Role and Reference Grammar
The Role and Reference Grammar (RRG) is a structuralist and functional grammar theory that was developed by the two American linguists Robert D. Van Valin, Jr. and William A. Foley in the 1980s.
The RRG is one of the best known grammar theories in Europe and the USA today. This is not only due to the large number of simple and compound sentence phenomena she describes. It is also due to the ability to conclusively combine the three major areas of grammar syntax , semantics and pragmatics in one theory. Last but not least, this has to do with the universality or typological adequacy of this theory, which tries to provide a descriptive apparatus for as many, ideally all languages worldwide.
According to Van Valin, the RRG is a structural-functional theory ("a structural-functionalist theory of grammar", Van Valin 1993: 1). It is thus on a continuum between extremely formal and extremely functional theories. According to an extremely formal theory such as that on which the various versions of N. Chomsky are based, language is an inventory of structural descriptions of sentences that indicate the sound shape and meaning of a linguistic expression. Language is reduced to grammar, whereby there is no longer any space for communicative functions and substantial semantics and syntax is seen as a separate, autonomous area. The radical functional conception of an “emergent grammar”, for example, which PJ Hopper represents, lies at the other end of the continuum in that it denies Saussure's conception of language as a structural system of signs and tries to reduce grammar to discourse. Language is then ultimately just a collection of fixed phrases and formulaic expressions that are coded by various information strategies and discourse patterns. The RRG lies between the two extremes. Compared to formal theories, the RRG emphasizes that the grammatical structure can only be understood and explained in relation to its semantic and pragmatic-communicative functions: "Syntax is not autonomous" (Van Valin 1993: 2). In relation to rigid functionalism, the RRG rejects the position that grammar is random and therefore unlearnable, and claims that grammar is determined semantically and pragmatically.
William A. Foley and Robert D. Van Valin, Jr. had just completed their dissertations on exotic languages at Berkeley, although Eurocentric grammar was often of little help (Foley 1976; Van Valin 1977) than they did in the second half of the 1970s Started thinking about a new grammar theory. The first document was a co-authored article from 1977. The RRG was thus developed in describing the Austronesian, Australian and Indian languages. This theory only achieved systematic consistency in the first half of the 1980s. Up until that time, both linguists worked together. Later, only Van Valin appeared with publications, while Foley became a leading field researcher and typologist.
At these beginnings it becomes clear what characterizes the RRG: on the one hand, the search for a linguistic theory that also has explanatory power for non-European languages such as the North American Lak (h) ota, the Austronesian Tagalog and the Australian Dyirbal, on the other hand the question of how the relationship and interaction of syntax, semantics and pragmatics can be adequately described.
The name of this theory, Role and Reference Grammar or German: Role and Reference Grammar, goes back to the two major areas of investigation: the level of semantic (case) roles and the level of the referential or pragmatic properties of noun phrases, i.e. H. the level of the Reference Grammar, as the Pragmatic Grammar was also called in the initial phase.
Sources on this grammar theory
The earliest publication on RRG is an essay written jointly by Foley and Van Valin on the subject concept in universal grammar, which appeared in 1977. Besides three presentations of the theory of normal scope (Foley & Van Valin 1980; Van Valin 2001, Van Valin approx. 2009) and a longer exposition of the theory (Van Valin 1993), the first major summary of the RRG is a joint effort by W. Foley and R. Van Valin written book (Foley & Van Valin 1984), which, however, still reflects an older status in some points and should therefore always be used together with newer publications. A later book (Van Valin & LaPolla 1997) forms the draft of a so-called typological grammar, undoubtedly the most important and most comprehensive publication on the RRG. An updated version is the book by Van Valin published in 2005. The following sections 3 to 6 are based on the two presentations by Foley & LaPolla 1997, Van Valin 2001 and Van Valin 2009. Some English terms have not been translated to avoid terminological confusion.
The semantic analysis of the sentence structure
In the RRG, the semantics is the most important level of description of the sentence, even if the syntax retains a certain independence and cannot be completely reduced to the sentence level as in generative semantics. Among other things, it is important to analytically prove semantic incompatibilities (sentences such as “The moon lit a surprise with peace”) to be unacceptable and, conversely, to ensure the semantic acceptability of a meaningful sentence.
