The volume is dedicated to the investigation of non-canonical passives, namely constructions with a passive ‘flavor’, displaying some, but not all, characteristics of the canonical passive. It contains contributions dealing with these structures in a wide variety of languages, with most authors focusing on semantic and syntactic aspects of the constructions.
The introduction, ‘Non-canonical passives’, by the editors Artemis Alexiadou and Florian Schäfer, presents the general issues which the volume aims to address. It begins with establishing the properties of canonical passives, which serve to identify those passives which are non-canonical: (i) argument suppression, (ii) Case absorption, and (iii) argument enhancement. The authors then include a short discussion of some of the better known cases of non-canonical passives across languages, notably the get-passive in English. The introduction ends with a chapter-by-chapter overview.
The contributions start with Chapter 1, by Andrew McIntire, ‘Adjectival passives and adjectival participles in English’. The chapter discusses a wealth of phenomena related to English adjectival participles, presenting some novel claims and providing new arguments for claims already present in the literature. The author follows recent suggestions that in different languages, adjectival passives can be divided into several subtypes, depending mainly on the implication of an event and/or an agent (Anagnostopoulou Reference Anagnostopoulou, Alexiadou, Rathert and Stechow2003, Embick Reference Embick2004, Meltzer-Asscher Reference Meltzer-Asscher2011). The two main original claims in the chapter are: (i) there exists a class of ‘situation-in-progress’ adjectival participles (e.g. ‘The car is badly driven’), not recognized before; and (ii) adjectival participles are productively formed from unaccusative verbs with a relatively stable result state (e.g. ‘recently departed people’). Additionally, new considerations are presented in favor of assuming that the theme argument is not generated as a complement of an adjectival participle, and that some adjectival passives may include implicit agents. The chapter ends with a proposal on how different types of adjectival participles are derived, specifying the semantics of the participle morpheme (Prt) in transitive resultative (‘The car is scratched’), ‘situation-in-progress’, and unaccusative resultative participles. The difference between the first two participle types is accounted for by assuming that Prt of the former, but not the latter, includes a causal relation yielding a result state. Unlike the author, I tend to believe that a uniform analysis, using the same participle morpheme, could be given to transitive resultative and ‘situation-in-progress’ participles, with the difference between the two arising only from the aspectual difference in their verbal base (e.g. ‘scratch’ has a result state, whereas ‘drive’ does not). Nevertheless, the chapter is extremely interesting and thought-stimulating, both in the data presented and in the analyses proposed.
Chapter 2, ‘The get-passive at the intersection of get and the passive’, by Anja Wanner, revisits some of the claims about semantic differences between get-passives and be-passives when checked against data from a corpus of American English. The author shows that neither the claim that get-passives have an adversative meaning, nor the claim that they have a ‘secondary agent’ meaning (with the surface subject interpreted as agentive) is corroborated by the corpus data. Moreover, get-passives may have an implicit argument, on a par with be-passives. Although the corpus study would benefit from contrasting the data regarding get-passives with the relevant data on be-passives (e.g. what percentage of be-passives are adversative?), the chapter suggests a refreshing – and evidently important – viewpoint, broadening the empirical basis of the discussion of passives.
Chapter 3, by Alexandra N. Lenz, ‘Three “competing” auxiliaries of a non-canonical passive: On the German GET passive and its auxiliaries’, is somewhat of an oddball in the volume, presenting a typological study of the so-called ‘recipient passive’ in German. The paper presents some interesting facts with regard to the use of three auxiliaries – bekommen ‘to get/receive’, kriegen ‘to get/receive’ and erhalten ‘to receive’ – in this construction, in different regions and registers. At times, though, one gets the feeling that the paper is not focused enough, as it encompasses data from different types of corpora from different periods, as well as experimentally elicited utterances and judgments. Additionally, the presentation could be clearer at certain points (e.g. it is unclear what the y-axis in Figure 4 represents).
Perhaps the most ambitious contribution to the volume is Chapter 4, ‘Variations in non-canonical passives’, by C.-T. James Huang. The chapter argues that many non-canonical passives can be analyzed as either raising or control constructions, and that this is due to the semantics of the passive auxiliary, which can cover a varying portion of the causative–unaccusative continuum, with the more causative readings giving rise to control structures, and the more unaccusative readings giving rise to raising structures. The analysis is exemplified with the Mandarin ‘give-passive’, and potentially can account for many phenomena cross-linguistically. Perhaps the weak point of the chapter, however, is that the decision to implement the polysemy of the auxiliaries in terms of lexical decomposition or cartography is left unexplained and unjustified.
Chapter 5, ‘How much bekommen is there in the German bekommen passive?’, by Markus Bader & Jana Häussler, returns to the German non-canonical passive. The authors ask whether the thematic properties of the lexical verb bekommen ‘get’, specifically the fact that it assigns the recipient role, are relevant in its function as a passive auxiliary. On the basis of a grammaticality judgment survey and a corpus study covering 12 classes of ditransitive verbs, the authors conclude that bekommen has undergone grammaticalization, allowing dative arguments which are not recipients in the passive. The experiment and the corpus study are both well designed, and the authors make a convincing argument, although submitting the data to a statistical analysis would undoubtedly make it even stronger.
