AJ 0-

an»? mm»

wt! 1‘ O

r 7-. a; svw.

 

9
.- ‘4 .
~...~--.

an..- _
'11

~wl:u
”Tm-..
u u...

3:...

t
- ‘ “.".'.:.:T:s-L~L

...

‘ MIL; .c...}.u
Q4

.i.- u.- 3;}; _

' “L.
...:.-
£1}: -..-...\. lug-
» -. . . 1 ,._
- .4 _ ~-v \ ,45‘q“ s‘.‘..a. 'u... ’9;
L» Z...-:a-.::‘::;-v~¢-J:EM 1; l « , . ' 0 xx.“ -u..~..fn-.‘..:_.:
HA. i J-7 ‘ “I5, 7.. .. U 9- ~' ' ~ w 4‘ y?" . "2.19:1.“
r..- ‘.‘ (m 1. . I; a. . . w

.5“
J. . --
"’1"... ~ 5 Vi‘ - ..
. , 7 ¢ 7.. --._-
.«.._~-n ’ ‘. , .4» .r . ‘ __-f;:__
.

:1; _-;;.&-~: 5 but?

i‘~\ ‘
.:".:";Cl 5,» .7». "was?

I.-~.L;.A

J.-

.3

t. " 15-77 .
‘i. Lhasa“. '1“-'-
.. i. . ‘s- ,

f" k? s 4

’ .5.-i‘;.1y ( ”

flu? '1 A -
1!

.(«v ”on.-

q; -‘-"~‘1".’ , , ‘_ L 3.3 ‘3' “5"“ as;
-J.‘-¢.-L ‘ .‘ 1“ 5"" "1"“. E‘ \J- “-
u. , ..
a g F L _ _ u .. kéz. v f;
‘7st ”:33” _‘ ,m... “m M “31.2%?
.. sno- Ag: , a» ‘ u ,0 .VL'aArJ
v‘r' . n d 'vx - ”9!:'fim "N
. ‘ W

' -..vx
2 x" .. .4...
1.

.t'
.,v
my: I

_.,-.

'4.’,..»:n--u“
:7

x. >
’E'.’ I”
>étflfl'. 51-‘
“r"f

t‘
..,.
47.-.," 5:. fi , , .2
(r 3:3"..‘A ‘F"“ .
,..~ :..,«.-‘h.“3—"

~rr

. .
3:3'.‘ , .. r
r-‘Zl‘l‘jt' “)‘I""'
.

 

"'-u
_....,

 

 

'A , yr”-
O-l‘ VvL’.

 

{Jain}

9; m."

>u p f _ 7 '1. ”5“ r
”A. -.>.- '. J... r- “ ‘
-~.< ‘5" ' ,7, ‘ . 3-, ”7 ’ N, 'K ' .""’ L515... rpv vqrrtu ‘-
' "' '1".

n!" _ __, a...»

5...".1
«-

. {‘55:

“rag:- ' f'

l,
. .u pus-1‘”
‘ :5...“-
"l

 

1...:59-

- 352;:7”;

7'on-4-‘s-1

-uL.:

M~wan§ii§h- A)
_.,-..~;‘w....
" 91"" Jun-a a
. ‘Ifitul‘lii Ac
1.2%.“. 5.3.5“.
. “W .33: .m....?;2

hug-me
uk -.55. ..
v Uni...

.Z‘i :sI-hr .
«Sup. - «A:

£21.,-

‘ 2‘
fi'.
‘ wv‘igiég

,2-

3;.9 :
L-J‘md “‘“'

«.41:

4

(.4.

‘th

, .m... . ’3 ..
m..-” . .
5»: ,.,,.‘..JL.»..1.-
m ,._..,,. ..
rm-.. :-
V
4-...

E NIVERSITY LIBRARIES

lllllllllllllllllll

ll l
iii 00900 0974

SAN 8T

"ll“llll

 

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled

A FUNCTIONAL ANALYSIS OF THE

CLAUSE STRUCTURE OF DARI
presented by

Linda Stump Rashidi

has been accepted towards fulfillment
of the requirements for

Ph.D. Linguistics

degree in

 

 

C
«g: ”3' /m%

 

Major professor

Date flfle 3/7/

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

__‘_.‘

7 T

LIBRARY
Michigan State
University

is. J

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before dde due.

I IE! I
ILJELJ
—7
ll

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU I. An Affirmative ActlorVEquel Opportunity Institution
chimeras-9.1

 

A FUNCTIONAL ANALYSIS OF THE CLAUSE STRUCTURE OF DARI
By

Linda Stump Rashidi

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Linguistics

1991

ABSTRACT
A FUNCTIONAL ANALYSIS OF THE CLAUSE STRUCTURE OF DARI

by
Linda Stump Rashidi

While individual languages pattern differently, all languages
share some central core of universal features. One area that holds
promise in a search for a universal base is the analysis of
functional relations. Systemic linguistics, which bases its
framework on functional relations, is an ideal theory from which to
work. Though many systemicists have been exploring these relations
in English, little research has been done on other languages. This
research investigates the functional structure of spoken Dari
(Afghan Persian). The major text studied is an historical narrative
told by a native speaker. The analytical focus is on the realization
of meaning within context. The purpose of the study is threefold: 1)
to explore universal notions, 2) to test the assumptions of systemic
linguistics, and 3) to describe the structure of Dari.

The data are analyzed from three perspectives: grammatical,
functional, and thematic. The grammatical analysis reveals four
clause types for Dari: relational, existential, transitive, and
intransitive. The functional analysis, following M.A.K. Halliday,
shows six distinct process types for Dari: material, mental, verbal,
identifying, attributive, and existential. The use of Halliday's
conceptualization of ergativity gives insight into the notion of
agent. The analysis of Theme/Rheme structure points to the

importance of Rheme as the core of the message, with Theme as the

more secondary component.
This analysis reveals areas of both similarity and difference

between English and Dari. As such, the study suggests some possible
areas where assumptions of universals, in general, and s’ystemics, in

particular, might be re-evaluated.

ACKNOWLEDGMENTS

I wish to express my appreciation to a number of people who
have been instrumental in various stages of this research from its
conception to preparation of the final manuscript. First of all, I am
grateful to M. Mehdi Aram whose narrative this is and who has so
wonderfully entertained us with his poetic storytelling. Then I
would like to thank Jan Firbas, whose lucid explication of Theme and
Rheme has greatly influenced my own thinking, for his insights into
the message structure of an earlier Dari text.

I am most appreciative of the guidance and true interest in my
work of my dissertation committee: Mutsuko Simon for her
willingness to absorb yet another theory and another language;
Carolyn Harford for her probing questions and insights from
generative theories; David Lockwood for his phonological assistance
and his astute contributions from the point of view of a
stratificationalist. Finally I would like to thank Ruth Brend, my
dissertation chairperson and constant supporter, who first
encouraged me to reach beyond my English confines and explore

another language and whose editorial skills are without match.

TABLE OF CONTENTS
I Introduction .......................................... 1

ll Background

1. Previous Persian studies ........................ 5
2. Functional studies of Persian .................... 6
3. Previous Dari studies ........................... 7
III The Hallidayan model of language
1. Introduction ................................... 8
2. Halliday's three metafunctions ................... 10
3. Clause as process: transitive view ............... 12
4. Clause as process: ergative view ................. 19
W Dari: grammatical structure
1. Dari as a distinct dialect ........................ 22
2. Basic typology .................................. 23
3. Grammatical properties of Dari clause structure . . . 23
4. Noun phrase .................................... 25
4.1 -ra .................................... 25
4.2 -haa ................................... 27
4.3 The ezafe ............................... 28
4.4 ke ...................................... 30
5. Verb phrase ..................................... 32
5.1 Tense and aspect ......................... 33
5.2 me- .................................... 33
5.3 be- .................................... 35
5.4 Perfect .................................. 36
6. Compound verbs ................................. 38

vi

7. Clause types .................................... 40
7.1 Relational clauses ........................ 41
7.2 Existential clauses ....................... 43
7.3 Transitive clauses ........................ 44
7.4 lntransitive clauses ...................... 47
7.5 Passive ................................. 48

V Transitive structure of Dari

1. Functional roles ................................. 52
2. The text ........................................ 53
3. Dari process types ............................... 54
4. Material processes ............................... 57
5. Mental processes ................................. 65
6. Verbal processes ................................. 69
7. Identifying processes ............................. 72
8. Attributive processes ............................ 75
9. Existential processes ............................ 78
10. Conclusion ..................................... 81

VI Ergative view of Dari
1. Traditional view of ergativity ..................... 83
2. Halliday's ergative interpretation of transitivity . . . . 85
3. An ergative view of transitivity structure in Dari . . . 85

4. Analysis of the data ............................. 87
4.1 Material processes ........................ 87
4.2 Mental processes ......................... 91
4.3 Verbal processes ......................... 92
4.4 Identifying processes ..................... 93
4.5 Attributive processes ..................... 94
5. Conclusion ...................................... 95

6. Some examples analyzed .......................... 96

v i i
Vll Theme/Rheme Structure

1. Introduction .................................... 99
2. Halliday's view .................................. 99
3. Firbas' view .................................... 100
4. Prague School foundation ......................... 101
5. Other interpretations: Theme and Topic ............ 105
6. Defining Theme and Rheme ........................ 1 06

VIII Theme and Rheme in Dari
1. Identification of Theme and Rheme ................ 1 1O
2. Theme and Rheme in Text B ....................... 1 13
3. Parameters of message structure in Dari ........... 1 14
3.1 Clauses without a Theme .................. 1 15
3.1.1 Existential ........................ 1 15
3.1.2 Ellipsis ........................... 116
3.1.3 Clauses without an ideational Theme . 118
3.1.4 Truly Themeless clauses ............. 1 19
3.1.5 Topic and Theme ......................... 120
3.2 Linear position ............................ 122
3.3 Given/New ................................ 124
4. Embedded clauses ................................ 126
5. Analysis of the text .............................. 128
6. Theme/Rheme chart .............................. 131

IX Summary and Conclusion

1. Summary of findings ............................. 138
2. Analysis across the three metafunctions ........... 140
3. Implications for systemic theory .................. 144
4 Implications for a theory of language universals ..... 146
5. Conclusion ...................................... 147
Appendix A: Text A; Text B; Text C ......................... 148
Appendix B: Theme/Rheme Structure: Text B ................ 164
Appendix C: Translation of the text ........................ 1 71

List of References ........................................ 1 7 5

LIST OF TABLES

Table 1: Process Type Distribution ......................... 57
Table 2: Material Processes ............................... 64
Table 3: Mental Processes ................................ 69
Table 4: Verbal Processes ................................ 72
Table 5: Identifying Processes ............................ 75
Table 6: Attributive Processes ............................ 78
Table 7: Theme/Rheme Chart .............................. 1 34

viii

LIST OF FIGURES

Figure 1: Halliday's English Process Types .................. 18
Figure 2: Halliday's System of Voice ....................... 20
Figure 3: Dari Verb Phrase ................................ 38
Figure 4: System of Transitivity and Voice for Dari ......... 86
Figure 5: Ergative Functions .............................. 96
Figure 6: Detailed Analysis of B47 ........................ 127
Figure 7: Detailed Analysis of 353 ........................ 127
Figure 8: Analysis Across Metafunctions ................... 143

KEY TO SYMBOLS AND ABBREVIATIONS
(Dari citations in bold face)

Glosses:

E - ezafe marker

BE = imperative, subjunctive, optative marker
ME = imperfective, durative marker
NEG = negative

PF = perfective, completive marker
PL a: specific plural marker

RA = definite direct ‘object marker
13 = first person singular

2s = second person singular

35 = third person singualr

1p first person plural

2p = second person plural

3p =- third person plural

Grammatical Symbols:

Ad =- Adjunct act - action verb

AP = adjective phrase b = budan

C = Complement 90ft = gotten

MP - modal particle me = mental process
NP = noun phrase v = verbal process
0 - Object s = sudan

0A a Object Appositive

P = Predicate

pp = prepositional phrase

S = Subject
SA = Subject Appositive
VP -= verb phrase

Functional Symbols :

A -= Actor
A0 = Accompaniment
Ag = Agent

Ar = Attribute

At = attributive process
Bn Beneficiary

Ca Carrier

Cr = Circumstance

En = Extent

Et = Existent

Ex .. existential process
G=Goal

l = identifying process

Other Symbols:

Gv = Given
Nw = New
Th = Theme
Rh = Rheme

Id = Identified

Ir = Identifier

Lc = Locaflon

Ma = material process
Md = Medium

Me = mental process
Mn = Manner
Ph = Phenomenon

R = Range

Se = Senser
Sy = Sayer

Tm = Time

V = verbal process

Vb = Verbiage

CHAPTER I

INTRODUCTION

The question of the universality of linguistic structures is
one that has long occupied many linguists in most linguistic
theories. While it is generally agreed that languages pattern
differently, in the final analysis theorists and practitioners
recognize that languages are more alike than they are different. As
a result, linguists are constantly searching for the commonality that
comprises the central core of language. Noam Chomsky, and the
diverse generativists whose theories have evolved out of his,
focuses his search on common grammatical structures that underlie
the different surface features of various languages. Sydney Lamb,
and the small but still active group of stratificationalists, has set
about to construct a cognitive model of linguistic structure.

Kenneth Pike, and the numerous practitioners of his tagmemic
theory, built a model of language based on a unified theory of
behavior. Michael Halliday, following in the tradition of JR. Firth
and the Prague School, has developed a sociolinguistic, systemic
model of language, where surface structures are viewed as
realizations of underlying systems.

One area that holds promise in this search for the common core

of language is that of the analysis of the logical relations of various

2

constituents to each other. Most theories of linguistic behavior have
explored, in one form or another, these more semantically-based
relationships. It is now generally acknowledged that syntactic
relations can only be fully understood within a framework of logical
relations. While along one parameter, clauses have subjects and
predicates, along another parameter, clauses also have such
constituents as agents and affected participants. These latter
relationships are often referred to as functional relationships
because they deal with how various constituent structures function
semantically in relation to each other and to the unit of language as
a whole.

While most linguistic theories today recognize the importance
of functional relations, systemic theory bases its very framework
in a notion of such logical relationships. These functional relations
have been, and are being, extensively explored for English. Halliday's
own work is rooted in English, and while he insists that he is making
no universal claims, he seems to be, nonetheless, propounding a
theory of language. If this theory is ultimately to have validity, it
must be broadened beyond its English boundaries.

This research investigates the functional structure of Dari, or
Afghan Persian. Though no universal claims are made here, the
exploration of the structure of language beyond English and within a
functional framework should broaden our understanding of how
language in general conveys meaning. The focus of this research is
on the nature of functional relations as being central to the
realization of meaning potential. Our particular interest is in how

various functional structures interrelate and influence each other.

3

While Dari is an lndo-European language, it is different enough from
English to make it an interesting testing ground for an exploration of

functional notions.

Methodology

Language is not a set of sentences; it is a system of meaning.
As a result, an attempt to truly understand language and how it
functions must consider language as it occurs in context. For this
reason, the data used for this analysis come from natural connected
speech. We are not looking for idealized sentence structure or
grammar, but for what real people say in real conversation. And
since Dari is essentially an oral dialect of Persian, the text ls
spoken colloquial Dari.

The major text is a thirty minute taped narrative by a single
informant. Though he now resides in the United States, he speaks
little English; he spent the first fifty years of his life in
Afghanistan, and his daily life is still conducted in his native
tongue. While this is not the perfect linguistic situation for
analysis of a language, there appears to be little English (or other
language) interference. The ideal informant, of course, would be an
Afghan villager who had never left Afghanistan or been exposed to
other languages, but the political situation in Afghanistan makes
this kind of data virtually impossible to obtain at present.

Afghans love to tell stories, and, in order to elicit as much
complete clause structure as possible, the informant was
encouraged to relate incidents or narrate events. The main text is a

rather rambling account of the history of Islam told within a casual

4

family setting. (For an English translation of the narrative, see
Appendix C.) These parameters, of course, affect the nature of the
language elicited. The casual setting helps to produce natural,
colloquial speech, but the informant was actively aware that he was
being recorded. In fact, Afghans take on a distinct performance
speech style under these conditions. It is this genre of speech that
is analyzed here.

The use of a single coherent text for this analysis was
deliberate. Though the data base has the disadvantage of being one
narrative by one speaker, it has the advantage of being consistent
and contextually-bound. In fact, the speaker himself was
deliberately chosen because he is considered by other native
speakers to be a particularly effective and fluent story teller. His
narrative is analyzed not as an isolated speech event, however, but
against the background of previous work on the texts of other
speakers (Rashidi 1987; 1988a; 1988b; 19880; 1989a; 1989b; in
press) and with the intuitive input of native speakers and my own
non-native feeling for the language obtained while living for three
years in Afghanistan. Data from texts other than the narrative under
consideration have been used for confirmation of generalizations

about Dari.

CHAPTER II

BACKGROUND ON PERSIAN

1. Previous Persian studies

Little work has been done on the clause structure of Persian,
and none, it appears, on that of Dari specifically. The major study
being done today on Persian ( at least in the West) is that of Gernot
Windfuhr and Gilbert Lazard. Windfuhr's W (1979) is
still definitive. In this work, Windfuhr describes and discusses the
basic grammatical structure of Persian, in general, but with
emphasis on the Persian of Iran and, more specifically, on standard
Tehranian Persian, the prestige dialect of the middle class educated
speaker. He gives us an overview of how various structures function
grammatically, his emphasis being clause structure and his model
being traditional structural linguistics. This work is an excellent
generalized base from which to work in exploring specific varieties
of Persian, as it probes most of the complex and puzzling features of
Persian grammar positing possible explanations for variation in
usage.

A more recent publication is Windfuhr's chapter on Persian in
Bernard Comrie's W (1987). In this work,
Windfuhr points out major problem areas (genericity and plurality,
definiteness, tense vs aspect, relative clauses, to name a few),
discusses the influence of Arabic, and distinguishes the three major

dialects (Iranian, Dari, and Tajik). The one major shift from his

6

W is the refinement of his discussion of the verb
phrase. In the more recent work, tense and aspect are seen as
equally basic categorical vectors. Here Windfuhr posits three verb
stems: present, perfect and aorist, reinstituting the aorist as a
distinct category. Windfuhr discusses this model for tense, aspect
and mood in detail in a 1985 EQLiLLjngu'szjga article. At present, he
is working on ergativity in Iranian dialects and diachronic typology
(personal communication, March 1990). His longitudinal and
intensive study of literary Iranian Persian makes an insightful
foundation for the study of spoken, colloquial Dari.

Gilbert Lazard works mainly on particular problems of Persian
grammar. His most recent known area of investigation is the
morpheme -ra (1982), usually designated as an object marker.
Lazard's hierarchy of conditions for the presence of -ra tells us
much about its complex usage as well as giving insight into the
underlying complexity that is beneath Persian's surface simplicity.
In addition, Lazard has briefly investigated Tajik (1970), the other
Eastern dialect besides Dari.

2. Functional Studies of Persian

Only two investigations of the functional grammar of Persian
appear to exist. Both are Ph.D. dissertations done by native Iranian
speakers: Ali Asghar Aghbar (1981) and Seyed-Ali Miremadi (1981).
Both are based on Walter Cook's case grammar matrix model (1979),
and both draw their data from literary texts and 'made-up' examples,
relying heavily on the authors' native intuitions (and, therefore,

competence) concerning the structure of Persian. Miremadi's

7

dissertation is basically an apologia for Cook's model. Aghbar's
dissertation, while also using Persian to support Cook's model, gives
an indepth look at the Persian verb system. Both are useful for their

native insights into Persian.

3. Previous Dari studies

Work on Dari itself is sparse and confined to practical guides
of Dari grammar for second language users. The most thorough of
these guides is M. Ehsan Entezar's WM produced
in 1964 as a manual for Peace Corps volunteers. Its very pragmatic
origins belie its firm linguistic underpinnings. Entezar's background
in linguistics is manifestly apparent in the structural orientation of
the manual. Though this is not an empirical investigation of Dari,
the contents illustrate and concisely comment on the structural
patterns of the language. This manual is a significant contribution
to our understanding of Dari.

The bulk of the work on Dari being done today is in Russian, but
even this seems to be of a pragmatic bent. There appear to be no
major analytical investigations of Dari, specifically, being
undertaken at present.

CHAPTER III

THE HALLIDAYAN MODEL OF LANGUAGE

1. Introduction

Because this research investigates the functional structure of
Dari, mainly from the viewpoint of systemic theory, an initial look
at the basic tenets of that theory is appropriate. Systemic
linguistics is not well known outside of Great Britain. In recent
years, however, systemic philosophy has spread to some degree to
other European countries and Australia, and to a lesser degree to
North America. The basic tenets of the theory are socioculturally-
oriented, making the theory of particular importance to applied
fields. Less extensive work has been done on a purely theoretical
level. Nonetheless, systemics offers a fresh alternative to
generative theories which are cognitively based.

Systemic linguistics is functional in approach. It views
language holistically and within context. While there is a specific
semological stratum, there is no attempt to separate semantics
from the other levels, particularly the grammatical level. Halliday,
in particular, has been much criticized for this 'fuzziness', but the
entire theory is, in fact, grounded on the infusion of meaning at
every level. Halliday states in his WWW
Grammar: ”there is no clear line between semantics and grammar,
and a functional grammar is one that is pushed in the direction of

the semantics" (1985zxix). Any attempt to view pieces of language

9

in isolation inevitably distorts because language can only be
understood as a whole system.

In systemic theory, following Hjelmslev, language is seen as a
process, not as a thing or a product. Language is a system of
meaning which is realized through form. Thus, a text is
fundamentally a semantic unit rather than a grammatical unit.
Systemic linguistics is a relational theory where the relationship
between semantics and grammar is one of realization.

Systemics is also a top down theory that starts from the
general and works down to the specific (known as degrees of
delicacy). Halliday sees the sentence as a “significant border post"
(1985: xxi). Above the sentence, non-constructional forms of
organization take over (i.e. grammar has little functional role).
Below the sentence, relationships are constructional ones (i.e.
grammar carries a significant meaning load).

Michael Gregory, a systemic linguist who has been specific-
r ally analyzing the structure of discourse, distinguishes between
discourse structure and grammatical structure in this way:
Discourse structure is not rule-based. It is dependent on culture
rather than being language specific. This is in contrast to grammar
at the sentence level and below which is rule-based and language
specific. Gregory's model of discourse structure, which he calls
phasal structure, deals with the structure of large units, but this is
not hierarchically parallel to the grammar of smaller units. Phases
involve realization directly, as opposed to sentence—level grammar
which is realization of an underlying system (Gregory 1967 and
1935). '

1O

2. Halliday's three meta-functions

While the sentence is the upper boundary of grammar (or
syntax), “the fundamental unit of organization is the clause"
(Halliday 1985: xxi). Systemic-functional grammar is anchored by
the clause, because, as Halliday puts it, it is at the clause level that
we have the most material with which to crack the code (which is
the goal of grammar). The clause can be viewed from a variety of
perspectives. It is not only organized tri-stratally (semology,
morphosyntax, phonology/graphology) but also tri-functionally.
Functionally, Halliday sees the clause as a unit of language "in which
meanings of three different kinds are combined” (1985: 38). These
are Halliday's three meta-functions of language: the textual, the
interpersonal, and the ideational. He contrasts these meta-
functions by use of the late nineteenth century concepts of
psychological subject, grammatical subject, and logical subject,
stating that while usually the three notions are conflated into a
single element, they are semantically distinct. He illustrates their
distinctness in the following example:

(1) this teapot my aunt was given by the duke

The first kind of semantic organization, the textual
component, is thematic structure or the clause as message. From a
Hallidayan point of view, the Theme (psychological subject) of the
clause is what the clause is going to be about, the starting point of
the message, and is realized in English in clause-initial position.
What is left over after Theme is identified is Rheme. In the example
above, this teapot is the Theme and the rest of the clause is the
Rheme.

11

The second kind of semantic organization of the clause, the
interpersonal component, is meaning as exchange, an interactive
event between speaker or writer and audience. This is the
propositional structure of a clause. The Subject (or grammatical
subject in traditional terms) is ”the element that is held
responsible” (1985: 37), or more traditionally 'that of which
something is predicated'. The Subject combined with the Finite
element (the verbal operator expressing tense or modality) forms a
single constituent which Halliday calls Mood. The remainder of the
clause is termed the Residue. In (1), the Subject is my aunt, the
element about whom the statements validity rests.

The third aspect of clausal meaning, the ideational component,
is clause as representation. This is meaning in the sense of content;
and it is here that Halliday fully develops his notion of clause as
process. A process consists of potentially three components: 1) the
process itself, 2) participants in the process, and 3) circumstances
associated with the process but not 'essential' to the clause. In (1),
was given is the process; the duke is the 'logical' or functional
subject, in this case, what Halliday terms the Actor; the teapot is
the Goal; and my aunt is the Recipient.

And so, while in English clauses, all three of these 'subjects'
are often mapped onto the same constituent, the notions Theme,
Subject and Actor represent different kinds of meaning: Theme is a
function of the clause as message; Subject is a function of the
clause as exchange; and Actor is a function of the clause as
representation (1985: 37). In the 'teapot‘ clause, the three functions

are manifested by separate elements.

12

this teapot / my aunt l was given I by the duke
Theme Subject Actor

(psychological) (grammatical) (logical)

3. Clause as process: transitive view

Halliday characterizes the clause in its ideational function as
a transitive interpretation of the clause. From this point of view,
clauses are distinguisted by process type. Associated with and
defined by the process are the various participants. In addition to
the central participants in the process, there may be circumstances
, associated with the process. These three components, process,

' participant, and circumstancial element, provide the frame of
reference for interpreting our experience of what goes on (Halliday
1985: 101). Halliday goes on to state: "The concepts of process,
participant and circumstance are semantic categories which explain
in the most general way how phenomena of the real world are
represented as linguistic structures” (1985: 102). Each language, of
course, represents reality in a different way, but there is certainly
some universal core to this representation.

Halliday, as he has refined his notion of process, has organized
and reorganized both the number and categorization of English
process types. In his Wm, which has
become the reference manual for most systemicists, Halliday puts
forth three major types and three minor types. The three principal
types are: 1) material, 2) mental, and 3) relational. In addition, he
recognizes the subsidiary types: 1) behavioral, 2) verbal, and 3)
existential.

13

Probably the major way that language structures experience is
through some kind of action process. This Halliday terms a material
process, a process of doing. We generally think of 'doings' as
concrete, physical events like running, hitting or throwing, but they
may also be more abstract doings and happenings like resigning from
office or dissolving a committee. In any event, every material
process involves an Actor, the participant that is the 'logical'
subject or the doer of the action. The Actor of the process need not
be the Agent. Halliday reserves the term agent for another use.
From a transitive point of view, the Actor could be a participant
such as the boat in the boat sailed or Jack in Jack fell down,
where agency is, at best, ambiguous. In addition to an Actor, some
material processes contain a second participant, a Goal. The Goal is
the participant toward which the action is directed. In (1), the
teapot is the Goal. If there is a third participant, it will be the
Recipient, as my aunt is in this teapot my aunt was given by
the duke. The notion of, material process seems fairly universal,
though the participants in the process may be more language
specific and certainly their particular realization will be.

As a process becomes more abstract, the distinction between
Actor and Goal begins to blur, the Actor becomes more involuntary,
and the process is less a doing than a happening. To better account
for this kind of process, Halliday has posited a subsidiary process
called behavioral. These are processes of physiological and
psychological behavior like coughing, laughing or watching. The
Behaver, the obligatory participant, is typically a conscious being.

Generally, these are one-participant processes, but occasionally

14

there will be an explicit Behavior such as a sigh of relief in he
heaved a sigh of relief.

The second major process is that of mental, the process of
sensing. As Halliday points out, most of the time we do not talk
about such active, concrete processes as hitting or falling
(19852105). We talk about our feelings, likes and dislikes, what is
good or bad; these kinds of processes are more abstract and require
a different functional interpretation. Mental processes are, in fact,
not only semantically different from material processes but are also
grammatically distinct. Halliday gives five criteria for a mental
process: 1) one participant is human (or embued with humanlike
perceptions); 2) the second participant may be a 'thing' (as in
material processes) or a 'fact' such as something had happened in
he sensed something had happened; 3) in English the unmarked
present tense in a mental process is simple present, 9.9. she likes
the gift, while in a material process it is present progressive, 9.9.
I'm going home; 4) mental processes are two-way processes, i.e.
either the Sensor or the Phenomenon can be the Subject of
semantically equivalent active clauses (e.g. she likes it; it
pleases her); 5) mental processes are not processes of doing and,
therefore, cannot be probed by do ; one cannot ask the question what
did John m; and get the answer he knew the material.

The third major process type is relational, the process of
being. "The central meaning of the clause is that something is"
(1985: 112).

(2) John is tall.

(3) Sarah is the leader.

15

Halliday delineates various ways of expressing being in
English, grouping them into two modes: attributive and identifying.
Both modes require two participants. In the attributive mode, one
participant is the Attribute, the other is the Carrier. In (2), John is
the Carrier while tail is the Attribute. In this mode, an attribute or
quality is ascribed to an entity. In the identifying mode, an entity is
used to identify another entity; these functions are labelled
Identifier and Identified, respectively. In (3), Sarah is the
Identified, the leader is the Identifier. Identifying processes are
reversible, whereas attributive processes are not. Thus, in (3), the
leader can become the Identified by putting the leader In subject
position: the leader is Sarah. The same is not true for (2).

