Topic of the Volume: Diatheses and Epitheses
Although verbs are usually associated with a specific argument structure, it is well known that the realization of their arguments can be affected in different ways. The German verb schenken ‘to give as a present’ in (1), for instance, has three arguments which can be termed ‘agent’ (der Junge ‘the boy’), ‘theme’ (einen Ball ‘a ball’), and ‘recipient’ (dem Mann ‘to the man’).
In the passive, the agent is realized as an adjunct within a von-PP. The theme is realized as the nominative subject argument and the realization of the recipient argument is unaffected. As German possesses different passive constructions, the one in (2) is termed ‘event passive’ (German Vorgangspassiv).
Passive might be one of the best-known instances of what Cysouw calls “diatheses.” A diathesis is roughly defined as “a clause alternation in which at least one of the lexical roles has a different form between the clausal alternants” (p. 7). In addition, the clausal alternants have to represent monoclausal constructions which contain the same verbal lexeme as their main predicate. This is an essential – but also not uncritical – aspect of Cysouw’s conception of a diathesis. I take up the issue of the identity of the main predicate at the end of the review. The Vorgangspassiv in (2) is a diathesis since it is a monoclausal construction which has the same main predicate as the active sentence in (1) but two lexical roles – agent and theme – are mapped differently.
In his monograph on German diatheses, Cysouw identifies 250 different diatheses most of which have received considerably less attention in the linguistics literature than the German Vorgangspassiv. The presentation of the different diatheses is organized into nineteen major types, for instance, “Passive diatheses,” “Inversive diatheses,” “Applicative diatheses,” and “Antipassive diatheses.” The diatheses of one type share a specific alternation pattern. A passive diathesis, for instance, “removes the role marked as subject and promotes an object to be the new subject” (p. 84). Cysouw lists seven passive diatheses in addition to the already mentioned Vorgangspassiv. Some of these passive diatheses have hardly been mentioned in previous literature, for instance, the Fortsetzungspassiv (‘continuation passive’) exemplified in (3).
This example shows that Cysouw goes beyond the traditional notion of “diatheses” and identifies rarely mentioned or even new types. The discussion of each diathesis is complemented by an identification of its formal characteristics. Given the huge number of diatheses, a detailed discussion of the individual diatheses goes beyond the limits of the volume. However, Cysouw goes beyond a sheer listing of formal properties and also mentions semantic properties and gives a hint on functional restrictions. The level of detailedness varies from diathesis to diathesis as some have received a proper treatment in the previous literature, but some have not.
Besides diatheses, Cysouw also discusses of what he calls “epitheses.” An epithesis is a monoclausal construction without any changes in the mapping of the verb’s arguments. An example of an epithesis is the use of modal verbs such as wollen ‘want’ (4). The modal verb is the finite verb in the sentence but it neither introduces a new argument, nor does it affect the mapping of the arguments. Epitheses are mostly associated with the periphrastic expression of grammatical notions like modality, evidentiality, and aspect (e.g., perfect aspect).
At first glance, the inclusion of epitheses looks like a digression from the book’s main topic. But it is not for two reasons. First, the discussion of epitheses allows for a clearer identification of the boundaries of the notion of a diathesis. Second, diatheses and epitheses interact and we observe stacking of the two types like in (4). The sentence in (4) is an instance of the so-called Pertinenzpassiv. The main predicate is schneiden ‘cut’ but it is not the verb’s agent argument which is realized as the subject but the beneficiary of the cutting. In (4), we observe a stacking of a modal epithesis and the Pertinenzpassiv diathesis. If we compare (4) with (5) – which is an instance of the Pertinenzpassiv without a modal epithesis – we observe that the acceptability of realizing the agent argument within a von-PP decreases. Also, while the epithesis does not affect the mapping of the agent argument, it increases the acceptability of having the von-PP realized.
The Structure of the Volume
Cysouw’s encyclopedia is a 663-page long book consisting of thirteen chapters. The book starts in chapter 1 with a definition of the central concepts which are basically “diathesis” and “epithesis.” In addition, the chapter presents the relevant background on the data collection and introduces how the author proceeded with his study. Also, the chapter is short (eighteen pages), but it introduces the basic concepts in a very clear way.
Chapter 2 presents a detailed discussion of the notion of a diathesis and identifies the various dimensions in which diatheses can differ from one another. This basically concerns the type of role-remapping associated with the individual diatheses. The central dimensions are: (i) the number of arguments affected (one or more than one), (ii) the type of argument affected (e.g., subject or object), and (iii) the remapping of the affected argument(s). Taking up again the Fortsetzungspassiv from example (3), we can describe it as a remapping of two arguments which are the agent argument (subject) and the theme argument (direct object) of the verb öffnen ‘to open’. The theme argument is remapped onto the nominative subject and the agent argument is – at least in (3) – unrealized.
