A LONGITUDINAL STUDY OF THE USEOF SYNTAX AND TEXT FORMS IN CHILDREN'SNARRATIVE AND REPORT WRITINGByJENNIFER JUKES MARTINB. Ed., University of British Columbia, 1976A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENT FOR THE DEGREE OFMASTER OF ARTSinTHE FACULTY OF GRADUATE STUDIES(Department of Language Education)We accept this thesisas conforming to e re• uired standardThe University of British Columbia, October 1993© Martin, J., 1993In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature) Department oThe University of British ColumbiaVancouver, CanadaDate ^&a-aut. Jo?, /”..jDE-6 (2/88)ABSTRACTThe purpose of this longitudinal study was to examine the useof syntax and text forms in the narrative and report writing of aselect group of early intermediate grade children using a descriptivecase study approach.Writing samples were obtained during weekly writing sessionsconducted by the researcher as participant observer with a group of49 students in two and latterly, three classes. Using a writingprocess approach (Graves, 1984), the students alternated writingnarratives and reports from November to May in each of three years.While all students submitted their writing to the researcher forperusal and comment, only the writing of 14 randomly selectedstudents was used to provide the data for analyses of syntax and textforms. Additionally, a closer examination of the writing of a subsetof three writers selected by the researcher provided the opportunityfor detailed comparison of the three writers and the means tospeculate as to factors which may affect use of syntax and text forms.The analysis of data was twofold: an analysis of the complexityof syntax based on mean T-unit length and frequency counts ofextended noun phrases, subordinate clauses and verb phrases and ananalysis of text forms including the complexity of beginning andending statements based on a classification system designed byLanger (1986) and overall length of writing samples.Results indicated the students had more diverse and complexiiuse of syntax in their writing in successive years. Further, use ofthese forms varied in the two genres. Students had a wider varietyof more complex syntax in their narrative writing in Grades 4 and 5than in their report writing, however, by Grade 6, there were higherfrequencies of more complex forms of syntax in report writing thanin narrative writing. Beginning and ending statements were morecomplex in narrative writing than in report writing in each year.Also, overall length increased in both genres over time andnarratives were approximately three times longer than reports.Results from the profiles of three writers indicate there appear to beseveral factors which may affect the use of syntax and text forms inthe writing of these students. These factors include forms oflanguage, stylistic preferences, types of writing, maturation andmotivation.The results indicating observable differences in students' use ofsyntax and text forms in narrative and report writing haveimportant pedagogical and methodological implications and presentsome interesting directions for further study.iiiTABLE OF CONTENTSPageABSTRACT ^iiTABLE OF CONTENTS ^ivLIST OF TABLES ^xLIST OF FIGURES^ xiiAKNOWLEDGEMENTS xiiiCHAPTER IINTRODUCTION ^1Background of the Study ^1Syntactic complexity ^3Text Forms in Writing ^4Research Question ^5Definition of Terms ^6Significance of the Study ^9Limitations of the Study ^11CHAPTER IIREVIEW OF THE LITERATURE^ 13Syntactic Complexity ^13Basic Constituent of Analysis - The T-unit^14Syntactic Structures Within the T-Unit ^24Extended Noun Phrasesand Nominal Clauses ^24The Verb Phrase ^32ivPageDevelopment of Writing in Various Genres ^37Knowledge of Text Forms ^38Narrative ^38Other Modes of Discourse ^42Length of Discourse ^57Summary and Considerations of thePresent Study ^63CHAPTER IIIDESIGN OF THE STUDY^ 68Subjects of the Study ^68Procedures of the Study ^69Initiating the Study ^69Conducting the Study ^70Analysis of Data ^73Syntactic Analysis ^73T-Unit Length ^74The Noun Phrase ^74Subject and Complement Clauses ^75The Verb Phrase ^75Analysis of Text Forms ^76Beginning and Ending Statements ^76Length of Writing Sample ^77Interrater Reliability ^77CHAPTER IVRESULTS OF THE STUDY^ 80Syntactic Complexity ^80T-Unit Length ^80Narrative ^81Report ^85The Extended Noun Phrase ^91vPageNoun Phrase with Modifiers ^ 91Narrative 91Report^ 96Noun Phrase with Prepositional Phrase ^ 97Narrative 97Report^ 100Noun Phrase with Appositive Phrase 102Narrative and Report ^ 102Noun Phrase with Relative Clause ^ 103Stage I - Narrative ^ 105Stage I - Report 107Stage II - Narrative ^ 109Stage II - Report 110Stage III - Narrative and Report 110Stage IV - Narrative and Report ^ 110Subject and Complement Clauses ^ 111Subject Nominal Clauses 112Narrative and Report ^ 112Complement Nominal Clauses.^• • 112Narrative and Report ^ 112The Verb Phrase ^ 114Modal Auxiliaries ^ 115Narrative 115Report 118Do Auxiliary ^ 119Narrative 119Report 120Catenative Verb Phrase ^ 121Narrative ^ 121Report 123Simple Verb Phrase 124Narrative ^ 125Report 128Progressive Verb Phrase ^ 128Narrative ^ 130Report 132Passive Verb Phrase 132Narrative ^ 133Report 135viPagePerfect Verb Phase ^ 136Narrative 137Report 137Perfect Progressive/Perfect PassiveVerb Phrases ^ 139Narrative and Report ^ 139Percentage of Total Verb Phrases ^ 139Modals^ 140Catenative Verb Phrase 144Simple Verb Phrase ^ 144Progressive Verb Phrase 144Passive Verb Phrase 145Perfect Verb Phrase ^ 145Other Verb Phrases 145Text Forms in Writing ^ 146Beginning and Ending Statements ^147Narrative Beginnings ^ 147Report Beginnings 151Narrative Endings 155Report Endings ^ 159Length of Writing Sample 162Narrative 162Report^ 166CHAPTER VPROFILES OF THREE WRITERS^ 172Description of the Students and Sessions ^ 172A Brief Description of the Students ^173A Brief Description of the WritingSessions ^ 175Syntactic Complexity 178T-unit ^178Extended Noun Phrases^ 189vi iPageNoun Phrase with Modifiers ^ 189Noun Phrase with Prepositional Phrase^191Noun Phrase with Appositive Phrase^194Relative Clauses ^ 195Subject and Complement Clauses ^ 199The Verb Phrase - Modal Auxiliaries ^199Do Auxiliary ^ 201Catenative Verb Phrase ^ 201Simple Verb Phrase 203Progressive Verb Phrase 205Passive Verb Phrase ^ 206Perfect Verb Phrase ^208Text Forms in Writing ^ 209Beginning and Ending Statements ^ 209Narrative Beginnings ^ 209Report Beginnings ^213Narrative Endings ^215Report Endings ^ 218Length of Writing Sample 220CHAPTER VISUMMARY AND DISCUSSION^ 226General Findings ^ 226Syntactic Complexity ^ 228T-Unit Length 229Extended Noun Phrases ^ 229Noun Phrase with Modifiers ^ 230Noun Phrase with PrepositionalPhrase^ 230Noun Phrase with AppositivePhrase 231Relative Clauses ^ 231Subject and Complement Clauses^232The Verb Phrase - Modal Auxiliaries^232Catenative Verb Phrase ^ 233viiiPageSimple Verb Phrase ^ 234Progressive Verb Phrase ^ 235Passive Verb Phrase 235Perfect Verb Phrase 236Perfect Progressive andPerfect Passive Verb Phrases . . . . 236Mean Percentage of Total Verb Phrases 236Text Forms in Writing ^ 237Beginning and Ending Statements ^ 237Beginnings Statements 237Endings Statements ^ 238Length of Writing Sample 239Discussion ^ 239Syntactic Complexity ^ 240Factors Affecting Syntactic Complexity ^ 245Text Forms in Writing 250Beginning and Ending Statements ^ 250Length of Writing ^ 252Factors Affecting Text Forms ^ 253Conclusion ^ 254Implications of the Study ^261Areas for Further Research ^ 264^BIBLIOGRAPHY ^ 267APPENDICES 285Appendix A^ 286Appendix B 289Appendix C^ 295Appendix D 297ixTABLESTable Page1. Mean T-unit Length inPrevious Studies ^ 192. Mean T-unit Length 823. Noun Phrase with Modifiers ^ 934. Noun Phrase with Prepositional Phrase ^ 995. Noun Phrase with Appositive Phrase 1046. Noun Phrase with Relative Clause ^ 1067. Subject and Complement Clauses 1138. Modal Auxiliary Verbs ^ 1169. Catenated Verb Phrases 12210. Simple Verb Phrases ^ 12611. Simple Verb Phrase Ratioof Main to Subordinate Clause ^ 12712. Progressive Verb Phrases 13113. Passive Verb Phrases ^ 13414. Perfect Verb Phrases 13815. Mean Percentages of TotalVerb Phrases ^ 14116. Percentages of Beginning andEnding Statements ^ 152xTable^ Page17. Length of Writing ^ 16418. Ratio of Length of Narrativeto Length of Report^ 168x iFIGURESFigure Page1. Mean T-unit Length^ 902. Grade Means of Noun Phrase withModifiers and Prepositional Phrases ^ 953. Grade Means ofRelative I Clauses ^ 10 84. Modal AuxiliariesUse in Main and Subordinate Clauses ^ 1 175. Percentage of Simple Verbs ^ 12 96. Ratio of Main to Subordinate Clause ^ 12 97. Mean Percentages of VerbType - Narrative ^ 14 28 Mean Percentages of VerbType - Report^ 14 29. Mean Percentages (excluding simple verb)of Verb Type - Narrative ^ 14 310. Mean Percentages (excluding simple verb)of Verb Type - Report 14 311. Narrative Beginning Statements ^ 14 912. Report Beginning Statements 14 913. Narrative Ending Statements ^ 15 814. Report Ending Statements 15 815. Median Lengths of Narrativeand Report^ 171 xi iAKNOWLEDGEMENTSI am very grateful to my advisor, Dr. W. Sutton for hertremendous support and the care and effort she so willingly investedin my project. I would also like to thank the members of mycommittee Dr. P. Duff for her most thoughtful advice andencouragement and Dr. S. Butler for his kind words and timelyassistance.Thanks are also extended to Dr. M. Crowhurst and Dr. K. Reederfor their help and involvement in the early stages of the project andto the many members of the Language Education Department whooffered assistance.I am indebted to the Richmond School District for itscooperation. My special thanks go to the many teachers and studentswho participated so readily in this study and to my principals, BobHolman, Marg Compo, and Deirdre Lintott who offered their supportand understanding.I would also like to thank Dean Mellow, Anne Deacon, and KeriStanger for their perseverance and invaluable assistance checkingthe data.The success of the project could not have been achievedwithout the help of my family and friends. I wish to extend mysincere thanks to them all especially, Cynthia, John and Janet for theunderstanding and continued support offered throughout.Finally, I wish to express my very special thanks to myhusband, Eric, who stood by through it all.1CHAPTER IINTRODUCTIONThe purpose of this longitudinal study is to examine thenarrative and report writing of children in their early intermediateschool years. The nature of the inquiry stems from a considerationof children's growing competency in written language as evidencedby their use of various structures in their writing, in particular,sentence forms (elements of syntax) and text forms (length and useof genre markers, specifically, beginning and ending statements). Itis postulated that data from this inquiry will reveal differences inthe cognitive and linguistic resources that a group of children, withvarying abilities, use in their narrative and report writing throughGrades 4, 5 and 6.Background of the StudyCountless numbers of studies have been conducted in the areaof the development of children's writing, yet this area of studycontinues to present a challenge to researchers. Judith Langer(1992) best describes this challenge: "to understand the knowledge,structures, and strategies that are called on during literacy activities. . . . [and] to know how knowledge shifts and grows over time"(p. 32). Understanding how children's writing, in particular theknowledge, structures, and strategies, develops over time becomesespecially important in view of recent trends in pedagogicalphilosophies regarding instruction and assessment. These trends2focus on the importance of viewing learners as individuals andnecessitate "learner-focussed" instructional and assessmentprocedures that are based on what children "can do" (Ministry ofEducation, 1990). Langer reflects this philosophy, that is, focussingon what children can do rather than what they cannot do, in herwork on children's reading and writing. She states, "my own work inchildren's literacy development has been motivated by . . . . whatlearners can do . . . ." (1992, p. 32).Understanding children's growing and changing development inwriting is important in terms of taking deliberate steps to aid themin the development of written competency. This approach isespoused by the Bullock Committee which stated, "we advocate, inshort, planned intervention in the child's language development"(DES, 1975, p. 67). As Perera (1984) states, "if teachers are going toplan activities to improve their pupils' command of the language,they need a clear understanding of the normal course of languagedevelopment" (p. 1). Thus, the challenge to understand children'swriting development becomes inextricably bound with the challengeto provide instruction which will meet the needs of each learner.Accordingly, understanding what children can do enables teachers todesign instructional activities that "better match classroominstruction to children's developmental journey" (Langer, 1992,p. 33).This study is motivated by a sense of purpose to promoteunderstanding of children's growing and changing languagedevelopment. While there are many aspects of children's writingworth studying, the intent in this study is to examine measurable3indices of written competency, in particular, students' use of syntax(indicating their linguistic resources) and their use of text forms(giving some indication of their knowledge of conventions andcontent in writing, more specifically their schemata of variousgenres). Schemata are generally thought to be composed of genericor abstract knowledge and reflect the prototypical properties ofvarious experiences encountered by an individual. Genre schemataare the experience and knowledge the writer uses to constructexpectations (such as encoding, organization, representation andretrieval of information) of what should occur in a specific genre(Stein and Trabasso, 1982).Syntactic ComplexityHistorically, studies in syntactic complexity have providedsubstantial evidence that certain grammatical structures indicatematurity in elementary school children's writing. The presence ofgrammatical elements such as the extension of noun and verbphrases, the varied and flexible use of subordinate clauses, andincreased T-unit lengths represent increased control over the use oflanguage (Harpin, 1973, 1976; Loban, 1963, 1976; O'Donnell et al.,1967; Perera, 1984). Consequently, teachers who are aware of thenatural sequence of language acquisition as indicated by children'suse of various syntactic structures in their writing can recognizelinguistic maturity, thereby increasing their knowledge of their1^See Section 4.0 Definition of Terms for a description of T-unit length.4students' writing abilities. Also, recognition of linguistic maturity orimmaturity through the analysis of syntax may enable teachers toassist developing writers by providing ". . . . advice and guidancewhich will help pupils to improve their written work" (Perera, 1984,p. 4).Previous research designs in the field have examined theeffects of mode of discourse (genre) on students' use of syntax.They have found that the use of various forms of syntax, as reflectedin T-unit length, for example, changes as the genre changes. Theseresearchers point to the need for up to date longitudinal studieswhich control for mode of discourse and seek to determine generalpatterns in children's use of syntax (Crowhurst, 1978, 1980a, 1980b,1983; Faigley, 1983; Harpin, 1976; Hunt, 1965, 1970; O'Donnell etal., 1967; Perron, 1977; San Jose, 1973, 1978; Witte and Davis,1980).Text Forms in WritingMore recent writing research which focuses on knowledge ofstructures such as form, conventions and content in different genres,specifically the acquisition of genre schemata, provides informationregarding children's sense of purpose for writing (Bereiter andScardamalia, 1982; Clay, 1983; Gundlach, 1981; Harris andWilkinson, 1986; Langer, 1986; Raphael et al., 1989; Schank andAbelson, 1977). Understanding students' use of text forms such asoverall length (indicating understanding of the type and amount ofcontent necessitated by the genre) and sentences which frame the5discourse (beginnings and endings), for example, provides teacherswith valuable information about students' developing understandingof the requirements for written competency in different genres. AsLanger (1986) writes, "the ways in which children handle beginningsand endings are important indices of their developing competencewith discourse level tasks" (p. 65). This understanding, then, enablesteachers to model strategies and provide experience with variousgenres which may assist children in writing more effectively.Evidence from analyses of syntax (which provide evidence ofstudents' linguistic knowledge) can be coupled with data fromanalyses of students' use of text forms (which provide information asto their acquisition of formal or conventionalized writing). Theacquisition and analysis of objective data from both syntax and textforms will provide information about the linguistic and cognitiveresources that children of differing ages and abilities call upon whenwriting narrative and reports. Additionally, the examination of thisdata over a three year period with the same subject group shouldyield pertinent information about these children's developing writtencompetency. Correspondingly, the interpretation of results shouldprovide indications for areas of further research.Research Question The analysis of children's use of elements of syntax and theexamination of children's use of text forms in narrative and reportwriting provides valuable knowledge about their writingdevelopment in these two genres. The examination of the data is to6address this question: Are there differences in the use of a) syntax,specifically: T-unit length, frequencies of extended noun phrases,subject and complement clauses, and extended verb phrases, and b)text forms, specifically: overall length and beginning and endingstatements, in the narrative and report writing of a select group ofchildren in Grades 4 through 6?Definition of TermsGenre - is the style of writing and the purpose for which it iswritten. For this study, genre is a production in which the writer'sattention is directed in one of the following ways:Narrative - using language that, in the main, is imaginative andtells a sequence of events that relates a story. The story may be anyone of a variety of types such as mystery, fantasy, or adventure butfor the purposes of this study is not in the form of a poem or play.Report - using language that, in the main, is factual and sharesinformation about a topic.^A report may tell about such topics assports, hobbies, people, special things, etc.Syntactical Analysis - is a grammatical description which consists ofidentifying regularly occurring words, phrases, and clauses in thelanguage and assigning labels to them (syntactic elements). Theidentification of these elements is based on their function within thesentence (Perera, 1984).Elements of syntactic complexity which are analyzed are:Mean T-unit Length - a T-unit is defined as "one main clause7plus any clausal or non-clausal structure which is embedded in it"(Hunt, 1970). Mean T-unit length is calculated by dividing the totalnumber of words by the total number of T-units in a sample ofwriting. (See Appendix A.)Extended Noun Phrases - any noun phrase with pre- or post-modification (four or more words in length). The specific structuresexamined are: noun phrase with modifiers (the dark, grey -hairedman), noun phrase with prepositional phrase (the dog in the house),noun phrase with appositive phrase (Mary, my older sister) andnoun phrase with a relative clause (see below).Relative Clause - a finite or non-finite clause which givesadditional information about the head noun of a noun phrase. Perera(1984) outlines four stages of relative clauses which indicateprogressive linguistic maturity: Stage I (finite clause modifyingobject - He told us about the people that come out here at night.)Stage II (non-finite clause modifying object - She looked at the manflying away.) or (finite clause modifying subject - All the work thatyou do is make my bed.) Stage III (finite clause modifying aprepositional object pronoun - Mary was the girl whom I hadauditioned at an early age.) and Stage IV (non-finite clausemodifying subject - The jacket worn by the coach was lost).Nominal Clause - a subordinate clause, either finite or non-finite which functions as subject (Playing soccer is fun.) orcomplement in the main clause (She is one lady who works veryhard).Verb Phrase - any finite verb phrase used as the main orsubordinate clausal element which is described as simple (She went8to the store.), progressive (She is going to the store.), passive (Shewas taken to the store.), perfect (She had gone to the store.), perfectprogressive She had been going to the store), or perfect passive (Shehad been taken to the store.) or any extended verb phrase whichuses auxiliaries (e.g. be and have), modals which indicate tense,certainty, possibility, or obligation (will, may, can, shall, could, might,would, should, must, ought to) or is catenated, uses other verbs toform a chain, (wanted to go, likes dancing) in either main orsubordinate clauses. 2Analysis of Text Forms - an examination of overall length of thewriting samples in addition to an analysis of use of beginning andending statements.Length - the total number of words in a sample of writing.Beginning and Ending Statements - An analysis of students'beginning and ending sentences based on a system of classificationused by Langer (1986). Statements in Langer's system arecategorized in one of the following categories in order of complexity:weak, formulaic or structural. Each of these categories has severalsub-categories.