Unaccusative verb
The term unaccusative verb (also: ergative verb ) describes a class of verbs in linguistics that are externally marked as intransitive verbs , i.e. have a subject in the nominative as the only grammatical addition , but which treat this addition according to various other grammatical criteria, how else the direct object is handled.
The term unaccusative comes from linguistic analyzes, according to which such verbs basically have no subject, but only a grammatical object, but cannot assign an accusative case to it, because a sentence can only have an accusative object if it is also transitive , i.e. a relationship between expresses an acting subject and its object. The visible nominative subject of the unaccusative verbs is then only superficially a subject, but not a subject in the full sense. This results in a dichotomy within the class of "intransitive" verbs. The distinction has far-reaching consequences for the grammatical behavior of the verbs; for example, the non-accusative type cannot be passivated .
In languages like German, unaccusativity is a hidden grammatical category that is only expressed in the behavior of verbs or their subjects, but is not displayed in word forms . Some other languages also show two groups of verbs in the allocation of cases or in the personal forms of the verb ( congruence ), namely those that mark their only complement as a subject, and others that mark their only complement as an object; in such cases, however, one speaks more of “split intransitivity” or an active-inactive language . In both cases, however, similar groups of verbs appear to be involved, so the problem arises as to which factor in the meaning of verbs explains this behavior.
Terminology problems
The term “unaccusative” goes back to the work of David M. Perlmutter (1978) and was first formulated in the context of his model of relational grammar . The term “ergative verb” is equivalent in the sense of Luigi Burzio (1986) (this term, however, has nothing to do with ergative as the name of a case).
Since the unaccusative thesis divides the class of verbs traditionally called “intransitive” into two parts, the problem arises of finding a clear designation for verbs with a real subject as the only argument. In connection with the terminology of Burzio (1986), the opposite of unaccusative (or “ergative”) verbs - ie the “normal” intransitive verbs - is often referred to as “unergative”. Another possibility is to replace the traditional term “intransitive” with the term “single-digit verb” (according to Levin & Rappaport 1995) and then to divide this category into “intransitive (in the narrower sense)” and “unaccusative”.
In the case of whole verb pairs, in which the same verb can be used both transitive and unaccusative, there is also the term unstable verb .
Relevant grammatical phenomena in German
Formation of the "past participle"
In German, participles can be derived from many verbs , which can then be used like an adjective. The so-called " past participle passive " usually refers to the object of the underlying verb (hence it is often referred to as "passive"):
der mitgebrachte Wein -- „der Wein“ = y aus: x bringt y mit die geleerten Mülltonnen -- „die Mülltonnen“ = y aus: x leert y das geputzte Bad -- „das Bad“ = y aus: x putzt y
Intransitive verbs cannot normally appear in this construction; This is explained by the fact that the participle cannot refer to the subject of the underlying verb, as can already be seen in the examples above.
*nicht: *der gesungene Mann mit „der Mann“ als x aus: x singt (y) *die geschlagene Uhr mit „die Uhr“ als x aus: x schlägt *die geputzte Hausfrau mit „die Hausfrau“ als x aus: x putzt (y)
However, among the intransitive verbs there is a group of verbs that are exceptions (so that the designation "past participle passive " is no longer applicable):
der eingeschlafene Hund -- „der Hund“ = x aus: x schläft ein die eingetroffene Post -- „die Post“ = x aus: x trifft ein das verrostete Scharnier -- „das Scharnier“ = x aus: x verrostet
The verbs fall asleep , arrive and rust are examples of verbs that treat their only complement (x) in a way that otherwise only objects can be treated (namely the y from the first group of examples) - these verbs are examples of unaccusative verbs.
