ‘crack’, p < 0.001) and with non-overt cause contexts (d = 0.714, p = 0.007), with madhao ‘pluck’, and zarono ‘wrinkle’, non-active voice is preferred for gremizo ‘crumble’, tsalakono ‘crumple’, dhialio ‘disperse’, zesteno ‘heat’, lerono ‘sully’, and skizo ‘tear’, and no significant preference is found for erimono ‘desert’. Regarding the effect of Voice in the triple interaction Voice × Context × Item, it is statistically relevant only for two verbs, i.e., erimono ‘desert’ and zesteno ‘heat’. In these cases, Voice plays a role in the reported acceptability only when no contextual information is provided (i.e., only in Experiment 1), but not when there is an overt or a non-overt cause context. Specifically, erimono shows a significant preference for active voice (d = 0.720, p = 0.026), and zesteno displays a significant preference for non-active voice morphology (d = −2.631, p < 0.001). A post hoc analysis was run. A single variable indicating the participants’ overall preference for active vs. non-active forms was obtained from the acceptability responses to the critical items of both Experiment 1 and 2. Using the function partykit::ctree(), the tested FIGURE 2 verbs were found to be best classified into two groups: Group A, Acceptability results of critical items in Experiments 1 and 2. featuring the verbs with high preference for active forms (erimono ‘desert’, madhao ‘pluck’, zarono ‘wrinkle’, rayizo ‘crack’), and Group B, TABLE 2 Mean (and standard deviation) values for the reported acceptability of each specific critical item across active and non-active voice morphology in both Experiment 1 and 2. Mean (SD) acceptability reported Pairwise contrasts Item Active voice Non-active voice Cohen’s d Value of p rayizo ‘crack’ 83.56 (23.23) 20.84 (28.75) 4.046 <0.001 gremizo ‘crumble’ 26.70 (32.76) 88.24 (18.94) −3.489 <0.001 tsalakono ‘crumple’ 27.02 (32.56) 87.07 (20.36) −3.936 <0.001 erimono ‘desert’ 81.73 (25.94) 76.88 (29.97) 0.286 0.283 dhialio ‘disperse’ 32.06 (36.01) 87.42 (20.38) −3.227 <0.001 zesteno ‘heat’ 55.16 (36.83) 77.73 (24.95) −1.116 <0.001 madhao ‘pluck’ 64.90 (34.49) 29.89 (32.60) 2.593 <0.001 lerono ‘sully’ 30.88 (33.46) 79.70 (29.02) −3.021 <0.001 skizo ‘tear’ 21.48 (29.77) 87.04 (18.84) −3.953 <0.001 zarono ‘wrinkle’ 74.65 (30.42) 38.08 (34.81) 2.465 <0.001 The last two columns indicate the results of the pairwise contrasts associated with the significant interaction Voice × Item. Statistically significant results appear in bold. Frontiers in Psychology 09 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 including the verbs with low preference for active forms (gremizo 2.3.2. Interpretation results ‘crumble’, lerono ‘sully’, dhialio ‘disperse’, skizo ‘tear’, zesteno ‘heat’, Figure 3 illustrates how salient the participants consider the two tsalakono ‘crumple’). possible interpretations offered in Experiment 2. While the first row The verb classification variable, labeled as VGroup, was modeled depicts the reported saliency of a cause interpretation for items along with Voice and Context (with all their possible interactions) in presented with an overt cause context (left) and a non-overt cause a new generalized linear mixed model. Contrasting the acceptability context (right), the second row shows the reported saliency of a between the overt and non-overt cause contexts for each combination no-cause interpretation in the same types of contexts. Overt cause of VGroup × Voice, Group A verbs with active morphological contexts favor a cause interpretation and disfavor a no-cause marking (felicitous combination, mean acceptability = 0.763) showed interpretation; non-overt cause contexts disfavor a cause no significant preference for any context; Group A verbs with interpretation, while being unclear as regards no-cause interpretations. non-active morphological marking (infelicitous, M = 0.419) showed A statistical model was run taking the reported saliency of the a preference for overt cause contexts vs. no context (d = 0.764, specific interpretation as the dependent variable. It included as fixed p = 0.008); Group B verbs with active morphological marking factors Voice (active, non-active), Context (overt cause context, (infelicitous, M = 0.321) showed a significant preference for context- non-overt cause context), Interpretation (cause interpretation, accompanied stimuli (d = 1.07–1.43, p < 0.001); finally, Group B verbs no-cause interpretation), and Item Type (critical, control). The with non-active morphological marking (felicitous, M = 0.846) were random effects’ structure included a random slope for both Context dispreferred when presented with a non-overt cause context and Interpretation by Subject, plus a random intercept for Item. (d = 0.75–1.07, p ≤ 0.002).11 Four fixed effects were found to be significant, i.e., two main effects and two interactions. The significant main effects were Context, 11 The reviewers