^See Appendix B for a detailed description of each ofthese categories and sub-categories.Structural - provide both a clear indication of the genre of thepiece and are integrated into the overall structure. Examples ofbeginnings would be a 'thesis' statement in report writing and a2^See Perera (1984) for a detailed description of catenative verbs.9statement about 'setting' in narrative writing. Examples of endingswould be a 'summary' statement in report writing and a 'solutionresolution' statement (provides resolution to the problem in a story)in narrative writing.Formulaic - traditional beginnings and endings based onconventional phrasings such as "once upon a time," "once," or "theend," for narrative and "this report is about" and "this has beenabout" (or approximations of same) for report.Weak - beginnings that require the reader to read further todetermine the genre. A weak beginning for a story would fail toestablish setting or character and indicate no attempt to entice thereader into reading the story. A weak report beginning might simplybegin with a fact without establishing the topic of the report. Weakendings in stories state a last event in a sequence which does notsignal closing and those for report simply end with a personalstatement or a last fact.Significance of the StudyThe significance of the study is that it will provide up to dateresearch in the area of syntactic complexity in elementary schoolchildren's writing, in particular, children's writing which has beenwritten without time or topic restrictions imposed by the researcher.Previous studies in syntactic complexity study children's writingproduced within a specified time period and school-sponsored 3 or3^School-sponsored writing is writing done for the teacher, that is, theteacher assigns the topic.1 0written in response to stimuli such as specific topics or pictures(Crowhurst, 1978, 1979, 1980a, 1980b, 1983; Golub & Frederick,1970; Hunt, 1964, 1965, 1970; O'Donnell et al., 1967; Perera, 1984;Perron, 1977; Richardson et al., 1976; San Jose, 1972). The analysesof syntax are detailed and are provided for two distinct genres. Thisstudy also provides analyses of children's use of text forms(indicating their knowledge of the requirements for competentwriting in each genre). Relatively few studies (especiallylongitudinal studies) have been undertaken to examine children's useof text forms, such as beginning and ending statements and overalllength, which indicate their growing knowledge of schemata forvarious genres. Similarly, although many studies have focused onchildren's narrative writing and some have provided information onchildren's non-fiction writing such as persuasive, comparison-contrast, and description (Bereiter and Scardamalia, 1982;Crowhurst, 1978, 1983, 1990; Gorman et al., 1988; Harris andWilkinson, 1986; Newkirk, 1987; Perron, 1977; Richardson et al.,1976; Wilkinson et al., 1980), few have looked at children's reportwriting, a genre used widely at the elementary school level.Much of the previous research has been confined to cross-grade studies of relatively short duration. Evidence fromlongitudinal studies in the area of the use of syntax is limited,especially from recent studies in which children have control overthe topic written and the amount of time needed to write. Similarly,there are no longitudinal studies presently available which focus onthe use of text forms in differing modes of discourse or that combinethe analysis of syntax and text forms with observations taken from1 1case studies using qualitative research design procedures.The accumulation of data and its subsequent analysis willprovide information about children's use of syntax (indicatinglinguistic maturity) and their use of text forms (indicating knowledgeof different genres). Further, the analyses will examine the use ofsyntax and text forms in the writing of the same children at varyingstages in their development over three years. The results from theanalyses will be combined with qualitative research data which willprovide valuable information as to students' use of syntax and textforms indicating children's growing competency in written language.The summary and discussion of the results will suggest general andpedagogical implications and may indicate areas for future research.Limitations of the StudyThe study is descriptive in nature and therefore results are notconclusive but are intended to suggest general and pedagogicalimplications and directions for further research. In conducting thestudy it was apparent there were certain limitations.Although some attempt was made to control the circumstancesunder which the writing samples were obtained, the study does notcontrol for differences in instruction received by the students, theirwriting environment, or differences in teacher effects. Similarly, thestudy does not account for individual subjects' backgrounddifferences such as socio-economic status or home environment.The sample size of 14 students is quite small and does notpermit reliable generalization to a wider population. In the writing1 2sessions an attempt is made to model regular classroom procedures;the researcher, using a process approach to writing (Graves, 1984)asks the subjects to draft, edit and proofread all writing samplesbefore they are handed in. Students are given as much time as theyneed to write and therefore it may take several sessions before apiece of writing is completed. However, as the students only meetwith the researcher once a week, the length of time between writingsessions may prevent the sessions from adequately modeling regularclassroom procedures. Also, as the students are providing only foursamples of writing a year, the sampling of writing may beconservative.The study deals primarily with objective measures ofidentifiable characteristics in children's writing. It does not evaluateword choice, cohesiveness, or rhetorical effectiveness and does notconcern itself with overall quality of writing. Growth is measuredthrough the acquisition of more syntactically complex forms ofwriting (Perera, 1984) and through increasing length and the use ofmore complex text forms which indicate knowledge of discoursetasks (Langer, 1986).The analysis deals exclusively with data provided from reportand narrative writing and is not intended to be generalized to othermodes of discourse. As indicated in the definition of terms, there aremany forms of narrative and report writing. The subjects wereallowed freedom of choice of topic so they could write from theirown experience. Failure to control for topic may prove to be aninherent weakness in the design as comparisons are made betweenthe analyses of different pieces of writing on different topics.13CHAPTER IIREVIEW OF THE LITERATUREThe major areas of research relevant to the present study arestudies examining children's syntactic development and thoseexamining children's developing ability to write in various genres.Syntactic ComplexityBeginning in the early 1900's researchers examined children'swritten language in an attempt to determine measurable indices oflinguistic maturity which would indicate patterns of normativecognitive growth. Since then an enormous amount of research hasbeen conducted into syntactic development in writing. The purposehere is not to repeat material to be found in other reviews (see, forexample, McCarthy, 1954; O'Donnell, Griffin & Norris, 1967) butrather to review research involving three structures of particularinterest, namely, T-unit length, extended noun phrases, and verbphrases. T-unit length is of interest because, over several decades, ithas proven to be a simple, yet reliable means of measuring syntacticdevelopment. Similarly, the use of extended noun phrases andextended forms of the verb phrase have been used with notablesuccess as indices of maturity in children's writing.14Basic Constituent of Analysis - The T-unitThe fact that children's writing changes as they grow older hasalways been apparent to both researchers and teachers. Researchersin the area of children's writing have sought ways to describe thesechanges as well as means of measuring them. Early studiesexamining children's writing used sentence length and classificationas the basis of their analyses (Anderson, 1937; Bear, 1939; Frogner,1933; Heider & Heider, 1940; McCarthy, 1954; Stormzand & O'Shea,1924). These analyses consistently indicated both that childrenwrite longer sentences as they mature, and that increased length isaccompanied by a decrease in the use of simple sentences and anincrease in compound and, especially, complex forms (Bear, 1939;Harrel, 1957; Heider & Heider, 1940; McCarthy, 1954). McCarthy(1954) summarized the relationship of sentence length to theincreased incidence of complex sentences stating that "whateverincrease in sentence length occurs at higher age levels is broughtabout largely through the addition of subordinate clauses" (p. 551).Thus, many of the researchers also studied incidence ofsubordination and clause length in children's writing but began toexpress difficulties using the sentence as the main constituent ofanalysis within which these structures were contained (La Brant,1933; McCarthy, 1954; Hunt, 1965).La Brant (1933) questioned using the sentence as the basicconstituent, observing that it is often "impossible to determine whatconstitutes a sentence in an individual's written composition unless ithas been perfectly punctuated" (p. 482). She preferred clause length1 5as the main unit of analysis. La Brant (1933) did seminal work inthis area of research by developing a subordination index wherebyshe divided the number of subordinate clauses by the number ofmain clauses. She found advances in age and grade were clearlyreflected in higher subordination ratios. Criticisms were levied on LaBrant's analysis, however, as her definition of clause did notdistinguish between coordinated predicates and was "not what manyschool grammarians nowadays would call a clause" (Hunt, 1970, p. 3).Several researchers modelled their analyses of children'swriting on La Brant's (1933) work, not only calculating subordinationratios, but adding transformational grammar approaches (modelledafter Chomsky, 1957, 1965) to their measures (Hunt, 1965; Golub &Frederick, 1970a; Loban, 1963; O'Donnell et al., 1967; Riling, 1965;San Jose, 1972). Detailed accounts of these important studies arepresented later. However, these systems of analysis provedexceptionally time-consuming and laborious. The study of syntacticcomplexity was in need of a simpler, more easily identified yetreliable measure. Such a measure was developed by Kellogg Hunt(1965).Hunt (1965) initiated his study of children's writing in Floridafollowing the work of his predecessors. He examined 1000-wordwriting samples of regular classroom writing from 18 children (nineboys and nine girls) in each of Grades 4, 8 and 12 and the writing ofskilled adults published in Harper's and The Atlantic. He looked atsentence length and clause length as well as a subordination ratiomodelled after La Brant (1933). Hunt (1965), like La Brant threedecades before him, described his difficulties identifying sentences.1 6Younger students' nonconventional use of punctuation and generaltendency to overuse coordinating conjunctions in their sentencesoften resulted in their having a higher mean sentence length thanolder children. He therefore sought another, more satisfactory unitof analysis which would reflect clause length and the use ofsubordinate clauses and yet would be unaffected by the difficultiesassociated with punctuation and excessive use of coordinateconjunctions. His search led him to the development of the "minimalterminable unit" or T-unit. He defined the T-unit as "one main clauseplus any subordinate clausal or non-clausal structure that isembedded in it" (1965, p. 9). Hunt (1965) found that mean T-unitlength steadily and significantly increased from grade level to gradelevel: at Grade 4 it was 8.6 words, at Grade 8, 11.5 words and atGrade 12, 14.4 words. Furthermore, statistical treatment of theresults of the four measures Hunt used (sentence length, clauselength, subordination ratio and T-unit length) led to the conclusionthat T-unit length was the best indicator of a student's grade level.Hunt (1970) conducted a subsequent study to determinewhether grade-related differences in mean T-unit length were aresult of different kinds of writing done at different grade levels. Headministered the same writing task to 50 students in each of Grades4, 6, 8 10, and 12 and two groups of adults (described as "averagedand skilled"). Subjects were asked to rewrite "in a better way" apassage consisting of 32 very short sentences. He found that olderstudents wrote significantly longer T-units when combining thesentences. Further, he found that across three of the four two-yearintervals the means were significantly different (Hunt, 1970).1 7Additional support for the validity of the T-unit was providedby Loban's (1963, 1976) longitudinal study of the languagedevelopment of 211 students in Berkeley, California. Loban,examining 1000-word writing samples yearly from the same groupof students in Grades 3 through 12, was the first to providenormative data on the development of children's writing. Lobanused the "communication unit" which he described as "the samemethod of segmentation [as] in Hunt's research" (1976, p. 9) as thebasis of his analyses.^He defined the communication unit (alsoknown as c-unit) as "each independent grammatical predication." Hefound the length of communication unit increased by an average of.63 words per year, with marked increases between Grades 7 and 8(1.43 words) and 11 and 12 (2.26 words).Verification of the T-unit's validity was also provided by acomprehensive cross-grade study which modelled the work of Hunt(1964, 1965). O'Donnell, Griffin and Norris (1967) examined writingsamples of 15 boys and 15 girls in each of Grades 3, 5, and 7. Theymeasured T-unit length and the rate of occurrence per 100 T-units ofthe numbers, kinds, and functions of sentence-combiningtransformations. They found that T-unit length and the rate ofsentence-combining transformations were related in measuringsyntactic complexity: as T-unit length increased in successive gradesso did increments in the number of sentence-combiningtransformations that occurred in each grade (O'Donnell et al., 1967).Therefore, in that T-unit length was easier to compute, theyconcluded that it had "special claim as a simple, valid indicator ofdevelopment in syntactic control" (p. 98).18Several other researchers found comparable mean T-unitlengths to those of Hunt's in their subjects' writing. Mellon's (1969)seventh grade students produced a mean T-unit length of 9.96 wordson pre-test compositions prior to sentence combining instruction.Rosen (1969) found that his 50 15- and 16-year-olds produced amean T-unit length of 14.6 words. Dixon (1970) analyzed 200words taken from narratives written by fourth-, eighth-, andtwelfth-grade and college seniors. He examined T-unit length, freemodifier variables, and intra-T-unit structures (Christensen, 1967)and found that T-unit length correlated more highly (.66) with gradelevel than any other variable.Golub and Frederick (1970a) in their study of 80 fourth-gradestudents and 80 sixth-grade students found mean T-unit lengths of9.5 (Grade 4) and 11.7 (Grade 6), a statistically significant difference.Richardson, Calnan, Essen and Lambert (1976) found a mean T-unitlength of 10.8 for 11-year-old students. Perron (1977) studied T-unit length among third-, fourth- and fifth-grade students and foundsteadily increasing T-unit length across the grades. Results of thesestudies are compared in Table 1. 5^The table indicates that T-unitlength has been used with notable success. In fact many of hiscontemporaries regarded Hunt as having provided normative data onthe development of syntactic maturity (Mellon, 1969; San Jose, 1972;Stotsky, 1975). The simplicity of using the T-unit plus its reliabilityhas made it one of the most widely used indices of syntacticcomplexity (e.g. Crowhurst, 1978, 1980a, 1980b; Crowhurst & Piche,5^This table does not adequately reflect the difference in writing task orconditions.^Nevertheless it provides some basis for comparison.19TABLE 1Mean T-unit Lengths in Various StudiesStudy N Type of WritingGr. 3 Gr. 4 Gr. 5GradeGr. 6 Gr. 7 Gr. 8 Gr. 12Hunt (1965)O'Donnell et al.(1967)Bortz(1969)Golub & Frederick(1970)San Jose(1972)Richardson et(1976)Loban(1976)Perron(1977)Crowhurst(1978)18305016040al.52121131120(variety of topics)(response to twoshort films)narrativeexposition(various topics)(response topictures)narrativeexpositionargument("imagine you are25 years old. . .")(typical examples)narrativeexpositionargument(assigned topics)narrativedescriptionargument(response topictures)7.677.67.28.1510.48.610.510.89.58.79.910.48.028.918.9812.89.348.9211.58.769.5610.413.19.6811.1511.710.89.0410.1310.4511.759.998.9411.510.314.413.272 01979; Hunt, 1964, 1965, 1970; Kroll, 1983; Langer, 1986; O'Donnellet al., 1967; Perron, 1977; Richardson et al., 1976; Rosen, 1969; SanJose, 1972; Smith & Coombs, 1980; Veal & Tillman, 1971; Yau &Belanger, 1985). While the T-unit has been used with someconfidence, it has also been subjected to criticism as it has beenfound to have certain limitations.Although early researchers who examined syntactic complexitydid not account for differences in the writing situation, particularlymode of discourse, many alluded to the problem. Anderson (1937)questioned the application of La Brant's (1933) subordination indexto written language when variables of subject matter and situationvaried. Other researchers pointed to the need to examine the effectsof genre and subject matter in writing (Hunt, 1964, 1965; Loban,1963, 1976; O'Donnell et al., 1967; Sampson, 1964; Seegers, 1933).Thus, a series of subsequent researchers studied the effects of modeof discourse on T-unit length (Bortz, 1969; Crowhurst, 1978, 1980a,1980b; Crowhurst & Piche, 1979; Langer, 1986; Perron, 1977;Rosen, 1969; San Jose, 1972).Bortz (1969) examined one example in each of descriptive,expository and narrative writing from 50 subjects per grade inGrades 4, 5, and 6. He found significant differences between modesfor all grades on mean T-unit length. Exposition was the mostcomplex, that is the students used the longest T-units followed bynarration and description.^Rosen (1969), who examined the effectsof eight different sets of writing assignments on the syntacticcomplexity of 15- and 16-year-olds, found that the two sets whichrequired narration had the lowest mean T-unit length, while the set2 1that urged students to debate and discuss produced the greatestmean T-unit length. San Jose (1972) studied the narrative,expositive, argumentative, and free choice writing of 40 Grade 4students. While some practice effects may have limited her findings(students wrote successively, a mode per week for four weeks), shefound statistically significant differences in mean T-unit lengthacross modes. Similar results were found by Perron (1977).Collected writings in the same modes from approximately 51students in each of Grades 3, 4, and 5 showed consistent, statisticallysignificant differences in mean T-unit length across all grades. Inboth studies the order of decreasing magnitude was argument,followed by exposition, followed by narration and description.Crowhurst (1978) found that differentiated effects on writing wereproduced by mode of discourse on 120 sixth-grade and 120 tenth-grade students' writing. The effect of mode of discourse was"persistent" across the three modes of argument, narration anddescription. At the Grade 6 level, argument was more syntacticallycomplex, as measured by T-unit length, than either narration ordescription, while there was little difference between the latter two.Grade 10 writing showed greatest T-unit length in argumentfollowed by description and narration. Crowhurst (1980a) continuedto study the effects of mode of discourse in a partial replication andextension of her previous study with 240 students in Grades 6, 10,and 12. Significant differences in T-unit length between the modesof argument and narrative were found at every grade level. Morerecently, Langer (1986) found significant genre effects on T-unitlength in the narrative and report writing of eight,- 11,- and 14-2 2year-olds. Overall, the ratio of words per T-unit was greater forreport than narrative. For example, in narrative, eight-year-olds hadmean T-unit lengths of 8.43 words, whereas 14-year-olds' T-unitlength was 12.26.^By contrast, the same age groups had mean T-unit lengths in reports of 9.03 and 13.37 respectively.Not only is T-unit affected by mode of discourse but there isalso evidence in the research that it cannot be used as a reliableindicator of writing complexity for individuals. The research in T-unit length is most often reviewed in terms of age- or grade-relateddata (Crowhurst, 1978; Crowhurst & Piche, 1979; Hunt, 1964, 1965;O'Donnell et al., 1967; Perron, 1977). Both Hunt (1965) andO'Donnell et al., (1967) refer to general trends recognizing that there"are no data to show how consistently [the index] measure[s] thestructural complexity of an individual's writing in various situations"(O'Donnell et al., 1967 p. 6). Crowhurst (1978, 1980a, 1980b), forexample, in her studies of writing samples of students in Grades 6,10, and 12, found substantial variation in mean T-unit length by thesame student in the same mode of discourse. She found differencesof up to ten words for individuals in each grade in both narrativeand argument. Therefore, Crowhurst (1978) cautioned againstapplying the so-called "norms of syntactic complexity" to individualstudents.T-unit length may be regarded as a measure that gives ageneral indication of syntactic development. It has been argued thatit obscures important structural differences in the writing of groupsof children (Endicott, 1973; Golub & Kidder, 1974; Loban, 1976;Richardson et al., 1976). In the 1970's, various researchers worked2 3to develop other indices that would reflect more precisely the degreeof complexity of writing in syntactic studies, for example, Endicott's(1973) transformational and morphemic analysis, Golub and Kidder's(1974) Syntactic Density Scale, and Loban's (1976) Elaboration Index.After reviewing these scales, however, O'Donnell (1976) points outthat, while they have potential to reveal developmental differences,they involve "rather complex and expensive procedures and since T-unit length is highly correlated with Syntactic Density Score. . . . [it] isstill the most useful and usable index of syntactic development overa wide age range" (p. 38).Mean T-unit length, then, has received widespread acceptanceand has been used extensively to measure syntactic maturity inchildren's writing. It has been shown to be reliable over a wide agerange and has been useful for predicting general age- or grade-related norms of complexity in a wide variety of research studies. Itmust, however, be used with some caution in that inconsistencies inT-unit production have been reported across different kinds ofwriting. Moreover, caution must be exercised if T-unit length is usedto examine the writing of individual students.As stated previously, other measures of syntactic complexityhave traditionally been used to accompany results provided by meanT-unit length. Other measures have potential for providing moredetailed information about individual and group differences than isprovided by the more generalized T-unit measure.24Syntactic Structures Within the T-unitSeveral early researchers conducted sophisticated analyses ofsyntax in children's writing (La Brant, 1933; Golub & Frederick,1970a; Harrel, 1957; Hunt, 1964, 1965; Loban, 1963, 1976;O'Donnell et al., 1967; San Jose, 1972). Subordination indices orratios of subordinate and main clauses to main clauses, analyses offrequency of subordinate clauses, and analyses of the noun phraseand verb phrase highlighted specific features of the syntax in writingwhich could be used to trace development, features indicative of"maturity" - a term Hunt (1965) described as, "the observedcharacteristics of writers in an older grade" (p. 5).^Excellent reviewsof the procedures and findings of these grammatical analyses areprovided by Hunt (1964, 1965), Loban (1976) and O'Donnell et al.(1967) and need not be repeated here. Results from these studies,however, indicate there are three grammatical structures inchildren's writing which have consistently indicated maturity, thuswarranting continued study: extended noun phrases (including theuse of relative clauses), subject or complement nominal clauses, andthe use of certain forms of the verb phrase.Extended Noun Phrases and Nominal Clauses The use of noun phrases (headed noun phrases pre- or post-modified) especially those which are extended through the use ofmodifiers, prepositional phrases and relative clauses as well as theuse of nominal clauses (subject and complement), has persistedthroughout the literature as an indicator of maturity in children's2 5writing (Harpin, 1976; Hunt, 1964, 1965, 1970; Lawton, 1968;Loban, 1976; 0' Donnell et al.,1967; Perera, 1984; Riling, 1965).Hunt (1964, 1965), in detailed analysis of 1000-word writingsamples of the same students' writing, studied all headed nounphrases within T-units including the various structures used for theirmodification. He found that younger students used largely simplenoun phrases: unmodified common nouns, personal pronouns andproper nouns. By contrast, older students wrote more and longernoun phrases using more adjectives, genitives, prepositional phrases,infinitives, participle phrases and relative clauses. Statisticallysignificant increases for grade were found in the use of prepositionalphrases and relative clauses. In reference to relative clauses, Hunt(1965) reported that they increased significantly with grade andwere a most significant factor in the increased length of T-unit. Thenumber of relative clauses per T-unit increased from .044 to .09 to.16 for Grades 4, 8, and 12 respectively. He also found that nominalclauses were frequent in the writing samples. The great majority,however, were functioning as object clauses and the incidence ofcomplement and particularly of subject nominal clauses was rare.For example, out of 247 nominal clauses produced by 13-year-olds,only 9 were functioning as subjects. Hunt (1965) concluded that theability to use the nominal clause in positions other than object was "amark of maturity" (p. 77).^Hunt also developed a count system toascertain the complexity of noun phrases and nominal clauses usedby the students. For each word or phrase used to modify a headednoun and for each nominal clause the student was awarded onecount. He tabulated the total number of counts per student for2 6extended noun phrases and nominal clauses (those with counts ofthree or more). He found significant increases in complexity, asmeasured by this index, at each grade level. An analysis of varianceshowed that this measure was significant for grade. Hunt (1965)concluded that the complexity count of noun phrases includingnominal clauses was "one of the best indices of maturity" (p. 116).Riling (1965), in her analysis of 25 sentences from the writingof 114 Caucasian and 107 Negro [sic] Grade 4 children and 110Caucasian Grade 6 children, found that children with verbalintelligence scores in the highest quartile (as measured bystandardized tests) used more phrases to elaborate the grammaticalsubject than students in the lower quartile. As Riling does notspecify how the subject noun phrases are elaborated, her results aredifficult to interpret. She also reported that none of her studentsused clauses much to elaborate the subject, a finding which isconsistent with other researchers at that time (Hunt, 1964, 1965;Lawton, 1968; Loban, 1976; O'Donnell et al., 1967). Yet, in view ofthe small size of students' writing samples, her finding is notsurprising.Reporting on writing samples from 30 students in each ofGrades 3, 5, and 7, O'Donnell, Griffin and Norris (1967) drew severalinteresting conclusions regarding extended noun phrases andmaturity in writing. They found the incidence of sentence-combiningtransformations in noun phrases increased significantly in bothGrades 5 and 7. All measures of headed noun phrase modificationmeasured as a rate of occurrence per 100 T-units, increased fromGrades 3 to 7. Significant increments were found in the use of2 7genitives, modifiers, and relative clauses in Grade 5 and the use ofprepositional phrases, participle phrases, gerund phrases andrelative clauses in Grade 7.