Syntax and Communicative Strategies in Intermediate German Composition - Virginia M. Coombs - 1986 - The Modern Language Journal

Syntax and Communicative Strategies in Intermediate German Composition
Author(s): Virginia M. Coombs
Source: The Modern Language Journal, Vol. 70, No. 2 (Summer, 1986), pp. 114-124
Published by: Wiley on behalf of the National Federation of Modern Language Teachers Associations
Stable URL: .
Accessed: 28/06/2014 12:27
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
/>
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact

.

Wiley and National Federation of Modern Language Teachers Associations are collaborating with JSTOR to
digitize, preserve and extend access to The Modern Language Journal.



This content downloaded from 46.243.173.175 on Sat, 28 Jun 2014 12:27:11 PM
All use subject to JSTOR Terms and Conditions


Syntax and CommunicativeStrategiesin
IntermediateGerman Composition
VIRGINIA M. COOMBS

WRITING

IN A FOREIGN

LANGUAGE

CONSTITUTES

an important part of language proficiency.'
Like speaking, writingshows that the individual can use the language to communicate, "to
formulatesequences oflogical thoughtin intelligible and appropriate syntax."2Good writing
requires choices on three separate levels: the
word, the sentence, and the paragraph. At the
word level the lexical itemsare selectedfortheir
suitability,clarity,and nuance. At the sentence
level these words are combined into syntactic
patternsthatmove the message forward.At the
paragraph level the sentences are sorted and
organized into a particularsequence to present
informationin a coherent manner.3 Writing
proficiencyin a foreignlanguage involvesmasteryat each of these three levels. When classroom foreignlanguage learnerswritetheyneed
to demonstratetheirabilityto use a varietyof
syntacticoptions, attendto the accuracy oflanguage forms, and communicate their ideas
coherently.Frequentlythe firsttwo tasks compete withthe thirdratherthan complementit.
Although native language writers and their
classroom foreign language counterparts
employ the same writingprocess, i.e., planning, composing,and revising,thelattergroup
focuses on sentence-levelconcerns for grammatical accuracy rather than on the function
of the text as a part of the writingprocess.
REVIEW

OF

SCHOLARSHIP

The studies in the past ten years which discuss how classroom learners of German as a
foreignlanguage develop their proficiencyin
writingGerman fallinto two major categories:
1) those which treat grammatical concerns on

The ModernLanguageJournal, 70, ii (1986)

0026-7902/86/0002/114
$1.50/0

?1986 The ModernLanguageJournal

the sentence level; 2) those which assess the
impact of the whole text.
Investigations of sentence-level grammar
focus on: 1) the accuracy of linguistic forms,
i.e., the typesof errorsstudentsmake; 2) syntactic proficiency,i.e., the ability to employ
both simple and complex constructionsin written sentences. LoCoco's analysis of errors in
writtenGerman and Spanish attributesthetype
of error to universals in language acquisition
or to language-specific constraints.4Lalande
presentsstrategiesforreducing the number of
errorsin writtencompositions of intermediate
level students.He advocates total correctionof
errors by studentsand instructionalfeedback
to make student writersaware of their recurring errors.5 Cooper's research assesses syntactic proficiencyat the sentence level identifyingwhen particularsyntacticpatternsare acquired.6 Cooper later presents instructional
materials employingsentencecombiningexercises to help writersacquire complex syntactic
structuresquickly.'

Studies of the whole text explore how ideas
are communicated withrespectto the purpose
of the piece of writing.What the writerhas to
say about a topic and what he assumes the
backgroundknowledgeofhis readerto be characterizes the informationalstructureof a text.
Walker'sapproach to improvedwritingin German is to have his students analyze the texts
of shortstoriesand other prose writingto discover the authors' syntacticand lexical strategies for presenting information.8 Kramsch
demonstrateshow studentswritingin German
fail to use the variety of syntacticstructures
otherthan the grammaticalsubjectto build discourse topics. She concludes that a "sensitization to the notion of discourse topic" needs to
be integratedinto the presentationof German
word orderrules.9Byrnesalso stressesthe need
forwhole textapproaches in evaluatingwriting
proficiency.10She advocates attentionto units

This content downloaded from 46.243.173.175 on Sat, 28 Jun 2014 12:27:11 PM
All use subject to JSTOR Terms and Conditions


Syntaxand Communicative
Strategies

115

oflanguage witha broader informationalrange
than the sentence, to the purpose of different
types of texts,and to the consequences of particular lexical and syntacticchoices. Attention
to the communicativefunctionofa textcreates
an awareness of language beyond mere grammatical accuracy.
PURPOSE

The purpose of this study is two-fold: 1) to
identifythose syntacticstructureswhich effectively communicate ideas in writtenGerman
compositions of fourth-semesterstudents and
2) to identifygrammatical structures which
writers can control on a syntactic level, but
which theydo not employ as a communicative
strategy.The varietyof syntacticstructuresand
their frequencyof occurrence in each of three
textsis presentedin a syntacticprofileforeach
student writer. The informationstructurein
selected sequences of sentences from three
pieces of writing for each writer is analyzed
within the frameworkof a discourse analysis
approach. Following the flowof informationin
sentences within a paragraph allows one to
identifywhich syntacticstructuressuccessfully
establish and maintain coherence throughout
the paragraph. Because the subjects in this
study are intermediate-levelstudents of German, the data also reveal to what extent particular syntactic choices fail to communicate
effectively.The problem areas which emerge
fromthe analysis of the informationstructures
suggest the need for pedagogical strategies
which coordinate the study of specific grammatical structureswithcommunicativegoals in
writing instruction.
CONCEPTUAL

BACKGROUND:

