A hierarchical approach to grammaticalization Kees Hengeveld.

A hierarchical approach to grammaticalization

Kees Hengeveld

Research questions

• Can Functional Discourse Grammar serve as a framework to predict, describe and explain processes of grammaticalization?

• What are the relevant processes of contentive change?

• What are the relevant processes of formal change?

• How do these processes interact?

2

Contents

1. Functional Discourse Grammar (FDG)2. Contentive change in FDG3. Formal change in FDG4. Contentive change and formal change in

FDG5. Conclusions

3

1. Functional Discourse Grammar

Conceptual Component

Contextual

Component

Articulation

Expression Level

Prosodic Contours,Sounds

Frames, Lexemes, Operators

Templates, Grammatical elements

Pragmatics, Semantics

Formulation

Encoding

Morphosyntax, Phonology

Grammar

Output


Contextual

Component

Articulation

Expression Level





Formulation

Encoding


Grammar

Output


Contextual

Component

Articulation

Expression Level





Formulation

Encoding


Grammar

Output

Frames, Lexemes,Primary operators

Templates,Auxiliaries, Secondary operators

Interpersonal Level

Representational Level

Formulation

Morphosyntactic Encoding

Morphosyntactic Level

Phonological Encoding

Phonological Level

Prosodic patterns,Morphemes, Tertiary operators

Frames, Lexemes, Primary operators


Interpersonal Level


Formulation




Phonological Level




Interpersonal Level


Formulation




Phonological Level




Interpersonal Level


Formulation




Phonological Level


12

Interpersonal Level

(π M1: [ Move

(π A1: [ Discourse Act

(π F1) Illocution

(π P1)S Speaker

(π P2)A Addressee

(π C1: [ Communicated Content

(π T1)Φ Ascriptive Subact

(π R1)Φ Referential Subact

] (C1)Φ Communicated Content

] (A1)Φ Discourse Act

] (M1)) Move

13

Representational Level(π p1: Propositional Content

(π ep1: Episode

(π e1: State-of-Affairs

[(π f1: [ Configurational Property

(π f1) Lexical Property

(π x1)Φ Individual

] (f1)) Configurational Property

(e1)Φ]) State-of-Affairs

(ep1)) Episode

(p1)) Propositional Content

2. Contentive change

Scope increase (layers)

Semantic units develop diachronically from lower to higher layers, and not the other way round (Hengeveld 1989)

Representational Level: p ← ep ← e ← f

15


Spanish haber ‘have’ (Olbertz 1993)

1. resultative, now replaced by tener:

Tengo prepara-d-a unacena fenomenal.

have.PRS.1.SG prepare-ANT-F.SG INDEF.SG.F meal(F)terrific

‘I have a terrific meal ready (for you).’

16


Spanish haber ‘have’

2. anterior

Había / he / habré preparado have.PST.1.SG/ have.PRS.1.SG / have.FUT.1.SGprepare-ANT

una cena fenomenal.INDEF.SG.F meal(F) terrific

‘I had/have/will have prepared a terrific meal.’

17



3. (recent) past

Me he levanta-do a las siete. 1.SG.REFL AUX.PRS.1.SG get.up-ANT at the seven‘I got up at seven o’clock.’

18



4. mirative (Ecuadorian Highland Spanish, Olbertz 2009)

Mire, compró estos, los probé ... y ..

Look bought.PF.3SG these them tried.PF.1SG and ¡han sido peras!have.3PL been pears

‘Look, she bought these, I tasted them ... and ... they are pears!’

19



p ← ep ← e ← fp ← ep ← e ← fp ← ep ← e ← fp ← ep ← e ← f

20


Pragmatic units develop diachronically from lower to higher layers, and not the other way round

Interpersonal Level: M ← A ← C ← R ← T

21


sort of (Hengeveld & Keizer 2009)

I keep sort of thinking about that and coming back to it. (Google)

I think I can more or less understand in general terms what happens up until sort of the impressionist time, maybe just post-impressionist. (BNC)

McCain backtracks on gay adoption, sort of. (Google)

22


sort of

M ← A ← C ← R ← TM ← A ← C ← R ← TM ← A ← C ← R ← T

23

Scope increase (levels)

Semantic units may develop diachronically into pragmatic units, and not the other way round (Hengeveld & Wanders 2007)

Interpersonal Level↑


24


RL: Providing food assistance is not easy because the infrastructure is lacking.