^Noun phrases extended by a relativeclause, for example, increased from .99 times per 100 T-units inGrade 3 to 3.37 times in Grade 5 to 4.46 times in Grade 7. Theyconcluded that increased exploitation of these transformationsmarked development of syntactic control in children's language(O'Donnell et al., 1967).Lawton (1968) studied the writing of 20 boys aged 12 and 15years to determine whether there were differences in writing abilityattributable to social class. His description of the writing depictedsubject, complement and appositive nominal clauses, as indicators ofmaturity in writing. Further, he found clauses of these types were"fairly uncommon" in middle class students' writing samples but"almost non-existent" in the same writing task in working classstudents' work.Hunt (1970) had subjects from Grades 4, 6, 8, 10, and 12 andtwo groups of adults ("average and skilled") rewrite "in a better way"a paragraph comprised of 32 very short sentences. He found that, aswriters matured, they reduced predication (i.e., represented theideas contained in short sentences by reducing the sentences toeither a word, words, or a phrase) by extending the noun phrase.Subject noun phrase modification by the use of appositives,prepositional phrases, prenominal adjectives and relative clauses wasrare for Grade 4, somewhat rare for Grade 6 and generally increasedsteadily for Grades 8, 10, 12, and skilled adults (Hunt, 1970).Averages of the number of incidences recorded for using an2 8appositive, for example, were .25 in Grade 4, 2.5 in Grade 6, and 9.0.,9.2, 9.0, and 11.5 times for Grades 8, 10, 12, and skilled adultsrespectively.Golub and Frederick (1970a) conducted a 63-variable linguisticanalysis of 80 students' writing in each of Grades 4 and 6. Amongtheir findings were several grammatical structures which they foundindicative of maturity in writing. Many of these structures modifiedand thus extended the headed noun phrase. Sixth-grade studentsused significantly more relative clauses than fourth-grade students.While not statistically significant, sixth-grade students also tended touse more adjectives and more prepositional phrases. An interestingfinding of Golub and Frederick's (1970a) was the correlation betweenlinguistic variables and quality of writing. While quality ratingvariables were not stipulated and interrater reliability betweenraters was not always high (.64, .68, .80) they reported "good" themewriters were distinguished from "low" theme writers by the use ofseveral linguistic variables found in the noun phrase: nounmodifiers, possessive markers and prepositional phrases.Harpin (1976) commented on nominal clauses in his two yearlongitudinal study of the writing of 7- to 11-year-olds. His used adescriptive research design to study students' factual and creativewriting. He found nominal clauses functioning as objects were usedabundantly in both factual and creative writing samples. Subject,complement, appositive and prepositional nominal clauses, bycontrast, were infrequently used. He reported that nominal clausesused as subject were a rarity at this age, never appearing more thanonce for every 150 nominal clauses. Harpin (1976) also found a2 9steady advance in students' use of relative clauses as they matured.Kinds of subordinate clauses represented as percentages of allsubordinate clauses revealed steady increases in the use of relativeclauses: Grade 3 used 11 percent, Grade 4, 16 percent, Grade 5, 19percent and Grade 6, 22 per cent.Loban (1976) analyzed the writing samples of the same 105students over a nine-year period. On the basis of teacher ratings andstandardized tests, he divided the group into 35 students of "high"language ability, 35 of "low" ability and 35 who were randomlyselected from his original sample of 211 students. While Loban didnot study noun phrases specifically, he did look at subordinateclauses which would function as part of an extended noun phrase,specifically relative clauses, and nominal clauses. All three groupsused an increasing proportion of relative clauses (per total number ofclauses) from Grades 5 through 12. The random group, for example,increased from 14 per cent in Grade 5 to 33 per cent in Grade 12. Inreference to nominal clauses, Loban's finding verified Hunt's (1965)and Harpin's (1976) results, in that those functioning as objects werevery common and learned early in life, but nominal clauses used assubjects and appositives were later, less frequent developments.Wilkinson et al., (1980) used a qualitative research design toexamine the autobiographical, narrative, descriptive, andargumentative writing of students aged seven, 11 and 13. Theycollected four pieces of writing from 30 children at each age level.While descriptions of the analysis of syntax (in their "stylistic mode")were limited, they found use of relative and noun clauses increasedas the students became older.3 0Perera (1984) provides a detailed summary of the syntacticcomplexity in children's writing using a corpus of material;children's writing samples from ages seven to 17 were collected fromtwo sources: the Child Language Development Project at BristolUniversity (Kroll & Wells, 1983) and the Children LanguageTranscripts Volumes I-IV (Fawcett & Perkins, 1980). She reportedthat extended noun phrases such as determiner-adjective-noun,noun-prepositional phrase, and noun phrase in apposition were usedwith increasing frequency from ages seven to 17. She found thatbeginning writers employed only simple noun phrases as subjectsbut, by the age of 10 they incorporated expanded noun phrases(through the use of words and phrases) which continued to increasein complexity as they matured. Further, nominal clauses used assubjects were especially rare and were more likely to occur in thewriting of older children. Perera (1984) concluded that "only writerswith some linguistic maturity are able to write subject phrases thattap some of the structural possibilities of the noun phrase" (p. 229).The increased use of modification in the noun phrase aschildren's writing matures has been referred to in a wide variety ofstudies. Mellon (1978) likens the more mature use of extended nounphrases and nominal clauses in writing to a growth in conceptualknowledge indicating maturity of the writer. He states that "part ofsyntactic fluency growth attributable to increasing elaboratedness inthe grammatically restrictive structure of dominant noun phrases isa direct and unavoidable consequence of conceptual knowledge"(p. 17).3 1Little research has been conducted where extended nounphrases and nominal clauses have been studied in different writingtasks and across different modes of discourse. It would seem logicalthat incidents of these would vary with the kind of writing in muchthe same manner as the T-unit. The few studies that report on nounphrases and nominal clauses used in different situations indicate thisto be the case.In Golub and Frederick's (1970a) study, Grade 4 and 6 writerswere asked to write in response to different instructions given with avariety of pictorial stimuli. Although complexity of syntax was notaffected by different instruction, it was affected by differentpictures. The effect of colour pictures versus black and white wassignificant for a number of linguistic variables. Black and whitepictures evoked more subordinate clauses, especially nominal andadverb clauses, while coloured pictures evoked more adjectives andprepositional phrases. Thus, in this instance, different stimuliaffected complexity of language use. San Jose (1972), in her study ofGrade 4 writing across four modes of discourse, found that certainlinguistic features, which resulted in extended noun phrases, wereaffected by mode. She found modification of the noun phrasediffered significantly among modes in eight of 10 categories, two ofwhich extend the noun phrase: the use of prepositional phrases andphrasal genitives. Similar effects of mode on syntax were reportedby Crowhurst and Piche (1979) in the descriptive, narrative andargumentative writing of Grade 6 and 10 students. They found thatclause length and ratio of clauses to T-units were affected by mode.While they did not differentiate among types of clauses affected,3 2they did find that clause length was significantly longer indescription than in argument or narration. Similarly, ratio of clausesto T-unit increased significantly in order of magnitude fromdescription to narration to argument.Thus, the studies reviewed support the finding that increaseduse of extended noun phrases accompanies advances in age. There isalso evidence that mode of discourse and writing task have beenshown to affect the use of extended noun phrases. The results ofthese studies suggest that study of frequency and use of extendednoun phrase is a viable method of describing syntactic complexity inchildren's writing.The Verb Phrase Several studies have examined the verb phrase in children'swriting to see if advances in age accompany changes in the types andstructures of the verb phrase (Golub & Frederick, 1970a; Hunt, 1964,1965, 1970; Kress, 1982; Loban, 1976; Perera, 1984). Reports fromthese studies outline several verb forms which are indicative ofmaturity in writing.One of the most extensive analysis of the verb phrase wasconducted by Hunt (1964, 1965). He tabulated frequencies of 20different finite verb auxiliaries and 24 classes of regular andirregular verbs per 1000-word sample. Hunt (1964) found that useof the auxiliary became more complicated with successively olderstudents. Statistically significant changes for grade level occurred inthe use of perfect tenses (have + past participle, have walked) and inpassives (be + past participle, is cooked). The frequency of perfect3 3forms increased substantially, especially from fourth to eighthgrades. The numbers of occurrences of perfect forms were 50, 128,and 144 for Grades 4, 8, and 12 respectively. Increases in thepassive form rose from 50 in Grade 4 to 142 in Grade 8 and finally185 in Grade 12. Use of progressive forms (be + present participle, isgoing) did not change from Grades 4 to 8 but decreased sharply inGrade 12. In his study of modal auxiliaries, Hunt found thatincreased use of six out of eight modals (will, would, shall, should,may and might) phrase was significant for grade level. Further,increased use of the word can was significant for grade. Regular andirregular verb analysis revealed three types or combinations of typeswhich showed significant differences by grade. Simple intransitiveverbs (they came quickly) decreased significantly from Grade 8 to 12and intransitives with complements of motion (he went away)decreased significantly from grade to grade. Hunt (1964, 1965) alsofound the use of the verb be (as distinct from be as an auxiliary)increased from Grade 8 to 12.Several researchers, while not studying the verb phrasespecifically, have noted the use of passives in children's writing.O'Donnell et al. (1967) found the use of passive constructionsinfrequent in samples of writing from Grades 3, 5, and 7. Althoughhis tables indicate infrequent use, they also show a slight increase inuse of passives across age groups. In Grade 3, the rate of occurrenceper 100 T-units was .35, in Grade 5, .65 and in Grade 7, 1.05. Lawton(1968) also commented on the use of passive verb forms in thewriting of 12- and 15-year-olds. He noted a higher incidence ofpassive verbs in the older students' writing. Hunt (1970) also3 4reported on passive forms in the rewriting of 32 short sentencesamong Grades 4, 6, 8, 10, 12 and skilled writers. He found increaseduse of the passive with increases in age of the writer. Total numbersof passives were two for Grade 4, 12 for Grade 6, 27 for Grade 10,84 for Grade 12 and 106 for skilled adults. Results of these studiesare comparable to Hunt's (1964, 1965) earlier work. Together theysuggest that increased use of the passive verb in writing begins inthe middle school years and becomes more prevalent in the highschool and adult years. These results suggest it is reliable as ameasure of syntactic complexity.Golub and Frederick's (1970a) studied verb forms in thewriting of Grade 4 and 6 students. They tabulated frequency of theuse of modals, have and be forms, as well as infinitives andcoordinate and "unique" verb forms. Their results indicated astatistically significant difference for grade in the use of modals (will,would, shall, should, must, may, might, can, could, ought to, have toand used to).^Other forms of the verb did not show statisticaldifferences in terms of grade. Interestingly, Golub and Frederick alsofound differences in the use of verb phrases when reporting onquality of writing. As stated previously, they had two groups ofwriters based on independent raters' unspecified quality ratings:those who wrote high quality themes and those who wrote lowquality themes. Their results show that all measures of the verbform were used significantly less by writers of low quality themes.In his summary of his longitudinal study of the same writersfrom Grades 3 to 12, Loban (1976) reported that the results of hisverb density index, which is a method of counting total verb words3 5and representing them as a percentage of words per communicationunit, were not significant. Loban (1976) also studied non-finiteverbs and found they, too, were not significant as a means ofdistinguishing maturity of writers. He did, however, note that whenhe compared verb usage between high language ability and lowlanguage ability groups, the high group used substantially moreinfinitive verbs than the low group. From this observation heconcluded that "those rated high in language make a conscious effortto use infinitive verbs in their writing" (p. 69). He suggested theneed for further research into the use of verbs stating "it is still ourobservation that few of our subjects during the elementary schoolyears spontaneously used tenses such as the past or future perfect"(p. 69). He also argued that tallying verb strings of five or morewords might provide more evidence in describing verb usage andlanguage maturity.Kress (1982) provides a detailed description and analysis ofchildren's writing from ages six through 12. He describes thechanges in the students' use of some forms of the verb phrase asthey grow older. In particular, he comments on children's increasingability to use more complex verb phrases as evidenced by the"temporal layerings" within their narrative writing. He writes, "thewriter establishes further temporal layerings within the past bymeans of a step from the past perfective [verb phrase] to the simplepast" (p. 117).Perera's (1984) summary presents a detailed account of thechanges in the use of the verb phrase as children's writing matures.In reviewing the language samples from several studies, she found3 6that younger children (aged seven and eight) "characteristically" usesimple active verbs (ie, two boys went fishing and they had somefish). Children of this age also used auxiliary forms (was coming),catenative forms (seemed to creak) and verb particle phrases (hadworn out) readily. Yet, while two and three word verb phrasesappeared commonly at this age, other forms such as modals (could,would, should, may, might,) and combinations of the catenative andauxiliary phrases (had been trying to go), did not appear frequentlyin writing until children were older.As with the T-unit and the extended noun phrase, use of theverb phrase may be affected by rhetorical context. Little evidencehas been advanced to either support or refute this statement. Onestudy conducted by San Jose (1972) indicates that verb usage isaffected by mode of discourse. She tabulated incidences of Grade 4writer's use of passive verbs, six modals (will, would, shall, should,may, and might) and two other modals (can and could) in fourdifferent modes of discourse: description, narration, exposition andargument. She found the use of the passive verb was unaffected bymode of discourse and was distributed equally over the four modes.The other two verb analyses, which involved using two sets ofmodals, were both significantly affected by mode. San Jose does notspeculate as to reasons why but she found argument elicited thehighest incidence of modals followed by exposition, followed bynarrative and description equally. As the passive is a latedeveloping feature of children's writing, it may be that San Jose'sresults were affected by infrequency of its use; however differencein use of modals is notable.3 7In summary then, certain forms of the verb phrase appearlater in children's writing. In particular, the use of perfect tenses,passive forms and modals have been shown to increase with age.Similarly, there is evidence that the use of extended verb formsinvolving combining verbal auxiliaries, catenative forms and modalsis also a later developing feature in children's writing. Littleresearch has been conducted which tests the effects of variousvariables which may affect verb usage; however, there is someevidence that use of modals is affected by mode of discourse.Development of Writing in Various GenresIn developing their writing ability children must learn theconventional forms of the established genres, namely the contentrequired, how to organize the content and the types of sentences thatare appropriate to frame the discourse. Several researchers theorizethat knowledge of these essential forms is a major requirement forcompetence in writing (Bereiter & Scardamalia, 1982; Clay, 1983;Dillon, 1981; Gundlach, 1981; Kress, 1982; Harris & Wilkinson,1986; Kintsch, 1982;^Martin & Rothery, 1986; Martlew, 1986;Scardamalia & Bereiter, 1986). It is thought that development of thisknowledge comes from experience with genres; through experience,writers gradually come to internalize knowledge of the forms andstructures used in various kinds of genres often referred to asschemata (Anderson, 1937, 1978; Anderson et al., 1976; Bartlett,1932; Black et al., 1982; Mandler & Johnson, 1977; Rumelhart &Ortony, 1977; Schank & Abelson, 1977). Knowledge of genre3 8schemata is used in conjunction with students' knowledge of surfacestructures such as syntax, to help govern the production of text(Kress, 1982). A schema for narrative is acquired early (Kroll &Anson, 1984; Hidi & Hildyard, 1983; Mandler & Johnson, 1977;Martlew, 1986; Rumelhart & Ortony, 1977). Schemata for othergenres are somewhat slower to develop (Applebee, 1984; Crowhurst,1983, 1989, 1990; Dillon, 1981; Durst, 1984, 1987; Harris &Wilkinson, 1986; Kintsch, 1982; Stein & Trabasso, 1982; Wilkinsonet al., 1980). Reasons for this slower development are debated andwill be discussed later.Knowledge of Text FormsResearch indicates that through experience with various genreschildren develop internalized scripts or schemata which enable themto use appropriate text forms in their writing. Studying the use ofthese forms in children's writing has proven to be a useful method ofdetermining children's competency in different genres.Narrative Several studies have shown that children, even prior toentering school, have well developed internalized scripts fornarrative (Mandler & Johnson, 1977; Stein & Glen, 1979; Stein &Trabasso, 1982). Stein and Glen (1979) had children, aged four andfive, retell a series of stories. Using a story grammar analysis of theconstituents of their stories based on Propp's (1968) work, theyconcluded that children this age tell remarkably well-formed stories.3 9Studies with slightly older children using the same method ofanalysis have demonstrated children's growing knowledge of anarrative schema (Kroll & Anson, 1984; Hidi & Hildyard, 1983;Mandler & Johnson, 1977; Rumelhart & Ortony, 1977; Schallert,1982). Mandler and Johnson (1977), in a widely quoted study, had21 subjects in each of Grades 1, 4, and college retell stories todetermine if there were measurable differences in recall. Theypostulated that the subjects' growing ability to access theirknowledge of story schema would enable older subjects to recallbetter than younger subjects. In their analysis they tabulated thenumber of propositions for each of settings, beginnings, reaction,attempts, outcomes and endings and assessed the total numbers ofmajor constituents (propositions) recalled. They found significantdifferences across age groups in ability to recall. Adults recalledmore than fourth-graders who recalled more than first-gradechildren. A comparison of groups revealed developmentaldifferences in recall. Many subsequent studies supported Mandlerand Johnson's conclusions. Excellent reviews of further research inchildren's comprehension and recall of stories are provided by Steinand Trabasso (1982) and Schallert (1982). These reviews attest tochildren's growing acquisition of narrative schemata and need not beelaborated upon here. Other studies at this time used similaranalyses to study children's growth in ability to write stories. Theirfindings suggest that growth begins at school age and continues untilthe upper intermediate and early high school years (Hidi & Hildyard,1983; King & Rentel, 1983; Kroll & Anson, 1984; Langer, 1986).4 0King and Rentel (1983) studied 36 children's oral and writtenstories as the children progressed from the end of first grade to thebeginning of grade two. Analysis of story structure based on 32functions of character action in stories (Propp 1968) revealed anincreasing range and number of functions as children matured (King& Rentel, 1983). In a study of the writing of 54 Grade 3 students,Kroll and Anson (1984) described children's ability to write coherentstories. They analyzed 25 completed narratives (many wereunacceptable for analysis) using an adaptation of Mandler andJohnson's (1977) story grammar. They found three characteristicpatterns indicating increasing degrees of complexity and developingcompetency in writing. Of the 25 narratives, ten were "linear"(events strung together with no causal connection), nine were"embedded," (described as events not sequentially ordered but timerelated) and six had "mixed linear embedded" patterns. Evidence ofthese patterns indicated variation in the development of storyschema used to facilitate writing.Hidi and Hildyard (1983) compared oral and written discourseof 20 Grade 3 and 22 Grade 5 children. Subjects were asked torespond orally and in writing to introductory statements intended toevoke narrative. Essays were analyzed by two independent ratersfor semantic well-formedness (degree of elaboration), cohesion(correct use of syntax, appropriate use of pronouns and conjunctions)and number of words. While no significant differences were foundon any measure between oral and written stories, ANOVA resultsindicated a significant main effect for grade in written stories; the4 1Grade 5 children produced essays which were better formed andlonger than those produced by Grade 3.Langer (1986) studied the development of narrative writing by67 8-, 11-, and 14-year-old above average children. Hercomprehensive, descriptive study included interviews, think-aloudprotocols, retelling of stories, and written samples. Her analyses ofthe text forms of narratives examined overall length, beginning andending statements, and organizational and structural devices usingtree diagramming of rhetorical predicates which were sets ofrelationships between individual T-units. She found there were age-related effects in length of story writing. Children aged 14 wrotesignificantly longer stories than 11-year olds who wrote more thaneight-year olds. Similarly, older children made more use of structuralbeginning and ending statements in their stories rather than theconventional statements preferred by younger writers. Rhetoricalstructural analysis, however, indicated little growth in the hierarchyof structures used. Langer (1986) found that the broadest anddeepest levels of structures remained relatively constant as childrenmatured which indicated limited development in surfaceinterrelationships. She concluded that limited development inrhetorical predicates may have been due to the fact that storystructures were firmly in place as early as age eight.In summary then, research in the acquisition of the ability torecall and write narrative indicates that there is rapid and continualgrowth during children's elementary school years. Older childrenshow increased maturity in their writing by writing longer, betterformed stories with more mature beginnings and endings. There is4 2also evidence to suggest that this growth in ability to write narrativemay slow as children reach their high school years (Langer, 1986).Bereiter and Scardamalia (1982) summarize the research thus: "fornarrative discourse it seems reasonable to accept the accumulatedevidence that children have well-developed closed schema for it thatcorrespond reasonably well with mature ideas of what constitutes astory and what are its necessary elements" (p. 34).Other Modes of Discourse In an attempt to discover more about and analyze differencesin the acquisition of ability to write in various genres, severalresearchers have studied children's use of text forms in other modesof discourse (Crowhurst, 1983, 1990; Durst, 1984; Newkirk, 1987).Many compare writing in other modes with narrative (Bereiter &Scardamalia, 1982; Crowhurst, 1990; Gundlach, 1981; Hidi &Hildyard, 1983; Kress, 1982; Langer, 1986; Martlew, 1983; Morgan,1984).