soundly raise questions regarding potential inter-speaker χ2(1) = 67.789, p < 0.001, and Interpretation, χ2(1) = 54.487, p < 0.001, variation. We report that the tendencies described above were found to though they are better explained by looking at their paired interaction, be consistent across subjects, with only zarono ‘wrinkle’ and dhialio ‘disperse’ which was also found to be significant, χ2(1) = 2185.895, p < 0.001. The showing some variability. paired interaction Context × Interpretation can be explained in two FIGURE 3 Reported saliency of the two possible interpretations offered in Experiment 2: cause and no-cause interpretations (by rows), for items presented with an overt cause context or a non-overt cause context (by columns), across Item Type and Voice. Frontiers in Psychology 10 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 complementary ways. First, cause interpretations are rated higher in the majority of alternating verbs, speakers know that the anticausative overt cause contexts than in non-overt cause contexts (d = 4.161, variant shows a tendency toward either active voice marking or p < 0.001), and no-cause interpretations are considered more salient in non-active voice marking. Consequently, participants systematically non-overt cause contexts than in overt cause contexts (d = −2.127, choose active voice morphology for some verbs and non-active voice p < 0.001). Alternatively, for overt cause contexts, cause interpretations morphology for a different group of verbs, thus indicating that Class receive higher ratings than no-cause interpretations (d = 4.340, C of Greek anticausatives is not productive (cf. Alexiadou et al., p < 0.001), whereas for non-overt cause contexts, no-cause 2015).13 That said, our results further show that membership of a interpretations receive higher ratings than cause interpretations given anticausative verb in a morphological verb class is not (d = 1.947, p < 0.001). Lastly, the other interaction found to categorical. In this sense, gremizo ‘crumble’ could be considered as a be significant was the triple interaction Context × Interpretation × more representative member of Class A than zesteno ‘heat’; see Item Type, χ2(1) = 8.352, p = 0.004, which can be related to the fact that, Table 2. in assigning a cause interpretation to a non-overt cause context, Experiment 2 tested the same active vs. non-active voice control items obtained higher ratings than critical items (d = 0.290, distinction in Class C anticausatives against a contextual setting that p = 0.013). No significant effect of Voice was found whatsoever. was phrased in such a way that favored a (non-)salient cause (i.e., Since we were interested in knowing whether there is a significant overt cause vs. non-overt cause context). Interestingly, with the difference between the two less preferred interpretations (i.e., cause exception of erimono ‘desert’ and zesteno ‘heat’ which showed a Class interpretation of non-overt cause contexts vs. no-cause interpretation C behavior in this second experiment (no significant preference for of overt cause contexts) and between the two preferred ones (i.e., cause either active or non-active voice marking), the results of Experiment interpretation of overt cause contexts vs. no-cause interpretation of 1 were replicated: gremizo ‘crumble’, tsalakono ‘crumple’, dhialio non-overt cause contexts), an additional statistical model was run in ‘disperse’, lerono ‘sully’, and skizo ‘tear’ behaved as Class A verbs, while which the combination of Context and Interpretation was modeled as rayizo ‘crack’, madhao ‘pluck’, and zarono ‘wrinkle’ displayed a Class a single variable with four levels (i.e., the four panels in Figure 3). The B behavior. The findings of Experiment 2 suggest that the speakers’ results indicate no significant differences between the two less preference for either active or non-active morphological marking in preferred interpretations (d = 0.178, p = 0.483), and significantly Greek anticausatives is not sensitive to the saliency of an external greater values for the cause interpretation of overt cause contexts initiator bringing about the described change of state (cf. Oikonomou, compared to the no-cause interpretation of non-overt cause contexts 2014). Moreover, by reproducing the preference pattern obtained via (d = 1.888, p < 0.001). The rest of the effects described above were Experiment 1, this second experiment offers additional evidence found intact.12 supporting the conclusion obtained from Experiment 1, namely that Class C is not productive in the grammar of native Greek speakers. The results of Experiment 2 lead to another interesting conclusion 3. Discussion: The role of voice in relative to the interpretation of Greek anticausatives. Recall that, in Greek anticausatives addition to rating the naturalness of the test sentence, we asked participants to evaluate the saliency of