FG Expression Rules: The Dynamic Approach

Dik Bakker. University of Amsterdam.

Anna Siewierska. Lancaster University.

At the 1998 FG Conference, a number of problems were presented with the current form of FG expression rules (ERs). One of the major points made is that ERs fail to express adequately phenomena, frequently observed in the languages of the world, where form and order are inextricable connected. In this sense the current ERs undergenerate. Another, more theoretical problem is that ERs as they stand overgenenerate, i.e. they provide no constraints on what may and what may not be a well-formed expression in a language thus failing to reflect typological adequacy. In order to come to grips with this, Bakker (1999) sketches a proposal for the reorganization of ERs, which includes the following features:

a. the component that generates form and the one that deals with linear order are integrated rather than sequentially ordered

b. templates are organizad into constituent structures ("trees")

c. the expression of an underlying representation is implemented dynamically, in a top down, left to right, depth first fashion.

Currently, this proposal is under discussion in a local FG working group. Our paper, which assumes some familiarity with Bakker (1999), will present a short overview of the above proposal, including some refinements involving morphosyntactic constraints on expression. We will also present a discussion of a worked out example of a complex morphosyntactic phenomenon that is representative of the problems signalled above, and of the way they are countered in our approach. More specifically, it will be shown that a dynamic rather than a static approach to expression is necessary for explaining certain synchronic and diachronic formal aspects of language.

The paper will introduce an informal notational convention for ERs that is more compatible with the familiar notation for underlying representations than the feature-value notation as used in Bakker (1999).

Reference

Bakker, D. (1999). 'FG Expression Rules: from templates to constituent structure'.

Working Papers in Functional Grammar

Indirect Object in Spanish

Silvia Becerra Bascuñán. University of Copenhagen.

Ole Nedergaard Thomsen. University of Copenhagen.

FG conceives of grammatical relations as 'vantage points' in perspectivizations of SoAs. It restricts their existence in a given language to the existence of different marked voices (perspectivizations). Thus, a language only possesses Subject (primary vantage point) if it also has a passive voice, and only an Object (secondary vantage point) if it has Dative Alternation. FG only recognises these two vantage points, Subject and Object. Grammatical relations can he identified by their coding properties (case and adpositional marking; agreement; position) and behavioural properties (e.g. reflexive control). Spanish possesses Subject because it has a passive voice. Subject is primarily signalled by pronominal case (nominative), and inflectional agreement. FG also speaks about a "basic perspective" inherent in a predicate frame, running from the first argument, to the second, and on to the third. Here, the numbered arguments are "basic" vantage points, irrespective of the existence of marked perspectivizations. Now, a study by Silva-Corvalán (1984) demonstrates that coreferential pronominal clitics in accusative and dative are agreement markers of direct and indirect object. That is, Subject in Spanish is signalled by inflectional agreement, whereas secondary and tertiary arguments of basic perspectives are signalled by clitic agreement. Generalising agreement as a coding device for grammatical relations in Spanish, the present paper argues that the secondary and tertiary arguments of the basic perspective are automatically assigned (basic) grammatical relations, i.e. Direct and Indirect Object-although the language does not possess Dative Alternation. Spanish thus shows that the FG Object (in some languages) may have to be split up into two relations, and that the existence of (marked) voice is not the sole criterion for the establishment of a grammatical relation.

References

Silva-Corvalán, C. (1984): Semantic and pragmatic factors in syntactic change, in: J. Fisiak (de): Historical Syntax, Berlin etc.: Mouton, 555-573.

Paola Bentivoglio. Universidad Central de Venezuela.

Mercedes Sedano. Universidad Central de Venezuela.

The purpose of this study is to explore Chafe's hypothesis (1994) that grammatical subjects obey to the light subject contraint (LSC). The data consists of twelve half-an-hour semiformal interviews with Caracas native speakers recorded in 1987.

Chafe's hypothesis is based on the claim that in any given clause one of the referents receives the unique and special status of grammatical subject. Subjects show two types of restrictions: i) from the information load perspective, they tend to code given or accessible rather than new information; and ii) from a referential viewpoint, they can have three degrees of importance: primary, secondary or trivial (88). Information load (high or low) and referential importance combined constitute the LSC: Chafe's findings prove that in English subjects tend, on the one hand, to be given or accessible and, on the other, of primary or secondary importance. Heavy subjects, on the contrary, tend to be of trivial importance. Broderick 1999 confirms the LSC's existence in a conversational English corpus.

All subjects in the corpus have been coded according to information load and referential importance and submitted to a statistical analysis conducted by means of Goldvarb 2.0 (Rand & Sankoff 1990). The analysis shows that spoken Spanish also conforms to LSC. The differences All subjects found between Chafe's results and ours are small and do not impair the overall conclusion about the validity of the proposed constraint for both languages.

References

Broderick, John P. 1999. Wallace Chafe´s light subject constraint in conversational discourse in the inmediate mode of consciousness. Word 50,2. 143-154.

Chafe, Wallace. 1994. Discourse, Consciousness and Time. Chicago: The University of Chicago Press.

Rand, David and David Sankoff. 1990. GOLDVARB 2.0. Montréal: Université de Montréal.

Chris Butler. (Retired Professor of English).

Considerable advances in accounting for the richness of lexicogrammatical patterning have been made in recent years within the Functional Lexematic Model (see Martín Mingorance 1998, Faber and Mairal 1999). However, with some notable exceptions (see Márquez Linares 1998, Faber and Márquez Linares in press, on polysemy), the proponents of the FLM have made rather little use of the rich insights provided by corpus analysis.

In this paper, I will bring evidence from dictionaries, corpora and informant testing to bear on one particular issue: the tendency of many high and middle frequency verbs in English to have a wide range of uses, many of which involve collocation and colligation with items of varying degrees of specificity. The meanings involved are systematically related to one another, but also to other quite different lexemes. The challenge is to capture the semantic, syntactic and pragmatic aspects of these complex relationships within an integrated model which does full justice to the complexity of the evidence.

The discussion will centre on the verb CATCH. Having first looked at how this verb has been handled so far within the FLM framework, I will present a schematic network for the meanings of CATCH, derived from the definitions given in the Longman Dictionary of the English Language, the corpus-based Collins COBUILD English Dictionary and the Collins English Dictionary, and also from informant testing on the relative salience of the meanings. I will then examine data from the British National Corpus which reveals the quantitative importance of the cooccurrence of CATCH with expressions referring to perceptual phenomena (e.g. catch a glimpse/whiff, catch + the/possessive eye). An attempt will be made to assess the extent to which such systematic collocations, and their relationships with single lexemes such as glimpse, see, can he captured within the FLM.

References

Faber, P. and Mairal Usón, R. (1999) Constructing a Lexicon of English Verbs, Berlin and New York: Mouton de Gruyter.

Faber, P. and Márquez Linares, C. F. (in press) El tratamiento de la polisemia en el Modelo Lexemático-Funcional. Actas del V Simposio Internacional de Lingüística Hispánica. Leipzig, November 1998.

Márquez Linares, C. F. (1998) La polisemia en el campo léxico el cuerpo humano: un estudio contrastivo inglés-español. Unpublished PhD dissertation, Universidad de Córdoba.

Martín Mingorance, L. (1998) El Modelo Lexemático-Funcional. (Ed. A. Marín Rubiales). Granada: Universidad de Granada.

A Systemic Functional Interpretation of French Grammar

Dr. Alice Caffarel. University of Sydney.

This paper will provide an overview of the clause grammar of French within the framework of systemic functional theory (Halliday 1994, Matthiessen 1995). It will explore how the theory of metafunctions is realised paradigmatically and syntagmatically in a particular description of French. What is proposed is an overview of the most general clausal systems (TRANSITIVITY, MOOD and THEME systems) and of their realizations in structure. This paper aims at highlighting both the systems and features of French which are assumed to be general across languages and those systems and features which tend to be specific to the French language. To establish what type of subsystems may be common across languages and what type of subsystems are "Language specific" one has to draw the line between theoretical and descriptive categories. The theoretical categories give us the means to describe and talk about any language. They provide a framework for the description of other languages and multilingual descriptions. As such they are categories that allow us to make cross linguistic generalizations. Thus, the notion of metafunction in SF theory is a theoretical category. The metafunctional hypothesis predicts that all languages will have some kind of experiential, interpersonal and textual organisation that will tend to be realised by TRANSITIVITY, MOOD and THEME systems in the grammar of a particular language. The way these systems are systemically organised and functionally realised will however vary across languages. Differences between French and English systems and structures will be foregrounded as the discussion unfolds.

References

Halliday, M.A.K. 1994. An introduction to Functional Grammar. Second Edition. Edward Arnold: London.

Matthiessen, C. 1995. Lexicogrammatical Cartography: English Systems. International Language Science Publishers: Tokio.

Mood and Modality in Modern Greek: Sentence Type and lllocution

Maria Chondrogianni. University of Westminster.

Simon Courtenage. University of Westminster.

In this paper we continue our investigations (Chondrogianni 1995,1997) into the relationship between Mood and Modality in Modern Greek from a Functional Grammar perspective. The Indicative, the Subjunctive and the Imperative are examined as well as secondary, non-fully grammaticalised moods.

Here, we further explore the typology of sentence type as an illocutionary operator. Special emphasis is given to Subjunctive and Imperative, in order to examine whether semantic or pragmatic parameters allow for the recognition of the illocutionary force, as in

(1) "Fwnaze oso theleis" You can scream' as much as you want.

(2) "Pame ?" Should we go'?

(3) "Pame!" Let's go'.

(4) "Fygame" Let's go2 .

(5) "Na rwtisw kati ?" May I ask 3 something?

We also investigate whether the use of Imperative/Subjunctive creates a conflict with the Interrogative (examples (2) and (5) and Exclamative (example (3) nature of these clauses in the assignment of sentence type. Finally, we explore further our earlier observations regarding the boundaries and mutual exclusiveness of the directness expressed in Imperative and Subjunctive.

----------

'Main verb in Imperative

2Main verb in Indicative/ Exclamative sentence type

3Main verb in Subjunctive

The Question of Discourse Representation in Functional Grammar

Dr. John Connolly. Loughborough University.

During the last few years the FG literature has reflected a considerable level of interest in the treatment of discourse. Writings on this subject include, among others, Dik (1997: chapter 18), Steuten (1998) and several of the chapters in Connolly et al. (1997) and in Hannay and Bolkestein (1998). However, there is, as yet, no consensus on how discourse should be incorporated into FG, or on the notation to he used in the FG description of discourse.

Thus, we are faced with the issue of how best to represent discourse in FG. This question constitutes the focus of the present paper. The aim is to contribute to the development of an adequate means of discourse representation within the overall framework of FG.

The paper begins with a survey of previous work both in FG and in related areas, such as Discourse Representation Theory (see e.g. Kamp and Reyle, 1993) and Critical Discourse Analysis (see e.g. Fairclough and Wodak, 1997). It is clear from this overview that numerous phenomena occur in discourse which need somehow to be accommodated within an adequate DR. These include: hierarchical structure, linear structure, adjacency sequences and rhetorical relations. General properties of discourse, such as those associated with style, genre and ideology, also need to be taken into account. Furthermore, the interfaces between discourse and grammar and between discourse and context need to be articulated explicitly.

In the light of all this, the main part of the paper is devoted to presenting a fairly comprehensive, FG-oriented approach to DR which offers a broader coverage of the various discourse phenomena identified above than is found in any individual existing FG work, and is also user-friendly, i.e. readable and relatively quick to write down. The conclusion contains evaluative comments and pointers to future search.

References

Connolly, J.H., Vismans, R., Butler, C.S. and Gatward, R.A. (eds.) (1997). Discourse and Pragmatics in Functional Grammar. Berlin: Mouton de Gruyter.

Dik, S.C., ed. Hengeveld, K. (1997), The Theory of Functional Grammar. Berlin: Mouton de Gruyter.

Fairclough, N. and Wodak, R. (1 997). Critical Discourse Analysis. In van Dijk, T. (ed.), Discourse as Social Interaction. London: Sage, 258-284.

Hannay, M. and Bolkestein, A.M. (eds.) (1998). Functional Grammar and Verbal Interaction. Amsterdam: Benjamins.

Kamp, H. and Reyle, U. (1993). From Logic to Discourse: Introduction to Modeltheoretic Semantics of Natural Language, Formal Logic and Discourse Representation Theory. Dordrecht: Kluwer.

Steuten: A.A.G. (1 998). A Contribution to the Linguistic Analysis of Business Conversations within the Language/Action Perspective. Delft: Steuten,

"Focus of Attention" in discourse : a comparison between the four FG Topic functions and the systems of "Focus" and "Deixis" in the Columbia School of Linguistics

Dr. Francis Cornish. Université de Toulouse-Le Mirail

In Columbia School (CS) Linguistics, the system of "Focus"(1) has two values: for transitive predications, "IN-FOCUS" for the A l (in FG terms, i.e. the referent of the subject expression), and "NOT-IN-FOCUS" for the A 2 (Reid, 1991:178). The instruction conveyed to the understander is "pay more attention to the referent at issue" in the former case, and "pay somewhat less attention to the referent involved", in the second. In languages possessing case-marked clitic pronouns, nominative clitics signal "HIGHER-FOCUS", while non-nominative clitics signal "LOWER-FOCUS". Moreover, in inversion constructions, the value "MORE-FOCUS (needed)" is said to be encoded by "A-pred" ordering, whereas "LESS-FOCUS (needed)" attaches to "pred-A" ordering.

"Deixis" in CS terms, on the other hand, is said to involve "degree of insistent pointing toward the intended referent", and is claimed to he encoded by clitic pronouns of different types. Reflexive pronouns, as well as subject or nominative clitic pronouns, signal the lowest degree of Deixis (in this sense of the term), namely that very little effort need be expended by the addressee to retrieve the intended referent; and non-subject, non-reflexive clitics signal a high(er) level of Deixis.

However, this distinction between the subsystems of "Focus" and "Deixis" is not wholly convincing, since it fails to capture what is common to both in a revealing way. Significantly, the very same morphemes constituting signals of LOW DEIXIS are said to be signals of HIGH FOCUS (reflexive pronouns, nominative pronouns), and those which correspond to HIGH DEIXIS signals are also markers of LOW FOCUS (non-reflexive complement pronouns). I will argue that there is indeed a single system involved here, namely FOCUS OF ATTENTION, and that the difference between the two sets of signals (which partially overlap) posited within CS analysis has to do with their discourse function.

Now, the four Topic statuses recognized in FG ("Given", "Sub", "Resumed" and "New") appear to subsume the distinction in CS between "Focus" and "Deixis". In the light of several textual extracts in both French and English, I will re-examine this four-way distinction in terms of the CS "Focus"-"Deixis" system, and will pinpoint the nature of the discourse-level phenomena being captured by the two approaches, in order to determine the extent to which each theory can benefit from the other’s insights. In particular, can FG´s framing distinction within a clause between what pertains to "topicality" and what to pragmatic knowledge state, be reinterpreted in terms of the CS notion of a "semantic substance" which corresponds to "Focus of Attention"? I will argue that it can, and should -once certain revisions are made as to which signals encode" "HIGH(ER)" and which "LOW(ER) FOCUS", in CS analyses.

-----------

1 This is not to be confused with the sense of "Focus" in FG terms. "Focus" in CS terminology means "what the speaker is assuming the hearer is already concentrating on, what s/he has at the forefront of his/her consciousness at the time of utterance".

References

Birner, B.J., & Ward, G. (1998) Information Status and Noncanonical Word Order in English, Amsterdan/Philadelphia: John Benjamins.

Bolkestein, A.M. (1998) 'What to do with Topic and Focus? Evaluating pragmatic information', in M. Hannay & A.M. Bolkestein (eds.), Functional Grammar and Verbal Interaction, Amsterdam/Philadelphia: John Benjamins, 193-214.