^These studies indicate that, although children have anunderstanding of genres other than narrative and their competencein writing increases as they grow older, these types of writing appearto be more difficult for them.Newkirk's (1987) study with 34 students in Grade 1, 31 inGrade 2 and 35 in Grade 3 indicates that while children may havedifficulty associated with expository writing, they do have strategiesfor organizing text. Using systems of analysis (the hierarchicalstructures of constituents) based on Langer (1986), Newkirkidentified eight basic structures: label, list-basic, reason list, couplet,attribute-series, hierarchical attribute-series, paragraph basic, and4 3paragraph ordered. The structures represented increasing levels ofcomplexity in the writing. The simplest writing samples werepictures with labels. The next, slightly more complex, samples werein the form of randomly ordered lists, a series of attributes, orcouplets which were statements with an accompanying sentence ofelaboration. More complex samples were specifically-orderedattribute series or were complete paragraphs. While the study islimited in that length was not mentioned and only one piece ofwriting was analyzed, Newkirk found students' writing maturedacross the grades. Grade 1 children used the label category but inGrade 3 none of the students simply labeled. Similarly, attributeseries was commonly found in Grades 1 and 2 and accounted for 21and 26 per cent of the samples but by Grade 3 only 6 per cent of thestudents used this less complex form. Additionally, whereas only 15per cent of first grade pieces were classified as paragraphs, 49 percent were paragraphs in Grade 3. Generally, Newkirk found thatchildren in the upper grades were able to sustain coherentrelationships and when given more opportunities to write expositionsthey made considerable advances in the direction of matureexpository writing (Newkirk 1987).In their summary of work with very young children Gundlach(1981) and Kress (1982) noted the differences in their students' useof text forms such as beginnings and endings and verb tenses innarrative and exposition. They reported that early in their writingdevelopment, children recognize that their stories, for example, musthave a coherent structure with specific beginnings and endings.They often frame their writing with "once upon a time" and "the end"4 4(Gundlach, 1981; Kress, 1982). Whereas beginning writers are ableto provide appropriate beginnings and endings for stories, Kress(1982) reports that young children's expository pieces have, "nobeginning or end" (p. 101). They are unable to use appropriatebeginnings and endings in exposition. Further, he points out thateven at a very early age children write narrative in the past tenseand non-narrative in the present tense indicating their knowledge ofa formal distinction between the two genres (Kress, 1982).Working with Grades 3 and 5 children who wrote narrative andopinion essays, Hidi and Hildyard (1983) found several statisticallysignificant differences between the two genres. They examinedchildren's competency in the use of text forms as measured bycohesion in their writing. Samples written in response tointroductory statements were independently rated on the variablesof semantic well-formedness, cohesion and number of words. Athree-way analysis of variance (grade by mode by genre) on well-formedness indicated significant main effects of grade and genre.The Grade 5 children wrote narrative and opinion essays which werebetter formed than those from Grade 3 children. Additionally, bothgrades were more competent in producing narrative than opinionessays: cohesion scores for narration were significantly higher thanopinion and differences in length were significant for both grade andgenre. Grade 5 children wrote more than Grade 3 children in bothgenres and both groups wrote longer narratives than opinion essays.They found support for their prediction that children have discourseschemata for narrative yet a similar schemata is lacking for opinionessays (Hidi & Hildyard, 1983).4 5Bereiter and Scardamalia (1982) investigated slightly older,middle grade children's knowledge of narrative, argument andexposition. They wanted to see if these children had schemata for"anything besides stories, whether the knowledge [was] accessible toconsciousness and whether they [could] make use of it in planning"(p. 32). They interviewed children and asked them to speculate as towhat kind of elements (constituents) would be found in each of thethree modes. One month later, children were asked to write in eachmode to provide a comparison of usage to naming. They report thatin the aggregate, children were able to name all the elements of story(setting, event, response, plan, attempt, consequence and reaction).Additionally, children were able to use more elements than theynamed. Similarly, regarding elements of argument (beliefs, reasons,elaborations, points of view, conclusions, personal and generalstatements) and exposition (destination, route, orientation,transportation, description, location, arrival complications), subjectson the average used more different discourse elements than theynamed but as a sum total they named all they used. Thus Bereiterand Scardamalia (1982) suggested that "children have potentially atleast conscious access to their discourse grammar knowledge" (p. 30)yet they could find no intra-individual consistency in the naming anduse of elements.A study by Morgan (1984) with first and second yearsecondary students compared their ability to order narrative andexpository pieces of text book writing. Subjects were asked tounscramble the order of sentences in a narrative and an expositorypiece of writing taken from school textbooks as a means of4 6discovering more about their discourse knowledge of the two modes.Subjects were found to have a definite sense of story beginning forthere was unanimity as to which sentence began the narrative. Inthe non-narrative task, the first sentence did not perform anorienting function; therefore none of the students put it first, insteadchoosing a sentence which did orient. Choosing the orientingsentence indicated that students were imposing a narrative formatupon an expository piece of writing. Whereas children had fewproblems discerning the order of and determining which sentenceconcluded the story, they had difficulty structuring and ending theexpository piece. The students had much greater difficultyreproducing the order of an expository text; story structure wasmuch more familiar to them. Morgan concluded that students of thisage were only just beginning to understand simple kinds of discourseother than narrative.Langer's (1986) study of the narrative and report writing ofeight-, 11- and 14- year- olds also revealed differences in knowledgeof the two genres. Differences were revealed in the way thatstudents' structured their writing. Differences in use of text formwere found in the use of many phrases which frame the discourseand grammatical forms in writing such as verb tense. She calledthese phrases and forms in writing "genre markers," and found theuse of them differed among age groups and between genres.^Langerfound that the way children handled use of genre markers such asbeginnings and endings was an important index of their developingcompetence with discourse level tasks. She classified students'beginning and ending statements for both narrative and report into4 7one of three categories: weak, formulaic and structural. Weakbeginnings for both genres failed to establish mode of discourse inany way. Similarly, weak endings did not represent any form ofclosure and were often a last personal statement in narrative or alast fact in report writing. Formulaic beginnings and endings beganand ended pieces in a traditional way; stories began with "once upona time" and ended with "the end" while reports began with "this isabout" and also ended with "the end" or "this has been about."Structural beginnings and endings represented the most mature formof beginning and ending in writing. These statements introduced thework, clearly indicating the type of genre, and were integrated intothe structure of the piece as a whole. Structural endings providedclosure often with reference to beginning statements. While themost popular narrative beginning for eight-year-olds was aformulaic beginning (57.1 percent), this trend was found to decreaseas writers matured: 32 percent for 11-year-olds and 0 percent for14-years-olds. The decline in use of formulaic beginnings wasreplaced by a corresponding increase in the use of structuralbeginnings: up to 75 percent by age 14. While 28 percent of eight-year-old report beginnings were considered weak, most beginningstatements, 50 percent and 44 percent respectively, used by eight-and 11-year-olds were topic starters such as "I know a lot abouthorses" or "Horses are animals." By contrast, most 14-year-olds (72percent) began reports with a thesis statement. Similarly, childrendecreased their use of formulaic beginning statements in reports asage increased. Children's endings to their stories showed moredevelopment across the eight to 14 year age span. In the eight-year-4 8old sample, 50 percent of the children used a formulaic ending either"the end" or "they lived happily ever after" while 14 percentconcluded with a strong resolution based on the theme of the story.By age 14, none used a formulaic ending, 42 percent stopped at anatural break in the action, and 25 percent provided a thematicallydriven resolution. Report endings showed a similar development.The eight-year-olds were most likely to end their reports with a finalfact about their topic (43 percent) or the formulaic ending, "the end"(43 percent). Older, 11-year-olds either had weak endings (30percent) or structural endings, mainly of the evaluative type (38percent). Structural endings predominated in the 14-year-old'ssamples with most conclusions written as summary statements (36percent) and evaluative statements (37 percent). Langer (1986) alsofound differentiation between modes and across age groups in theuse of the verb tense in the writing of her subjects. In general, shefound children treated the past tense as most appropriate for storiesand the present tense as most appropriate for reports. In narrativewriting, 71 percent of eight-year-olds, 76 percent of 11-year-oldsand 83 percent of 14-year-olds used past tense. Report writingsamples indicated 86 percent, 59 percent and 73 percentrespectively of eight-, 11-, and 14-year-olds were written in presenttense.In a later study conducted with students of the same age,Langer (1992) studied the written reports of 16 third grade, 36 sixthgrade and 15 ninth grade students. She reported on the difficultiesthat many students encounter with this genre. She writes, "childrenpresent ideas (often listing descriptions, events, or directions about4 9some general issue or activity) with no overall structure holdingthem together - much like the "and then. . . and then" connectionsyoung children use to move their stories along" (p. 36). She foundthat sixth graders wrote longer papers which used more "developed"vocabulary, syntax, and mechanics than third graders but they oftendid not present their ideas in a coherent or well-elaborated manner.In a descriptive study which examined the narrative,expository and argumentative writing of approximately 28 childrenat each of ages seven, 10 and 13, Wilkinson et al. (1980) describeage-related differences in ability to write competently in each genre.Generally, students across the age groups had little difficulty withand steadily increased their competency in narrative writing. Seven-year-olds wrote simple, active, chronologically ordered stories whichhad little elaboration of detail. Stories written at age 11 were alsochronologically ordered and showed increased attention toelaboration of detail and knowledge of text form of the genre, inparticular, setting, characterization, plot and conventional beginningsand endings. Stories written at age 13 indicated continueddevelopment of elaborateness and attention to text form anddemonstrated a high level of overall effectiveness (Wilkinson et al.,1980). To provide an example of exposition each child was asked towrite an explanation of a simple game. Wilkinson et al. (1980)reported seven-year-olds' explanations of the game were dominatedby partial information. One half of the written pieces offered noexplanations and only one third could be classified as reporting,described as a full account from beginning to end. They concludedthat the task for seven-year-olds was a particularly difficult one.5 0Explanations from 10-year-old writers, described as much moreexplicit, included more descriptive and explanatory sentences.Whereas one quarter of the 10-year-olds gave partial accounts, threequarters gave full chronologically ordered accounts. A considerabledegree of competence marked 13-year-old explanations. Incidenceof partial explanations decreased while full classificatory accounts,which included details with concrete examples and explanations,increased. They also noted that chronological ordering of texts wassuperseded by classificatory order indicating growing knowledge ofform and ordering of this mode of discourse (Wilkinson et al., 1980).The argumentative task required children to defend whether or notthey could do what they liked at school.^Results were similar tothose in the explanatory task. Fewer than one third of seven-year-olds stated their position or gave a reason to support it. Further,reasons were highly personal and context-bound when given. Also,few seven-year-olds wrote more that two sentences which indicatedthis was a particularly difficult task for this age group (Wilkinson etal., 1980). Ten-year-olds also experienced difficulty with this mode.Three-quarters of the 10-year-olds stated a position elaborated withone or more reasons, most often personal, but only three out of 31had a generalized conclusion. By age 13, however, almost allstudents stated a position and substantiated it with logical reasons.Further, over two-thirds of their texts contained concludinggeneralizations. Generally, Wilkinson et al.'s (1980) cognitivemeasures of drawing inferences, generalizing, hypothesizing andspeculating showed increases from age group to age group. Theyfound 13-year-olds drew more inferences and generalized more5 1often than 10-year-olds who inferenced and generalized more thatseven-year-olds. The same pattern emerged with the two othercategories of hypothesizing and speculating. Older students werefound to elaborate upon context more, and generalize andhypothesize more than younger children (Wilkinson et al., 1980).Unlike many of the descriptive studies conducted at this time,Crowhurst (1983) used an experimental design to quantifydifferences in students' ability to use appropriate text form inpersuasive compositions. She used an adaptation of Wilkinson et al.'s(1980) cognitive measure for argumentative writing namely,reporting, generalizing, hypothesizing and speculating. She studiedthe writing of 40 students in each of Grades 5, 7, and 11. Significantdifferences were found across grades, particularly in the category ofreporting (which reflected either narrative writing or reporting on"what is happening now" rather than persuading). Grade 11 studentsused significantly less of this type of writing than students in Grades5 and 7. Also, 17 out of 40 fifth-grade children, when asked to writepersuasive compositions, wrote narratives on two or more occasionsout of six. Additionally, seven out of 40 seventh-graders wroteentirely in narrative indicating not only a lack of schema forpersuasive discourse but the pervasiveness of narrative schema onthe ability to write in other genres.In a later report, Crowhurst (1990) summarized thecharacteristic patterns of over 1200 argumentative compositionswritten by Grades 5, 6, and 7 students on a variety of topics inseveral different studies (1978, 1980a, 1980b, 1983, 1987). Shefound the majority of students wrote recognizably argumentative5 2pieces in response to assignments given in this mode. She also found,however, that many of the argumentative compositions deviatedfrom expected forms of the genre indicating students' difficulty withand lack of schemata for argument. A number of students, forexample, wrote non-arguments - either narratives, descriptions ordialogues instead of the required form - a finding which otherresearchers had noted in their studies (Gorman et al., 1988; White,1989). Also, students often wrote essays which were relatively briefand had little elaboration of topic. Some resembled a series of wishesor read like a list, a structure which characterizes younger students'argumentative writing (Newkirk, 1987). Specific features ofcharacteristic deviations from conventional argument also involved alack of structural organization; few compositions were organized intoparagraphs and many lacked conventional beginning and endingstatements. She found students had difficulty starting their essays,often beginning with statements such as "I think" or "no, I don'tthink." Students had equal difficulty ending argumentative pieces,either failing to include an ending statement or using anunconventional one. Concluding statements were most likely to besome kind of an appeal or a single statement repeating the writer'sopinion rather than a conventional ending statement (Crowhurst,1990).In the United Kingdom, major assessments of writing wereconducted from 1979 to 1983 by the Assessment for PerformanceUnit of the National Foundation for Educational Research (Gorman etal., 1988). As part of the study, they used holistic and analyticratings to assess 11- and 15-year olds' performance on53argumentative and descriptive/comparative writing. For theargument task, students were asked to select a strongly heldpersonal opinion and to persuade another person to share it.Students' arguments were assessed on a standardized scale ofimpression scores based on structures such as the following: content,organization, knowledge of stylistic conventions, knowledge ofgrammatical conventions, and knowledge of orthographicconventions. They found that 11-and 15-year old students'performance in argument was comparable to performance across alltasks (i.e., narration, argument, descriptive/comparative essay andselected writing from regular classroom activities) as evidenced bydistribution scores that were close to average for both groups. Therewere differences, however, in the methods the two groups used intheir arguments. Fifteen-year-olds used more written conventionssuch as beginning and ending statements and patterns oforganization which indicated knowledge of appropriate text form ofthe genre. On the other hand, 11-year-olds often relied upon spokenlanguage conventions basing their writing on their knowledge of howarguments are managed in spoken exchanges. Gorman et al., (1988)also noted several difficulties which students in both groupsexperienced while writing argument. Over 50 scripts presented inthe qualitative analysis indicated several difficulties found in non-narrative writing. Generally, students had difficulty generatingcontent for their writing. Also, they often relied on chronologicalordering, typical of narrative and reporting styles, indicating theiruncertainty as to the organization of argumentative writing.Similarly, younger children lacked the variety of connectives needed5 4for coherent argument, often monotonously overusing "and," "or," and"then." Some writers found it difficult to sustain a consistent opinionthroughout their writing, frequently stating unrelated sets ofopinions or changing their opinion midstream. Further, withuncertainty of form of genre, grammatical clarity and coherencesuffered; students failed to use paragraphs and few provided titlesor contextualized openings, beginning instead with "I am going to tellyou about" or "I feel that" (Gorman et al., 1988). Similar results werefound for a task which asked the same students to compare twoinsects. Gorman et al. (1988) found that students were able to writedescriptive/comparative essays but structuring the genre createddifficulties for them. Problems with selection and organization ofcontent and appropriateness of style marked many of the scriptswritten by both 11- and 15-year-olds. Often students presented anunfocused listing of similarities and differences of the two insects(they were to compare the two insects) or provided vague, inaccuratedescriptions. Stylistic problems included lack of any comparativeframework indicated by the absence of beginning and endingstatements and problems with marking sentence boundaries. Whilethe students seemed to have difficulties in writing argument anddescriptive/comparative essays in this comprehensive study, theresearchers found they had a basic grasp of the requirements of thegenres.. Gorman et al. (1988) therefore attributed the students'difficulties in written argument and descriptive/comparison not toan inability to understand the genre but to a lack of instruction inuse of the conventional forms associated with the genres and overalllack of experience with these modes of discourse in their schooling.5 5Durst (1984) and Applebee (1984) also provided descriptiveaccounts of non-narrative writing through the examination of highschool students' collections of elementary school and high schoolwriting. They underscore several difficulties students encounter asthey develop schema for informational writing in the form ofthesis/support essays, comparison/contrast essays, reports, andinferential reports. Durst (1984) examined the collections of writingsof three students from their third through twelfth grades. Organizingthe collections into English, Social Studies and Science related topics,he traced the development of these students' informational writing:report, argument, critical essays. Though the limitations of a select,very small sample size must be considered, Durst's work offers someinsight into the development of knowledge of structuring thesemodes. He found incidence of informational writing such as report,argument or critical essays rare in the elementary school. Mostwriting of this kind was report writing which was often a summationof information organized chronologically. Students were often givena format for writing reports which provided the framework for theirwriting. As students entered high school, however, they wereexpected to write other forms of informational writing which tendedto create difficulties for them. Durst (1984) found that studentsinitially relied on narrative structure when asked to writecomparison/contrast essays, arguments or interpretive summaries;texts were chronologically ordered, written as narratives orsummaries and were only occasionally interspersed with explanatoryor interpretive comments. Nonetheless, students gradually movedfrom narrative and summary forms to more logical analytical forms5 6such as interpreting and hypothesizing. Characteristic features of thetext of narration such as chronological ordering were replaced by awider variety of organizational patterns within the written pieces(Durst, 1984).Applebee (1984) described the development of informationalwriting after completing a 16 month study of 15 students: six inGrade 9 and nine in Grade 11. Through his qualitative analysis ofwriting samples, interviews, and observations of the students in theclassroom, he posited several generalizations as to the students'growth of knowledge of this kind of writing, that is, argument,critical essay and report. He found there were clear patterns in thedevelopment of informational writing at the high school level.Different students adopted similar strategies for coping with thedifficulty of new structures in various forms of informational writing.Applebee found most students wrote fluent narratives when theyentered high school. Thus, as they began using writing to analyze orexpress opinion, they often initially included long stretches ofnarration within the global analytic frame. When confronted with anargumentative task, he found less experienced writers oftenrandomly listed a series of points rather than organizing pointsaround a related idea. Similarly, some argument samples containedimbalanced elaboration; some parts of the essay were explained atgreat length while others remained unexplicated. He described thistime as "a time of transition from reliance on primarily time-orderedand descriptive methods of presentation toward more analytic modesof organization" (1984, p. 185). Applebee (1984) also surmised that5 7students' writing skills may have been limited as a result ofinsufficient experience with these forms of writing.The results of these studies suggest that children of variousages have knowledge of genres other than narrative but that thisknowledge is either not fully developed and/or not accessible(Applebee, 1984; Bereiter & Scardamalia, 1982; Crowhurst, 1990;Durst, 1984; Kress, 1982; Gorman et al., 1988; Langer, 1986;Newkirk, 1987; Wilkinson et al., 1980; White, 1989). Further, lackof development and apparent inability to access knowledge of modesother than narrative manifest themselves in students experiencingdifficulty in structuring non-narrative discourse competently;students have difficulty reproducing the appropriate form andconventions of these modes.Length of Discourse Length of