Another German passive construction, this one using the auxiliary ‘have’, is addressed in Chapter 6, ‘Haben-statives in German’, by Martin Businger. The author makes two claims: (i) that the participle in this construction is adjectival, and (ii) that the construction includes a small clause complement of haben ‘have’. While the author's arguments are very clear, the chapter would benefit from reference to parallel structures in other languages, notably English (Harley Reference Harley, Guéron and Zribi-Hertz1998, Kim Reference Kim2012). Additionally, while the chapter convincingly shows that the participles in haben-statives must have an adjectival reading, it does not rule out the possibility that they may, in addition, have a verbal reading. Since small clauses can have verbs as well as adjectives as predicate, it is at least possible that the participle is categorically ambiguous.
The haben-passive is also the focus of ‘Another passive that isn't one’, by Helga Gese (Chapter 7). The chapter accepts Businger's syntactic analysis of the structure, and investigates the semantics of the participle in it. Its main claim is that this participle is interpreted like the adjectival passive in sein-passives (‘be’-passives), in that (i) its meaning is underspecified by the lexical information, specifically the result state, associated with the corresponding event, and (ii) its verbal base receives a generic interpretation. The conclusion that the participle has the same semantics in the two structures is in fact the null hypothesis given that both structures involve the adjectival passive. With regard to the diagnostics showing that the verbal base's agent is inert, recent studies have shown that more accurate application of these can in fact lead to different results (McIntire in this volume, Meltzer-Asscher Reference Meltzer-Asscher2011), and reference to this literature would contribute to the chapter.
The remainder of the volume is dedicated to non-canonical passives in less-studied languages, beginning with Chapter 8, ‘Passives and near-passives in Balto-Slavic’, by James E. Lavine. The chapter aims to capture the difference in the availability of accusative Case in passive-like structures in Ukrainian, Russian, Polish and Lithuanian. The general idea advocated here is that accusative may be available even in the absence of an external argument (i.e. in the passive) only when the language has separate VOICE and CAUSE heads. Though this suggestion can account for the observed facts, I find that the chapter lacks independent motivation for assuming a bundled VOICE-CAUSE head in some languages, and two heads in others. It is not clear what ramifications this difference has, besides the availability of accusative in the passive, and hence it is not clear how this hypothesis can be reinforced or falsified.
Chapter 9, ‘How do things get done’, by Fredrik Heinat & Satu Manninen, takes up the passive in Finnish. The authors show that Finnish agreeing past participles, apparently traditionally viewed on a par with simple adjectives, in fact exhibit what the authors call ‘eventivity/agentivity’ passive properties. They suggest that these form a type of syntactically formed adjectival passives. The authors distinguish between these agreeing passives and a class of participles which, according to them, do not show passive properties (examples (7) in the chapter), as they do not pass tests detecting the existence of an implicit agent. However, they fail to notice that the latter participles' verbal base (e.g. ‘admire’) does not have an agent to begin with. I find that the presentation and the conclusions could have been clearer if different aspects of these passives were examined separately, namely, their status as adjectives, their entailment of an event, and their agent entailment.
The passive se faire construction in French is the topic of ‘Anticausativizing a causative verb’, by Marie Labelle (Chapter 10). The aim of the chapter is to develop an analysis of this structure that does not involve faire introducing a CAUSE operator, which is immediately removed by SE. This is achieved by assuming that faire is a change of state verb, taking as a complement an infinitive clause containing a moved null operator. The VP headed by faire is a sister of a [–Active] Voice head occupied by SE. The subject of the passive is merged in spec,VoiceP, and interpreted via predication as the argument abstracted over in the embedded clause. One unsettling point is that under this analysis, the subject of any predicate derived with SE is assumed to be merged outside of the VP. It is not clear how this can account for the fact that subjects of unaccusatives behave like internal arguments.
Chapter 11, ‘On the syntax-semantics of passives in Persian’, by Fatemeh Nemati, aims to present a typology of passive constructions in this language, maintaining a unified analysis for the auxiliary/copula šodœn. It discusses structures with the auxiliary including past participles, nominals or simple adjectives. With regard to the participial structures, the author's claim seems to be that they are ambiguous between a verbal and an adjectival reading, though this is not completely clear. Beyond the analysis proposed, the chapter's main merit is that it presents a large array of data from Persian, which has hardly been studied in the generative tradition, and can stimulate future research of the passive in this language.
In Chapter 12, ‘Two indirect passive constructions in Japanese’, Masanori Deguchi examines affective passives formed with (r)are and morau. The author's main claim, which I find convincing, is that the ‘benefactive’ meaning in the latter structure arises since it involves causation of the event by the surface subject, the affected argument, who will tend to only cause events that will benefit him/her. In contrast, in the (r)are passive there is no causation meaning component, and the affected argument is simply subjected to whatever event occurs, which is often adversative. The chapter would be more complete had it included a reference to Pylkkänen's (Reference Pylkkänen2008) influential analysis of adversative constructions, discussing whether the current analysis is compatible or at odds with it.