A basic typology for semantic sentence analysis is based on Z. Vendler (1967) and D. Dowty (1979) and consists of four types of states of affairs according to the semantic parameters of dynamics and control:
These types of facts correspond to types of action that have been expanded to include “semelfactives” and “active accomplishments”. Types of action are the inherent temporal properties of verbs and, to that extent, linguistic properties, while the facts relate to the extra-linguistic. There is a causative counterpart to each of these six types of action types. This creates twelve semantically defined types of action:
The twelve types of action (English examples)
a. State: The boy fears the dog.
a '. Causative state: The dog frightens / scares the boy.
b. Achievement: The balloon popped.
b '. Causative achievement: The cat popped the balloon.
c. Semelfactive (repeated, selective events): The light flashed.
c '. Causative semelfactive (kaus. Wied. Point. Event): The conductor flashed the light.
d. Accomplishment: The ice melted.
d '. Causative accomplishment: The hot water melted the ice.
e. Activity: The dog walked in the park.
e '. Causative activity: The girl walked the dog in the park.
f. Active accomplishment: The dog walked to the park.
f '. Causative actives accomplishment (kaus. Active process): The girl walked the dog to the park.
There are (according to Dowty) five tests to determine which type of action it has for a concrete verb (e.g. does / does not occur with a progressive verb form (- static, - punctual; only in languages such as English, Spanish and Icelandic), does / does not occur with adverbs like fast, slow etc. (+ - dynamic), does / does not occur with X in an hour (+ - telic);).
The crux of this theory of action types is the verb, which is the most important part of a sentence in semantic terms. The semantic representation of a sentence is based on the lexical structure of the verbs. What is important in the RRG compared to other theories is that the lexical entry of a verb does not contain a list of thematic case roles. Rather, the RRG uses a system of lexical decomposition to reduce the verb meanings in a metalanguage to a few predicates (in the logical sense) on the basis of the types of action and to trace related meanings back to grammatical modifications of the predicates in the sense of the types of action. So die and kill have the decompositional structure: y becomes dead (the predicate); x causes [y becomes dead]. This analysis, carried out in metalinguistic terminology, is cross-lingual and applies to English, German and Latin as well as Georgian. Grammatical modifiers such as BECOME ('will') stand for a temporal change and CAUSE ('cause') for a causative relationship between two predicates. According to the representation of the RRG, a simple transitive sentence would have the following semantic structure:
a) The boy broke the glass
b) [ do ′ (boy, Ø)] CAUSE [BECOME broken ′ (glass)]
Such a system makes semantic and lexical analysis more tangible because only the primitive predicates have to be defined, but not the grammatical elements of the meanings such as BECOME and CAUSE. For example, with the three verbs of the same word field already quoted, death needs to die, kill and be dead (or in Latin mori, interficere and mortuum esse or in Georgian mok'vda and daixoca vs. movk'ali and davxoce or in Old Hebrew muth vs. qatal) only the predicate to be defined dead (e.g. as (- living)), while the other two verbs can easily be described as semantic derivatives. In addition, a distinction is made between the core meaning that finds its way into the lexicon and its contextual modifications (e.g. eating versus eating a slice of pizza ), which result from the specific analysis of the facts. Polyseme verbs have multiple lexical entries.
With the decompositional analysis, which forms a core area of the semantic analysis of the RRG, it is also easier to assign verbs to the types of action. Conversely, this typology can be clearly described analytically using the lexical decomposition, as the following table from Van Valin & LaPolla (1997: 109) shows:
Verb class - Logical structure
State: predicate ′ (x) or (x, y)
Activity: do ′ (x, [ predicate ′ (x) or (x, y)])
Achievement: INGR predicate ′ (x) or (x, y), or
INGR do′ (x, [predicate′ (x) or (x, y)])
Accomplishment: BECOME predicate ′ (x) or (x, y), or
BECOME do′ (x, [predicate′ (x) or (x, y)])
Active Accomplishment: do ′ (x, [ predicate 1 ′ (x, (y))] & BECOME predicate 2 ′ (z, x) or (y)
Causative: α CAUSE β, where α, β are LSs of any type
The participants or arguments of the individual types of facts are linguistically represented by different types of noun phrases. This role structure is based on the "inherent lexical content" (M. Silverstein), i. H. on person, number, humanity, liveliness and concreteness. Also important in this conception is the expression of the different roles that the participants can play in different situations.