This reversibility may be peculiar to English where word order
carries a heavy semantic load and grammatical agreement is limited.
Also, identifying clauses have passive counterparts, while
attributive clauses do not. Again, it is not so apparent that this
distinction will carry over into other languages.

In addition to behavioral processes, Halliday proposes two
other minor process types: verbal and existential. Verbal processes
are processes of saying. Halliday interprets 'saying' in a broad
sense, including such symbolic 'sayings' as the clock says half
past nine. The Sayer, then, need not be human. Besides the Sayer,
there is the verbalization itself which Halliday terms the Verbiage.

The Verbiage can be either a thing or a proposition.

(4) he told Later! (thing)
(5) he said W (Proposition)

16

In addition, there may be a Recipient: e.g. he told me a
story.

Existential processes are one-participant processes, the
meaning of which is that something exists or happens. The Existent
may be a thing, e.g. there's ”mm, or a phenomenon such as an
event, e.g. there's W or it's raining. In English,
these clauses typically begin with a dummy 'there' or 'it' followed by
the copula and then a noun phrase expressing the Existent.

Existential processes may be particularly variable among languages,
in part because the necessity to fill the subject slot is not a
language universal. Dari does not, for example, have dummy
subjects.

Permeating all of the process types are two additional
participant functions: Beneficiary and Range. The Beneficiary is, as
its name implies, the participant for whom the process is taking
place. In material processes, the Beneficiary is the Recipient or the
Client. In verbal processes, the Beneficiary is the one being
addressed, the Receiver. Beneficiaries sometimes occur in
attributive processes as well, the typical example being: she made
him a good wife.

Range specifies the scope of the process. Typical examples
are:

(6) play 19.0.an

(7) give LIQQJS

(8) sins 3.19.09

(9) climb a_moumajn

17

Range is either an entity that exists independently of the
process but defines the domain of the process, eg. a mountain, or
expresses the process itself, e.g. tennis, a look, a song. English is
fond of shifting the action to the noun phrase and making the verb
contentless as in give a look. Other languages may or may not use
this device to the same extent or at all. Range may occur in
material, behavioral, verbal or mental processes. In mental
processes, however, Range conflates with the Phenomenon and,
therefore, does not have a unique function.

In addition to direct participants, clauses may contain various
circumstantial elements. These elements are peripheral
grammatically in that they are not necessary for the completion of
the proposition. In English, they are typically realized as adverbials
or prepositional phrases. Circumstantials may express: extent or
location in time or space, manner, cause, accompaniment, matter or
role.

Extent or Location in time or space
(10) he traveled Miles
(11) we mu arrived

Manner (how, with what)

(12) he beat the dog W

Cause (why, how, for what, for whom)

(13) he died ”mun

Accompaniment (with what/whom)

(14) he came M
Matter (what about)

(15) I'm worried w

18

Role (what as)

(16) I say this MM

Halliday sets up this model specifically for English, but the
underlying concepts, if not the realizations, seem to have universal
implications. Other languages will certainly embody the same types
of functions in different ways, distribute them differently, and may
even have different basic process types. As Dari is a language far
removed from English both socio-culturally and in terms of origin
and structure, it will most certainly show some variation from
Halliday's description of the transitivity system of English. But Dari
is also an lndo-European language and shares a common base with

English even beyond any universal linguistic claims.

PROCESS PARTICIPANTS
Material: Actor; Goal; Beneficiary; Range
Behavioral: Behaver; Behavior
Mental: Senser; Phenomenon
Verbal: Sayer; Receiver (Beneficiary); Verbiage
Attributive: Carrier; Attributor; Attribute; Beneficiary
Identifying: Identified; Identifier
Existential: Existent

Figure 1

Halliday's English Process Types

19

4. Clause as process: ergative view

Though traditional Western linguistics has divided languages
into transitive and ergative, Halliday sees the transitive/ergative
distinction as one of point of view. He argues rather persuasively
that English is more profitably viewed from an ergative perspective;
that is, the distinction in English clauses is not one of
transitive/intransitive contrast where the variable is one of
extension, but ergative/non-ergative where the variable is one of
causation. For example, the pair the tourist hunted/the tourist
hunted the lion forms an intransitive/transitive contrast; in both
the tourist is the agent, but in the second clause the process has
been extended to a second participant, the undergoer of the action,
the lion. In the pair the tourist woke/the lion woke the
tourist, the contrast is an ergative one; in both cases the tourist
stops sleeping. The variable is that which brought the process
about; the essential key element in the process, the tourist, is the
same in both clauses. Halliday claims that an examination of the
lexicon of English shows that the majority of high frequency verbs
yields pairs of clauses where the contrast is an ergative one rather
than a transitive one (1985:145).

Looked at from an ergative point of view, all processes are
alike; that is, they have the same functional roles. Every process
has associated with it a key participant which Halliday terms the
Medium. This is the participant without which the process would
not exist, the essential element. The Process and the Medium, then,
come together to form the core of the clause. In English, the Medium

is the only element that always participates directly in the process,

20

and, therefore, can never be introduced by means of a preposition
(Halliday 19852147).

Halliday represents the ergative point of view visually as a
nucleus encircled by an inner ring as well as an outer ring
(1985:147). The nucleus is ‘Process + Medium’; the inner ring
consists of additional participants; the outer ring is composed of
more peripheral circumstances. While the Medium is the core
participant, three other participant functions may be involved in the
process (the inner ring). The most crucial of these is an external
causer which Halliday labels Agent. This notion of Agent transcends
all transitive process types.

(17) leinn awoke the tourist (material)

(18) mm choked Mary (behavioral)

(19) W convinced me (mental)

(20) mm dried the clothes (relational)

This interpretation allows a clear and unified system of
voice to be set up (Halliday 1985:151):

..___ middle

clause as
representation
acfive

 

—— effective ——;[:
passive

Figure 2

Halliday's System of Voice

21

A clause with no feature of 'agency' is middle voice. A clause
with agency is effective voice; active if the Subject is Agent,
passive if the Subject is Medium. One needs to note that the
middle/effective distinction. is one based on agency and not, strictly
speaking, iAgent. In English, it is possible to have clauses that
involve agency but lack the specification of an overt Agent. Halliday
uses the following example:

(21) the glass broke (middle)

(22) the cat broke the glass (effective: active)

(23) the glass was broken (effective: passive)

The other two participants in the inner ring are Beneficiary
and Range. We have already described their functions in transitive
terms and they are similar from an ergative point of view:
Beneficiary is the one for whom a process is being carried out;
Range is the scope or domain. Agent, Beneficiary and Range may
participate in the process directly as noun phrases or indirectly in
prepositional phrases. Other elements in the clause are more clearly
circumstantial, and these form the outer ring. They are expressions
of extent, location, manner, accompaniment, matter, cause, or role.

Probably the greatest advantage of an ergative interpretation
is its ability to involve the separate feature of agency. Among other
insights, it provides for a secondary Agent:

(24) we let the wind dry our bodies

(25) 1 got John to boil the rice

(26) they call him Beau

This broad extension of Agent allows for indepth analysis of

otherwise troublesome constructions in English.

CHAPTER IV

DARI: GRAMMATICAL ANALYSIS

1. Dari as a distinct dialect

The main classification of Persian dialects is by general
agreement into a western and an eastern group (eg. Lazard 1970).
The various dialects of Iranian Persian comprise the western group;
Tajik and Dari comprise the eastern group. Though all these dialects
are mutually understandable, to quote from Ehsan Entezar's 13sz
W, "the dialect variations between Afghan Persian
and Iranian Persian and the concomitant cultural implications are
such that one may not be substituted for the other" (1964zx).

Literary Persian derives from the educated spoken language of
Tehran which is now contemporary standard Persian, the accepted
standard in all but Tajikistan which, because of Soviet influence,
has developed its own literary language written in Cyrillic script
(Windfuhr 1987).

In Afghanistan the distinction between the written and spoken
language is still a wide one. Dari is essentially a spoken dialect of
Persian. While the influence of colloquial Dari is perceivable to
some extent in the literary language of Afghanistan (Lazard
1970:70), the written language is standard Tehranian Persian, not
Dari. For example, the Dari word for 'water' is au and is used in all

spoken Dari, even formal. The written word is ab. The word au is

22

23

never written (indeed, one would not know how to do it) and in
Afghanistan ab is never spoken, except when reciting a written text.
This separation of written and spoken language is maintained at
least in part because Afghanistan is basically a non-literate
society; few Afghans read or write. Those people that are literate
are more often literate in Arabic than in Persian because of the
pervasive influence of Islam. The language of this investigation is

spoken colloquial Dari, not literary Persian.

2. Basic typology

Dari is what Joseph Greenberg (1963) has termed a rigid verb-
final language; its basic Clause typology is subject-object-verb
(SOV). Johanna Nichols calls Persian a 'split-order' language (1986):
head-first at the phrase level, head-last at the clause level. This is
only partially true for Dari; while the noun phrase has mostly
(though not exclusively) post modifiers (Rashidi 1989a), the
adjective phrase and the adverbial phrase have preposed modifiers.

In addition, Dari is a prepositional language.

3. Grammatical properties of Dari clause structure

Following Halliday, the clause, and not the sentence, will be
viewed as the basic grammatical unit of organization. This has
functional implications that will be apparent, especially in the
discussion of Theme. A clause will be viewed functionally,
._ following Halliday, as a process along with both the direct
participants in the process and the circumstantial elements

associated with the process. A process is loosely defined as a 'going-

24

on'; a doing, happening, feeling, or being (Halliday 19852101), which
may be realized in a variety of ways in a language. For our purposes,
any process, whether dependent or independent, embedded or 'main',
will be defined as a clause.

In Dari, a clause is typically realized grammatically as a
Predicate (P), its only obligatory element, preceded by either a
Complement (C) or an Object (O), which is in turn preceded by a
Subject (S). In addition, there may be various Adjuncts (Ad) of
Location (Lo), Time (Tm), Manner (Mn), Range (R), Beneficiary (Bn),
Extent (Et), or Accompaniment (Ac). Predicates are realized as verb
phrases (VP); Subjects as noun phrases (NP); Objects and
Complements as noun phrases, adjective phrases (AP), or
prepositional phrases (PP); Adjuncts as noun phrases, adjective
phrases, prepositional phrases, or ke, 'that', clauses. Two other
clausal elements were found in the data under investigation: 1) the
modal particle (MP) bayad, 'must', and 2) appositives, both subject
(SA) and object (OA).

The word 'typical' is being used here in the psychological sense
of that which the native speaker feels is the normal way of saying
things. Unless there is some particular reason for doing otherwise,
a structure will be realized in a typical manner. This may or may
2 not be numerically significant. Halliday uses the term 'congruent';
most grammars refer to 'unmarked' structures. The less technical
term 'typical' is preferred here to avoid any indication of formalism.
In this light, the typical constituent order for Dari is (Subject)-
Object/Complement-Predicate. Adjuncts are very mobile but

typically appear immediately preceding the verb, with the exception

25

of ke clauses which typically appear following the verb. This

typical order holds for all clauses. Modal particles are also mobile,
but their most common position is clause-initial. Appositives appear
immediately following theconstituents to which they are

appositive.

4. Noun phrase

There are no grammatical case markings as such in Dari. Case
(subject versus object generally) is distinguished by context, word
order or affixation of -ra. Nouns are generally not marked for
number and never marked for gender. Two morphological markers,
however, do occur with regularity in Dari: 1) -ra, often referred to
as the definite direct object marker, and 2) -haa, the specific plural
marker. Neither of these morphological markers has an equivalent in

EngHsh.

4.1 -ra

The most thoroughly investigated but still the most elusive
morphological marker in Persian is -ra (glossed RA). Gilbert Lazard
has conducted, perhaps, the most intensive research on this
morpheme. In general he found that -ra accompanies some direct
objects but not others. Its presence is conditional on a complex
number of factors including: degree of definiteness, degree of
humanness, degree of amplification of the process, semantic
distance of the object and the verb, relative weight of the syntactic
group, and the aim of the communication (Lazard 1982). -ra most

often occurs when the object needs to be distinguished without

26

ambiguity from the subject, but there is an assortment of diverse
exceptions, and grammarians have not yet found a satisfactory

solution of infallible criteria for determining when -ra is used.

-, What is readily apparent is that the use of this morpheme depends on

complex semantic and grammatical conditions; a better
understanding of its nature will take into account both paradigmatic
and syntagmatic factors.

Windfuhr indicates that the fact that -ra is elusive to
grammatical analysis suggests that it is at least partially
conditioned by rules that transcend clause boundaries, in particular
rules of reference and specificity (1979253). Underlying all of these
conditions appears to be speaker intentionality and point of view;
the encoder marks what s/he views as a unique referent within the
context of the utterance. Thematic emphasis seems to play a role,
and topicalization is definitely a factor.

Entezar only briefly mentions the use of -ra in his Earsj
WM (1964). He indicates that -ra marks some direct
and indirect objects. His main criteria for use are specificity and
verb type. His first criterion is well-substantiated; his second
seems more questionable. The following examples show the

contrast of a specific versus a non-specific object:

(1) qalam me-xaay-om 'Iwantapen’
pen ME-want—1s

(2) qalam-a me-xaay-om 'I want the (a particular) pen'
pen-RA ME-want-1s

27

The use of -ra with indirect objects, as well as direct
objects, in Dari is confirmed by the data under analysis in this

study. For example:

(3)C100b1 arab-aa-ra goft 'he told the Arabs‘
arab-PL-RA said

4.2 -haa

The second morphological marker that occurs regularly in Dari
is -haa (glossed PL), usually realized as -aa in the data.2 The
basic function of this morpheme is one of amplification (Windfuhr
FA 1979232). Though -haa is usually referred to as a plural marker, the
category of grammatical plural does not exist as such for noun
phrases in Dari. The marker is not obligatory when more than one
item is indicated, only when the speaker wants to specify one-plus.
As with -ra, specificity plays a role here. Note examples (4) - (6)

and their English translations:

(4) malim ist 'he is a teacher'
teacher is

(5) malim ist-an 'they are teachers'
teacher is-3p

(6) malim-haa ist-an 'they are the teachers'
teacher-PL is-3p

In addition, subject agreement with the verb, which is marked
for a singular/plural distinction, is vague. Noun phrases modified by

a number are never marked with -haa.

28

(7) se malim ist-an 'there are three teachers'
three teacher is-3p

-haa often seems equivalent to such concepts as 'plenty of',

'many', or 'all kinds of'. Even 'mass nouns' may be marked with -haa.

(8) Sir-haa 'plenty of milk'
milk-PL

(9) nan-haa 'all kinds of bread'
bread-PL

4.3 The ezafe

A third morpheme (though not strictly a morphological marker)
that plays an important role in the noun phrase is the clitic -e. In
contrast to the clause, the noun phrase in Dari has a relatively rigid
word order (Rashidi 1989a). Noun phrases are head-first with the
modifiers connected to the head (or the preceding modifier) by the
clitic -e (glossed E), referred to as the ezafe or izafef in
literature on Persian and Arabic. This particle is phonologically
attached to the head or the preceding element but is semantically
part of the post-modifier.

(10)C99 mardum-e waSi

people-E uncivilized
'uncivilized people'

The ezafe is realized in several ways, [e], [i], O, all
phonologically predictable. The presence or absence of -e, however,
is not difficult to distinguish even in casual oral data such as the
text under investigation. The psychological reality of this connector

is readily and consistently apparent to native speakers; seldom is

29

there ambiguity on this score. When the ezafe is not realized as a
separate phonological entity, eg. (11), it has been coded paren-

thetically as (e). Otherwise, it has been coded consistently as -e.

(11) poi-(e) xub-e-s
leg-E good-E-he
'his good Ieg'

At times the ezafe can be translated 'of' as in (12), but, for
the most part, -e simply serves to attach a modifier to a preceding
element. Greenberg notes that the distinction between genitives and
adjectives in Persian is fuzzy because both ”are marked by exactly
the same formal means" (19632100). This is an accurate description
of Dari. There are no separate possessive forms for pronouns (11),
nor are there separate morphological markers for possessive nouns
as in English (13); any modifier is simply attached to the head in a
string, each element connected to the previous one by the ezafe .

(12)A7 yak qismat-e S’ar

one section-E city
'one section of the city'

(13)884 pirambar-e xoda
prophet-E god
'prophet of God'
Because of the ezafe connectors, noun phrases are easily
distinguished as single clausal constituents (14) as opposed to, say,

a subject noun plus an adjective complement (15).

(14) kitab-e kalon ist 'it is a big book'
book-E large is

30

(15) kitab kalon ist 'the book is big'
book large is

The ezafe connector is strictly a property of the noun phrase,
which is head-first. Adjective and adverb phrases are head-last and

do not have ezafe connectors (16).

(16) besior kawi
very powerful

4.4 ke

A final aspect of noun phrases in particular, but also Dari
clauses in general, is the use of the relativizer ke, glossed 'that'.
This morpheme introduces both restrictive and non-restrictive
clauses at all levels of grammar. (In fact, the narrator in this text
uses ke as a general connector of clauses, giving the narrative at
times the flavor of one long run-on sentence.) Windfuhr says that ke
is a surface particle and not a pronoun, a 'dummy' word, as it were
(1979269). This seems accurate. Though ke is not an empty word in
any real sense, the 'meaning' of ke is highly grammatical and often
difficult to render into English. Entezar lists the main functions of
ke in Dari:

1. After the main verb, ke introduces a clause that answers

the question 'what'.

(17) me-fam-om ke Suma az kabul ist-in
ME-know-1s that you from Kabul are-2p
'I know that you are from Kabul'

31

2. After a noun, ke introduces a modifier.

(18) baée ke erat raft
boy that Herat went
'the boy that went to Herat'

3. ke introduces clauses that are circumstances of cause or
reason

(19) amad-om ke Suma-ra be-ben-om
came-Is that you-RA BE-see-1s
'I came in order to see you'

4. ke introduces dependent time clauses

(20) ke amad-om Suma raft-a bud-in
that came-1s you went-PF was-2p
'when I came you had gone'

5. ke introduces indirect reports

(21) éuma goft-in ke awaa xub ist
you said-2p that weather good is
'you said that the weather was nice'

(22) golam-a bo-go ke sabr ko
Ghulam-RA BE-say that wait (BE)-do
'tell Ghulam to wait'
(Entezar 19642198)

A further function of ke is as an emphatic. Sometimes the
best translation seems to be 'but', 9g. (23); at other times ke
follows a topicalized element, further highlighting the 'focused'
status of that element, eg. (24); or ke may serve to simply

emphasize a noun phrase or entire clause [see text BI 8; B41]. The

32

position of emphatic ke appears to be mobile, though this particle

often is a part of fixed idiomatic expressions as in (25) and (26).

(23)B70 tana kason-e ke qabul kard
only peOple-E that accept did
'only some people accepted'

(24)B63 i but-e ke xud-e-tan . . .taiar me-kon-in
this idol-E that self-E-you . . . construct ME-do-2p
'these idols you yourselves construct with your own hands'

(25)B16 mohamad wafat ke yaft-a bud
Mohammed death that found-PF was
'Mohammed died'

(26)B§4a l ke kalon Sud
he that large became
'he grew up'
5. Verb phrase
The verb system in Dari seems easily describable but is
underlaid with a subtle complexity. Tense and aspect are so
intricately intertwined as to make classification difficult at best.
Modality seems more formally separate (see discussion of the modal
particle bayad below) but is, in fact, also infused into the verb
phrase itself (see discussion of be- below). In his latest analysis
of tense, aspect and modality in Persian (1987), Windfuhr fully
develops a model that he first presented in 1985. Proposed in
reference to Persian, this is a universal model where aspect is as
basic a categorical vector as tense. From this perspective, Persian
verb forms are based on three stems: present, aorist, and perfect.

This is a highly involved model beyond the scope of this study.

33

Suffice it to say that it does pointedly show the difficulties
involved in trying to untangle the Persian verb system. A simpler,

though less delicate and insightful, analysis will be presented here.

5.1 Tense and aspect
Dari has two formal tenses: present and past. Each has a

separate, often phonologically disparate, stem form. For example:

guess mascot east
see ben did
be ist bud

Entezar notes that many past tense stems and in [t] or [d]
(19642128), but the present and past stems must be learned as
separate entities. Future is not a tense, but at best a modality (see
discussion of me- below).

The major distinguishing feature of the Persian verb system is
probably aspect rather than tense (Windfuhr 1979:86). There are
two main aspect markers in Dari, me- (glossed ME) and be- (glossed
BE). These markers have been glossed simply ME and BE rather than
descriptively in this analysis because their core meanings are still a
topic of much discussion.

5.2 me-
me- is most often described as an imperfective aspect
marker, but it has also been variously identified as durative,

habitual action, progressive action, and future action. Windfuhr

34

(1979287) states that me- ”appears to express the absence of any
limitation of an event in terms of its beginning or end, be it
temporal or factual." It is used with both the present and the past
tense. Entezar lists three meanings for me- in Dari with the
present stem and three with the past stem. Time expressions are

often needed to disambiguate meaning. The following are

retranscribed illustrations from Entezar's MW:

Present:

1. continuation of an action over a period of time
(27) S’uma aali farsi me-xan-in
you now Farsi ME-study-2p
'you are now studying Farsi'

2. habitual action
(28) me ar roz farsi me-xan-om
l every day Farsi ME-study-Is
'II study Farsi every day'

3. future action
(29) sabaa farsi me-xan-a
tomorrow Farsi ME-study-3s
'tomorrow he will study Farsi'
(Entezar 1964236)

Past

1. wishful thinking
(30) kaéke ma-m me-raft-om
wish l-also ME-went-1s
'I wish I had gone also'

35

2. continuation of an action in the past
(31) diroz piéin radyo gos me-kard-om
yesterday afternoon radio hear ME-did-1s
'yesterday afternoon I was listening to the radio'

3. habitual action in the past
(32) parsal ar roz sinemaa me-raft-om
used to every day cinema ME-went-1s
'I used to go to the movies every day'

(Entezar 19642165)

5.3 be-

Modern Persian does not really have a perfective counterpart
to me-. In general, the absence of me- indicates action seen as
punctual or highly circumscribed. The prefix be- (glossed BE) has
been described as the perfective marker, but in modern Persian it
has disappeared from all forms except imperative (33) and (34),
optative (35), and subjunctive (36), (37) and (38). Except for the
copula, BE does not appear in the negative (37). In Dari, not all verbs
actually have the prefix be-. For example, the verb kardan, 'to do',
a common compounder, does not take the prefix be- but simply uses
the present stem, or more familiarly, the abbreviated ko in the
imperative (34).

(33) gap be-zan-ln

speak BE-hit—2p
'speakf

(34) kor kon-in OR kor ko
work do-2p
'workl'

36

(35)C123 dar maj'id bur-in
in mosque go-2p
'you should go to the mosque'

(36)B58 bayad Suma ma-ra parastes kon-in
must you I-RA worship do-2p
'you must worship me'

(37)C117 bayad doktar-aa-tan-a na-kus-in
must daughter-pl-your-RA NEG kill-2P
'you must not kill your daughters'

5.4 Perfect

A third aspect marker appears in Dari, the suffix -a (glossed
PF). Entezar (19642164) calls this morpheme the perfect tense
marker. Its meaning is probably closer to completive. Though
referred to as the perfect marker, -a is not the perfective
counterpart to me-, as a verb can be (though rarely is) marked with
both (of. Windfuhr 1979290). For the past perfect, -a is suffixed to
the main verb which is followed by the past form of the copula
inflected for person and number.

(38) raft-a bud-in

went-PF was-2p
'you had gone'

The present perfect uses -a only for the third person singular,
as present perfect is basically formed by stressing and/or
lengthening the final person-number suffix. In effect, present
perfect forms remain indistinguishable from simple past forms,
except for the first person singular which takes on the plural marker

37

(phonologically a nasal is added) and the third person singular (O)
which becomes a stressed [a]. Context is important for
disambiguation in all but the third person singular, but a slightly
heavier final stress does occur in the present perfect (Entezar
19642164).

W W
15 did-om did-im
1p did-im did-im
25 did-i did-i
2p did-in did-in
35 did did-a
3p did-an did-an

The following illustrations are verb phrases from the text:

(39)A8 fatha kard 'he conquered'
conquer did (past; 35)
(40)0131 hast kard-a 'he has created'
create did-PF (non-past; PF; 35)
(41)B28b aftad-a bud-an 'they had fallen'
fall-PF was-3p (past; PF; 3p)

(42)BS2 parastes me-kon-an 'they are worshipping'
worship ME-do-3p (ME; non-past; 3p)

(43)B69 qabul na kard-an 'they did not accept'
accept NEG did-3p (NEG; past; 3p)

(44)095 be-rason-an 'they should deliver'
BE-deliver-3p (BE; non-past; 3p)

(45)C124 ebadat kon-in 'you should pray'

pray (BE)-do-2p (BE; non-past; 2p)

38

(46)C100 na kuS’t-in 'you should not kill'
NEG kill-2p (NEG; BE; non-past; 2p)

The following is a system network of the verb phrase in Dari:
{NEG
6

NE
-—- aspect—I‘
BE

-——o

past
VP tense —{:
non-past
PF 1
O ~person——E2
3

affix—V plural (p)
— non-plural (5)

 

 

 

 

 

 

Figure 3

Dari Verb Phrase

6. Compound verbs

Most verbs in Dari are phrasal verbs, or more commonly stated,
compound verbs. Windfuhr indicates that diachronically Persian is
becoming increasingly 'synthetic' with the expansion of compound
verbs, "virtually the only source of verbal innovation for many
centuries” (19792113). Surely, in Dari the use of compound verbs as
a tool of invention is prominent. It is not uncommon for the entire

semantic loading of a clause to lie in the verb phrase.

39

Dari has basically two types of phrasal verbs. The first type
is similar to phrasal verbs in English, consisting of a verb plus a
closely-linked adverb.

(47)A23 pas amad

again came
'returned'

(48)B82 payan andoxt
down dropped
'threw down'

But it is the second type of compound verb that is 50 common
in Dari and of particular interest here. This verb phrase is best
termed a complex verb to distinguish it from an object plus a verb.
A complex verb is composed of a content word followed by a 'dummy'
auxiliary verb. It is the auxiliary that receives the inflection.
Though there are variousunusual auxiliaries, the vast majority of
complex verbs are formed from one of three verbs: sudan 'to
become', kardan 'to do', or dardan 'to have'. Some examples:

(49) xau me-s-om 'I sleep'
sleep ME-become-1s

(50) kor me-kard 'he worked'
work ME-did
(51) qalam kor dar-om 'I need a pen'

pen work have-1s

(52)834c zlarat me-kard-an 'they paid religious respect'
respect ME-did-3p

40

Some complex verbs are quite transparent (49) and (50), but
others have become opaque and native speakers generally cannot
deconstruct them (51) and (52). Some complex verbs are transitive
(51) and, therefore, take direct objects, but others are intransitive
(49). In other words, complex verbs behave as any other verb phrase
would. The line between a complex verb and an object-plus-verb is a
fuzzy one, however. In analyzing the data, several criteria were
used in establishing a verb's status as a complex verb:

1. The content noun or adjective is not movable.

2. Transitive verbs can take direct objects of their own.

3. Functionally, the constituents have a better semantic fit if

the process is viewed as a complex verb rather than a process-

plus-goal.

4. The verb a5 a unit is relatively non-transparent.

5. The auxiliary is semantically 'empty'.

7. Clause types

In Dari, all clauses have the same basic syntactic format
regardless of semantic function: statement, question or command.
Thus, there are not separate declarative, interrogative, or
imperative clause structures. Questions are indicated by intonation;
imperatives use the be- prefix on the verb. In light of this, clause
classification is relatively simple. The following clausal typology
was based on the text under investigation. The transcribed oral
narrative was divided into grammatical clauses, yielding
approximately 150 separate clauses. Each clause was then analyzed

grammatically, identifying the basic clausal constituents:

41

Predicate, Subject, Object/Complement, and Adjunct. This analysis
yielded four separate major clause types, based solely on
grammatical criteria but with functional and semantic correlations.
Each major clause type (except existential) was further

distinguished into subtypes.

7.1 Relational clauses

ReICI - is +C +P

Relational clauses are all middle voice, that is, minus agency.
There is an obligatory, overt Complement plus a Subject, which is
optionally overt but always in the underlying structure; that is, the
verb is marked for this subject and it can also be retrieved from the
context. Verbs comprising the Predicate are of two types: 1)
budan, 'to be' and 2) Sudan, ' to become'. The Complement can be a
noun phrase (RelA), an adjective phrase (RelB), or a prepositional
phrase (ReIC). No relational clauses in the data contained Adjuncts,
only Margins of Time.

RelC clauses are prominent in Dari, occurring where English
would express the same idea as a possessive phrase.