Impersonal passive
In German, transitive and intransitive verbs can appear in a passive form . Each time in the passive voice the underlying (“logical”) subject is not taken over into the sentence; if the verb was transitive, it moves the underlying object into the subject position instead. If the underlying verb was intransitive, after the deletion of the underlying subject there is no supplement and an impersonal construction without a subject is created:
x leerte die Mülltonnen --> die Mülltonnen wurden geleert x reparierte den Motor --> der Motor wurde repariert
x rauchte --> (Also) wurde geraucht. x putzte --> (Den ganzen Tag) wurde geputzt
However, some intransitive verbs do not allow such an impersonal passive, including precisely those verbs that turn out to be unaccusative with regard to the participle formation. So the examples show the effect that these verbs do not treat their complement the way a subject is otherwise treated:
x schlief ein --> *nicht: Dann wurde eingeschlafen x traf ein --> *nicht: Da wurde (von der Post) eingetroffen x verrostete --> *nicht: Überall wurde verrostet
Resultative adjectives
In the resultative construction , a verb that describes an activity is combined with an adjective that names the result of this activity:
den Tisch trocken reiben -- Aktivität reiben bewirkt das Resultat, dass der Tisch trocken ist den Teller leer essen -- Aktivität essen bewirkt das Resultat, dass der Teller leer ist das Taschentuch nass weinen -- Aktivität weinen bewirkt das Resultat, dass das Taschentuch nass ist
In general, the result is stated by the object of the construction, whereby the object can be newly introduced together with the adjective and does not have to be a supplement required by the verb (as in the examples above). According to this rule, a result cannot be predicated of the subject. However, if this contextual reference is still to be established, there is the option of additionally introducing a meaningless reflexive pronoun . This has the effect that, in a purely formal manner, a grammatical object is available that can provide the carrier of the result:
Der Kranke hat sich wund gelegen -- Aktivität liegen (des Kranken) bewirkt das Resultat dass der Kranke wund ist Sie lachte sich tot -- Aktivität lachen bewirkt das Resultat dass „sie tot“ ist (in einem übertragenen Sinn).
Some verbs do not allow such a reflexive, instead they appear in constructions in which the resultative adjective apparently does refer to the subject. Such constructions are relatively rare, but examples can be found in both German and English.
der Motor lief heiß -- *nicht: der Motor lief sich heiß
Englisch: the door slid open -- *nicht: the door slid itself open „die Tür glitt auf“
The verbs in this latter construction are again to be described in such a way that they treat their subject in the same way as the grammatical object is otherwise treated. This also corresponds to the possibility of forming a participle like the overheated engine . The rule that the resultative adjective must refer to the object is observed if the subjects of this construction are to be seen as the underlying objects. At the same time this also explains why the reflexive not even used in addition may : The grammatical function of the object, which would have to take it, is occupied by this analysis already from the unit, which is apparently subject.
Auxiliary verbs in the perfect tense
Various other criteria were named for different languages that characterize non-accusative verbs, but are not applicable or controversial in German. Often z. For example, the choice of the auxiliary verb in the perfect tense is mentioned as a criterion: Unaccusative verbs should tend to be formed with a sein -Perfekt, genuinely intransitive with a haben -Perfekt. However, it is not clear why this connection should exist, and furthermore there are cases in various languages where a sein -Perfect is extended to verbs which should not be unaccusative, e.g. B. at verbs of active movement such as running, swimming , etc. in the German or weather verbs such as rain in Italian (the Italian form ha equivalent to German has , and è correspond German is ):
| German | Italian |
|---|---|
| He is running | Ha corso |
| It has been raining | È piovuto |
Grammatical structure
The question of whether the unaccusativity effects mean that the nominative complement of these verbs is actually an object is controversial.
One possible position is that the effects that a supplement is treated like an object fall within the realm of semantic interpretation and need not be viewed as grammar rules.
The opposite view, held by many followers of transformational grammar , is that in an underlying representation the additions in question are actually used as the object of the verb and, through a transformation, get into the subject position, where they receive the nominative case. Against this background, the term “unaccusative” is to be understood: They are verbs that actually only have one grammatical object, but which they cannot assign to an accusative case (as a result of Burzio's generalization ). As can be seen in the examples of the resultative construction, normal intransitive verbs (see the examples with eating, crying, lying, laughing ) are in principle capable of appearing with an accusative object, only the verbs (hot) can run, glide, etc. not this.
