Givon, Talmy [Givón];
Syntax: an introduction, Volume 1
J. Benjamins, 2nd ed. 2001 (1st: 1984), 500 pages [gbook]
ISBN 1588110656, 9781588110657
topics: | language | grammar | linguistics-functional | syntax
Argues for a functionalist view, that is opposed to structuralist models like Chomsky's. But Function and structure go hand in hand - if structures without function are plainly senseless, functions without structures are downright lame.
When the first volume was written in the early 1980s, I thought it was possible to treat grammar responsibly, in terms of both its adaptive motivation and typological diversity, without an explicit account of the more formal aspects of syntactic structure. These aspects — constituency, hierarchy, grammatical relations, clause-union, finiteness and syntactic control — were matters I took for granted but chose to defer. In retrospect, it was a bad mistake. My sins of omission were perhaps understandable as an over-reaction to the formalism of my early training. But the damage was real enough. What I took for granted but for the moment chose to ignore, others cheerfully discarded. While the chasm between the formal Generative approach to grammar and the adaptive perspective I pursue here seems at times unbridgeable, it can be narrowed down to a relatively small number of issues that are, in principle, empirical. ABSTRACTNESS: Does phrase structure pertain only to concrete entities (words, phrases, clauses), or also to semantic, pragmatic and sundry formal ghosts devoid of surface manifestation? MORPHOLOGY: Should constituency and hierarchy apply equally to grammatical inflections as to lexical words, or does a profound structural re-organization occur during grammaticalization? LINGUA MENTALIS: Is the mental representation of yet-to-be communicated thought syntactic, or is grammar a coding function that maps cognitive input onto syntactic output? GRAMMATICAL RELATIONS: Are subject and object universal primes, or are they derived from constituency and hierarchy? IDEALIZATION: Are rules of grammar, or the universals that license them, manifestations of some abstract Platonic competence, or are they performance constraints on the language processor? GENERATIVITY: Are rules of grammar 100% rigid, or can they still function as neuro-cognitive processing instructions with a certain measure of leakage? EMERGENCE: Are universals of syntax synchronic, or do they constrain first and foremost the possible pathways of evolution, acquisition and grammaticalization? Both the extravagantly abstract structures proposed by extreme formalists and the obsessively concrete structures embraced by extreme functionalists are due, in large measure, to pre-empirical philosophical predilections. For over the discordant din loom the twin towers of reductionism — Plato with his predilection for idealization and abstraction; the Positivists with their reflexive abhorrence of all invisible mental entities. Overthe din also looms, still, larger than ever, the paradoxical figure of Noam Chomsky; who in 1965 licensed semantically-motivated deep structure but in 1968 and 1970 reined us all in and reaffirmed the Saussurean dogma of autonomous syntax; who between 1957 and 1965 celebrated the reality of concrete syntactic structures, but in 1992 decreed them figments of the methodologist’s imagination. It is convenient to forget that the 1965-vintage Aspects model did proclaim, explicitly, the isomorphism between syntactic deep structure and propositional-semantic function, a dispensation that was tweaked to its logical conclusion two years later by Ross and Lakoff. It is just as easy to forget that even the most extravagant mad-hatter abstractness of Minimalism is still inspired by a germ of perfectly good intentions — the search for universals that are not about uttered concrete structures, but about the neuro-cognitive organization that accounts for such structures. That the carcasses of perfectly good intentions still litter linguistics’ road to, well, whatever, is a sad legacy of Platonism’s methodological indifference; in this case, the disdain for the tiresome process of discovery, a distant echo of Socrates’ frustration with expert knowledge. From whence flows the reification of competence and the mirage of generativity. The Devil is, alas, still, in the fine detail. The research program outlined here pays heed to Chomsky’s exhortation to seek universal principles, while affirming the mental reality of syntactic structures. For while not the real end of the theoretical enterprise, concrete structures are real enough. As theoretical entities, perceived and produced by the language processor. As methodological entities, grist to the mill of discovery and theory building.