discourse, another aspect of students' form of writingtraditionally used to distinguish competency and maturity, alsopoints to the difficulties children encounter when writing in variousgenres. Research provides evidence that, as children mature, theirwriting increases in length (Crowhurst, 1978, 1980, 1983; Harpin,1973; Golub and Frederick, 1970; Harrel, 1957; Langer, 1986).Similarly, length has been found to significantly increase withinmodes and across modes as children mature (Crowhurst, 1978, 1980,1983, 1987, 1990; Hidi & Hildyard, 1983; Langer, 1986; Perron,1977; San Jose, 1972). Further, children's narratives are often muchlonger than their writing in other forms of discourse perhapsindicating less difficulty with narrative than other genres.5 8In her study of Grade 4 children's writing, San Jose (1972)found mean number of words higher for narratives (441) thandescription (361), exposition (309), or argument (303). Similarly,Perron (1977) found significant differences in length between modesfor his Grade 3, 4 and 5 writers. Grade 3 writers had mean lengthsof 75 words for narration and 33 words for description, Grade 4'shad mean word lengths of 470 for narration and 118 for descriptionand Grade 5's wrote 628 words on the average in their narrativesand 157 words in descriptions.Crowhurst (1978) found differences in mean length in herstudy of the narrative, descriptive and persuasive writing of 40Grade 6 and 10 students. The Grade 6 students had mean numbersof words per piece of 465.5, 397.5 and 386 respectively fornarration, description and argument while the Grade 10 studentswrote pieces of 574.5 words in narrative, 503 words in descriptiveand 461 words in argumentative writing. The data indicatedifferences occurred not only in grade by grade comparisons butmode by mode comparisons.Hidi and Hildyard (1983) found length of piece increased forboth narration and opinion essays as writers matured and thatnarratives were significantly longer than opinion essays. Grade 3writers produced an average of 65 words in narratives and 21 wordsin opinion essays and Grade 5 writers had mean lengths of 125words in their narratives and 43 words in opinion essays.Langer (1986), in her study of the narrative and report writingof students aged eight, 11, and 14, found that eight-year-oldsproduced an average of 108.26 words in narratives and 72.61 words5 9in report. Eleven-year-olds had mean lengths of 202.19 and 175.61words for stories and reports respectively, but 14-year-olds hadmeans of 225.27 and 243.86 respectively. For ages eight and 11, thechildren's stories were longer than their reports suggesting thatstorytelling was easier relative to report writing; by age 14, however,this difference had disappeared and perhaps even begun to reverseitself (Langer, 1986).Thus, there is evidence that length as well as text forms inwriting (such as beginnings and endings, cohesion, length andordering) are affected by mode of discourse. The research indicatesthat throughout their elementary school years, children write longernarratives than expositions, arguments or reports (Crowhurst, 1978;Hidi & Hildyard, 1983; Langer, 1986; Perron, 1977; San Jose, 1972).Further, there is evidence to indicate that this difference in length ofmodes continues in high school (Crowhurst, 1978). There is alsosome evidence, however, that indicates that as children enter highschool their narrative texts are not longer than in other modes ofdiscourse (Langer, 1986). Reasons for the differences in length ofchildren's writing in narrative as compared to other genres areunclear; however, it would seem that they may be attributable tostudents' lack of difficulty in narration as compared to other forms ofdiscourse.In their descriptive and qualitative studies, several researchershave hypothesized as to the reasons children experience difficultiesin modes of discourse other than narrative (Bereiter & Scardamalia,1982; Clay, 1983; Christie, 1986; Dillon, 1981;^Harris & Wilkinson,1986; Martlew, 1983; Martin & Rothery, 1986; Taylor & Beach,6 01984). Some contend that it is the students' lack of experience withvarious genres that inhibits their growth in ability to write(Applebee, 1984; Christie, 1986; Gorman et al., 1988; Martin &Rothery, 1986; Newkirk, 1987; White, 1989).Newkirk, who demonstrated that young children makeconsiderable advances in the direction of mature expository writing,challenges the model which views children's writing as deficientadult writing and that children's inability to write is due to "cognitiveoverload." He suggests that schools have excluded serious instructionin expository writing, leaving this instruction until secondary schoolyears, a point that is also argued by Applebee (1984), Gorman et al.,(1988) and White (1989). Similar views are expressed by Martinand Rothery (1986) who observed that very young childreninstinctively write recounts and reports (a description of someobject) in the initial stages of writing. They maintain that teacherstreat this early report writing as "less prestigious" than narrativewriting. They argue that teachers who are willing to "break traditionby accepting and even encouraging writing in a genre besides thevicarious experience narrative [which] will undoubtedly affectchildren's abilities to respond to a variety of written languagesituations as future adult writers" (p. 262).Christie (1986) argues that failure to master the skills neededto produce varied patterns of discourse is a result of teachers' lack ofguidance when their students are learning these patterns. He feelsteachers must design programs that will enable children to gainaccess to the various forms of discourse and teach them from anearly age. Other researchers point to children's lack of personal6 1experience and knowledge of content that makes writing in othermodes difficult. Clay (1983), who described the structure of primarychildren's writing, posits that story writing is easier because childrencan write on the basis of their own feelings and experiences andfrom an awareness of story. Non-fiction, however, requirestruthfulness and logic which create problems for the young writer.Thus she concludes, "children are more likely to have a scatter ofideas and be casting around for order and structure in writing of thiskind" (p. 270).Many researchers point to the increased cognitive demandsthat non-narrative writing places upon writers and their general lackof knowledge of these types of discourse. Martlew (1983) considersthe young child's needs for immediacy and personal significancewhen writing. She notes that the impersonal nature of expositorywriting precludes these needs, requiring explicit and unambiguouswriting which increases the cognitive demands of the communicationtask.^This increased cognitive demand, in turn, creates difficulty foryoung children. Commenting on their work with middle gradechildren, Bereiter and Scardamalia (1982) found that narrativewriting is well established by age eight but expository writing, due toa lack of knowledge of the discourse, develops much later. Similarly,Taylor and Beach (1984) report on adolescents' continued difficultywith expository writing, stating, "in comparison with narrativewriting, however, adolescents have difficulty comprehending andproducing expository texts" (p. 135). They postulated thatdifficulties were associated with high levels of abstraction,transforming thoughts into conceptually integrated knowledge, and6 2lack of detail (Taylor & Beach, 1984). Difficulties in writing non-narrative texts persist even at the college level. Dillon (1981), in hiswork with college students, comments on the "developmentalcommonplace that we master narrative both as readers and writersbefore we master the more abstract analytic patterns" (p. 61).Further, many researchers state that the structure of non-narrativediscourse presents problems for the writer, in particular, the lack ofchronological ordering of the discourse (Harris, 1980; Kress, 1982;Perera, 1984). These genres require high levels of abstract thinkingin order to attend to the overall organization of the writing.Thus the theoretical description in these studies indicates thatwriting non-narrative discourse requires experience, accessibilityand growth of knowledge, and maturity in cognitive processes.Research indicates young children have knowledge of the form ofvarious genres but lack the experience and guidance needed todevelop this knowledge (Christie, 1986; Martin & Rothery, 1986;Gorman et al., 1988; White, 1989). Further, children must developthe ability to access their knowledge and experience in order to usethe appropriate structure, convention and form for the genre(Bereiter & Scardamalia, 1982; Clay, 1983; Kress, 1982; Kintsch,1982). As Kintsch (1982) contends, students must have, "arepresentation of content plus a macrostructure (organizational planor gist) that is determined by text specific organizations" (p. 100).Similarly, growth in cognitive processes is needed to enable studentsto attain the level of abstraction needed to transform thoughts intoan organized, coherent piece of writing (Bereiter & Scardamalia 1982;Dillon, 1981; Martlew, 1983; Taylor & Beach, 1984). While these63theories are posited, conclusions as to whether differences indevelopment of writing in genres other than narrative arepedagogical or developmental are, as yet, unclear.Summary and Considerations of the Present StudyHistorically, research in the development of children's writingability was primarily based upon the measurement of syntacticcomplexity. Extensive analyses of the grammatical structures usedby children provided reliable information as to their growing controlover written language. Early analyses, however, proved time-consuming and expensive and eventually many researchersabandoned elaborate analyses in favour of the much simpler andequally reliable measurement developed by Kellogg Hunt, the T-unit.Similar results over a wide variety of age groups indicatedstatistically significant increases in T-unit length accompaniedadvances in age. While use of T-unit length has been notablysuccessful, its limitations must also be recognized: it cannot beequated with quality in writing; it is affected by mode of discourse;it may obscure important structural differences in the writing ofgroups of children; used alone it is not reliable as an indicator ofmaturity in a child's linguistic development.The T-unit's limitations dictate that it must be used with somecaution yet, despite its lack of precision, it continues to be used inresearch as it is "the most useful and useable index of syntacticdevelopment over a wide age range" (O'Donnell, 1976, p. 38). The6 4T-unit, then, is used with confidence in this study as a generalindicator of maturity in children's writing.In order to lend support to the T-unit as a measure ofdevelopment this study also draws upon traditional syntacticresearch which has outlined specific grammatical features indicatingmaturity in children's writing, namely, the use of extended nounphrases, nominal clauses and elements of the verb phrase. Theliterature suggests that comprehensive analyses of syntax within theT-unit reveal statistically significant age-related changes in the nounphrase structure. Older children use significantly more prepositional,appositive, and participle phrases, more infinitives and adjectives,and incorporate more relative clauses than younger children.Similarly, as children age, they incorporate more subject andcomplement nominal clauses. Changes in the use of the finite verbphrase also accompany development in writing. Research examiningverb phrase formation indicates children use more perfect, passiveand less progressive forms in their writing as they grow older.Increased use of modals generally, and modals accompanyingcatenative forms, are increasingly evident in more syntacticallycomplex writing. Little research has been conducted recently whichexamines these variables especially in writing process situations. Itis hoped that measurement of these variables in the present studywill provide supporting data as to individual development ofsyntactic complexity. While there is some evidence in the literaturewhich indicates these grammatical elements are affected bysituational factors, it is at best inconclusive. It appears that controlover mode of discourse is important to the design of a study6 5examining this factor and thus it is accounted for in the presentstudy.More recently, researchers have examined differences in thedevelopment of writing in various genres. Investigations haveattempted to describe children's knowledge or schemata of certaingenres as evidenced by their use of text forms such as conventions,organizational patterns and length in writing. While it is apparentchildren have well-developed schema for narrative even at an earlyage, comparable schemata for other modes appear less developed.Similarly, there is rapid and continual growth in ability to writenarrative through to the secondary school level yet growth in othergenres seems comparatively slow. Further, this apparently less-developed schemata and slower development are manifested inchildren experiencing difficulty in structuring their writing. Studiesexamining the expository, argumentative and descriptive/comparative writing of students from elementary to high schoolconfirm several difficulties with these modes: lack of knowledge anduse of conventional beginnings and endings; inability to achieve anadequate overall organization; difficulty sustaining coherent, logicalanalyses or argument; maintaining the juxtaposition of conflictingopinion. In addition to the problems students experience structuringnon-narrative writing, they also have difficulty generating content.Comparative studies of children's narrative and non-narrativewriting indicate children write significantly longer narratives thanother forms of writing. Some work in the field has attempted toquantify certain structural differences in children's writing,particularly use of rhetorical structures, beginning and ending6 6statements, and control over verb tense (Langer, 1986). This workhas indicated that there are important, age-related differences in theuse of these variables. Accordingly, this study examines similarvariables as a means of complementing syntactic measures todescribe children's writing more precisely.Reasons for differences in ability to write narrative ascompared to other modes of discourse have been posited by anumber of researchers.^Some theoretical assumptions are based onthe premise that students lack the experience with other genre; thatis, their schooling has excluded instruction in the appropriateconventions, forms and organization. Others are based on thestudents' general lack of knowledge and personal experience whichappear to create difficulty with form of genre and content. Stillothers base their assumptions on the immature cognitivedevelopment of the students maintaining that high levels ofabstraction needed to organize the discourse are not yet developed.The literature suggests that knowledge of the essential formsand conventions of various genres is a major requirement forcompetence in writing. This knowledge, in turn, must be used inconjunction with knowledge of surface structure or syntax to governthe production of text.^It is the purpose of this study to extend theexisting research by providing information on the acquisition ofability to write in the genres of narrative and report byincorporating previous techniques used to measure maturity inwriting, specifically elements of syntax and text form. It is also theintent of this study to supplement this data using qualitative designprocedures; that is, with description obtained from observations in67the naturalistic setting of the classroom. Previous studies of thiskind are relatively scarce (e.g. Applebee, 1984; Crowhurst, 1990;Kress, 1982; Gorman et al., 1988; Wilkinson et al., 1980) in theliterature and, as this study is examining these variables andconducting observations using the same subjects over a three yearperiod, it is unique.68CHAPTER IIIDESIGN OF THE STUDYThe purpose of the study is to examine the narrative andreport writing of the same 14 students through Grades 4, 5 and 6. Itis postulated that data from this inquiry will reveal differences inthe use of syntax and text forms by a group of children with varyingabilities over a three year period.Subjects of the StudyThe subjects of the study were selected in 1989 from twoGrade 4 classes in a lower to middle class area of Richmond, a suburbof Vancouver, B. C. The initial sample of 49 students was reduced to30 students who used English as their first or preferred language,had scored average or above average in reading and language instandardized tests and had received their parents' consent. The 30students were divided into groups based on ability (high ability,above average ability and average ability according to teacherevaluations) and gender (male and female). To ensure arepresentative sample of ability and gender, a subset of 14 students(seven boys and seven girls) was randomly sampled from thesegroups. Of the 14 students, 10 had attended the school sincekindergarten, two had attended since Grade 1 and two since Grade 3.At the outset of the study in September of 1989, the children rangedin age from 9.0 to 9.8 years. All 14 students had generalperformance scores in Language Arts on recent standardized tests6 9that were equal to or above the national average for their age level.The tests scores were taken from the Gates McGinite and CanadianBasic Skills Tests given to the students in either Grade 2 or 3. Thestudents' scores are provided in Appendix C. While most of thestudents spoke English as a first language, two did not. One spokeFilipino and the other spoke Cantonese. Both of these students,however, used English as the preferred language at home. Allchildren but one were born in Canada; 11 in Richmond, one inEdmonton, one in Saskatoon and one was born in the Philippines.Procedures of the StudyThere were two stages of procedures which were necessary inorder to conduct the study: the initiation of the study andconducting the writing sessions.Initiating the StudyIt was necessary to obtain permission from the appropriateindividuals and agencies having jurisdiction over the subjectpopulation and researcher. Two principals and seven teachersinvolved in the study gave their consent and provided continualcooperation throughout the three year period. Written consent wasobtained from both the Richmond School Board and the University ofBritish Columbia Behavioural Sciences Screening Committee forResearch and Other Studies Involving Human Subjects. All writtenrequests for permission are provided in Appendix D. All of theoriginal 49 Grade 4 students perceived they were a part of the study7 0for the three year period. During the last two years of the study,however, some of these students were in combined age classes forGrades 5 and 6. In these cases, all of the students in the combinedclasses participated in the writing sessions. The children in the gradelevel other than the subjects became aware they were not a part ofthe sample group but this did not appear to affect the subjects in anyway.Conducting the StudyTraditionally, students in this school are taught to write using awriting process approach as part of their writing program (Graves,1984; Burke et al., 1980). Once their writing has progressed from anemergent to consolidated stage (Ministry of Education, 1990) in theearly primary years, students learn to draft, edit, redraft andproofread theirs and others' work as part of the writing process.Generally, students choose their writing topics and have as muchtime as is needed to complete their work. At the time of the study,the school was instigating integrated programs of study at everygrade level. Teachers planned integrated units based on themeswhich incorporated all curriculum areas. Thus, students' writingoften became part of the integrated theme that the class wascurrently studying. As a result, topic restrictions were often placedupon the students' written work, a practice which did not reflectcurrent educational theory. This theory espouses the need forstudents to choose their own topics during the writing process inorder that they exercise the strongest control over their writing(Bereiter & Scardamalia, 1986; Graves, 1984; Martin & Rothery,7 11986).^Similarly, due to the nature of the program, students oftenhad time restrictions placed upon their writing. Pieces had to becompleted within a certain time limit, a practice which is notconsistent with writing process approaches to writing (Graves, 1984).Research has also demonstrated the need to control for mode ofdiscourse when analyzing writing samples (Crowhurst, 1978).Comparisons of writing should not be made across genres. Therefore,it was important to have students write in specified genres.In order to provide the students with a writing situation thatwould not impose topic or time restrictions and that would controlfor mode of discourse, the researcher became a regular part of thestudents' classroom situation. During the first year, on every Fridaymorning from the end of October until the end of May the researcherconducted writing sessions of 45 minutes duration with each of thetwo classes. Once this time was established it continued for theensuing two years. The only exception was when the subjects werein Grades 5 and 6 and the researcher worked with three classesinstead of two due to combined age groupings.The researcher, who was a teacher at the school, was quite wellknown to the students and was easily assimilated into the classroomsettings. The classroom teacher was welcomed to stay for the writingsessions but seldom did, preferring instead to use the time formarking or preparation. All writing was done in the writing sessions.In the beginning of the study, the researcher gave identicalexplanations to the classes which included information about theproject, a review of the writing process, and an outline of theexpectations for writing. Students were told to draft, edit and7 2rewrite all writing in order to provide their best work. They had notopic or time restrictions imposed on their writing but were requiredto write in two genres: narrative and report. Regular classroomprocedures for writing were followed, such as brainstorming ideas,discussing students' ideas for writing and providing support andencouragement, with the exception that neither the researcher northe teacher provided the students with reference material or help inlanguage usage or structural aspects of their writing such asbeginnings, endings or organizational patterns. Each student took asmuch time as needed to complete a piece of writing.The students began each year with a narrative. The researcherasked them to "write a story about any topic you like." After all thestudents had completed at least one narrative, the researcher thenasked them to write a report using the following statement: "writeabout something you know a lot about." The students continuedalternating narrative and report writing throughout the year keepingthe two genres distinct. The students submitted all pieces of writingto the researcher, but chose only one report and one narrative perterm to be photocopied for analysis. The piece of writing that theychose was one they felt was indicative of their best work. Thestudents were told all work would be reviewed by the researcherand then returned to the classroom teacher for assessment. Actualassessment was at the discretion of the teacher. The students werenot aware of the types of analyses to which their writing would besubjected.To satisfy the need for feedback on the students' work, theresearcher commented in writing on every submitted piece of7 3writing. The researcher commented on the meaning of the piece andprovided overall positive comments to each student. Care was takenthat all students, subjects or otherwise, were treated equally. For all14 subjects in the study, one narrative and one report werephotocopied in late November and in early March in each of threeyears. Thus four samples of writing per student per year generated168 writing samples for analysis. In addition to the samplecollection, whenever pertinent, the researcher recorded generalobservations about the writing sessions and individual students.Analyses of DataInitially all samples of writing were read for coherency beforethey were subjected to syntactical and structural analysis. Inaccordance with the segmentation rules for T-unit analysis outlinedby Crowhurst (1978), which are provided in Appendix A, all tagquestions, incoherent language, interjections, sentence fragments,and colloquial expressions were struck out (although were includedin overall length tallies). In almost all cases no words wereeliminated. The text was then analyzed using the followingprocedures.Syntactical AnalysisSeveral features of the students' writing which have beenfound to indicate syntactic complexity were used in the analysis:T-unit length, extended noun phrase modification, subject andcomplement clauses, and extension of the verb phrase. Syntactic7 4analysis was used to provide a comparison of students' writingwithin each year (comparing November and March samples), overthe three years and between the two genres of narrative and report.All pieces of writing were coded in the following order: a) total wordcount, b) T-unit analysis, c) colour coding of all noun phrases, d)coding of all relative, subject and complement clauses, e) colourcoding of all verb phrases (separately for main and subordinateclauses). After all the students' writing samples were analyzed andthese features labeled, mean T-unit length and frequency counts per100 words of each of the grammatical structures were calculated.Descriptions of these grammatical structures follow:T-unit LengthAfter all incoherent language, tag questions, sentencefragments and interjections were eliminated, each piece wassegmented into T-units. The mean T-unit length for all 14 students'November and March narrative and report samples for each of threeyears were calculated.The Noun PhraseFurther analysis involved examining the extension of the nounphrase, which has often been used as a measure of syntacticcomplexity. The use of extended noun phrases, that is those whichinclude four or more words, is one indicator of maturity in writing(Golub & Frederick, 1970a; Hunt, 1965; O'Donnell et al., 1967;Perera, 1984). Noun phrase analysis in this study was based onHunt's (1965) work, which outlined several categories for analyzingextended noun phrases: noun phrases with modifiers, noun phrases7 5with prepositional phrases, noun phrases with appositive phrases,and noun phrases with relative clauses. Noun phrases with relativeclauses were further subcategorized into four levels of maturity asoutlined by Perera (1984, p. 236). The writing samples wereanalyzed for each of these categories and frequency countsdetermined. Separate frequency counts were calculated forNovember and March samples in both narrative and report.Subject and Complement Clauses As children age, they incorporate more subject andcomplement clauses into their writing (Harpin, 1976; Hunt, 1965;Loban, 1976; O'Donnell et al., 1967; Perera, 1984).