two different interpretations at Let us now take stock of our findings. Experiment 1 tested the the end of each item –one that was linked to the specific cause active vs. non-active voice morphology distinction in 10 Class C mentioned in the contextual setting, and one that denied the existence (Alexiadou et al., 2015) anticausative verbs. Crucially, all the verbs of such a link. Irrespective of the voice morphology on the that formed part of our experiment were found to behave as members anticausative verb, participants provided low saliency ratings for those of either Class A (gremizo ‘crumble’, tsalakono ‘crumple’, dhialio interpretations that did not match their respective context, i.e., cause ‘disperse’, zesteno ‘heat’, lerono ‘sully’, skizo ‘tear’) or Class B (rayizo interpretations under non-overt cause contexts and no cause ‘crack’, erimono ‘desert’, madhao ‘pluck’, zarono ‘wrinkle’). That is, for interpretations under overt cause contexts. These ratings arguably reflect the clash between the information provided by the context and 12 Four separate one-sample t-tests were performed to the four different the information provided by the interpretation and, concretely, the combinations of Context × Interpretation. The category of data pertaining to clash between the explicit specification of an external initiator the non-overt cause context × no cause interpretation condition shows a bringing about the described change of state [the earthquake in our bimodal distribution, with a set of responses close to 0 and another set close example (16)] and the inferred absence of such an initiator for the to 100, which cannot be explained by means of the variable Voice. To explain same change of state. some of this variation, several conditional classification trees were run. Initially, On the other hand, our participants provided high saliency it was observed that the data was better classified into multiple groups by ratings in those cases where the information contributed by the considering the specific verb used; in this regard, lerono ‘sully’, peristrefome/ interpretation and the context matched: cause interpretations under yirizo ‘rotate’, mionome/lighostevo ‘diminish’, madhao ‘pluck’, tsalakono overt cause contexts and no cause interpretations under non-overt ‘crumple’, and veltionome/kaliterevo ‘improve’ were grouped for being classified cause contexts. Crucially, the preference for the no cause as having a no cause interpretation, in contrast with the rest of the verbs. If interpretation in non-overt cause contexts was significantly lower we disregard the effect of the specific verb, two different demographic variables than the preference for the cause interpretation in overt cause allow to explain part of the variation as well. Specifically, subjects with contexts. This result cannot be explained under the assumption that linguistics-related studies and subjects with a postgraduate or a doctoral degree the speakers’ saliency ratings were exclusively based on the tended to accept more a no cause interpretation for the anticausatives that interaction between context and interpretation. Instead, it can were used under a non-overt cause context (compared with those participants with a tertiary level of education not related to linguistics). 13 See also Fotiadou (2022), for indirect evidence in this direction. Frontiers in Psychology 11 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 be accounted for if a causative semantics for both active and information), especially whenever the provided context and non-active Greek anticausatives, independently argued for in the interpretation match. literature (Alexiadou et al., 2015), is adopted.14 The question raised next is how these generalizations relate to the Under the view that the saliency ratings represented in Figure 3 theory of Greek anticausatives. In this sense, we next examine what are based on the interaction between (i) the contextual setting, (ii) the they tell us about the hypothesis made in the literature (Alexiadou given interpretation, and (iii) the causative semantics of anticausatives, et al., 2015) according to which the VoiceP projected in anticausatives the obtained pattern can be straightforwardly derived: The overt cause with non-active voice morphology is expletive. context – cause interpretation condition is the optimal condition due Our empirical generalizations (i) and (ii) suggest that the to the fact that the context, the interpretation, and the semantics of the so-called Class C is not an appropriate contrast field on which to anticausative included in the test sentence coincide in favoring the pursue some specific meaning ascription to the voice of Greek existence of a specific cause bringing about the described change. anticausatives. Researchers attempting to do that have compared the Therefore, in this condition we get the highest ratings. Moving to the