Chapter 1: The functional approach to language and the typological approach
to grammar
The history of the functional approach to language:
Edward Sapir:
Language is a purely human and non-instinctive method of communicating
ideas, emotions and desires by means of a system of voluntarily produced
symbols... " [Sapir, Language, 1921 p.8]
Hence we have no recourse but to accept lg as a fully formed functional
system witin man's psyche or "spiritual" constitution. We cannot define it
as an entity in psycho-physical terms alone, however much the psycho-physical
basis is essential to its functioning. p.10-11, Language
Otto Jespersen:
The essence of lg is human activity -- activity on the part of one
individual to make himself understood by another, activity on the part of
the other to understand what was in the mind of the first... "
[Jaspersen The philosophy of grammar 1934, p.17)
George Zipf:
language is primarily a representation of experience. It may repr
experience as [in an account of a football game]. Or it may repr
tendencies to act representative of potential activity, such as an
oration to persuade others to modify their behaviour... a function of the
linguistic representation is to preserve or restore equilibrium. This
equilibrm may be of two types: a) inter-personal, and b)
intrapersonal...
Zipf, The psycho-biology of language, 1935, p.294-5
Ultimately, however, the best point of departure for functionalism is to be
found in biology, which has been profoundly functionalist for 2K years.
Aristotle ~single-handedly dislodged the two structuralist schools that had
dominated Greek biological thought up to his time. Both schools sought to
understand live organisms componentially, the way they did inorganic matter.
Thus Empedocles proposed to explain organisms by their component elements.
While Democritus opted for understanding organisms through their compoent
parts, or structure.
In his De Partibus Animalium, Aristotle first argues against Empedocles'
elemental approach, pointing out the relevance of histological and anatomical
structure:
"... But if men and animals are natural phenomena, then natural
philosophers must take into consideration not merely the ultimate
substances of which they are made, but also flesh, bone, blood and all
the other homogeneous parts; not only these but also the heterogeneous
parts, such as face, hand, foot... Does configuration and color
constitute the essence of the various animals and their several parts?
... No hand of bronze or wood or constituted in any but the appropriate
way can possibly be a hand in more than a name. For like a physician
in a painting, or like a flute in a sculpture, it will be unable to do
the _office which that name implies..." [McKeon ed. 1941 p.647]
"... if a piece of wood is to be split with an axe, the axe must of
necessity be hard; and indeed if hard, it must of necessity be made of
bronze or iron. Now exactly in the same way the body, which like the
axe is an instrument - for both the body as a whole and its several
parts individually have definite operations for which they are made;
just in the same way, I say, the body if it is to do its work, must
of nec be of such andd such character... " [650]
[Arbitrariness of grammar - rejected by Givon.
Aristotle's doctrine of the arbitrariness of the linguistic sign --
thus the arbitrariness of cross-language diversity -- refers only to the
coding of concepts by words. Latter-day structuralists unreflectively
extended the arbitrariness doctrine to grammar.
]
In the early 20th c. structuralism re-surfaced in the nascent social
sciences. To the infant disciplines of psychology, anthropology and
linguistics, two towering exponents of logical positivist philo of sci,
Russell and Carnap - sold the deceptive analogy of physics. from Carnap's reflections:
The thesis of physicalism as orig accepted in the Vienna Circle, says
roughly : Every concept of the lg of sci can be explicitly defined in
terms of observables; therefore every sentence of the lg of S is
translatable into a sentence concerning observable properties... " [Philo
of R Carnap 1963, p. 59]
CHOMSKY's ASPECT
The functional approach to grammar pursued here owes its direct descent,
somewhat paradoxically, to Chomsky's fatal indecision, in his
Aspects of the Theory of Syntax (1965) about the relationship between
syntax and meaning. While insisting on the arbitrariness and autonomy of
grammar, Chomsky nonetheless conceded that syntactic deep structuure (in
our terms here, the structure of simple clauses) was isomorphic to
propositional meaning (his logical structure). This crypto-functionalist
lapse, soon discarded by Chomsky himself, was embraced by Ross and Lakoff
(1967) and others who under the banner of Generative Semantics, spurned the
dogma of autonomous syntax.