Contini-Morava, E., & Goldberg, B.S. (eds.) Meaning as Explanation: advances in linguistic sign theory, Berlin : Mouton-de Gruyter.

Cornish, F. (1998a) 'The FG conception of (discourse) anaphora: a (sympathetic) critique', paper presented at the 8th International Conference on Functional Grammar, Vrije Universiteit, Amsterdam, The Netherlands, 9 July, 1998.

Cornish, F. (1998b) 'Les "chaines topicales" : leur role dans la gestion et la structuration du discours', Cahiers de Grammaire 23: 19-40.

Cornish, F. (1999) Anaphora, Discourse, and Understanding. Evidence from English and French, Clarendon Press: Oxford.

Davis, J. (1995) 'Italian pronouns and the virtue of relative meaninglessness', in E. Contini & B.S. Goldberg (eds.), 423-440.

Dik, S.C. (1997a) 'Pragmatic Functions', ch. 13 in The Theory of Functional Grammar Part I: The structure of the clause (2nd edition), Berlin/New York: Mouton-de Gruyter.

Dik, S.C. (1997b) 'Anaphora', ch. 10 in The Theory of Functional Grammar Part II: Complex and derived constructions (ed. K. Hengeveld), Berlin/New York: Mouton-de Gruyter.

Diver, W. (n.d.) 'Avoidance of the obvious : the pronoun as a minimax solution', course text for Linguistics G 6801.

Diver, W. (1995) 'Predictions from the hypotheses about the characteristies of the texts' (§ 3.4.4.l), and 'Measuring the strength of the correlation' (§ 3.4.4.2), extracts from 'Theory', in E. Contini-Morava & B. Goldberg (eds.), 34-5 and 35-7.

Huffman, A. (1993) 'Cognitive and semiotic modes of explanation in Functional Grammar', talk given at Rutgers University, 10 October, 1993.

Huffman, A. (to appear) Ch. 1 'The Control-Focus interlock', in Modern Englis : A Columbia School Grammar.

Lambrecht, K. (1994) Information Structure and Sentence Form: topic, focus, and the mental representations of discourse referents, Cambridge: Cambridge.

Reid, W. (1991) Verb and Noun Number in English: a functional explanation, London and New York: Longman.

Valduvi, E. (1992) The informational Component, New York: Garland.

Walker, M.A. Joshi, A., & Prince, E.F. (eds) (1998) Centering theory in Discourse, Oxford: Clarendon Press.

Zubin, D.A. (1979) "Discourse Function of morphology: the focus system in German", in T. Givón (de.), Syntax and Semantics Vol. 12: Discourse and syntax, New York: Academic Press, 469-504.

Simon Dik's Functional Grammar As The Linguistic Foundation For The Construction Of A Brazilian Portuguese Lexical Knowledge Base

Bento Carlos Dias da Silva. Universidade Estadual Paulista. Brazil.

Dias-da-Silva (1998) demonstrates that NLP (natural language processing) systems can be conceived of as a particular type of knowledge processing systems (Durkin, 1994) where the complex of linguistic and extra-linguistic knowledge is represented and applied electronically to exploit and to perform a number of linguistic as well as metalinguistic tasks such as "check" spelling and grammar, "analyse" morphological and syntactic structures, "understand" and "produce" texts, "translate" words, sentences and texts, "make" and "answer" questions, and "help" linguists develop their own linguistic models, just to mention the most impressive ones. In particular, such knowledge-based approach to NLP research program claims that the process of designing and implementing a computer environment that is capable of building and applying a lexical knowledge base (Saint-Dizier & Viegas, 1992; Copestake, 1992; Briscoe, Paiva, and Copestake, 1993) should comprise at least three iterative and evolutionary phases of analysis in three complementary domains: (i) Linguistic Domain - With the recourse to explicit models proposed within the linguistic theory, the task here is to gather a body of knowledge about the linguistic phenomena one is trying to model; (ii) Representational Domain - In this domain, the system conceptual design is pursued. It involves the selection and/or the proposal of computational tractable lexical representation languages whose degree of expressiveness is powerful enough to encode the body of lexical knowledge constructed in the previous phase; and (iii) Implementational Domain - Finally, the previous formal encoding is "translated" into a suitable programming language and the overall system planning is developed. Accordingly, this paper intends to explore the use of Functional Grammar concept of "extended lexicon", called the fund, with its predicate frames, terms, and the lexicon itself, to underpin the design of Linguistic Domain of the development of a Brazilian Portuguese Lexical Knowledge Base (Dik, 1978 and 1989; Neves, 1997 and 2000).

Bricoe, T, V. Paiva & A. Copestake (1993) Inheritance, defaults and the lexicon. Cambridge University Press, Cambridge.

Copestake, A. (1992). The representation of lexical semantic information. Ph. D. thesis, Sussex University. Cognitive Science. Research Paper CSRP 280.

Dias-da-Silva, B.C. (1998). Bridging the gap between linguistic theory and natural language processing. In: Proceedings of the 16th International Congress of Linguists. Pergamon, Oxford, Paper No. 0425.

Dik, C.S. (1978) Functional grammar. Foris Publications, Dordrecht.

Dik, C.S (1989) The theory offunctional grammar. Foris Publications, Dordrecht

Durkin, J (1994) Expert systems, design and development. Prentice Hall International, London.

Neves, M.H.M. (1997). A gramática funcional. Martins Fontes, Sáo Paulo.

Neves, M.H.M. (2000). Gramática de usos do portugués. in press.

Saint-Dizier, P. & Viegas, E. (1995) Computational Lexical Semantics, Cambridge University Press, Cambridge.

A Diachronic Approach To Optionality

Javier E. Díaz Vera. University of Castilla-La Mancha. Spain.

Within the Functional-Lexematic Model (F-LM; Faber & Mairal 1998), meaning definitions (Dik 1978) have been brought to the forefront in Functional Grammar, as far as semantic information is projected onto syntax. Accordingly, a verb's complementation pattern is not randomly selected, but is rather a systematic reflection of its semantic structure. This relationship can he mapped out by means of syntactic-semantic (synsem) parameters, which operate throughout the lexicon in the various areas of meaning and constitute a determining factor in the actual process of constructing an underlying clause structure (Faber and Mairal 1998).

Synsem parameters can he divided into three main types, according to their scope of application:

1 Lexically-realized grammatical parameters, which determine what complementation patterns a certain verb can accept.

2 Lexically-realized optional parameters, which explain why certain arguments, though semantically present, are not syntactically activated in the actual linguistic expression.

3 Lexically-realized contextual parameters, which are not syntactically projected, but serve as clues for contextual settings (Faber and Mairal 1998: 38).

Lexically-realized optional parameters explain why certain arguments, though semantically present, are not syntactically activated in the actual linguistic expression (Faber and Mairal 1998: 58). In this paper, I am going to analyse some theoretical implications of these parameters and their application to the reconstruction of linguistic change: broadly speaking, optionality is going to he considered as a relative term, so that arguments that used to be true became shadow arguments (or vice versa) due to the continued effects of change. In order to illustrate some of these changes, I will analyse some examples from the history of English, that can be explained as changes of optional parameters:

1. OE/ME loss of casual distinctions

2. ME loss of impersonal constructions.

3. ModE periphrastic to do.

References:

Dik, S. C. 1978. Stepwise Lexical Decomposition. Lisse: Peter de Ridder Press.

Faber, P. and R. Mairal. 1998. Towards a Semantic Syntax. Revista Canaria de Estudios Ingleses 36: 37-64.

Lexicalization Patterns And Naming Strategies. Position Verbs, Motion Verbs And Placement Verbs In Russian, Danish, And English.

Per Durst-Andersen, Copenhagen Business School, Denmark

If the concrete uses of position verbs in Russian, Danish and English are compared, a striking picture emerges. It appears that whereas English tends to use a pure existential verb instead of a specific position verb (e.g., The tree is (*stands) in the backyard and He is (*lies) in the hospital, Russian and Danish prefer a specific position verb, but not the same: while stojat´ "stand" is the natural choice in Russian, ligge "lie" is the natural choice in Danish. In Russian a town, a house, a shop, a mountain, a regiment, a problem, a prize, temperatures, etc. all STAND, while in Danish they all LIE. The only natural conclusion to be drawn is that what is vertical is unmarked in Russian, whereas what is horizontal is unmarked in Danish. This way of thinking is not confined to position verbs, but is also characteristic of so-called motion verbs like walk, lead, crawl, etc.) and placement verbs like put, sit down, lie down, etc. If we take Russian as an example, it appears that those motion verbs that entail a vertical position (e.g. idti/xodit´ "walk, go") are unmarked in relation to those entailing another position (e.g. polzti/polzat´ "crawl") and that those placement verbs that involve a vertical position (prijti/prixodit´ "come" and stavit´/postavit´ "put") are unmarked and therefore far more frequently used than verbs involving another position. There is almost a hundred percent correspondence between the subjects (X) that require stojat "stand", lezat´ "lie", sidet´ "sit" and viset´ "hang", and the objects (Y) that take stavit´/postavit´ "put (=X DO SMTH+Y STAND)", klast´/polozit "put (=X DO SMTH + Y LIE)", sazat´/posadit´ "put (=X DO SMTH + Y SIT)" and vesat´/povesit´ "hang up (=X DO SMTH + Y HANG)". This kind of symmetry is also observed in Danish (although to a lower degree), but is non-existing in English. I shall argue and present further evidence for the hypothesis that Russian takes its starting point in the figure (which will be vertical unless the object is permanently anchored in a ground) and then places it on a ground, whereas Danish takes its starting point in the ground (which is the earth - not water or the air) and then places this figure on the ground. In the former case, we have what could be called "from-figure-to-ground strategy", in the latter we have what could be named "from-ground-to-figure strategy". It will be demonstrated that these two strategies have several reflexes in their case system and their syntactic structures.

Foregrounding, Backgrounding, And The Linearity Of The Clause

Lisbeth Faister Jakobsen, University of Copenhagen.

Most discourse worlds consist of a string of sentences modelling events in the non-linguistic world. in the discourse world it is compulsory that the events are linearized, whereas the relations of non-linguistic events are far more differentiated. The linguistic concept of foregrounding and backgrounding has mainly been used as a way of distinguishing between the events of the main story line and events depicting collateral circumstances (e.g. Hopper). However, the same concept might be applied to the sentence. Contrary to the smaller information units of sentences, Chafe regards sentences as such as superfoci of consciousness, and says that "sentence boundaries appear when a speaker judges, during a particular telling, that a coherent centre has been verbalised at that point". A non-linguistic event consists of a set of simultaneous factors, an action with some actors (arguments), in a setting of LOC, TEMP, MOD, CAUS, etc.; roughly the set of factors discussed in Dik's Layered Model. In language, the factors are necessarily linearized, but grounding seems to be another, non-linear structure principle applied at the same time. Through the valency of the verb some factors are chosen as standing out, as foregrounded (the arguments), whereas the rest, the satellites, is the background, the setting. Mostly, the argument factors are human or animate beings or things, but valency can promote other factors, normally satellites, to arguments: In Chicago (satellite) ist es im Winter sehr kalt / Chicago (argument) ist im Winter sehr kalt.

The explanation of this otherwise superfluous choice of construction might be the chance of manipulating factors between being part of the foreground and setting. (German examples).

Reformulation Of The Domain-Level Semantic Pattern Of Axiological Evaluation In The Lexicon Of English Verbs

Ángel Felices Lago. Universidad de Granada. Spain.

Recent developments in the Functional Lexematic Model (FLM) provide an account of the relationship between lexical structure and cognition. One of the key issues has been the introduction of a cognitive axis and a typology of predicate schemas at different levels of the lexicon [lexeme, sub-domain and domain] (cf. Faber & Mairal, 1999). Among the four domain-level semantic patterns proposed, axiology is considered to appear in a wide range of domains and, consequently, if the domain structure is polarised, positiveness and negativity are structured in three levels (sensory, life and health, spiritual) in accordance with Tischner three-level hierarchy of values.

In this paper, it is claimed that the axiological parameter needs further clarification and decomposition. Its structure is multidimensional, internally hierarchical and canonical, as it is proved in the recent literature of axiological linguistics (Krzeszowski 1993, 1997; Pauwels & Simon-Vandenvergen 1993, 1995; Felices Lago 1997a, 1997b). In consequence, the three-level hierarchy of values in the lexicon of English verbs should be reformulated and largely expanded considering the following suggestions:

(a) The axiological feature (positive or negative) is a permanent component in the definitional structure of many lexical entries (at lexeme-level predicate schemas) and should be specified in a systematic way. A specific classification or taxonomy is needed.

(b) From a linguistic perspective, the axiological levels are not hierarchical according to the religious or ideological beliefs of philosophers or individuals (i.e. Tischner), even if their ideas are extremely well-presented. The only hierarchy that can be assumed for general purposes is built in language and depends, for its relevance (positive or negative), on what is perceived by the vast majority of speakers of a linguistic community as well as on the result of an exhaustive scrutiny of empirical data.

(c) The axiological parameter is divided into multilevel categories and crossed by two layers of canonical axes (cf. Felices Lago, 1997a). Consequently, the insertion of an axiological formula incorporating this might improve the understanding of this parameter within the lexical architecture of the verbal lexicon.

References

Faber, R. & Mairal Usón, R. (1999) Constructing a Lexicon of English Verbs. Berlin: Mouton de Gruyter.

Felices Lago, A. (1997a) "'The Integration of the Axiogical Classemo in an Adjectival Lexicon Based on Functional-Lexematic Principles", A Fund of Ideas: Recent Developments in Functional Grammar, C.S. Butler, J.H. Connolly, R.A. Gatward, R.M. Vismans (eds.). Amsterdam: IFOTT, pp. 95-112.

Felices Lago, A. (1997b) "La Compilación de un Diccionario de Términos Valorativos desde la Perspectiva Lexemático-Funcional", Toward a Funetional Lexicology/Hacia una lexicología funcional. Gerd Wotjak (ed). Frankfurt am Main: Peter Lang, pp. 159-184.

Krzeszowski, T.P. (1993) "T'he Axiological Parameter in Preconceptual Image Schemata", Conceptualizations and Mental Processing in Language. R.A. Geiger & B. Rudzka-Ostyn (eds.). Berlin: Mouton de Gruyter, pp. 307-330.

Krzeszowski, T. P. (1997) Angels and Devils in Hell. Elements of Axiology in Semantics. Warsaw: Energeia.

Pauwels, P.; Simon-Vandenhergen, A. (1993) "Value Judgment in the Metaphorization of Linguistic Action", Conceptualizations and Mental Processing in Language. R.A.Geiger & B.Rudzka-Ostin (eds.). Berlin: Mouton de Gruyter, pp. 331-367.

Pauwels, P. & Simon-Vandenbergen, A. (1995) "Body Parts in Linguistic Action: Underlying Schemata and Value Judgments", By Word of Mouth: Metaphor, Metonymy and Linguistic Action in a Cognitive Perspective. L.Goossens, P. Pauwels et alii (cds.). Ainsterdam: John Benjamins, pp. 35-70.

A Proposal for a Simplification of the FG Model at the Basic Level of the Predication

Michael Fortescue. University of Copenhagen

It can be argued that the complex FG bracketing formalism involving the interplay of predicate restrictors and term variables at the lowest level of clause structure has served its, historical purpose and that now, with cognitive rather than formal considerations coming to the fore, it would be timely to bring that part of the FG model more in line with the process-orientated higher levels of the model. The first-order predicate calculus aspect of the formalism can surely be jettisoned for the purpose of maintaining interest in the model amongst a wider audience of functionalist-cognitivists. One way of doing this would be directly to integrate Whiteheadian propositions (approximately= FG predications) at the lowest level, with the choice of what is to be conceptualised as predicate and what is logical subject (s) already made there. This would bring the model closer to Strawson´s point than the variables should be regarded as the basis of propositional meaning.