DISCOURSE

ANALYSIS

Discourse analysisdescribeshow participants
in spoken or writtenexchanges constructand
presentinformation.The syntacticstructureof
sentencesin discourse provides the mechanism
to encode specificlexical choices and select appropriate word orders. One can argue with
Givon that an explanation of surfacesyntactic
phenomena includes their relative ability to
communicate information." Givon demonstratesthat syntacticcomplexitycorrelateswith
presuppositional complexityand that the high
frequency of the neutral (main, declarative,
affirmative,active) clause type is related to the

low degree of presuppositional complexity
found in such clauses. This suggests that successfulwrittenand oral discourseis not a matter
of so-called "syntacticmaturity"but the ability
to build logically on presupposed information
while conveying new content.'2
Informationis commonly divided into that
which is known or shared by the participants
(topic/theme)and that informationwhich the
speaker/writer
presentsas new, i.e., whichsubsequently becomes part of the shared pool
(comment/rheme).'3The topic/theme,generally stated at the beginning of utterances, establishes a frame; it states what the sentence
is about. The comment/rheme
followsand conflow
veys new content, creating a left-to-right
of information.In his study of textual analysis, Werthexamines the motivationforthe leftto-rightflow of informationin discourse and

the functionofanaphora as an indicatoroftext
coherence.'4 Anaphora permits concepts to
remain active throughoutthe discourse by reto informationin previous
peating or referring
utterances or by emphasizing one member of
a set of semanticallyrelated concepts fromthe
common pool of informationin order to provide contrast.Anaphoric material servesas the
locus of accumulated informationup to a particular point in the discourse and as a point of
departureto which new (non-anaphoric) information is added.
Kramsch uses the theme/rhemedistinction
to demonstrate that students inappropriately
use subject nouns and pronouns as themes.'5
This syntacticstrategypreventsthese inexperienced writers from effectivelyinterlocking
themes and rhemes in textsentencesto sustain
a topic of discourse over several utterances.
The accurate identificationof surface structures which carry new informationpresents a
major difficultyin Werth'sand Kramsch's approaches. Neither method provides adequate
explanation fordeterminingthe scope/rangeof
the elementsin the right-mostpositionin a text
sentence,the positionoccupied by the non-anaphoric materialor rheme.16One can inferfrom
Givon's discussion of the main declarative sentence that this type presents the bulk of new
information in discourse. The subject-verbobject word order, characteristicof this clause
type, may suggestwhich surface grammatical
categories functionas the locus of new information in text sentences."

This content downloaded from 46.243.173.175 on Sat, 28 Jun 2014 12:27:11 PM
All use subject to JSTOR Terms and Conditions


116

VirginiaM. Coombs

This studyexploresthe relationshipbetween
surfacesyntacticstructuresand theircommunicative functions.My analysisproceedsfromthe
assumption thatwritersemploy a communicative strategyin which new informationfollows
the informationwhichthe participantsshare up
to a particularpoint in the discourse. Identifying those surface syntactic structureswhich
make effectiveuse of anaphoric expressions as
well as those instances in which a strategyis
not selected by the writerreveals the extentto
which informationhas been successfullycommunicated.
METHODOLOGY

Data fromsix writingsamples are analyzed
below to show: 1) variation and frequencyof
surfacesyntacticstructures;2) the information
structurein textsentences. Relating these two
types of evidence to each other demonstrates
which syntacticstructurescan functionas effective communication strategiesin writtenGerman textsproduced by intermediate-levelstudents.
The surfacesyntacticstrucStructure.
Syntactic
tures found in each of threecomplete pieces of
writingfortwo studentsis presented in a syntactic profile. The profileemerges froma descriptionof seven featuresand theirfrequency
of occurrencerepresentedin percentages. The
simple sentence (1) refersto the neutral sentence type as defined by Givon, the main,
declarative, affirmative, active clause. All
clause types which fit this description are
counted separately; the writer'suse of coordinate conjunctionstojoin neutralsentencetypes
is ignored. The complex sentence (2) consists

of a main clause plus subordinate clause, subclausal unit, or an infinitivephrase. Because
the clause order in complex sentences affects
the informationstructureof textsentences,the
sequence of clauses is also addressed (2 and 3).
The descriptionof surfacesyntacticstructures
also considersthe grammaticalpropertiesofthe
initialelementin simple sentencesand in main
clauses in complex sentences. Since German
word orderconstraintsadmit both subjectsand
non-subjects in this position, the presence of
either option is accounted for (4 and 5). The
presence of the copula sein as main verb (6)
limitsthe syntacticoptions in the remainderof
the sentence and affectsthe transmissionof informationacross sentenceboundaries. Because

the passive voice constructionstands in a particularsemanticand syntacticrelationto active
voice constructions,its discourse functionsare
also considered(7). The percentagesexpressed
forfeatures1-3 are determinedusing the ratio
of clause types to total sentences in the compositions. The numerical values for features
4-7 relate the specificgrammaticalcategoryas
a percentageofthe totalnumberofoccurrences
for the sentence/clausetype.
The contentof each indiStructure.
Information
vidual sentencein the data is classifiedintothat
informationwhich has been established up to
a particular point in the discourse and that
which is new. Following the principle of communicativedynamism,the establishedinformation appears in the left-mostposition in an
utterance, while the new informationfollows

it in the right-mostslot.18 The information
structureof discourse initial sentences varies
slightly;the left-mostpositioncontains contextual informationwhich helps the writerestablish the discoursetopic. The subject-verb-object
word order of the neutral clause type offersa
syntacticstrategyformatchingthe information
structureto the surfacesyntaxof the discourse
sentences.In simple sentencesthe shared information segment corresponds to the sentenceinitial unit. Grammatical structures which
carrythisinformationinclude subjects,objects,
and adverbial phrases which establish spatial,
temporal,or causal frames.Grammaticalstructures in the predicate including objects, adjectives,and spatial, temporal,or causal adverbial
phrases conveythe new information.When the
grammatical subject does not occupy the sentence initial position, it can also carrynew information.
Particular lexical strategiesare selected for
their ability to maintain the flow of information in discourse. Shared informationin the
left-mostsegmentof the sentence is located by
identifyinginstancesof repetition,pronominal
reference, and paraphrase, and then traced
backwards throughpreceding sentences to establish the antecedentlinks. Clause sequences
in complex sentences determine the informational structure.The procedure used to assess
the informationalstructurein simple sentences
is also applied to main clauses in sentenceinitial
position. New informationis frequentlylocated
in the subordinate clause or infinitivephrase
when either of these functionas the object of