Early functionalism focus on the reln between grammar and propositional
semantics:
[Chafe 70; Lakoff 70; Foley vanValin 84; Langacker 87,90]
But the realization that the bulk of the machinery of grammar was deployed in
the service of discourse pragmatics did not lag far behind
[Givon 79; Chafe 1980, 1994; Hopper ed. 1982; Tomlin ed 1987]
1.1 Perspective 1
1.1.1 Antecedence 1
1.1.2 Structuralism 4
1.1.3 Direct descent 7
1.2 The role of grammar in human information processing 7
1.2.1 The cognitive representation system 7
1.2.2 Peripheral sensory-motor codes 11
1.2.3 The grammatical code 11
1.3 Grammatical vs. pre-grammatical communication 13
1.4 The limits of conscious reflection 16
1.5 The theme-and-variation approach to syntax 18
1.6 The typological approach to grammar: Functional domains,
cross-language diversity and universals 20
1.6.1 The functional basis of grammatical typology 20
1.6.2 The diachronic underpinnings of grammatical typology 23
1.6.3 Typology and universals of grammar 24
1.6.4 Ranges of typological variation 25
1.7 Generativity and variation 26
1.7.1 The bounds of rule-governedness 26
1.7.2 Categories, continua and prototypes 28
1.8 Iconicity, naturalness and markedness 34
1.8.1 Iconicity vs. arbitrariness 34
1.8.2 Markedness, complexity and frequency 37
1.9 On the use, and uses, of this book 41
2.1 Preliminaries 43 2.1.1 Recapitulation: Meaning, information, discourse 43 2.1.2 Semantic features and semantic fields 43 2.2 Lexical vs. grammatical vocabulary 44 2.2.1 Words and morphemes 44 2.2.2 Morphemic status, morphotactics, and writing systems 48 2.3 Majorle xical word-classes 49 2.3.1 Membership criteria 49 2.3.2 Pragmatic limits of semantic taxonomies 50 2.3.3 Semantic criteria: An overview 50 2.3.4 Morphotactics: Affixes, clitics and inflections 54 2.4 Nouns 55 2.4.1 Semantic characterization 55 2.4.2 Syntactic characteristics 59 2.4.3 Morphological characteristics 60 2.5 Verb s 69 2.5.1 Semantic characterization 69 2.5.2 Syntactic characterization 69 2.5.3 Morphological characterization 69 2.6 Adjectives 81 2.6.1 Semantic characterization 81 2.6.2 Syntactic characterization 84 2.6.3 Morphological characteristics 85 2.7 Adverbs 87 2.7.1 Preamble 87 2.7.2 Semantic classification 88 2.8 Minor word classes 94 2.8.1 Preamble 94 2.8.2 Adpositions 95 2.8.3 Determiners 97 2.8.4 Independent subject and object pronouns 99 2.8.5 Inter-clausal connectives 99 2.8.6 Quantifiers, numerals and ordinals 100 2.8.7 Auxiliary verbs 102 2.8.8 Interjections 102
3.1 Preliminaries 105 3.1.1 Orientation 105 3.1.2 States, events, and actions 106 3.1.3 Semantic roles 106 3.1.4 Grammatical relations 108 3.1.5 Transitivity 109 3.2 Phrase structure and tree diagrams 110 3.2.1 What the formalism is good for 110 3.2.2 What the formalism is not good for 114 3.3 The classification of verbs and simple clauses 118 3.3.1 Dummy-subject verbs 118 3.3.2 Copularc lauses and copularv erbs 120 3.3.3 Simple intransitive verbs 125 3.3.4 Simple transitive verbs 126 3.3.5 Intransitive verbs with an indirect object 136 3.3.6 Bi-transitive verbs 141 3.3.7 Verbs with clausal (verbal) complements 149 3.3.8 Multiple membership in verb classes 160 3.4 Optional participant case-roles 161 3.5 Distributed lexical verbs 162 3.5.1 Complex predicates distributed across the clause 163 3.5.2 Multiple stems distributed across the verbal word 167