(53)A4 kabul / da roi éand bazar-aa-e bezug I bud
Kabul in way many bazaar-pl-E big was
'Kabul had many big bazaars'

(54)BS1 in-s l da tasaruf-e 32 i I bud
key-its in possession-E from he was
'he had its key'

Relational clauses are functionally either identifying or

attributive. Variation in basic word order was almost non-existent

42

in the data; just one clause, A13, had a Complement-Subject order.
Twenty-three of the thirty-six relational clauses had overtly
expressed subject-slot fillers. Formulas for and illustrations of the

three types are as follows:

RelA - iS:NP +C:NP +P2Vb,s (b=budan, §=§udan)

SzNP CzNP P2Vb
(55)855 ma I piqambar-e xoda / -st-om
l prophet-E god is-1s

'I am the prophet of God'

CZNP P3Vg
(56)B38 but parast I Sud-an
idol worshipper became-3p
'they become idol worshippers'

RelB - _1-S:NP +C:AP +P2Vb/g

SzNP C2AP P2Vb
(57)B57 i but-aa / qabil-e paraste§ / n-est
this idol-PL worthy-E worship NEG-is
'these idols are not worthy of worship'

SzNP , C2AP P2Vg
(58)093 akwam-e xud I bira I Sud-a bud-an
tribes-E self lost become-PF was-3p

'the tribes themselves had become religiously lost'

RelC =i82NP +C2PP +P2Vb3

CzPP szb
(59)A12 da roi éor darwaza / bud
in way4 door was

'it had four doors'

43

7.2 Existential clauses

ExCI-+S +P

Purely existential clauses are quite common in Dari. Where
English would use a ’dummy' subject construction (e.g. 'there’s a
small hotel'), Dari uses simply a Subject plus the copula (in essence,
'a small hotel is'). Dari also uses existential clauses extensively for
ideas expressed in English by a Subject plus a Complement, 9.9. 'the
dog is big'; in Dari, the phrase would be the equivalent of 'a big dog
is', 'big dog' being realized as a single clausal constituent [see (14)].
Though the adjective can be expressed as a separate clausal
constituent (without the ezafe ) [see (15)], the construction is
unusuaL

Existential clauses are all middle voice. In fact, there is only
one participant in the process. This participant is realized as an
obligatory Subject (NP). There may be, in addition, various
circumstantial elements of Time or Location. The verb is either a
form of the budan or Sudan. The process type is always
existential, thus, the single participant is the Existent. Since there
are only two clausal elements, both obligatory and one the
Predicate, there is no variation in element order. These clauses
cannot be further differentiated into subtypes. The detailed formula

is:

ExCI - +S:NP +P2Vb/g

The following are examples of existential clauses in the data:

44

SZNP PZVg
(60)C135 amr-e namaz I Sud
edict-E prayer became
'the edict of prayer came into being'

SINP PZVb
(61)C88 éor kitab-e asman I -ist
4 book-E heaven is

'there are the four books of heaven'

7.3 Transitive clauses

TrCl =- 18 +0 +P

Transitive clauses contain an obligatory Object, usually, but
not always, marked by the definite object marker -ra. Verbs are all
action verbs, most complex. As in relational clauses, there is an
underlying subject which may or may not be overtly realized.
Process types include: material, mental, identifying and attributive.
These are effective/active voice clauses with agency in subject
position.

Transitive clauses have been differentiated into two subtypes.
Type A clauses are by far the most common; they are the typical
transitive clauses following the basic formula. A detailed formula
and examples follow:

TrA :- _-_I_S:NP +O:NP +P2Vact

OINP PZVact
(62)321 besior mamlakat-haa-ra / graft-an
very country-PL-RA took-3p

'they took many countries'

45

OINP SzNP P2Vact
(63)C115 i kuI-e éis-aa-ra / xoda I mana kard-a
this all-E thing-PL-RA god prohibit did-PF

'all of these things God has prohibited'

SzNP OzNP P2Vact
(64)361 ma / Suma-ra / hast kard-im
I you-RA create did-1p
'I created you'

Type C clauses are a restricted subset; the predicate is the
verb gotten, 'to say' or 'to call', and there is an obligatory object

appositive that is essential to the completion of the proposition.
TrC - _-I_-S2NP +O:NP +OA2NP +P2V90fl

OzNP OAzNP szgofl
(65)B27c u-ra I ajarl aswad / me-g-an
that-RA [Arabic for 'black rock'] ME-say-3p
'they call that Ajarl Aswad'

Transitive clauses show a good deal of variation in element
order. They most commonly are minus an overtly expressed Subject
(34 out of 51 clauses). The Object is often fronted, and when it is,

it is always marked with -ra. This marked Object precedes even the
modal particle, bayad.

0 MP S P
(66)B47c i-ra I bayad/ maa I parastes kon-im
this-RA must we worship do-Ip
'we must worship it'

All Adjuncts immediately precede the Predicate. These may
be Adjuncts of Location, Manner, Beneficiary, or Time, eg. (63).

46

O Ad/Mn P
(67)C104 doktar-aa-e xud-a / da qimor I me-baxt-an
daughter-PL-E self-RA in gambling ME-lost—3p
'they lost their daughters in gambling'

An expanded transitive clause formula is as follows (element

order is tentative as no one clause contains all elements):
TrCI= _«I_-MP :5 18A +O 10A _-I_-Ad +P

Some examples are:

0 S Ad/Lc Ad/Lc P
(68)B31b i-ra / xoda / az asman I be zamin / andoxt-a
this-RA god from heaven to earth dropped-PF
'this God dropped from heaven to earth'
0 OA P
(69)A6 yak-e-éa / darwaz-e lauri I me-goft-an
one-E-this door-E bazaar ME-said-Sp

'one of these they called the Door Bazaar'

The text contains one exceptional element order clause, (70).
This is formally a ke clause, but is, in fact, an independent clause.
In this clause, the Object, marked with -ra, follows even the
Predicate, and is Itself followed by another relative clause. The

reason for the unusual element order is unclear.

(70)C129 (ke) u har musilman/me-fam-ale sura-haa-ra
that every Muslim ME-know-PF this sura-pI-RA
misli ke maa me-fam-im
like that we ME-know—1p
'every Muslim knows these suras like we know (these suras)‘

47

7.4 lntransitive clauses

IntCl =- _-_I-_S +P

lntransitive clauses have only one obligatory constituent, the
Predicate. There is an implicit Subject, which may or may not be
overtly realized as a separate subject element. The process
expressed can be: material, mental, behavioral, or verbal. These are
middle voice clauses containing only an Affected Participant, not an
Agent. In addition, some mental and verbal processes (lntC) contain
a Phenomenon or Verbiage, but these participants are never realized
as noun phases, only as ke clauses or direct reports in the form of
independent clauses.

A third subtype, labelled lntB, has an obligatory Adjunct
realized as a prepositional phrase. These Adjuncts act to extend the
process. lntB clauses are what Pike and Pike would call bi-
intransitive: a second obligatory participant is necessary to
complete the proposition (Pike and Pike 1982243).

lntransitive clauses may contain Adjuncts of Manner, Location,
Range, Beneficiary, Time, or Extent. In addition, the data contain one
IntC clause with an indirect object marked with -ra (71). Most
Beneficiaries are realized as prepositional phrases.

(71)C100 mohamad ke amad I arab-aa-ra I goft I

Mohammed that came Arab-pl-RA said
ke Suma aulad-aa-e-tan-a na kuSt-In
that you child-pl-E-your-RA NEG kill-2p

'Mohammed when he came told the Arabs that they should not
kill their children'

48

Adjuncts in all intransitive clauses are very mobile. While
most directly precede the Predicate, Adjuncts of Location and Extent
appear before the Subject, and one Beneficiary appears clause-
finally. Clause-level ke clauses, eg. (71), and direct statements,
eg. (74), all appear clause-finally. Range, however, always appears
just before the Predicate, eg. (73). An overt Subject is expressed in
subject slot a little more than half the time. Following are subtype

formulas and an example of each:
lntA - iS:NP +P:Vact

S.NP PIVact
(72)B19 ln-aa ham / xelafat kard-an
this-PL also khalifing did-3p
'these also acted as khalifs'

lntB - 182NP +Ad2PP +P2Vact

S:NP AdzPP szact
(73)B36 akwam-e arabistan-hama I az dln / gan-an
tribes E Arabia-all from religion abandoned-3p

'all the tribes abandoned religion'
IntC - iS:NP +AD2ke/cl +P2VV/me

S:NP P2Vme Adzcl
(74)C101 un-aa I fikar me-kard-an / doktar bad ast-an
that-PL think ME-did-3p daughter bad is-3p
'they thought that daughters were bad'
7.5 Passive
The notion of passive is a Western construct used by European
linguists and grammarians. Passive in the sense of being a

counterpart to an active clause has no basis in Persian. John Moyne

49

(1974) argues that there is no lexical or syntactic category of
passive in Persian. Hessam Tabaian (1979) argues that there is no
opposition between active and passive, only an opposition between
state (budan) and process (Sudan). Lazard (as cited in Tabaian
1979) states that Sudan forms the intransitive verbs which
normally serve as the passive of the verbs formed with kardan, 'to
do'. But the term passive is not even used by Persian scholars.
So-called passives in Dari are formed by a two part verb
consisting of a content verb plus Sudan. These clauses have an
obligatory Subject which is the Affected Participant. Within the
system set up for Dari clauses, these are simply intransitive
clauses. There is no Agent and most clauses are also minus agency.
If one wants to state unambiguously that a participant in a process
is Agent, one must put that participant in the subject slot. Any
clause marked for agency then becomes a transitive clause. Thus
there is no need for a separate grammatical clause type to
differentiate active from passive; from a grammatical point of view
the distinction is unnecessary. The following are examples from the

data of minus agency intransitive clauses:

S:NP AdzPP szact

(75)B46a yak but-e az in-aa/az yaaqud/saxt-a Sud-a bud
one idol-E from this-pl from ruby make-PF become-PF was
'one idol from these had been made from ruby'

(76)093 akwam-e xud / ber-a Sud-a bud-an
tribes-E self lost-PF became-PF was-3p
'the tribes themselves had become religiously lost'

50

Other clauses that might be interpreted as passive in a
different system are better seen as simply clauses of change of
state. These clauses have stative counterparts. Compare (77), a

stative, with (78), a process.

(77) xué Ist-om 'I am happy'
happy is-1s
(78) mt? Sud-om ' I became happy'

happy became-1s
In the text this opposition is found in:

(79)Ct34 elamre/bud/ke bare piqambar awal dafar amad
this edict was that for prophet first time came
'this was the edict that came first to the prophet'

(80)C135 bad az u / amre-(E) namaz I 's'ud
after from that edict-E-prayer became
'after that came the edict to pray'

Passive does not seem to be a distinctive notion in Dari. It is
argued here that there is not a separate passive clause type for Dari.
The so-called passive construction occurs when the speaker cannot
or does not wish to mention an agent (or there is no agency). The
speaker then puts the affected participant in subject position by
default, not necessarily as a focus. These clauses must be viewed in
their own terms within the system in which they occur. Within the
Dari clause system, the so-called passive, in its various

manifestations, is simply a variation on other clause types.

51

Endnotes

1. Examples are of two types: 1) those that actually occur in the
text under analysis and 2) those that occur in other texts or have
been solicited from native speakers. Those that occur in the text
are cross-referenced to one of three sections of the text: B is the
main section of the text; section A precedes this section and section
C follows the main section. The entire text, glossed and translated
can be found in Appendix A.

2. The phoneme [h] is rarely realized in casual spoken Dari, though
its underlying presence istestified to by native speakers who can
supply it without fail and, indeed, often insist on its physical
presence even when the sound is not articulated.

3. It is reasonable to posit Vg for RelC clauses as well, but no such
examples were found in the data.

CHAPTER V

TRANSITIVE STRUCTURE OF DARI

1. Functional roles

Every linguistic theory has struggled with the notion of
functional roles, and while all are basically exploring the semantic
aspect of clausal constituents, each approach is different enough
that a comparison is at best misleading. Most models of functional
roles have been independentlydeveloped as an integral part of the
theory to which they belong. For this reason it is difficult to
transpose from one model to another. However, most 'case' theories
involve the notion of the verb as central, with various constituent
relations dependent upon that verb.

Halliday's functional grammar is no exception. He focusses on
what he terms processes, thus avoiding an explicit linkage to a
grammatical form and leaving open the possibility that processes
may be realized in various ways. In his discussion of grammatical
metaphor, he gives three steps for the realization of meaning: 1)
selection of a process, 2) construction of a configuration of
transitivity functions (process, participants, and circumstances),
realized as 3) a sequence of grammatical classes (noun phrases, verb
phrases, etc.) (Halliday 19842321).

This view of the process of creating language starts with
meaning. In this analysis, we have done precisely that. Though the
text under investigation was divided into processes (realized as
clauses by definition), the analysis of the individual clauses was

52

53

always done in terms of their relationship to the whole text and the
reader's judgment of what the narrator intended to mean at that
particular point in the narrative. To a certain extent, however, a
clause by clause analysis is forced to ignore the whole because the
very act of dividing the whole creates individual chunks of language.

This dilemma seems to be unavoidable.

2. The text

In a casual, oral narrative such as the one under investigation,
language flows in a wave-like pattern (to use a Pikean concept),
where clauses are often incomplete as grammatical entities and
inextricably connected to the next clause. In this particular text,
the question of what is a dependent (beta) and what is an
independent (alpha) clause is vastly complicated by the narrator's
use of ke as a general linker of ideas. Nonetheless, the text has
been divided into individual processes. Each verb of any form in any
relationship, main or subordinate, was called a process and given a
number or letter. Rather loosely, main clauses are numbered and
subordinate clauses are sublettered, but the coding system should
not be given significance beyond a mere reference tool. Though the
text is numbered consecutively throughout, it has been divided into
three sections: A, a short beginning section; B, the main section
which has been analyzed at the discourse level and will be used for a
contextual analysis; and C, the concluding section. All three
sections have been reproduced in their entirety, glossed and
translated, in Appendix A. An English recapitulation of the entire
narrative can be found in Appendix C.

54

3. Dari process types

The analysis began with an application of Halliday's three
major (material, mental and relational) and three minor (verbal,
behavioral and existential) process types for English. The
classification of Dari processes into these categories was a
relatively smooth one; there seemed to be no particular reason to
drastically change the system. Some adjustments were made,
however. First, relational processes were divided (following
Halliday) into identifying and attributive. These two sub-
classifications were analyzed separately, as in Dari there are some
fairly distinct things to say about each. Second, existential
processes are important and distinct in Dari, both from a
grammatical standpoint and from a functional one. In Dari, this is
not a minor process type. In fact, Dari processes seem to run along a
cline from identifying through attributive to existential. Third, in
contrast to the relative importance of existential processes, the
data contain only one process that might be classified as behavioral.
As a result, behavioral is not posited as a separate process type for
Dari.

This distribution in Dari is interesting in light of Halliday's
original classification of process types obtained from his ”Class
Handouts” of 1977 (unpublished). In this mimeographed paper, there
are four principal types of processes for English: material, mental,
verbal, and relational. Relational is then sub-divided into
attributive, identifying, and existential. Behavioral does not exist
as a separate process. This earlier classification seems to fit the

precise status of the Dari processes in this text.

55

Halliday's reasons for his re-classification to include a
behavioral process type are clear for English and have proven useful
in the analysis of English text.1 There seems little evidence in the
Dari text, however, of a separate behavioral category. The one

example classified as such, (1), is highly metaphorical.

(1)A16 mohamad / wafat ke yaft-a bud
Mohammed death that found-PF was
'Mohammed had died'

Literally, the clause is 'Mohammed had found death” (ke here is
used as an emphatic). If one analyses the clause at its literal,
rather than metaphorical, level, the process is not a behavioral one
but either a material or mental one, depending on one's view of death
and its reality as a physical entity as opposed to a mental concept.
This is, of course, highly metaphysical, but, whatever the argument,
on a literal level the process is not a behavioral one. On a metaphor-
ical level, the core meaning, 'Mohammed died', is behavioral, but
this is hardly a solid base for positing the category of a behavioral
process for Dari. This piece of data does not support a separate
behavioral process type.

The other re-classification from Halliday's earlier model is
really only a reorganization, not a positing of a new type. In
Halliday's original system, existential is a sub-type of relational.
This makes sense grammatically for English as English existential
clauses have 'dummy' subject fillers. In Dari, existential clauses
are, however, not relational, even from a grammatical standpoint

(there is only one participant). Thus, it makes sense to separate

56

them grammatically from truly relational clauses, identifying and
attributive (which have two participants). We have done this in the
grammatical analysis. But functionally, there does not seem to be a
dichotomy between relational on the one hand and existential on the
other. There seems to be as much functional difference between
identifying and attributive as between attributive and existential.
In fact, the distinction between the latter two was often
problematic; because of the structuring of the Dari noun phrase, the
chunking of a noun and a modifier into one or two clausal
constituents was not always clear. The difference hinges on the use
of the ezafe which is not always phonologically apparent. This
grammatical fuzziness sometimes obscured the semantic intent of
the narrator. The following is a good example:

(2)A11 dewaI-aa-e gali-(e) kalon kalon/saxt-a Sud-a bud

wall-PL-E mud-(E) large large made-PF became-PF was

The clause may mean 'very large mud walls were made', or it
may mean 'mud walls were made very Iarge'. The difference function-
ally is that the former is simply a statement of the existence of the
walls, a characteristic of which is that they are very large, while
the intent of the latter is to state that the walls, the existence of
which is not in question, are very large. The ambiguity arises
because in this phonological environment the ezafe would elide with
the final [i] of gali, 'mud', and not be a physically separate entity;
this ezafe can only be posited on semantic, contextual, or
grammatical grounds. This example is not a unique one; the data
abound with cases of such ambiguity.

57

Because of this realization of these three functional
processes, three distinct process types of equal status are being
posited for Dari: identifying, attributive, and existential. In total,
Dari processes can be classified into six distinct types, with no
division into major or minor, though some types are clearly more
prevalent than others: material, mental, verbal, identifying, attribu-
tive, and existential. All the types but existential cross grammat-
ical clause type boundaries; existential alone seems to coincide

grammatically and functionally (with one possible exception).

Table 1
Process Type Distribution

Ma Me V I At Ex
Tr 38 5 7
Int 24 11 11 1 6 1
Rel 17 18
Ex 17

TOTAL 62 16 1 1 25 30 18 .- 162

4. Material processes

Material processes are the core of a narrative. These are the
action processes that account for the events that take place, the
'doings' in Halliday's words. Numerically, material processes
dominate this text. This seems normal for a narrative, even an
historical one such as this. Material processes have, by definition,
an obligatory Actor, the 'logical' subject of the clause but not

58

necessarily the agent. In Dari this Actor may be realized as the
overt grammatical Subject, or it may be realized only as a verbal
suffix. Material processes may have a second direct participant, a
Goal toward which the action is directed. In Dari, this optional
participant is never the grammatical subject of the clause; if there
is a Goal, it is most often realized as an Object, but may, in rare
instances, be realized more indirectly as an Adjunct. Sixty-two
processes (out of approximately 162), including (2), were analyzed
as material. This gives a nice broad basis from which to examine
the nature of Dari material processes.

Though most grammars of Persian state that the language has
a flexible word order, this does not hold true for the text under
study. Participant elements were rigidly ordered with one
exception: Goal may be fronted. Overt Actor52 appear in twenty-
five processes; only three of these follow the Goal. Thirty-eight
clauses contain Goals (all overt, of course), twenty-six of which are
in clause-initial position. This means that if there is a Goal
participating in the process, it is twice as likely to appear in clause-
initial position as in some other position. Many of these clause-
initial Goals are there by default, however, as the Goal is the only
overt participant. In only fourteen processes do both a Goal and an
overt Actor participate; of these, Goal appears before Actor in only
four. Ten processes have overt Actors preceding a Goal, but in two
of these, (3) and (4), the Actor is not an actual clausal constituent
as the Goal is embedded in a clause modifying the Actor; thus the
process is a dependent clause acting as a modifier of the noun
phrase of which the Actor is the head.

59

A: G20 Ma2P
(3)0131 (xoda-e bazurg-e ke) tu-ra I hast kard-a
god E powerful-E that you-RA being did-PF
'the powerful god that has created you'

A: G20
(4)C133 (parwardegor-e ke) u tamam-e maxluqat-e
creator -E that that whole-E creations-E
Ma2P
dunya-ra/ paida kard-a
world-RA create did-PF
'the creator that has created the entire creations of the world'

The modal particle, bayad, 'must', is a mobile element, but it
seems to directly precedes an overt Actor (B58, B62, B470). The
modal particle's status as a true clausal constituent is vague. It
appears five times in this text in nearly every position in the clause.
It is most likely one of those free-floating elements, like 'perhaps'
in English, the position of which is not semantically relevant. In
functional terms, the modal particle is part of the process.

Other direct participants (Beneficiary or Range) or circum-
stantial elements follow Actor and Goal, if either is overtly present.
Most immediately precede the Predicate. One clause, 380, has a
clause final circumstantial element of Location. The data contain no
instances of clauses with more than two direct participants or
multiple circumstantial elements. It is, therefore, not possible to
posit an order of participants beyond Actor-GoaI-other. Range would
most likely appear directly preceding the verb, as it is often
difficult to distinguish between Range plus Predicate and a complex

Predicate. Two Beneficiaries appear in material processes, one as

60

an Adjunct directly following the Goal (5), the other as the single

overt participant in the process (6).

6:0 Bn2Ad Ma2P
(5)0107 zan-e xud-a / bar u I me-dad-an

wife-E self-RA for them ME-gave-Sp

'they gave their own wives to them'

anAd Ma2P
(6)095 baroi omat / be-rason-an
for faithful BE-deliver-3p
'they must deliver for the faithful'

If there is an overt Actor, it is always a Subject. In fact,
every material process has an Actor, overt or otherwise, and it is
always the grammatical subject of the clause. In Dari, unlike
English, the logical subject and the grammatical subject map one on
one, at least for this text, and probably for the language in general.
Goal is realized as an Object, with one exception, (7), where Goal is

realized as an Adjunct.

GzAd Ma2P
(7)835 ba dist-e pimmbar-e hazrat-e ebrahim I xord
for hand-E prophet-E hazrat3-E Ebrahim hit

'it hit the hand of the prophet Ebrahim'

Dari makes the distinction between an actor as an active
aggressor and an actor as a mutual participant. In (7), the Actor
('it') is a rock that fell from heaven and in falling came in contact
with the hand of Ebrahim. dist, 'hand', is functionally the Goal but it
is not the grammatical object. One could have said: dist-e
ebrahim-re xord, hand-E Ebrahim-RA hit, in which case, the rock

61

would be not only Actor but also Agent. The difference is much like
the difference in the English clauses: the car hit the bus and the
car collided with the bus. Such English clauses as Ralph
married Josephine are often ambiguous when decontextualized;
Ralph could be the bridegroom or the cleric. In Dari, this type of
ambiguity does not occur; if Ralph were the bridegroom, he would
marry 'with' Josephine, becoming a co-Affected Participant; if Ralph
were the cleric, he would marry Josephine, acting as Agent. A more
semantically accurate, though grammatically more awkward,
translation of (7) would be: 'it touched with the hand of Ebrahim'.

Goal realized as Object is nearly always marked with -ra. The
exceptions are enlightening; there are five in the data:

A:S G20 Ma:P .
(8)0109 arab-haa / aadam / me-kust-an

Arab-PL people ME-kill-3p
'the Arabs were killing people'

G:O Cr(Mn):Ad Ma:P
(9)0112 doktar-aa—e mardum-e I be zurg I me-graft-an
daughter-PL-E people-E with power ME-took-Sp
'they were taking people's daughters by power'

G20 Ma:P
(10)0105b zen-e yak-e degar l paisa me-dad-an
wife-E one-E other money ME-gave-3p
'they bought another's wife'

G20 Ma:P
(11)0106 zan-e-s-aa I aalis me-kard-an
wife-E-his-PL exchange ME-did-3p

'they exchanged wives'

62

6:0 Ma:P

(12)830 ami sang I graft o wardost
this rock took and lifted up
'he took and lifted up this rock'

The most glaring exception is (8), 'Arabs were killing people'.
Both Lazard and Windfuhr indicate that the primary function of -ra
is to distinguish the Object from the Subject, but the narrator did
not feel it necessary to do so here. All other instances of the verb
kustan, 'kill', (C100, C102, 0117) have Goals marked with RA. In all
of these other clauses, the Goals are specific, however, while in (8),
the Goal is a general 'people'. It is probably the specificity that has
triggered the use of -ra. In addition, there is really no need to
distinguish the Goal from the Actor here; if 'Arabs' were the Goal,
arabhaa would, as a fronted Goal, be marked with -ra. (9) also has
a generalized Goal, 'people's daughters'. The explanation for lack of
-ra marking in (10) and (11) is less clear. Perhaps in (10), because
the phrase-final degar ends in /r/, -ra is phonologically precluded,
but this is not the case in (11). A grammatical argument can also be
made for (10). The entire sentence includes an initial 'if‘ clause:

(13)C105a agar xué-e-san me-amad
if like-E-their ME-came
'if they liked
It is not entirely clear to which process zan-e yak-e degar,
'another's wife', belongs. The sentence is either: 1) 'if they liked,
they bought one another's wives' (zan, 'wife', as direct object of
'bought'), or 2) 'if they liked one another's wives, they paid money'
(zan as direct object of 'like'). Semantically the grouping of

63

constituents in C105 is irrelevant, but the fuzzy grammatical status
of the noun phrase coupled with the verb, literally 'gave money',
make the Goal not very object-like. It is possible that some verbs
simply do not lend themselves to participants marked with -ra.
Entezar hints at this (1964:64). In (11), 'exchange' seems a likely
candidate for a verb that does not take -ra, but this is only
speculative here.

(12) is not marked for quite a different reason: aml sang,
'this rock', is already marked for high specificity by use of the
deictic ami, which means something like 'this very one'. In addition,
there is no overt Actor to distinguish from, and a rock is clearly not
the Actor of 'took and lifted up'.

The one instance of an Object as Range is also marked with RA:

RzO Ma:P
(14)C133 e6 roi-ra / na-ro / bajust . . .

no way-RA NEG-go except
'go no way except . . .'

This marking seems grammatical. Objects of prepositions are
never thus marked. -ra is still a very elusive notion, but the
variable in usage in this text seems to be specificity, along with the
role of the participant in the process and the semantic nature of the
process itself.

Material processes are realized grammatically as both
transitive (TrA) and intransitive clauses (lntA and lntB). This is not
surprising. They can then be either middle voice (with only an
Affected Participant, no Agent) or effective voice (with an Agent as
Subject).

64

Table 2

Material Process

E I' . | B l' |'
Actor (62) overt Subject (24)

Goal (36) Object (35) / Adjunct (1)
Beneficiary (2) Adjunct (2)

Range (7) Adjunct (4) I Object (3)

Circumstance (15) Adjunct (15)

Some examples of material processes follow:

A:S RzAd
(15)322 hazrat-e omar / taa qlsmat-aa-e éaom
hazrat-E Omar until section-PL-E Amman
Ma:P
o filistino baitl moqadas / fatha kard
& Palestine & Jerusalem conquer did
'Omar conquered up to sections of Amman, Palestine, and Jerusalem'

RzAd Ma:P A:
(16)C123 dar majid / b-ur -in
in mosque IE-go 2p

'you should go to mosque'

R:O Ma:P A:
(17)C125 sur-e al hamd-a I bo-xan-in
verse-E RA BE-read-Zp

'you should read the verse al Hamd‘
A:S Cr(Ac):Ad Mazp

65

(18)BBO in-aa I amroi mohamad I raft-an
this-Pl with Mohammed went-3p
Cr(Lc):Ad
I dar xon-e kaba-e but-aa-ra
in house-E rock-e idol-PL-RA
'they went with Mohammed to the Hona Kaba of idols'

Cr(En):Ad :MP Ma:P A:
(19)C122 pain] waxt I bayad I uzu be-gar-In
5 time must ablutions BE-take-2p

'you must make ablutions five times'

G:O A:S Cr(Lc):Ad Cr(Lc):Ad Ma:P
(20)B31b Ha I xoda I az asman / ba zamin / andoxt-a
this-RA god from heaven to earth dropped-PF
'this God dropped from heaven to earth'

G:O Cr(Mn):Ad Ma:P A:
(21)C112 doktar-aa-e mardume I be zurg Ime-graft-an
daughter-PL-E peeple by power ME-took-3p

'they took people's daughters by power'

5. Mental processes

Halliday lists the second major process as that of mental, the
process of sensing. These processes deal with our feelings, likes
and dislikes, how we view and interpret the world. Though it is
probably true as Halliday states that much of the time we are
talking about what we are thinking and feeling rather than concrete
events, mental processes are scant in this Dari text; only sixteen,
including dependent clauses, were found in the data: A2a, A14a,
B29a, 853a, B69, B70, B72 (or B75 or B79), BB1, 090, C91, C94,
C101, 0105a, C129a, C129b, C134c. This is not surprising, given the

nature of the narrative as an historical account rather than a

66

personal relating of events. Nevertheless, within these limited
data, patterns emerge on the composition of mental processes in
Dari. In a mental process, there are two central participants: 1) the
Senser, a human (or humanlike) perceiver, and 2) the Phenomenon,
the 'thing' or 'fact' perceived. In addition, there may be various
circumstantial elements. The typical element order for Dari is:
Senser-Phenomenon-Circumstance-Process.

Unlike Actors in Dari material processes, Sensers nearly
always appear as overt constituents, in clause-initial position. Like
Actors in Dari, but unlike Sensers in English, Dari Sensers are
always Subjects (with one exception, (22), discussed below). An
overt Senser occurs as the initial element in all but four clauses.
C94, C105a, and A14a have no overt Sensers. This leaves (22) as the

only clause with a fronted Phenomenon.