The basic system in the RRG distinguishes between semantic macro roles, i.e. H. between participants who carry out, cause, set in motion or control the situation indicated by the predication, and those participants who do not do all of this. The first type is the perpetrator ("actor"), the second type the sufferer or the affected person ("undergoer"). Compared to these two macro roles, the case roles that can be traced back to Charles Fillmore, such as agent, patient, experiencer, have no specific theoretical status in the RRG. They are designations for argument positions in logical structures and result from the meaning of the verb. While one can predict these case roles from the verb without a syntactic context, the reverse is not possible from a noun or prepositional phrase: "Since the role of a participant is a function of the state of affairs it is involved in, the semantic function of an argument referring to a participant should follow from the representation of the verb or any other predicate coding the state of affairs. ".
The choice of the semantic macro roles of the actor and the undergoer determines the perspective from which the situation (with which the different types of facts are described more generally) is described. The actor is not simply to be equated with the syntactic subject, how easily an English sentence pair of an active and passive sentence shows where in the passive sentence it is not the syntactic subject but the by-phrase that is the actor. The opposition of actor and undergoer is best shown in transitive sentences, where the actor is the subject and the undergoer is the direct object. It is fundamental for the semantic case structure of the sentence.
Typologically, the two basic categories actor and undergoer are universal, while the semantic case roles such as agent, patient and experiencer are individual language categories and can vary in number. They are therefore of little interest to the RRG. (The same applies, by the way, to the syntactic categories subject, direct and indirect object.)
The pragmatic analysis of the sentence structure
"The second, larger system in sentence level grammar deals with the organization of the sentence in terms of the discourse role of its constituents." (FVV 1980: 338 transl.) While Foley and Van Valin initially called this system "referential structure", it later became denotes pragmatic structure of the sentence. (This also explains the meaning of the second term in the term Role and Reference Grammar.) This later version is largely based on the approach of Knud Lambrecht (1986; 1994). In the endeavor to ensure the grammaticality and acceptability of sentences, pragmatics tries to recognize disturbances in the transmission of information. Such disruptions come about when an utterance is in the wrong context or the reference is incorrect. So the sentence you said he said nothing is unacceptable if it is at the beginning of the text, so that the two personal pronouns have no reference. In the pragmatic analysis, the RRG tries above all to pay attention to the contextual appropriateness.
The most important terms are the pair of terms topic and focus , which is used in the sense of the pragmatic approach of K. Lambrecht (1994). So topic means "that entity which the sentence or proposition is about." and to that extent is the subject of a larger linguistic unit. In contrast, focus means "the semantic component of a pragmatically structured proposition whereby the assertion differs from the presupposition."
Both terms, which were used for the first time by the Prague School of Linguistics, and especially by J. Firbas, are used in many more recent theories and u. a. Central parameters in the Functional Grammar by Simon C. Dik (different from the Functional Discourse Grammar , according to which both terms are not complementary). According to the RRG, topic and focus are based on Lambrecht “the two primary information statuses that referring expressions may have in an utterance”. Like Lambrecht, the RRG differentiates between different forms of focus structures: the predicate focus, the sentence focus and the narrow focus. In the first type, the unmarked case, there is a topic, the predicate expressing a comment about the topic: Q .: What happened with your car? - A.: My car / It broke DOWN. In a sentence focus, the entire sentence is in focus, so there is no topic: Q .: What happened? A .: My CAR broke down. In contrast, in a narrow-focus domain, the focus is limited to a single constituent: Q .: I heard your motorcycle broke down. A .: My CAR broke down.
The pragmatic structure is marked either by case marking (as in German) or by word order (as in English), whereby case marking means any morphological marking that describes the grammatical function or the pragmatic conspicuousness or both of a noun phrase in a sentence. Since the pragmatic structure is a second system of sentence organization, the question arises whether there are any relationships to the system of role structure. Even if both systems are independent of each other and the choice of actor is not determined pragmatically, there is an interaction that z. B. means that the actor is the unmarked choice of the focus and for the English a hierarchy runs from the actor via the undergoer to other constituents: ACTOR> UNDERGOER> OTHER.
It is important that the RRG also graphically illustrates the pragmatic structure in the sentence images. The focus structure is shown separately as a projection. Examples of such typesetting can be found in Van Valin & LaPolla (1997: 215-217).
There are some intra-clausal and interclausal tests that can be used to determine whether or not a language has a pragmatic structure. These tests are: the choice of subject, which must not be semantically determined, the existence of passive and anti-passive constructions, the formation of relative clauses and the deletion of co-referentiality in complementary constructions. Foley and Van Valin use such tests to prove negatively that, alongside many other languages, Choctaw is a language that does not have a pragmatic structure. Some other languages without a pragmatic structure are also mentioned. If such languages have a formal passive and anti-passive construction, these systems have a different function, e.g. B. the suppression of actors and undergoers.