Based on these criteria, Cysouw comes up with a typology of diatheses which results in classifying the Fortsetzungspassiv – like the other types of passive constructions – as a “chained subject diathesis.” The notion “chaining” refers to the fact that “any diathesis involving the nominative subject typically includes two remappings, namely one from something else to nominative and a second remapping of the erstwhile nominative to something else” (p. 42). Thus, the nominative subject is demoted (i.e., to an adjunct) and some other argument expression is promoted to the subject. However, the notion of “chained subject diatheses” covers other diatheses as well, for instance, “anticausative” but also what Cysouw terms “novativ.” The latter notion is a cover term for different diatheses such as “causative” or “experientive” (illustrated in (6)). The experientive introduces a new argument – the first-person experiencer – and demotes the agent to an accusative object argument.
Chapters 3 and 4 present the summary of major diatheses or epitheses already mentioned above. The remaining chapters of the book – 5 to 13 – are concerned with specific types of alternations, that is, different ways of expressing diatheses. Cysouw distinguishes between case-marking alternations (ch. 5), prepositional alternations (ch. 6), reflexive pronoun alternations (ch. 7), preverb alternations (ch. 8), adverbial alternations (ch. 9), light-verb alternations with Partizip (ch. 10), light-verb alternations with Infinitiv (ch. 11), light-verb alternations with zu-Infinitiv (ch. 12), and light-verb alternations with Präpositionsinfinitiv (ch. 13). For reasons of space, I cannot discuss the different alternations in detail but would just like to point out that the “experientive” in (5) is considered to be an instance of a light-verb alternation with Infinitive. Cysouw treats the infinitive schenken ‘to give as a present’ as the main predicate and analyzes the finite verb sehen ‘see’ as a light verb. I come back to this issue in my overall evaluation of the volume.
Overall Evaluation
The volume presents an impressive summary of language data and hence provides a rich resource for any study on diatheses and epitheses – as well as their interaction – in German. Given the huge number of language examples, Cysouw decided to omit any interlinear glossing. Readers unfamiliar with German might miss some details but a sufficient number of examples are discussed in the text. So, I do not think that the omission of the glossing makes the book a less valuable resource for readers without any knowledge of German.
Although I think that Cysouw’s analysis is largely convincing, I identified two critical issues. First, I do not think that his use of the term “light verb” is a good choice. The term is already used in linguistics and although it might not be used in a clear and consistent way, adding an additional use is not really helpful. I think that Cysouw does not need the term “light verb” as he seems to equate light verbs with auxiliaries. On page 10, he states that he simply refers “to the whole group of these auxiliary verbs as light verbs.” Why is it necessary to introduce an additional term if the verbs referred to as “light verbs” are just auxiliaries?
Cysouw proposes that the classification of a verb as a light verb is based on the notion of “monoclausality.” Thus, a light verb is a finite verb realized together with a nonfinite main predicate within a monoclausal construction. An example of such a light verb is sehen ‘see’ in (6). The verb – like other perception verbs – functions as a light verb in the experientive diathesis. It is the finite verb realized together with the infinitive schenken which, according to Cysouw, is the main predicate.
If we adopt Cysouw’s analysis, verbs like lehren ‘teach’ and helfen ‘help’ qualify as light verbs as well. In fact, the two verbs are light verbs in Cysouw’s analysis. The example in (7) presents another instance of the category “light-verb alternation with infinitive” and exemplifies a diathesis Cysouw calls “assistive.” As in the previous example, Cysouw considers the infinitive to be the main predicate of the clause.
The status of the main predicate is the second critical issue I identified with Cysouw’s analysis. As mentioned, he treats the infinitives rather than the finite verbs as the main predicates in (6) and (7). As a consequence, Cysouw considers the examples in (6) and (7) to represent diatheses of the verbs schenken and schwimmen, respectively. But why should schwimmen rather than lehren be the main predicate of (7)? The sentence refers to an event of teaching rather than to an event of swimming. The sentence is even compatible with the mother teaching the boys swimming without actually swimming. Thus, I think there are good semantic reasons – the denoted event – which speak in favor of treating lehren (7) and sehen (6) as the main predicates. Thus, I would follow the more traditional analysis and treat the infinitives as object complements of the finite verbs. Correspondingly, the examples in (6) and (7) do not represent any diathesis. This reduces both the number of light verbs as well as the number of distinct diatheses.
Despite any criticism, Cysouw develops an interesting data-based analysis of German diatheses and epitheses. The volume will appeal to a readership interested in German morphosyntax and it provides a perfect starting point for further analyses of individual diatheses. Overall, the book presents a very valuable and rich contribution to a central aspect of German morphosyntax.