Ph:O Se:S Me:P
(22)829a I sang-a I hazrat-e ebrahim ke I me-xost
this rock-RA hazrat-E Ebrahim ME-want

'this rock Ebrahim wanted'

In (22), the Phenomenon is highly emphatic; it is fronted and
marked with -ra, as well as having the specific deictic i. Except
for this highly specified instance, Phenomenons follow Sensers. If
the Phenomenon is a 'thing', it comes before the Process; if the
Phenomenon is a 'fact', it comes after the Process, usually realized
as a ke clause. 'Things' are marked with -ra, if objects, or
realized as prepositional phrases. Four of the clauses have no overt
Phenomenon:

67

Se:S Me:P
(23)B70 tana kason-e ke I qabul kard
only person-E accept did

'only some people accepted'

Se:S Me:P
(24)B72 i / qabul kard
she accept did
'she accepted'

Se:S Me:P
(25)BB1 anuzl I qabila I xabar na dast-an
yet tribes information NEG have-3p

'yet the tribes did not know'

Se:S Me:P
(26)C129b maa I me-fam-im
we ME-know-1p

'we know'

In all four processes the Phenomenon is implied and can be
supplied. All clauses, except (27), (28), and A14a, have overt
Subjects, and all of the Subjects are Sensers, with the exception of

(29), a dependent clause, and the highly idiomatic (28).

Me:P Se: thke
(27)C94 nazer sud-an I ke . . . .
observer became ~3p that
'they observed that . . . .'

Me:P
(28)C105a agar / xué-e-san I me-amad
if liking-E-their ME-came

'if they liked'

68

SezAd Cr(Tm):Ad
(29)C134c (i amre) (ke)/bare pimmbar/awal dafar/ amad
this edict that for prophet first time came

'(this edict) (that) came for the first time for the prophet‘

(29) is metaphorical and has been interpreted on that level.
Literally, of course, the process is a material one. (28) is a common
idiom in Dari, the semantic interpretation of which, 'they liked‘,
bears little resemblence to the grammatical realization 'their liking
came'; as a result, there is little to say about the functional
structure of this dependent clause.

All mental processes (except (29), perhaps) show the potential
for Halliday's bi-directional criterion for mental clauses (where
either the Sensor or the Phenomenon can be the Subject of
semantically equivalent active clauses, see Chapter III). In this
text, however, only the 'active' (or 'like') type exist. It is quite
possible that, like true material passives, this kind of passive also
does not exist, or is rare, in Dari. Such phrases as the English 'the
book pleases me' have no equivalent in Dari. If the Phenomenon is
the Subject, the process is realized as a metaphorically material
one as in (29). One would have to say something equivalent to
'reading the book pleases me' (30). Bi-directionality does not appear

as a general pattern in this text, but the data are few.

(30) xandan-e in kitab / ba roi-m I xusi meta
reading-E this book for-me pleasure give

Mental processes are realized grammatically as both

transitive (TrA) and intransitive (lntA, lntB, IntC) clauses.

6 9
Table 3

Mental Processes

E I' . | B l' l'
Senser Subject (15) I Adjunct (1)

Phenomenon Object (4) I Re (3) I Adjunct (3) /Subject (1)
Circumstance Adjunct (1)

6. Verbal processes

Verbal processes are processes of saying, either directly or
indirectly, literally or metaphorically, e.g. 'the clock says five
o'clock'. The Sayer need not be human; it can be an inanimate object
or even the document itself, e.g. 'the report states'. In addition to a
Sayer, a second direct participant is what Halliday calls the
Verbiage, that is, the verbalization itself. A third direct participant
is the Beneficiary or Recipient of the message. Halliday classifies
verbal as a minor process type, and, indeed, in this Dari text, it is
the least exemplified process. There are only eleven verbal
processes in the data: A9a, A13, B31a, B47b, BS4c, BS6, BB3, C98,
C100, C127, C130. The constituent order is highly predictable:
Sayer-Beneficiary-Process-Verbiage. Only (31) differs from this.

Vb:S SyzAd anAd
(31)C98i farman-haa/tawasut-e pitpmbarlbare mardum
this decree-PL via-E prophet for people
V:P
me-ras-an
ME-deliver-3p

'these decrees were delivered via the prophet for the people'

70

(31) has been interpreted as a verbal process, but it is quite
possible that the 'delivery' was a material action rather than a
verbal one. The clause is, at any rate, not a usual verbal process.
God is giving decrees for the people, but the prophet is the
messenger who actually delivers these decrees. Translation of this
clause into English is difficult. 'These decrees' is the grammatical
subject but it is not the agent; it is, instead, the affected
participant. A closer semantic rendering into English might be:
'these decrees arrived for the people through the prophet' or 'the
prophet delivered these decrees for the people'. Functionally, the
participant roles are clear: 'these decrees' is the Verbiage; the
prophet is the Sayer; and 'the people' is the Beneficiary.

A Beneficiary occurs in four of these clauses, (31) - (34),
either as a bare phrase (in three) or as an indirect object marked
with -ra (32). In (33), the Beneficiary follows, rather than
precedes, the Process.

Sy:S Bn:O V:P Vb:ke

(32)C100 mohamad ke amad I arab-aa-ra I goft I ke . .

Mohammed that came Arab-PL-RA said
'Mohammed who came told the Arabs . . . .'

Sy:S V:P Bn:Ad Vb:ke
(33)BS6 xoda I amlr kard-a I bare ma / ke . . . .
god order made-PF for me

'God instructed me that . . . .'

Bn:Ad V:P Sy: Vb:ke
(34)C127 misli ke pe§tar I/ bare tan I goft-om I ke . .
like earlier for you said-1s

'like I told you earlier . . . .'

71

All verbal processes except (35) have a Verbiage, realized as

either a ke clause or a direct statement:
RzAd V:P Sy:

(35)A9a ba estila / be-goy-Im

by expression BE-say-1p
'let's say'

(35) is idiomatic, literally 'we say by expression', but having a
meaning similar to the English 'let's say' or 'as the expression goes'.
It is treated as an indivisible unit and not deconstructed, though
functionally ba estila is Range.

There are few overt Sayers. Sometimes the Sayer is an
impersonal, unspecified general 'they', e.g. (36) and (37).

V:P Sy: Vb:ke

(36)A13 me-g-an I ke . . . .

ME-say-3p
'they say that . . . .'

V:P Sy: Vbzcl

(37)B47b me-goft-an I i-ra bayad maa parasteé kon-im
ME-said-3p this-RA must we worship do-1p
'they said, ”This we must worship."

One native informant translated (37) as, 'the word was that '.
Other sayers are more specific but simply not expressed as overt
Subjects. It is interesting to note here that while mental processes
usually have specific overt sensers, verbal processes do not usually
have overt sayers. This may be true for just these data rather than

the language as a whole. A wider investigation would be necessary
to come to any firm conclusion on this. Though there are few overt

72

Sayers, all verbal processes in the data have Sayers realized as
verbal suffixes, e.g. (35) - (37). Two clauses, however, have Sayers
realized as Adjuncts,(31), already discussed, and (38):
SyzAd V:P Vbzcl
(38)C130 dar u sura I nawlsta kard-a I [Arabic]
in that verse written did-PF
'in that verse was written . . .'

This is an intransitive clause, the Subject of which is not
clear but is most likely the Arabic verbalization that follows the
predicate. The Sayer, however, is clearly the sura, the meaning of
the process being 'the verse says'. Here we have an example of the
document itself being the Sayer and the actual words being the
Verbiage.

Table 4

Verbal Process

E I' . | B I' l': 1
Sayer Subject (9) I Adjunct (2)
Verbiage ke or cl (9) I Subject (1)
Beneficiary Adjunct (3) I Object (1)

7. Identifying processes
In Dari there are three ways of expressing being: identifying,
attributive, and existential. Identifying processes involve two

73

obligatory participants, one the Identified (Id), the other the
Identifier (Ir). In this relational process, one entity is used to
identify another in a relationship of token and value, phenomenon and
circumstance of time, place, etc., or of owner and possession
(Halliday: 1985:113). Identifying processes function to identify or
define an entity while attributive processes merely assign an
attribute to that entity. Halliday says that in English, ”The
fundamental difference between the two is the fact that identifying
clauses are reversible, whereas attributives are not" (1985:114). In
Dari, this reversibility (see discussion Chapter III, p. 18) does not
seem applicable; merely putting one entity in clause-initial position
does not make it the Identified as opposed to the Identifier. Clause
position does not have the same kind of grammatical significance in
Dari as it does in English.

The order of Identified-Identifier is almost built into
Halliday's definition of the two participants and is tied in with the
notion of Theme which will be discussed later. But in Dari, the two
participants are also realized in rather different ways. Identifiers
may be realized as prepositional phrases, e.g. (39) and (40), while all

ldentifieds, if they are overtly realized, are noun phrases.

Id:S lrzAd
(39)B45 but-aa-e kalon-e az I I ba nam-e Iaut o malaut
idol-PL-E large-E from this by name-E Laut & Malaut
l:P

o ozu I yad me-éud

& Ozu memory ME-became
'the large idols of these were called by the name of Laut,
Malaut and Ozu'

74

Id:S lr:Ad l:P
(40)328a u sang I az sang-haa-e asman I -is
that rock from rock-PL-E heaven is

'that rock is from the rocks of heaven'

In addition, there is a sub-group of identifying processes
where the Identified is the Object and the Identifier is an Object
Appositive in a transitive C type clause. These clauses have an
additional participant, the Assigner. In the data, an Assigner is
realized only as a verbal suffix and the verb is always goftan, 'said'.

Two examples follow:

Id:O lr:OA l:P
(41)Bz7c u-ra I ajarl aswad / me-g-an
that-RA Ajarl Aswad ME-say-3p
'they call that Ajarl Aswad (Arabic for Black Rock)‘

ld:O lr:OA l:P
(42)A5 yak-e-éa I §aar bazar I me-goft-an
one-E-those city bazaar ME-say-3p
'one of those they called City Bazaar'

There are 25 identifying processes in the text. A typical
identifying process has the following constituent order: Identified-
Identifier-Process. Most identifying processes are relational
clauses, e.g. (43) - (46), but the data also include transitive (TrC),
e.g. (41) and (42), and intransitive (lntB), e.g. (47), clauses.

Id:S Ir:C l:P
(43)B24 markaz-e xalaf—aa I sar-e madina I bud
center -E khalif-PL city-E Medina was

'the center of the khalifs was the city of Medina'

75

Id:S Ir:C l:P
(44)B84 ma I phambar-e xoda I -st-om
I prophet-E god is-1s

'I am the prophet of god'

Ir:C l:P Id:
(45)B74 xalif-e awal I bud-Q
Khalif-E first was-3s

'he was the first khalif‘

Ir:C l:P Id:
(46)826a bad dar awaxar l/ make I sud-6
later in the end Mecca became-33

'later in the end it became Mecca'
Id:S Ir:SA l:P
(47)B71 yak zan-e hazrat-e mohamad/xadija/nam dast

one wife-E Hazrat-E Mohammed Khadija name had
'one wife of Mohammed was named Khadija'

Table 5

Identifying Processes

E|"| BI'I'

Identified Subject (19)
Identifier Complement (17) IOA (5) IAdjunct (2) ISA (1)
Assigner Subject (6)

8. Attributive processes

Attributive processes differ from identifying processes in
that the latter are equative while the former are not. In an

76

identifying process, the two participants are of equal status, and
which is the Identified and which is the Identifier is often a matter
of perspective based on context and grammatical and thematic
realization. In an attributive process, one participant is an

attribute of the other, not to be equated with it but a single property
of that participant.

Halliday distinguishes in English between 'John is the leader',
an identifying process, and 'John is a leader', an attributive process.
Dari has no articles as such, so other ways must be found to indicate
specificity, or to distinguish between a member of a class
(attributive, e.g. 'Ford is a car') and an identity (e.g. 'Ford is the car
to buy).

In the clauses in this text, if Identifiers are realized as noun
phrases, these noun phrases are highly specific, e.g. (45) and (46). In
contrast, Attributes are usually realized as adjective phrases or as
general noun phrases after da§tan, 'had'. Some examples are:

Ar:O At:P Ca:
(48)C111a zurg I daét-an
power had-3p
'they had power'

Ca:S Ar:O At:P
(49)B42 ar qabila I yak but I da§t
every tribe one idol had

'every tribe had one idol'

Ar:C At:P Ca:

(50)C113a saxi o garlb / bud-an
propertiless and poor was-3p

'they were propertiless and poor'

77

Ca:S Ar:C At:P

(51)B68 I I qarat I -ast

this wrong is
'this is wrong'

As in identifying processes, there is usually an overt Subject
in attributive processes. The Subject is always Carrier and there
are no instances of fronted Attributes in the data. The invariable
constituent order is: Carrier-Attribute-Process. In one clause, the
Carrier is modified by a ke clause that contains both the Attribute
(realized as an Adjunct) and the Process.

Ca: ArzAd At:P

(52)A14b (en-aa nam . . . ke)/da tarix-e/tazakuri yaft-a

this-PL name that in history-E mention found-PF
'these names that were mentioned in history'

The data contain 30 instances of attributive processes,
realized grammatically by all clause types except existential:
transitive (TrA), e.g. (57), intransitive (lntB), e.g. (55) and (56), and
relational (ReIA, RelB, ReIC), e.g. (53) and (54); most attributive (18

out of 30) processes are relational clauses.

Ca:S Ar:C At:P
(53)CB7 zabur I az hazrat-e daud I -ist
Old Testament from Hazrat-E David is

'the Old Testament is from David'

Ca:S Ar:C At:P
(54)C93 mam-e xud I bira I sud-a bud-an
tribes-E self lost became-PF was-3p
'the tribes themselves had become (religiously) lost'

78

Ca:S Ar:Ad(Lc) At:P
(55)385 ma I az taraf-e xoda I amad-im
I from direction-E god came-1P
'I came from God'

Ca:S Bn:Ad Ar:Ad At:P
(56)C84 Ina awal sura/bare ma/az qoran-e sariflamad-a
this first verse for me from Koran-E holy came-PF
'this first verse came to me from the holy Koran'

Ca:S Ar:O At:P

(57)A1 kabul I éakle emroz-ara I na daét
Kabul like today-RA NEG had
'Kabul wasn't like today'

Table 6

Attributive Process

E I" | B I. I.
Carrier Subject (30)

Attribute Complement (18) / Object (6) I Adjunct (6)
Beneficiary Adjunct (1)

Existential processes

Existential processes are one participant processes, the

meaning of which is that something exists or happens. This type of

process is much more common in Dari than in English. In the data,

there are eighteen. All grammatically existential clauses are also

79

functionally existential, and with one exception, the reverse is also

true. The exception is:

Et:S Ex:P
(58)A11dewaI-aa-e gall-(e) kalon kalon/saxt-a sud-a bud
wall-PL-E mud large large made-PF became-PF was

'very large mud walls had been made'

Structurally, (58) is not unusual; this seems to be a common
way of stating an existential proposition in Dari. On one level, the
clause appears to be a material middle-voice process, but the sense
of the clause is that there were very large mud walls in old Kabul, a
statement of existence. There is no agency involved even though our
knowledge of the world tells us that the walls were brought into
existence through some conscious outside source. Thus,
grammatically, this is an intransitive A clause. (59) is a similar
kind of clause; it is functionally an identifying process but

grammatically an intransitive B clause.

Id:S lr:Ad l:P
(59)B46a yak but-e az In-aalaz yaqud/saxta sud-a bud
one idol-E from this-PL from ruby made became-PF was
, 'one of these idols had been made from ruby'

These seemingly material middle-voice clauses warrant a
more extensive look that is beyond the scope of this analysis.
Another existential clause that is difficult to classify is:

Et:S Ex:P
(60)C116 bedat I -is
sin is

'it is a sin'

80

Semantically bedat is an attribute of all the actions
previously listed, but the sense here (reinforced grammatically) is
not 'all these things are sins', an attributive process, but a more
unspecified ’it's a sin'. This interpretation is in accord with native
intuition, and the narrator intonationally isolates the clause by
large pauses on either side of it.

All other existential processes are quite uniform in structure.
The only participant is the Existent, which is the Subject of the
clause realized as a noun phrase. This noun phrase may be simply a
noun as in (60), a noun with a post-modifier, e.g. (61) and (62), or a
noun modified by a ke clause following the verb, e.g. (63). Some
clauses have clause-initial circumstantial elements, e.g. (61). With
the exception of (58), the verb is a form of either budan or "sudan.

Cr(Lc):Ad Et:S Ex:P

(61)A2b dar markaz-e kabul II yak herb-e kalon I -is

in center-E Kabul one fortress-E large is
'in the center of Kabul is a large fortress'

Et:S Ex:P
(62)B27b yak sang-e sia I -ist

one rock-E black is

'there is a black rock'

Et:S Ex:P Et:S

(63)086b injil I -ist I ke az hazrat-e esaa-s
Bible is that from Hazrat-E Jesus-is
'there is the Bible that is from Jesus'

Cr(Tm) Et:S Ex:P
(64)C135 bad az u II amr-e namaz I sud
after that edict-E pray became

'after that the edict of prayer came into being'

81

10. Conclusion

This functional analysis of the clause structure of Dari has
revealed some interesting features of that language specifically, as
well as some generalizations about all languages. While it is probab-
ly true that all languages realize the same basic semantic processes
in one way or another, these processes are distributed differently in
each language. Dari seems to place considerable weight on grammati-
cal realization for expression of functional roles and little weight
on constituent order. Morphological markers (such as -ra) also play
an important part in the determination of functional structure. Dari
seems to make greater use, than does English, of existential pro-
cesses for the expression of concepts; it is common in Dari to state
the existence of something as a separate proposition. This is not as
usual in English. Dari also makes relatively frequent use of range.
This is seen in the semantic loading of the predicate as well as in
the preponderance of Range as an important functional participant.
Arguments in a proposition are often more indirect than in English;
for example, Dari uses prepositional phrases where English would
use direct or indirect objects. All of these features affect the

expression of ideas and the way in which one views the world.

82

Endnotes

1. I found the addition of a behavioral category to be highly
insightful in my analysis of Lawrence Durrell's W
(Rashidi 1983a; 1983b, 1985).

2. An 'overt' element here, as elsewhere, is defined as one which is
expressed as a separate clausal constituent, as opposed, say, to
being marked as a verbal suffix.

3. Hazrat is a general religious term of address that has no
meaningful English gloss.

CHAPTER VI
ERGATIVE VIEW OF DARI

1. Traditional view of ergativity

Traditionally ergativity has been viewed as a syntactic coding
system; a language either codes its clausal participants according to
an ergative classification or an accusative one. In an accusative
system, grammatical subjects are distinguished from grammatical
objects. In an ergative system, however, the basis for classifica-
tion is functional: an agent is distinguished from an affected
participant. Thus in the following English clauses the similarity of
participants is viewed differently in an accusative system than in
an ergative system:

1) John hit the ball.

2) John ran.

From an accusative viewpoint, John has a similar function in
both clauses, that of subject, the door of the action. If the language
actually marks for function, or case, John will take the same case
in both clauses. This system is the basis of the traditional typology
of clauses into transitive and intransitive; a transitive clause, 1),
has an object while an intransitive clause, 2), does not.

From an ergative viewpoint, John in 2) has the same function
as the ball in 1); each is the affected participant of the clause, the
core constituent without which the proposition would not exist. If a
so-called ergative language actually marked for case, John in 2) and

83

84

the ball in 1) would take the same case markings. The dichotomy
from an ergative viewpoint is not one of transitive versus intransi-
tive (1 object) but one of :l: agency. Clause 1) has an agent while
clause 2) does not. Thus, from an accusative point of view (or in an
accusative language), every clause has an obligatory, at least
underlying, subject and an optional object; while from an ergative
viewpoint (or in an ergative language), every clause has an
obligatory affected participant and an optional agent. Traditional
linguistics does not appear to have a specific label for the two types
of clauses that would be analogous to the transitive/intransitive
distinction. In essence, of course, a semantically transitive clause
is one that has both an agent and an affected participant.

Traditionally, then, languages have been classified as either
accusative or ergative. An accusative language is one that codes,
either on the surface or as an underlying principle of organization,
the grammatical functions of subject versus object, regardless of
the logical roles of these constituents. In contrast, an ergative
language is one that codes the roles of agent versus affected
participant regardless of the status of the grammatical constituent
as subject or object.

With the recent recognition of the importance of functional
roles in the analysis of grammatical structure, the concept of
ergativity has been used increasingly in the exploration of various
languages (see Bashir 1986; Cooreman 1982 and 1983; Cooreman,
Fox, and Givon 1984). Most of this work recognizes not only the
importance of functional roles, but also the contribution that

discourse structure makes to the selection of surface realization

85

features. The way a clause is realized is dependent upon the
function of the clause as a whole in the discourse as much as the

resources available to the user of that particular language.

2. Halliday's ergative interpretation of transitivity structure

Halliday has been in the forefront of this recognition of the
importance of functional roles, making use of the notion of
ergativity, not as a classification device, but as an additional way
to view the structuring of any language. Though Halliday's
development of ergativity has been treated within systemic
linguistics less extensively than his major model of functional
transitivity (process typology), his concept of ergativity can prove
profitable and insightful in the analysis of a language. From this
viewpoint processes are characterized as being either self-
engendering (- agency) or engendered from without (+agency). It is
this factor of iagency that Halliday calls the ergativity option
(1985:147). From an ergative viewpoint each process has an
affected participant which Halliday calls Medium (Md). In addition,
there may or may not be an Agent (Ag). If there is agency involved,
either explicit or implied, then the process is effective voice; if the
process is minus agency then the process is middle voice. An agent
need not be overtly realized as the structural function Agent for a
process to be effective, but agency must semantically underlie the
process.

3. An ergative view of transitivity structure in Dari

This ergative distinction based on agency seems an appropriate

86

one for Dari, as in Dari there is no active/passive distinction; Agent
is always realized as Subject. Thus, Halliday's system of transi-
tivity and voice for English (1985:151) (see Figure 2, p.20) can be
modified for Dari as shown in Figure 4:

 

middle (Medium/Subject)
process \ [

effective (Agent/Subject)
+Agent (Medium/Object or Adjunct)

Figure 4
System of transitivity and voice for Dari

Example B15 is middle voice; B31a is effective voice:

Md:S
B15 . . . rakbiléaa l hukumat me-kard
Rakbil Shah govern ME-did
'Rakbil Shah was governing'

Md:O Ag:S
I31a i-ra I xoda / az asman I ha zamin I andoxt-a
this-RA God from heaven to earth dropped-PF

'this god dropped from heaven to earth'

In our analysis, we have started with the core notion of
Medium. Every clause has a nuclear participant in an ergative view
of process. If the Clause has a single participant, then that

participant is Medium. If the process has some outside engendering

87

force, then that participant is Agent. We will be defining agency
here in a, perhaps, narrower, more semantic sense than Halliday does
for English, as will be apparent. Other direct participants from an
ergative viewpoint are Range and Beneficiary. Range is the scope or
domain of the process, a participant that seems of more prominent
concern in Dari than in English. Beneficiary is the one for whom the
process is explicitely enacted, a concept we have discussed

elsewhere.

4. Analysis of the data

In analyzing the processes in the data, it was not always clear-
cut which participant was Medium. The notion of Medium, like other
functional notions, is a combination of semantic and grammatical
considerations. At the core is the semantic intent within the
proposition, but there should be some consistency of grammatical

realization within a particular language.

4.1 Material processes

Halliday says that in a material process in English, Medium is
Goal if there is a Goal; If there is no Goal, then Medium is Actor
(1985:147). This works most of the time for Dari. There are
processes, however, where Range seems to be the core participant.
The most obvious of these are clauses that involve Adjuncts of
Range that are very object-like, e.g. (2), (3), (4). Dari often realizes
notions indirectly using prepositional phrases, where English would
express the same notion more directly as a direct object. We have

already noted the distinction earlier with such ideas as 'the car hit

88

the bus' or 'Ralph married Josephine'. We see the same distinction in
the data between (1) and (2).

Md:S:O Ag:A:S
(1)821 besior mamlakat-haa-ra I graft-an
very country-PL-RA took-3p
'they took many countries'
Md:A:S R
(2)322 hazrat-e omar/taa qismat-aa-e . . . I fatha kard
Hazrat-E Omar until section-PL-E conquer did

'Omar conquered up to sections of . . .'

In (1), there is an obvious, clear-cut agent, 'they', realized as
the verbal suffix -an. There is also a clear-cut Goal realized as a
direct object marked with ~ra. The Medium is then 'many countries'.
But the ergative roles in (2) are not as clear-cut. Semantically,
clauses (1) and (2) seem similar. In (2), however, the verb fatha
kard, 'conquered', cannot take a direct object; it is, in effect, an
intransitive verb. The 'object' then must be realized indirectly as
the object of a preposition. This participant is grammatically an
Adjunct and functionally Range. But it seems that the core
participant, as in (1), is the domain of the conquering. It also
seems that the Actor is Agent, not Medium. One might wish to re-
analyze the Adjunct in (2) as Goal, but there are less goal-like
examples that follow a similar pattern. In (3), the verb gasdan,
'abandon', like fatha kard, 'conquered', takes only an adjunct, not a
direct object, but az din, 'from religion', seems more range-like
than the Range of (2); 'religion' is not a concrete object. 'Abandon
religion' differs from 'hit a ball' or even 'conquer Amman'. In (4),

'ab

are
wl"
co
Re
an
die

89

'abandon the way of God' is even more range-like. Thus, it seems to
make sense in (3) and (4) to analyze the Actors as Mediums. These
are, in addition, processes that are very close to mental processes
where the senser is Medium. (2) remains a problem, but the most
consistent analysis seems to be the designation of omar as Medium;
Range then remains the same notion both from a transitive and from
an ergative viewpoint. This analysis also recognizes the semantic

distinction between an Adjunct and an Object.

Md:A:S R
(3)B36 akwam-e arablstan I hama I az din I gaéd-an
tribes-E Arabia all from religion abandon-3p

'the tribes of Arabia all abandoned religion

Md:A:S R

(4)B53b ln-aa I az roi-e xoda I gasd-an
this-PL from way-E god abandon-3p
'they abandoned the way of God'

Two other material processes in the data present an opposite
problem, grammatical direct objects that seem more range-like than
goal-like, (5) and (6).

R Md:A:S
(5)C125 sur-e al hamd-a I bo-xan-ln
verse-E al Hamd-RA BE-read-2p
'you should read the verse al Hamd'

R Md'Azs
(6)6126 sur-e exlas-a I bo-xan-in
verse-E EkhIas-RA BE-read-2p

'you should read the verse Ekhlas'

90

Both of these clauses are optatives. In each, sura indicates
the domain of the reading, not a goal that is being acted upon, even
though the noun phrase is marked with -ra. Thus, 'you' is not a true
agent but more of an affected participant. Again, these clauses are
not very material; these are close to verbal processes. In this light,
we will interpret 'you' as the Medium and the verse as the Range of
the process 'read'. Again, this is consistent with the transitive
interpretation of these clauses.

This brings us to a third group of material processes where
Range seems to play an important role. In Dari, it is not uncommon
to have clauses that consist only of a predicate, where the Subject
is realized only as a verbal suffix. The entire semantic loading is in
the predicate. Many of these predicates are complex verbs where the
content verb is very close to being range, e.g. (7), (8), (9).

(7)B34c zlarat me-kard-an

obeisance ME-did—3p
'they paid religious obeisance'

(8)0110 duzdi me-kard-an
thievery ME-did-3p
'they were robbing'

(9)0103 qimor me-zad-an

gambling ME-hit-3p
'they were gambling'

Though the subjects here come close to being agents and,
particularly in (9), the content verb is quite like a Medium, it seems
best to analyze these as middle voice with only a Medium realized as
the verbal suffix -an, 'they'.

91

4.2 Mental processes

There is little to say about mental processes in Dari that
differs from Halliday's analysis for English. Senser is Medium and
the Phenomenon is Range in the typical consciousness to phenomenon
type process, the 'like' type in Halliday's terms (1985: 149). If the
process is encoded in the other direction, however (the passive
'please' type), the Phenomenon is Agent. We find two of these
clauses in the data:

Ag:Ph:S Md:SezAd
(10)0134c (i amre)/bare piqambar/awal dafarlamad

this edict for prophet first time came
'(this edict) came the first time for the prophet'

(11)0105a agar I xrg-e-san I me-amad

if liking-E-their ME-came
'if they liked'

(10) presents no problems: amre, 'edict', which is the
Phenomenon encoded as Subject, is the Agent, while plmm bar,
'prophet', the Senser, becomes Medium. The problem arises with (11)
where there is no Senser. If we designate xué-e-san as Agent, the
process has no Medium. This is a rather common way of expressing
mental processes in Dari, so we cannot just dismiss the clause as an
unusual metaphor. The process is, of course, highly metaphorical.
One solution might be to split the clausal constituent and call san,
'their', Medium. This seems, however, to ignore too blatantly the

grammatical realization of the idea expressed by (11). A better,

92

though not completely satisfactory, solution is to call the single

participant the Medium. We have chosen this solution.