actively and the non-actively marked anticausative variant of the non-overt cause context – no cause interpretation condition, this can same verb in different syntactic or interpretative contexts. However, be characterized as suboptimal; the context and the interpretation we got evidence that speakers most usually prefer clearly one of the coincide in favoring the absence of a specific cause, but they both go two variants. We further found that, even if they accepted both, against the causative semantics of the anticausative in the test sentence. context cannot affect or determine the choice between one variant or It is this suboptimality that explains the ratings obtained. Lastly, the the other. overt cause context – no cause interpretation and the non-overt cause It seems, then, that Greek grammar gives speakers two options in context – cause interpretation conditions trigger extremely low forming an anticausative: either to mark the verb with active voice saliency ratings due to the fact that the information clash between the morphology or to mark it with non-active voice morphology. Whether context and the interpretation immediately gives rise to an an alternating verb falls in the scope of the former or the latter incongruent output.15 grammatical rule appears to be something that needs to be learnt by Summing up the discussion so far, our experimental findings can those acquiring/learning the language (see also Alexiadou et al., be compressed into the three empirical generalizations that follow: (i) 2015).16 The verb-specific and, therefore, learnt status of the active and Class C of Greek anticausatives is not productive; the majority of non-active voice marking of Greek anticausative verbs suggests that alternating verbs falls either into Class A or Class B. (ii) The speakers’ actively and non-actively marked anticausatives are semantically preference for active or non-active voice marking on Greek equivalent in terms of event structure. Consequently, our findings anticausatives does not depend on the presence of some explicitly appear to be consistent with the hypothesis that the VoiceP of stated external initiator in the contextual setting. (iii) Native Greek non-actively marked anticausatives is interpreted as introducing an speakers prefer to semantically compute/accommodate a cause when identity function over predicates of events and, in this sense, can interpreting anticausative events (irrespective of the voice morphology be dubbed to be expletive (Schäfer, 2008; Alexiadou et al., 2015; on the verb and, partly, irrespective of the available contextual Oikonomou and Alexiadou, 2022). We should note that the experimental results related to our control 14 See Koontz-Garboden (2009) for the same claim regarding languages items can also be viewed as evidence pointing to the expletiveness other than Greek. direction. We remind the reader that our controls, in both Note that Alexiadou et al. (2015) argue in favor of a causative semantics for experiments, involved 10 pairs of synonymous anticausative verbs, anticausatives based on the fact that the latter are compatible with modification each pair consisting of one Class A verb, that was marked with by Prepositional Phrases carrying causer information: non-active voice morphology, and one Class B verb, that was morphologically marked as active. When comparing the ratings given to the members of each pair (see the Results section 2.3.1.1), this (i) To spiti gremistike apo to sismo. synonymy was experimentally supported for 7 out of the 10 pairs the house crumbled.nact from the earthquake tested. The very existence of synonymy across the morphologically ‘The house crumbled from the earthquake.’ distinct anticausative Classes A and B points toward the view that anticausative Voice does not affect the truth-conditional meaning of 15 A reviewer suggests that the interpretation results of Experiment 2 are the sentences in which the verb appears and, thus, the expletive status compatible with an alternative view, according to which the causative meaning of anticausative Voice are strengthened by our findings.17 component of Greek anticausatives is an implicature. However, if it were indeed In a recent study on expletiveness across several functional an implicature, the causative component would be easily canceled by both categories, Tsiakmakis and Espinal (2022) reach the following the context and the offered interpretation in the non-overt cause context – no cause interpretation condition, leading to saliency ratings as high as in the 16 For the very few verbs that alternate freely between non-active and active overt-cause context – cause interpretation condition. This is not what we find. voice morphology, children acquiring Greek will also have to learn the double The implicature alternative cannot account for the significant difference morphology as part of the respective lexical entries. We