This content downloaded from 46.243.173.175 on Sat, 28 Jun 2014 12:27:11 PM
All use subject to JSTOR Terms and Conditions


Strategies

Syntaxand Communicative
the main clause verb. When the subordinate
clause or infinitivephrase introducesthe complex sentence,the informationalstructureis described in terms of its presuppositional complexity and/or functionin establishing a contextual frame for the utterance.
DATA

PRESENTATION

The writingsamples are drawn fromthree
separate assignments completed between the
second and eighth weeks in a fourth-semester
college-levelGerman course. Each writercompleted three draftsfor each assignment. First
and second draftsreceivedungraded writtenresponses frompeer editorsand fromthe instructor. A lettergrade evaluation givingequal considerationto language accuracy and mattersof
organization and content was assigned to the
third draft. Prior to each rewriteopportunity
studentswere encouraged to rethinkthe presentation of theirideas as well as correcterrorsin
grammar and vocabulary. Due to limited
space, only data fromthe graded draftsare presented here. In the present analysis the data
sentencesare numbered consecutivelyforeach
writer. The samples convey the writers'language complete withgrammaticalinaccuracies
and incorrectvocabulary.
Essay topics relateto assigned materialin the
19All
principaltextforthe courseZurDiskussion.
students wrote on the same topic for the first
two assignmentsbut were offereda choice from
two suggested topics for the third essay. The
topics were stated as questions, a formatwhich
would ideally elicitwrittenresponses using the
informationfrom the readings, class discussions, and personal knowledge.20
CASE

1: DOUG

SyntacticProfile. An even distribution of
simple and complex sentences characterizes
Doug's syntacticstyle(see Table I). The simple
sentences are frequentlyjoined by coordinate
conjunctionscreatingsentencesequal in length
to many of the complex structures.When complex sentence structuresare selected, the main
clause of that sentence occurs overwhelmingly
in the sentence initial position. Subordinate
clauses and infinitivephrases preceding the
main clause in complex sentences occur only
five times in all three samples.
Just as a preferredorder of clauses in com-

117
plex sentencescharacterizesa writer'sstyle,so
does the position of grammatical units within
clauses and simple sentences.Althoughthe first
writingsample shows a balance in the use of
the grammatical subject as well as non-subject
material in initial position in both simple and
complex sentences, the initial syntacticslot in
both simple and complex sentences is dominated by the grammaticalsubject in the second
and thirdsamples. The strongpresence of the
copula sein in the firsttwo samples contrasts
with minimal use in the third essay. Passive
voice structuresoccur predominantlyin simple
sentences.

The firstparagraph in
Structure.
Information
a piece of writtendiscourse establishesthe discourse topic. The structureof the shared or
background informationis expressed by subject nouns in sentences 4, 5, and 6 (see Appendix 1). New informationis introduced by
the accusative object (2), a predicate nominative (4), predicate adjectives (4, 5), and a compound adverbial expression (6). The topic of
pollution is introduced and set in a temporal
frame in sentence one.
The second and third sentences carry this
topic forward across sentence boundaries by
means of the logical connectorsdarumand aus
diesemGrund.Anaphoric expressionsin the leftmost position in sentences 4-7 forman interlockingchain of information.Examples ofrepetition include 4 Gesetze (2 Gesetze) and 7
LewisburgundBucknell(6 hierin Lewisburgundan
Bucknell).The anaphoric material in the grammatical subject in 6 (Ein TellvondieserAktionen)
demonstrates the contrastive feature of anaphora, since the noun Teil singles out by
example one member of the set referredto in
5 (Aktionen
vonLeutenundGesellschaften).
The subject noun phrase in 5 is a paraphrase ofthenew
in dermenschinformationin 4 (eine Veranderung
lichen
Mentalitiit).
Simple sentencesin the active voice alternate
with agentless passive voice constructionsto
convey informationin sentences8-16. Subject
nouns and pronouns occupy theleft-mostposition in sentences 11, 12, 13, 14, 15, and 16.
The presenceofthe copula as main verb in several of these sentenceslocates the new information in the predicatenominative(11), predicate
adjectives (10, 16), and adverb (8). The informational content expressed by these surfacelevel grammatical categories cannot be easily

This content downloaded from 46.243.173.175 on Sat, 28 Jun 2014 12:27:11 PM

All use subject to JSTOR Terms and Conditions


118

VirginiaM. Coombs

TABLE I
SyntacticProfile:Doug

TABLE II
SyntacticProfile:David
Essay

1. Simple sentence
2. Complex sentence
Main clause + ----:
3. Complex sentence
----+ Main clause
4. Subject-initiated
a) Simple sentence
b) Main clause
(complexS)
5. Non-subjectinitiated
a) Simple sentence
b) Main clause
(complexS)

Essay

Essay

1 (%)

2 (%)

3 (%)