4.1 Orientation 173 4.2 Semantic roles vs. grammatical relations: The dissociation test 173 4.3 Formal properties of grammatical relations 175 4.3.1 Overt coding properties 175 4.3.2 Behavior-and-control properties 177 4.3.3 Universality and variation: The prototype approach to grammatical relations 194 4.3.4 Functional correlates of grammatical relations 195 4.4 The typology of case-marking systems 197 4.4.1 Preliminaries 197 4.4.2 Topicality and grammatical relations 198 4.4.3 Majort ypes of case-marking systems 200 4.5 The typology of direct objects 219 4.5.1 Overt coding properties: Degree of grammaticalization 219 4.5.2 Morphological promotion-to-DO and the topicality hierarchy 220 4.5.3 Promotion to DO and verb-coding of semantic roles 225 4.6 Serial verbs, case-marking and grammatical relations 227 4.7 Verb-coding of case-roles 230
5.1 Preliminaries 233 5.2 Scalarity, consistency and purity of types 234 5.3 Rigid word-order types 235 5.3.1 Rigid word-order in simple (main) clauses 235 5.3.2 Rigid word-order in the noun phrase 242 5.4 Mixed word-orders 246 5.4.1 Mixed clausal order 246 5.4.2 Mixed word-order in the NP 254 5.5 Word-order and bound morphology 260 5.5.1 Reorientation 260 5.5.2 Case-marking morphology on nouns or NPs 260 5.5.3 Tense-aspect-modal morphology 263 5.5.4 Negation morphemes 267 5.6 Flexible word-order 270 5.6.1 Reorientation 270 5.6.2 Types of flexible-order languages 271 5.6.3 The pragmatics of word-order flexibility 277 5.7 So-called non-configurationality 279
6.1 Introduction 285 6.2 Tense 285 6.3 Aspect 287 6.3.1 Perfectivity: boundedness and duration 287 6.3.2 Perfective vs. imperfective 288 6.3.3 The perfect 293 6.3.4 The immediate aspect: Remote vs. vivid 297 6.4 Modality and mood 300 6.4.1 Preamble: Propositional modalities 300 6.4.2 Epistemic modalities 301 6.4.3 The distribution of modality in grammar 302 6.4.4 Irrealis and the subjunctive mood8 312 6.4.5 Evidentiality and epistemic space 326 6.5 Communicative and cognitive aspects of tense, aspectand modality 329 6.5.1 Markedness 329 6.5.2 Frequency text-distribution 330 6.5.3 Cognitive considerations 332 6.5.4 The discourse correlates of aspect: Perfective/imperfective vs. simultaneous/sequential 333
7.1 Reorientation 337 7.2 The Creole prototype 338 7.3 T-A-M auxiliaries and their grammaticalization 340 7.4 The expansion of tense 341 7.5 The scope-of-assertion aspect 343 7.6 Perfective/imperfective-based systems 345 7.7 A four-way dichotomy 348 7.8 The conflation of perfect, durative and immediate 352 7.9 The grammaticalization of modality 355 7.9.1 Marking presupposed vs. asserted information 355 7.9.2 Split vs. uniform marking of irrealis 358 7.9.3 Unified vs. split subjunctive 360 7.9.4 The use of perfective markers in subjunctive forms 362 7.10 The grammaticalization of evidentiality 366 7.11 Primary vs. secondary grammaticalization of T-A-M 366
8.1 The multiple facets of negation 369 8.1.1 Negation and propositional logic 369 8.1.2 Negation and subjective certainty 370 8.1.3 The communicative pragmatics of negation 370 8.1.4 Negation as a speech-act 372 8.2 The cognitive status of negation 372 8.2.1 Change vs. stasis 372 8.2.2 The ontology of negative events 373 8.2.3 The ontology of negative states 375 8.3 Negation and social interaction 376 8.4 The scope of negation 378 8.4.1 Presupposition, assertion and negation 378 8.4.2 Negation and contrastive focus 380 8.4.3 Negation and optional constituents 381 8.5 The morpho-syntactic typology of VP negation 382 8.5.1 Grammaticalization of negation markers 382 8.5.2 De-verbal negation markers 382 8.5.3 De-nominal negation markers 383 8.5.4 Negation and word-order 385 8.6 Further topics in the syntax of negation 386 8.6.1 The coding of negative scope 386 8.6.2 Diverse negation markers across grammatical contexts 388 8.6.3 Emphatic orNP negation 392 8.6.4 Negation in main vs. complement clauses 393 8.6.5 Depth of embedding: Syntactic, morphological, and inherent (‘lexical’) negation 395 8.6.6 Negative polarity and levels of negation 396