note that, to our between the ratings in the optimal condition and the ratings in the knowledge, there is no acquisition study focusing exclusively on the suboptimal condition. anticausative readings of Class C verbs. The reader is referred to Fotiadou and As regards specifically the suboptimal condition, another reviewer suggests Tsimpli (2010) for an experimental study on the acquisition of transitivity that the obtained results may also be attributable to the participants’ uncertainty, alternations in Greek. or the clash between a cause accommodated via the context and the absence 17 See Schäfer and Vivanco (2016) for similar argumentation in relation to of a specific cause imposed by the provided interpretation. languages other than Greek. Frontiers in Psychology 12 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 generalization: semantically expletive categories, apart from Following this approach, Voice in Greek anticausatives introducing an identity function at the truth-conditional level, need ultimately introduces no semantic argument; it ends up being to stand in a local syntactic relation to an element with respect to interpreted as an identity function, as a result of an emerging which they initially encode some redundant meaning. If Voice in semantic redundancy. In this way, we correctly predict that there Greek non-actively marked anticausatives is interpreted as expletive, is no traceable truth-conditional asymmetry whatsoever between it is expected to satisfy this precondition. We claim that this prediction an actively and a non-actively marked anticausative, that is is borne out, and that this is exactly where our empirical generalization between a Voice-less anticausative and an anticausative projecting (iii), namely that speakers prefer to interpret a cause when faced with a non-active VoiceP. The meaning of Greek anticausative Voice is anticausative structures, becomes relevant.18 represented formally in (21) (Schäfer, 2008, 2017; Wood, 2014, Let us adopt for our discussion the main aspects of the syntactic 2015; Schäfer and Vivanco, 2016), where 〈s,t〉 stands for the account by Alexiadou et al. (2015). In the authors’ view, anticausatives situation (or event) in which the proposition is true. The partial correspond primarily to causative little vPs, as represented in (19b); semantics for (19) and (20) is given in (22) and (23), those that are marked with non-active voice morphology further respectively. project a non-active VoiceP lacking a specifier above vP, as shown in (20b). (21) Anticausative Voice: ⟦-thik-CAUSE⟧20 = λP〈s,t〉.P〈s,t〉 (22) ⟦[v cause [VP I fusta zaron]]⟧ = λeCAUS [wrinkle(e) & theme(e) = the skirt] (19) a. I fusta zarose. (23) ⟦[Voice -thik-CAUSE [v cause [VP I thalasa zesten]]]⟧ the skirt wrinkled.act. = (λP〈s,t〉.P〈s,t〉) (λeCAUS [heat(e) & theme(e) = the sea]) ‘The skirt wrinkled.’ = λeCAUS [heat(e) & theme(e) = the sea] b. [vP [v v √zaron] [DP i fusta]] (20) a. I thalasa zestathike. As stated throughout the paper, the idea that non-actively marked the sea heated.nact anticausatives project an expletive Voice is not new; it is present in ‘The sea heated.’ Schäfer (2008) and Wood (2014) and is put forth explicitly in relation b. [VoiceP-NACT [Voice-NACT thik-] [vP [v v √zesten] [DP i thalassa]]] to Greek in Alexiadou et al. (2015). The finer insight that this expletiveness is dependent on the syntactic context is not new either; As regards the semantics attributed to such structures, it is see Oikonomou and Alexiadou (2022). What is novel in the present postulated that, up to the vP level in both actively and non-actively proposal is that the expletiveness of Voice is legitimized by its local marked anticausatives, the interpretation is one of a causative relation to the causative v, not by unspecified information attributed event, because v has a causative flavor. Crucially, this is in to the verbal root (cf. Alexiadou et al., 2015; Oikonomou and accordance with our experimental finding that, regardless of voice Alexiadou, 2022).21 This departure from previous approaches is not morphology, speakers interpret/accommodate a cause for trivial and is motivated by the inclusion of Greek anticausative Voice anticausatives (see Figure 3). The question that is at the heart of the present theoretical analysis is in what way the projection of the the agent elsewhere. We speculate that this is what happens in Greek deponent non-active VoiceP in (20b) contributes compositionally to the verbs (verbs with non-active voice morphology but transitive syntax; see meaning of the sentence. Zombolou and Alexiadou, 2014 for a finer categorization), which have been We postulate that the non-active Voice that emerges as non-active argued to introduce agents at the specifier of a functional projection below morphological marking in Greek anticausatives encodes a cause- VoiceP (Grestenberger, 2018). We provide an example and its partial derivation, related meaning, i.e., that it is initially non-expletive. On this basic in the spirit of such an analysis: tenet, we coincide with Oikonomou (2014). This non-active Voice head merges with a vP that also involves a cause formal feature. The (i) a. I Maria iperaspistike ton mikro tis adherfo. resulting redundancy of cause-related information in the structure – the Maria defended.nact the small her brother that is, the multiple representation of some cause semantic feature within a syntactically local domain (see Tsiakmakis and Espinal, ‘Maria defended her younger brother.’ 