The proposal will be illustrated by sentences drawn from the Wakashan languages of western Canada, in which distinction between lexical nouns and verbs is notoriously slippery. In these languages the choice of predicate as opposed to subject often goes contrary to European expectations. What corresponds to an adverbial or a quantifier for instance might be chosen as predicate, while a nominalized verbal expression is chosen as subject. Rather than attempting to treat such sentence types as produced by higher level syntactic or pragmatic level choices (corresponding to clefts in typical European languages) it will be demonstrated that for languages of this type it is preferable to regard such choices as being made already at the deepest (psychologically meaningful) level. One advantage is the avoidance of overburdening the description of such languages with literally hundreds of term and predicate formation rules and the necessity of invoking such processes in order to describe the structure of perhaps a majority of sentences. My earlier proposal for representing "predicate satellites" in Kokuyon and other polysynthetic languages finds its natural place within such a revised formalism. The proposal opens up the possibility of introducing focality and presupposition in the deepest level of clause organisation in languages of this type, where the decision as to what to regard as predicate cannot he postponed to the level where focal operators normally apply, since one cannot assume the same kind of anchoring of lexical items in default predicate templates as in European languages.

References:

Fortescue, Michael. 1992. Aspect and superaspect in Koyukon: an application of the Funclional Grarnmar model to a polysynthetie language. In: Layered Structure and Reference in a Funclional Perspective, eds. M. Fortescue, P. Hardor & L. Kristoffersen. Atnsterdam- John Beniamins, 99-141

Daniel García Velasco (University of Oviedo)

Carmen Portero Muñoz (University of Córdoba)

It is generally assumed that arguments are those participants in a predication whose presence is required by a predicate in order for an expression to be meaningful. Adjuncts (or satellites in FG) are those elements whose presence in a linguistic expression is optional (i.e., not required by the predicate). Most grammatical theories formalise this fact by expressing the obligatoriness of arguments in the lexical entries of the relevant predicates.

Understood arguments pose a significant problem for linguistic description because they blur the argument/adjunct distinction: under the appropriate contextual or extralinguistic conditions arguments may be left unspecified as long as their absence does not lead to an uninterpretable structure. Even though most scholars are well aware of this problem, so far there have been few studies which attempt to make explicit the contextual or pragmatic conditions under which a given argument can be left unspecified or can be retrieved from the context, and researchers tend to express the potential optionality of a given argument in lexical entries.

In FG, understood arguments are treated as empty argument positions in predicate frames. Although this is an elegant solution for some cases, and, conforming to FG principle no elision of lexical material is posited, it is not at all clear that all understood arguments can be treated in this way. Some scholars (Fillmore 1986; Groefsema 1995) claim that there is more than one type of understood arguments, a fact which would force FG to extend its notational system to account for those cases. Furthermore, the standard of pragmatic adequacy demands from FG grammarians to make explicit the contextual factors which allow or forbid understood arguments. By examining real data, we shall try to shed some light on the discoursal conditions which permit the existence of understood arguments and, hopefully, we will show how the FG model should be extended to capture all relevant cases.

References:

Fillmore, C.J. 1986. Pragmatically controlled zero anaphora. BLS XII: 95-107.

Groefsema, M. 1995. Understood Arguments: A semantic/pragmatic approach. Lingua 96: 139-61.

Functional Grammar Representations as Descriptive Algorithms in Computerized Adaptative Texts

Margarita Goded Rambaud. Universidad Autónoma. Madrid.

Cornputerized adaptative tests are based on IRT (Item Response Theory) which is subsequently founded on the idea of item's interchangeability. Recent psychometric devolopments within the field of IRT include componential models which try to inform about the level of difficulty of each item due to the complex psychological processes involved in the item's solution. Computerised adaptative tests are the result of a link between recent computer developments and the psychometric models contributions involved in IRTS. Consequently a very detailed and accurate description of items components is needed to enhance items prediction capacities. One component refers to its linguistic representation. Another component has to do with level assignment which in turn is also related to linguistic complexity. Therefore the more accurate the descriptive algorithm the higher the item's interchangeability.

The purpose of this paper is to show how a functional-cognitive type of linguistic representation with its acknowledgement for semantic and contextual information contributes to a more refined item description in coniputerised adaptative tests.

The hypothesis upon which I am working is based on recent lexical semantics theories (Levin, 1991, Mairal 1999) which claim that the grammatical configuration of a sentence is affected by its lexico-semantic and cognitive configuration. As a result a detailed analysis of such configurations is presently being undertaken to help defining items´ descriptors and consequently to refine items´ prediction capacities. A selection of items taken from the CATEG (Computerized Adaptative Test of English Grammar), now under the pre-implementation stage, are analysed to show how such cognitive- functional representation allows for a highly accurate descriptive algorithm thus contributing to a more refined items interchange possibilities and items prediction capacities.

# CATEG is a computerized adaptative test of English grammar constructes as part of the research project: "Tests adaptativos informatizados: efectos de la revisión y restricciones en la selección de items" DGICYT PB 97-0049

References

Faber, P. and R. Mairal Usón (forthcoming) Constructing a Lexicon of English Verbs. Mouton de Gruyter.

Olea, J. V. Ponsoda, G. Prieto (1999) Tests informatizados. Fundamentos y aplicaciones. Madrid: Pirámide.

Levin, B. (1991) Lexical and Conceptual Semantics. Cambridge Ma: Blackwell

Muñiz, J. (1997a) Introducción a la teoría de Respuesta a los Ítems. Madrid: Pirámide.

Rethinking Predicational Be: From Copula Support To A Constructional Account

Francisco Gonzálvez García. Universidad de Almería. Spain.

This paper is basically concerned with revisiting the theoretical and descriptive adequacy of the Copula Support hypothesis advanced in Dik (1980, 1983, 1997 (1989)), Hengeveld (1990a,b) and Dik & Hengeveld (1991) to account for both the occurrence and the semantico-pragmatic import of the copula in non-finite complement clauses after verbs of cognition and volition, as in (1) below:

(1) (a) "I consider John (to be) a fool"

(b) "I want you (to be) standing by my side"

At a theoretical level, it is argued that the Copula Support rule is decidedly ad hoc in the constructions at hand here, since it relies on whether an (infinitival) predicate operator can be specified for the complement in the underlying structure. Moreover, it is also a-semantic in that it ignores the fact that the feasibility of introducing the copula cannot be aptly described independently from its semantic contribution to the syntactic frame and its semantico-pragmatic interaction with the meaning and form properties of the matrix verb, the intervening nominal and the XP complement.

At a descriptive level, it is shown that the predicability conditions at work here should be best handled in terms of a continuum, depending on whether the copula is (i) always obligatory, (ii) potentially obligatory, and (iii) always optional. The existence of such a continuum is illustrated with special focus on (a) the identifying character of the XP, (b) the clausal or non-clausal realization of the XP, (c) tense-aspect-voice marker, (d) occurrence of existential "there" in the postverbal NP slot, (e) occurrence in a reduced participial relative clause environment, (f) Tough movement, (g) extraposition of a clausal object clause, and (h) placing of a "by-agentive adjunct" between the matrix verb and the XP.

References

Dik, S.C. (1980), Studies in Functional Grammar. London: Academic Press.

Dik, S.C. (1983), "Auxiliary and copula be in a functional grammar of English', F. Heny & B. Richards (eds.), Linguistic Categories: Auxiliaries and Related Puzzles. (Vol. II), 121-143.

Dik, S.C. (1989), The Theory of Functional Grammar Vol. l.: The Structure of the Clause. Dordrecht: Foris. (1997 edition by K. Hengeveld. Berlin: Mouton de Gruyter).

Dik, S.C. & K. Hengeveld. (1991), "The hierarchical structure of the clause and the typology of perception-verb complements", Lingua 29, 231-259.

Goldberg, A.E. (1995), Constructions (A Construction Grammar Approach to Argument Structure). Chicago & London: The University of Chicago Press.

Gonzálvez García, F. (1997), "A modality view of predicate selection in small clauses", Texas Linguistic Forum 38, 103- 7

Gonzálvez García, F. (in press), "To be or not to be again: A constructional view". To appear in Proccedings of the 22nd AEDEAN International Conference. Lleida: Spain.

Goossens, L. (1992), "Graded predicationality, semanticization and "be" in a functional grammar of English", Belgian Journal of Linguistics, 7, 57-73.

Hengeveld, IC (1990a), "Semantic relations in non-verbal predication", Nuyts, J., A.M Bolkestein & C. Vet (eds), Layers and Levels of Representation in Language Theory: A Functional View. Amsterdam: Benjamins, 101-122.

Hengeveld, K (1990b), "Non-verbal predicability", van der Awera, J. & M. Kefer (eds.), Meaning and Grammar. Berlin: Mouton de Gruyter, 77-94.

Two English Perfects, Content Syntax, And Functional Grammar

Louis Goossens, University of Antwerp

This paper explores the implications of a content-syntax (Harder 1996: 376ff.) or compositional (Langacker 1991: 5.2.3) approach to two English perfects (Present-day English and late Old English). The focus is on the specific contribution of HAVE, resp. HABBAN, in relation to their lexical counterparts.

The empirical testing ground is provided by two samples of 500 random instances with HAVE (from the LOB corpus), resp. HABBAN (from AElfric via the Toronto Concordance). A detailed investigation reveals:

(1) an inverse relation between the "lexical" and the "perfect" uses (17. 8 % vs. 76. 8 % for HAVE', as opposed to 82.8% vs. 17.2% for late Old English);

(2) a significant difference in the make-up of lexical HAVE and HABBAN (76.4 % of HAVE is purely relational, with an A 1 which is best captured in FG by the semantic function Zero, and an A2 which is a Reference; for HABBAN, on the other hand, the A 1 is typically experiential, sometimes /+ Control/, and rather rarely purely relational);

(3) whereas there appear to be no restrictions in Present-day English on the nuclear predications with which perfect HAVE combines, its late Old English counterpart shows up a number of restrictions: the combining predication always involves a two-place (though not necessarily a fully transitive) predicate, and the first argument of the perfect combination is as good as always animate and experientially involved in the perfect state of affairs.

It will be argued that these observations are mutually connected and that a compositional approach is considerably enriched if we take the overall make-up of HABBAN/HAVE into account. Also some of the consequences for the treatment of the perfect in Functional Grammar will be considered.

----------

1 The remainder of the Present-day uses is constituted by instances of HAVE TO, which is lacking in the Old English sample.

References

Harder, Peter (1996) Functional Semantics. A Theory of Meaning, Structure and Tense in English. Moton de Gruyter.

Langacker, Ronald W. (1991) Foundations of Cognitive Grammar. Volume II: Descriptive Application. Stanford University Press.

On The Typology Of Interrogatives And Their Operators

Casper de Groot. University of Amsterdam

This paper discusses the approach to interrogative clauses as set out in chapter 12 of the second volume of Dik (1997). Dik distinguishes between:

(1) a. Yes-No questions "Will John come to Madrid?"

b. Disjunctive questions "Do you present a paper or a poster?"

c. Q-word questions "What is the title of your talk?"

In his account, Dik represents the following general structure of questions:

(2) Int E: X: [Extended Predication]

where the illocutionary operator Int is pragmatically interpreted along the lines of:

(3) S wishes A to provide him with a true proposition X with a content as specified in the Extended Predication.

Dik, however, only discusses type (1c),the Q-word questions. For the analysis he introduces the Q operator in addition to the Int operator. Q indicates which part os the predication is being questioned. The Q-word questions take the following schema, where, in fact, the Int operator is redundant:

(4) Int E: X: ( ... Qxi ...)

If (2) were the appropriate analysis for Yes-No questions and Disjunctive questions and (4) the appropriate analysis for Q-word questions, the two analyses suggest that Q-word questions would widely differ from the other two. Moreover, how does (2) account for Yes-No questions and Disjunctive questions and their differences?

The paper will argue that a separation should be made between the illocution of questioning (accounted for by the operator Int) and the contents of what is questioned (accounted for by different representations of the Extended Predication). The paper will present typological data about Yes-No questions and their expressions in relation to the expression of Focus in the various languages.

References

Dik, S.C. (1997) The Theory of Functional Grammar. Part 2: Complex and derived constructions. Berlin: Moton de Gruyter.

Reconsidering Aspectuality: Interrelations Between Grammatical And Lexical Aspect

Pilar Guerrero Medina. University Of Córdoba

As pointed out by Comrie (1976), there is no generally accepted terminology in treatments of aspect. On the one hand, different authors apply different labels to the same concept; on the other hand, the same term is often used to refer to different concepts.

In the first part of the present paper, I will try to clarify the terminological confusion concerning aspectual distinctions in the linguistic literature, where the pair of terms "telic"/"atelic", "bounded"/"unbounded" are interchangeably used on many occasions. In discussing aspect, it is necessary to make a first distinction between "Aktionsart" (or "mode of action") and "Aspect". Following Dik (1997), "I will reserve the term "Aspect" for the aspectual distinctions which are expressed by grammatical means. For those distinctions which concern the internal semantics of the predication I will use the term "Aktionsart", which Dik regards as synonymous with "type of state of affairs". It is precisely the classification of states of affairs which determines the general approach to aspect followed in FG, specifying the nature of the interaction between lexical and grammatical aspect (cf Siewierska 1991).

Drawing upon Dahl (1981) and Depractere (1995), I contend that it is necessary to draw a further distinction between (a)telicity and (un)boundedness: telic events are those that possess an inherent terminal point, but they are not necessaryly presented as bounded.

Telicity and atelicity will here be considered as semantic parameters included in the "Aktionsar" category, while boundedness and unboundedness will be subsumed under the grammatical category of "Aspect"

Finally, in the second part of this contribution, I will argue that it is only in a given clause pattern that aspectual distinctions manifest themselves (cf. Durst-Andersen and Herslund 1996). The argument will be illustrated using specific material from the English language.

References

COMRIE, BERNARD (1976) Aspect. Cambridge: Cambridge University Press.

DAHL, ÖSTEN (1981) "On the definition of the Telic-Atelic (Bounded-Nonbounded) distinction'. In Philip Tedeschi and Annie Zaenen (eds), 79-90.

DEPRAETERE, ILSE (1995) "On the necessity of distinguishing between (un)boundedness and (a)telicity". Linguistics and Philosophy 18. 1-19.

DIK, SIMON C. (1997) The Theory of Functional Grammar. Part 1: ' The Structure of the Clause. 2nd ed. Berlin/New York: Mouton de Gruyter.

DURST-ANDERSEN, PER & MICHAEL HERSLUND (1996) "The syntax of Danish verbs. Lexical and syntactic transitivity".In Elisabeth Engberg-Pedersen et al. (eds), 65- 102.

ENGBERG-PEDERSEN, ELISABETH el al. (eds) (1996) Content, Expression and Structure. Studies in Danish Functional Grammar. Studies in Language Companion Series. Amsterdam: John Benjamins.

SIEWIERSKA, ANNA (1991) Functional Grammar. London:Routledge. Tedeschi, Philip & Annie Zaenen (eds) (1981) Syntax and Semantics. Tense and Aspect, New York: Academic Press.

Grammar, Function And Online Processing: Iconicity And Structure

Peter Harder, English Department, University of Copenhagen.