37

49

53

57

48

45

3

1

6
77

82

9

35

79

88

62

18

1

45

21

6

6. Sein

a) Simple sentence
b) Main clause
(complexS)
7. Passive voice
a) Simple sentence
b) Main clause
(complexS)

54

39

15

30

33

18

23

18

20

5

6

18

reintroducedin subsequent anaphoric expressions.
Sentence 18 advances the topic, sexual explicitnesson television,by referringto particular programs to support the claim made in the
finalsentenceofthe precedingparagraph, sentence 17. The informationin sentences 17-20
is conveyedpredominantlythroughsimple sentence structures,all beginning with the grammatical subject. Since the left-mostposition in
a simple sentence should form the locus of
known or shared informationrelevant to the
topic of discourse, these subjects ought to provide the coherentlinks which sustain information throughoutthe paragraph. As such they

should all be characterizedby some sortof anaphoricfeature.Such connectionsare completely
missing between the object clause in 18 (da?f
jeder . . springt)and the subject nouns in 19
and 20 (Erwachsene,Kinder).
CASE

2: DAVID

Profile.David employs both simple
Syntactic
and complex sentences in equal numbers in
each of the three samples (see Table II). The
main clause introduces complex sentences in
all but one instance in the firsttwo samples.

Essay

1. Simple sentence
2. Complex sentence

Main clause + ----:

3. Complex sentence

Essay

2(%)

3(%)

43

40

53

54

60

28

4

-

19

50

50

82

60

73

44

50

50

18

7

13

22

8

30

12

7

20

33

-

10

11

-

-

-

----+ Main clause

4. Subject-initiated
a) Simplesentence
b) Main clause
(complexS)
5. Non-subjectinitiated
a) Simple sentence
b) Main clause
(complexS)

Essay

1(%)

6. Sein

a) Simple sentence
b) Main clause
(complexS)
7. Passive voice
a) Simple sentence
b) Main clause
(complex S)

In the third sample subordinate clauses or infinitivephrases precede the main clause about
as often as they follow.
The grammaticalsubject in simple sentences
appears equally as oftenin non-initialposition
in the firsttwo samples as it does in initialposition. The thirdsample shows a clear preference
forthe grammatical subject in sentence initial
position. In complex sentencesthe grammatical subject in initial position is preferredin all
threesamples. The copula as main verb is used
sparinglythroughout.The passive voice constructionappears only twice.
Structure.David uses different
Information
types of contrastto shape the presentation of
information(see Appendix 2). The repository
of shared informationis expressed by a subject
pronoun (3), by spatial adverbial phrases (2,
5), and in one instance by the grammatical
object ofthe esgibtconstruction(4). New information is packaged in the direct object (5),
prepositional objects (2, 3, 4), and a predicative nominative (1).
Anaphoric material in the spatial reference
foundin 2 (Hier an Bucknell)qualifiesmore narrowlythe genitivemodificationin 1 (desAlltags).
Withinthisestablishedspatial framethe reader

This content downloaded from 46.243.173.175 on Sat, 28 Jun 2014 12:27:11 PM
All use subject to JSTOR Terms and Conditions


Syntaxand Communicative
Strategies
is presentedwithtwo contrastingpieces of new

informationin 2 (nichtaufdieUmwelt)and 3 (nur
an sichselbst).The contrastiveanaphoric expression in 4 (mehrere
Leute) can be traced back to
the pronoun referencein 3 (sie) and its antecedent in 2 (mancheStudenten.).
Contrast is also
the
inclusion
of
the
lexical item
signaled by
in
4 (an die
The
jedoch.
prepositional object
new
offers
because
cominformation,
Umwelt)
municativelyit differsfromthat already introduced in 3 (nuran sichselbst).The spatial frame
established in 4 (hier) is maintained by repetition in the left-mostpositionof 5 (in Lewisburg).
The new informationin 5 is not found in the
object (die Umwelt),which has been established
as part of the context already in the firstsentence, but in the verbal activity(pflegen).
Of the five sentences (6-10) which David
groups togetheras a paragraph, only the first
two illustraterelevant discourse. The opinion
marker in 6 (ich bin derMeinung)establishes a

context for the presentation of the topic- the
threat of (nuclear) war. The content of the
opinion is sustained in the followingsentence
by new informationwhich substantiates the
claim (Rohstoffe,
Raum, Geld,fir die. . kdmpfen
werden).Each of the remainingsentencesintroduces a new subtopic (8 dieseDrohung,9 Schutzand 10 Frieden),but thereare no coherent
mittel,
elements,because the left-mostpositionof each
sentence contains neither anaphoric material
nor intersententiallogical connectors. The use
of doch in 13, normally a lexical signal for
contrast, is unjustified.
The parallel syntacticstructuresin 11 and
12 establish conditional frames in which the
new informationis to be understood. Both conditions presuppose that the reader associates
positiveresultswiththe conditionin 11 and less
favorable resultsas a resultof the condition in
12. The informationin sentence 13 contributes
nothingnew to the topic, because it merelyrepeats what the presupposition in 11 has established. If the reader accepts the presupposition
in 14, that unhealthy food and medications
exist (ungesunden
Lebensmitteln
undMedikamenten)
and that one must protect oneself fromthem
(uns . . . schiitzen)then this infinitivephrase
refersto the same context established in the
conditional clause in 12. Pillenschlucken
(12) is
the antecedent forungesunden

Medikamenten
(14)
and nichtrichtigessen(12) the antecedent for
Lebensmitteln
ungesunden
(14).