4.3 Verbal processes

In Dari verbal processes, as in English, the Sayer is Medium,
the crucial participant, and the Verbiage is Range. Many verbal
processes in the data have Beneficiaries as well, the receiver of the
message. As in other processes, metaphorical clauses have been
analyzed at that level rather than at a non-metaphorical level,
usually material. As a result, there is one clause, (12), that does
not fit the typical description of a verbal process. The Verbiage is
Subject and the Sayer is realized as an Adjunct. Nonetheless, an
analysis of this process as verbal instead of material helps to
highlight the distinction in Dari of a facilitator from a true agent.
On a semantic level, the true agent is God. It is not the picambar
that has sent the decrees; he is only the means through which the
people receive them. Thus, it seems appropriate that plcpm bar is
Medium, 'decrees' is Range, and 'people' is the Beneficiary. This is

consistent with the concept of Agent in Dari that is being espoused

here.
R:Vb:S Md:SyzAd
(12)098 I farman-haa I tawasut-e picpmbar /
this decree-PL via-E prophet
Bn

bare mardum I me-ras-an
for people ME-deliver-3p
'these decrees were delivered through the prophet for the people'

93

4.4 Identifying processes

Halliday argues that in English identifying processes, Value
(for all practical purposes, the Identifier) is Medium and Token
(essentially the Identified) is Agent (1985:148). This seems to
stretch the notion of agency. To analyze Token as Agent helps
account for the grammatical reversibility of identifying clauses in
English (e.g. Sarah Is the leader versus the leader Is Sarah).
The notion of reversibility and, therefore, active/passive identifying
clauses, makes little sense for Dari. Though the Identifier in theory
may be fronted in Dari (there are no examples in these data), the
Identifier is never Subject as it may be in English. Since in Dari,
Agent, if present, is always the Subject of the clause and
identifying clauses have two obligatory participants, the Identifier
can never be Subject, and, therefore, there is no 'passive' identifying
process as in English.

A more illuminating analysis for Dari identifying clauses
would be one similar to that Halliday posits for English attributive
clauses: Identified is Medium, the core participant, while the
Identifier is Range. An Agent would only appear as an Assigner (As)
(e.g. 10.91 made Sarah the leader). There are a number of these
clauses in the data, e.g. (13), (14).

Md:ld:O RzlrzOA Ag:As:S

(13)B27c u-ra I ajarl aswad I me-g-an

that-RA [rock black]Arb ME-say-3p
'they call that Ajarl Aswad'

94

Md:ld:O R:lr:OA Ag:As:S

(14)A7 yak qismat-e Star I demazan I me-goft-an

one section-E city Demazang ME—said-3p
'one section of the city they called Demazang'

The unsatisfying aspect of this analysis is that often the
Medium is merely realized as a verbal suffix, e.g. (15), and the core
of the message seems to lie in the Value (Identifier), not in the
Token (Identified). Nonetheless, it is not the message structure as
such that is the concern here, but the participant that is 'the one
that is critically involved' (Halliday 1985:147).

R:Id:0 Md:ld:S

(15)B74 xalif—e awal / bud o

khalif-E first was 33
'he was the first khalif'

4.5 Attributive processes

The above analysis also makes the interpretation of
identifying and attributive clauses more parallel, reflecting their
closeness grammatically, functionally, and semantically. In an
attributive process, the Carrier is Medium and the Attribute is
Range. If there is an Attributor, it is Agent There are, however, no
Attributors in the data, and they are probably rare in the language in
general. There is, for example, no Dari equivalent to the English I

am finished (with something). In Dari, one says, 'it is finished'

instead:
*xalas-ast-om 'I am finished'
finish-is-1s
xalas-ast 'it is finished'

finish-is

95

5. Conclusion

The Medium, then, is the core participant in a nuclear
(ergative), as opposed to a linear (transitive), view of process.
Other participants radiate out from this core: Range, Agent,
Beneficiary. Every process has a Medium. In Dari, the next most
prominent participant is Range. Agency is not a dominent feature of
Dari clauses. Though we have defined Agent more narrowly than
Halliday, the distinction between plus Agent and minus Agent is an
important one in Dari. It remains a consistent notion across process
types. All process types except existential have the potential for
agency (though we can only posit this for attributive processes). An
ergative interpretation of process has both highlighted the
centrality of the core participant, Medium, and shown the
prominence in the language of Range. As Halliday points out
(1985:149), all languages are probably a combination of ergative and
transitive structure. This is certainly true for Dari; while it is
enlightening to view Dari clause structure in terms of transitivity
process type as we did in Chapter V, it is also profitable to view
Dari clauses from the consistent viewpoint of ergativity. The

ergative functions are summarized in Figure 5.

96

 

 

 

 

 

 

 

 

 

 

 

 

 

function Transitive

Ergative Ma Me V At I Ex

Medium G;A (mid.) Se Sy 0a Id Et

Agent A (eff) Ph('please') Attributor As

Ben Bn Bn Bn

Range R Ph('like') Vb Ar l r
Figure 5

Ergative Functions

6. Some examples analyzed

Some examples of an ergative interpretation of clauses in the
data follow:

R:lr:C Md:ld:S
(16)B52 lxtlador-e xon-e kaba I bud -6
authority-E house-E rock was 3s
'he was the authority of the Hona Kaba'

Md:ld:S R:lr:C
(17)B24 markaz-e xalaf-aa I éar-e madina I bud
center-E khalif—PL city-E Medina was
'the center of the khalifs was the city of Medina'

Md:ld:O R:lrzOA Ag:As:S
(18)A5 yak-e éa I sar bazar I me-goft-an

one-E this city bazaar ME-said-3p

'one of these they called the City Bazaar'

R:Ar:0 Md:Ca:S
(19)C113a saxi o garlb I bud-an
propertiless & poor was-3p

'they were propertiless and poor'

97

Md:Ca:S R:Ar:O

(20)B57 i but-aa I qabil-e parasteé I n-est
this idol-PL worthy-E worship NEG-is
'these idols are not worthy of worship'

Md:Ca:S Bn R:ArzAd
(21)084 Ina awal sura/bare malaz qoran-e sariflamad-a
this first verse for me from Koran-E holy came-PF
'this first verse came to me from the holy Koran'

R:Ar:O Md:Ca:S

(22)0111a zurg I daét-an
power had-3p
'they had power'

Md:Sy:S R:VbzAd
(23)B73 elan kard-a I ke pbambar-e xoda sud
announce did-3s that prophet-E god became
'he announced that he had become the prophet of god'

Md:Sy: S Bn R:VbzAd
(24)B56 xoda I amlr kard-a I bare ma I ke . . .
god order did-PF for me that
'God instructed me that . . .'

Md:Se:S R:Ph:O
(25)B69 pader-aa-e jad . . .Ii-ra/awal/qabul na kard-an
ancestors this-RA first accept NEG did-3p
'the ancestors . . . at first did not accept this'
Md:Se:S R:thAd
(26)A2a maa I fikir kon-im / ke . . .
we think do-1P that

'I think that . . .'

98

R:Ph:O Md:Se:S

(27)B29a i sang-a I hazrat-e ebrahim I (ke) me-xost
this rock-RA hazrat-E Ebrahim (that) ME-wanted
'this rock Ebrahim wanted'

Md:G:O Ag:A:S Cr(Lc) Cr(Lc)

(28)B31b l-ra I xoda I az asman I ba zamin I andoxt-a
this-RA god from heaven to earth dropped-PF
'this rock god dropped from heaven to earth'

MdezO Cr(Mn) Ag:A:S
(29)C112 doktor-aa-e mardume/ba zurg/me-graft-an
daughter-PL-E people by power ME-took-3p
'they took people's daughters by force'

Md:G:O Ag:A:S
(30)0132a tu-ra I baujud award-a
you-RA brought into being-PF(3s)
'he brought you into being'

Mdfizs
(31)B19 in-aa ham I xelafat kard-an
this-PL also khalifing did-3p
'they also acted as khalifs'

Cr(Lc) Md:A:S
(32)A13a dar zamon-e qadim I (ke) hindu-ham da kabul
in old-E section (that) Hindu-all in Kabul
zendagi me-kard-an
live ME-did-3p
'all the Hindus in Kabul were living in the old section'

R:Ad Md:A:S
(33)C123 dar majid I b-ur-in

in mosque BE-go-2p

'you should go to the mosque'

CHAPTER VII
TI-IEMEIRHEME STRUCTURE

1. Introduction

Halliday sees three principal kinds of meaning embodied in the
structure of a clause: clause as message, clause as exchange, and
clause as representation. The realization of each of these meanings
is a configuration of functions, the significance of which lies in the
relationship of these functions to other functions. These various
structures mapped on to each other constitute the clause which is
the basic meaning unit of language. Of these three kinds of meaning,
message structure, or Theme/Rheme structure, is perhaps the most
central to understanding the nature of functional relationships and
their various realizations, because it is the message structure that
gives the clause its status as a communicative event (Halliday 1985:
38).

2. Halliday's view

The Hallidayan view of message structure is one of focus on
Theme and its role as an organizer of the message. One element in
the clause is given special status and "enunciated as the theme”
(Halliday 1985: 38), that which the clause is about. This
announcement of Theme has different realizations in different
languages. In English, Theme is signalled by putting it first. Though
Halliday is quick to point out that sentence position is not a

definitional aspect of Theme, in fact, work on Theme in English

99

100

assumes that Theme is the first ideational element of the clause and
proceeds from there. Rheme is generally ignored as a notion to be
explored. Indeed, Rheme is simply what is left over after Theme is
identified. From a Hallidayan perspective, Theme seems to be the
crucial organizer of message structure and, as such, the element to

focus on in any analysis of text.

3. Firbas' view

Jan Firbas views message structure from the opposite
direction of Halliday. His theory of Functional Sentence Perspective
(1966a; 1966b; 1987) focuses on Rheme instead of Theme. He
proposes the criterion of communicative dynamism (CD), whereby
any clausal element that has meaning, and, therefore, participates in
the development of the communication, is assigned a value of CD. CD
is a relational concept. The degree of 0D carried by a sentence
element is the extent to which it pushes the communication forward.
The element that carries the lowest degree of 0D is the starting
point of the communication, or Theme. The element carrying the
highest degree of CD is the Rheme proper which completes the
development of that proposition.

Distribution of CD is determined by an interplay of factors
including: linear modification, context-dependence/independence,
semantic content, and intonation. For Firbas, ”the foundation-laying
elements set up the theme and the core-constituting elements the
non-theme, within which the transition and the rheme can be
established" (1987: 34). Within this framework, Rheme is the

101

element towards which the sentence is prospected. Every element is
in competition for the position of Rheme.

For Firbas, Theme/Rheme is essentially, in Hallidayan terms,
ideational and highly content-based. It is also a notion that has firm
psychological underpinnings. In the normal course of conveyance of
information, human speech works to produce a linear string of
elements, and these elements follow each other in ascending order
of the degree of 0D contained, provided there is no overriding
interference such as contrast or marked intonation. So the usual
order is theme-transition-rheme (Firbas posits a unit of message
structure between Theme and Rheme which he calls transition.) The
order, however, may vary. Different linear arrangements can
produce different orientations, but do not necessarily do so (Firbas
1987:40). Unlike Halliday, Firbas does not tie Theme to sentence-
initial position. Firbas would not, for example, distinguish between
the following:

A flyI settled/ on his hair

Rheme Theme

On his hair/ a fIyI settled

Theme Rheme
(The verb semen adds little beyond the mere fact of 'appearance on
the scene' and lacks 0D altogether.)

4. Prague School foundation
These two views of message structure are not really at odds;
Halliday and Firbas simply approach Theme/Rheme from different

102

angles. Both views are based on the Prague School notions of Theme
and Rheme (whence the terms derive) and are quite semantically-
based. Most of our current understanding of Theme/Rheme arises
from the work of Vilem Mathesius, the founder of the Prague School,
and later Frantisek Travnicek's exposition and criticism of the ideas
of Mathesius (Firbas 1966a).

Mathesius' notion of Theme/Rheme is tied to the criterion of
Known/Unknown information from the decoder's point of view, or
probably more accurately, from the encoder's assessment of the
decoder's point of view. But this concept of Known is a broad one,
including what is accessible from the text, either anaphorically or
cataphorically; the context of situation; or even the gnostology1 of
the decoder. Travnicek objected to the inclusion of this broad
concept of Known as a criterion for themehood. He narrowed but
delineated the concept of Theme to ”the sentence element that links
up directly with the object of thought, proceeds from it and opens
the sentence thereby” (as cited in Firbas 1966a: 269).

Both Halliday and Firbas separate the notions of Theme/Rheme
and Known/Unknown, but neither divorces the two concepts
completely. For each theorist, the concept of Known/Unknown is not
irrelevant to message structure but neither is it a definitional
feature.

Halliday prefers the terms given and new instead of Known
and Unknown. The interplay of Given and New makes up the
information unit which is to be distinguished from the message unit,
conceptually at least. Given/New is decoder-oriented while
Theme/Rheme is encoder-oriented. Both are, however, as Halliday

103

points out, speaker-selected (1985: 278). The two notions often
coincide; the unmarked pattern is for Theme to be mapped on to
Given, and for Rheme to be mapped on to New.2 Given for Halliday is
a broad notion, indeed, including not only that which is retrievable
from the text or the context of situation, but also ”something that is
not around at all but that the speaker wants to present as Given for
rhetorical purposes" (Halliday 1985: 277).

Firbas does not use either the Known/Unknown dichotomy of
the early Prague School or Halliday's Given/New. Instead, he speaks
of context-dependency or context-independency. A sentence
constituent is context-dependent if it is retrievable from the
immediate context, either verbal or situational (Firbas 1987).
Context-dependency is a complex and graded phenomenon; an item,
such as a pronoun, that is retrievable from the immediate situation
contributes more to CD than a textually given item but less than a

context-independent item.

5. Other interpretations: Theme and Topic

Theme, in the early Prague School sense of sentence-level
Theme, seems to receive little attention outside of Systemic and
European Functional linguistics. Most not immediately connected
with either school equate the concept of Theme with Topic. Little is
said about Rheme at all. Wallace Chafe, for example, in an article in
Charles Li's book, W discusses the various packaging
statuses that a noun may have. He uses Halliday extensively but in
the end treats Topic and Theme as synonymous terms (Chafe 1976:

49). It is interesting to note that Topic is the only packaging device

104

that Chafe discusses that has varying definitions cross-linguistic-
ally. For him, Topic in a topic-prominent language such as Chinese
is a different notion than Topic in a subject-prominent language
such as English. He states that what the term topic means ”seems to
differ from language to language" (1976: 55). And none of these
notions of Topic is the same as the notion of Theme, as conceived by
the early Prague School and developed by Halliday and Firbas.

Another linguist who has equated not only Topic and Theme,
but also Theme and Given, is Talmy Givon. He begins his introduction
to W by noting the shift of attention in
linguistic study ”from the purely structural notion of ‘subject'
toward the more discourse-functional notion of 'topic', or under
some other guises 'theme" (1983:5). He goes on to state that the
sources most often cited were--and are--the Prague School (Firbas
1966a and 1966b), the Firthian tradition (Halliday 1967), and Dwight
Bolinger. He describes this collective influence:

In one form or another, the various strands of this tradition
tended to divide sentences ('clauses') into two distinct components,
one of them the 'focus' ('rheme', 'comment', 'new information'), the
other the 'topic' ('theme', 'old information'). And it was the second,
the topic, which all early practioners would then link to discourse
structure, communicative dynamism, functional sentence
perspective etc., in ways that tended to be often both vague and
mysterious” (1983:5).

While Halliday's 1967 exposition of Theme is less detailed
than his more recent (1985) descriptions, his conceptualization and

analysis of the realization of Theme in English is quite specific.

Firbas' exposition, even in 1966, was quite specific and detailed.

105

Both theories have heavy semantic underpinnings, however, which
make them not entirely acceptable to more quantitatively and
surface-oriented linguists such as Givon. At any rate, Halliday and
Firbas have separate conceptualizations, and if one lumps their
theories together, as Givon has done, it is understandable that the
result is 'mysterious‘.

Givon's definition of Topic differs in a number of ways from
either Halliday's or Firbas' notion of Theme. Though for Givon Topic
is a functional entity, it is not a discrete entity. In other words,
Givon is looking at 'topicality'. This view of Topic is one of
information flow tied firmly to Given/New or retrievability (he uses
the concepts interchangably). Givon speaks of process-ibility and
measures this ease of processing by setting up a cross-language
scale. His is a quantitative model where such things as length of
absence and interference of other Topics are used to determine the
'availability' in the decoder's cognitive 'file'. While the various
analyses based on Givon's model (see Cooreman 1982 and 1983;
Cooreman, Fox and Givon 1984) yield some interesting findings, what
they show has little to do with Theme as conceptualized in present
Systemic or European Functional theory.

A recent Ph.D. dissertation by Ali Asghar Aghbar on case in
Persian also uses Halliday's conceptualization of Theme in
discussing 'topicalization.’ Like Chafe, Aghbar makes use of the
notion of 'psychological subject.’ He restates Halliday's definition
of Theme as ”that part of the clause about which a message has been
asserted and which usually occurs in sentence-initial position”
(1981: 179). He adds that he will call this notion 'topic' and

106

proceeds to discuss tapicalization, not thematization. Aghbar's
points on topicalization in Persian are valid and interesting but they

may or may not be pertinent to thematic selection in Persian.

6. Defining Theme and Rheme

In the present attempt to come to some kind of understanding
of this elusive notion, several parameters of Theme/Rheme
structure have emerged. First, Theme and Rheme are encoder-
oriented as opposed to decoder-oriented. Theme is, in Halliday's
words, "what I, the speaker, choose to take as my point of departure"
(1985: 278). Rheme is what the speaker chooses to impart as the
essential message. This is a psychological notion, not just a
packaging strategy. Second, Theme and Rheme are essentially clause-
level constituents. The clause from a textual point of view is the
unit of information processing, and while Theme and Rheme may
exist on other levels, the basic message unit is the clause. Third,
both Theme and Rheme are discrete elements represented by clausal
constituents (i.e. one can delineate a unit as Theme or Rheme).
Fourth, Theme and Rheme are universal notions; that is, all
languages have Theme/Rheme structure and every clause has at least
a Rheme, or the clause would have no communicative value.

These parameters distinguish Theme/Rheme from other related
notions. First, Theme is not to be equated with Given (or Known)
which is decoder-oriented. The two notions, of course, often
coincide and are realized by the same element, but Given is not a
definitional aspect of Theme and, as we will see, Theme is quite

often new information. Given will be defined here as that which is

107

already in the mind of the bearer. New then is that which is not at
the moment of speaking already in the mind of the hearer. Rheme is
more closely tied to the concept of New than Theme is to Given.
Rheme will most likely be new information or the speaker would not
be mentioning it, but New is not a definitional aspect of Rheme.
Second, Theme is not the same as Subject. Halliday distin-
guishes the two concepts by relating Theme to the late nineteenth
century concept of the psychological subject and Subject to the
notion of the grammatical subject of the same period (1985: 33-34).
Subject is more intimately tied to surface features of grammati-
cality than is Theme. Chafe describes Subject as ”the hitching post
for the new knowledge" (1976: 44). The primary result of hearing
the assertion is that you know something new about the Subject.
This is different from ”point of departure of message” which is more
textually-oriented and more intimately linked to discourse flow.
Third, Theme is not synonymous with Topic, and Rheme is not
synonymous with Comment. As Halliday points out (as well as
others; see Chafe 1976), Topic/Comment, and even more particularly
Topic, has taken on many guises and is used in too many ways to be
understood in any universal sense. But several common aspects of
Topic emerge: 1) Topic is a sentence-level constituent, not a clause-
level constituent. It often does not participate in case frame
relations. 2) Topics, like Themes, but unlike Subjects, are highly
discourse-oriented. 3) Topics are not selectionally related to the
verb (Li and Thompson 1976: 466). Chafe says that Topic is "the
frame within which the sentence holds“ (1976: 51). Topic seems to

set up a frame of reference for the predication.

108

With this in mind, we can set up a working definition of Theme
and Rheme that is cognitive and substantive. Theme is the clause-
level constituent that the encoder uses as the starting point for the
message, the constituent that begins moving the decoder toward the
core of the communication. Theme is the essential ideational
jumping off point directing the decoder's attention to the ultimate
goal of the communication, the kernal of the message, the Rheme.
Rheme, then, is that which develops the Theme. Following Firbas'
notion of communicative dynamism, that element which does the
most for pushing the message forward will be the Rheme proper.
Rheme is the essential ideational purport that the speaker wishes to
get across.

We will treat Theme and Rheme as universal and as obligatory,
at least as underlying notions; that is, all languages have
Theme/Rheme structure. The realization of Theme and Rheme will
be overt discrete clausal elements. Theme may be mapped on to the
same constituent as Subject, Topic, or Given but is not synonymous
with any of these. In fact, all four notions may be realized by the

same clausal element.

109

Endnotes

1. The term 'gnostology' is being used here following Michael
Gregory. Gregory uses gnostology to refer to the knowledge base of
the encoder/decoder. For his latest exposition, see Gregory, 1988.

2. For a fuller discussion of the interrelationship of Theme and
Given see Fries 1983.

CHAPTER VIII

THEME AND RHEME IN DARI

1. Identification of Theme and Rheme

Though there are clauses where functional and even
grammatical roles are not entirely clearcut, most are readily
apparent and not open to divergent analyses. Such is not the case
with message structure, or thematic roles. There are few formal
criteria for the assignment of Theme and Rheme, only semantic
criteria. Even Halliday's very explicit clause-initial rule for Theme
in English is a realization rule, not a definitional one. Firbas' rather
mathematical assigning of numbers for degree of CD is both
intricate and subjective. If we wish to delineate message structure
in any meaningful way for Dari, we must come up with some criteria
for determining which element carries the most communicative
value and is, therefore, Rheme proper, and which element carries the
least (but not zero) 0D, acting as a 'jumping off' point for the core of
the message and is, therefore, Theme.

Based on an earlier analysis (Rashidi in press), some
parameters of the message structure in Dari have emerged. First,
linear realization of a proposition plays an important role in the
structure of this language. Where other languages, such as English,
use intonational prominence for focus and emphasis, Dari is more

likely to use element order. Linear position as a universal should

110

111

correlate with Theme/Rheme structure. In the normal course of
producing a sentence, one element follows another. From a cognitive
standpoint, this is significant, especially for spoken text. It would
seem natural that the orienting part of the communication would
appear early in the linear sequence and before the core idea. While
intonation markedness may have the effect of counter-balancing this
tendency, in general Theme should appear before Rheme. This is
particularly true for Dari where word order is relatively flexible
from a grammatical standpoint and intonation is highly predictable
(Bing 1980) and plays a minor role in marking prominence.

Second, message structure correlates with information
structure (or Given/New in Halliday's terms). For language in
general, we are most likely to orient our listeners to our message by
a piece of content that is already in the consciousness of the
listener, that is, has been activated either linguistically or
extralinguistically. As such, Theme is most likely to be Given, or at
least not entirely news. Because an element labelled New may
contain some given informationl, many Themes may be labelled New
but not be completely new information. Rheme, by contrast, is most
likely to be not only New but entirely new information. While
previous research on Dari (Rashidi in press) did not show as high a
correlation between Theme and Given as expected, Rheme was
always New.

A third possible correlate lies in the grammatical structure of
the clause. It would seem reasonable, given Chafe's definition of
Subject as ”the hitching post", that the Subject of a clause would
have a tendency to be Theme. Halliday states that in English the

112

mapping of Theme on to Subject is the norm and, therefore, the
unmarked situation. Indeed, Halliday defines 'marked Theme' as
”something other than the Subject“ (1985:451). As shown in the
present analysis, however, Dari does not have obligatory overt
subjects, overt subject pronouns only occur in marked situations,
and the subject inflection on the verb appears clause-finally. These
three factors make the linking of Subject with Theme less
attractive in Dari than in English.

In addition, the grammatical realization of a constituent in
Dari seems more loosely tied to its status as an argument in the
proposition than does a similar constituent in English. Objects and
Complements, for example, may be realized as prepositional phrases
in Dari. There is no formal distinction between direct and indirect
objects. Beneficiaries in material processes are always realized as
prepositional phrases. A Beneficiary in a verbal process may be
realized as an object, but then there is no other object. For example,
one cannot say in Dari the equivalent of 'he told John a lie'. One can
only say 'he told John that . . .' or 'he told a lie to John'. Because the
function of a constituent is not strictly tied to its grammatical
realization, the grammatical status of a constituent is not a factor
in determining the message structure of Dari. If grammatical
function does correlate with Theme or Rheme, it is probably
indirectly as the result of the correlation of grammatical function
with linear order.

In identifying the Theme and Rheme of each clause, then, only
two correlates seem to have bearing: linear position and

information structure. In the final analysis, however, the

113

recognition of Rheme and Theme is a semantic one based on their
respective values as the core of the message and the orienter of that
message core. From a methodological standpoint, it is important to
first establish Rheme, the essential ideational purport, and then
identify Theme, the jumping off point directing the decoder's
attention to the ultimate goal of the proposition, the kernal of the

message, the Rheme.

2. Theme and Rheme in Text B

Theme, then, following Halliday, is the element which
"organizes the clause as a message“ (1985:39), 'it is that with
which the clause is concerned“ (1985:39). Theme is what the clause
is about. Rheme is the essential thing the encoder wishes to impart
about the Theme. But it is best to reverse the emphasis on these
two parts of message structure and say that Rheme is the
development, the core of the message, while Theme is a statement
of what is to be developed, that about which something is said.

To investigate the message structure of Dari, we have taken
the main section of the text, labelled Text B, and divided it into
major independent clauses. Because the main contribution to the
development of a text comes from the message structure of
independent clauses, we will take into account essentially these; we
will only peripherally consider: 1) embedded clauses which because
of their 'down-ranking‘ are themselves constituents of a clause or,
more likely, another clausal constituent such as a noun phrase (e.g.
B29b); 2) minor clauses which have no mood or transitivity
structure (e.g. saudi in B25; e6 kas in B81); 3) subordinate (beta)

114

clauses (e.g. B27b), the options for the formation of which are
constrained by the main (alpha) clause.2 Appendix B shows the
message structure of Text B. The Theme of each clause is underlined
and the Rheme proper is in italics. In general, any constituent not
marked as Theme is part of the Rheme. The exceptions to this are
budan and sudan which, because they carry no communicative value
(following Firbas, see discussion of CD, Chapter VII, p. 100), are

outside the Theme/Rheme structure.

3. Parameters of message structure in Dari

Thematic structure is only meaningful if there are choices to
make. An element that is highly mobile will have thematic value
because the choice of where to place it in the structuring of the
proposition will be significant in terms of how that message is
communicated. On the other hand, an element the linear position of
which Is fixed will have little thematic value. Choices besides
linear position also affect message structure. Can a particular
argument in a proposition be realized, either grammatically or
functionally, in a number of ways? What choices does the encoder
have open to express a particular concept?

Dari, like any other language, places restrictions on the way
in which an idea can be expressed, and these restrictions both
determine and constrain the Theme/Rheme structure of the encoding
of that message. The following aspects of Dari point out the
problem of a universal notion of message structure because they
highlight differences in the possibilities open to Dari encoders as

compared to those open to English encoders.

115

3.1 Clauses without a Theme

In the analysis of Dari text, one problem that consistently
arises is the possibility of a clause without a Theme. This dilemma
is the result of a preponderence of one-participant clauses in Dari.
Because English requires an overt subject, clauses with a single
element of communicative value are probably rare. In Dari, they are

common.

3.1.1 Existential

One example of a one-participant clause in Dari is an
existential clause, which has by definition a single participant
followed by a form of budan or s’udan, elements which add little or
nothing to the communicative dynamism of the message and,
therefore, following Firbas, carry no 0D. In these clauses, then,
there is no choice in structuring the proposition. In any case, there
is only one element carrying CD and that element is designated
Rheme. The clause, in effect, has no Theme.

Fries and Francis (1990) suggest the semantic notion of
existential as Theme for English 'dummy' constructions (e.g. 'it is
raining'). But this makes little sense for Dari as there is no 'dummy'
element to carry the notion. A possible solution would be to attach
the notion of existential to the predicate and call that Theme, but
that seems to be stretching the concept of message structure beyond
any usefulness. It would appear, instead, that there are clauses that
do not have a Theme/Rheme structure. An existential clause is one

such clause.

116

3.1.2 Ellipsis

We find another type of structure in Dari that is problematic
for Halliday's conceptualization of thematic structure. Is elliptical
Theme possible? Because message structure is a surface concept
involving the presentation of a proposition, the notion of ellipitical
Theme seems questionable. Yet we find in Dari, clauses that consist
only of a Predicate, usually a complex verb. It is important to
remember that in Dari, subject pronouns are encoded only in marked
situations; their overt appearance is not the norm. Thus, while a
zero Subject clause has an underlying subject (and actor or senser or
sayer, for example), that subject is encoded only as a verbal suffix
which appears in clause final position, hardly a location for
thematic material. In addition, third person singular, the most
common inflection, is realized as a null element (i.e. not physically
realized at all). As a result, it is often not clear if a clause is a
compound clause with a single overt element that could be labelled
Theme of both or two separate clauses, one without a Theme/Rheme

structure. Several such examples appear in Text B. One is BS4:

834a WI
all-E Muslim-PL that ME-came-3p

834b sang-a / tamas me-kard-an II
rock-RA touch ME-did-3p
B34c zlarat me-kard-an II
respect ME-did-3p
'all the Muslims that came touched the rock; they paid respect'

B34a is a noun phrase (with an embedded ke clause) that is the
Subject and Actor of B34b. That much is straightforward. The

117

semantic content encoded in B34a is also the subject and actor of
834c. But B34b and 3340 are not connected with a conjunction, an
atypical occurrence for this text. B34c can stand on its own as an
independent clause. Our translation reflects this. So while we
might posit B34a as the Theme of a compound Rheme, it is important
to remember that the encoder chose not to encode an overt subject
for 834C or connect it explicitly to the preceding clause.
Nonetheless, while the notion of an elliptical Theme does not seem
appealing, the notion of compound Rhemes seems to be plausible for
examples like B34. Less obvious, but possible, examples of a
compound Rheme occur in 859-61 and again in 864-66.