For functional grammarians, the question of the exact relationship between grammar on the one hand and cognitive and discourse function on the other is the most existential issue on the agenda. Over the last decade, work based on Simon Dik's theory of Functional Grammar has given rise to several different approaches, recently exemplified and discussed in Hannay & Bolkestein (1998). Even more recently, the question of how to relate the grammatical model to issues of online language production have been raised, cf. Bakker (1999) and Mackenzie (fc). When these three different types of phenomena are brought together, as they should be in a functional perspective, it becomes necessary to face the issue of the exact relationship between the regularities that are brought into contact. A functionalist will expect there to be a significant amount of iconicity between the hierarchical structures that are found in discourse, in grammar, and in processing procedures. But it remains a point of contention how far one should expect the iconicity to extend, and how to cope with the type of phenomena where there are divergencies.

In the paper I will discuss some linguistic phenomena that illustrate the questions that arise at the interfaces, summed up in terms of the slogan 'iconicity and structure' that the Copenhagen functionalists use as the title of a running umbrella project. With respect to the discourse/grammar interface I take up a discussion by Vet in Hannay & Bolkestein 1998 on the relation between speech act, utterance and grammatical levels: while agreeing with his central point I argue that his (referential) approach to meaning leaves aspects of the problem unresolved. Mackenzie (1998 and fc) has proposed a characterisation of Pl that means an almost complete identification between 'discourse' firstness and 'grammatical' firstness that also invites an identification between discourse structure and the structure of the online processing procedure. While carefully pointing out some necessary qualifications, Mackenzie suggests that an overall need for redundancy may be the only functional explanation for such divergencies. As against that, I shall argue that the collaboration-cum-tension between discourse-level relations and grammatical relations in the incremental perspective is necessary to the kind of semantic complexity that we take to be characteristic of human languages.

Central in the understanding of grammar in relation to discourse and processing that I propose is the status of grammar as a socially sanctioned norm for how to relate utterance fragments into wholes. This means that it constrains both function and processing in a way that is specified in advance of any actual utterance event, and so grammatical relations are in principle different from relations in ongoing events, whether described as discourse acts or acts of cognitive processing.

References

Bakker, Dik (1999) FG Expression Rules: from templates to constituent structure. WPFG 67, Universiteit van Amsterdam.

Hannay, Mike and A. Machtelt Bolkstein (eds) (1998) Functional Grammar and Verbal Interaction. (Studies in Language Companion Series 44).

Mackenzie, Lachlan fc. First Things First. Towards an Incremental Functional Grammar. To appear in Acta Linguistica Hafniensia.

Subjects In A Functional Grammar -A Danish Perspective

Lars Heitoft. Roskilde University

In Dik's functional grammar, subjects are assumed to perspectivize states-of-affairs: "Subj. assignment allows for altenative specifications of the "perspective", the "vantage point" from which the SoA is to be presented", as in:

(1) John earned the money

(2) The money was earned by John

Dik´s view is a version of the categorical subject (Sasse 1987). In my view, this is not a universal notion of subjecthood, but rather a reflection of normal subject grammaticalization in European SVO languages, as "the entity predicated about". I shall assume that subjecthood is grammaticalized from two main semantic domains: semantic roles and predicational structure. Examples of the former are agentive and ergative subjects, and the unmarked nominative subjects of older Indo-European languages. In Danish, subjects have been reinterpreted from the latter type to the categorical type.

Criteria for subject structure will include dependency tests as a robust tool for setting up constituent structure. At the content end of the criteria, subject type and focus structure will be claimed to he intimately related..

Old Danish nominative subjects were sensitive to iconic information structure rules. Like adverbials and other NPs, they acquired textual status through position:

(3) AEn vm thaet wil dylia arfui hans. at hanum dothum.

but if this will contest heir-NOM his. at him dead.

(4) Warthcer thrceli hand af hoggen bote..

If is thrall-DAT hand-NOM off cut.. pay as compensation..

'if a thrall has his hand cut off (not a foot, nor his nose.. etc.), he must pay as

compensation..,)

Being in the zone Negation X V marks any constituent X as the iconic focus, categories preceding negation are sheltered from its scope. Subjects are not per se bound up with focus structure.

In modern Danish, subjects and objects are no longer accessible to iconic focus. Iconic focus has been replaced by open focus. The VP is the focus domain, within which actual focus will normally fall. Objects are readily accessible to focus, subjects are not.

References

Dik, Simon C. 1989. The Theory of Functional Grammar. Part 1. The Structure of the Clause. Dordrecht: Foris.

Dik, Simon C. 1997. The Theory of Functional Grammar. Parts I & II. Ed. by Kees Hengeveld. Berlin: Mouton de Gruyter.

Engberg-Pedersen, Elisabeth 1996. Iconicity and arbitrariness. Engberg-Pedersen et al., 453-468.

Engherg-Pedersen, Elisabeth et al. 1996. Content, Expression and Structure. Studies in Danish Functional Grammar. Amsterdam: John Benjamins, 469-494.

Faarlund, Jan Terje 1990. Syntactic Change. Toward a Theory of Historical Syntax. Trends in Linguistics. Studies and Monographs 50. Berlin and New York: Mouton de Gruyter.

Heltoft, Lars 1996. Topics in Danish and in universal pragmatics. Lars Heltoft and Hartmut Haberland eds., Proceedings of the 13th Scandinavian Conference of Linguistics. Roskilde, Department of Languages and Culture, 251-262.

Hengeveid, Kees 1998. Review of Engberg-Pedersen et.al. Acta Linguistica Hafniensia 30, 206-219.

The Architecture Of A Functional Discourse Grammar

Kees Mengeveid, University of Amsterdam

As noted in Hannay & Bolkestein (1998), two approaches have been taken in developing a grammar of discourse on the basis of FG: an approach involving 'upward layering', as defended in Hengeveld (1997), and a modular approach involving separate modules for the grammatical and pragmatic domains, as defended in Kroon (1997). In this paper I claim that the development of a grammar of discourse on the basis of FG requires that both approaches be combined. The resulting Functional Discourse Grammar (FDG) has the following properties:

1. FDG distinguishes three levels (modules): the pragmatie level, the semantic level, and the morphosyntactic level, in that hierarchical order.

2. Upward layering applies at each level separately (cf. Gómez Soliño 1996), i.e. each level is organizad hierarchically.

3. Complex interfaces determine the choices that can he made from a layer at one level to a layer at the next lower level: assignment rules link the pragmatic to the semantic level; expression rules link the semantic to the morphosyntactic level.

4. Once a choice has heen made for a layer at the next lower level, only that layer and the ones included within it are relevant.

5. Each level has its own set of functions: pragmatic, semantic and syntactic functions, respectively.

6. T'he pragmatic and morphosyntactic level are always present. The semantic level is absent in those cases in which the pragmatic level is mapped directly onto the morphosyntactic one.

The development of FG into FDG has far-reaching concequences for the architecture of current FG. Apart from the arrangement of the levels and layers and the content of the interfaces, this approach involves the conversion of FG from a bottom up grammar to a top down grammar, starting from intentions running all the way down to articulations (Levert 1989).

References

Connolly, John H., Roel M. Vismans, Christopher S. Butler & Richard A. Gatward cds. (1 997), Discourse and pragmatics in Functional Grammar. Berlin: Mouton de Gruyter.

Gómez Soliño, José S. (1 996), La organización jerárquica de los textos desde una perspectiva textual. In: M. Teresa Caneda Cabrera & Javier Pérez Guerra eds., Os estudios ingleses no contexto das novas tendencias. Vigo: Universidade de Vigo, 43-68.

Hannay, Mike & A. Machtelt Bolkestein (1998), Introduction. In: Mike Hannay & A. Machtelt Bolkestein eds., Functional Grammar and verbal interaction. Amsterdam: John Benjamins, vii-xii.

Hengeven Kees (1997), Cohesion in Functional Grammar. In: Connolly et al. eds.1-16.

Kroon, Caroline (1997), Díscourse markers, discourse structure and Functional Grammar. In: Connolly et al. eds., 17-32.

Levelt, Willem J.M. (1989), Speaking.- From intention to articulation. Cambridge: MIT Press.

Left-Dislocation And Voice: Pragmatic And Syntactic Coding In Contemporary Spoken English And Spanish

Raquel Hidalgo. Universidad Complutense de Madrid. Spain.

Typologic functional approaches have described left-dislocation as a topic-marking constructor which exhibits certain functional and formal properties (Givón 1992, Dik 1989, 1997, Lambrecht 1981, 1994). In the Literature, it is normally assumed that left-dislocation is universal, in the sense that all languages seem to have identical or similar constructions, presumably to cover parallel functions. Although the universality of the Theme constituent is discussed in Dik (1989, 1997), and cross-language considerations in the marking of topic are presented in Gundel (1988, 1987), neither of them really get into the details of a comparative examination. In this paper, I will address some of these questions in a comparative study between spoken English and Spanish, based on the examination of a sample of data from Svartvik and Quirk's (1980) corpus of English Conversation, and from Marcos Marin's (1992) Corpus oral de referencia del español contemporáneo. The data show that similarities between the use of topicalizing constructions in the two languages are only superficial. On the one hand, spoken Spanish exhibits a higher frequency and greater variety of topic constructions than spoken English. For instance, there are certain types of topic constructions, such as pronominal and non-anaphoric detachments, which are common in Spanish but are rarely found in English. Very often, such constructions are pragmatically but not syntactically linked to the predication and serve functions which have to do with the speaker's perspective or subjectivity on discourse topics (Morris 1998). The data also reveal that in similar contexts English grammar is syntactically oriented, and chooses alternative subject-predicate constructions, where the passive is a particularly prominent device. An observation which follows is that left-dislocation and passive share certain functional properties; the data show that the two constructions present similarities in the way they rearrange topicality and semantic-case assignment relationships. However, they are also different in the way such relationships are coded, either by favouring a pragmatic or a syntactic coding (Givón 1979).

References

DIK, S. 1989. The Theory of Functional Grammar. Dordrecht: Foris.

DIK, S. 1997. The Theory of Functional Grammar. Part II. Dordrecht: Foris.

GIVÓN, T. 1979. "From discourse to syntax: grammar as a processing strategy". In T.Givón (ed), Syntax and Semantics. Discourse and Syntax. Vol. 12. New York: Academic Press pp.335-384.

GUNDEL, J. 1988. "Universals of topic-comment structures". In M.Hammond et. al., Studies in Syntactic Typology,. Amsterdam, John Benjanúns, 209-239.

GUNDEL, J. 1987. "Shared knowledge and topicality". Journal of Pragmatics 9, 1:83-107.

LAMBRECHT, K. 1981. Topic, antitopic and verb agreement in non-standard French. Amsterdam: John Benjamins.

LAMBRECHT, K. 1994. Information structure and sentence form. Topic, focus, and the representation of mental referents in discourse. Cambridge: Cambridge University Press.

MARCOS MARÍN, F. 1992. Corpus oral de referencia del español contemporáneo. Madrid: Universidad Autónoma de Madrid.

MORRIS, T. 1998. "Topicity s. thematicity: Topic-prominence in impromptu Spanish discourse". Journal of Pragmatics 29, 193 -203.

Brazilian Portuguese Conditional Constructions in the Layered Structure of Functional Grammar

Flavia Bezerra de Menezes Hirata. Universidade Federal de Goiás (UFG). Brasil.

The aim of this paper is to establish a classification of written Brazilian Portuguese conditional clauses according to Functional Grammar(FG) typologies of layers and satellites. In FG, a sentence is represented as a layered structure. The highest layer is the clause layer. The following layers are the proposition and the predication layers and the last one is the predicate and terms layer. In each of these layers one can introduce operators or satellites, grammatical and lexical means, respectively, which function in the different levels, and modify or qualify these levels until they reach the effective expression, the speech act. According to Dik (1989) and Hengeveld (1989) there are three types of satellites: predicate, propositional and illocutionary. Conditional constructions are considered satellites in FG as they are adverbial constituents which give additional information about the layers they are attached to. In this sense, we are going to analyse a corpus composed by 346 occurrences of conditional constructions in written Brazilian Portuguese, taking into account studies made by Dik et al (1990), Dik (1990), Sweetser (1990) and Wakker (1992, 1996) which focus the satellite function performed by conditionals.

References

DIK, S. C. Theory of functional grammar. Dordrecht: Foris Publications, 1989.

DIK, S. C. On the semantics of conditionals. In: NUYTS, J., BOLKESTEIN, AM.,

VET, C. (Eds.) Layers and levels of representation in language theory: a functional view. Amsterdam/Philadelphia: John Benjamins, p. 233-261, 1990.

DIK, S. C. et al. T'he hierarchical structure of the clause and the typology of abverbial satellites. In: NUYTS, J., BOLKESTEIN, A.M., VET, C. (Eds.) Layers and levels of representation in language theory: a funclional view. Amsterdam/Philadelphia: John Benjamins, p. 25-70, 1990.

HENGEVELD, K. Layers and operators in functional grammar. Journal of Linguistics, v. 25, p. 127-157, 1989.

SWEETSER. E. E. From etymology to pragmatics: metaphorical and cultural aspects of semantic structure. (Cambridge Studies in Linguistics, 54). Cambridge: Cambridge University Press, 1990.

WAKKER, G. Conditionals in the layered structure of functional grammar. In: FORTESCUE, M., HARDER, P., KRISTOFFERSEN, L. (Eds.) Layerered structure and reference in a functional grammar perspective. Amsterdam: John Benjamins, p. 369-386, 1992.

WAKKER, G. Conditional at different levels of the clause. In: DEVRIENT, B., GOOSSENS, L., AUWERA, J. (Eds.). Complex Structures: a finctionalist perspective (Functional Grammar Series 17). Berlin/New York: Mouton de Gruyter, 1996.

Cognitive Linguistic Aspects in Functional Grammar. The case of Cleft and Pseudocleft Constructions

Roman Kalisz, University of Gdansk. Poland.

In the final chapters of the second volume of Dik 1997 it is possible to note notions and tools which are also employed in cognitive linguistics i.e. frame, iconicity (Dik 1997 II: 334, 335 ) and most importantly and extensively prototype. In his analysis of cleft and pseudocleft constructions Dik presents a prototypical approach based on matching of properties devised for a prototypical cleft/pseudocleft construction. He claims that pseudocleft constructions are more prototypical than cleft constructions since the former more faithfully represent the underlying structure for both clefts and pseudoclefts. Prototypicality is also considered with respect to pragmatic functions of those constructions.

Polish pseudoclefts differ from pseudocleft constructions in English or German since they reduplicate pronouns not the verb be:

Co jest jasne to, to ze Piotr nie ma wyboru.

'What is clear it it that Peter not has choice'.

What is clear is that Peter has no choice.

In spite of the peculiarity of Polish pseudocleft constructions, it is more obvious in Polish that pseudoclefts are more prototypical since Polish clefts (or dummy clefts in Dik's terminology ) are on the verge of ungrammaticality. This rather neatly corroborates Dik's ( 1997 11: 252 ) stance.

Prototypicality as an operating notion appears very late in Dík's 1997 work. If it applied to earlier levels e.g. instead of meaning postulates or stepwise lexical definitions, it could have serious consequences upon the overall theory of FG, not necessarily negative. This possibility is considered in the paper, especially on the basis of analysis of cleft and pseudocleft constructions in Polish and English at all levels. When prototypicality is accepted in FG to a slightly wider extent then cognitive linguistics and functional grammar analyses are quite compatible in spite of the absence of underlying structure in CL and equally revealing.

References

Dik, Simon C. 1997. The Theory of Functional Grammar. Parts I and II. Edited by Kees Hengeveld. Berlin: Mouton de Gruyter.