9.1 Introduction 399 9.2 The semantic organization of pronominal paradigms 400 9.2.1 Pronominal classificatory features 400 9.2.2 Initial selected examples 401 9.3 Pronoun and grammatical agreement 407 9.3.1 Optional clitic anaphoric pronouns (Ute) 407 9.3.2 Pronominal agreement on the verb 408 9.4 The implicational hierarchies of pronominal agreement 416 9.5 Foundations of the grammar anaphoric reference 417 9.5.1 Preamble 417 9.5.2 Anaphoric zero vs. unstressed anaphoric pronouns 417 9.5.3 Anaphoric vs. stressed independent pronouns 418 9.5.4 Pronouns and zero anaphors vs. definite full-NPs 419 9.6 Explaining the rise of pronominal agreement 420 9.7 Head-modifierag reement in the noun phrase 426 9.8 Multiple functions of grammatical agreement 430 9.8.1 Verb-type and transitivity (Melanesian Pidgin) 430 9.8.2 Marking direct vs. indirect objects (KinyaRwanda, Lunda) 431 9.8.3 Marking topical objects (Amharic, Machiguenga) 432 9.8.4 Marking definite objects (Swahili) 433 9.8.5 Marking main vs. subordinate clauses (Bemba) 433 9.8.6 Marking existential-presentative clauses (KinyaRwanda) 433 9.8.7 Binding NP constituents together 434 9.9 Indefinite and non-referring pronouns 434
10.1 Introduction 437
10.2 Reference 437
10.2.1 The RealWorld vs. the Universe of Discourse 437
10.2.2 Referential intent 439
10.2.3 The universal quantifier and non-reference 439
10.3 The semantics of indefinite reference4 440
10.3.1 Referential opacity 441
10.3.2 Reference and propositional modalities 441
10.3.3 Gradations of referentiality 449
10.4 Grammatical marking of indefinite reference 450
10.4.1 Preamble 450
10.4.2 The numeral ‘one’ as reference marker 450
10.4.3 The demonstrative ‘this’ as reference marker 452
10.4.4 Noun classifiers and reference markers 453
10.4.5 Grammatical devices that mark non-reference 454
10.5 The pragmatics of indefinite reference: Denotation vs.
topicality 455
10.6 Definiteness 459
10.6.1 Definiteness and anaphoric reference 459
10.6.2 Grounding referents to the shared current speech
situation (working memory, current attention focus) 460
10.6.3 Grounding to shared generic-lexical knowledge (permanent
semantic memory) 460
10.6.4 Anaphoric grounding to the shared current text (long-term
episodic memory) 462
10.6.5 Proper names and global access 464
10.6.6 Generic subjects 465
10.6.7 Gradation of definite description 467
10.7 Grammatical marking of definite NPs 468
10.7.1 Distal demonstratives 468
10.7.2 Noun classifiers 469
10.7.3 Definite-accusative markers 470
10.7.4 Direct object and dative shifting 471
10.7.5 Topicalization and definiteness 472
10.7.6 Pronominal object agreement 473
10.7.7 Definiteness, reference and case-roles 473
10.8 The grammar of referential coherence as mental processing
instructions 474
Bibliography 479
Index 493
These two volumes are introductory in two distinct senses. First, they start from zero and assume no priorkno wledge of syntax, howeverbr iskly they may then proceed. Second, they remain primarily a survey of the phenomenology of grammar. In the process, they introduce the reader first to concrete morpho-syntactic structures and their semantic and pragmatic correlates. They then map out the likely bounds of cross-language typological variability. Out of this primal phenomenology emerge, gradually, some of the universal principles that govern both the functional and structural organization of grammar. Complex as this enterprise may seem, it is but a preliminary to the real task, that of elaborating the universal principles that govern the emergence of grammar as a language processor — in evolution, in development, in diachrony. I would like to dedicate these volumes to an old friend whose work at times seems just as paradoxical as my own; who shares both my passion for the incredible diversity of languages and my faith in the ultimate commonality of language; Ken Hale. Abrazos, Jefe. To Ken Hale, a linguist’s linguist, a cowboy’s cowboy.