^Thus, thenumber of subject and complement clauses in each piece of writingwas determined. Then, frequency counts per 100 words of eachclause type were used to compare students' writing over the threeyears in the two genres.The Verb PhraseAs children mature they make use of longer and more complexverb phrases in their writing (Golub & Frederick, 1970a; Hunt, 1965;O'Donnell et al., 1967 Perera, 1984). Perera (1984) writes, "youngchildren use simple active verb phrases. As they grow older theytend to use more sophisticated verb phrases generally through theextension of the verb auxiliary" (p. 54). In this study, nine separatecategories from Perera's (1984) work were used to analyze the verbphrase: modal auxiliary, do auxiliary, catenated, simple, progressive,passive, perfect, perfect progressive, and perfect passive. Thesecategories were used to study the verb phrase in both main clauseand subordinate clauses in students' writing. Separate frequency7 6counts per 100 words of each category were calculated for main andsubordinate clauses in all pieces of writing. This further examinationof the verb phrase was to test the researcher's supposition thatstudents will increase their use of extended verb phrases insubordinate clauses as they mature.To provide an overall picture of the use of the verb phrase in agrade-by-grade comparison, an additional analysis of the verbphrase within each genre is provided. The percentage of type ofverb per total verb use within each sample was calculated. Themeans of these percentages for each genre for each grade were thenused to compare general verb usage in narrative and report writingover the three year period.Analysis of Text Forms Subsequent analyses looked at the text forms children use intheir writing, particularly beginning and ending statements andoverall length. Several researchers have recognized that studentsstructure their writing in recognizably different ways and indifferent ways for different modes of discourse (Bereiter &Scardamalia, 1982; Crowhurst, 1990; Gorman et al., 1988; Gundlach,1981; Kress, 1982; Langer, 1986; Wilkinson et al., 1980).Beginning and Ending Statements The use of certain kinds of structural statements which framethe discourse such as beginning and ending sentences indicatematurity in writing (Crowhurst, 1990; Langer, 1986). Langer (1986)developed a system of categorizing beginning and ending statements7 7to show differing levels of maturity in children's writing. Sheclassified statements in order of increasing difficulty as eitherbelonging to weak, formulaic or structural categories. Each of thesecategories was subcategorized into three or four types of weak,formulaic, or structural beginnings and endings. All beginning andending statements in the narrative and report writing samples wereclassified according to the types used in Langer's work (see AppendixB).Length of Writing SampleResearch provides evidence that as children mature, theirwriting increases in length (Crowhurst, 1978, 1980, 1983; Harpin,1973, 1976; Golub and Frederick, 1970a; Langer, 1986). Further,there is some indication that length is affected by mode of discourse(Crowhurst, 1978, 1980a, 1980b; Hidi & Hildyard, 1983; Langer,1986; Perron, 1977; San Jose, 1972). Therefore, total word countwas determined for each student's samples of narrative and reportwriting as this index might provide useful information to comparestudents' writing competency in the two genres over the three years.Interrater ReliabilityTo provide a reliability check for the analyses, three separateraters were used. Two independent raters who were trained in theprocedures for coding checked the T-unit analysis and the analysis ofsyntax. To obtain a representative sample of a little more than tenpercent (18 samples), the following procedures for random samplingwere conducted. All students were ordered according to analphabetic listing of their first names. Their compositions were7 8ordered chronologically alternating report and narrative for eachyear and numbered one to 160. Eighteen numbers were selectedfrom a list of random numbers, once for T-unit analysis and once foranalysis of syntax. The compositions that corresponded to thosenumbers were checked by the raters.A teacher, who has completed her master's degree in a relatedfield, conducted the T-unit analysis. She had a sample of eightreports (two Grade 4, four Grade 5, and two Grade 6) and 10narratives (three Grade 4, four Grade 5, and three Grade 6). Out of atotal of 517 T-units the raters agreed on 511, or 99 percent.The analysis of syntax was completed by a doctoral candidatein the area of linguistics and second language acquisition. He studieda total of 11 reports (two Grade 4, five Grade 5, and four Grade 6)and seven narratives (two Grade 4, two Grade 5, and three Grade 6).He examined all forms of the extended noun phrase includingrelative subject and complement clauses. The percentage agreementwith the principal rater for noun phrases was 88 percent: out of 223occurrences the rater agreed on 196. The verb phrase analysis wasslightly higher at 91 percent: the raters agreed on 753 out of 830incidences.A third rater was trained to test the reliability of the coding ofbeginning and ending statements. This rater, who was a teacher atthe same school, examined all beginning and ending statements.Both the rater and the researcher agreed that the subjectivecategories in the classification system used (Langer, 1986) weresometimes difficult to apply to the students' writing. Nevertheless,decisions were made as to the beginning and ending statements' 'best7 9fit.' Interrater reliability decreased slightly over time as thestudents' sentences became more complex. In Grade 4 the ratersagreed on 94 out of 102 samples, or 92 percent; in Grade 5, 91percent (94 out of 104 sentences); and in Grade 6, 89 percent (93 outof 104 sentences). An overall average of agreement for the threegrades was 91 percent.80CHAPTER IVRESULTS OF THE STUDYThe purpose of this study was to examine the complexity of 14students' writing samples using analyses of syntax and text form. Itis postulated that, as the students mature, they will use differentelements of syntax and different text forms in the two genres ofnarrative and report. The findings of these two analyses aresummarized in the text and tables that follow.Syntactic ComplexityThe study examines three syntactic elements of the students'writing, in particular, T-unit length, extended noun phrases, and verbphrases. The findings highlight several interesting comparisonsbetween the genres of narrative and report in the students' writing.T-unit LengthThis study initially examined mean T-unit length in a grade bygrade comparison. In narrative, the mean T-unit length for the 14students in Grades 4, 5, and 6 remained virtually the same at 9.55,9.18, and 9.85 respectively. The means in Table 1 provide guidelinesby which to examine these students' work. These T-unit lengths aresimilar to those found in Table 1. The means for report show longerT-unit lengths and overall increases from Grades 4 to 6. Mean T-unitlengths for report were 10.24, 11.30, and 12.13 which appear to be8 1slightly higher than the means for exposition in Table 1. They arealso higher than the students' means in narrative writing. Grademeans of T-unit length, however, may actually obscure some veryinteresting features of individual student's growth in the use ofsyntax in narrative and report writing. Further, these means mayalso obscure differences in syntactic complexity between the twogenres.NarrativeIn narrative writing, for example, seven out of 14 students (#2,#3, #4, #6, #7, #9, #12) produced their longest T-units in Grade 4.(See Table 2.) T-units lengths, for all eight students, ranging from10.09 - 16.50, were substantially longer than the grade mean (9.78).Six of these students never produced as long a T-unit in anysubsequent writing sample. Furthermore, three students (#3, #4, #6)used their shortest T-units in Grade 6. These results demonstratethe fluctuation in T-unit length in individual student's writing fromyear to year.^Only five students (#5, #8, #10, #11, #13) showedsomewhat steady increases in T-unit length in their narratives overthe three year period. Fluctuations in T-unit length also occurredwithin the year in writing samples at every grade level. Student #1,in a Grade 4 narrative written in November, had a mean T-unitlength of 7.7 words. In March of that year, his T-unit lengthincreased to 10.1 words. The following year, his T-unit lengthdecreased to 6.7 in November and increased to 7.7 in March.^InGrade 5, 64 percent of the students had reduced T-unit length intheir March narratives while the remaining 36 percent increased T-unit length. By contrast, in Grade 6 well over half the studentsTABLE 2Mean T-unit LengthStudentNarrativeGrade 4Report NarrativeGrade 5Report NarrativeGrade 6ReportNov. Mar. Mean Nov.^Mar. Mean Nov. Mar. Mean Nov.^Mar. Mean Nov. Mar. Mean Nov.^Mar. Mean#1 7.70 10.11 8.90 10.80 9.43 10.11 6.73 7.70 7.22 8.52 9.72 9.12 10.30 10.00 10.20 11.00 12.50 11.75#2 9.11 9.16 9.14 * 8.55 8.84 8.00 8.42 10.83 12.45 11.64 8.95 8.44 8.69 12.56 9.88 11.22#3 10.21 11.58 10.90 7.69 7.22 7.46 10.88 10.24 10.56 10.47 13.39 11.93 9.48 10.76 10.12 13.41 16.28 14.85#4 10.19 7.82 9.01 11.14 8.68 9.91 8.07 7.89 7.98 11.30 9.09 10.20 6.77 7.90 7.34 11.08 10.13 10.61#5 8.74 9.92 9.33 10.41 8.28 9.35 10.59 10.62 10.61 10.95 15.90 13.43 10.16 11.28 10.74 14.25 10.96 12.60#6 16.5 14.11 15.31 13.12 14.42 13.77 9.21 8.28 8.75 10.00 11.44 10.72 9.23 9.15 9.19 11.43 8.93 10.18#7 12.0 12.16 8.94 8.40 15.0 11.70 11.88 9.42 10.65 17.60 10.46 14.03 11.57 9.31 10.44 14.00 9.61 11.81#8 8.39 8.64 8.43 8.40 6.24 9.61 7.93 * 11.16 * 10.31 11.12 10.71 14.20 11.72 12.96#9 11.37 9.50 10.44 6.44 8.20 7.30 10.20 10.47 10.34 10.83 10.25 10.54 9.38 10.14 9.76 9.03 9.00 9.01#10 8.80 6.66 7.73 13.00 15.25 14.13 11.73 9.26 10.50 12.83 14.00 13.42 8.92 10.92 9.92 11.50 12.40 11.95#11 7.29 9.20 8.25 6.64 10.00 8.32 10.50 8.09 9.30 8.46 11.66 10.06 8.87 11.42 10.15 15.96 15.13 15.54#12 10.09 9.32 9.71 10.22 7.85 9.04 6.72 8.13 7.43 7.54 12.35 9.95 8.81 10.09 9.45 16.00 11.54 13.77#13 9.41 10.02 9.72 13.75 11.44 12.59 10.67 10.42 10.55 14.25 8.53 11.69 12.93 10.56 11.74 16.45 12.94 14.70#14 6.46 9.35 7.90 13.6 11.89 12.74 9.28 7.17 8.28 10.80 8.77 9.79 7.94 10.84 9.39 8.80 9.00 8.90Mean 9.73 9.83 9.78 10.40 10.30 10.50 9.40 9.00 9.18 11.10 11.40 11.30 9.54 10.10 9.85 12.80 11.40 12.10not suitable for analysis8 3increased T-unit length from November to March while theremaining students had shorter mean T-unit length.Reasons for such fluctuations in students' T-unit lengths are notwholly clear. In examining the students' writing samples, however,there are some factors which appear to affect T-unit length. Students'use of language, and in particular the use of dialogue, skew T-unitlength in narrative writing. In Grade 4, for example, students #1, #4,#11, and #14, had shorter mean T-units in narratives which includeddialogue. Student #14 wrote this 'typical' section of dialogue in anarrative in November of Grade 4:'Speedy went up to the blob and asked, "what is your name?"The Blob said, "Blob."Speedy said, "My name is Speedy."Shaggy said, "My name is Shaggy.""Good," said Blob."Let's go to my house," Speedy said."O.K." So they went.'Dialogue often includes short phrases and one word answers. Manyof these students used dialogue in their narratives in Grades 4through 6. In almost every case where a student's T-unit lengthdecreased between two and three words from one narrative toanother, it appears to be the result of using dialogue. 5^Further, itwas also apparent that the content of these dialogues showed littlechange over time; shorter T-units continued to be evident even inGrade 6 dialogues.Fluctuation in T-unit length also occurs when students appearto model language they have heard in other narratives (such as thoseThese are apparent differences and were not subjected to statistical testsfor significance.8 4read aloud during class). For example, in a November narrative inGrade 4, student #6 had an extraordinarily long mean T-unit lengthof 16.5 words. In reading the story it is apparent that she may havemodelled some of the phrases heard in an Indian legend aboutTsongua as she used several relative and nominal clauses that werenot found in her other narratives at this time. Two other students inMarch of the same year modelled their writing after a pattern bookread in class called Alexander and the Terrible, Horrible, No Good,Very Bad Day. Repetition of the phrase, "I knew it was going to be aterrible, horrible, no good, very bad day," in their narrativesincreased their mean T-unit length. Similarly, many students appearto have modelled language from classroom activities, particularly inGrade 4. In view of these findings, T-unit length may not be asreliable a measure of syntactic complexity in narrative writing asmany research studies have claimed (O'Donnell et al., 1967, 1976;Hunt, 1970).Other reasons for increased T-unit length seem related to thestudents' ability to use more complex phrases when describingpeople, actions, or events in their stories. These complex phrases doindicate increased syntactic complexity. Clearly, Grade 6 students,as illustrated by the following examples (from students #13 and #2),have the ability to add detail and figurative language such as simileand metaphor which provide the reader with a clearer picture of thestory event:"Some people of the village said he could jump so high, for sucha long period of time, he was like a hawk." (student #13)8 5"They gave the knight (that escaped) a real nice suit, made ofSiberian, crystal white, silver so he could go into battle."(student #2)Many of the sentence constructions which add detail also contributeto longer T-units. Some of these constructions are examined in thesubsequent sections which analyze extended noun and verb phrases.Comparing grade means of T-unit length can somewhat obscurethe ranges of T-unit length which illustrate the individualcapabilities of the students. The ranges of T-unit length in Grade 4were 5.71-15.60, in Grade 5, 6.24-11.88 and in Grade 6, 6.77-12.93.These ranges demonstrate the variety of T-unit lengths of thestudents, particularly in Grades 4 and 5. T-unit length in Grade 6,however, tended to be more consistent: 10 of 14 students werewithin one or less T-unit length of the grade mean. It is alsointeresting to note that in March of Grade 6 all but 4 students hadmean T-unit lengths of 10.2-11.4 words. These lengths, although notnecessarily statistically significant, are higher than the grade meansquoted in previous studies. Langer (1986), for example, found thatT-unit length in narrative increased from 9.38 to 12.26 with hersubjects aged 11 (Grade 6) to 14 (Grade 9). It would be interestingto see whether the longer T-units produced by students in this studywould increase proportionately in subsequent years.Report Whereas in narrative most students produced their longest T-units in Grade 4, the reverse was true for report writing; moststudents had their longest T-units in Grade 6. T-unit length in reportwas consistently higher than T-unit length in narrative, particularlyin Grades 5 and 6. Interestingly, in Grade 4, almost half of the8 6students produced shorter T-units in report than narrative writing.This occurrence of shorter T-units appears to be attributable to thedifficulty students had writing reports. Some students, #2 and #8 forexample, were unable or unwilling to write a report and wrote astory instead. This appears to be a common occurrence in writingresearch as several researchers in previous studies comment thattheir subjects (of similar ages) wrote narratives or partial narrativesin response to non-narrative writing tasks (Crowhurst, 1983, 1987,1990; Gorman et al., 1988; White, 1989; Wilkinson et al., 1980).Another student (#13) wrote a series of four statements for hisreport entitled "Basketball Rules." These writing samples wereexcluded from the analyses. An example of one of his statementswas, "1. technical foul - when a player slams the ball against theground or pushes someone on purpose."Most students in Grade 4, included very little information intheir reports even though they were asked to provide examples oftheir best work. (See Table 17.)^Students #3, #7, #9 and #11 onlyproduced between three and 10 T-units which appears to indicatethe difficulty they had in generating content. This difficulty, in turn,might have created shorter T-units as students struggled forinformation to write in their reports. By Grade 5, only one studenthad shorter T-unit lengths in report writing. While the students' T-unit length increased in each year, results also showed fluctuations inT-unit length within each year and over the three years. Differencesin T-unit length in the same student's reports varied well over fourwords and in some cases up to seven words per T-unit in one year.(See Table 2, student #7, Grade 4.) In each year, half the students87either decreased or increased T-unit length in their reports.Reasons for the fluctuations in T-unit length in report writingare not clear although there are some possible explanations. Many ofthe students' Grade 4, 5, and 6 reports have what might be termed"inflated" T-unit lengths. Longer T-units were achieved by includinglong lists of items in the report, a strategy which Kress (1982) andCrowhurst (1990) commented upon in their research on expositiveand persuasive writing. In Grade 4, for example, students #4, #6, #7#10, and #13 included lists in their reports. One student (#12) had amean T-unit length of 10.2 words. His sample "Nintendo" illustrateslisting of this sort:"The triforces are triangles of power. I have the Magicalboomerang, the Blue ring, the Red candle, the Wand, the Map,the Meat, the Whistle, the Bombs, the Bow and arrow, theMagical key and that is all."Without this list of words, his mean T-unit length would have been7.8 words. Use of lists continued in Grades 5 and 6. Even though thestudents had no access to reference material and were required towrite only during the writing sessions, some were able to rememberand include a surprising amount of detail in these lists. One student(#7) demonstrated his increasing memory capacity when he wrotethis list of items:"These cars can be customized like my dad's with large tailwings, air vents, nitron ocsides, supertrapp mufflers, bundleof snake headers, radar detectors, cuton Pirelli P-7 tires,steering wheels and much much more."Using T-unit as a measure of syntactic complexity must be cautionedas a result of this kind of construction which was found in several of8 8the students' reports.One of the difficulties in having students write reports is thedifferent kinds of writing the term "report" elicits. These studentswrote reports which described personal items, sports, or games,explained certain aspects of sports, games, and special topics andgave instructions on how to do certain things such as care foranimals, paint with latex, or play a game. Some topics appear torequire longer T-units than others. Giving instructions, for example,seems to require longer T-units than simple explanations aboutspecial items. Generally, students in Grades 4 and 5 wrote reportswhich explained about things. Topics included: Indy Cars, Dancing,Owls, Teddy Bears, Flute, School, and Dogs. These types of reportshad shorter T-units than reports which gave instructions on how todo something. By Grade 6, many students wrote "how to do" reportswhich, in turn, had longer T-units. Student #13 demonstrated thisincreased T-unit length in the beginnings of his two Grade 6examples:"I'm going to talk about football, a rough but fun sport.// Ithink I like football because of the glory and the action.//Myself, I love the sport of football// and I hope it neverchanges."// (9.2 words per T-unit)"If you are dribbling the basketball and you stop and then youstart dribbling again that is a double dribble.// If you aredribbling and somebody hits it and then you start dribblingthat is not double dribble.// When you are dribbling with bothhands at exactly the same time that is double dribble."//(18 words per T-unit)In November of Grade 6, students #3, #6, #10, #11, #12 and #13wrote reports of this kind. Figure 1 illustrates the increase in mean8 9T-unit length. These reports appear to have longer T-units becauseof the causal situations described by the students in theirexplanations. They often wrote something like "If you do this thenthis will happen . . . ." These types of reports demonstrate thelanguage elicited by the situation and the students' increasingly morecomplex thinking as they explain intricate instructions. Students #3and #7 wrote these T-units in reports on volleyball and painting withlatex which seem to illustrate this ability:"So when you volley you need to have your hands above yourhead with your hands slightly curved and your knees slightlybent." (student #3)"First to get your paint brush ready, you have to get a wirebrush and push and push all the little parts of dried paint out,from when you used the brush a previous time." (student #7)As with narrative writing, mean T-unit length calculated as agrade mean does not reveal the wide range of T-unit lengths in thestudents' reports. The ranges were 6.44-15.25, 8.46-17.60, and8.80-16.28 respectively for Grades 4, 5, and 6. It is interesting tonote that the range of T-unit length decreased in Grade 6 indicatingoverall a more consistent mean T-unit length at that time.On the whole, the results show a consistent pattern ofincreasing T-unit length in report writing over the three years inspite of fluctuations within each year for all students. Further, thestudents' mean T-unit lengths appear to be consistently andsubstantially longer in report writing than in narrative writing asdemonstrated by Figure 1. These results must be viewed with somereservation, however, in view of the "inflated" and, therefore,skewed T-unit lengths in several students' writing, especially in90Mean T-unit Lengthin Narrative and Report Writing91Grades 4 and 5.The Extended Noun PhraseThe use of an extended noun phrase in writing is indicative of amore economical use of words in writing, hence a more maturegrammatical construction (Hunt, 1970). Hunt (1965) writes, "thechief factor which lengthens clauses appears to be the increase ofnon-clause modifiers of nouns and the nominalization of clauses.This factor and the increase in adjective clauses account in the mainfor the increased length of T-units" (p. 143). The results of thisstudy show that as these children mature, their writing has increaseduse of modification to extend the noun phrase.Noun Phrase with Modifiers Research indicates older writers not only produce moremodifiers but attach more modifiers to a single head noun (Hunt,1965). In particular, students tend to reduce coordinated nounphrases or short T-units and combine them to form one longer T-unitwith a pre- or post-modified noun phrase. Results of this studyindicate that these 14 students had increased frequencies of nounphrases with modifiers as they grew older.Narrative.