2022)– leads to the interpretation of this Voice as an identity function b. [VoiceP-NACT [Voice-NACT tik-] [XP [DP i Maria] [Xˈ [X] [[vP [v v √iperaspiz] [DP over the predicate of events introduced by the causative v.19 In other ton mikro tis adherfo]]]]]] words, the expletive interpretation of Voice in anticausatives is a by-product of it merging with a causative v. 20 We have chosen -thik- as the morphological exponent of anticausative Voice in our definition because it is the one that appears in our examples. For discussion on the full paradigm of non-active voice morphology, the reader 18 Tsiakmakis and Espinal (2022) take the allegedly expletive voice marker is referred to Holton et al. (1997), Ralli (2005), and Manzini et al. (2015). se in Romance anticausatives to be a syntactically expletive category, not a 21 It should be clarified at this point that, of all the possible interpretations semantically expletive one. However, these authors do not take into account of Voice in Greek, our analysis aims to account only for the interpretation of Greek expletive voice, which arguably has different characteristics. the causative Voice projected in non-actively marked anticausatives as an 19 Notice that an explanation of the expletiveness of Voice in this spirit creates expletive, which we consider as the result of a cause-related information the expectation that a Voice head carrying agent information can also end up redundancy. The interpretation of Greek Voice in the rest of its instances is being interpreted as an identity function, if merged in a structure introducing beyond the present experimental study. Frontiers in Psychology 13 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 in the typology of semantic expletives, which also include expletive cause-related information merges with a causative little v. The negation and expletive plural number on mass nouns (see Tsiakmakis resulting redundancy of cause information allows the Voice of and Espinal, 2022).22 Greek anticausatives to behave as a typical semantically expletive category (Tsiakmakis and Espinal, 2022): it stands in a local relationship with an element that also encodes causative meaning 4. Conclusion and is, ultimately, interpreted merely as an identity function over causative events. In this paper we investigated voice in Greek anticausatives. We ran In light of the above, the present study can be regarded as a solid an experimental study that explored Greek native speakers’ preferences source of both empirical and theoretical evidence in support of the with respect to two different judgment tasks that aimed to shed light expletiveness hypothesis regarding the voice of Greek anticausatives. on the distribution and interpretation of anticausative voice in Greek, While extending our view to other languages requires further a topic clouded by a heated debate. research, a paradigm has been set to pursue relevant queries The results of our experiments support the following conclusions: across languages. (i) The so-called Class C of Greek anticausative verbs is not productive; for the majority of alternating verbs, speakers have a clear preference for either the anticausative variant that bears active voice morphology Data availability statement (Class B) or the variant that displays non-active voice morphology (Class A). (ii) The speakers’ preference for either form of the The raw data supporting the conclusions of this article will anticausative verb is not significantly affected or determined by the be made available by the authors, without undue reservation. presence/absence of a linguistically phrased external initiator bringing about the described change of state. (iii) Regardless of voice morphology, speakers tend to interpret/accommodate a cause when Ethics statement faced with anticausative event descriptions, provided that the information surrounding the event description is not contradictory The studies involving human participants were reviewed and (see Graphs 1 and 4 in Figure 3). approved by the Ethics Committee on Animal and Human Our first and second empirical generalizations suggest that the Experimentation of the Universitat Autònoma de Barcelona, under emergence of active or non-active voice morphology in Greek the approved experimental protocol number