The syntactic analysis of the sentence structure
In the area of syntax, the grammatical relations that nouns have in relation to the verb are important for the RRG. This theory has a completely different attitude towards these grammatical relations than other theories. Mainly because of certain description difficulties in non-European languages, the RGG does not ascribe any cross-lingual validity to the traditional relations of subject, direct and indirect object and therefore does not use the terms as theoretical or analytical constructs. This different attitude towards traditional clauses is also evident in the various syntactic operations such as the formulation of the passive and antipassive, the rules of case assignment and the rule of congruence, which all relate to the two semantic macro roles. Universal, on the other hand, is the nucleus of a sentence, a predicative argument and the arguments, which are usually noun phrases. Since these sentence constituents are semantically defined (see Section 3 above), the universality of the RRG lies in the field of semantics (see Wikipedia article Robert D. Van Valin, Jr. ).
Instead, the RRG has a construction-specific conception of grammatical relations and only postulates a single relation, the privileged syntactic argument. This term also includes so-called syntactic pivots, e.g. B. the checkers for verb congruence or the precursors of reflexive words. There is also a hierarchy for choosing the privileged syntactic argument that corresponds to the actor-undergoer hierarchy. There are also selection principles for syntactic accusative constructions and for syntactic ergative constructions.
In addition, there is the conviction expressed in the syntactic operations that the semantic level, including the definition of the macro roles, is universal, while the syntactic level, which means the formation of the semantic case roles, remains an individual language. All of this also shows that the syntax in the RRG is not an autonomous area, but is subject to semantic rules.
In the concrete sentence analysis, the RRG assumes that only those trains are recorded that are universal. For example, the verb phrase is ruled out as a universal quantity. Therefore, the RRG also rejects the representations of the Relational Grammar and the Lexical Functional Grammar, which are based on grammatical relations, as well as the constituent structures of the X-bar type, because they are not universally founded.
Like almost all newer syntax theories, the RRG uses tree graphs to illustrate sentence structures. The sentence image connects aspects of constituency and dependency by representing constituents, some of which are semantic arguments. In addition, there are so-called operators, which represent the grammatical morphemes aspect, tense, negation and illocutive power in a sentence image, which is then differentiated according to constituent projection and operator projection. The privileged syntactic argument has no direct structural representation; it is therefore more a property of the linking algorithm than of the sentence structure. The RRG only assumes a single syntactic representation for each sentence, which corresponds to the surface shape of the sentence. There are no abstract representations like in the Lexical Functional Grammar or deeper levels of representation like in transformation grammar, relational grammar and government and binding theory. Many sentence descriptions of the RRG have a. a heuristic value because they allow languages to be assigned to specific types.
The analysis of complex sentences
Unlike some other newer grammar theories, the RRG also pays great attention to complex sentence structures. It differentiates between different types of sentence linking, called nexus relations, and in addition to coordination and subordination also include co-subordination, which occurs when the dependent and matrix sentences share important operators such as aspect or when there is switch reference between the two sentence units. At the same time, the sentence can be linked on different levels of the sentence: "There are only three types of units involved in complex sentences in universal grammar, and these are the clause, the core and the nucleus". The type of integration of the dependent sentence in the matrix sentence forms a hierarchy from clausal coordination as the lowest integration to nuclear cosubordination as the closest integration:
Figure: The hierarchy of interclausal syntactic relations (Interclausal syntactic relations hierarchy)
Ebenen und Satzverbindung Beispiele Grad der Integration Nucleas Cosubordination: Max made the woman leave. Strongest: Tightest integration into a single unit Nuclear Subordination: z. B. im Koreanischen Nuclear Coordination: F: Je ferai manger les gâteaux à Jean. Core Cosubordination: Ted tried to open the door. Core Subordination: David regretted Amy’s losing the race. Core Coordination: Louisa told Bob to close the window. Clausal Cosubordination: Harry ran down the hall laughing loudly. Clausal Subordination: John persuaded Leon that Amy had lost. Clausal Coordination: Anna read for a few minutes, and then she went out. Weakest: Least integration into a single unit
These syntactic sentence linkage relations are used to express certain semantic relations between the units of the linkage. Such semantic relations are e.g. B. Causality, Intention, and Time Sequence.