B59 153/ ml s'uma-ra I hast kard-im II
that l you-RA create did-1p
B60 asman o zamin-a l hast kard-im II
sky and earth-RA create did-1p
861 afta o sifa-aa-ra I hast kard-im I/
sun and star-PL-RA create did-1p
'I created you; I created the sky and earth; I created the sun
and stars'

864 u I MI mend-in II
and that-place leave it-2p
865 mer-in II

90-29
866 salda me-kon-In II
bow down ME-do-2p
'you leave it there; you go; you bow down'

865 and 866 are probably connected to unja (864) as Theme.
In like manner, the thematic status of the overt ma in 859 probably
carries over to 860 and 861. The nature of Dari makes it highly

118

unlikely that, within the particular contextual environment of these
clauses, the overt subject pronoun ma would be repeated. The
presentation as it exists seems unmarked, making the concept of a

compound Rheme very appealing here.

3.1.3 Clauses without an ideational Theme

Halliday distinguishes textual, interpersonal, and ideational
theme and says that every clause has an ideational Theme because
textual and interpersonal Themes do not use up the thematic
potential of a clause. Though, in this data, clauses without a Theme
of any kind are rarer than expected, some clauses have only a textual
Theme. This is because the narrator connects most clauses using
textual elements, generally we, 'and', or ke. We will follow Halliday
here, assuming that a textual Theme, because its clause-initial
position is not a matter of choice, does not use up all of the
thematic potential, thus leaving open the possibility of an additional
ideational Theme. Only a few clauses in the data had no potential
candidates for an additional ideational Theme. One instance, 818, is
discussed later on. The other clauses with only textual Themes are:
821, 823, 835, 839, B46b, 883.

821 9_I besior mamlakat-haa-ra I graft-an II
and very country-PL-RA took-3p
'and they took many countries'

823 jug/pas amad I/
and again came
'and he returned'

119

835 tun/ha dist-e pkpmbar-e hazrat-e ebrahim Ixord
since by hand-E prophet-E hazrat-E Ebrahim hit
'since it hit the hand of the prophet Ebrahim'

839 “I but-aa-e I daét-an I ke . . . I I
and idol-PL-E had-3p that
'and they had idols . . .'

B46b Lg/ éiéma-aa-e zamarud / bud II
that eye-PL-E jade was
'with eyes of jade'

883 ml elan kard II
and announce did
'and announced'

B21 and 823 are material process clauses. 823 consists only
of a Predicate, offering no choice in constituent presentation. 821,
though it consists of an Object in addition to the Predicate, also
offers no real choice of constituent presentation. In any event, the
Rheme proper is besior mamlakat-haa-ra, 'many countries', the
Object, and the other constituent is the Predicate. Predicates in
general seem unlikely candidates for thematic status in Dari, and we
have not posited any. A similar argument can be made for B35 and
B39. B46b is an existential clause with a single participant. B83 is

a verbal process consisting of only a Predicate.

3.1.4 Truly Themeless clauses

There remain clauses in the data without any Theme whatever:
831a, B70, 825, 837, 838, 849, 852, B74, B77. 831a is a verbal
process clause consisting of only the Predicate, goft, 'said'. in turn

followed by the verbalization which has its own thematic structure.

120

In 870, tana, 'only', seems the core of the message, and the entire
Subject is connected to the Predicate by ke, making the clause as a
whole seem like a single constituent, though the ke is probably an
emphatic. The other clauses all consist of a single constituent
followed by a form of budan or sudan. They are either existential
clauses or relational clauses with non-overt subjects. These latter
clauses are truly clauses without a Theme/Rheme structure; they

consist only of the core of the message, the Rheme.

831 a gaff I I
said

870 tana kason-e Ire I qabul kard /I
only people-E that accept did
'only some people accepted'

3.1.5 Topic and Theme

An interesting aspect of Dari is its tendency for topicalization
(Rashidi in press). Encoders often begin an idea by first stating the
topic of that idea, its central participant, even if that participant
does not formally then participate in the proposition that follows. A
topicalized element is most often the Agent. In another Dari text
about a motorcycle accident (see Rashidi in press), the following

sentence appeared:

ebrahim poi-xub-lé-a élklstan
Ebrahim leg-good-his-RA broken
'Ebrahim broke his (someone else's) Ieg'

Ebrahim is the Agent, the cause of the broken leg, but it is not
Ebrahim's leg that is broken. The closest English translation is:

121

'Ebrahim, his (someone else's) good leg was broken'. In Text 8, there
are similar topicalized elements, though none is an Agent (851,
853a, 863). In analyzing these clauses, we have considered a
topicalized element as part of the Theme but not the entire Theme;
there are still choices left open to the encoder to organize the
following proposition independent of the topicalized element which
does not directly participate in the proposition. Thus, we have
chosen to call this element Topic as distinct from the Theme of the
clause that follows. In all three examples in Text 8, the clause
following the Topic contains at least a textual Theme in addition to
the Topic of the sentence.
Topic
851 Ml LIL-ii da tasruf-e az i I bud I/

Hona Kaba key-its in possession-E from him was
'he had the key to the Hona Kaba'

Topic
853 Maxim/mike. .--/raft II
this Mohammed grandson-his time-E that saw

'this Mohammed, his grandson, when he saw that . . . he left'
Topic
863W

this idol-E that self-E-you in hand-E-you construct ME-do-2p

864 wa I un-ja I mand-in II
and that-place leave it-2p
'these idols that you yourselves construct with your own hands; you
leave them there'

122

851 is quite typical of topicalized elements in Dari. The
narrator repeats the xone kaba, 'House of the Rock', from the
previous sentence, establishing it as the referent for -§, 'it', and the
topic of the clause that follows. In 853, Mohammed is semantically
the subject and actor of raft, 'left', and, therefore, the Theme, but
is physically separated from the Predicate by so much material (see
Appendix B for full text) that its effect is topical. Given the
structure of Dari, it is logical to posit this sentence-initial element
as Topic. 853 as a whole, however, has been deconstructed in a later
section. As viewed there, the English equivalent would be: 'this
Mohammed, his grandson, when he saw that they, abandoning the way
of God, were worshipping idols, left'. Even in English, however, a
speaker would probably insert the pronoun 'he' before 'left', making
Mohammed a topicalized sentence element. In Dari, because subject
pronouns are not obligatorily overt, it is ambiguous whether I
mohamad or o, the null third person singular marker, is the formal
subject of raft, 'left'. 863 is probably the Topic for the four
clauses that follow it, 864-67, but is at least the Tepic of 864.
mandln is an intransitive verb, the meaning of which has the notion
of object built into it, 'leave it'. 864 is 'about‘ the idols mentioned
in 863, even though but, 'idol', never directly participates in the
clause.

3.2 Linear position

Unlike Halliday's analysis for English, in Dari linear position is
not, a priori a realization feature of thematic structure. But we
would also vary from Firbas' looser association of Theme and linear

position. Linear postion is not irrelevant; the order in which an

123

encoder choses to structure the elements of a proposition is highly
relevant to the message structure and the identification of Theme

and Rheme in Dari. Two major factors contribute to the importance
of linear sequence in this language.

First, the verb in Dari has two counter-tendencies. Most
Predicates fall into one of two categories: a relatively semantically
empty verb (e.g. budan, sudan) or a complex verb with heavy
semantic loading. These latter Predicates often contain the entire
semantic content of the proposition. In addition, the verb is
obligatorily marked for subject person and number, often the only
realization of subject in the clause. This results in clauses with the
entire semantic content mapped on to a single constituent, as we
have seen. A second characteristic of Dari verbs is that, with few
exceptions, they are clause-final. Thus, the Rheme proper, by the
very nature of the language, will appear at the end of the clause.

Second, except for the Predicate, element order in Dari is
grammatically flexible, and, therefore, carries thematic
significance. The order of the Subject and the Object is particularly
variable. Other elements, Circumstances and Modal Particles, are
rather free to occur in different positions. A fronted Object or an
overt Subject is definitely thematic, as are clause-initial
Circumstances and Medal Particles. By contrast, clause-final
circumstantial elements are not prominent as they
often are in English, but have more of a feeling of being 'tacked on'
and are, therefore, not rhematic.

The significance of linear position in Dari seems to be as

follows. If an element is in clause-initial position by choice, it is

124

probably Theme. Clause-final position does not carry the same
significance. Predicates are nearly always clause-final because of
the grammatical structure of the language. In Dari, clauses do not
build up to the end as in English (see Fries unpublished) but build up
to the Predicate. The most prominent linear position in Dari in
terms of message focus (as well as information focus) is
Immediately preceding the Predicate. This is where the Rheme

proper is most likely to occur.

3.3 Given/New

Like linear position, Given/New also correlates highly with
Theme/Rheme in Dari: Theme is usually Given and Rheme is usually
New. But, message structure and information structure do not
always coincide. There are clauses where there is only given
information. At first glance, it may appear pointless to encode a
proposition with no new information, but in reality we do this all
the time. We may wish to emphasize or clarify, as the narrator of
this text does in 819 (see Appendix B); or the various pieces of
information may already be in the mind of the decoder, but their
combination is what is news (e.g. the l qabul kard clauses B72,
B75, 879).

There are also clauses with only elements labelled New. There
may be two reasons for this. The most obvious is that, in fact, all
the information is news. This happens most often at the beginning
of a text or a new section of a text. Text 8 begins with a transition
sentence that connects Text 8 to Text A, what went before, but the

second sentence (816) in Text 8 announces a new idea and is all new

125

information. A less obvious reason for a clause with only New is
that elements labelled New may not, in fact, be entirely news. In
other words, some New elements may be more new than others. An
element that is entirely news will be more likely to be Rheme than
an element that is only partially news. For example, in 871, the
Theme contains the given information mohamad, but yak zan, 'one
wife', is new information. Her name, however, is entirely news and

is the Rheme.

B16 w_a_I Wu! pIcpmbar-e Islam mohamad /
and in Arabia prophet-E Islam Mohammed

wafat ke yafta bud I/
death that found was
'and in Arabia the prophet Mohammed had died'

B71 Wamn-Xadila Inam daét
one wife-E hazrat-E Mohammed .. . Khadija name had

'one wife of Mohammed was named Khadija'

If a clause has both a Given and a New element, the element
that is Given will be the Theme and the element that is New will be
the Rheme. This overrides even linear position. In 869, such a
conflict between linear position and information structure occurred.
ira, 'this‘, is quite definitely given information, while pader-aa-e
. ., 'ancestors', is at least partially news. The encoder could have
fronted ira. The clause as it stands is typical word order for Dari
but seems atypical in this context for this narrative. The narrator
has often fronted object pronouns in other similar contextual
environments. In this case, he chose not to do so. ira is still the

Theme of the clause and pader-aa-e . . . is the Rheme proper.

126

869 pader-aa-e [ad a kaka-e mahamad / L-La I awal I
father-PL-E old and uncle-E Mohammed this-RA first
qabul na kard-an II
accept NEG did-3p
'the ancestors and uncles of Mohammed at first did not accept this'
4. Embedded clauses

Embedded clauses have Theme/Rheme structure of their own,
and this embedded message structure is intricately interwoven with
the message structure of the clause of which it is a constituent. It
is instructive to look at the detailed analysis of two such examples:
B47 and 853.

In 847, the Predicate of the main clause is bud which is out-
side of the Theme/Rheme structure. I, the Subject and Carrier, is
the Theme of the entire sentence, the Rheme of which is the Attri-
bute as Complement, realized as the noun phrase 'such an expensive
idol that the word was we had to worship it' (all of 847 except I and
bud). The ke clause, a down-ranked modifier of besior buta
qimati, has a textual Theme, ke, as well as an ideational Theme,
ira, the initial Verbiage. The verb, megoftan, is the Rheme proper,
and the Verbiage that follows the verb is more rhematic material.
This final part of the Verbiage is a clause in and of itself. ira, the
Goal and Object of this clause, is the Theme. bayad, a Modal Part-
icle, is an interpersonal Theme due mostly to its linear position
before the Rheme. We have posited maa as Rheme for three reasons:
1) it is an overtly expressed pronoun, giving it prominence; 2) it is
New, and 3) it has focussed linear position immediately preceding

the Predicate. The Predicate itself, parastes konim, is additional

127

rhematic material.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Th Rh Rh
tx Th Rh rhem
TI1 lnTh Rh them
847 i | besior buta qimati | bud | Re I ira | megoftan | ira | bayad | maa | parastes k.
this very idol expensive was that this said this must we worship do
A A 6:01 MP A:S P-Ma
Vb P-V Vb
Ca:S Ar:C P-AtI Ar:C
'this was such an expensive idol that the word was we had to
worship it'
Figure 6
Detailed Analysis of B47
Th Rh
text Rh rhem '
tx Th Rh rhem
tx Th Rh rhem

 

 

 

 

 

 

 

 

 

853 m | waxte ke I did I ke | inaa | az roi xoda | gastan | butaara | parastes mek. | raft
M. when saw that they from way god abandon idols worship do left

 

A:S R P-Ma
P- A:S G:O P:Ma
Me Ph 1 P-
Ma

 

 

 

 

 

 

 

 

 

 

 

A:S
'Mohammed, when he saw that they, abandoning the way of God, were
worshipping idols, left'

Figure 7
Detailed Analysis of 853

128

853 can be viewed formally as a single clause (for a different
interpretation see Chapter VIII, p. 121) with progressively embedded
internal clauses. In this interpretation, the first two elements
(represented above by mohamad but including the appositive
nawasls, 'his grandson') are the Theme of the entire clause and the
final element is the Rheme proper. The mental process clause
embedded between these two elements has a textual Theme, waxte
ke, but no ideational Theme; the verb, did, is the Rheme proper. The
Phenomenon of this clause is an event, actually two events, one
embedded in the other with the whole introduced by ke, a textual
Theme. inaa is the Subject, Actor and Theme of both the main
clause, inaa_butaara parastes mekonan, and the further
embedded infinitive clause, az roi xoda gaétan. In the outer of
these two material process clauses, inaa is the Agent and butaara,
the Rheme, is the Medium; of theinfinitive clause, inaa is the
Medium and the other participant, az roi xoda, is the Range.

5. Analysis of the text

Not all decisions in the Theme/Rheme analysis are clear-cut
and without dispute. To illustrate some of the reasoning behind the
choices made, a discussion of the initial section of Text 8 (probably
the first semantic paragraph) follows.

Text 8 begins in a usual fashion for a narrative. 815 has two
thematic elements, one a Circumstance of Time, the other a
Circumstance of Location, that orient us to the core of the message,
'Rakbil Shah was governing', Rakbil Shah being the Rheme proper.
dar in waxt, 'in that time', refers back to the events in Text A and

129

is, therefore, Given; dar kabul, 'in Kabul', is also Given from Text A.
Rakbil Shah is New as is the idea of governing. 816 follows a
similar pattern without the initial Circumstance of Time. The
narrator is orienting us to an event in time, but he puts the locations
in parallel sentence-initial position. 815 and 816 are probably, in
fact, a single compound sentence (or clause complex in Hallidayan
terms), dar in waxt being a sentence-level constituent, not a
clause-level one. 'At this time', two events were occurring: 1)
Rakbil Shah was governing and 2) Mohammed had died. The first
occurred in Kabul, the location of Text A, what has gone before, and
the second occurred in Arabia, the location of the text to follow.
This is a transition sentence.

B17 begins with the textual element 0, 'and'. This 'and'
connects the listener back to 816, what was happening in Arabia.
But the narrator also orients us timewise: bad az u, 'after that',
'that' being the death of Mohammed. The Rheme of B17 is the
existence of the four khalifs of Rashid.

818 does not have an ideational Theme, only the textual
connector ke. This presents some problems for the notion of
message structure as proposed by Halliday. Grammatically, Re is a
subordinate clause connector, but 818 is not a subordinate clause.
We find throughout this narrative the use of ke as a general
connector of ideas. The enumeration of the names of the four khalifs
in 818 is, of course, closely linked semantically to 817. 818 is an
identifying process, 'they were Abu Bakar, Omar, Usman, and Ali'. In
English, the clause-initial 'they' would act as Theme, but Dari does

not require an overt subject pronoun here; indeed, the presence of

130

one would be marked. Thus, we have a clause with only a textual
Theme, no ideational Theme.

819 illustrates another aspect of Dari that affects message
structure: the penchant for complex verbs. While in English, the
core of the message seldom lies in the verb, in Dari the semantic
load of the proposition is often shifted to the verb. 819 does have a
Theme with the highly focussed initial element inaa ham, 'they
also', but in another context the clause might have consisted of only
the Predicate. Thus, we have in 818 and 819 two clauses that
potentially do not have Themes.

There is little new to say about 820. The clause-initial o,
'and', is the textual Theme; dar zaman-e-éan, 'in their time', a
circumstantial element, is the ideational Theme. The semantic
intent of the remainder of the clause is rather ambiguous. ziat sud
is a common idiom meaning 'increased'. If interpreted as such here,
the clause would mean: 'and in their time conquering increased'. But
this seems strange as futuhat, 'conquering', is new information. In
addition, futuhat is connected to ziat by the ezafe. Thus, we have
interpreted futuhat-e ziat as one constituent meaning 'a lot of
conquering', and sud alone as the Predicate meaning 'happened'. The
closest English translation is: 'a lot of conquering occurred'. This
seems logical within the context. futuhat-e ziat, then, is the
Rheme.

In 822, the narrator again uses the time frame bad az u, 'after
that', as Theme. The Rheme is not so apparent. The rest of the
clause following bad az u is rhematic, but the Rheme proper seems
to be taa qismat-aa-e . . ., 'up to sections of' . This is the only

131

constituent that is New, and it is in linear position immediately
preceding the Predicate. Though its gramnmatical realization is an
Adjunct, it functions directly in the process as Range.

823 has only two constituents. we, 'and' is a textual Theme
and pas amad, 'returned', is Rheme. It is possible that 823 is a
compound with 822, in which case, bad az u, 'after that', would also
be the ideational Theme of 823; it is not possible to tell here. In
such cases of ambiguity, we have selected the most overtly obvious
analysis.

6. Theme/Rheme chart

Appendix B shows the designation of Theme and Rheme in
clausal context. There we can see the flow of thematic structure
from one clause to the next and in relationship to the choices not
made. The Theme/Rheme Chart, Table 7, shows in condensed form
the interrelationship of Theme/Rheme structure to: information
(Given/New) structure, grammatical structure, transitive functional
structure, ergative structure, and linear position. In this chart we
can easily compare the parameters of Theme versus Rheme in this

text.

There are 77 clauses tallied in this chart. All 77 had a Rheme.
We are positing Rheme as obligatory; without at least a Rheme, a
proposition would not exist. All Rhemes are clausal constituents,
occurring in the following order of frequency: Complement (22),
Predicate (21), Subject (15), Object (9), Adjunct (9), and Subject
Appositive (1). By contrast, only 55 of the 77 clauses have

132

ideational (or topical) Themes. There are four possible reasons for
this: 1) there is only a textual Theme; 2) the Theme consists only of
Topic and, therefore, is not a clausal constituent; 3) a constituent is
the Theme of more than one clause (i.e. has a compound Rheme); and
4) a clause has no Theme/Rheme structure he has only a Rheme). If
a Theme is a clausal constituent, it is most likely to be the Subject
(31); if it is not a Subject, it is either an Object or an Adjunct (11
each)

Theme and Rheme vary in functional structure as well as
grammatical structure, but not equally for all process types. In a
material process, both Actor and Goal are equally likely to be Theme
(A=7, G=8) or Rheme (A=6, G=8). In a mental process, an overt
Senser may be Theme (4) or Rheme (3), but the Phenomenon may be
only Theme (2), never Rheme. An overt Sayer may be Theme (1) as
may the Verbiage (1). In verbal and mental processes, the Rheme is
most often the Predicate (the process itself) which often includes
the Sayer or Senser as a verbal suffix but not a separate constituent.
In relational clauses, the correlations are invariable: Attributes and
Identifiers may be Rheme but not Theme, while Carriers and
ldentifieds may be Theme but not Rheme, perhaps by definition.
Existents are, of course, always Rhemes. Circumstantial elements
are far more likely to be Theme (11) than Rheme (3).

From an ergative perspective, Rheme is most often Range (32),
while Theme is most often Medium (36). In fact, the Theme is
seldom anything but Medium: three Themes are Range and four are
Agents. Nineteen Rhemes are constituents that act as Mediums but

other Mediums are indirectly Rheme as part of the Predicate.

133

As noted earlier, there is a high correlation between message
structure and information structure. A Theme is only New if the
entire clause is New, and Rheme is only Given if the entire clause is
Given. The correlation of Theme and Rheme with linear position is
also virtually invariable. Only one Theme (B69) is not the clause-
initial element. With few exceptions (853, 863), the Rheme is the
clausal constituent that immediately precedes the Predicate or is

the Predicate itself.

Theme PosGr
15. dar in waxt/

darkabul 1 I’d
16. we (text)

dar arabstan1 Ad
17. 0 (text)

badazu 1 Ad
18. ke (text)
19. inaaham 1 S
20. 0 (text)
darzamanesan 1 Ad
21. 0 (text)
22. badazu 1 Ad
23. wo (text)
24. markaze xa 1 S
25. _____
263. bad/dar awa1 Ad
26b.éun (text)

dar Inja 1 I’d
26c.wo (text)

make 1 S
27. 0 (text)

mazjid ke. 1 S
28. ke (text)

usang 1 S
29. 0 (text)

Isanga 1 O
30. ami sang 1 0
31a. _____
31b.ira 1 O
32. we (text)

I 1 S
33. Ira 1 O
34. hame musll 1 S
34c. "
35. (Sun (text)
36. badazhaz 1 Ad
37. _____
38. ______
39. we (text)

134

Table 7

Theme/Rheme Chart

Tr Er Inf

Cr Gv
Cr
0r Gv
A Md Gv
Cr

Cr

IdeNw

Cr
Cr
Id Md 0/

IdeNw

ldeGI

93
E1:
9?

§>QQ a
£55 5
ggee @

9
9

Nw

Inf Er Tr

NdeA
NdeA

NdeEt
NwR lr
GI

NdeEt
NdeG
NwR R
Nw

NwR Ir
(31R Ar
NwR Ir

NdeEt
NwR Ir
NwR Ir
NwR Ir

Nw Md$
Nw
Nw
Nw Pg A

NwR Ar
Nw 0r
Nw

NwR G
NdeA
NwR Ar
NwR Ar
NwR Ar

GrPos

ooomappao mppm o o o m ooovaom pom m m

Rheme

1P rakbiléa
1P piqtmbare . . .

1P xalifaae . . .
1P hazrate . . .
xelaafat kardan

1P futuhate ziat
1P besior mamlakathaara
1 P taa qismataae . . .
pas amad
1P sare madina
1P dar arabstan
1P mdta

1P mazjidl haram
1P jol_ke . . .

1P yak jol_ke . . .
1 P az sanghaae asman_ke . .

1P hazrate ebrahim
graft o wardoét
goft

3P xoda

1Pmoqada

1P dar deware xone kaba
tamas mekardan
zlarat mekardan

1P ba diste pinmbare . . .

1P akwame . . .

1P kafire

1P but parast

1P butaae_ke . . .

41. arqabila 1 S (thGI
42. ar qabila Gv
43. we (text)
44. sesodosas1 O G MdGI
45. ke (text)
butaae kalone azi1 S Id Md Nw
46a. wa (text)
yakbuteazinaa 1 S CthNw
46b.ke (text)
47a. wa (text)

I 1 S (h Md Gr
47b.ke (text)

ira 1 O \IbR GI
47c.lra 1 O G Md GI
48. waxte bud ke (text)

ibutaara 1
49. _____
50. abdulmotal 1 S Id Md GI
51 . xone kaba (Topic)

klis 1 S (h Md Nw
52. _____
53a.i mothawasié (Top)

53b.

53c.
53d.

54.
55.
56.
57.
58.
59.
60.

61.
62.

waxte ke (text)

ke (text) (dep cl)

lnaa 1 S A Md GI
" Po
[53a-c]

0 (text)

i 1 S (h Md GI
ke (text)

ma 1 S Id Md GI
wo (text)

xoda 1 S Sy Md GI
ke (text)

ibutaa 1 S Ga Md GI
bayad (interp)

suma 1 S A Ag GI
ke (text)

ma 1 S A Ag GI

bayad (interp)

hamagi 1 S A Ag GI

135

Tfih7

NwR Ar
Gv
Nw
GIR Cr
NwR Ir

NwR Ar
NdeEt

NwR Ar
Nw

NwAgA
NwR Ir
NwR Ir

NwR Ar
NwR Ir

Nw

Nw

NwR R
Nw

NwR Ar
GI MdG
GI MdG
NdeG
NdeG

GIMdG

0 IF yak but

yak but

se sod o éast qabila
Ad 1P dar ami rafaae x. k.

Ad 1P banamelaut...

Ad1Pazyaqud

S 1P cisma-aae zamarud
1P besior buta qimati

megoftan
1 P m

(0'00

tawasute mohamad
1P pisare abdula wa . ..
1P mujawir wa . ..

1P da tasarufe az l
1P ixlardare xone kaba

did
1 P az role xoda

1P butaara
raft

p03 p on 00

dawi kard
1P pimmbare xoda
+ amlr karda
1P qabile parastes
1P mara
1 P 'sumara

1 P asman o zamina
1P afta o sita-aara

0000001101)

1 P mara

63.
64.

65.
66.
67.
68.
69. '
70.
71.
72.
73.
74.
75. '
76.
77.
78.

79.
80.

81.
82.
83.
84.

85.

i bute 1
wa (text)
unja 1
ura 1
i 1
Ira 2
yak zane 1
i 1
bad az hazra 1
I 1
bad az u 1
ke (text)

i 1
i 1
0 (text)

inaa 1
anuzl (text)
qabila 1
kule butaara1
wa (text)

ke (text)

ma 1
ma 1

o
c.)
5
Q

R
9
9

39"
EE ”ES

30) 80303 0030
996
99 99; 999

98’
E

00) (I) (00)
mg: > 95
EE E 55
99 9 99

(DU)
95
E5
99

136

Table7
GIPgA
Nw
Nw
Nw
Nw
NwR Ar
Nde&
Nde&
NwR Ir
GI
NdeEt
NwR Ir
GI
NdeEt
NwR Ir
NwR Ir
GI
GI Cr
Nw
Nw
Nw
GIR Ir
GIR Ar

0 'O'O'D a 'DO OWUOCD'DQCDUDO'O'OTJ'U (D

Pd

2 P xudetan

mandin
merin
sajda mekonin
ebadat mekonin
1P qarat
1 paderaae jad o . ..
1 P tana kasone ke
1P hazrate xadija
qabul kard
1P abu bakar
1P xalife awal
qabul kard
1 P ast sola baéa
1P hazrate ali . . .

1P bacae kakae picambar
qabul kard

1P amroi mohamad
xabar na dastan
sikastand o . . .
elan kard

1P piqtmbare xoda
1P az tarafe xoda

137

Endnotes

1. Firbas feels that if part of a constituent is irretrievable, the
whole constituent is considered irretrievable (i.e. New) [from
discussion at the Second Nottingham International Systemic
Workshop (on Rheme)]. I would agree with Firbas on this.

2. A subordinate (beta) clause stands in a hypotactic relationship to
the main (alpha) clause; this is a tactic relation between clauses.
An embedded clause functions as a constituent within a group which
is itself a constituent of a clause (e.g. a modifier within a noun

phrase). For further discussion of this distinction, see Halliday
1985: 219.

CHAPTER IX

SUMMARY AND CONCLUSION

1. Summary of findings

In this analysis we have looked at some of the clausal patterns
that run through this text, tracing various features and showing how
these features are interrelated and dependent upon both other
features and the text as a whole. Some features are more signficant
than others and, consequently, affect other features of the text and
the meaning of the entirety in larger proportion. But all strands--
grammatical, transitive, ergative, thematic--interact and pattern
with each other to create the essence of the text. We have pulled
apart the various strands, as we must, in order to gain some insight
into the separate functioning of the various parts. In the end,
however, the individual systems only have meaning in relationship to
each other and to the whole.

By basing our analysis within a particular theory, we are able
to establish a consistent framework for both an internal look at the
nature of Dari and a comparative view of Dari in relationship to
other languages, forming some tentative assertions about the nature
of Language. We have begun our analysis with a number of long-
established basic assumptions, leaving open the possibility that
some of these assumptions may not be borne out.