IL: Watch out, because there is a bull in the field!

RL: Providing food assistance is not easy exactly because the infrastructure is lacking.

IL: *Watch out, exactly because there is a bull in the field!

25


Semantic units may develop diachronically into pragmatic units, and not the other way round



26


Semantic units may develop diachronically into pragmatic units, and not the other way round



27

From lexeme to operator

Goossens (1985), Olbertz (1998), and Keizer (2007).

π ← Lexeme

28


fail to (Mackenzie 2009)

π ← LexemeHe failed to win the race.The bomb failed to explode.

fail (fc)(neg fc)

29


decir (Olbertz 2005, 2007; Grández Ávila 2010)

π ← LexemeThey say (dicen que) Juan is ill.Juan apparently (dizque) is ill.

decir (C) (Rep C)

30

Contentive change in FDG

31

Lexicon Lex Lex Lex Lex Lex

↓ ↓ ↓ ↓ ↓Interpersonal Level M ← A ← C ← R ← T

↑Representational Level

p ← ep ← e ← fc ← fl

↑ ↑ ↑ ↑ ↑Lexicon Lex Lex Lex Lex Lex

Contentive change in FDG: haber

32







33







34







35







36







37







38







39







40






Contentive change in FDG: sort of

41







42







43







44







45







46







47






Contentive change in FDG: because

48







49







50






3. Formal change in FDG

Main issue

There cannot be a one-to-one relation between formal changes and layers/levels, as lexical elements may enter the grammatical system at any layer/level

52

Grammaticalization scales

inflectional affix < clitic < grammatical word < content item

but: isolating vs. agglutinative vs. fusional languages

53

A scale of formal change in FDG

Keizer (2007)

lexemes (xi: – man – (xi): – old – (xi))

‘the/an old man’lexical operators (that xi: – man – (xi))

‘that man’operators (1 xi: – man – (xi))

‘a man’

54

Formal categories in FDG

Criteria:

lexemes: modification:an extremely old man

lexical operators:focalization(which man?) THAT

manoperators: neither

55

A grammaticalization scale in FDG

operators < lexical operators < lexemes

56

4. Contentive and formal change in FDG

Linking the scales

Each of the contentive parameters can be linked to the formal parameter to provide a more coherent view of the interplay between contentive and formal aspects of grammaticalization processes

58

Linking the scales

contentive scale:p ← ep ← e ← f

formal scale:operators < lexical operators < lexemes

As elements move up the contentive scale, they cannot move down the formal scale

59

Linking the scales

Allowed:



60

Linking the scales

Not allowed:



61

Linking the scales

contentive scale:M ← A ← C ← R ← T

formal scale:operators < lexical operators < lexemes

As elements move up the contentive scale, they cannot move down the formal scale

62

Linking the scales

Allowed:

M ← A ← C ← R ← T


63

Linking the scales

Not allowed:

M ← A ← C ← R ← T


64

5. Conclusion

Conclusions 1

FDG offers a framework within which known processes of grammaticalization can be captured

Contentive changes are restricted in terms of the hierarchical relations between layers and levels

Formal changes can be captured in a crosslinguistically valid way by adopting Keizer’s grammaticalization scale rather than traditional ones

66

Conclusions 2

Contentive and formal scales can be linked by defining a relative rather than absolute relationship between them

67

this presentation downloadable fromhome.hum.uva.nl/oz/hengeveldp

A hierarchical approach to grammaticalization Kees Hengeveld.

Documents

primary operatorstemplates

tertiary operatorsframes

fdgformal change

grammatical elementspragmatics

representational level

contentive changescope

propositional content

interpersonal level