Operators and focus structures are also described in complex sentences (Van Valin & LaPolla 1997: 455-68; ibid .: 484-92) and the connection between syntactic and semantic representations (Van Valin 1993: 124-51). As with the pragmatic focus structures, the RRG uses corresponding sentence images to analyze complex sentences.
The relationship of the RRG to other theories
As should be clear from the above, the RRG differs in many points empirically and conceptually from many other linguistic theories. Nonetheless, it is an indirect or critical successor to Generative Transformation Grammar because, on the one hand, it is directly dependent on Generative Semantics and Case Theory , both of which come from transformation grammar, and on the other hand, it differs significantly from transformation grammar. The RRG is also influenced by the relational grammar, which also goes back to the transformation grammar. Nevertheless, this functional theory differs in some not insignificant points from the transformation grammar and its successors up to the X-bar theory and the government binding theory : above all in the renunciation of the autonomy of the syntax in favor of a syntax controlled by semantics and pragmatics -Conception and a reduction of the syntactic analysis to the syntactic surface.
However, the RRG is closely related to the Functional Grammar by Simon C. Dik , with which there was an exchange from the beginning: In the book from 1984 Dik's standard work from 1978 was cited; his original clause layering theory was inspired by the RRG. Common features of the RRG and the FG are: Both theories strive for typological adequacy, in both theories semantics and pragmatics are the primary areas of language description. Both theories use a four-fold semantic typology of predicate frames for the sentence analysis, and both theories explain, albeit with z. Sometimes there are subtle differences, the pragmatic structure of a sentence and text with the same concept of known message (topic) and most important information (focus). However, there is a crucial difference between the two theories: the FG has no real syntax or syntactic representation, its sentence level model is semantic, while that of the RRG is syntactic.
In contrast, the RRG is less closely related to the newer Functional Discourse Grammar, which is derived from the Functional Grammar. In addition to general similarities (functional approach, semantics and pragmatics as central levels of description), two main differences are that with RRG, semantics is the most important description area, while with Functional Discourse Grammar, pragmatics is more important, and that the latter theory has more levels and Components is designed.
Current and future trends
The RRG has been analyzed and developed a lot in linguistic research to this day. Van Valin has a research grant from the Max Planck Society to fund a research group on syntax, typology and information structure for five years at the Max Planck Institute in Nijmegen. Considerable work is done in integrating the RRG into computer programs. To facilitate this research, conferences have been introduced that have been held annually in university cities around the world since 1999, for example in Santa Barbara in 2001, in Taipei in 2005 and in Leipzig in 2006 (see first web link). Supporters of this theory not only worked in the field of universals and language typology, but also, like the representatives of the Lexical Functional Grammar, tried to incorporate their theory into a language-psychological theory of language understanding and language generation. And as for adherents of the Principles and Parameters approach, the question of language acquisition is of great interest to them, even if they do not assume the existence of an autonomous language-generating apparatus.
Three more recent anthologies deserve special attention here:
- Van Valin (ed.): Investigations of the syntax-semantics-pragmatics interface . 2008, Amsterdam
- Kailuweit u. a. (Ed.): New applications of Role and Reference Grammar . 2008, Cambridge
- Guerrero et al. a. (Ed.): Studies in Role and Reference Grammar . 2009, Mexico City
This grammar theory, which has now been asserting itself for over 30 years without major revisions, has a number of advantages which should favor its further dissemination: It has a relatively simple formal description apparatus. It considers the components syntax, semantics and pragmatics halfway equally, even if the last two are slightly overweight. She tries to describe as many natural languages as possible. From a European point of view, it could be perceived as a disadvantage that the traditional clauses subject and object, which are stigmatized here as individual linguistic parameters, are not very well used in this theory. On the other hand, the dissolution of the conventional subject concept in ergative languages either leads to the abandonment of a universal grammar component or to the adoption of a grammar theory such as the RRG, which however also requires concessions and rethinking. As long as empirical and theoretical typological research continues to advance, theories such as the RRG and the functional grammar are ideal descriptive apparatuses.
- William A. Foley & Robert D. Van Valin, Jr .: On the viability of the notion of "subject" in universal grammar. In: Proceedings of the Third Annual Meeting of the Berkeley Linguistic Society. Berkeley 1977, pp. 293-320.
- FVV 1980: William A. Foley & Robert D. Van Valin, Jr .: Role and Reference Grammar. In: EA Moravscik & JA Wirth (eds.): Current approaches to syntax. New York: Academic Press 1980, pp. 329-352.