Non-canonical passives in Swedish are the topic of Chapter 13, ‘Få and its passive complement’, by Eva Klingvall. Similarly to the structures described above in Mandarin and Japanese, as well as to the English get-passive, Swedish forms a passive with the verb få ‘get’, which is ambiguous in that its subject can be interpreted either as a cause or as a recipient. The author presents several environments which differentiate between the two readings, and proceeds to suggest that these differences can be captured by assuming that recipient få is a realization of incorporation of the functional heads BECOME and HAVE, whereas its causative counterpart is the spell-out of a little-v incorporated into V. The details of the structures are somewhat schematic. More importantly, I do not see how the differences between the two readings motivate a different structure, rather than lexical polysemy.
Chapter 14, ‘The Danish reportive passive as a non-canonical passive’, by Bjarne Ørsnes, discusses structures involving passive raising predicates, e.g. ‘Peter is said to be out of town’. The author shows that these predicates are indeed passive, as they entail an implicit argument. He proposes an analysis in the framework of Head-driven Phrase Structure Grammar, which contains a lexical operation turning an active verb selecting a finite clause into a passive verb selecting a VP. One question which immediately comes to mind is whether the passive predicates are not better analyzed as adjectival passives, which may explain their ‘grammaticalized’ flavor, an option not considered by the author. Another issue which is left open is why the finite clause is converted to a non-finite clause by the lexical operation. In Government and Binding, raising out of a finite clause is banned based on general considerations (‘improper movement’). Under the current analysis, this seems to be a random feature of the structure, and one could imagine a similar operation which does not change the finiteness of the clause, contrary to fact.
Finally, Dalina Kallulli, in ‘(Non-)canonical passives and reflexives’ (Chapter 15), investigates deponent verbs in Greek and Albanian and ‘inherent reflexive’ verbs in Germanic and Romance, suggesting a parallel analysis for the two verb classes. Both deponents (i.e. dergjem ‘I linger’ in Albanian) and ‘inherent reflexives’ (e.g. si arabia ‘gets angry’ in Italian) are verbs with non-active morphology that lack corresponding active forms. Additionally, both verb types are primarily denominal or deadjectival. The author proposes that the function of ‘non-active’ morphology in both cases is actually that of a verbalizer, specifically, one carrying the features [–external argument] and [+actor-initiated], also shared by passives. The analysis seems to capture much cross-linguistic synchronic as well as diachronic data. A minor comment is that I find the term ‘inherent reflexive’ unfortunate. There is nothing reflexive about e.g. ‘get angry’, neither in meaning, nor in syntactic behavior, as true reflexives do not exhibit unaccusative behavior (Reinhart & Siloni Reference Reinhart, Siloni, Alexiadou, Anagnostopoulou and Everaert2004).
One of the biggest merits of the volume as a whole is that it covers a large number of languages and language families. This provides the reader with multiple structures exemplifying the differences in non-canonical passives across languages, and, perhaps more interestingly, the many traits that they in fact share. One common property that seems to stand out is that these structures tend to involve a grammaticalized light verb such as ‘get’, which can accommodate agentive/causative as well as recipient/beneficiary readings for the subject.
To make the volume more coherent, however, I believe that a more accurate definition of what makes a passive ‘non-canonical’ should have been included in the Introduction. As mentioned above, the editors propose that canonical passives are defined by argument suppression, Case absorption, and ‘argument enhancement’. The latter is an uncommon term which the authors do not define. Possibly, it refers to the externalization of the original internal argument (i.e. ‘argument promotion’). Assuming this, and given that, for example, the get-passive involves argument suppression, Case absorption, and argument enhancement, it is not at all clear why it is considered to be non-canonical. This lack of a definition for passive can make it difficult at times to understand the claims of the authors. For example, in Chapter 9, Heinat & Manninen discuss adjectival participles, saying that a certain analysis ignores the fact that ‘many of these sentences receive “passive” readings’ (p. 215), but it remains unspecified what is meant by a ‘passive’ reading. Similarly, in Chapter 14, Ørsnes dedicates several subsections to showing that reportive passives are in fact passives, but it is never explicitly stated what being a passive entails. Given that the passive is in fact the topic of the volume, it seems that an accurate definition of this term would have been useful.
Also, perhaps because the volume is a selection of papers from a workshop, the level of sophistication of the different contributions is not homogeneous, with some chapters defending more ambitious claims or presenting stronger argumentation than others.
Despite these flaws, this is an important volume, particularly due to the breadth of the phenomena discussed and analyses put forward. It raises the question of how the similarities and differences between structures in the different languages can be captured under a unified, perhaps parameterized account. Hopefully, it will incite further research in this direction.