^In November of their Grade 4 year, fewer than 50percent of the students used extended noun phrases by adding pre-or post-modifiers in their narrative writing. (See Table 3.) ByMarch, however, all 14 students used this kind of construction. Thenumber of occurrences per 100 words ranged from .28-2.53.9 2Generally, the use of this form increased as students referred morespecifically to a single head noun. Student #1 referred to "a little,old woman" instead of 'a woman' in his March narrative. As with T-unit length, the students sometimes increased their use of modifiersbecause they modelled language from other sources. Students #5, #7,and #9 had higher frequencies of modifiers in their March narrativesthan most of the students. These three students modelled the phrase"terrible, horrible, no good, very bad day" which increased thenumber of modifiers found in their narratives. (Figure 2 indicatesthis increased usage in March of Grade 4.)Each year the number of students that used noun phrases withmodifiers grew. In Grade 5 and 6, nine and 11 students respectivelyhad an increased rate of usage of this form. The results also showsome fluctuations in numbers per 100 words between the students'November and March writing samples. In each grade fromNovember to March more than 50 percent of the students' narrativesshowed an increase in the use of modifiers with the remainder eitherunchanged or decreased. Fluctuations in the use of modifiers weremost apparent in Grades 5 and 6. Student #1, for example, had oneof the highest frequencies of noun phrase and modifiers in hisNovember narrative and yet used no modifiers in his Marchnarrative. It may be that the type of narrative influenced his use ofmodifiers. In November, for example, the student wrote a verydescriptive Hallowe'en story whereas in March he wrote about afamily dance party. The Hallowe'en story (perhaps through practicefrom previous years) appeared to evoke more description, such as"an old broken down house" and "all these dead bodies."93TABLE 3Frequencies per 100 Words ofNoun Phrase with ModifiersStudentGrade 4NP & Mod.Grade 5NP & Mod.Grade 6NP & Mod.Nov. Mar. Nov. Mar. Nov. Mar.#1 (n) - 0.58 1.71 - 0.29 0.42(r) - - - - 0.55 0.94#2 (n) - 0.52 0.69 0.71 0.43 1.27(r) 1.30 - - 0.73 1.49 1.68#3 (n) - 0.18 0.85 0.23 0.70 0.90(r) - 0.77 0.91 1.95 - 1.32#4 (n) - 0.28 0.30 0.46 0.48 0.85(r) 1.28 - - 1.50 - 0.66#5 (n) - 2.53 0.54 - 0.78 0.43(r) - 1.72 - - - 0.31#6 (n) 0.28 - - 0.48 0.28 0.46(r) - - - - 0.54 1.39#7 (n) - 2.28 0.70 1.14 0.66 1.48(r) - 1.68 - 3.18 0.39 1.60#8 (n) - 0.83 0.39 0.37 1.34 0.27(r) - - 0.98 0.78 -#9 (n) 0.33 1.32 0.38 0.75 1.17 0.58(r) - - 1.53 0.61 1.38 1.01#10 (n) - 0.50 - 0.38 1.05 1.23(r) - - - - 1.59 2.15#11 (n) 0.21 1.30 2.54 0.40 0.62 1.66(r) - - 0.91 2.14 0.25 2.64#12 (n) 0.73 1.79 0.45 0.42 0.97 1.08(r) 1.08 1.82 1.20 1.73 4.72 2.08#13 (n) 0.33 - 0.26 0.74 1.89 1.85(r) - - 1.77 0.78 1.64 1.96#14 (n) 0.29 0.57 1.20 0.75 0.82 1.11(r) 0.98 - 1.45 0.73 0.54 -(n) = narrative(r) = report^- null frequency9 4Several students in Grade 5 and 6 used twice as many modifiers inone narrative than in the other. As suggested previously, the use ofdialogue appears to lower the incidence of noun phrases withmodifiers. Again, the many short sentences in dialogue seem topreclude the use of modifiers. The type of story also appears toaffect the use of this construction, particularly at these grade levels.Student #11, for example, in a Grade 5 narrative, used an unusuallyhigh number of modifiers in a fantasy called "I Am Doll." She used aseries of modifiers to describe herself: "I had short, orange, frizzlyhair, a striped pink and white hat, a plain, pink and white clown suitand white shoes." Student #2 in a Grade 6 narrative used somemodifiers in his action-packed adventure story "Hostage" but heincluded almost three times as many noun phrases with modifiers ina fantasy called "Sleeping Beauty 2 The Witch Still Lives." Heincluded many phrases such as "a sharp-edged bone" and "the cold,dark ground." These examples indicate that fantasy, which is a typeof narrative, appeared to evoke more modifiers than other types ofnarratives such as mystery or adventure. Perhaps type of writingwithin a genre may affect the use of certain elements of syntax.Over the three year period the use of noun phrases withmodifiers increased steadily: 12 of 14 students used two to 10 timesas many noun phrases with modifiers in their Grade 6 narratives asin their Grade 4 narratives as previous researchers would predict(Hunt, 1970; O'Donnell et al, 1967; Perera, 1984). (See Figure 2.)Interestingly, while use of this form increased over the years, therange of occurrences per 100 words remained fairly constant. InGrade 6, for example, the range was .29-1.89 which was only slightly95FIGURE 2Grade Means of Noun Phrase withModifiers and Prepositional Phrases9 6lower than in Grade 4.Generally, as students grow older they appear to use moremodifiers to include specific detail in their narratives. As somestudents appear to use more of this construction in certain kinds ofnarratives, it is important to consider type of narrative whenmeasuring use of this form. Also, it seems that modelling languagefrom other sources is a strategy which many children use to assistthem in writing stories. However, repetitive use certain phrasesfrom other stories such as Alexander and the Terrible Horrible No Good Very Bad Day may inflate the frequency of elements of syntax,particularly the use of modifiers.Report.^Use of modifiers to extend the noun phrase wasmuch less frequent in report writing, particularly in Grade 4. By theend of March in Grade 4 only seven students used this form.Further, the use was sporadic between November and March. As thestudents matured, however, usage became more frequent. Tenstudents used modifiers to extend the noun phrase in Grade 5 withall 14 students using this form by Grade 6. Students reportsexhibited some variation in the use of this construction within theyear. Reports with higher numbers of modifiers are descriptive innature. These reports tell about "special things." Student #2, forexample, used modifiers to explain about "Sea Urchins." He wrote "itssuction cupped feet" and "the female, European, edible sea urchin."As students matured, frequencies of modifiers increased. Theyappeared to use more modifiers to be precise in detail or to labelspecific things. Student #13, for example, wrote about "Football."Phrases such as "at exactly the same time," "the offensive, forward9 7play " or "the over and back rule" added specificity to theexplanation. Similarly, student #7 wrote "the world's fastest milerace cars."Numbers per 100 words of noun phrases with modifiers wereslightly higher in report than in narrative especially in Grades 5 and6. Only four students in Grade 4 had higher numbers of this form intheir reports. Yet in Grades 5 and 6, nine students had higherfrequencies of modifiers in their reports than in their narratives.In Grade 4 numbers of occurrences per 100 words (.77-1.82) werefewer than in narrative. The range increased in Grade 5, becominghigher in report than in narrative. By Grade 6 the range increased to.31-4.72, representing the use of up to four times as many nounphrases with modifiers in report writing than in narrative writing.(See Figure 2.) Some caution must be used in interpreting these data,however, for the length of reports is often very short. Thus, althoughan assumption is made that this construction (or other forms ofsyntax) would be used in the same proportion if the reports werelonger, this may not always be true.Noun Phrase with Prepositional Phrase Hunt (1965) found significant increases in the use ofprepositional phrases to modify the noun phrase in successivegrades. The results of this study support his findings. As thesestudents matured, they used more prepositional phrases. There arealso some apparent differences in usage of this form betweennarrative and report.Narrative.^All students in Grade 4 made use of prepositional9 8phrases to modify nouns in their narrative writing. (See Table 4.)Most students were consistent in their use of this form betweenNovember and March. The range of number of occurrences at thisgrade level was .24-2.53 per 100 words with a mean rate of usage of.90. Some students had fluctuations in frequencies of prepositionalphrases from one narrative to another. Often a higher incidence ofprepositional phrases was due to the students' use of a repetitivephrase. Student #5, for example, used "the most wonderful day ofmy life" several times in one narrative. Another student (#2), usedthe construction "all of my things" several times.^In Grade 5, moststudents had increased rates of use of prepositional phrases.Interestingly, almost all the students in Grade 5 had higherfrequencies in March. The range of use was .32-5.68 with moststudents using just over 1.20 per 100 words. Most of the studentswho had used higher numbers of prepositional phrases increased thisusage in Grade 5. (See Figure 2.) At this time they appeared tobecome more specific in their descriptions. These students primarilyused prepositional phrases that began with "of." Often they usedprepositional phrases to make approximations such as "a couple ofbags," "lots of candy," "big bunch of rocks" and "a whole jug of beer."Numbers continued to increase in Grade 6. At this time, therate of use of this form increased in 13 of the 14 students'narratives. Several students who had previously used fewprepositional phrases began to incorporate more into their writing atthis time. The range increased to .40-3.81 with most students usingapproximately 2.0 per 100 words. The grade mean also increased to1.85. Reasons for increased usage appear to be attributable to the99TABLE 4Frequencies per 100 Words ofNoun Phrase with Prepositional PhraseStudentGrade 4NP & Prep.Grade 5NP & Prep.Grade 6NP & Prep.Nov. Mar. Nov. Mar Nov.^Mar#1 (n) 0.84 0.58 1.14 1.62 1.72 2.38(r) 3.70 - 1.80 0.93 2.18 3.29#2 (n) 0.81 1.04 - 0.71 1.42 3.25(r) - 1.30 4.61 2.19 2.98 2.24#3 (n) 0.77 1.24 0.64 2.31 2.99 1.09(r) 1.63 1.54 1.36 3.57 2.20 3.07#4 (n) 0.47 0.98 0.60 0.98 1.19 1.42(r) 1.28 - 0.89 0.50 3.13 3.94#5 (n) 0.24 2.53 0.90 1.27 1.44 1.99(r) 3.20 - 2.81 - 2.94 2.83#6 (n) 0.28 0.83 0.56 0.48 0.83 0.46(r) - - 2.91 0.89 -#7 (n) 1.04 0.46 2.10 5.68 3.81 1.18(r) 1.19 2.50 2.27 3.82 2.75 4.00#8 (n) 0.85 1.38 0.78 1.86 2.42 2.45(r) - - - 2.98 0.78 0.70#9 (n) 1.63 1.32 0.65 2.01 1.17 2.11(r) 6.98 3.45 3.08 1.22 1.38 5.05#10 (n) - 1.00 1.14 1.32 2.80 1.41(r) 1.54 3.28 1.30 2.38 2.54 3.23#11 (n) 0.84 - 0.32 0.73 1.61 1.38(r) - 3.33 4.54 2.14 1.77 6.16#12 (n) 1.45 0.90 1.35 0.85 0.48 3.42(r) - - 3.61 1.16 2.36 -#13 (n) 0.66 0.97 0.52 1.84 2.23 2.77(r) 0.97 4.42 4.69 3.82 3.31#14 (n) 0.57 1.57 0.68 1.49 0.40 1.90(r) - 2.80 3.80 3.62 2.16 2.59(n) = narrative(r) = report^- null frequency1 0 0students' overall ability to add descriptive detail which reducesambiguity in their writing.^Further, it was particularly noticeablethat several students (#2, #3, #7, #10, #12, and #13) whose workexhibited high incidences of prepositional phrases increased thevariety of prepositions that they used to include "along," "through,""about," "in," "with," "around," "behind," and "before." Also, theybegan to use series of prepositional phrases. Student #3, forexample, wrote "The Old Hickory house on top of the hill behindEmily's House."In each year, these students increased their use ofprepositional phrases. By Grade 6, the majority of students usedbetween two to four times as many prepositional phrases in theirnarratives as they did in Grade 4.Report.^Eight of 14 students used prepositional phrases tomodify the noun phrase in their November Grade 4 reports. At thisgrade level, usage of this form was inconsistent as half the studentshad prepositional phrases in only one report during the year. Therange of numbers per 100 words was .97-6.98 with a grade mean of1.85.^This range was twice as high as that found in Grade 4narratives, indicating higher frequencies of prepositional phrases inreport writing. This finding, however, must be viewed with somecaution due to the shorter lengths of reports. 6 One student had avery high frequency of prepositional phrases in her November reportyet this report was only 43 words long. As she used threeOther researchers, for example Langer (1986), have based their data(T-unit length) on short reports such as these. They must also bemaking an assumption that the same ratio of complexity would occur inlonger reports.101prepositional phrases within the report, the result is a very highfrequency rate of 6.98 per 100 words. As stated previously, theassumption that this proportion would continue in longer reportsmay be faulty.In Grade 5, all students used prepositional phrases to modifynouns in at least one of their reports during the year. As innarrative writing, there were fluctuations in usage as eight studentsused fewer per 100 words in March than in November. Overall,however, numbers per 100 words continued to increase in each year.Ten students in Grade 5 had between two and five times as manyprepositional phrases in their reports as they had in Grade 4. Therange of .89-4.69 and grade mean of 2.44 per 100 words were againtwice as high as in the students' Grade 5 narratives. In November ofGrade 5, six of the students had frequencies of approximately 3.0 to4.5 per 100 words in their reports. (This accounts for the highermean rate in Figure 2.) While these reports were still quite short,the students appeared to have higher frequencies of prepositionalphrases when they included more detailed explanations aboutspecific items such as "go carts," "pianos," "hockey cards," "stuffedanimals" and "school." Student #13 reported about his go cart andused phrases such as "a sparkly type of plastic," "two wings on theside," "a fin at the back," and "more fins on the top with blades."Reasons for fluctuations in frequency within the years aredifficult to ascertain. As stated previously, one reason studentsreduce prepositional phrases may be due to the inclusion of lists intheir reports. Students #12, #4 and #6 included lists in their reportswhich seemed to reduce the use of prepositional phrases. Another1 0 2reason that student #11 appeared to reduce prepositional phrasesseems to be due to a stylistic preference. She replaced some of thesephrases with non-finite relative clauses. For example, instead of"two tailed feathers . . . ." [on their back] she wrote "two tailedfeathers curled into their back."By Grade 6, use of prepositional phrases continued to increase.Twelve out of 14 students used between two and six times as manyprepositional phrases to modify the noun phrase as they did in Grade4. Interestingly, one student who rarely used this form prior toGrade 6 had over three times as many prepositional phrases in herreport as she did in Grade 5. Thus, while frequencies of this formwere rare and sporadic in Grade 4, they rapidly increased insuccessive years; numbers per 100 words in report were more thantwice that in narrative in each year as is indicated by Figure 2.Noun Phrase with Appositive Phrase As students mature they use appositive phrases to modify thenoun with increasing frequency (Perera, 1984). Results of this studysuggest that, while the use of this form is somewhat rare throughoutthe three year period, there is a pattern of increased usage as thestudents grow older.Narrative and Report.^There appears to be little difference inusage of the noun phrase with appositive phrase in narrative andreport writing. (See Table 5.) Numbers per 100 words are verysmall and are not affected by genre. The only apparent differencebetween the two genres is in the number of students who use thesephrases and the consistency with which they are used, especially in1 0 3Grades 5 and 6. Only two Grade 4 students had incidences of anappositive phrase: one in a narrative and one in a report. In Grade5, nine students used this form in a narrative and five in a report.By Grade 6, 11 students had appositive phrases in their narrativewriting and eight students had them in their report writing. Allstudents had small numbers of this form in their narrative writingby March in Grade 6. Generally, the students appeared to use thisform to introduce a new character into their narratives. For example,students #5, #4 and #11 introduced these characters midwaythrough their stories: "my stuffed dog, Snoopy," "my best friend,Shannon," and "the newlyweds, Trevor and Melanie."In report writing, students seemed to add an appositive phraseto increase clarification such as: "our favourite, Super Nintendo"(student #1) and "her father, John, her mother, Judy and her sister,Sarina" (student #3). Generally, frequencies of this form were higherin report writing. Numbers per 100 words in narrative ranged from.07-.90 over the three years whereas in report writing they werehigher and ranged from .44-2.19. Overall, however, students tendedto use appositive phrases to extend the noun phrase moreconsistently in narrative than report, particularly in Grade 6. Whileit is interesting that students are able to use appositive phrases atthis age, the low frequencies of this form suggest it may not be areliable measure of syntactic complexity.Noun Phrase with Relative ClauseLinguistic maturity is signalled by the use of subordinateclauses, in particular, the relative clause (Hunt, 1970; Loban, 1976;TABLE 5Frequencies per 100 Words ofNoun Phrase with Appositive PhraseStudentGrade 4NP & Appos.Grade 5NP & Appos.Grade 6NP & Appos.Nov. Mar. Nov. Mar. Nov. Mar.#1 (n) - - - - - -(r) - - - - - 0.94#2 (n) - - - - - 0.14(r) - - - - - -#3 (n) - - 0.42 0.23 0.18 -(r) - - 0.45 - - 2.19#4 (n) - 0.14 - 0.16 0.24 0.14(r) - - - - - 0.66#5 (n) - - - - 0.39 0.14(r) - - 0.70 - - -#6 (n) - - - - 0.28 0.15(r) - - - - - 2.09#7 (n) - - - 0.38 0.33 0.59(r) - - - 1.27 - 0.80#8 (n) - - 0.39 - - -(r) - - - - 0.78 -#9 (n) - - 0.98 0.50 - 0.58(r) - - 1.53 - - 1.01#10 (n) - - - 0.19 - -(r) - - - - - -#11 (n) - - 0.32 0.20 0.37 0.14(r) - - - - - 0.88#12 (n) - - - - - 0.36(r) 1.09 - - - - -#13 (n) - - 0.26 0.37 0.34 0.46(r) - - - - 0.55 0.44#14 (n) - - - 0.25 0.40 -(r) - - 0.63 - - -(n) = narrative(r) = report^-null frequency1047105O'Donnell et al., 1967). Not only do the numbers of relative clausesincrease in students' writing in successive years but the type ofrelative clause that is used changes since "some are noticeably latedevelopments" (Perera, 1984, p. 236). The analysis in this studyclassifies relative clauses as belonging to one of four hierarchicalstages, Stages I, II, III, or IV. 7^Results indicate students used moreand different stages of relative clauses as they matured from Grade 4through 6.Stage I - Narrative.^All 14 students used Stage I relatives (afinite clause modifying an object noun phrase, e.g. She went to thehouse that Mary liked) in narrative writing in Grades 4 through 6.(See Table 6.) Generally, the number of occurrences was quite smallranging from .14-1.74 per 100 words for all students in each of thethree years. Only three students had more than 1.0 relative clauseper 100 words in their narratives over the three years. The datashow that students used this form inconsistently between Novemberand March in each year, particularly in Grades 4 and 5. By Grade 6,however, all students had occurrences of Stage I relative clauses andmost were used quite consistently; nine of 14 students used theseclauses in both narratives. As students mature they seemed to beginto use Stage I relative clauses in their narratives. Use of theseclauses adds more precise detail and description in narrative writing.For example, student #10 used this form to describe a part of his treehouse, ". . . . and a trap door that he would cut in the floor" andstudent #13 wrote, "homeless dogs that scavenged for food aroundSee Definition of Terms for a detailed description of relative clauses.TABLE 6Frequencies per 100 Words ofRelative ClausesStudentGrade 4^Grade 5^Grade 6Relative I ClauseGrade 4^Grade 5^Grade 6Relative II ClauseGrade 4^Grade 5^Grade 6Relative IV ClauseNov. Mar Nov Mar. Nov.^Mar. Nov. Mar. Nov. Mar. Nov.^Mar. Nov. Mar. Nov. Mar. Nov.^Mar.#1 (n) .42 1.74 - .57^.84 .84 1.16 1.14 1.62 .57^.14 - .57 - .86^.14(r) 3.70 - - - 1.63^.94 .93 1.52 .90 1.87 -^- . • - - -^.94#2 (n) .41 .26 - .28 .28^.28 - .26 .46 .14 -^.28 - - .23 .14 -^.14(r) - .77 - .99^- - - - - 1.49^- - - - - -^0#3 (n) - .18 - .69 .18^.18 - 1.24 .21 .23 .18^.36 - .35 - .11 -^-(r) • - 1.82 .97 1.38^.44 - - .45 - -^- - - - - -^.44#4 (n) .47 .14 .60 .39 -^.14 .47 .14 - .39 .24^.14 - - - - -^.14(r) - .61 1.77 - .39^1.31 - .61 - 1.50 -^- - - .89 - -^-#5 (n) .73 .36 - .12 .39^.57 - - .54 - .66^.14 - - - -^.43(r) .80 - 2.11 - 1.26^.63 - - .70 - -^.94 - - - - -^-#6 (n) .55 .83 .56 .48 -^.31 .28 .42 .56 .96 .28^.15 - - .56 .48 .56^-(r) - .99 1.43 .97 1.48^.70 - - - .97 1.19^.70 - - - - .30^-#7 (n) - .35 - .33^- 2.08 - .70 .38 .49^.30 - - .35 -^-(r) - .83 1.13 - 1.18^- - - - - -^- - - - .64 -^-#8 (n) - .43 .39 - .54^.27 - - - - .27^.27 - - - - -^-(r) - - - 1.49 .78^.70 - 4.76 - - -^- 2.38 - - - -^-#9 (n) .33 - .65 .50 .67^.77 .65 - - - .33^- - .66 - - -^.17(r) - - - 1.83 .35^- - 1.15 - .61 .35^1.01 - - - - .69^1.01#10 (n) - .50 - .57 .18^1.23 .54 - - 1.14 .35^.18 - - - - .18^.73(r) - - - 1.59 -^1.07 4.61 1.64 - - -^- - - - - -^-#11 (n) .21 - .32 .33 .12^.55 1.30 - .07 - .37^28 - - - .25^-(r) - - -^.44 - .91 .71 1.27^.88 - - - - .50^-#12 (n) .24 - - - -^.18 .97 - .90 .85 .24^.36 .24 - .45 - -^-(r) 2.17 - - - -^- - - - - -^- - - - - -^-#13 (n) - 97 .26 - 1.20^.92 - - - .74 .86^.23 - - .26 .37 -^.23(r) - - - - -^- - - .88 - -^- - - - - .55^-#14 (n) - - .17 .25 -^.27 - - - - -^.14 - - - - -^.14(r) .98 .94 .63 -^.54^.52 1.47 - - - .54^.52 - - - -^-(n) = narrative(r) = report^- null frequency1 0 7the area."In each year, the frequency of noun phrases with Stage Irelative clauses remained quite small. Although more studentsappeared to be able to use this form as they got older, they did notuse it frequently.Stage I - Report.^In each year fewer students used Stage Irelatives in their report writing than in their narrative writing. Only50 percent of the students used these clauses in Grade 4. In Grades4 and 5 usage was very inconsistent between November and March.By Grade 6, however, usage became more consistent and morestudents included Stage I relative clauses in their reports; 12students used these clauses at this time.Students tend to use Stage I relative clauses in reports whichexplain about something. Student #10 in a Grade 5 report wroteabout "Baseball," "There are 5 leagues that you go through like . . . ."Similarly, student #5 used a number of relatives in a Novemberreport about School, ". . . . you get a report card that tells you if youpassed the grade." This same student, however, did not include StageI relative clauses in her March report on "How to Care for Puppies."Reports which give instructions such as those written by students#10, #11, and #12 in Grade 6, do not have Stage I relative clauses.While fewer students used Stage I relative clauses and usagewas more varied in report writing than in narrative writing, thenumbers per 100 words were consistently higher, particularly inGrades 5 and 6. The range of frequency for these clauses in reportwas .35-3.76, which was two to four times higher than in narrative.The difference in frequencies of Stage I relative clauses in narrative108FIGURE 3Grade Means of Stage I Relative Clauses1 0 9and report writing is reflected in Figure 3. 8Stage II - Narrative.^While there were fewer incidences ofStage II relatives (finite clauses modifying the subject or non-finiteclauses modifying the object, e.g. 'The house that Mary lived in wastorn down.' or 'He hit the boy using his stick.') than Stage I relativesin narrative writing over the three years, more students used thisform in successive grades. (See Table 6.) In Grades 4, 5, and 6, thenumber of students who used this form increased: 9, 11, and 14students respectively employed this form in either a November orMarch writing sample. While more students in each year used moreStage II relatives in their narratives, the number of incidencesactually decreased, especially in Grade 6. Over 50 percent of thestudents had reduced numbers of these clauses in their Grade 6narratives. Reasons for this reduction are unclear. It may be that asthe length of student writing samples increases, the use of Stage IIrelative clauses does not proportionately increase. For students #1,#7, #10 and #11 this seems to be the case. Each of these studentsused these clauses somewhat frequently in proportion to the numberof words they wrote in early narratives. As they grew older, theycontinued to use the same number of Stage II relative clauses buttheir stories became longer which reduced the frequency of use ofthe clauses.Interpreting the results of the students' use of Stage II relativeclauses is difficult due to the relatively low frequencies andinconsistency of use, especially in Grades 4 and 5. These resultsThis figure also indicates the inconsistency and infrequent use ofStage I relative clauses in Grade 4.1 1 0suggest that use of Stage II relative clauses is not a good measure ofsyntactic complexity in narrative writing at these grade levels.Stage II - Report.