REMARKS ON HOW An Analysis Which Is Better Than The One Proposed BECOMES A Better Analysis Than The One Proposed

Alex Klinge Copenhagen Business School

Syntactic descriptions of noun phrase structure tend to be polarised into two tradition, viz. the tradition of descriptive grammar and the tradition of generative grammar. Traditional descriptive grammar aims to record correlations of forms and distribution in a simple head- dependent NP structure. Seeking comprehensiveness through lists of correlations, descriptive grammars attain substantial adequacy in the range of data treated. New data may simply be added to the list. However, comprehensiveness is achieved at the expense of sophistication of theory and economy of description. If the grammar does not postulate basic governing principles, no generalisation is possible. New observations have to be included ad hoc. Recent exponents of the descriptive approach to NPs are Quirk et al. (1985), Bache and Davidsen-Nielsen (1997), and Flämig (1991).

Conversely, in generative grammar careful crafting of precise generalisable hypotheses is a primary concern.

Comprehensiveness of data is only a secondary concern. Data not currently amenable to the state of the theory will have to await their turn. The result is that generative grammar strives for accuracy and generalisable hypotheses at the cost of comprehensiveness. And indeed, in places, accuracy has forbidding descriptive complexity as a side effect (a recent generative approach comparable to those above is found in Haegeman and Guéron 1999).

It is my intention to probe into the middle ground, occupied by functional grammar. By appealing to general principles of syntactic organisation and their interface with semantic structure, I will seek to identify the motivation underlying the structure of Germanic NPs. By developing the postulate that NP-structure is motivated by basic syntactic and semantic principles, it is possible to do away with the constraints of the autonomous syntactic rules of generative grammar. The principles may be formulated with sufficient precision to allow generalisation while a concurrent goal of comprehensiveness is pursued. At this mediating level of description it is possible to make observations which are obscured in either of the two above-mentioned prevailing traditions and to bring together apparently disparate data, such as the nature of the syntactic side of the kinship holding between an analysis which is better than the one proposed and a better analysis than the one proposed. Rijkhoff (1998) is an excellent recent exponent of a functional approach to the syntactic structure of Nps.

Discourse Type And Its Implications For A Modular Functional Grammar Of Discourse

Caroline Kroon Vrije Universiteit Amsterdam

It is a well-recognised fact, also within FG literature, that discourse type (or, with another term, text type) has an effect on the internal structure of clauses. Cf. e.g. Dik (1997: 417): "The choice of a genre (Dik's term for 'discourse type', CK) has important implications for the way it can be linguistically construed, both globally (at the level of macrostructure), and locally, at the level of the microstructure of individual clauses or even lower-level constituents". It is for this reason that Hengeveld (1997), in his proposal for an FG model of discourse structure, introduces the notion 'discourse type' at the highest, rhetorical level of this model. A slightly different proposal is Moutaouakil (1998).

The accounts by Hengeveld and Moutaouakil have in common that they approach the issue from an 'upward layering' point of view, i.e. an approach whereby the current layered clause structure is expanded by means of one or more discourse levels. In this paper I will argue that discourse type (or text type) is a quite complex phenomenon and that the accounts given by Hengeveld and Moutaouakil are in fact an oversimplification of what we actually find in discourse. I will argue that in order to account for discourse type (and the effects of the various types on the level of the clause structure) in a more appropriate way, a 'modular' approach (i.e. an FG discourse model which encompasses various interrelated modules, including a grammatical module) might be more promising. For the debate referred to here, cf. the introduction in Hannay and Bolkestein (1998).

The paper indirectly draws on the investigation of various narrative text types in Latin, but is mainly theoretically oriented. It incorporates insights from other theories than FG, especially from the Geneva discourse model (see e.g. Roulet 1997).

References

Dik, Simon C. (1997) The theory of Functional Grammar. Part 2: Complex and derived constructions. Berlin/New York: Mouton de Gruyter

Hannay, Mike and A. Machtelt Bolkestein (1998) Functional Grammar and verbal interaction. Amsterdam: Benjamins.

Hengeveld, Kees (1997) Cohesion in Functional Grammar. In: Connolly, J.H. et al. Discourse and pragmatics in Functional Grammar. Berlin/New York: Mouton de Gruyter, p. 1-16.

Moutaouakil, Ahmed (1998) Benveniste's "Récit" and "Discours" as discourse operators in Functional Grammar. In: Hannay and Bolkestein (eds.).

Roulet, Eddy (1997) A modular approach to discourse structure. Pragmatics 7.125-146.

Sentencing The Sentence

Kwee Tjoe Liong. Universiteit van Amsterdam. The Netherlands.

Study of language has long been identified with grammar, grammar with syntax, and syntax with sentence structure. FG aims at a wider perspective, encompassing language use, and giving pride of place to semantics and pragmatics. So far, however, its fundamental unit of description is the complete sentence, be it under the name of utterance and underlying clause, as the top level of a multi-layered structure.

Dik's (1997) final chapter, "Towards a functional grammar of discourse", is a modest first attempt to expand the actual sentence grammar into a discourse grammar. Ideas were already around for some time and lively debates will continue for quite a few years. As the sequel of Connolly et al. (1997), Hannay-Bolkestein (1998) distinguish two approaches: upward-layering versus modularity.

The roots may be traced back to the treatment of complex structures (Devriendt et al. 1996). Kwee (1999), sowing that the multi-layered model encounters scope problems when a sentence contains two or more clausal satellites of the same order, regrets that insights of RST (Thompson-Mann 1987, Matthiessen-Thompson 1988), while being discussed by Dik (1997), are not incorporated into FG. Earlier, Kroon (1995, 1997, 1998) already approvingly refers to RST.

The discussion of holophrases (Mackenzie 1998) and problems with variables in underlying structures, especially the e-variable for SoAs (Wanders 1999), also contribute to the suggestion of a dichotomy between a micro-grammar for single (possibly layered) units, and a macro-grammar. for combinations. Units need not be complete clauses - terms, for instance, are allowed as well. Single units are not the same as simple structures- subordination of complement clauses as arguments wouldn't lead to complex structures, clause combining will. The two modules essentially differ in that micro-units are expressed in rather standard patterns, but a large choice is offered for the realisation of macro-structures.

References

Connolly, John H., Roel M. Vismans, Christopher S. Butler, and Richard A. Gatward (eds) 1997. Discourse and Pragmatics in Functional Grammar. Berlin : Mouton de Gruyter.

Devriendt, Betty, Louis Goossens, and Johan van der Auwera (eds) 1996. Complex structures : a functionalist perspective. Berlin Mouton de Gruyter.

Dik, Simon C. 1997. The Theory of Functional Grammar. Ed. by Kees Hengeveld. Part 1, The Structure of the Clause. 2nd rev. ed. Part 2, Complex and Derived Constructions. Berlin : Mouton de Gruyter.

Hannay, Mike, and A. Machtelt Bolkestein (eds) 1998. Functional Grammar and Verbal Interaction. Amsterdam: Benjamins.

Kroon, Caroline 1995. Discourse particles in Latin : a study of nam, enim, autem, vero and at. Amsterdam: Gieben.

Kroon, Caroline 1997. Discourse markers, discourse structure and Functional Grammar. In Connolly et al (eds), 17-32.

Kroon, Caroline 1998. Acts Unlimited ? Towards a Functional Discourse Grammar. Paper read at 8th International Conference on Functional Grammar, Vrije Universiteit Amsterdam, July 6-9, 1998.

Kwee Tjoe Liong 1999. Adverbial clauses, Functional Grammar, and the change from sentence grammar to discourse-text grammar. Working papers in Functional Grammar 70. Amsterdam : IFOTT.

Mackenzie, J. Lachlan 1998. Tle basis of syntax in the holophrase. In Hannay-Bolkestein (eds), 267-295.

Matthiessen, Christian, and Sandra A. Thompson 1988. The Structure of Discourse and 'Subordination'. In John Haiman, Sandra A. Thompson (eds), Clause combining in grammar and discourse. Amsterdam: Benjamins. 275-329.

Thompson, Sandra A., and William C. Mann 1987. Rhetorical Structure Theory : A Framework for the Analysis of Texts. IPRA Papers in Pragmatics 1: 1, 79-105.

Wanders, Gerry 1999. Variables and entities. Talk given at Seminar on Selected Topics in Functional Grammar, Universiteit van Amsterdam, Spring term 1999.

Forms And Functions In The Grammar Of Spoken English

J.Lachlan Mackenzie

According to corpus research (Biber et al. 1999: 1071), 38.6% of units in British and American spoken English are non-clausal. Whereas, with some interesting exceptions (Barton 1990; Haegeman and Guéron 1999: 614-625) formal grammarians have tended to ignore non-clausal utterances, Functional Grammarians cannot afford to do so, committed as they are to understanding how language is used. The standard model of Functional Grammar (FG), however, has tended to assume the clause, invested with illocutionary force and propositionality, as the sole unit of description. The present paper will argue that that position need not he totally abandoned, but that a discourse- grammatical approach will bring order to the analysis of both non-clausals and clausals in FG. Two scales, formal and functional, will he identified and brought into correspondence:

Formal: non-verbal > formulaic > phrasal > elliptic > elisional > complete

Functional: expressive > interactional > illocutionary > propositional

Conceptual unity between discourse and grammatical analysis will he achieved by adopting an incrementalist stance to the sequencing of conversational units and the constituents of those units (Mackenzie forthcoming).

References

Barton, E. (1990). Nonsentential constituents: a theory of grammatical structure and pragmatic interpretation. Amsterdam. Benjamins.

Biber, D. Johansson, S., Leech, G., Conrad, S., and E. Finegan (1999). Longman Grammar of spoken and written English. Harlow: Longman.

Hageman, L. and J. Guéron (1999). English Grammar: a generative perspective. Oxford and Malden MA: Blackwell.

Mackenzie, J.L. (forthcoming). First things first: towards an incremental Functional Grammar of English. Acta Linguistica Hafniensia.

FG And RRG On Old English Quirky Case

Javier Martín Arista. Universidad de La Rioja. Spain.

Descriptive approaches to old English syntax (Visser 1963-1973 ; Mitchell, 1985) have stressed the atypical government of accusative and dative case where the complementation pattern of the verb seems to require nominative and accusative respectively. This paper is meant to provide a comparison between the more lexically-oriented approach to quirky case favoured by FG (Dik 1989; Dik and Hengeveld 1997) and the more syntactically-based view adopted by RRG (Van Valin and LaPolla 1997) . Data dealt with in this paper have been drawn from a concordance of the Helsinki Corpus (AD 850-1050) Additional information comes from Faarlund (1990) Van Valin (1991) and Roberts (1995) The conclusion is drawn that an adequate treatment of the assignment of functional relations allows for a better understanding of quirky case phenomena than formal studies on old English case assignment like Denison (1993) and Allen (1995) have provided.

References

Allen, C. L. (1995). Case marking and reanalysis: grammatical relations from Old to early modern English. Oxford, oxford University Press.

Helsinki Corpus. (1991). The Helsinki Corpus of English Texts. Helsinki, Department of English, University of Helsinki.

Denison, D. (1993). English historical syntax: verbal constructions. London, Longman.

Dik, S. C. (1989). The Theory of Functional Grammar I. The Structure of the Clause. Dordrecht, Foris Publications.

Dik, S. C. and K. Hengeveld, ed. (1997). The Theory of Functional Grammar II: Complex Constructions. Berlin, Mouton de Gruyter.

Faarlund, J. T. (1990). Syntactic change : toward a theory of historical syntax. Berlin, M. de Gruyter.

Mitchell, B. (1985) old English Syntax. Oxford, Oxford University Press.

Roberts, L. (1995). (3)pivots, voice and macroroles: from Germanic to universal grammar.(2) Australian Journal of Linguistics 15: 157-214.

Van Valin, R. (1991). (3)Another look at Icelandic case marking and grammatical relations. (2) Natural Language and Linguistic Theory 9: 145-194.

Van Valin, R. D. and R. J. LaPolla (1997). Syntax: structure, meaning, and function. Cambridge, Cambridge University Press.

Visser, F. (1963-1973). An Historical Syntax of the English Language. Leiden, Brill.

Derived Intransitivity Revisited. A Cognitive Perspective

Juana I. Marín Arrese. Universidad Complutense de Madrid. Spain.

In the literature on derived intransitives or thematic-subject constructions - so-called inchoative or anticausative and middle or patient-subject constructions - it is widely held that these constructions differ syntactically, and that, in cross-linguistic terms, we find the same split as in English between the inchoative and the middle (Keyser & Roeper 1984; Davidse 1992; Fagan 1992; Mendikoetxea 1999, inter alia). Langacker (199l), however, considers the difference between these constructions a matter of degree, as the existence of blends and non-prototypical cases seems to point to. Certain puzzling facts about the middle construction, such as the existence of middles derived from intransitives and of middle constructions with instrumental subjects, have also been pointed out (Van Oosten 1986; Heyvaert & Davidse 1999).

Within the FG framework, these constructions are considered to be derived by First Argument Reduction and further characterised by argument shift and satellite absorption, and involve a virtual/potential interpretation (Kahrel 1985; Dik 1997).

In this paper we will provide a characterisation of these constructions in English and Spanish with respect to:

(a) the properties of and constraints on these constructions in both languages;

(b) the occurrence of blends and non-prototypical instances.

We aim to examine the explanation provided by the FG model for the linguistic facts involved. Following Langacker (1991), we will offer an alternative explanation for the characteristics of these constructions in terms of profiling and conceptual autonomy/dependence alignment.

References:

Davidsc, K. (1992) Transitivity/ergativity: the Janus-headed grammar of action and events. In: M. Davis, J. Martin & L. Ravelli (eds.) Advances in Systemic Linguistics.- Recent Theory and Practice, 105 -13 S. London & New York: Pinter.

Dik, S. (1997) The Theory of functional Grammar. Parts 1 & 11. Ed. by K. Hengeveld. Berlin: Mouton de Gruyter.

Fagan, S. (1992) The Syntax and Semantics of middle Constructions. Cambridge: Cambridge University Press.

Heyvacrt, L. & K. Davidsc (1999) On the so-calied 'middle' constraction in English and Duteh. Paper presented at the Symposium on Contrastive Linguistics and Translation Studies.- Empirical Approaches. U Catholique de Louvain, Louvain-la-Neuve.

Kahrel, P. (1985) Some aspects of derived intransitivity. Working Papers in Functional Grammar 4.

Keyser,S.J. &T.Roeper(1984)On the middle and ergative constructions in English.Linguistic lnquiry 15:381- 416.

Langacker, R. (1991) Foundations of Cognitive Grammar. Vol. 11. Stanford, Ca.: Stanford University Press.

Mendikoetxea, A. (1999) Construcciones inacusativas y pasivas. In: 1. Bosque & V. Demonte (eds.) Gramática Descriptiva de la Lengua Española, vol. 2, 1575-1629. Madrid: Espasa Calpe.

van Oosten, J. (1986) The Nature of Subjects, Topics and Agents: A Cognitive Explanation. Bloomington, IN. Indiana University Linguistics Club.

A Discourse-Cognitive Perspective of the Functional Notion of Theme

Anne M. McCabe Saint Louis University, Madrid Campus. Spain.

Isabel Alonso Belmonte. Universidad Autónoma de Madrid. Spain.

In this presentation, we look at Theme choice in terms of its experiential content, focusing on its function in discourse from a cognitive perspective. We believe that Theme choice comes about because of influences from several factors, due to pressures which are connected to the three metafunctions of language: the interpersonal, the textual and the experiential. These pressures apply in varying degrees throughout texts depending on the text producers´ main concerns at any given point in the text, and depending on other co-and con-textual factors.

We focus in this presentation on the experiential metafunction, i.e. the text producer's cognitive perspective of reality which is made manifest by topical Theme choices. We will broaden the functional conception of Theme by adding a discoursal cognitive perspective, and then review other Theme analysts´ work (e.g. Ghadessy, 1993; Fries, 1995a; inter alia). We also provide evidence from our own corpus studies on Theme in Spanish and English to show how an analysis of experiential Theme selection portrays the text producer's cognitive view of reality. This leads us to the conclusion that, given that Theme choice is linked to the contextual configuration of the text, the way in which text producers view the field of discourse, i.e. their cognitive representation of the field, is a contributing factor in motivation for Theme selection.