119
RESULTS

Case 1: Doug. AlthoughDoug's syntacticprofile shows that he uses a variety of syntactic
structureschoosing between simple sentences
and complex sentences,perhaps the most striking featurein his writingis the frequencywith
which the grammaticalsubject occurs in initial
positionin both simple and complex sentences.
Although the grammaticalsubject can serve as
the locus of shared informationto advance the
information chain forward, problems arise
when new informationin the predicateposition
is not integratedinto the pool of shared knowledge throughreiterationin the followingsubject. This subject-initialword order excludes
the effectiveuse of logical markersto connect
the semantic content of two propositions.
To illustrate the reduced communicative
effectiveness
which overuse ofthe grammatical
in
left-mostpositioncauses, one can
the
subject
consider how effectivelyinformationis transmitted in the sentences 1-4 without the presence of the grammaticalsubject in theleft-most

position. The effectiveuse of logical markers
darumand aus diesemGrundconnects ideas between sentences and sustains the topic of discourse over three sentences.
In contrastto these examples, the sequence
of sentences 8-16 lacks the connective devices
or spatial and temporalreferencesto maintaina
contextforproductivediscourse. Failing to use
sententialconnectorsand subordinationin the
appropriate contexts creates communicative
roadblocksforthe reader. Furthermore,monotonous sentence structurecaused by loyaltyto
the grammatical subject in initial position results in missed opportunitiesto use the nonsubject elements to cement semantic relationships in sentences 17-20. If the demonstrative
das in 19 were frontedto sentence initial position, it would provide the link to the information in the object clause in 18. Frontingthe dacompound in 20 would establish a similar link
with the object clause in 18.
An additional problem which surfaces in
relation to the syntacticstructurein sentences
8-16 is caused by the alternationof copula sein
as main verb in 8, 10, 11, and 16 with passive
voice constructionsin 9 and 14. In the subjectinitiated sentences containing the copula, the
new informationis packaged as predicatenoun
(11), adjective (16), or adverb (8). The new in-

This content downloaded from 46.243.173.175 on Sat, 28 Jun 2014 12:27:11 PM
All use subject to JSTOR Terms and Conditions


VirginiaM. Coombs

120
formationin the subject-initiatedpassive voice
sentences appears in the past participle which
functions as a predicate adjective. Subject

nouns and pronouns in subsequent sentences
do not referto the informationfromthe antecedent adjectiveor past participle.Two or more
of these types of syntactic constructions in
sequence hinder the flowof informationacross
sentence boundaries.
Case2: David. David's syntacticprofilereveals
his ability to constructboth simple and complex sentences correctlyand to vary syntactic
structurewithin sentences. When these facts
are scrutinizedin termsof effectivecommunicative strategies at least three problem areas
appear. The firstrelates to the position of the
thematic or shared material. Examples from
David's firstsample illustratewhat typesof informationprocessingdifficulties
arise when the
thematicmaterial does not appear in sentence
initial position. The surface accusative object
of the es gibt(4) constructioncan functionin
information structure as either the locus of
shared informationor new information.In the
formercase the syntacticconstructionis clumsy,
oftenrequiringa relativeclause to statethe new
information.Alternatively,a straightforward
simple sentence in which the accusative object
of es gibt becomes the grammatical subject
streamlines the information process by not
pushing the thematic anaphoric information
any furtherto the right than necessary. The
contrastin 4 failsto achieve its fullimpact due
to the intrusionof the esgibtconstruction.The
lexical jedochwould carry more informationif
repositioned furtherleft in the sentence.

The second problem area can be illustrated
in sentences 6-10. Communication problems
arise when the reader does not have access to
the same pool of shared or presupposed information as the writer.The adverbs allerdings
(8)
and doch(10) in sentence initialposition might
be regarded as examples of what Kramsch has
termedthe"randomuse ofunjustifiedelements"
in sentence initial position.21An explanation
forthe surfaceappearance of these"unjustified"
adverbs emerges fromthe bits of information
available to the reader in the noun phrase diese
Drohung(8). The demonstrativemodifierseems
to indicate that the writerconsiders this concept partofthe sharedinformationon thetopic.
Possible antecedent expressions include the
three accusative objects in 7 Rohstoffe,
Raum,

and Geldand the relative clause modification.
What is missing that would justifysentence 8
in its presentformis a clear connection establishingthe items listed in 7 as the components
of the referenceto a threatin 8. That informational link is not immediatelyaccessible to the
reader even though it seems to exist for the
writer.A similar link is missing in the surface
structurebetween 9 and 10, so that the use of
doch appears equally out of place. Linda
Flower's workon writers'composing strategies
refersto this kind of phenomenon as "writerbased prose."22 The reader is excluded from
participationbecause all parts of the information structureare not fullyspecified.
The complex nature ofpresuppositionlevels

associated withdifferent
surfacesyntacticstructures presents a third problem area. The sequence of sentences11-13 illustratesthatDavid
fails to understand how presupposition functions communicatively.The less complex presuppositional structurein sentence 13 presents
the problem. The reader comes to thisstraightforwarddeclarative assertion afterprocessing
the presupposition involved with the two preceding conditions. What should be processed
as new informationin 13 has already contributed to the presuppositional structurein
sentences 11 and 12. Reordering the sequence
so that 13 introducesthe paragraph would establish a context for the conditional in 11.
DISCUSSION

This studyshows thatclassroom foreignlanguage learners are able to control the grammatical aspects of syntactic structures, i.e.,
word order and agreement,beforetheycan use
these structureseffectivelyin discourse situations. The writingsamples contain a varietyof
clause structuresand infinitivephrases in addition to simple sentences which contain relativelyfewgrammaticalinaccuracies. The data
also suggestthatwriters'skillin using language
structuresto communicate effectively
develops
independentlyof theirknowledge and control
of grammarrules. For example, one mightexpect the successfulstrategiesthat both writers
employed in the firstassignment to be used
again in subsequent assignments,but the data
do not supportthisin spiteofregularclassroom
instructionin writing. Students not only received writing instructionin class and individual feedback fromthe instructoron drafts,