138

139

Our most primitive assumption has been that the basic
message unit of language is the clause, and we have founded our
analysis on this assumption. Indeed, even in a casual, oral discourse
such as this historical narrative, the division of the text into
clauses has been a relatively simple one. Our next assumption has
been that clauses are composed of discrete constituents that
function in various ways in relationship to each other to form the
whole. We have used essentially Halliday's model to identify these
constituents, but the foundation is a traditional hierarchical one:
clauses are composed of phrases (or groups) which are composed of
words which are composed of morphemes.

These constituents are simultaneously arranged according to
various functional concepts, or modes of meaning, to create a single
structure or output, and they take on significance structurally in
their syntagmatic relationship to each other. But these constituent
choices are also a result of the paradigmatic systems from which
they are selected. Thus we have explored the description or output
of this Dari narrative, but in doing so we have had to consider both
the system and the text. Our Interpretation of Dari, or of a single
text, must take into consideration the choices available, the
possible forms, and why an encoder would chose this particular form
at this particular time and within this particular framework.

Halliday sees these choices as coming from three separate but
interlinking metafunctions or choice systems: the interpersonal
(clause as exchange), the ideational (clause as representation), and
the textual (clause as message). We have analyzed the Dari text

along all three of these modes of meaning, both paradigmatically and

140

syntagmatically. This analysis along various parameters and within
various systems allows for the richness of the language to be
exposed. We gain one insight from analyzing a clause into its
grammatical structure, another insight from viewing its transitivity
structure, and still another insight from exploring its thematic

structure.

2. Analysis across the three metafunctions

Figure 6 and Figure 7 have shown in detail the interrelated
systems across the three metafunctions for two clauses. Figure 8
shows the flow of structures within the three metafunctions for
clauses 815-823, the first nine clauses of the main text.

815 has two clausal constituents marked as Theme. Both are
Circumstantial elements, one of Time and one of Location. There is
nothing remarkable about their order and both seem to carry equal
CD. Linearly, the next element is rakbilsé, both Actor and Subject
of the clause as well as Medium. Its status as Rheme is firmly
established on two counts: 1) most importantly, rakbilsa seems to
be the core of the message, and 2) it is in linear postion directly
preceding the Predicate. In addition, it is the Medium or core
participant. hukumat mekard, 'governed', the Predicate, is more
rhematic material but does not carry as much CD as rakbilsé.

B16 begins with a textual Theme wa, 'and'. In addition, this
clause, like 815, has an ideational Theme that is a Circumstance of
Location. Again as in 815, the Rheme is the Actor, Subject, and
Medium of the clause, the core participant, while the Predicate is

more rhematic material. It seems clear that the narrator is

141

focusing the listeners' attention by orienting them to first the vital
people of that time frame and secondly their 'activity', as it were.

It is not death or governing that is the essential news but the
existence of Rakbil Shah and Mohammed.

B17 and B18 follow a similar pattern with the introduction of
prominent figures in history in Rhematic position. There is, in fact,
little other content material in either of these clauses beyond the
Rhemes. 817 has, besides the textual Theme, an ideational Theme of
Time. 818 Is most likely a dependent relative clause acting as a
postmodifier of the Rheme of 817. The English equivalent would be:
'and after that there were the khalifs of Rashid. four khalifs, who
were . . . '. Both the ke following raéuda and the ke that begins 818
are functioning in the same manner, as relativizers of post
modifiers. It is difficult to capture this use of ke in English as ke
functions more in the capacity of a connector than the English
relative pronoun 'who'. B18 is only formally a dependent clause; 817
and 818 form something closer to a compound sentence (two alpha
clauses) than a complex sentence (an alpha clause plus a beta
clause), the relationship being more one of parataxis than hypotaxis.
Nonetheless, bad as u, 'after that', might be viewed as the
ideational Theme of 818 as well as 817. 818's close linkage to 817
is reflected in the realization of the Medium as the verbal suffix -
an, 'they', the referent for which is the Subject of 817. While B17 is
an existential process establishing the existence of the four khalifs,
B18 is an identifying process, establishing who they were by name.

The core of the message of B19 lies in the complex verb,
xelaafat kardan, 'acting as khalifs'. This is a mysterious

 

142

sentence, the purpose of which is unclear. The only bit of
information that is 'new' is ham, 'also', but ham has no apparent
referent (who else was acting as khalif or what else were the four
khalifs doing?). The sentence merely repeats information already
given. The overt subject pronoun is marked but necessary for the
use of ham; otherwise the narrator would probably have said simply
xelaafat kardan. This is the first clause in this section (indeed,
the only clause) in which the Medium is the Theme.

B20 is more typical: it has first a textual Theme, then an
ideational Theme of Circumstance. Rheme is Subject, Existent and
Medium, the only direct participant in the clause. sud, 'became',
carries no communicative value. 821 follows a similar pattern but
has only a textual Theme. In addition to the Medium (Goal realized
as Object), 821 has an Agent realized as the verb ending -an, 'they'.

822 has a Medium that is neither Theme nor Rheme, a very
unusual occurrence. The rationale for the Theme/Rheme analysis has
been discussed in Chapter VIII and we will not repeat it here. The
Rheme is Range realized as an Adjunct. Range plays a prominent role
in Dari clause structure, and it is not surprising that Range would be
the core of the message.

In 823, again there is only a textual Theme, and the Rheme is
the Predicate. In this clause, we find the Medium, the central
participant in the process, realized only as a null element. This,
perhaps, presents a dilemma for the theory as it is a situation that
occurs frequently in this text.

143

Th:Cr'Ad Th:Cr:Ad Rh:A(Md):S
15. dar in waxt | dar kabul | rakbllsa | hukumat me-kard |
in that time In Kabul Rakbil Shah govern ME-did

text Th Th :Cr°Ad Rh :A(Md):S

16. we | dar arabstan | plqambar-e Islam mohamad | wafat ke yafta bud |
and in Arabia prophet-E Islam Mohammed death that found was
text Th Th:Cr:Ad Rh:Et(Md):S

17. o | bad az u | xallf-aa-e ra§uda ke cor xallf-a | bud |

and after that khalif-PL—E Rashid that 4 khalif-RA was

text Th Rh:lr(R):C :ld(Md):
18. kc l hazrat-e abubakar wo hazrat-e omar . . . | bud-an I
that hazrat-E Abu Baker and hazrat-E Omar was-3p

Th:A(Md):S Rh:Ma:P
19. In-aa ham | xelaafat kard-an |
this-PL also khalif duties did-3p

text Th Th:Cr:Ad Rh:Et(Md):S

20. o | dar zaman-e-San | futuhat-e ziat | Sud |
and in time-E-their conquering-E alot became
text Th Rh:G(Md):O :A(Ag):

21. o | besior mamlakat-haa-ra | graft-an |

and very country - PL -RA took-3p

Th:Cr:Ad A(Med):S Rh:R(R):Ad
22. bad az u | hazrat-e omar | taa qlsmat-aa-e Sam 0 flllstln . lfatha kard
after that hazrat-E Omar until section-PL-E Amman and Palestine conquer did

textTh Rh:Ma:P
23. we pas amad
and back came

Figure 8
Analysis Across Metafunctions

144

3. Implications for systemic theory

Any theory has rough edges, and systemic theory is no
exception. As can be seen even in this brief analysis, the various
parts do not always fit together smoothly and certain dilemmas,
even contradictions, arise. The analysis and discussion of 815-823
reveal a few places in particular where some criteria or
classification categories might be re-evaluated.

Within the textual metafunction, Dari shows some variation
from English. For example, textual Themes are prominent, often
filling the entirety of the thematic potential. In some cases, there
are simply no other clausal constituents available. In other cases,
the Predicate might meet the criteria for an ideational Theme, but
that would mean a 'split Theme‘, one part before the Rheme and one
part after. This is worth considering, but in doing so we might have
to rethink the whole notion of message structure. It seems unlikely
that an encoder would structure a message in such a way that the
core of the message lies between two pieces, each of which is the
organizer of the message.

Ideationally, while the process types and participants set up
by Halliday for English differ for Dari essentially in their
distribution and emphasis, some areas of systemic process theory
seem to have more universal implications than other areas.
Halliday's most recent classification of process types is a refining
of the theory based on English. His original thinking on this may be
closer to a universal model. As a tentative observation based on
this analysis of Dari and work on English, three major process types
might be posited: material (including behavioral), mental (including

145

verbal), and a 'being' process including relational and existential.

If such a system were set up, participants directly involved in
each particular process type would necessarily take on a more
universal flavor and, therefore, be more applicable across languages.
This would, of course, be moving away from Halliday's efforts at
greater delicacy. But Halliday already posits two participants that
cross process boundaries: Beneficiary and Range. This analysis of
Dari has already given new insight into the notion of Range; because
of the more prominent role Range plays in Dari, its potential has
been more fully explored than was possible from an analysis of
Enghsh.

Grammatically, of course, languages differ, but some of the
grammatical properties of Dari that differ from English make Dari
difficult to fit into the model that Halliday sets up for English. This
is to be expected. The most obvious divergence is the lack of true
passives in Dari, which sometimes leaves Dari without the contrasts
that Halliday uses to distinguish process types in English. in light
of this difference in the grammatical realization of agency, we have
set up some differing criteria for identification of participant
functions. In fact, the system set up here seems more consistent
across processes than Halliday's system for English.

A second grammatical property that marks Dari is the optional
nature of subjects; the subject slot in Dari need not be filled. This
seems to open the way for true existentials which are both
grammatically and functionally distinct from other 'being' clauses.
But the null realization of participants also presents problems for

the theory. This is especially apparent in an ergative interpretation,

146

where it is possible for the Medium to not be realized at all, even as
a verb suffix. We need to address the question of whether a core

participant in a proposition can have no surface realization.

4. Implications for a theory of language universals

What can we say about the nature of Language? First, the
traditional grammatical categories of Subject, Object, Complement,
Adjunct, and Predicate seem to work well as basic clausal
constituents. By identifying these units, a language can be
categorized into basic clause types. It would seem reasonable that
all languages would in some way encode the basic units of meaning
as transitive, intransitive, relational and existential.

Grammatical patterns are realization patterns and, therefore,
are probably not universal. A more likely area to look for broad
generalizations would be in functional patterns, the notions of
which are more probably universal. For example, the much explored
passive is clearly not universal. Languages such as Dari do not have
passives in the sense of their being counterparts to active clauses.
But all languages most likely have the notion of agency encoded in
some way. In English, an Agent can be encoded as a direct
participant in the process (i.e. as Subject) or more obliquely as an
Adjunct. ln Dari, an Agent can only be encoded as a direct
participant. Thus there is no active/passive contrast.

The universality of process types is more open to variation,
given both cultural differences and differences in the resources of
each language, but all languages will most likely encode meaning

using some process or another and Halliday's major categories of

147

material, mental and relational seem fairly universal. The emphasis
that a particular language places on a particular process type,
however, will vary. In Dari, for example, meaning is frequently
encoded as an existential process, a process type rarely used in
English. On the other hand, meaning in Dari is seldom realized as a
behavioral process. While the details of participant relations will
vary from language to language, the basic participants themselves
seem fairly universal.

Even the thematic or message structure of a language seems
quite consistent across languages. Given a substantive definition of
Theme and Rheme, a language will consistently structure messages,
guiding the decoder toward the core of the message. But the way in
which each language does this may vary. In English, the message
generally builds to the end of the clause. In Dari, however, the mes-
sage builds up to the position preceding the Predicate and then trails
back off; the clausal position that typically carries the highest CD is
that immediately before the Predicate. In light of this, we might
want to posit a message structure for Dari where Rheme is preceded
by a lead-in element and followed by a lead-out element.

5. Conclusion

What has been described here is the nature of this text. From
this exploration, some assertions have been made about the nature
of Dari. Though these particulars are only a small piece of the
puzzle, they bring us one step further in our quest to understand the
nature of language in general. As we explore more and more
particular instances of language, we gain a better insight into this
most complex of human activities.

APPENDICES

Appendix A
Text A; Text B; Text C

Text A
1. dar zamon-aa-o qadim kabul bakll omroz-ara na dait
in time-Pl-E old Kabul like today-RA not have
"In the old times Kabul was not llke today'

2a. maa fiklr kon-lm ke
we think do-1p that

2b. dar markaz-e kabul yak-e herb-e kalon Is
in center-E Kabul a-E fortress-E large ls
"I think that in the center of Kabul is a large fortress'

2c. dar rol-e dar-Ii burja is o tlkaéasu
around in-its turret is and rifle hole
”All around it are turrets and rifle holes”

3. (faqat)
okay?
4. kabul dar roi-e éand bazar-aa-e bazurg bud

Kabul in way-E many bazaar-PL-E great was
'Kabul had many great bazaars”

5. l bazar-aa-ra yak-e-é-aa §ar bazar me-goft-an
this bazaar-Pl-RA one-E-this-PL city bazaar ME-said-ap
'These bazaars--one of these they called City Bazaar'

6. wa yak-e-é-aa darwaz-e Iaurl me-goft-an
and one-E-this-PL door-E Lahore ME-said-ap
"And one of these they called doors of Lahore"

7. yak qlsmat-a i-aa demazan mo-goft-an
one section-E this-PL Demazang ME-said-Sp
"One section of these they called Demazang'

8. yak qlsmat-e é-aa do afgn-aa me-goft-an
one section-E this-PL from Afghan-PL ME-said-3p
”One section of these they called Deafghana'

9a. wa yak qlsmat éor éata [9b. ba estlla be-goy-lm] bud
and one section four comer in speak BE-say-fp was
”And one section was, let’s say, four comers“

148

10. ba sarat-e kabul ba carat-a saqtuman-Ii bud kc a traf-e kabul
in kind-E Kabul In kind-E structure-this was that surround-E
"It was this kind of structure that surrounded Kabul”

11. dawaI-aa-a gall kalon kalon saxta sud-a bud
wall-PL-RA mud large large made become-RA was
”Very large mud walls had been made'

12. wa dar roi-o éor darwaza bud
and in way-E four door was
”And it had four doors“

13. kc I darwaza-e az I-ra mo-g-an ko
that this door-E from this-RA ME-say-3p that
dar zamon-e qadim ke hlndu-ham da kabul zendagl me-kard-an
In old-E section that Hindu-all In Kabul live ME-did-3p
”This door here, they say that in the old section, all the Hindus in Kabul
were living“

14. un-aa-ra ba nam-e paéa-haa-e ratblléa zatblléa aratsa
Ihat-PL-RA by name-E king-PL—E Ratbil Shah Zatbil Shah Arat Shah
zamburaksa on-aa nam me-bar-an ka da tarlx etazakurl yaft-a
Zamburak Shah this-PL name ME-take-ap that in history mention -PF
”They called them by the names of kings, . . ., that were mentioned

In history'

Text B

15. dar In waxt dar kabul rakblléa hukumat me-kard
In that time In Kabul Rakbil Shah govern IMP-did
'At that time in Kabul Rakbil Shah was governing.”

16. we dar arabstan plqambar-e Islam mohamad
and in Arabia prophet-E Islam Mohammed
wafat ke yafta bud
death that found was
”And in Arabia the prophet of Islam, Mohammed, had died."

17. 0 bad az u xaIIf-aa-e raéuda kc éor xaIIf-a bud
and after that khalif-PL-E Rashid that 4 khalif-RA was
"After that were the khalifs of Rashid, the four khalifs.“

150
18. ka hazrat-a abubakar wo hazrat-e omar hazrat—e usman

that hazrat-E Abu Bakar and hazrat-E Omar hazrat-E Osman

wo hazrat-c all bud-an

and hazrat-E Ali was-3P

"They were Hazrat Abu Bakar, Hazrat Omar, Hazrat Osman, and Hazrat Ali."

19. ln-aa ham xelaafat kard-an
this-PL also khalif duties did-3P
”These also acted as khalifs."

20. o dar zaman-e—éan futuhat-o zlat find
and in time—E-thelr conquering-E alot became
”And in their time a lot of conquering occurred.“

21. o besior mamlakat-haa-ra geraft-an
and very country - PL —RA took-3P
”And they took many countries.”

22. bad az u hazrat-e omar taa qismat-aa-e éam o fillstln
after that hazrat-E Omar until section-PL-E Amman and Palestine
0 baltl moqadas fatha kard
and holy house (Jemsalem) conquer did
”After that Omar conquered Jerusalem up to sections of Amman.”

23. wo pas amad
and back came
”And he returned.“

24. markaz-e xalafaa éar-e madina bud
center-E khalif place city-E Medina was
"The center of the khalifs was the city of Medina.”

25. dar arabstan-Is saudl
in Arabia-is Saudi
“It is in Saudi Arabia.”

26a. bad dar awaxar maka find
after in the end Mecca became
"Later in the end it became Mecca.“

26b. éun dar In-ja mazjidl haram bud
since in this-place mosque Haram was
”Since in this place was the Haram Mosque.”

1 51
26c. wo maka on-Ist kc badatgaa-c mucllman act-a

and Mecca place-is that house of worship-E Muslims is—PF
"And Mecca Is the place that is the Muslim's house of worship."

27a. 0 mazjld kc act yak dar bcrun-Ié yak jol-Ist
and mosque that is one in outside-its one place-Is
“And there is a mosque outside "

27b. kc yak sang-c cIa-Ist nasb ast
that one rock-E black-is install is
"of which Is Installed a black rock'

27c. kc u-ra ajarl aswad mc-g-an
that that-RA [rock blacklkrb IMP-said-aP
”they are calling that Ajarl Aswad.“

28a. kc u sang az cang-haa-c asman-Is
that that rock from rock-PL-E sky-ls
“That rock is from the rocks of the sky”

28b. kc ba zamin aftad-a bud-an
that to earth fell-PF was-3P
”that had fallen to earth."

29a. 0 c sang-a hazrat-c cbrahlm kc mc-xost
and this rock-RA hazrat-E Ebrahim that IMP-want
maszd-c maka-ra abad kona
mosque-E Mecca-RA build
”And this is the rock that Ebrahim wanted to build the mosque of Mecca.”

30. amI sang geraft o wardost
this rock took and lifted up
“He took and lifted up this rock.”

31. goft I-ra xoda az asman ba zamin andoxt-a
said this-RA god from sky to earth dropped-PF
"He said, ”This rock God has dropped from the sky to the earth.“

32. wo I moqada act
and this holy is
”And this is holy."

33. I-ra dar dcwar-c xon-c kaba gcraft nacb kard
this-RA in wall-E house-E rock took Installed
"And he took and installed this in the wall of the Hona Kaba.‘

34a.

34b.

34c.

35.

36.

SP

37.

38.

39.

40.

41.

42.

152

ham-c mucllman-aa kc mc-amad-an

all-E Muslim-PL that lMP-came-3P
"All the Muslims that came“

sang-a tamas mc-kard-an

rock-RA touch lMP-did-3P
“they touched the rock'

zlarat mc-kard-an

respect lMP-did-3P
”they paid respect'

éun ba dict-c plqambar-c hazrat-c cbrahlm xord
since by hand-E prophet-E hazrat-E Ebrahim hit
'since it hit the hand of the prophet Ebrahim.”

bad az hazrat-c cbrahlm akwam-c arablstan hama az dln gaéd—an

from hazrat-E Ebrahim tribesNb-E Arabia all from religion abandoned-

"After Ebrahim the Arab tribes all abandoned religion.”

kaflr-c §ud-an
infidel-E became-3P
”They became infidels.“

but parast {ud-an
Idol worship became-3P
“They became Idol worshippers."

wa but-aa-c dait-an kc dar care raf-aa-c xon-c kaba-re
and Idol-PL-E had-3P that In upon shelf-PL-E house-E rock-RA
”And they had Idols on the shelves of the Hona Kaba.“

jol bud kc but-aa-ra dc car - - -
place was that Idol-PL-RA on shelf
(Incomplete proposition)

ar qabila yak but daét kc taquband cc cod 0 sast but
every tribe one idol had that around 3 100 and 60 idol
”Every tribe had one idol, around 360 Idols.”

ar qabila yak but das't
every tribe one idol had
”Every tribe had one idol."

, 153
43. we cc sod 0 cast qabila

and 3 100 and 60 tribe
”And 360 tribes“

44. cc cod 0 cast but-a dar aml raf-aa-c xon-c kaba élnd-a bud-an
3 100 and 60 Idol-RA In this shelf-PL-E house-E rock arrange-PF was-3P
”They had arranged 360 Idols on these shelves of the Hona Kaba.”

45. kc but-aa-c kalon-c az I be ham-c laut o malaut o ozu
that idol-PL-E large-E from this by name-E Laut and Malaut and Ozu

yad mc—éud

remember IMP-became

”The large idols of these were called by the name of Laut, Malaut, & Ozu."
46a. wa yak but-c az ln-aa az yaqud caxt-a sud-a bud

and one idol-E from this-PL from ruby made-PF became-PF was
"And one of these had been made of ruby'

46b. kc clima-aa-c zamarud bud
that eye-PL-E jade was
“with eyes of jade.“

47a. we I besior but-a qimati bud
and this very idol-RA expensive was
”And this was such an expensive Idol“

47b. kc I-ra mc-goft-an
that this-RA lMP-said-3P

47c. I-ra bayad maa parastci kon-Im
this-RA must we worship do-1P
”that the word was we must worship It.

48. waktl bud kc I but-aa-ra tawasut-c mohamad
time was that this Idol-PL-RA through-E Mohammed
kc plqambar-c Islam-Is
that prophet-E Islam-is
'At this time these Idols through the agency of Mohammed who is the
prophet of God“

49. plsar-c abdula wa nawas-c abdul motallb bud
son-E Abdulla and grandson-E Abdul Motalib was
“was the son of Abdulla and the grandson of Abdul Motalib.”

50.

51.

52.

53a.

53b.

53c.

53d.

54a.

54b.

54c.

55.

154
abdul motaIIb mujawlr wa palrador-c xon-c kaba bud

Abdul Motalib curator and guard-E house-E rock was
”Abdul Motalib was the religious curator and guard of the Hona Kaba.”

xon-c kaba kII-i da tacaruf-c az l bud
house-E rock key-Its In possession-E from this was
”He had the key to the Hona Kaba.“

Ixtiardor-e xon-c kaba bud
authority-E house-E rock was
“He was the authority of the Hona Kaba.”

mohamad name-I5 waxt-c kc dId
this Mohammed grandson-his time-E that saw
”This Mohammed, his grandson, when he saw“

kc In-aa az roI-c xoda gait-an

that this-PL from place-E god abandon-3P
’that they abandoned God'

but-aa-ra parastci mc-kon-an

idol-PL-RA worship lMP-do-3P
“were worshipping idols'

raft

left
“he left."

I kc kalon Sud
and he that big became
”And he grew up'

be slnc écl solagl racid

for age 40 years arrived
”he reached the age of 40"

dawi ka rd

claim did
”He claimed”

kc ma plqambar-c xoda-ct-m
thatl prophet-E god-be-1P
”that he was the prophet of God.“

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

1
we xoda amlr kard-a bare rm
and god order did-PF for me
"And God had instructed me'

kc I but-ca qabII-c parasteé n-cct
that this Idol-PL worthy-E worship not-Is
”that these Idols are not worthy of worship.”
bayad suma ma-ra paractci kon-In
must you I-RA worship do.2P

"You must worship me.“

kc ma suma-ra hast kard-Im
thatl you-RA create did—1P
”that (because) I created you.”

asman o zamin-a hast kard-lm
sky and earth-RA create did-1P
"I created the sky and the earth.”

afta o slta-aa-ra hast kard-lm
sun and star-PL-RA create did-1P
”I created the sun and the stars."

bayad hama-gl ma-ra parastes kon-an
must all l-RA worship do-3P
”All of these must worship me.”

I but-c kc xud-c-tan da dlst-c-tan talar mc-kon-In
this idol-E that sclf-E-2P In hand-E-2P construct IMP-do-2P
”These idols that you yourselves are constructing with your hands”

wa un-ja mand-ln
and that-place left it-2P
”and you left it there'

m c r - I n
go-2P
'you are going (there)"

sajda mc-kon-In
bow down lMP-do-2P
'you are bowing down (before them)"

67.

68.

69

70.

71.

72.

73.

74.

75.

76.

77.

156
u-ra ebadat mc-kon-In

that-RA worship IMP-do-2P
'you are worshipping them.“

I qarat act
this wrong is
”This is wrong.”

. padcr-aa-c jad o kaka-c mohamad I-ra awal qabul na kard-an
father-PL-E old and uncle-E Mohammed this-RA first accept not did-3p
"The ancestors and the uncles of Mohammed at first did not accept this.“

tana kason-c kc qabul kard
only people-E that accept did
“Only some people accepted.“

yak zan-e hazrat-c mohamad hazrat-c xadija nam das't
one wife-E hazrat-E Mohammed hazrat-E Khadija name had
"One wife of Mohammed was named Khadija."

l qabul kard
she accept did
”She accepted.”

bad a: hazrat-c xadija az mard-aa-c bezurg-c qabila

after hazrat-E Khadija from person-PL-E important-E tribes

abu baker cadlq bud

Abu Bakar Sadiq was

”After Khadija from the Important people of the tribes was AbuBakarS.“

xaIIf-c awal bud
khalif-E first was
”He was the first khalif.”

I qabul kard
he accept did
"He accepted.”

bad az u az baéa-aa-c xord kc ast sola baéa bud
after that from boy-PL—E little that 8 years boy was
“After that from the young sons was an 8-ycar-old boy.”

hazrat-c all karam allah bud
hazrat-E Ali [term of respect] was
”He was Ali“

78.

79.

80.

re

81.

82.

. 157
kc I baca-c kaka-c plqambar bud

that he son-E uncle-E prophet was
”who was the son of the uncle of the prophet."

I qabul kard
hcacccptdld

"He accepted.“

0 In-ac amroi mohamad raft-an dar xon-c kaba-c but-ca-ra kulda-
and they with Mohammed went-3P in house-E rock-E idol-PL-RA all them-
"And they went with Mohammed to the Hona Kaba of idols, all of them.“

anuzl qabila xabar na dost-an c6 kac

yet tribes information not had-3P no person

”Yet the tribes did not know-nobody.“

kul-c but-aa-ra §Ikastand o payan andoxt

all-E idol-PL-RA broke and down dropppcd
”He threw down and broke all the Idols.”

83a. wa clan kard

and announce did
”And announced"

83b. kc plqambar-e xoda iud

that prophet-E god became
”that he was the prophet of God.“

83c. ma az taraf-c xoda amad-Im

I from direction-E god came-1P
'I came from God.”

Text C

84.

85.

we Ina awal sura barc ma az qoran-c iarlf amad-a
and this first verse for me from Koran-E holy came-PF
"And this first verse came to me from the holy Koran.“

cor kitab-c asman-let
four book-E heaven Is
"There are the four books of heaven.“

158

863. taurat-Ict
Torah Is
”There is the Torah."

86b. Injll-Ist kc hazrat-c ccau-c
Bible Is that Hazrat-E Jesus-ls
“There Is the Bible that is from Jesus."

87. zabur az hazrat-c daud-Ict
Old Testament from Hazrat-E DavId-Is
"The Old Testament Is from David."

88a. faqan az hazrct-c mohamad-let
Koran from Hazrat-E Mohammed-Is
”The Koran Is from Mohammed.“

88b. kc l cor kitab—c asman-lat
that this four book-E heaven-Is
'These are the four books of heaven'

89. kc az taraf-c xoda amad-a
that from direction-E god came-PF
“that came from the direction of God.“

90. we amagl be I kitab-c Iman dora
and everyone by this book-E faith has
”And everyone has faith In this book."

91. musllman-aa-am ba hazrat-c musau ba hazrat-c ccau
Muslim-PL-all by Hazrat-E Moses by Hazrat-E Jesus
be hazrat-c daud ba hazrat-c mohamad Iman dor-an
by Hazrat-E David by Hazrat-E Mohammed faith have-3p
”All Muslims have faith in Moses, Jesus, David, Mohammed.”

92. kc In-aa pIqmbar-c morsal actan kc az taraf-c xoda
that this-PL prophet-E genuine are that from direction-E god
baroi cdoyat-e ad
for religious-E direction
"They are the genulne prophets from God for religious direction.”

93. aqm-c xud bcra ind-a bud-an
tribes-E self lost became-PF was-3p
”The tribes themselves had become religiously lost."

 

94.

95.

96.

97.

98.

99.

100.

101.

159

nazer sud-an kc bayad plrewl ez u-aa-re kon-en
observer became-3p that must follow from that-PL-RA do-3p
ewemlI-c xoda-re

affairs-E god-RA

“They observed that they must follow those that do the deeds of God."

barol omat bare mardum bc-reeon-en
for follower for people should-deliver-3p
“They should deliver for the followers of God, for the people."

we mardum-c ez rol qalet az rol duzdi we edam kuél o

and people-E from way wrong from way thievery and man kill and

e: mi qimor o az rol kor-aa-c bezet o icrab xur-c zlet

from way gamble and from way matter-PL-E and alcohol drinking-E alot

kc be mardum zerer mc-recone

that by people Injury ME-deliver

“And the people from the way of thievery, murder, gambling, and excessive
alcohol drinking that Injures people . . . "

o mardum-e czar mete ez I-re-ae e2 duzdi o screfat
and people-RA bother glve from thls-RA-PL from thievery and "b

az heme élc bayad mane kon-e

from all thing must prohibit do-PF

"Bothering people from these, from thievery, from everything--

it must be prohibited.”

be: I farman-hea tawasut-c plqmber berc mardum mc-ras-an
then this decree-PL via-E prophet for people ME- arrive-3p
”Then these decrees arrived through the prophet for the people."

dar waxt erab-ae mardum-c wail bud-en
In time Arab-PL people-E uncivilized was-3p
"At that time the Arabs were uncivllized people."

mohamad kc amad areb-ea-ra goft kc

Mohammed that came Arab-PL-RA said that

sume auled-ee-c-ten-a na kust-In

you chlld-PL-E-your-RA not kill-2p

”Mohammed who came told the Arabs not to kill their children.“

éere un-ee flker mc-kerd-en doktar bed ast-an
because that-PL think ME-dId-3p daughter bad is-3p
“Because they thought daughters were bad.“

1028.