Discourse Coherence: A Case Study of the French Connectors Car, Parce Que and Puisque

Mohammed Jadir. Hassan II University. Morocco.

This paper aims to explore formal, semantic, pragmatic and textual properties of French connectors car, parce que and puisque commonly presumed as interchangeable and synonymous.

Building on the result of earlier studies on complement and adverbial clauses (Bolkestein 1990, 1992; Dik-Hengeveld 1990; Hengeveld 1996; Dik 1997 a,b) I will first try to examine the extent of the differences between the three "cause" connective particles can be accounted for in terms of the differences between the layers of the hierarchical structure of the clause as conceived of within the theory of Functional Grammar (FG). The entity types referred to by (simple and complex) terms of clauses introduced by markers at issue co-determine semantic functions of such clauses.

Second, if the information marked by car, parce que and puisque is focal and topical respectively, puisque- clauses, as will be argued, will be assigned the pragmatic function /e.g. Topic).

The last part of the paper will be devoted to an issue that has so far not received a great deal of attention in the FG literature: the textual function of discourse markers (cf. Kroon 1997; Risselanda 1998). I will try to give an FG account of the contribution of car, parce que and puisque in creating the degree of the French discourse coherence in the light of Dik´s (1997b:ch 18) discourse model (cf. Hannay-Bolkestein eds. 1998).

References

Bolkestein, A. M, 1990, 'Sentential complements in Functional Grammar ; embedded predications- propositions and utterances in Latin', in Nuyts el al. (eds.), Layers and levels of representation in language theory. Amsterdam: Benjamins.

Bolkestein, A. M. 1992, 'Limits to layering: locatability and other problems', in Fortescue et al. (eds.), Layereded structure and reference in funcitonal perspective. Amsterdam: Benjamins.387-407.

Dik.S.C. 1997a. The Theory of Functional Grammar. Part l.- the structure of the clause. Berlin´s Mourton de Gruyter.

Dik,S.C. 1997b. The Theory of Functional Grammar. Part 2: complex and derived constructions. Berlin - Mouton de Gruyter.

Dik, S. C. & K. Hengeveld. 1990, 'The hierarchical structure of the clause and the typology of perception verb complement', WPFG 37.

Hannay, M- E. Vester. 1997, 'Non-restrictive relatives and the representation of complex sentences'. in Auwera- Goossons (eds.), Ins and outs of the predication. Dordmcht: Foris, 39.52.

Hannay, M- M. Bolkestein (eds,), 1998. Functional Grammar and verbal interaction. Berlin : Mouton de Gruyer.

Hengeveld, K. 1996. 'The internal structure of adverbial clauses', in Devriendt et al. (eds.), Complex structures: a functional perspective. Berlin - Mouton de Gruyer, 119-147.

Kroon, C, 1997. 'Discourse markers, discourse structure and Functional Grammar', in Connolly et al. (eds.), Discourse and pragmatics in Functional Grammar, Berlin : Mouton de Gruyer, 17- 32.

Risselada, R, 1998. 'Tbe discourse function of sane : Latin marker of agreement in description, interaction and concession', in Journal of Pragmatics 30, 225-244.

Discourse Structure and the Parallelism Hypothesis

Ahmed Moutaouakil

University Mobammed. Morocco.

One of the most extensively discussed issues in nowadays FG literature is the way in which the current model of FO - which is, as well known, basically a sentence-grammar - can be extended so as to deal with discourse phenomena. In this study, I will concentrate on, argue for and further develop the idea put forward in Dik (1997 Chap 18) that the hierarchical structure and the functional relations postulated for the clause can be projected onto the discourse levels. In the light of this hypothesis, which I will call the "Parallelism Hypothesis", I will try to show that an archetypical structure with two levels (a representational and an interpersonal levels) and five layers (a quality, a quantity and a locality layers at the representational level and a modal and an illocutionary layers at the interpersonal one) can he assumed to be involved, although with decreasing degrees of explicitness from text to term-phrase, by any communicative unit (a term-phrase, a clause, a whole text). The implications of such an hypothesis for linguistic universals, discourse typology as well as language typology will he examined. Throughout the study, special attention will be paid to the still debated question whether the contextually bounded properties (Illocution, Subjective Modality, Pragmatic Functions etc.) are to be represented in the grammatical module or in a separate "pragmatic" module.

References

DIK, Simon C., 1997, The Theory of functional Grammar, Part 11. Complex and Derived Constructions. Mouton De Gruyter, Berlin.

Theme/Topic Relationship In News Items And Tourist Brochures

A. Jesús Moya Guijarro. Universidad de Castilla-La Macha. Spain.

We aim to study the differences in the thematic/topical progression of two subgenres, news items and tourist brochures. The general hypothesis proposed for this research is that the news items and tourist brochures should, according to the characteristics of genre and register, show a significant difference in their location of the topical elements in the clause structure and in their topical/thematic organisation.

Firstly, after a brief definition of theme as point of departure of the message (Halliday, 1994), the concept of topic will be defined as a semantic-pragmatic notion which makes reference to the entity, proposition or main idea which a sentence or discourse is about. Secondly, following van Dijk (1981), van Oosten (1985), Dik (1989) and Downing (1997), two basic levels of topicality will be proposed, distinguishing between discourse topics and sentence or local topics. Finally, the correlation between theme and topic will be analysed in 20 tourist brochures and 20 news items, taken from a selection of magazines and quality newspapers.

The persuasive function of the tourist brochure frequently leads the writer to move the new topical entities further away from the initial position in the sentence. However, in the news items the local topics tend to be located in thematic position.

Meaning Construction Within Spanish Nominal Compounds

Henrik Hoeg Maller. Copenhagen Business School

This presentation will focus on the mental integration of concepts - also known as blending - that takes place in the process of forming and interpreting phrasal nominal compounds (Turner & Fauconnier 1995, 1996 and 1998).

The principal method of combining nouns in Spanish is by placing the secondary noun last, being the leftmost noun the head, and connecting them with a preposition, which very often is de. Consequently, by phrasal nominal compounds are here intended two (or more) nouns connected by preposition used together to denote one single object or concept.

According to the lexical properties of the head noun, I will distinguish between predicative structures (la transferencia de genes) and non-predicative structures (el sistema de control), the subject of this paper being the latter.

Characteristic of the non-predicative structures is that no overt specification of their meaning is imposed by neither the two nouns nor the preposition in question. The construction is semantically highly underspecified and it seems as if any type of association can be encoded between the two nouns.

In order to account for the general logic that enables us to generate and interpret nominal compounds, I propose a design for a theory of meaning construction which is based on the ideas of world-knowledge structuring and lexical semantics expounded by Moravesik (1990) and Pustejovsky (1995).

Taking as point of departure the four explanatory categories of the Qualia-structure: CONSTITUTIVE, FORMAL, TELIC and AGENTIVE (Pustejovsky: ibid.), the interpretation of a compound construction depends not only on the vague concepts of context and world-knowledge, but more accurately on which parts of the Qualia-structure are activated by the modifier. According to this meaning construction model, the preposition functions as a cognitive primitive that makes us perform the basic mental operation of integrating the two concepts denoted by the compound nouns (Müller: 2000).

References:

Moravesik, J.M. 1990. Thought and Language. London: Routiedge.

Müller, H.H. 2000. Noun phrases in specialized communication. The cognitive processing of the Danish s-genitive construction. Text, Language and Knowledge. Mental Models of Expert Communication. Berlin: Mouton de Gruyter. To appear.

Pustejovsky, J. 1995. The Generative Lexicon. Cambridge, Massachusetts. London, England: The MIT Press.

Turner, M. & G. Fauconnier. 1995. Conceptual Integration and Formal Expressions. M. Johnson (ed.), Journal of Metaphor and Symbolic Activity, volume 10, number 3.

Turner, M. & G. Fauconnier. 1996. Blending as a Central Process of Grammar. A. Goldberg (ed.), Conceptual Structure, Discourse and Language. Stanford: Center for the Study of Language and Information (CSLI) [distributed by Cambridge University Press]. Pages 113-129.

Turner, M. & G. Fauconnier. 1998. Conceptual Integration Networks. Cognitive Science, 22(2). Pages 133-187.

Valency Reducing Rules: The Case Of Reciprocal Predicates In Moroccan Arabic

Abdelfattah Nacer Idrissi. Faculté del lettres Agadir. Maroc.

As it’s admitted in FG framework, predicate formation rules can have different effects on the input predicate frame: they modify not only its quantitative valency but also its qualitative one. Both of these modified valencies can be found in Moroccan Arabic Language (MA). My aim in this paper is to present one type of valency reducing rules, i.e. reciprocal predicate formation rules by dealing with the following question:

"To which type of reduction reciprocal predicates do pertain: to second argument reduction or to predicate frame reduction as it's advanced by Moutaouakil (1988a-b)?"

References :

Dik, S.c., 1997-b The Theory of Functional Grammar, Part 2: Complex and Derived Constructions, Mouton de Gmyter, Berlin, New York.

Moutaouakil, A., 1988-a Essais en Grammaire Fonctionnelle, SMER, Rabat.

Moutaouakil, A., 1988-b Qadaya mucjamiya fi llugati Icarabiya, SMER, Rabat.

Degrees Of Auxiliarization: The Case Of Spanish Periphrases

Hella Olbertz. Leiden University

To assess the degree to which a given verb is auxiliarized, the following questions should be answered.

A. Does the verb impose any constraints (inherited from its lexical function), and if so, how strict are they?

B. How specific/concrete or general/abstract is the auxiliary function the verb fulfils?

C. How frequently does the verb occur in its auxiliary function, i.e. how relevant is it within the grammatical system?

The following examples serve to illustrate each of these questions.

(1) Illustrates question A: ponerse a can only be used as an expression of Ingressive Aspect when the referent of A1 is animate, whereas there is no such constraint on the synonymous empezar:

1) La cosa *se pone/empieza a funcionar.

"The thing begins to work."

Examples (2) and (3) illustrate question B: while pensar in (2) is an expression of Volitional Modality with a specific (i.e. "near-lexical") meaning, poder in (3) expresses Epistemic Modality, which is a more abstract concept:

(2) No piensa volver a España.

"He does not intend to return to Spain."

(3) No descarto lo que dice, puede ser verdad.

"I don't discard what he says, it may be true."

The relative frequency of dejar de and cesar de, both expressions of Egressive Aspect, may serve to illustrate question C: in a corpus of approximately 300.000 words dejar de occurs 38 times, whereas cesar de occurs only once.

Although with many Spanish periphrases the answers to all of the three questions point into the same direction, there are quite a few periphrases that can fulfil their "grammatical" function without any restriction (A), while the "grammaticality" of the function may be doubted (B), and the relative frequency is low (C).

In my talk I will present an overall account of the auxiliarization of Spanish periphrases in the above sense and consider how FG can be applied in this context.

A Cognitive Approach To Adverbial Subordination.

Johan Pedersen. Universidad de Copenhague. Denmark.

Defining the notion of subordination and the way in which it differs from coordination is a notorious problem. In addition, the grammatical tests used are highly inconsistent. Three continua reflect the complexity of the adverbial subordinator = Q (Kortmann 1997):

D) Adposition - adverbializer - adverb.

E) Complementizer - adverbializer - relativizer.

F) Coordinator - subordinator - conjunctional adverbs.

Recent research in cognitive grammar offers a framework that allows us to see A, B and C as interrelated continua by means of the link-concept (Croft 1999, Deane 1992, Johnson 1987), distinguishing between a locative and a configurative link. Idea: C is motivated by a specific Q-predicate (locative link, reflected in A) construing - "verb like" - an idealised link- structure (a link-configuration, reflected in B). This idea suggests that the habitual distinction between subordinator and coordinator (cf. C) is manifested in two different ways: 1) as a ± "verb like" predicate caracter (Langacker 1991) of the connecting particle, since not all clause-linking devices code a specific link-structure, cf. A (ex. that=que), nor an idealised link-configuration, cf. B (ex. say=es decir); 2) in a "verb like" functioning of the connective predicate: constructional perspectives/alternations.

Based on empirical evidence, I will try to show the validity of this thesis, demonstrating the connection between adverbial subordination/coordination and the "verb like" grammatical relation: in the first place I demonstrate a clear correlation between the way recognised tests for subordination/coordination group the corpus, and the way it is divided in terms of alternative predicate constructions. In the second place, I show the existence of a connective control function (cf. the verb control) - in principle similar to the right frontier restriction in discourse analysis (ex. Webber 1991) - which in complex clause-linkage constructions, based on two connective predicates, group the corpus in a very similar way. Particularly, it demonstrates that aunque represents a special, more flexible, control function, in conformity with its more complicated behaviour as a subordinator/coordinator. This behaviour can thus de described very precisely. 1350 expressions with aunque have been studied. I conclude that this connective predicate grouping motivate our intuitive distinction between adverbial subordinators and coordinators.

References:

Clausner, T. & Croft, W. (1999): Image schemas and domains, Cog. Linguistics 1999, 10-1.

Deane, P.D. (1992): Grammar in the mind and brain, Mouton de Gruyter, Berun-New York.

Johnson, M. (1987): The body in the mind, University of Chicago Press, Chicago.

Kortmann, B. (1997): Adverbial subordination, Mouton de Gruyter, Berlin-New York

Langacker, R (1991): Foundations of cognitive grammar, vol.II: Descriptive Application, Stanford University Press, Stanford.

Webber, B. L. (1991): Structure and ostension in the interpretation of discourse deixis. Lan- guage and Cognitive Processes, 6(2).

Towards An Inventory Of Lexical Templates In A Functional Lexematic Derived Lexicon

María Jesús Pérez Qintero. Campus de Guajara. La Laguna-Tenerife

Francisco José Cortés García. Campus de Guajara. La Laguna-Tenerife

In many respects, the Functional Lexematic Model can be described as a naturally enriched version of the lexical component -or Fund- of Functional Grammar. However, as regards word-formation it is quite heterodox since it advocates the existence of an autonomous component for the treatment of derivation and composition, and defends the segregation of these processes from others that are described as predicate formation phenomena within the Functional Grammar tradition.

The Functional-Lexematic Word-Formation Component is organised internally in three subcomponents: (i) an affix lexicon; (ii) a fund of basic word formation patterns; and (iii) a set of rules with two functions: to expand those basic patterns and link the expanded ones with the appropriate lexical material (affixes and bases) to output a new word.

The proposal of this paper is to consider those patterns as templates or constructions susceptible of expansion. An attempt will he made also to show their connection with different expression formats by means of linking mechanisms; such linking rules will account for the origin of morphologically complex units and for their syntactic behaviour as well.

Towards A Cognitive-Functional Transduction Of A Machine-Readable Dictionary Of Spanish

J. Carlos Perinán Pascual

Most researches conclude that the best method for building computational lexicons is the automatic analysis of machine-readable dictionaries. All the information needed for NLP lexicons can be found in a dictionary text. We have to decode the information presented by the human lexicographer in order to make it accessible in a computational way. In the transduction of a dictionary text into a Functional Grammar format we propose an enriched model of the basic component of Dik's theory, the Fund. The meaning postulate associated to each predicative frame will take the form of a semantic density list, which enables the system to create a conceptually- motivated network connecting the predicates of the lexicon. This cognitive structuring of lexicon, also advocated by the Functional-Lexematic Model, is grounded on the Stepwise Lexical Decomposition approach. In this paper we therefore describe the automatic construction of predicative frames and semantic density lists for verbal predicates in Spanish. We chose to use the Diccionario del Español Actual (1999) as the machine-readable source for the construction of our verbal lexicon. This dictionary has some unique characteristics which make it remarkably suitable for use as the knowledge base of a NLP system of Spanish.