This content downloaded from 46.243.173.175 on Sat, 28 Jun 2014 12:27:11 PM
All use subject to JSTOR Terms and Conditions


Syntaxand Communicative
Strategies

but they also acted as readers foreach other's
intermediatedrafts.The students,workingin
pairs, completed a set of editorial tasks which
focused on topic organization at the paragraph
level and on the syntacticstructureof sentences
within the paragraph.
The syntacticstructuresin writtenGerman
at the intermediatelevel stillappear to remain
closely tied to memorized patterns fromtextbook grammar explanations. The frequentuse
of subject-initiated sentences to illustrate a
variety of grammatical points maycontribute
to the frequency with which this pattern appears in studentcompositions.23Intermediatelevel texts present the topic of subordination
relativeto word orderrules and the clause-final
position of the finiteverb ratherthan as a communicative strategy.Subordinate conjunctions
appear primarily as vocabulary items. Since
logical markers and text structuringphrases
belong to no single grammaticalcategory,they
do not receive systematictreatmentin grammar texts.24
At least three examples illustrate that the
writers' ability to select particular syntactic
variants fortheircommunicative functionlags
farbehind theirabilityto manipulate language
forms.First, syntacticstrategiesforemploying
eithercoordinate or subordinate structuresare
not used consistentlyor effectively.
The coordinate conjunctions undand aberfrequentlyjoin
simple sentences,not to establish a logical connection between the informationalcontent of
the two clauses, but to create the effectof a
"longer" sentence. Appropriate subordinate
structureswhich show relationships between

the ideas in adjacent sentences are missing despite the presence of several examples of subordination in each sample. Second, the infrequent use of logical markersand text structuring devices forthe readers' benefitreduces, if
not totallyobscures, the communicativeintent
in many of the data sentences. Third, the syntactic structurein simple sentences and main
clauses of complex sentencesreveals the dominant presence of the grammatical subject in
sentence initial position. This text feature,
which is described in detail in Kramsch's study
on topic construction,restrictsthe transmission
of informationto one or two sentences.
PEDAGOGICAL

IMPLICATIONS

Language instruction at the intermediate

121
level must take discourse level into account.
Teachers oftenperceive poorly writtenessays
as resultingfromstudents'failureto use thelanguage in a communicative manner, although
sentence grammar is essentially correct. To
remedythisproblem the grammar component
at the intermediatelevel must emphasize language structuresas a means to communicate
and not simplyas a collectionof rule-governed
patterns.
Data in this studypoint to three specificinstances in which grammar topics can be integrated with writingassignments. First, a new
perspectiveon the functionofword order rules
in German emerges from an approach that
stressescommunicative effectrather than the
mechanical manipulation of sentence components. The simple sentence remains a powerfultool forcommunicationwhen syntacticpatternsare utilized withregard forinformational
content. Kramsch describes a variety of oral
and writtenexercises designed to focus on the

communicative function of non-subject elementsin sentenceinitialposition.25Second, the
review of subordinating conjunctions progresses beyond the introductorylevel that
stressesthe manipulation of the finiteverb and
the memorization of lexical definitions.Exercises that focus on the semantic contentof the
subordinatestructureand itslogical connection
to the main clause emphasize the importance
of the intendedmessage ratherthan the clausefinalposition ofthe finiteverb. The lattertype
of exercisesincludes the traditionaldehydrated
sentence exercise which requires students to
construct clauses from elements provided in
order to test conjunctions as vocabulary items
and to reinforcethe word order rules in subordinate clauses. Exercises which develop an
awareness of the logical relationshipsbetween
ideas are constructedfromsequences of topically related sentences. When some ofthe sentential connections are removed, but not the
finiteverb form,the exercisechallengesthe student to reconstructa message ratherthan move
formsaround withina clause. The reconstructed
message emerges fromthe student'sinterpretation of the logical relationships which the
addition of a particularconjunctionestablishes.
The point of such exercisesis thatdifferent
conjunctions create differentlogical relationships
which resultin different
messages, even though
all subordinate conjunctions are governed by

This content downloaded from 46.243.173.175 on Sat, 28 Jun 2014 12:27:11 PM
All use subject to JSTOR Terms and Conditions


122

VirginiaM. Coombs

the same word order constraint.Third, logical
markersand textstructuringphrases are functional propertiesof texts.Contextualized exercises based on model texts increase students'
awareness of the communicativevalue of such
language forms.
Composition assignmentsmustclarifyforthe
students that text purpose is as important as
linguisticaccuracy. Since studentsat the intermediate level are stillgrapplingwithlanguage
formsas well as theirfunctions,theyshould be
encouraged to rewriteand revise their work.
They rarely get everything"right"in the first
draft.Most studentsunderstand"rewrite"and
"revise"to mean correctinggrammar mistakes
and vocabulary inaccuracies. They need some
practical strategiesforreviewingthe information structurein textsso thattheycan see where
the informationalconnectionshave not yetbeen
made. One such strategyfocuses on the structureofinformationat theparagraphlevel. Each
paragraph is reconstructedas a list of the sentences it contains. In this formatit is easier to
identifyvisually grammatical structuressuch
as subjectsand objects. The studentthentraces
the path of the paragraph topic fromone sen-

tenceto thenext,notingwhichstructuresmaintain the topic and which introducethe new information.The ideal patternfollowsthe principle of communicativedynamismwithleft-torightmovementin each sentence. At the point
thatthe informationchain failsto move forward
the studenthas identifieda problem sentence.
Each paragraphcan be analyzed in thisfashion.
Not only will thisproceduremake the information structuremore visible, but it also justifies
paragraph divisions as new topics are introduced in the text. Since most compositions
rarelyexceed 250 words in length,this strategy

can be applied quickly and effectively.
Successful writing exercises at the intermediate level address the functionalnature of
language structuresto emphasize yet another
way that language learning involves learning
to communicate effectively.For those students
who continuetheirlanguage studythisemphasis provides a foundationfordeveloping writing proficiency in traditional composition
courses and provides a more realisticbasis for
the kindsofwritingrequiredin advanced literature and culture courses.