CEEAH-4442. The anticausatives is verb-specific and, thus, encoded in the lexicon. patients/participants provided their written informed consent to Furthermore, these generalizations can be regarded as evidence participate in this study. consistent with the theoretical proposal that the non-active Voice that is realized as non-active morphology in Greek anticausatives is expletive, interpreted as introducing an identity function over Author contributions predicates of events (Schäfer, 2008, 2017; Wood, 2014, 2015; Alexiadou et al., 2015). Taking into account also the third ET ran the bibliographical search and designed the materials for empirical generalization, we can further refine this proposal: in the experiments. ET, JB-C and ME designed the experiments. JB-C is Greek anticausatives, a non-expletive Voice head carrying responsible for the statistical analysis of the results. ET and ME are responsible for the whole research. All authors contributed to the 22 Let us try to demonstrate briefly the parallel between the analysis proposed article and approved the submitted version. here for the expletiveness of Greek anticausative Voice and Tsiakmakis and Espinal’s (2022) account of the expletiveness of Catalan negation, with the help of the following example: Funding This research was supported by the Spanish Ministerio de Ciencia (i) Tinc por que no es mengin el pastís. e Innovación (FFI2017-82547-P, PID2020-112801GB-I00) and the I.have fear that not cl eat.subj the cake Generalitat de Catalunya (2021SGR00787). ‘I am afraid that they might eat the cake.’/‘I am afraid that they might not eat the cake.’ Acknowledgments Following the authors, in this case there is a negative semantic feature in the fear predicate tinc por (Espinal, 1992) and another negative semantic feature We thank the reviewers and the editorial board of the Frontiers in the negative marker no. When the complement clause is in subjunctive and, in Psychology. Language Sciences journal for their constructive therefore, transparent to the matrix fear predicate (Picallo, 1985), the two comments. We would also like to thank Florian Schäfer for categories carrying a negative feature are in a local syntactic relationship and stimulating discussion on the topic. Our last but not least thanks semantic redundancy of negative meaning arises. This semantic redundancy go to our colleagues at the Center for Theoretical opens the possibility that the negative marker no is interpreted merely as an Linguistics of the Universitat Autònoma de Barcelona for fostering identity function, giving rise to the interpretation ‘I am afraid they will eat the the environment that allowed us to develop the ideas cake.’ (see Tsiakmakis and Espinal, 2022, for details). here discussed. Frontiers in Psychology 14 frontiersin.org Tsiakmakis et al. 10.3389/fpsyg.2023.1068058 Conflict of interest or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its The authors declare that the research was conducted in the absence manufacturer, is not guaranteed or endorsed by the publisher. of any commercial or financial relationships that could be construed as a potential conflict of interest. Supplementary material Publisher’s note The Supplementary material for this article can be found online All claims expressed in this article are solely those of the authors at: https://www.frontiersin.org/articles/10.3389/fpsyg.2023.1068058/ and do not necessarily represent those of their affiliated organizations, full#supplementary-material References Alexiadou, A., and Anagnostopoulou, E. (1999). “Tests for unaccusativity in a Levin, B., and Rappaport Hovav, M. (1995). Unaccusativity: at the syntax–lexical language without tests for unaccusativity” in Proceedings of the 3rd international semantics interface (linguistic inquiry monograph 26). Cambridge: MIT Press. conference on Greek linguistics. ed. Amalia Mozer (Athens: Ellinika Grammata), 23–31. Manzini, M. R., Roussou, A., and Savoia, L. M. (2015). Middle-passive voice in Alexiadou, A., and Anagnostopoulou, E. (2004). “Voice morphology in the causative– Albanian and Greek. J. Ling. 52, 111–150. doi: 10.1017/S0022226715000080 inchoative alternation: evidence for a non-unified structural analysis of unaccusatives” Oikonomou, Despina. (2014). “In the middle of passive: middle voice in modern in The unaccusativity puzzle. eds. A. Alexiadou, E. Anagnostopoulou and M. Everaert Greek vs. passive voice in English,” in Ms. MIT Generals Paper. (MIT). (Oxford: Oxford University Press), 114–136. Alexiadou, A., Anagnostopoulou, E., and Schäfer, F. (2015). External arguments in Oikonomou, D., and Alexiadou, A. (2022). Voice syncretism crosslinguistically: the transitivity alternations: A layering approach. Oxford: Oxford University Press. view from minimalism. Philosophies 7:19. doi: 10.3390/philosophies7010019 Embick, D. (1997). “Voice systems and the syntax/morphology interface” in Papers