A Functional Description Of Adverbial And Prepositional Homomorphs In English

Santiago Posteguillo Gómez. Universitat Jaume I de Castelló. Spain.

Jordi Piqué. Universitat de Valéncia. Spain.

In this study, we understand homomorphy as the linguistic phenomenon that describes how

some words function most typically like words of a given class (for example that of adverbs), but that occasionally they realise syntactic functions which are normally realised by words of a different class (for example, adjectives). (Downing and Locke, 1992: 563)

Downing and Locke (1992: 564) contend that the most frequent instance of homomorphy in English is the overlapping of adverbs and prepositions.

In fact, this conflicting relationship between adverbs and prepositions is a well-acknowledged linguistic fact which has been analysed in a substantial number of Indo- European languages, such as Spanish (Bosque, 1992: 210-211), Slavic languages in general (Kopecny, 1970, 197la, 1971b), Russian (Ermolenko, 1963, Kalamova, 1955), Latin (Hiltbrunner, 1962), Rumanian (Bejan, 1973, 1976), Italian (Herczeg, 1969), French (Bel'skaja, 1955; Gamillscheg, 1952; Notholt, 1913; Remacle, 1956), and German (Holmlander, 1973; Levitiene, 1965; Litvinov 1974; Marx-Moyse, 1979; Okkerhaug 1952), to mention a few; and also in Non-Indo-European languages, namely Palola´s (1975) study in Finnish or Vanoverbergh's (1949) in lloko, a Philippine language.

In English, adverbs and prepositions are described as inter-related word categories in a number of studies (Allen, 1964; Avramenko, 1960; Baily, 1973; Bel'skaja, 1955; Berlizon, 1958; Blass, 1965; Brahde, 1938; Bruton, 1969; Bryant, 1946-47; Dietrich, 1956; 1960; Ellinger, 1938-39; Fairclough, 1964; Heaton, 1965; Hill, 1968, 1969; Kim, 1963; Jacobsson, 1977; Jatel, 1960; Lambert, 1929; Makeenko, 1955a, 1955b; Nenjukova, 1953; Owen, 193 1; Prigorovskaja, 1963; Schweisthal, 1967; Sroka, 1962, 1968, 1969; Veselitskij, 1955, 1958, Zlutenko, 1958; Burton-Roberts, 1991). As opposed to these accounts, Halliday (1985: 214) separates the prepositions and the adverbs in his description of word classes. He considers that the preposition is more related to the verb category whereas the adverb, together with conjunctions, constitutes what he labels as the adverbial category. We certainly agree with him when he underscores the close relationship between prepositions and verbs, but it is our contention that Halliday's classification oversimplifies the problem and that, for instance, the overlapping of prepositions and adverbs should not be overlooked, as the various studies (see above) describing this linguistic phenomenon in many different languages have shown. To illustrate our point of view, the objectives of this study are an attempt to:

a) classify the different adverbial-prepositional homomorphs according to whether they are more frequently used as either adverbs or prepositions-using WordSmith Tools Software (Scott, 1996) program for this analysis;

b) to draw an ideational diagram -showing the intersection between the adverbial and the prepositional word categories-- where the different homomorphs can he placed according to their different usage

Extending The FG Discourse: Adjacency Sequences

Carmen Portero Muñoz. University of Córdoba.

Daniel García Velasco. University of Oviedo.

With a view to develop into a grammatical model which takes entire discourses as input, most recent work in FG has paid attention to the systematic description of the various features of discourse that have to be considered in the creation of such a model. In its most recent version FG supports the existence of a parallelism between clause and discourse in the sense that different functional layers can be distinguished in both kinds of units, namely interpersonal and representational layers. It is generally acknowledged that participants contribute to create and maintain the structure of the discourse event they are involved in with their moves, and that they do so by using two types of strategies corresponding to each of the layers which have been distinguished in the discourse event. At the interpersonal level, the smallest unit is the speech act, and the crucial actions in establishing coherence in conversation are not such speech acts as requests and assertions, but rather challenges, defences and retreats, which have to do with the status of the participants, their rights and obligations, and their changing relationships in terms of social organization.

The aim of this paper is to illustrate the use of interpersonal strategies for both interaction management and attitude specification by examining a selection of conversational stretches. The focus will be on specific types of conversational exchange where communicative interaction is particularly conflictive, more specifically on the strategical use of questions by someone who conversationally superior, which results in a number of recurrent patterns. Although the interactional unit Question + Answer is basic to the type of discourse event in case, by using questions in a strategical way conversational interaction will be shown to develop in such a way that a structure is found beyond adjacency pairs over longer stretches of conversation, adjacency sequences being detected.

A Dynamic Perspective On Semantic Fields Relationships From Motion To Possession And Beyond

Raissa Rozina. Russian State University. Moscow.

The categorisation and representation of knowledge of the outer world revealed by the semantic field of possession have been the object of a number of studies (see, e.g. Jackendoff 1983, 1992; Faber, Fernandez 1996; Fernandez 1998; Rozina 1998). It has been shown that possession is closely related to the domains of movement and position (Jackendoff 1983, Fernandez 1998) and that these relationships are ensured by such cognitive abilities as metonymy and metaphor. I argue that possession is related to other domains as well, and that there are patterns of semantic derivation of predicates ensuring the transfer from one semantic field to another.

In Fernandez 1998 the etymological analysis of verbs of possession has revealed that their former meanings were more concrete and belonged to other domains of experience, from which verbs of possession were constructed by metaphoric or metonymic shifts. The present paper studies the synchronic relationships between the meaning of possession and other meanings in the semantic structure of polysemous verbs, in particular Russian verbs of motion. A verb's semantic structure is analysed in dynamic perspective, i. e. one meaning is seen as derived from another by means of certain rules.

The meanings of verbs of motion, e.g. Russ. vziat' 'to take' fall into the following semantic domains: MOTION (to take smth from smth) ; POSSESSION (to take smth from smb); POWER (to take smb 'to capture by force'); EMOTION ('to be taken by emotion'). As a result of the analysis, a list of derivational patterns has been compiled.

References

Faber, Pamela and Eulalio Fernandez. 1996. The Lexical Field of Possession as a Construction of Conceptual Primitives. Proceedings of the XIXth International conference of AEDEAN. Vigo: Universidad de Vigo, 63-268

Fernandez, Eulalio, 1998. Cognitive Abilities from a Functional-Lexematic Perspective. The Structure of Lexicon in Functional Grammar. Amsterdam; Philadelphia: John Benjamens, 65-84,

Jackendoff, Ray. 1983. Semantics and cognition. Cambridge, Mass, MIT Press

Jackendoff Ray 1992. Languages of the Mind; Essays on Mental Representation. Cambridge, Mass.: MIT Press.

Rozina, Raissa. 1998. Meaning Definitions and Grammatical Properties of Predicates: Russian Verbs of Possession. The Structure of Lexicon in Functional Grammar, Amsterdam; Philadelphia: John Benjamens, 85-95.

Some Remarks On Spanish Copula Constructions In A Functional Approach

Ventura Salazar-García. University of Alicante. Spain.

Within the framework of Dik's Functional Grammar (FG) (Dik, 1983; 1997,1: 193-216), copular verbs are considered as semantically empty supportive devices, functioning as carriers for tense, mood, and aspect distinctions. Hengeveld (1986; 1992) showed that the existence of more than one copular verb in Spanish (ser and estar) is compatible with the FG's theory of copula support. He concluded that, where both copulas may appear under the same conditions, the choice of one or the other copula can be attributable to aspectual differences. Hengeveld's hypothesis provides a very adequate explanation for most of the Spanish copula constructions. There are, however, a few others, not directly treated by Hengeveld, which require more detailed attention. For instance, the construction exemplified by the second sentence of the following pair: Carmen es camarera ('Carmen is a waitress') vs. Carmen está de camarera ('Carmen works as a waitress'; lit.: *'Carmen is of waitress'). This paper discusses whether or not an explanation for these constructions can introduce any changes in the current FG's copula support theory. Specifically, it suggests that not only aspect. but also modality are involved in Spanish copula constructions. It also pays attention to the grammatical status of the Spanish verb parecer ('to look like', 'to seem'), which is considered by many grammarians to be a copular verb.

References

Dik, Simon C. (1983). "Auxiliary and copula be in a Functional Grammar of English". F. Heny and B. Richards (eds.), Linguistic Categories: Auxiliaries and Related Puzzles, vol. 2. Dordrecht: Reidel, 121-143.

Dik, Simon C. (1997). The Theory of Functional Grammar (2 vols.). Ed. by Kees Hengeveld. Berlin: Mouton de Gruyter.

Hengeveld, Kees (1986). "Copular Verbs in a Functional Grammar of Spanish". Linguistics, 24, 393-420.

Hengeveld, Kees (1992). Non-Verbal Predication; Theory, Tipology, Diacrony. Berlin: Mouton de Gruyter.

Functional Pragmatics, The Natural Language User, And Functional Grammar

Ole Nedergaard Thomsen, University of Copenhagen

The ultimate object of investigation of the theory of FG (TFG, 1997) is the Natural Language User (NLU), i.e. the communicatively relevant aspects of a human being. FG is not particularly clear about the nature of NLU: even though an "ideal speaker-listener in a homogeneous speech community" does not qualify as a Dikian NLU, it is not stated that FG would endorse a variational (psycho- and sociolinguistic) NLU. FG strives to be an adequate theory of the clause, and owing to its aim of being pragmatically and psychologically adequate FG understands the structure of the clause as determined by psycholinguistic processing and by the clause functioning as an instrument of verbal social interaction. The assumption is that the clause is the basic unit of communication. This clause-centricity perhaps stems from an individualist stance of classical speech act theory (I-intentionality). However, Searle's recent methodological individualism includes a concept of irreducible collective (We-) intentionality, manifest in conversation, and the oasis of social conventions. This development warrants a discourse-based FG.

Functional Pragmatics (FP) is a name for an integrative model of grammatical and communicative competence and behaviour, adopting a Peircean trichotomy between language as law (capacities), as fact (processing hic et nunc), and as result (produced/interpreted discourse). The FP model of NLU includes a psycholinguistic processor (containing processing constraints), a Universal Functional Grammar (specifying the shape of a possible human language, functioning in language acquisition), and a universal communicative competence (pragmatic universals), in addition to the core grammar (typological groundplan and constitutive, productive rules), the regulative norms of usage, and the specific communicative competences. NLU is variational and this implies the typological adequacy of FP to involve variational typology.

The present paper will report on the theory of FP.

Degrees Of Auxiliarization: The Case Of Spanish Periphrases

Hella Olbertz. Leiden University. Amsterdam.

To assess the degree to which a given verb is auxiliarized, the following questions should be answered. A. Does the verb impose any constraints (inherited from its lexical function), and if so, how strict are they? B. How specific/concrete or general/abstract is the auxiliary function the verb fulfils? C. How frequently does the verb occur in its auxiliary function, i.e. how relevant is it within the grammatical system? The following examples serve to illustrate each of these questions.

(1) illustrates question A: ponerse a can only be used as an expression of Ingressive Aspect when the referent of A1 is animate, whereas there is no such constraint on the synonymous empezar:

(1) La cosa *se pone/empieza a funcionar.

"The thing begins to work."

Examples (2) and (3) illustrate question B: while pensar in (2) is an expression of Volitional Modality with a specific (i.e. near-lexical") meaning, poder in (3) expresses Epistemic Modality, which is a more abstract concept:

(2) No piensa volver a España.

"He does not intend to return to Spain

(3) No descarto lo que dice, puede ser verdad.

"I don't discard what he says, it may be true" '

The relative frequency of dejar de and cesar de, both expressions of Egressive Aspect, may serve to illustrate question C in a corpus of approximately 300.000 words dejar de occurs 38 times, whereas cesar de occurs only once.

Although with many Spanish periphrases the answers to all of the three questions point into the same direction, there are quite a few periphrases that can fulfil their "grammatical" function without any restriction (A), while the "grammaticality" of the function may be doubted (B), and the relative frequency is low (C).

In my talk I will present an overall account of the auxiliarization of Spanish periphrases in the above sense and consider how FG can be applied in this context,

On First Order Entities, States Of Affairs And Propositions In The Theory Of Functional Grammar

Paul Otto Samuelsdorff. University of Cologne.

In a chapter headed "Entities are mental constructs" (Dik 1997:129) gives three reasons why entities are not "things in reality" but "things in the mind". These reasons merely state that entities may be mental constructs, but none of them exclude their existence in reality. This is supported by Dik's definition of first order entities as spatial entities (Dik 1997:137) and his definition of State of Affairs (SoA) as "entities which can be said to occur, begin, last and end; they can be perceived watched heard, felt, etc. (Dik 1997:292).

Propositions are defined as "things that people can be said to believe, know or think about; they can be reason for surprise or doubt, they can be mentioned, denied, rejected, and remembered, and they can be said to be true or false." (Dik 1997:52). This definition is in accordance with the concept of proposition as it is generally used in philosophy and logic. Later on the same page Dik defines the proposition as a higher-order structure which designates a "propositional content" or a "possible fact". This would be a legitimate step if Dik did not include the attitude of the speaker in the propositional content This contradicts the above definition of "proposition". Predicate operators and satellites are not included in the predicate, nor should proposition operators and satellites be included in the proposition.

I shall justify the above points in more detail by giving a general explanation of semantic representation in a functional theory of language.

Reference

Dik, Simon C. 1997. The Theory of functional Grammar (Part l). Berlin: Mouton de Gruyter.

Cognitive Grounding And Pragmatic Control Of Semantic Functions

Hans-Jörg Schmid, University of Bayreuth, Germany.

Ever since Fillmore introduced deep cases into syntactic theory, this concept - and its relatives such as Dik's semantic functions, Halliday's participant roles, Langacker's archetype roles or the theta roles in generative grammars - has been plagued by a number of central, but still open, questions. Of these the most fundamental ones are:

a) What is the status of these roles; are they syntactic, semantic or cognitive in nature?

b) Which roles are needed for syntactic description and how can they be operationalised?

c) How is the mapping from these roles to syntactic surface functions determined?

With regard to the first two questions the claim is revived that at least some deep cases are grounded in our cognitive system. The best candidates for such a status are: agent, theme (a fairly general role, which is considered relevant and useful despite Dik's reservations) and experiencer, and location, source and goal. This claim will be justified by providing syntactic, semantic, cognitive-linguistic, psycholinguistic and typological evidence.

If such a stance is taken, question c) becomes even more pressing, because mappings that deviate from the cognitive principles that are captured in the Semantic Function Hierarchy are still harder to explain. With regard to this question, Dik's (1989: 210ff.) pioneering view that the assignment of semantic functions to surface constituents is partly determined by pragmatic factors such as givenness of information, empathy, and politeness conventions is taken up and developed further.

References

Dik, Simon (1989), The Theory of Functional Grammar. Part 1: The Structure of the Clause, Dordrecht- Providence: Foris.