Derivation from the meaning of the verb and questions of language comparison
Since the phenomenon of unaccusativity was discovered, attempts have been made to explain it in terms of verbs. A simple approach is to fix the phenomenon to semantic roles and to suspect that unaccusative verbs are those for which the semantic role of the argument is the same as for grammatical objects of transitive verbs. For example, there is a variant of the verb to break as a transitive verb and one with only one argument that is unaccusative. The latter actually only has the argument that corresponds to the object in the transitive construction . Hence, it might be expected that it should behave like an object in some ways too:
Das Kind zerbricht die Tasse Die Tasse zerbricht
Accordingly, if the only argument has a role akin to that of the patient , then unaccusativity would follow; if it has a role related to an agent , it would result in a truly intransitive (unergative) verb. Very early on, however, attention was drawn to differences in the behavior of verbs between languages that are at least equivalent in translation, for example the different classification of verbs like die, sweat or blush in different languages. Since then, other authors have suggested that subtle differences in the meaning of such verbs could explain the fluctuation between unaccusativity / intransitivity. B. the Italian arrossire (blush, translated as English "blush") from the Dutch blozen (also translated into English as "blush") in that arrossire inevitably denotes a change, while ndl. blozen a permanent state of affairs. Such a difference could lead to the fact that the Italian verb attributes object properties to its argument (i.e. change) and the meaning-like Dutch verb does not (because body states count as a kind of caused situations, i.e. result in agent-like verbs in the relevant sense). An explanation of why the Italian verb for "blush" (or "blush") is unaccusative while the Dutch verb is not, could therefore depend on differences that are finer than conventional semantic roles.
In summary, the following grouping of verb meanings results (whose assignment, as indicated, can still be fluctuating or controversial in some cases):
Real intransitive verbs :
- Agentive verbs (conscious causation), e.g. B. run, work
- Verbs that denote involuntary physical processes, e.g. B. sneezing, bleeding, sweating (controversial)
- Non-agentive verbs whose only supplement otherwise contains the internal causal factor of a situation, e.g. B. shine, bubble, stink
Unaccusative verbs
- Verbs of the change of state, e.g. B. be red, break (itr.) Arrive
- Verbs of directed change (without a fixed final state), e.g. B. cool down, rise
- Agentive verbs, if they also denote a change in the agent, e.g. B. stand up
- Various verbs that denote neither a causal factor nor a change of state or a targeted change, e.g. B. roll, exist .
The existence of an unexplained remainder of “other” verbs leads Levin & Rappaport (1995) to the position that the status of a superficial complement as a real subject depends on the presence of certain semantic factors (which were referred to above as “agent-like”); Incidentally, however, unaccusativity even represents the normal case of how a single argument relates to a verb. Accordingly, there would be no uniform semantic factor that unaccusative verbs would have to have in common, because it is essentially a residual class that simply remains after the determination of “real” subjects. However, the factor “directed change” must still be added as an independent trigger of unaccusativity, since this may still have priority over agent-like properties (as the example shows, stand up ).
literature
- Burzio, Luigi (1986): Italian Syntax . Dordrecht: Kluwer
- Kaufmann, Ingrid (1995): “O- and D-Predicates: A Semantic Approach to the Unaccusative-Unergative Distinction”, Journal of Semantics 12 (4): 377-427. doi : 10.1093 / jos / 12.4.377
- Levin, Beth & Malka Rappaport Hovav (1995): Unaccusativity . Cambridge MA: MIT Press
- Levin, Beth & Malka Rappaport Hovav (2000): “Classifying Single Argument Verbs.” In: Peter Coopmans et al. (eds.): Lexical Specification and Insertion , Amsterdam, John Benjamin. pp. 269-304
- Perlmutter, David & Carol Rosen (eds.) (1984): Studies in Relational Grammar. Chicago, University of Chicago Press. ISBN 0226660508
- Reuland, Eric: Arguments and Case: Explaining Burzio's Generalization . ISBN Amsterdam 2000. ISBN 902722755-1
- Rosen, Carol (1984): “The Interface between Semantic Roles and Initial Grammatical Relations.” In Perlmutter & Rosen (eds), pp. 38-80
- Van Valin, Robert (1990): "Semantic Parameters of Split Intransitivity." Language 66, 221-260
Individual evidence
- ↑ See Levin & Rappaport (1995)
- ↑ z. B. Kaufmann (1995)
- ↑ So already Perlmutter (1978) in his original description in the model of the "Relational Grammar", cf. Perlmutter & Rosen (eds.) (1984). Likewise Burzio (1986), which almost the entire literature followed in the generative grammar model.
- ↑ So also the short description in: DUDEN, the grammar. 8th edition 2009, p. 412.
- ↑ Rosen (1984); Data from American Indian languages that were cited here were, however, described elsewhere as possibly misanalysed, cf. Van Valin (1990)
- ↑ Levin & Rappaport Hovav (2000), pp. 297f.
- ↑ Levin & Rappaport (2000), p. 280
- ↑ see e.g. Rosen (1984), Levin & Rappaport (1995, 2000)