1
APPENDIX
Information Structure: Case 1

1. In den letzten Jahren wird viel fiber die Verpestung
und die Verschmutzung, die unsere Welt bedrohen,
gesprochen.
2. Darum gibtes fiirdie Industrienund fiirAutos Gesetze,
die gegen das Problem kiimpfensollen.
3. Aus diesem Grund sind unsere Landschaft und unsere
Luft jetzt sauberer und sicherer als friiher,
4. aber Gesetze sind selbst keine Ldsung des Problems,
dag eine Veriinderungin der menschlichen Mentalitdit
braucht.
5. Aktionen von Leuten und Gesellschaften sind sehr
wichtig, um die Natur unaufh6rlich zu beschfitzen.
6. Ein Teil von dieser Aktionen ist hier in Lewisburg und
an Bucknell,
7. und obwohl Lewisburg und Bucknell eine kleinenNachbarschaft in der Weltgemeinde sind, hier findetman
einige kleinen, aber wichtigeBeispiele von dem Schutz
der Umwelt.

8. Die Einflufite
der Drohung auf unser Leben sind fiberall,

9. aber sie werden nicht immer gesehen.
10. Ffir die Wirtschaftsind die Kosten sehr hoch.
11. Waffenim allgemein sind nichtwie andere Wirtschaftwaren.
12. Unsere Wirtschafthandelt sich um das Verhiltnis der
Firmen und der Verbraucher;
13. aber die Waffensproduktiongibt nichts dazu.
14. Arbeitsplitze werden durch diese Produktion
geschaffen,
15. aber diese Waren haben in der Wirtschaftkeine richtige
Benutzung;
16. sie sind endlich nicht verbraucht (Gottseidank).
17. Diese Leute und ihre Aktionen scheinen Kindern,
typisch von unserer Gesellschaft zu sein.
18. Programme wie "Dallas" und "Dynasty" zeigen, dag
jeder mit irgendjemand auger seiner Frau oder ihrem
Mann ins Bett springt.
19. Erwachsene verstehen das als Unterhaltung,
20. aber Kinder k6nnen davon verwechselt werden.

This content downloaded from 46.243.173.175 on Sat, 28 Jun 2014 12:27:11 PM
All use subject to JSTOR Terms and Conditions


Strategies
Syntaxand Communicative

123

APPENDIX 2
Structure:Case 2
Information

1. Die Umwelt, in der die Menschen leben, ist ein
bedeutender Teil des Alltags.
2. Hier an Bucknell achten manche Studenten nicht auf
die Umwelt;
3. sie denken nur an sich selbst.
4. Es gibtjedoch mehrere Leute, die hier an die Umwelt
denken,
5. und in Lewisburg pflegen die meisten Leute die
Umwelt.
6. Ich bin der Meinung, dab ein Krieg wieder passieren
wird, weil Menschen nattirlichselbstsiichtigsind.
7. Leute verlangen immer mehr Rohstoffe, Raum und
Geld, fuirdie sie kiimpfenwerden.
8. Allerdings beeinflust diese Drohung unser tdigliches
Leben.

NOTES

1I wish to thank my colleagues Professor Lois Huffines
and Catherine Blair who read earlier draftsof the manuscript and made helpful suggestions.
2See Claire Gaudiani, TeachingWritingin theFL Curriculum (Washington: Center for Applied Linguistics, 1981),

p. 43.

3Gaudiani (note 2 above), p. 1.

4See Veronica Gonzales-Mena LoCoco, "An Analysis of
Spanish and German Learners' Errors," Working
Papersin
Bilingualism,7 (1975), pp. 96-124.

5See John Lalande, "Reducing Composition Errors: An
Experiment," ModernLanguageJournal,66 (1982), p. 147.
6See Thomas C. Cooper, "Measuring WrittenSyntactic
Patternsof Second Language Learners of German,"Journal
of EducationalResearch,69 (1976), pp. 176-83.
7See Thomas C. Cooper, "Sentence Combining: An
Experiment in Teaching Writing,"ModernLanguageJournal,
65 (1981), pp. 158-65.
8See Ronald Walker, "Text Manipulation Techniques and
15
Foreign Language Composition," Unterrichtspraxis,
(1982), pp. 232-39.
9See Claire J. Kramsch, "Discourse Function of Grammar Rules: Topic Construction in German," ModernLanguageJournal,67 (1983), p. 18.
1oSee Heidi Byrnes, "Discourse Analysis and the Teaching of Writing," ADFL Bulletin,15, ii (1983), pp. 30-36.
"See Talmy Givon, On Understanding
Grammar(New
York: Academic Press, 1979), p. 49.
12Earlierstudiesof syntacticcomplexityequate the greater
number of syntacticallycomplex sentences with syntactic
maturity, i.e., a signal of greater facilitywith language

9. Wir brauchen Schutzmittel, um sicher zu sein.
10. Doch denken wir immer an Frieden - Frieden, den die
Welt noch nicht gesehen hat.
11. Wennjemand genug Schlaf und kbrperlichBewegung

bekommt, dann wird er gesund.
12. Aber wenn wir Pillen schluckenund nichtrichtigessen,
dann wird unser K6rper zerfallen.
13. Schlaf und Bewegung sind fiir unsere Gesundheit
wichtig wie richtiges Essen.
14. Um uns vor ungesunden Lebensmitteln und Medikamenten zu schiitzen, sollen wir die Etikette auf den
Lebensmitteln und Medikamenten lesen.