Picallo, C. (1985). Opaque domains. dissertation. New York: City University of from the UPenn/MIT roundtable on argument structure and aspect 32. ed. H. Harley New York. (Department of Linguistics, Massachusetts Institute of Technology), 41–72. Ralli, A. (2005). Morphology [Μορφολογία]. Athens: Patakis. Espinal, M. T. (1992). Expletive negation and logical absorption. Ling. Rev. 9, 338–358. Rappaport Hovav, M. (2014). Lexical content and context: the causative alternation in doi: 10.1515/tlir.1992.9.4.333 English revisited. Lingua 141, 8–29. doi: 10.1016/j.lingua.2013.09.006 Fotiadou, G. (2022). Processing of transitivity alternations and frequency-based Rappaport Hovav, M. (2020). “Deconstructing internal causation” in Perspectives on accounts in Greek adult language. Languages 7:216. doi: 10.3390/languages7030216 causation. eds. E. A. Bar-Asher Siegal and N. Boneh (Cham: Springer), 219–255. Fotiadou, G., and Tsimpli, I. M. (2010). The acquisition of transitivity alternations in Schäfer, F. (2008). The syntax of (anti-)causatives. External arguments in change-of-state Greek: does frequency count? Lingua 120, 2605–2626. doi: 10.1016/j.lingua.2010.06.011 contexts. Amsterdam: John Benjamins. Grestenberger, L. (2018). Deponency in finite and nonfinite contexts. Language 94, Schäfer, F. (2017). “Romance and Greek medio-passives and the typology of voice” in 487–526. doi: 10.1353/lan.2018.0034 The verbal domain. eds. R. D'Alessandro, I. Franco and Á. J. Gallego (Oxford: Oxford Haspelmath, M. (1993). “More on the typology of inchoative/causative verb University Press), 129–152. alternations” in Causatives and transitivity. eds. B. Comrie and M. Polinsky (Amsterdam: Schäfer, F., and Vivanco, M. (2016). Anticausatives are weak scalar expressions, not John Benjamins), 87–120. reflexive expressions. Glossa: J. Gen. Ling. 1:18. doi: 10.5334/gjgl.36 Haspelmath, M. (2016). Universals of causative and anticausative verb formation and Theophanopoulou-Kontou, D. (2000). “O/−me alternations in MG patient oriented the spontaneity scale. Lingua Posnaniensis 58, 33–63. doi: 10.1515/linpo-2016-0009 constructions: Anticausatives and passives [Δομές τηε ΝΕ με προοπτική του δέκτη και Holton, D., Mackridge, P., and Philippaki-Warburton, I. (1997). Greek: A η κατανομή -ω/−ομαι: Τα αντιμεταβιβαστικά και παθητικά]” in Studies in Greek comprehensive grammar of the modern language. London: Routledge. linguistics. Vol. 20. eds. Theodosia Pavlidou and Christos Tzitzilis (Thessaloniki: Insititouto Neoellinikon Spoudon), 146–157. Horvath, J., and Siloni, T. (2011). Anticausatives: against reflexivization. Lingua 121, 2176–2186. doi: 10.1016/j.lingua.2011.09.006 Triantafyllidis, M. (1941). Modern Greek Grammar [Νεοελληνική Γραμματική]. Athens: Organismos Ekdosis Didaktikon Vivlion. Kallulli, Dalina (2006). “A unified analysis of passives, anticausatives and reflexives,” in Empirical issues in formal syntax and semantics 6. Eds. O. Bonami and P. Cabredo- Tsiakmakis, E., and Espinal, M. T. (2022). Expletiveness in grammar and beyond. Hofherr (Colloque de Syntaxe et Sémantique à Paris), 201–225. Glossa J. Gen. Ling. 7, 1–36. doi: 10.16995/glossa.5807 Koontz-Garboden, A. (2009). Anticausativization. Nat. Lang. Ling. Theory 27, 77–138. Tsimpli, I. M. (2006). “The acquisition of voice and transitivity alternations in Greek doi: 10.1007/s11049-008-9058-9 as native and second language. In paths of development in L1 and L2 acquisition” in Honor of Bonnie D. Scwartz. eds. S. Unsworth, T. Parodi, A. Sorace and M. Young- Kratzer, A. (1996). “Severing the external argument from its verb” in Phrase structure Scholten (New York/Philadelphia: John Benjamins), 15–55. and the lexicon. eds. J. Rooryck and L. Zaring (Dordrecht: Kluwer), 109–137. Winter, Y. (2016). Elements of formal semantics. Edinburgh: Edinburgh Labelle, M. (2008). The French reflexive and reciprocal se. Nat. Lang. Ling. Theory 26, University Press. 833–876. doi: 10.1007/s11049-008-9053-1 Wood, J. (2014). Reflexive -st verbs in Icelandic. Nat. Lang. Ling. Theory 32, 1387–1425. Lavidas, N., Fotiadou, G., and Tsimpli, I. M. (2012). “Active vs. nonactive voice in the doi: 10.1007/s11049-014-9243-y Greek diachrony. Real or apparent optionality in the use of voice morphology?” in Selected papers of the 10th international conference of Greek Linguisctics. eds. Z. Wood, J. (2015). Icelandic morphosyntax and argument structure. Dordrecht: Gavriilidou, A. Efthymiou, E. Thomadaki and P. Kambakis-Vougiouklis (Komotini: Springer. Democritus University of Thrace), 390–400. Zombolou, K., and Alexiadou, A. (2014). “The canonical function of deponent verbs” Levin, B. (1993). English verb classes and alternations: a preliminary investigation. in Morphology and meaning. eds. F. Rainer, F. Gardani, H. C. Luschützky and W. U. Chicago: University of Chicago Press. Dressler (Amsterdam: John Benjamins), 331–344. Frontiers in Psychology 15 frontiersin.org