- William A. Foley & Robert D. Van Valin, Jr .: Functional syntax and universal grammar. Cambridge 1984.
- Robert D. Van Valin: A synopsis of Role and Reference Grammar. In: Robert D. Van Valin, (Ed.): Advances in Role and Reference Grammar. 1993, pp. 1-164.
- Robert D. Van Valin, Jr. & Randy J. LaPolla: Syntax: Structure, meaning, and function. Cambridge University Press, Cambridge 1997.
- Robert D. Van Valin, Jr .: An introduction to syntax. Cambridge 2001, pp. 205-218.
- Robert D. Van Valin, Jr .: Exploring the Syntax-Semantics Interface. Cambridge: Cambridge University Press 2005.
- Robert D. Van Valin, Jr .: An Overview of Role and Reference Grammar. 2009 ( PDF; 858 kB on buffalo.edu)
- Simon C. Dik: Functional Grammar. Dordrecht 1978.
- David Dowty: Word Meaning and Montague Grammar. Dordrecht 1979.
- William A. Foley: Comparative syntax in Austronesian. Unpublished doctoral dissertation. University of California, Berkeley 1976.
- William A. Foley: The Papuan languages of New Guinea. Cambridge 1986.
- Lilián Guerrero, Sergio Ibáñez Cerda & Valeria A. Belloro (Eds.): Studies in Role and Reference Grammar. Mexico City 2009.
- Rolf Kailuweit, Björn Wiemer, Eva Staudinger & Ranko Matasović (eds.). New applications of Role and Reference Grammar. Cambridge 2008.
- Knud Lambrecht: Topic, Focus, and the Grammar of Spoken French. Dissertation. University of California, Berkeley, 1986
- Knud Lambrecht: Information Structure and Sentence Form: Topic, Focus, and the Mental Representations of Discourse Referents. Cambridge University Press, 1994
- Robert D. Van Valin, Jr .: Aspects of Lakhota syntax. Unpublished doctoral dissertation. University of California, Berkeley 1977.
- Robert D. Van Valin, Jr .: Recent Developments in the Role and Reference Grammar Theory of Clause Linkage. In: Language and Linguistics. 8, 2007, pp. 71-93.
- Robert D. Van Valin, Jr. (Ed.): Advances in Role and Reference Grammar. Benjamin, Amsterdam 1993.
- Robert D. Van Valin, Jr. (Ed.): Investigations of the syntax-semantics-pragmatics interface. Amsterdam 2008.
- Robert D. Van Valin, Jr. & David P. Wilkins: Predicting syntactic structure from semantic representations: remember in English and Mparntwe Arrernte. In: Van Valin (Ed.): Advances in Role and Reference Grammar. 1993, pp. 499-534.
- Zeno Vendler: Linguistics in philosophy. Ithaca / New York 1967.
- Full bibliography at: Role and Reference Grammar [Bibliography]. Version: November 12, 2009, online at buffalo.edu, (Status: November 12, 2009, PDF (English); 1.1 MB)
- Official platform of the RRG (as of September 1, 2009), including a. a. an “Overview of RRG”, a “Bibliography of Work in RRG”, an overview of Conferences 1999–2009 and instructions for creating typesetting with a graphics program (RRGDraw)
- RD Van Valin's website at Heinrich Heine University Düsseldorf, as of September 1, 2009
- RD Van Valin's website at the University of Buffalo, as of September 1, 2009
- ↑ cf. z. B. Foley 1986
- ↑ cf. Van Valin 2009: 10
- ↑ Van Valin & LaPolla 1997: 94
- ↑ cf. Van Valin & LaPolla 1997: 90
- ↑ Van Valin & LaPolla 1997: 90
- ↑ Lambrecht 1986: 84; see. Van Valin 1993: 23
- ↑ Lambrecht 1994: 213, cf. Van Valin & LaPolla 1997: 202
- ↑ K. Hengeveld et al. a. (2008). Functional Discourse Grammar. Oxford: 92
- ↑ Van Valin 1993: 23
- ↑ Van Valin & LaPolla 1997: 206-209
- ^ Van Valin 1993: 106
- ↑ Van Valin 1993: 110; Van Valin & LaPolla 1997: 455
- ↑ Van Valin & LaPolla 1997: 458f.461.463-67
- ↑ s. e.g. Van Valin 2007 and the detailed bibliography of the RRG from 2008