structure. See Kellogg W. Hunt, GrammaticalStructures
Written
at ThreeGradeLevels, Research Report 3 (Urbana:
NCTE, 1965). Recent criticismof this approach is found
in StephenJ. Gaies, "T-UnitAnalysisand Second Language
Research: Applications,Problems and Limitations,"TESOL
14 (1980), pp. 53-60, and Stephen P. Witte, "The
Quarterly,
Reliability of Mean T-Unit Length: Some Questions for
Research in WrittenComposition," Learningto Write:First
Language/Second
Language,ed. Aviva Freedman, Ian Pringle
&Janice Yalden (New York: Longman, 1983), pp. 171-77.
13The terms "topic-comment" are used by Charles
Hockett, A Coursein ModernLinguistics(New York: Macmillan, 1958), p. 201. The "theme-rheme"distinctionoriginates with the Prague School's Functional Sentence Perspective approach to informationstructure. See Paperson
FunctionalSentencePerspective,
ed. F. Danes (Prague: Academia/The Hague: Mouton, 1974).
and Emphasis(London:
14See Paul Werth,Focus, Coherence
Croom Held, 1984). Werth qualifies anaphoric material
with respect to discourse initial sentences. Since this sentence introduces the discourse topic, it representsthe one
exception to the principle of the anaphoric-non-anaphoric

sequence of information.
15Kramsch (note 9 above), p. 16.
16Both approaches segment the information structure
according to intonationcontourscharacteristicof spoken discourse. Neither approach establishes appropriate signals
for the onset of new information in written discourse.
17Wordorderconstraintsin German declarativesentences
do not require the grammatical subject to appear in sentence position. When it does occupy this position and is
notthelocus of new information,the new informationoccurs
in the right-mostposition in the sentence. Accusative and
dative objects, prepositional objects, adverbial phrases as
well as predicate nominatives, adjectives, and adverbs can
occupy this right-mostslot.

This content downloaded from 46.243.173.175 on Sat, 28 Jun 2014 12:27:11 PM
All use subject to JSTOR Terms and Conditions


124

VirginiaM. Coombs

18SeeJan Firbas,"Some Aspectsof the Czechoslovak
ApproachtoProblemsofFunctionalSentencePerspective,"
in Danes (note 13 above).
19SeeDieterSevin& IngridSevin,ZurDiskussion
(New
York: Harper & Row, 1979).
20Theexacttitlesoftheassignedessaytopicswere1) Sind
undBucknell
inderGestaltung

und
Lewisburg
fortschrittlich
genug
derUmwelt?
eines
Verschiinerung
2) WiewahristdieDrohung
im
Atomkrieges?
3) Welchen
EinflufhabenSex undBrutalitiit
Fernsehen
orSchlucken
wirzu vielePillen
aufunsere
Gesellschaft?
undessenwirzuvielFarbeundanderes
Gift?
21Kramsch(note 9 above), p. 16.
22SeeLinda Flower,"Writer-Based
Prose: A Cognitive
Basis forProblemsin Writing,"
41 (1979),
College
English,
pp. 19-37.

23See the syntacticstructureof sentencesin the grammar
explanations and accompanying exercises in intermediatelevel texts, e.g., Robert Helbling & Franz Kempf, Deutsche

Gegenwart
(New York: Holt, Rinehart, 1985); Jack Moeller,
Helmut Liedloff & Helen S. Lepke, Kaleidoskop(Boston:
Houghton Mifflin, 1983); Charles Barrack & Horst M.
Rabura, Mosaik. DeutscheGrammatik
(New York: Random
House, 1982).
24This study makes no attempt to account for other factors which may have an impact on writing proficiencyin
German such as writingstrategiesused in English, personal
interestin the topic, or motivationto improve writingskills.
25Kramsch (note 9 above), p. 21.

Conference on Language Development
THE

BOSTON

UNIVERSITY

SCHOOL

OF EDUCATION

invitespapers forthe eleventhAnnual Conference on Language Development, to be held on
17-19 October 1986. Papers on the topics of
firstand second language acquisition, bilingualism, language disorders,writing,literacy,
narratives,ASL and sign language, neurolinguistics, sociolinguistics,theoreticallanguage

acquisition, and universal grammar will be
considered forpresentationat the conference.

This year'skeynotespeakeris Noam Chomsky,
Massachusetts Institute of Technology. For
more informationwrite: Language Development Conference,School of Education, Boston
University, Boston, MA 02215.

Conference: Second Language Acquisition
and Foreign Language Learning
A CONFERENCE

ON THE

RELATIONSHIP

BETWEEN

second language acquisition and foreignlanguage learning will be held on the campus of
the Universityof Illinois, Urbana-Champaign,
3-4 April 1987. Papers are solicited that deal
with the relationship between SLA and FLL
vis-a-vispsycholinguistictheoryand research,
with special emphasis on: 1) input and interaction; 2) interlanguage data and developmental patternsin the acquisition of grammatical structureand general communicativecom-

petence; 3) processingstrategiesin comprehension and production; 4) the relationship between informationprocessingand language acquisition; 5) fossilization; 6) research design
and methods; 7) research agenda for the next
ten years. Send inquiries or three copies of a
one-page abstractto the conferenceorganizers
by 1 October: William VanPatten & James F.
Lee/SLA-FLL/Spanish, Italian & Portuguese/
4080 Foreign Languages Building/707 S.
Mathews/Urbana, IL 61801.

This content downloaded from 46.243.173.175 on Sat, 28 Jun 2014 12:27:11 PM
All use subject to JSTOR Terms and Conditions