102b.

103.

104.

105.

106.

107.

108.

109.

110.

111.

160

nen daét-en
pride had-3p
“They had pride.“

doktar-ee-c xud-e dar xerdl mc-kuét-an
daughter-PL-E self-RA In youth ME-kiII-3p
“They were killing their daughters In their youth.”

we qimor mc-zad-an
and gamble ME-hit-3p
“They gambled.”

doktar-ea-e xud-a de qimor mc-boxt-en
daughter-PL-E self-RA in gamble ME-lost-3p
"They lost their daughters In gambling.”

agar xué-e-éen mc-emad zen-c yak-c degar paisa mc-dad-an
if liking-E-their ME-came wife-E one-E other money ME-gave-Sp
”If they liked, they bought another’s wife.“

zan-c-i-ea elii mc-kerd-en
wife-E-theIr-PL exchange ME—did-3p
"They exchanged wives.“

zen-c xud-a bar u mc-ded-an
wife-E self-RA for that ME-gave-3p
”They gave their wives to them.”

zen-c maqbul-c ez u-re mc-greft-en
wife-E beautiful-E from that-RA ME-took-3p
"They took the beautiful wife of another.”

arab-ae edam mc-kus‘t-en
Arab-PL people ME-kill-3p
"They were killing people."

duzdi mc-kerd-en
thievery ME-dId-3p
"They were robbing.“

meI-c kecc-ra kc zurg dait-en be zurg me-greft-an
property-E people-RA that power had-3p by power ME-took-3p
“They took by force the property of the people that had power.“

112.

113.

114.

115.

116.

118.

119.

120.

121.

122.

161

doktar-ea-c mardum-c be zurg mc-greft-en
daughter-PL-E people-E by power ME-took-3p
"They took people's daughters by force.”

we merdum-ee-c ereb kc out o garlb bud-en
and people-PL-E Arab that propertiless and poor was-3p
doktar-ee-c xud-e mc-fruxt-an

daughter-PL-E self-RA ME-selI-3p

”And the Arab people who were propertiless and poor sold their daughters.

mohamad emed
Mohammed came

goft I kuI-c éls-ea-ra xoda mane kard-a
said this all-E thing-PL-RA god prohibit did-e
“He said, 'God has prohibited all of these things.”

bcdat-Ie
sin-is
“It is a sin.”

bayad doktar-ee-ten-e ne kuet-In
must daughter-PL-your-RA not kill-2p
"You must not kill your daughters.“

zen-ae-e-tan ne frui-In
wife-PL-E-your not sell-2p
”You (must) not sell your wives.“

qlmor ne zen-In
gamble not hit-2p
”You (must) not gamble.“

§ereb na xur-In
alcohol not drink-2p
"You (must) not drink.“

lIbei-c-ten-e Ien-c-ten-e peq neqe deét-e b-ei-In
clothes-E-your-RA body-E-tan-RA clean keep have-PF BE-is-2p
"You should keep clean your clothes, your body.”

pain] waxt bayad uzu bc-ger-ln
tive time must ablutions BE-teke-2p
"You must take ablutions five times."

123.

124.

125.

126.

127.

128.

129.

130.

131.

162

der majid bur-In
in mosque BE-go-2p
"You should go to mosque.”

wo der belnc mejld cbedet kon-In
and In inside mosque pray do-2p
“And Inside the mosque you should pray.”

cur-c el hemd-e bo-xen-In
verse-E -Ra BE-reed-2p
"You should read the verse al hamd.‘

wo cur-c axles-e bo-xen-In
and verse-E -RA BE-read-2p
"And you should read the verse eklas.”
mlell kc pester berc ten goft-um
like that earlier for you said-ts
”Like I told your earlier"

kc awal-In sure kc amad bare plqmber anemu cur-c [Arabic]
that first-very verse that came for prophet that one verse-E
”that the very first verse that came to the prophet, that verse [name]."

kc u her musllmen mc-fem-e l sure-hee-re
that that every Muslim ME-know-PF this verse-PL-RA
misli kc mee mc-fem-Im

like that we ME-know-tp

“Every Muslim knows these verses like we know (them)."

der u sure newléte kard-e [Arabic]
in that verse written did-PF
”In that verse was written [Arabicl'

yenc boxan be nem-c xoda-c bazurg-c kc tu-re heat kard-a
means read by name-E God-E powerful-E that you-RA create did-PF

az yak eu high

from one water

”Translation: In the name of the powerful God that created you from water"

132a.wo tu-re baujud award-e

end you-RA bring Into being-PF
"And brought you into being'

163

132b.wo tu-re n-cet mc-kon-e

133.

134.

135.

and you-RA not-be ME-do-PF

'And takes you out of being'
we cé rol-ra nero be just parestce-c perwerdcgor-c
and nothing way-RA don't go except worship-E creator-E

kc u temam-c mexluqet-c dunye-re paIde kard-e

that that whole-E creations-E world-RA create did-PF

”And go no way except worship of the creator that has created the
entire creations of the world."

I emrc bud ye cure bud kc berc plqmber awal defer amad
this edict was or verse was that for prophet first time came
”This was the edict or verse that came the first time for the prophet.”

bed ez u emr-c namaz sud
after that edict-E pray become
"After that come the edict to pray."

APPENDIX B

Theme/Rheme Structure: Text B

Theme is underlined; Rheme is in italics

15. mm W rekbllie I hukumat mc-kard I/
In that time In Kabul Rakbil Shah govern ME-dld

16. 13.! Ml plqember-c Islam mohamad l wafat kc yafta bud
//
and In Arabia prophet-E Islam Mohammed death that found was

17. _g__/ M I xellf-ee-c reéude kc d‘or erIf-e I bud //
and after that khalif-PL-E Rashid that 4 khalif-RA was

18. .KQJ hazref-c abubakar we hezrat-c omer hazret-c asman wo hazrat-
c all I bud-an

that hazrat-E Abu Baker and hazret-E Omar hazrat-E Osman and hazrat-E
Ali was-3p

19. NJ xelaafat kard-en II
this-PL also khalifing did-3p

20. MWI futuhat-c ziat I Sud II
and in tImc-E-thclr conquering-E alot became

21. _g_/ besior mamlaket-hea-re I graft-en //
and very country - PL -RA took-3p

22. m I hazret-c omer / tee qlcmet-ae-c 5am a flllstln o baltl
moqadas I

after that hazrat-E Omar until section-PL-E Amman and Palestine and
Jerusalem

fatha kard //

conquer did

23. _w_g_/ pee amad //
and backcame

164

165

24._m3m1;g__xalgj_-_n leer-c madina I bud I/
center-E khalif-PI clty-E Medina was

25. der arabstan I -Ic I ceudl II
In Arabia- Is Saudi

26a..b.ed_I_der_menLI make I ind //
after in the end Mecca became

26b. _§_u_n_lda:_]n:]a_l mazjidl harem I bud II

since In this-place mosque Haram was

26c._w_g_ I_mg]§a_/ lol I -Ict I kc bedefgea-c musllmen est-e I I
and Mecca place -ls that house of worship-E Muslim is-PF

27a. 9. Inna—IWWI
and mosque that is one in outside-its
27b. yak 10! I -Ist I kc yak sang-c cle-Ist nesb est
one place -is that one rock-E black-is install is
2 7c. kc u-re elerl aswad mc-g-en II

that that-RA [rock blacijrb ME-said-3p

28a. kg I_u_gang_/ ez eeng-hee-c asman / -Is I
that that rock from rock-PL-E sky Is

28b. kc be zemln efted-e bud-en II
that to earth fell-PF was-3p

29a._g_I __|_§_ang;gl hezrat-c cbrahlm I kc mc-xost I
and this rock-RA hazrat-E Ebrahim that ME-want
29b. mezjld-c make-re abed kona II
mosque-E Mecca-RA build

30. MI graft o werdost II
this rock took and lifted up

31. goft II Ha I xoda I an asman I be zamin I andoxt-a II
said this-RA god from sky to earth dropped-PF

32. we I_Lj moqede I act II
and this holy Is

33. _I;La_l der dcwer-c xon-c kaba I graft nesb kard II
this-RA In wall-E house-E rock took Installed

166

Maw
all-E Muslim-PL that ME-camc-3p

34b. sang-a I tamas mc-kard-an II
rock-RA touch ME-did-3p

34c. zlarat mc-kard-en II
respect ME-did-3p

35. £1.an be dist-c plqember-c hazrat-c cbrahlm / xord II

since by hand-E prophet-E hazret-E Ebrahim hit
36W akwem-c arablsten heme Iez dInI
from hazrat-E Ebrahim tribesArb-E Arabia all from religion
gesd-an II '

abandoned-3p

37. kefIr-c I sud-en II
infidel-E became-3p

38. but parast I sud-an II
idol worship became-3p

39. u Ibut-aa-c / dest-an I kc der sarc ref-ae-c xon-c kaba-re I I
and idol-PL-E had-3p that In upon shelf-PL-E house-E rock-RA

40. lol I bud I ke but-aa-re do care - - -/I (unanalyzed)
place was that idol-PL-RA on shelf

41.j:__gab,|1a_j yak but I deét I kc taqubend ac cod 0 seat but II
every tribe one idol had that around 3 100 and 60 Idol

42. MI yak but I deét II
every tribe one idol had

43. ml cc cod 0 iect qabile II
and 3 100 and 60 tribe

44. WW I der eml ref-ea-c xon-c kaba I éind-e bud-en II

3 100 and 60 Idol-RA In this shelt-PL—E house-E rock arrange-PF was-3p
4&ng WI be nam-c leaf 0 malaut o ozu I yad mc-
‘éud //

that idol-PL-E large-E from this by name-E Laut and Malaut and Ozu
remember

46a. ELI W I a: yaqud I eexte sud-a bud I
and one idol-E from this-PL from ruby made became-PF was

167

46b._|§_c_ éls'ma-ee-c zamarud bud II
that eye-PL-E jade was

47a. 11 I_|_/ besior but-a qlmetl I bud I
and this very Idol-RA expensive was

47b. nl [-[g l mc-goft-en I
that this-RA ME-said-3p

47c. hm I bayad I me I parastes kon-Im II
this-RA must we worship do-1p

48._w_a]§t]__b_u_d_kg_l MI tewecut-c mohamad kc plqamber-c

Islam-Is
time was that this idol-PL-RA through-E Mohammed that prophet-E
Islam-is

49. pleat-c abdula we newes-c abdul mofallb I bud II
son-E Abdulla and grandson-E Abdul Motalib was

50._ahd_u_|_mgja|,|.b_l mulawlr we palrador-e xon-c kaba I bud II
Abdul Motalib curator and guard-E house-E rock was

51. mm Ilsll-é Ida tasaruf-c ez II bud II
house-E rock key-Its In possession-E from this was

52. Ixtlardor-c xon-c kaba / bud II
authority-E house-E rock was

53a._]__mghgmgln_g_w_g_g_-_L§_ I waxt-c kc I did II
this Mohammed grandson-his time-E that saw

53b. kc I In-aa / az roi-c xoda gait-en I
that this-PL from place-E god abandon-3p

53c. but-aa-re I pareetci mc-kon-en II
idol-PL-RA worship ME—do-3p

53d. raft //
left

54a._g_/_L [I kc kalon sud II
and he that big became
54b. be slnc écl solagl I resld II]
for age 40 years arrived
54c. dawi kard II
claim did

168

55._k_g_ I ma__/ plqembar-c xoda I -et-om II
that I prophet-E god Is-1s

56._w_g_/x_g_d,a_/emlr kard-e I bare me I
and god order did-PF for me

57._[§_c_ I| hut-ea I qebII-c parastes I n-cst II
that this idol-PL worthy-E worship NEG-is

58. bayad/WI ma-re I pereetcé kon-In I
must you l-RA worship do-2p

59. kg/m/éuma-ra I hast kerd-Im II
that I you-RA create did-1p

60. asman a zamIn-e I hast kerd-Im /I
sky and earth-RA create did—1p

61. afta a sIfe-ee-ra I hast kard-Im II
sun and star-PL-RA create did-1p

62. mulnmullme-re I paraetci kon-en II
must all l-RA worship do-3p

63._|_m Ikc xud-c-tan I de dIet-c-ten I teler mc-kon-ln I
this idol-E that self-E-you in hand-E-you construct ME-do-2p

64. minute I mend-In II
and that-place leave -2p

65. mar-In II
QO'ZP

66. sajda mc-kon-ln I I
bow down ME-do-2p

67. u-[a/ ebadat me-kon-In II
that-RA worship ME-do-2p

68._[_/ qarat I act II
this wrong is

69. pedcr-aa-c [ad a kaka-c mohamad I_|;m I awal I
father-PL-E old and uncle-E Mohammed this-RA first
qabul ne kard-en II
accept NEG did-3p

169

70. tana kason-c kc I qabul kerd II
only people-E that accept did

71. WWI hazret-c xadija / nem dast II
one wife-E hazrat-E Mohammed hazrat-E Khadija name had

72._1/qabul kard I/
she accept did

73 mm In I mard-aa-e hazard-e qabila I
after hazrat- E Khadija from person PL-E important-E tribes

abu baker sequ I bud II
Abu BakarSadiq was

74. xaIIf-c awal I bud II
khalif-E first was

75._|_/ qabul kard II
he accept did

76._hag_az_u_ I az baée-ee-c xord kc / est cola baéa I bud II
after that from boy-PL-E little that 8 years boy was

77. hezrat-c all karam allah I bud I
hazrat-E All [term of respect] was

78. kcl LI baée-c keke-c plqamber I bud II
that he son-E uncle-E prophet was

79. L] qabul kard II
he accept did

80._g_I1n;aa_/emrol mohamad Ireft-en/
and this-PL with Mohammed went-3p
dar xon-c kaba-c but-aa-re I kulda-ra II
in house-E rock-E idol-PL-RA ell them-RA

81 ._a_n_u_z_[/ mum xabar ne dasf-en I cc kec I/
yet tribes Information NEG had-3p no person

82W élkastend o payan andoxt II
all-E idol-PL-RA broke and down dropppcd

83a. ml clan kerdl
and announce did

170

83b._k_e_l mu pIqambar-c xoda I -st-om II
that l prophet-E god ls-1s

83c._ma_/az taraI-c xodel emed-lm II
I from direction-E god came-1p

APPENDIX C

Translation of the Text

Text A:

In the old times Kabul was not like today. I think, in the center
of Kabul was a fortress with turrets and rifle holes around it. Kabul
had many great bazaars: Shar Bazaar, Darwaze Lauri Bazaar,
Demazang, DeAfghana, Char Chata. These kinds of structures
surrounded Kabul, with very large mud walls. In one old section, all
the Hindus lived. They were named after the kings mentioned in
history: Ratbil Shah, Zatbil Shah, Arat Shah, Zamburak Shah.

Text B:

At that time, Ratbil Shah was governing in Kabul. And in
Arabia, the prophet of Islam, Mohammed, had died. After Mohammed,
came the four khalifs of Rashid: Abu Bakar, Omar, Asman, Ali. And in
their time, a lot of conquering occurred and they took many
countries. Omar conquered Palestine up to Jerusalem and sections
of Amman, Jordan. Then he returned.

The center of the khalifs was the city of Medina which is in
Saudi Arabia. Later Mecca became the center because the Haram
Mosque is in Mecca. It is outside this mosque that a black rock
called Ajarl Aswad is situated. This rock fell to earth from heaven.
Ebrahim wanted to build the mosque of Mecca with this rock. He
took and lifted up the rock and said, "This rock God dropped from the

171

172

heaven to earth, and it is holy." He took and installed the rock in the
wall of the House of the Rock. All the Muslims that came, touched
the rock and paid respect because it had hit the hand of the prophet
Ebrahim.

After Ebrahim, the Arab tribes all abandoned religion. They
became infidels. They became idol worshippers. They had idols on
the shelves of the House of the Rock. There were 360 tribes and
each tribe had one idol; 360 idols were arranged on the shelves. The
large idols had names: Laut, Malaut, Ozu. One of these was made of
ruby with eyes of jade. This was such an expensive idol that the
word was that we must worship it.

Mohammed was the son of Abdulla and the grandson of Abdul
Motalib, the religious curator and guard of the House of the Rock. He
had the key to the House of the Rock and was the authority. When his
grandson, Mohammed, saw that they had abandoned God and were
worshipping idols, he left. When he grew up and reached the age of
forty, he claimed that he was the prophet of God, and God had
instructed him that these idols were not worthy of worship: ”You
must worship me because I created you. I created the heaven and the
earth. I created the sun and the stars. All of these must worship
me. You construct idols with your own hands and, leaving them
there, you go and bow down and worship them. This is wrong.”

At first the ancestors and relatives of Mohammed did not
accept this. Only some people accepted. One wife of Mohammed was
named Khatija. She accepted. After Khatija, from the important
people of the tribes, was Abu Bakar Sadiq. He accepted. From the
young sons was an eight-year-old named All, the son of the uncle of

173

the prophet. He accepted. They all went with Mohammed to the
House of Rock. Yet the tribes did not know--nobody knew. He threw
down and broke the idols. And he announced, "I am the prophet of
God. Icome from God.”

Text C:

”And this is the first verse that came to me from the holy
Koran.” [Arabic]

There are the four books of heaven. There is the Torah. There
is the Bible which is from Jesus. The Old Testament is from David.
The Koran is from Mohammed. These are the four books of heaven
that came from God. Everyone believes in these books. All Muslims
have faith in Moses, Jesus, David and Mohammed. They are the
genuine prophets of God for religious direction.

The tribes themselves had become lost religiously. They saw
that they must follow those that do the deeds of God. They should
deliver as followers of God for the people. They should deliver the
people from stealing, murder, gambling and drinking. lnjuring people
in this way must be prohibited. These decrees were delivered for
the people through the prophet.

At that time the Arabs were uncivilized people. Mohammed
came and told the Arabs not to kill their children because they
thought it was bad to have daughters. They had pride so they killed
their daughters when they were young. They gambled and lost their
daughters in gambling. If they liked someone else's wife, they
exchanged wives. They gave their wives to others. They took a

beautiful wife from another. They were killing people. They were

174

robbing people. They took by force the property of the powerful.
They took people's daughters by force. And the Arab people who
were propertiless and poor sold their daughters.

Mohammed came. He said, ”God has prohibited all of these
things. It is a sin. You must not kill your daughters, sell your wives,
gamble, or drink. You should be clean in body and appearance. You
must make ablutions five times. You should go to the mosque, and
inside the mosque you should pray. You should read the verse al
Hamd. You should read the verse Eklas." Like I told you earlier, that
was the very first verse that came to the prophet. Every Muslim
knows these verses like we know them. In that verse is written
[Arabic]. Translated, that means: in the name of the powerful God
that created you from the waters and brought you into being and will
take you out of being. There is no way except worship of the creator
that has created the entire creations of the world. This was the
edict or verse that came for the first time to the prophet. After

that came the edict to pray.

LIST OF REFERENCES

LIST OF REFERENCES

Aghbar. AliAsghar- 1981. Wanna. Ph.D.

Dissertation, Georgetown University.

Berry, Margaret. 1990. "Thematic analysis and stylistic
preferences.“ Paper presented at the 17th International
Systemic Congress, University of Stirling, Scotland.

Bashir, Elena. 1986. ”Beyond split-ergativity: subject marking in
Wakhi" mammmmmmm
Winn 22:1 14- 35

Bing, Janet Mueller. 1980. ”Linguistic rhythm and grammatical
structure in Afghan Persian. W 11 .:3 437- 463.

Bolinger, Dwight. 1975. W, 2nd ed. New York:

Harcourt Brace Jovanovich, Inc

Chafe, Wallace. 1976. 'Givenness, contrastiveness, definiteness,

subjects, topics, and point of view." In W, ed.
Charles N. Li. New York: Academic Press, Inc. pp. 25-56.

Cook. Walter A. 1979. We):
Made] (1970-1978). Washington, DC: Georgetown University

Press.

Cooreman, Ann. 1982. 'Topicality, ergativity, and transitivity in
narrative discourse. evidence from Chamorro." W

Language 6-3-

175

176

---------- . 1983. "Topic continuity and the voicing system of an
ergative language: Chamorro.” In Givon, 1983, p.425-489.

Cooreman, Ann, B. Fox and T. Givon. 1984. ”The discourse definition

of ergativity." W 8.1.
Entezar, M. Ehsan. 1964. W31. Putney, Vermont:

The Experiment in International Living.

Fillmore, Charles J. 1968. "Case for case." In W
W, ed. Emmon Bach and Robert T. Harms. New
York. Holt Rinehart and Winston, Inc. pp.1- -90.

Firbas, Jan. 1966a. "On defining the theme in functional sentence

analysis.” In W ed Josef
Vachek. University of Alabama Press. pp. 267- 280.

---------- . 1966b. "Non- thematic subjects in contemporary

English” In W ed Josef
Vachek. University of Alabama Press. pp. 239- 256.

---------- . 1987. "On two starting points of communication. " In
-_|-_-.-: H ' I .0“ 0 u l- : u-. u, VO|.,1
eds., R. Steele and T. Threadgold. Amsterdam. John Benjamins.
pp. 23- 46.

Fries, Peter. 1983. "On the status of Theme in English: arguments

from discourse” In W eds

Janos S. Petofi and Emel Sozer. Hamburg. Buske.

---------- . forthcoming. ”Patterns of information in intial position

in English " In W
eds. PH. Fries and M. Gregory. NOI'WOOd, NJ. Ablex.

---------- . unpublished. "Toward a discussion of the flow of
information in a written English text.”

177

Fries, Peter and Gillian Francis. 1990. ”Exploring Theme: problems
for research [1].” Paper given at the 17th International
Systemic Congress, University of Stirling, Scotland.

Glassman. Eugene. 1972. QonxersationaLDanntLodumom
W Qafl, revised edition. Kabul: The Language and

Orientation Committee, International Afghan Mission.

Givon Talmy. ed. 1983. mm
W Amsterdam John

Benjamins.

---------- . 1989. MW

Hillsdale, NJ: Lawrence Erlbaum Assoc.

Greenberg, Joseph. 1963. ”Some universals of grammar with
particular reference to the order of meaningful elements." In

W, second edition, ed. J. H. Greenberg. pp.
73-113.

Gregory, Michael. 1967. "Aspects of varieties differentiation."

MW 3 177- 198

---------- . 1985. ”Towards communication linguistics: a

framework." In W. vol1. eds
J. Benson and W. Greaves. Norwood, NJ. Ablex. pp. 119-134.

Haider, Hubert. 1985. "Chance and necessity in diachronic syntax--
word order typologies and the position of Modern Persian

relative clauses." In W 34
Ed. Jacek Fisiak. John Benjamins. pp. 199- 216.

Halliday, M.A.K. 1961. "Categories of the theory of grammar." mm
17:241-292.

---------- . 1966. "Lexis as a linguistic level.” In WE,
Elfin, eds. C. E. Bazell, J.C. Catford, M.A.K. Halliday, R.H. Robins.

London: Longman's. pp. 148-162.

178

---------- . 1967. "Notes on transitivity and theme in English

(Parts 1-3)." W 3.1: 37-81; 3.2: 199-244;
4.2: 179-215.

---------- . 1970. "Functional diversity in language as seen from a
consideration of modality and mood in English." EQundaflm

Language 6.1: 322-361.
---------- . 1976. W Ed. GB. Kress.

London: Oxford University Press.

---------- . 1979. "Modes of meaning and modes of expression:
types of grammatical structure, and their determination by
different semantic functions." In W

W. eds. D..J
Allerton, Edward Carney, David Holdcroft. Cambridge.

Cambridge University Press. pp. 57-79.

---------- . 1981. "Text semantics and clause grammar: some

patterns of realization." In W.
eds. J. E. Copeland and P. W. Davis. Columbia, SC: Hombeam
Press. pp. 31-59.

---------- . 1984. "On the ineffability of grammatical categories."

In W. eds. A. Manning. P. Martin. K.
McCalla. Columbia, SC: Hombeam Press. pp. 3-18.

---------- . 1985. WWW London:
Edward Arnold.

Henderson, Michael M ..J 1978. "Modern Persian verb morphology."
W 98: 375

Hodge, Carleton. 1958. "Review of W
mm by Gilbert Lazard." Language 34:119.

179

Hook, Peter E. 1986. "Null valents in the expression of impersonal

action in Kashmiri and Russian. W1
22:179- 194.

Hook, Peter E. and Omkar Nath Koul. 1987. "Subject versus agent: a
study of the Kashmiri phrasal verb nye 'begin to'. " JeumeLeI
WM 36 1 -:4 115-132.

Jordan, Michael P. 1985. "Non-Thematic re-entry: an introduction to
and extension of the system of nominal group
reference/substitution in everyday English use" In 5151911116

W, Vol.1 e,ds. J. Benson and W. Greaves.
Norwood, N. J. Ablex. pp 322-332.

KatZ. JJ and PM Postal 1964. AtheoLatethmgLLmomstIc
Qeeefletjene. Cambridge: MIT Press.

Langandoen. D. Terence. 1970. W
New York: Holt.

Lazard, Gilbert. 1965 . "L'enclitique nominal :1 en Persan: un ou

deux morphemes?" WNW
Earls 61.1:249-264.

---------- . 1970. "Persian and Tajik." In W11
Linguistics 5364-96-

---------- . 1982. "Le morpheme [.6 en Persan et les relations

actancielles" WW8
77.1: 177- 207.

Li, Charles and Sandra Thompson. 1976. "Subject and topic: a new

typology of language." In W, ed. Charles N. Li.
New York: Academic Press, Inc. pp. 457-489.

---------- . 1981. MW
Grammar. Berkeley: U. of California Press.

180

Longacre, Robert E. 1983. InefiramrnaLgLQiseeuree. New York:

Plenum.

Martin, James R. 1985. "Process and text: two aspects of human

semiosis." In WWII. eds.

J. D. Benson and W. S. Greaves. Norwood, N.J.: Ablex. pp. 322- 332.

Miremadi, Seyed- Ali. 1981. W
W Ph..D

Dissertation, Michigan State University.

Moyne, John A. 1974. "The So-called Passive in Persian."
Eoundatiomuanouage 12: 249- 67

Nichols, Johanna. 1986. "Head-marking and dependent-marking
grammar." Language 62.1: 56-119.

Pike, Kenneth L. and Evelyn G. Pike. 1982. firamrnatieaLAnaiyeie.
Dallas: SIL.

Rashidi, Linda S. 1983a. "Linguistic signals of activeness and

passiveness in Durrell's Alexandnafiuactet" In W
Eenrm, ed. J. Morreall. Columbia, SC: Hombeam Press. pp.405-
412.

---------- 1983b WWI:
Lawrenge_D_u_rrei_L'§_The Alexandria Quartet. M A. Thesis,
Central Michigan University.

---------- . 1985. "Complexity of reality in Durrell's Aiexansiria

Quartet." MW. Vol. 2. eds J.
Benson and W. Greaves. NonIvood, N ..:J Ablex. pp. 204-224.

---------- . 1987. "Participant reference in a Dari narrative."
Manuscript.

181

---------- . 1988a. "Functional co- occurrence restrictions in the

Fairy Tale." In W.
eds. J. Benson and W. Greaves. NonIvood, N..:J Ablex. pp. 113-

135.

---------- . 1988b. "Greenberg's syntactic universals: an
investigation of Dari." Manuscript.

---------- . 1988c. "A preliminary look at the Dari noun phrase: a
stratificational view." Manuscript.

---------- . 1989a. "The experiential structure of the Dari nominal
group." Paper presented at the 16th International Systemic
Congress, Helsinki, Finland.

---------- . 1989b. "Retrogressive devoicing in a Dari narrative."
In the 151h_LAC_U_S_Eo_rurn, eds. Ruth Brand and David Lockwood.
Lake Bluff, II: LACUS. pp. 312-318.

---------- . In press. "Toward an understanding of the notion of

Theme: an example from Dari." In W
W. eds M Davies and L.

Ravelli. London: Pinter.

Tabaian, Hessam. 1979. "Persian compound verbs." Lingua 47: 189-
208.

Windfuhr, Gernot L. 1979. Eersiarnfirarnrnar. The Hague: Mouton.

---------- . 1985. "A spatial model for tense, aspect, and mood."

EQII'LLinouistica XIX 3-4: 415- 461.

---------- . 1987. "Persian." In WW. ed.
Bernard Comrie. New York: Oxford University Press. pp. 523-

546.

Yar—Shater, Ehsan. 1970. "Iran and Afghanistan." In QULLQDIJIQIJQS
In_Linouistlss 6: 669- 689.

   

"Illlllllllllllllllllliilllfl