working papers in functional grammar

working papers in functional grammar
wpfg no. 79
August
2004


Morphology in Functional Discourse Grammar
Edited by Kees Hengeveld
University of Amsterdam

---

ii

---

Introduction

This collection of short papers is devoted to the treatment of a number of morphological
issues in Functional Discourse Grammar (FDG). The first three contributions are
concerned with the characterization of morphological types in terms of the interaction
between the pragmatic and semantic components on the one hand and the
morphosyntactic and phonological ones on the other: Kees Hengeveld gives a general
overview of the treatment of morphological types in FDG, Maria Mos then discusses a
number of morphemes in the agglutinating language Tarma Quechua in more detail, and
Marleen van de Vate does the same for a number of grammatical particles in the isolating
language Saramaccan. The last paper, by Suzanne Dikker, goes into the problem of the
treatment of semantic versus grammatical agreement within the context of FDG.

The papers in this volume are the outcome of a research tutorial organized within the
context of the Research Master in Linguistics of the Graduate School for Humanities of
the University of Amsterdam.
iii

---

---

Contents

Morphological types in Functional Discourse Grammar
Kees Hengeveld 1

Tarma Quechua morphology in the FDG model
Maria Mos 11

Saramaccan: Grammatical particles in an isolating language
Marleen van de Vate 23

On the whereabouts of gender and number agreement: Location and accessibility
Suzanne Dikker 35
v

---

---

1. Morphological types in Functional Discourse Grammar

Kees Hengeveld
Department of Theoretical Linguistics, University of Amsterdam


1.1. Introduction

The aim of this paper is to give an overview of the way in which the differences between
morphological types can be accounted for systematically in terms of the various modules
that make up the model of Functional Discourse Grammar. After summing up the most
relevant aspects of FDG in section 1.2, a classification of morphological types is given in
section 1.3. Section 1.4. then reviews the way in which FDG handles each of these types
separately. Section 1.5. summarizes the paper.


1.2. Outline of the FDG model

Figure 1 gives a general overview of the FDG model. A summary of the various
properties of this model may be found in Hengeveld (forthcoming); a full presentation of
the model will be given in Hengeveld & Mackenzie (in preparation). An important
property of the model in the context of the present discussion is that it distinguishes an
interpersonal, a representational, a structural, and a phonological level of linguistic
organization, and that each of these levels is built up using different sets of primitives.
The interpersonal and representational levels of organization are structured on the basis
of pragmatic and semantic frames, into which lexemes and primary operators (i.e.
operators that are defined in terms of their meaning) are inserted. The structural level is
organized in terms of morphosyntactic templates, into which, apart from lexical material
from the preceding levels, grammatical words and morphosyntactic secondary operators

---

2 Kees Hengeveld
(i.e. operators anticipating bound grammatical expressions) are inserted. The
phonological level, finally, is organized in terms of prosodic patterns, into which, apart
from the lexical and grammatical words from the preceding levels, bound morphemes
and phonological secondary operators (i.e. those anticipating acoustic effects of certain
morphosyntactic configurations) are inserted.

Figure 1. Outline of FDG



Frames
1
Formulation
Lexemes
Primary Operators
3
2


(A1: [ILL (P1)S (P2)A (C1: [(T1) (R1)]
(C1))] (A1))
(Interpersonal Level)
5 4
[(p1: [(e1: [(f1) (x1)] (e1))] (p1))]

(Representational Level)
6
7
Templates
8
Morphosyntactic Encoding
Auxiliaries
Secondary Operators
9

[[[lexeme

Adj]ModP lexemeN]RefP
[lexeme

V [lexemeAdv]ModP]PredP]CL
(Structural Level)
10
11
Prosodic Patterns
Phonological Encoding
Morphemes
12
Secondary Operators
/ xxx#XXX#xxx#XXX \ /

(Phonological Level)

---

Morphological types in Functional Discourse Grammar 3


Free grammatical morphemes have to be introduced at the structural level, since, unlike
bound grammatical morphemes, they occupy slots in the syntactic configuration, which is
determined at this level. Bound grammatical morphemes are introduced at the
phonological level since in many languages the form of grammatical morphemes may be
affected by the syntactic configuration in which they occur. Therefore morphosyntactic
secondary operators are inserted at the structural level in the appropriate position,
anticipating the morphological means of expression that will eventually be selected at the
phonological level.


1.3. Morphological types

Morphological types can be defined along two parameters: semantic transparency, and
synthesis. Along the first parameter one can distinguish isolating, agglutinating, and
fusional languages. Isolating languages are semantically transparent in the sense that
there is a one-to-one relation between a word and a unit of meaning, whereas in
agglutinating languages there is a one-to-one relation between a morpheme and a unit of
meaning. Fusional languages are semantically opaque, in the sense that there is no one-
to-one relation between a unit of form and a unit of meaning. Along the second parameter
one may distinguish between polysynthetic and non-polysynthetic languages.
Polysynthetic languages allow the presence of more than one lexical element within a
single word, non-polysynthetic languages do not. The two parameters are basically
independent of one another: the first has to do primarily with the status of grammatical
elements in the language, the second one with the status of lexical elements. As a result,
polysynthetic languages can be fusional or agglutinating just like non-polysynthetic
languages. The only restriction in terms of combinations of the two parameters is that a
polysynthetic language cannot at the same time be isolating. Note furthermore that many
languages exhibit features of more than one morphological type.

Examples from languages from these different types are given below. Fijian (1) is
an isolating language, Turkish (2) an agglutinating language, Spanish (3) a fusional

---

4 Kees Hengeveld
language, and Southern Tiwa (4) a polysynthetic language. The glosses clearly reveal the
morphological structure of the languages involved: in (1) the gloss is a word-by-word
translation, in (2) a morpheme-by-morpheme translation, in (3) a one-to-many
translation, and in (4) the gloss reveals the presence of two lexical elements. Note that (4)
is a case of syntactic incorporation, as the incorporated object is cross-referenced on the
verb.

Fijian (Milner 1972: 42)
(1) Mo dou
kauta mada yani na
cina.
IMP
2PAUC take
MIT
away
ART
lamp

'Take the lamp away.'
Turkish (van Schaaik p.c.)
(2) Anlı-y-abil-ecek-miş-im.
understand-y-ABIL-IRR-INFER-1.SG

'I gather I will be able to understand.'
Spanish
(3)
Lleg-ó.
arrive-IND.PAST.PF.3.SG
‘He/she/it
arrived.’
Southern Tiwa (Gerdts 1998: 88)
(4)
Te-shut-pe-ban
1.SG>PL-shirt-make-PAST
'I
made
(the)
shirts.'


1.4. The representation of morphological types in FDG

The application of the FDG model, and particularly its division of labour between the
various components, to the examples just given, leads to the following analyses of the
examples just given.

---