Ethic Pronouns as Illocutionary Convertes in French and Galician

John Charles Smith University of Oxford (St. Catherine's College)

In this paper, I analyse the so-called 'ethic dative' clitic pronouns found in contemporary French and Galician in the light of recent work in Functional Grammar. The syntactic distribution of these items is strikingly different from that of other clitics - they may occur only in positive declarative main clauses, yet within such clauses may freely violate the co-occurrence constraints which hold for other clitics. I shall account for these data by suggesting that, unlike indirect objects, which are arguments and form part of the predication, and unlike datives of interest, which lie outside the predication, but form part of the proposition, ethics lie outside both predication and proposition, forming part of the clause or speech act. More specifically, ethics may he viewed as illocutionary converters' of the type discussed by Dik (1997:245): these are 'linguistic elements and devices through which the basic illocution of a clause may be grammatically converted into some other illocution'. Ethics have the function of converting Declaratives into Exclamatives (a grammatical conversion which is intuitively plausible, although not mentioned by Dik in his (non-exhaustive) list of such operations (ibid.:243)). In keeping with the core value of the dative, which is to encode an 'affectee', it is also plausible to suggest that ethic pronouns encode the 'affectee' of the speech act. I shall relate this notion to Rijkhoff's proposals (Rijkhoff 1995; 1998) concerning 'bystander deixis'.

References

Dik, Simon C. 1997. The Theory of Functional Grammar. Part 2: Complex and Derived Constructions (ed. Kees van Hengeveid). Berlin & New York: Mouton de Gruyter (Functional Grammar Series 21).

Rijkhoff, Jan (1995). 'Bystander and social deixis: some programmatic remarks on the grammar/pragmatic interface.' Working Papers in Functional Grammar 58.

Rijkhoff, Jan (1998). 'Bystander deixis.' In Yaron Matras (ed.), The Romani Element in Non-Standard Speech. Wiesbaden: Harrassowitz.

On Transitivity In FG And In Cognitive Linguistics

Svilen Stanchev. University of Veliko Turnovo. Bulgaria.

In pre-FG structuralist and transformational linguistic studies transitivity and related categories like voice and reflexivity were treated as purely syntactic phenomena in terms of direct objecthood and transformations. With the further development of formalist approaches to grammar (notably GB) some lexico-semantic 'correctives' were introduced in an attempt to cope with the difficulties that the search for 'purely syntactic' solutions had led to. It was only with the launch of functional linguistics that the precedent of semantics and pragmatics was recognised and linguistic structures got to be treated as instrumental in verbal interaction rather than as an abstract object of study in their own right.

Although transitivity as such has never been a major issue in Functional Grammar its various aspects have been dealt with in terms of predicate frames, predicate formation, and function assignment among others. In this way lexico-semantic and pragmatic considerations were duly highlighted at the expense of purely structural discussions of direct objecthood, passivization etc. Cognitive linguistics has focused on the notion of construal and has taken a multifactorial approach to transitivity involving parameters with structural as well as extralinguistic relevance (cf. e.g. Hopper & Thomson (1980), Rice (1987)). Within this framework a theory-neutral and prototype-based explanation of transitivity has been sought within a scope broad enough to encompass both linguistic and extralinguistic factors.

In my paper I will argue for a complementary rather than mutually exclusive treatment of the functionalist and cognitive accounts of various aspects of transitivity supported by linguistic data from Bulgarian and English. The blending of FG explanatory formalism with the broader spectrum of cognitive parameters is seen as advantageous for an enriched and more detailed typology of states of affairs and some reconsideration of function assignment and predicate formation.

References:

Hopper, P.J. & S.A. Thomson. 1980. Transitivity in Grammar and Discourse. Language 56: 251-295.

Rice, S.A. 1987. Towards a Cognitive Model of Transitivity. Ph.D. diss., University of California at San Diego.

Four Types - Four Functions

Rita Therkelsen. Roskilde University.

According to Simon C. Dik (1997) Functional Grammar adopts the traditional subdivision of relative clauses into restrictive and non-restrictive (relative clauses being classified as a subclass of Verbal Restructures). In this subdivision the status of the relative clause is justified in and related to the layering of the clause, and the form of the head noun is not taken into consideration. From another functional view Givón (1984) also adopts the traditional dichotomy, although assigning different functions to the two kinds of relative clauses based on the 'givenness' and the 'newness' of the entity referred to by the head noun. Lehmann (1984) takes into account both the grammatical form of the head noun and the textual status of the entity referred to by the head noun, and he points out that both restrictive and non-restrictive relative clauses, as far as function is concerned, should be further subdivides into two types of restrictive and two types of non- restrictive clauses. In my research on Danish relative clauses (Therkelsen 1999 and 2000) Lehmann's further subdivision proves grammaticalized to some extent. The main point in my paper is that Lehmann's and my own results show that in accounting for the functions of relative clauses, it is necessary to go beyond the boundaries of the clause and include the textual layer. One should, however, carefully investigate the degree of grammaticalization of functions related to the textual layer.

References

Dik, S.C. 1997. The Theory of functional Grammar. Vol. I, II. Kees Hengeveld (ed.). Mouton de Gruyter: Berlin, New York.

Givón, T. 1984. Syntax - a functional-typological introduction. Vol. I, II. Benjamins: Amsterdam.

Lehmann, C. 1984. Der Relativsatz. Gunter Narr Verlag: Tübingen.

Therkelsen, R. 1999. Betydningsforskellen mellem der- og som-relativsaetninger. Ny forskning i Grammatik. SHF, faellespublikation 6. Odense Universitetsforlag

Therkeisen, R. 2000. Om klassifikation af ledsaetninger. Ny forskning i Grammatik, SHF, faellespublikation 7. Odense Universitetsforlag (forthcoming).

Old English Verbs For Giving Revisited: Turning Social History Into Lexeme- And Dimension-Level Schemata

Juan Gabriel Vázquez González. Universidad de Málaga. Spain.

This paper aims for the validation of the Functional-Lexematic Model from a historical perspective, and purports to do so by adapting the Germanic notion of the comitatus to a formalised Functional-Grammar basis (Dik, 1997). The theoretical postulates of this approach, developed by Martín Mingorance (1984, 1985a, 1985b, 1987a, 1987b, 1987c, 1990 & 1995) and his school (Faber & Mairal, 1994, 1997 & 1999), are now applied to a group of nearly three hundred verbal lexemes ascribed to GIVE in the Anglo-Saxon period (Vázquez González, 1999). In order to describe the particular structure of this lexical subdomain, it is our intention to show significant samples of lexeme-level schemata and concentrate on the description of its dimension-level schema, where attention will be paid to the display of interfield connections and the particular codification of pragmatic features. Both levels eventually reflect the hierarchical structure of Anglo-Saxon society and the existing clash between warrior-like aristocracy on the one hand and the rising economic needs on the other (Stenton, 1943; Whitelock, 1967). The results obtained may be also approached from a diachronically regressive perspective (Benveniste, 1969) by means of etymology (Klein, 1971), which may throw some further light on the continental phase (Lass, 1994).

References

BENVENISTE, E. 1969. Le vocabulaire des institutos indo-europénnes. Paris: Les éditions de Minuit.

DIK, S. 1997. The Theory of Functional Grammar Parts I & II. Ed. K Hengeveld. Berlin: Mouton de Gruyter.

FABER, P. & MAIRAL USÓN, R. 1994. "Methodological Underpinnings for the Construction of a Functional Lexicological Model" Miscelánea. A Journal of English and American Studies. Vol. 15. 193-217.

FABER, P. & MAIRAL USÓN, R. 1997. "Towards a typology of predicate schemata in a Functional-Lexematic Model" Toward a Functional Lexicology/Hacia una Lexicología Funcional. Ed. G. Wotjak. Frankfurt: Peter Lang. 11-36.

FABER, P. & MAIRAL USÓN, R 1999. Constructing a Lexicon of English Verbs. Functional Grammar Series, 23. Amsterdam: Walter de Gruyter.

KLEIN, E, 1971. A Comprehensive Etymological Dictionary of the English Language. Dealing with the origin of words and thus illustrating the history of civilization and culture. Amsterdam: Elsevier.

LASS, R. 1994. Old English. A historical linguistic companion. Cambridge: CUP.

MARTÍN MINGORANCE, L. 1984. "Lexical Field and Stepwise Lexical Decomposition in a Contrastive English-Spanish Verb Valency Dictionary". LEX'eter' 83: Proceedings of the International Conference on lexicography. Ed. R. Hartmann. Tübingen: Niemeyer. 226-37.

MARTÍN MINGORANCE, L. 1985a. "La Semántica Sintagmática del Adjetivo: Parámetros para la Organización de un Lexicón Inglés/Español de Valencias Adjetivales". Actas del 11 Congreso Nacional de Lingüística Aplicada. Madrid AESLA-SGEL. 329-340.

MARTIN MINGORANCE, L. 1985b. "Bases Metodológicas para un estudio contrastivo del léxico derivado". Revista Española de lingüística aplicada. Vol. 1. Madrid. SGEL-Educación. 37-54.

MARTÍN MINGORANCE, L. 1987a. "Classematics in a Functional-Lexematic Grammar of English". Actas del X Congreso de la Asociación Española de Estudios Anglo-Norteamericanos. Zaragoza: Publicaciones de la Universidad. 377-382.

MARTÍN MINGORANCE, L. 1987b. "Semes, Semantic Classemes and Dimensions: The Lexicological and Lexicographic Perspectives". Paper presentes at the XIV Congress of Linguists. Berlin.

MARTÍN MINGORANCE, L. 1987c. "Pragmatic Features in the Lexicon of a Functional Grammar" Paper presented at the International Pragmatics Conference. Antwerp.

MARTÍN MINGORANCE, L. 1990. "Functional Grammar and Lexematics in Lexicography". Meaning and Lexicography. Eds. J. Tomaszczyk. and B. Lewandowska-Tomaszczyk. Amsterdam: Benjamins. 227-53

MARTÍN MINGORANCE, L. 1995. "Lexical Logic and Structural Semantics. Methodological Underpinnings in the structuring of a lexical database for natural language processing". Panorama der Lexicalischen Semantik. Thematische Festschrift aus AnlaB des 60. Geburtstags von Horst Geckeler. Ed. U. Hoinkes. Tübingen: Gunter Narr Verlag. 461-74.

STENTON, F.M. 1943. Anglo-Saxon England. London: Clarendon Press.

VÁZQUEZ GONZÁLEZ, J.G. 1999. El Campo de la Posesión en Inglés Antiguo y Medio: Aplicación del Modelo Lexemático-Funcional al Cambio Lingüístico. Tesis Doctoral. Publicada en microfichas. Córdoba: Servicio de Publicaciones.

WHITELOCK, D. 1967. The Beginnings of English Society. London: Penguin.

Subjective Vs Objective Modality In Functional Grammar: Criteria, Types And Unmarked Options

Jean Christophe Verstraete. Fund for Scientific Research. Flanders.

In this paper, I will propose three modifications to the analysis of subjective vs. objective modality as it has been developed in the FG framework (Hengeveld 1988 & 1989, taken over in Dik 1997). The English modal auxiliaries will serve as a test case for the proposed modifications.

I will first show that the grammatical criteria presented in Hengeveld (1988) can only be understood correctly if we look at them from the perspective of the performative-descriptive distinction which underlies the subjective-objective dichotomy (Nuyts 1993). More particularly, I will argue that it is not the case that subjective modals cannot occur in interrogatives (Hengeveld 1988), but rather that there is an interaction between subjective modality and interrogation which transfers responsibility to the interlocutor. Further, subjective modals are not unacceptable in conditionals (Hengeveld 1988), but rather receive a specialized echoic interpretation which must be regarded as a propositionalized resumption of a performative use. On the basis of the performative-descriptive distinction, I will also propose two new criteria: the interpretation of the modal in embedded contexts and specific restrictions on modal verb-adverb coocurrence.

Secondly, I will apply these criteria systematically to the deontic, epistemic and dynamic (ability & volition) categories of the English modals. The resulting classification is markedly different from Hengeveld's (1988) proposal. Whereas in Hengeveld's model only epistemic modality can be both subjective and non-subjective, and deontic modality is always non-subjective, the classification resulting from the criteria is the other way round: deontic modality can be both subjective and non-subjective, and epistemic modality can only be subjective.

Finally, I will argue that the category of subjective modality should not be restricted to formally marked instances of modality. I will show that subjective modality is a grammatical system which consists of a formally unmarked option (simple indicative for epistemic, imperative for deontic) in addition to the formally marked options. Semantically, the absence of explicit modal verbs still takes up a function in the system by signalling default commitment to truth (epistemic) or necessity of action (deontic). Structurally, the inclusion of the unmarked options within the subjective-modal system is reflected in the fact that they share a number of modality-related properties with the explicitly marked types of modality

References

Hengeveld, Kees. 1988. Illocution, mood and modality in a functional grammar of Spanish. Journal of Semantics 6:227-269.

Hengeveld, Kees. 1989. Layers and operators in Functional Grammar. Journal of linguistics 25: 127-157.

Dik, Simon. 1997 (2nd ed). The Theory of Functional Grammar. Part 1: The Structure of the Clause. Berlin: Moutonr.

Nuyts, Jan. 1993. Modality and the layered representation of conceptual and linguistic structure. Linguistics 31: 933-969.

Accentuation and the Order of Constituents in Dutch

Roel Vismans. University of Hull.

It is widely recognised that in speech new information is generally marked by prosodic prominence. In Dutch the timing of a constituent in the clause also plays a part in differentiating between given and new information. In a neutral Dutch accentuation pattern the prosodic prominence occurs late in a clause, affecting the timing of new information. On the other hand, in a clause with a marked intonation pattern, new information may occur much earlier. Building on earlier investigations into the interaction between accentuation and constituent in the FG-framework,' this paper reports on an ongoing study of this interaction.

Whilst the literature on Dutch syntax, especially seen from a formal perspective, is fairly extensive, this is much less the case for prosodic phenomena. Authoritative experimental work on the phonetic aspects of Dutch prosodic structure exists (cf. 't Hart et al. 1990)2, but very little has been written about its pragmatic aspects and none of this is based on experimental research, which is essential for this kind of work.

For a pilot study, six sentences were recorded describing the same state of affairs but varied in terms of their constituent order. Each was then manipulated in terms of three accentuation patterns. We also recorded three questions to which the test sentences could be possible responses, resulting in 54 stimulus-response pairs. These pairs were arranged in random order with suitable pauses. The resulting output was then recorded for an experiment in which eight subjects were asked to grade each response in terms of its acceptability in the relevant context.

This paper will discuss the extent to which such acceptability depends not only on the accentuation pattern of a clause, but also on the timing of the accented constituent.

-----------------------

1 Cf. Vismans, Roel (1997) 'Pragmatic functions, special positions and accent in Dutch' in John H. Connolly et al. (eds) Discourse and Pragmatics in Functional Grammar Merlin: Mouton de Gruyter, pp. 109-119

2 Hart, Johan 't, René Collier & Antonie Cohen (1990) A perceptual study of intonation: An experimental-phonetic approach to speech melody Cambridge: CUP

Adverbial Clauses in P1

Ewa D. Zakrzewska. University Library of Amsterdam.

The Theory of functional Grammar gives rather scanty information on the subject of adverbial clauses. They are treated as complex terms occupying argument positions (TFG2:95) for which the usual rules of semantic and pragmatic functions assignment should apply. Recently this view has been challenged by Kwee T. L. with respect to the scope of adverbial clauses ('Adverbial clauses, FG, and the change from the sentence grammar to discourse-text grammar', WPFG 70, 1999). I will address a similar issue from another perspective: that of the positioning of an adverbial clause with regard to the matrix clause. My argumentation will be based on the example of proposed temporal clauses in Coptic.

In Coptic the unmarked (= postfield) position for adverbial expressions, both phrases or clauses, is following the verb/matrix clause:

P 1 V S 0 AdvP/AdvCl

Both types of adverbial expressions can also be put in front of the verb.

I will first discuss the discourse function of proposed temporal clauses in Coptic and their possible pragmatic function in the sense of FG. Then I will consider whether they can be described as P1 constituents of the matrix clause just as adverbial phrases can:

Pl ->S/O/ AdvP /AdvCl? In particular, I will discuss two notions in this context: recursion and syntactic integration.