^Approximately half the number ofstudents who used Stage II relatives in their narratives includedthem in their reports. The numbers of students who used this formare: five in Grade 4, eight in Grade 5, and six in Grade 6. Use of thisform was very inconsistent in each year. One student (#8) in Grade4, used a high number of Stage II relative clauses in his report on the"Pirates of the Caribbean." He described the actions of the figures inthe ride using non-finite clauses with the present participle "ing"which appears to be common at this age. He wrote:"You see pirates chasing people and lighting towns on fire""There are people shooting guns and putting people in jail."His report, however, was very short resulting in a very highfrequency rate. He did not have this same rate of Stage II relativeclauses in subsequent reports.While the number of students who used this form is less thanin narrative, the range of numbers of occurrences was slightly higherin report. The range was .52-1.87 (which excludes an unusually highrate of usage, 4.61, by student #8). This range of numbers remainedconstant over the three years. Generally, however, frequencies werequite low and suggest caution in using Stage II relative clauses as ameasure of syntactic complexity in report writing.Stage III - Narrative and Report.^The incidence of Stage IIIrelative clauses is so rare that results are inconclusive. Therefore,data for use of this clause are not included.Stage IV - Narrative and Report.^These students used Stage1 1 1IV relatives (non-finite clauses modifying the subject or a series ofrelatives, e.g. 'The lady sitting in the chair is my aunt.' or 'Youwouldn't like a person whose appearance was appalling, whose teethhadn't been brushed or whose hair was dirty.') very sporadicallyover the three years. Frequencies in both genres remained very low(.11-1.01).^Thus, this element of syntax is not a good measure ofsyntactic complexity at this age. It is important to note, however,that the number of students who used this form increased in eachyear. Only four students included Stage IV relative clauses in Grade4: three in narrative and one in report. By Grade 5, seven studentsused this clause in their narratives and six in their reports. It wouldseem from this data that students include these relatives more innarrative writing than in report writing. Further, while numbers arevery small, many of these students were able to use this complexform of relative clause. Therefore, use of this clause may provideinsight into the beginnings of more complex writing by somestudents.Subject and Complement ClausesIn her description of the writing of students' aged seven to 17,Perera (1984) writes, "nominal clauses functioning as clauseelements other than object are never frequent; indeed subjectclauses are especially rare and are more likely to occur in the writingof older more able children" (p. 234). Results of this study addfurther evidence to Perera's (1984) findings. Subject andcomplement clauses are used infrequently. The data also indicate,1 1 2however, that use of these clauses increases as students mature.Subject Nominal Clauses - Narrative and Report.Subject nominal clauses were rarely found in Grade 4 or 5writing samples. 9 (See Table 7.) Only one or two students includedsubject nominals in their narrative writing in each year. Slightlymore students used subject nominals in their report writing than intheir narrative writing: two, four and six students in Grades 4, 5, and6 respectively. While usage was slightly higher in report writing,low frequencies in every year in both genres suggest use of thisclause is not a viable means of assessing syntactic complexity.Examining the use of these clauses does, however, offer someinteresting glimpses of beginning attempts at complex linguisticstructures in certain students' work. Student #14 included two ofthese clauses in a report about "Drawing." An example of one of hersentences was, "What I like to do is drawing."Complement Nominal Clauses - Narrative and ReportComplement clauses were also extremely rare in thesesubjects' narrative writing. (See Table 7.) In each year, however,more students began to use complement clauses in their work. Justover half the students had at least one complement clause in one oftheir narratives by Grade 6. Frequencies were very low and use wasvery inconsistent. Only two students had complement clauses inPrevious research in the field has indicated that the use of these clausesdoes not often occur in writing until adolescence. (Hunt, 1970; Loban,1976; O'Donnell et al., 1967; Perera, 1984). The fact that the students inthe present study use these clauses at all indicates more complex syntaxin their writing and their increasing maturity.113TABLE 7Frequencies per 100 Words ofSubject and Complement ClausesStudent Grade 4Subject ClauseGrade 5 Grade 6 Grade 4Complement ClauseGrade 5 Grade 6Nov. Mar. Nov.^Mar. Nov. Mar. Nov. Mar. Nov.^Mar. Nov Mar.#1 (n) .57^.28(r) .93^.46 -^.94 .55^.47#2 (n) -^.14(r) -^.76 .49#3 (n) -^.35 -^.35 -^.35 -^.54(r) .55 .44^.44#4 (n) .14^- -^.14 -^.71(r) -^.39 .44^2.63#5 (n) .13^- -^.32(r) -^1.57#6 (n)(r) .60 -^.99 .89^1.39#7 (n) -^.38 2.28 .70(r) -^.64#8 (n)(r)#9 (n) .33 -^.38(r) .35#10 (n) .18^.53 .54 -^.35(r)#11^(n) -^.28(r) .25#12 (n)(r) 1.20 -^.69 -^1.73#13 (n) -^.37 -^.23(r) -^.78 1.09^-#14 (n)(r) .49^1.8 .54^.52 .98(n) = narrative(r) = report^-null frequency114their reports before Grade 6. In Grade 6 eight students' reportscontained complement clauses. Examples of these clauses are: "Inskating there are many different kinds of things to learn" (student#6) and "That is when you read a book" (student #4).Generally, frequencies of complement clauses were higher inreport writing than in narrative writing; there were twice as manycomplement clauses in students' Grade 6 reports as Grade 6narratives. However, as students exhibited low frequencies andwere inconsistent in their use of complement clauses over the threeyears, this element of syntax is not a reliable measure of syntacticcomplexity. Yet use of this form may indicate beginning attempts atcomplex linguistic structures by specific students.The Verb PhraseResearch has indicated students' use of the verb phrasechanges as they grow older (Loban, 1976; Golub & Frederick, 1970a;Langer, 1986; Hunt, 1970; Perera, 1984). Characteristically,younger students use simple, active verbs whereas older studentsuse significantly more perfect tenses, progressive forms, morepassives, and more modals (Hunt, 1970). Older students also useconstructions that combine both auxiliary and catenative phrasesmore frequently (Perera, 1984). Results of this study show thatthere are observable differences in the use of the verb phrase in thestudents' writing as they grow older.^Further, these differencesdevelop in dissimilar ways in narrative and report writing.1 1 5Modal Auxiliaries Findings in this study support the previous research whichshows increased use of modals by older students. The data alsoindicate this increase is only in report writing.Narrative.^All students in Grade 4 had modal auxiliaries in atleast one narrative writing sample. (See Table 8.) Some students'March narratives contained higher frequencies of modals than theirNovember narratives. Students #1, #6, #7, #8, #9 and #10 hadhigher incidences of modals than other students. Many of thesestudents used the modals "could" and "would" predominantly.Student #10, for example, used the modal "would" to makepredictions within his story: "We would have to fly" and "it wouldcost."^Other students often included the two words as contractions("couldn't" and "wouldn't") which increased the frequency of modalsin their Grade 4 narratives. As mentioned previously, students whomodelled phrases from the story Alexander and the Terrible, Horrible No Good Very Bad Day used the phrase "I could tell it wasgoing to be." Each time they used this phrase it increased thefrequency of modals in their March narratives. (The use of thisphrase accounts for the skewing of the data related to the use of thisform in March of Grade 4 as illustrated in Figure 4.) Students tendedto have reduced numbers of modals from Grade 4 to Grade 6. InGrade 4, for example, the range of numbers per 100 words was .21-2.00. This range was reduced in Grade 6 to .12-1.39. All but twostudents had fewer modals in their Grade 6 narratives than they hadin earlier grades. Reasons for the reduction in the use of modals instudents' narratives are unclear. Many of the students appeared toTABLE 8Frequencies per 100 Words ofModal Auxiliary VerbsGrade 4Modal AuxiliariesGrade 5 Grade 6 Grade 4Do AuxiliaryGrade 5 Grade 6Student NOV^MAR NOV MAR NOV MAR NOV^MAR NOV MAR NOV MARMAIN^SUB MAIN SUB MAIN SUB MAIN SUB MAIN SUB MAIN SUB MAIN SUB MAIN SUB MAIN SUB MAIN SUB MAIN SUB MAIN SUB#1 (n) .42^.42^1.16 - - .57^.54 .54 .29 .86^.28 .42 .42^-^- .42 - -^- - .86 -^.14 -(r) -^-^1.52 .76 - -^2.80 - - -^.94 .47 -^-^- -#2 (n) -^.41^.52 .52 .46 -^- - .57 .57^.28 .28 -^-^- - .46 -^- - .28 -^.57 .14(r) - - -^.73 .73 .50 1.00^.56 2.25 -^-^- - .78 1.54^- - .50 -^- -#3 (n) .26^.77^.35 .71 - .64^.93 .69 .58 .88^.18 .36 .51^.53^- .17 .21 .42^.12 .24 .18 .36^- .18(r) 1.63^-^- - - -^- - 1.38 .28^1.75 -#4 (n) .47^.47^.28 .42 2.42 1.81^.82 .99 .48 .48^.57 - -^.28^- - - -^.08 - .24 .48^.43 -(r) 3.85^1.28^3.03 .61 - -^1.00 .50 1.96 -^1.97 .66 1.28^-^- - - -^- - .39 -^- -#5 (n) .49^.24^.36 - 1.80 .36^.64 .80 .13 .92^.57 .28 -^1.45^- - .36 -^.16 .32 .52 .26^.14 -(r) 4.00^-^3.45 - - -^1.71 1.71 - -^1.57 .63 -^-^- • - -^- • .44 .44^- .31#6 (n) -^.83^1.25 - - .56^- .48 .28 -^.31 .31 .28^-^- .42 - -^- - .56 .83^.92 .15(r) 5.71^.95^4.95 1.98 1.43 1.43 - 3.28 1.09^- - -^1.90^- - - -^1.43 - .55 -^- -#7 (n) -^1.04^1.83 .46 - -^- - .39 -^.80 .80 -^-^- - - -^.38 - - -^.80 -(r) -^-^.83 .83 4.54 -^3.19 - - .50^.30 .30 -#8 (n) .85^-^1.10 .28 .78 -^- .37 1.08 -^- - -^-^- - - -^- - - -^- -(r) -^-^- - - -^- - 1.55 .77^2.11 - -^-^- - - -^- - - -^- -#9 (n) 1.30^.98^1.32 .67 1.63 .65^-^1.01 1.33 1.17^.77 .39 .33^-^- - .33 -^- .75 .17 .50^1.16 .58(r) -^•^- - 3.07 -^- - 2.42 .35^1.01 - -^-^- - - -^- - - -^- -#10 (n) -^1.08^2.00 - 1.14 -^.19 .38 .88 .88^1.06 .70 -^-^1.62 - - -^.19 - .18 .36^- -(r) -^1.54^3.28 - 3.89 2.60^2.38 - 2.23 • - - -^-^- - - -^- - - -^- -#11^(n) .63^.21^- .44 - -^.33 .66 .37 .12^- .14 .21^-^.63 - .32 -^06 .27 .12 .12^.14 -(r) -^-^- - - -^.71 - .75 -^1.76 - -^-^- - - -^• - - -^- -#12 (n) .73^.97^.97 - - .45^1.27 .42 .24 .24^.36 .18 45 24-^-^- • - -^- - - -^- -#13 (n) .33^1.00^.24 - .52 -^.74 - .17 .69^1.39 .69 -^.73^.33 .49 .26 -^- - .34 .17^.23 -(r) -^-^.97 .97 1.77 -^.78 - .55 1.10^1.77 .22 -^-^- - - -^- - .55 - -#14 (n) -^.86^1.26 .31 - .68^1.00 .25 - .55^.28 .28 .31^-^.28 - - -^.50 - .28 .14^.14 -(r) 2.94^-^- - - -^.63 - .52 .52^1.08 • -^1.47^- - - -^.36 - - -^- -(n) = narrative (r) = report^-null frequency117FIGURE 4Modal Auxiliaries:Use in Main and Subordinate Clauses118eliminate contractions from their stories which may, in part, accountfor this reduction in use. Results also indicate there was nodiscernible difference in the use of modal auxiliaries in narrativewriting between the main and subordinate clauses or fromNovember to March.Report.^Findings in report writing suggest overall increasesin the number of students who used modals in successive grades. InGrades 4, 5, and 6, the number of students who used this formincreased from nine to 11 to 13 respectively. Interestingly, twice asmany Grades 4 and 5 students used modals in the main clause inreports than in the subordinate clause. (See Figure 4.) One reasonstudents have higher numbers of modals in the main clause may bedue, in part, to the voice they use when they write their reports.Students who had higher frequencies of modal auxiliaries tended towrite reports in the second person. Students #3, #4, and #5 in Grade4 and #7, #9 and #10 in Grade 5 wrote their reports in this style.These reports had phrases such as "you can wash your dog," "youshould keep your hand on the leash" or "you can steal second base."Students continued to write reports of this kind in Grade 6. At thistime, the pattern of predominant use in the main clause wasapparent but lessens as more students used modals in subordinateclauses. One student (#2) had twice as many modals in thesubordinate clause as the main clause in Grade 6. An example of oneof his sentences is:"They have contests to see who can go the farthest and to seewho can get it on the grill."In each grade, numbers per 100 words were substantially1 1 9higher in report writing than in narrative writing. The range ofnumbers per 100 words was .61-5.71 in Grade 4 which was nearlythree times greater than the range in narrative writing in that year.For most students numbers per 100 words were slightly less inGrade 6 where the range was .30-3.28. Three students in Grade 4,whose numbers per 100 words were 3.85, 4.00 and 5.71, forexample, reduced usage by one half in Grade 6. Nevertheless, evenwith the reduced usage of this form in Grade 6, numbers per 100words continued to remain higher in report writing than narrativewriting.Do Auxiliary - Narrative.^Most students in Grades 4 and 5(10 of 14) used the "do" auxiliary in their narrative writing. 1 0 (SeeTable 8.) Students most often used this form as "don't" or "doesn't."Student #3, for example, in a Grade 4 narrative wrote "He replied,'Why don't you use my back yard." Numbers of occurrences of thisform per 100 words were fewer in Grade 5 than in Grade 4. Therange of numbers was .08-1.62 in Grade 4 which was reduced inGrade 5 to .08-.75. By Grade 6, 13 of 14 students used this modal.Most of these students incorporated this form in dialogue in theirwriting. Student #9 had high frequencies of the "do" auxiliarywritten in several conversations in her Grade 6 March narrative. Anexample of her use of this form was: "Don't worry we know who didit and where they are . . . ." Generally, the number of students whoincluded this modal auxiliary increased in each year, yet frequencieswere low. There were no differences in usage between November10^The "do" auxiliary is often called a 'dummy' auxiliary and is used to formnegatives and questions.1 2 0and March and no discernable patterns between usage in main andsubordinate clauses.Do Auxiliary - Report.^This modal auxiliary was rarely usedin report writing, particularly in Grades 4 and 5. (See Table 8.) Onlyfour of 14 students used this form in each of these two grades andnumbers per 100 words ranged from .81-1.90 (which were higherthan in narrative). Table 8 shows a slight increase in the number ofstudents (eight) who used this modal in Grade 6. Numbers per 100words, however, were less as the range decreased to .44-.55 whichwas comparable to narrative writing. Each of these students usedthe auxiliary "do" as "don't " or "doesn't " in their reports. Most oftenthe report was also written in second person. In these reports manystudents used the auxiliary "do" in a hypothetical situation or whenthey wrote an imperative. Students #5 and #6, for example, usedthese sentences:"If you don't go to school now you will have a hard timefinding a job as an adult." (student #5)"Don't forget to walk your dog." (student #6)Throughout the three years, usage was sporadic as there was nodifference between the number of "do" modals used in Novemberand March in each year. The data show, however, that frequencies ofthe "do" auxiliary were higher in the main clause (which may be dueto the hypothetical and imperative sentences discussed previously).In Grades 4, 5 and 6, for example, all but two students who used thisform used it in the main clause.1 2 1Catenative Verb Phrase Results of this study indicate there are observable changes inthe use of the catenative verb phrase in narrative and report writingas the students grow older. Further, it is apparent that use of thisverb phrase is handled differently in each of the two genres.Narrative.^All 14 students had catenative verb phrases intheir narrative writing in Grade 4. (See Table 9.) Numbers ofoccurrences per 100 words tended to be balanced between the mainand subordinate clauses and between November and March in thestudents' writing samples. The range of numbers per 100 words was.24-3.14 with most students using slightly more than 1.0 per 100words. At this time students often combined two lexical verbs toform catenative verb phrases. Many of the students used "started"and "got" as part of these phrases. 11 For example, students #3, #5,#8, #10 and #12 had high frequencies of catenated verbs due to theuse of phrases such as "started driving," "started hitting," and startedrunning." Similarly, student #7 used "got" in the same manner: "gotdumped," "got killed," and "got swished." Another very commonform of catenated verb phrase in Grade 4 was, "wanted to go."In Grades 5 and 6, 12 of 14 students had reduced numbers ofcatenative verb phrases in their narratives. This reduction wasdemonstrated by a decreased range of .14-1.76.^At this time, manyof the catenative verbs appeared to change. Students' narrativescontained fewer two word verb phrases and more three and four11^Some linguists might consider a phrase such as 'got dumped' to be apassive verb construction, that is, 'got' is substituted for 'was.'^For thepurposes of this study, however, the interpretation has been that 'got' isa lexical verb.122TABLE 9Frequencies per 100 Words ofCatenated Verb PhrasesGrade 4Catenated Verb PhraseGrade 5 Grade 6Student NOVMAINMARSUB MAIN SUBNOVMAIN SUBMARMAIN SUBNOVMAIN SUBMARMAIN SUB#1 (n) .84^.58 -^1.14 1.62 - .57 .28 -(r) 1.85 1.85^2.27 .76 - .94 .55 .55#2 (n) 1.22 .81^.52 1.30 1.16 .92 .14 .86 - .14 .28(r) .78 - .73 .50 -^1.12#3 (n) 1.29 -^1.42 1.06 .86 .21 .35 .23 1.76 .35 .73 .54(r) 1.63 - 3.63 .65 1.38 .28 .44#4 (n) - .47^.98 .70 .91 .30 .91 .25 .24 .24 1.00 .14(r) 1.77 3.00 1.95 - .66 1.20#5 (n) 1.70 .24^1.81 2.54 .72 .18 .64 .48 .66 .26 .85 .43(r) 2.40 -^3.45 - .70 -^1.71 .63#6 (n) .55 .28^.83 - 1.08 .31(r) - .95 2.86 1.43 .82 1.09 1.40 .70#7 (n) .91 1.83 .38 .38 .79 - .80(r) 1.14 1.43 .66 .17 1.80#8 (n) 1.70 .43^2.20 .83 - 1.49 .37 .27 .27 1.09 .54(r) - .98 .78 - .70 1.40#9 (n) .98 1.30^- .66 1.31 .25 .25 .83 .67 .39 .19(r) -^3.45 .61 1.04 .69 1.01 1.01(n) = narrative(r) = report - null frequency1 2 3word phrases as the students began to extend the verb phrase andincorporate a wider variety of catenative verbs. Student #3, forexample, in her November Grade 6 narrative used, "wanted to know,""were going to be " and "decided to stop off. " In each of these twogrades, students had higher numbers of catenative verb phrases inthe main clause than in the subordinate clause. In Grade 6, forexample, 13 of 14 students used more of these verbs in the mainclause. Reasons for predominance of catenative verbs in the mainclause are unclear. One reason may be that students in Grade 6 whouse these verbs appeared to use them in less complex sentences.They seemed to use catenative verbs in simple sentences such as "Ineeded to find Snoopy" or "Emily started to run" or "So Speed andBubbles decided to swim back." Another reason may be thatstudents' repertoire of verb phrases expands as they get older andthus catenative verbs become less common as the students begin touse a wider variety of verb forms.Report.^Fewer students used catenative verb phrases in theirreport writing than in their narrative writing although the number ofstudents who had catenative verbs in their reports increased insuccessive years. Whereas only eight students used catenative verbphrases in Grade 4, 12 students used them in each of Grades 5 and 6.Students in Grade 4 had twice as many catenative verb phrases inthe main clause as in the subordinate clause. This number increasedto three times as many in Grade 5. Higher incidences of catenativeverb phrases in students' Grade 4 and 5 reports often seem to be dueto repetition of a particular phrase in the report. Student #1 wroteabout playing Nintendo and used the phrases, "you have to get," "you1 2 4have to know," and "you have to blow up . . . ." Similarly, student #5repeated phrases beginning with "starts": "starts to come off," "startsshredding" and "starts ripping."By Grade 6 students had slightly lower frequencies ofcatenative verb phrases which were distributed more evenly overthe main and subordinate clauses. Grade 6 students did not tend torepeat the same phrases which seemed to reduce the number ofcatenatives that they used. Also, they began to write longer reportswhich appeared to have more clauses. Frequencies of catenativeverbs in the subordinate clause increased as some of these verbswere used in clauses. Students #4 and #14, for example, usedcatenative verbs in adverbial clauses:"Other jumps like the lutz, and axle are harder because youhave to do a full rotation." (student #4)"When he first started playing, Michael Jordan got his ownshoes called Air Jordans." (student #14)Generally for those students who included catenatives,numbers per 100 words were substantially higher in report than innarrative over the three year period. Although the overall rate ofusage of catenative verbs was reduced in students' Grade 6 reports(from .76-3.45 in Grade 4 to .44-1.95 in Grade 6), numbers continuedto remain higher in this genre than in narrative.Simple Verb Phrase As Perera (1984) states, "characteristically young children usesimple, active verbs" (p. 228). The data in this study support thisstatement in that the students used simple verb phrases1 2 5predominantly. (See Table 10.) Use of simple verb phrasesdominates in all 14 students' Grades 4, 5, and 6 narrative and reportwriting. Results also indicated that over the three years there weredifferences in the students' rate of use and use in main andsubordinate clauses of these verbs in the two genres.Narrative.^Results showed that students, in all three grades,had high frequencies of simple verb phrases in their narrativewriting.^Comparing individual use of the simple verb phrase wasvery difficult due to the variation in numbers per 100 words;therefore, mean percentages of total verbs (per writing sample) werecalculated to compare simple verb phrase use to use of other formsof the verb. (See p. 141, Table 15.) Most students continued to usethe same percentage of simple verbs in their narratives over thethree year period. Figure 5 indicates this lack of change. Reasons forthe lack of change in percentage of verb phrase use are unclear. Itwould seem, however, that narrative writing requires a higherpercentage of simple verbs as there was no change in the frequenciesof simple verb phrases in the students' narratives as they grewolder.A comparison of the use of simple verb phrases in the mainand subordinate clauses indicated students used simple verb phrasespredominantly in the main clause. (See Table 11.) Ratios of mainclause to subordinate clause also remained fairly constant over thethree years: 3.6 to 1, 4.8 to 1 and 3.4 to 1. (See Figure 6.) Over time,there were no differences in the number of simple verb phrases usedin either the main or subordinate clauses.126TABLE 10Frequencies per 100 Words ofSimple Verb PhrasesGrade 4Simple Verb PhraseGrade 5 Grade 6 Student#1 (n)(r)#2 (n)(r)#3 (n)(r)#4 (n)(r)#5 (n)(r)#6 (n)(r)#7 (n)(r)#8 (n)(r)#9 (n)(r)#10 (n)(r)#11 (n)(r)#12 (n)(r)#13 (n)(r)#14 (n)(r)NOV^MARMAIN SUB MAIN SUB10.46 .84 7.56 4.656.48 5.56 9.85 -10.16 1.22 10.39 2.599.09 3.898.24 2.32 7.78 2.309.76 3.25 13.85 -11.68 2.80 11.35 4.075.12 3.84 8.48 6.0610.95 3.89 5.07 2.171.60 3.70 8.62 -6.06 1.38 12.08 1.68^3.81 1.91^1.98 2.9712.50 2.08^3.65 2.7413.09 1.19 6.67 -8.94^.43^8.54 1.659.52 ^ -^3.58 4.56^9.86 1.9816.27 5.74 1.4910.27 1.62 12.00 4.007.69 6.15 1.6410.97 5.06 10.0013.70 1.37 10.0010.39 3.14^9.86 1.358.70 2.19 12.7310.30 2.33^9.73 2.1910.91^4.85^.974.30 4.87^9.75 1.263.92 1.97^6.54 3.73_NOV^MARMAIN SUB MAIN SUB14.28^1.71^8.10^1.0814.41^.90 7.48 4.6711.08 3.00 11.36 .999.23 4.62 5.84 4.389.19 2.78 10.29 3.827.73 2.73 6.17 1.959.06 5.74 11.10 1.8911.50 1.77 8.00 3.005.95 2.52 10.37 2.558.45 2.11 5.71 3.4312.77 2.78 11.48 1.447.14 2.86 7.76 6.808.07 1.40 8.33 .382.27^8.2816.22^1.56 10.37^.746.86^.98 7.46 1.498.17 1.96 9.30 2.513.07 1.54 9.15 1.227.39 1.14 11.93 2.465.20^- 4.76 4.7611.43^1.91 11.69^1.538.18 3.63 8.57^.7111.71^1.80^9.32 -^1 ^- 5.78 2.319.12 2.34 5.52^1.844.42 2.65 9.3812.14^1.20 10.67^3.729.43^.93 8.86^.63NOV^MARMAIN SUB MAIN SUB8.59 2.86^9.93 3.0710.30 3.28^6.57 5.1612.75^.72^9.07 2.555.97 1.49^7.30 2.818.61^3.87^9.98 2.545.80 3.59^2.63 .8814.05^3.57 11.14 3.435.09^.39^7.89 1.978.78 3.54^7.68 3.417.46 1.75^7.86 3.4612.50 3.06 10.63 2.164.00 7.00^8.00 1.396.30 1.58^6.40 -^9. . 2^7.98 1.7810.51^1.62^9.54 2.456.20 1.55^6.34 2.828.17 4.00^8.32 4.069.00 4.50^7.07 1.0110.50 2.98^7.39 2.825.41 2.23^4.84 1.089.05^1.98^8.32 1.943.26 1.00^2.64 1.3214.01^1.45 10.09 2.886.94 4.17^9.45 -7.7 ^7.85 3.463.27 2.73^5.96 3.9712.12 4.63^8.24 2.658.80 1.04^8.65 1.08(n)^narrative(r) = report - null frequencyMain18.0220.5516.0322.0316.0216.1416.1517.4913.3422.2720.9720.1520.0314.0518.099.04Sub.5.493.814.626.876.063.064.822.086.545.625.064.494.596.134.952.47Ratio3.285.393.473.212.645.273.358.412.043.964.144.494.362.293.663.66127TABLE 11Ratio of Main to Subordinate ClausesSimple Verb PhraseStudent#1#2#3#4#5#6#7#8#9#10#11#12#13#14MeanRatio8.005.702.902.605.305.709.202.803.905.406.7011.703.504.604.804.80NARRATIVEGrade Four^Grade Five^Main^Sub.22.38^2.7922.44^3.9919.48^6.6020.16^7.6116.32^3.0724.25^4.2216.37^1.786.59^2.3017.47^4.4719.32^3.6023.02^3.4421.13^1.8014.63^4.1822.81^4.9219.03^3.91^9.51^1.96 REPORTGrade Four^Grade FiveGrade SixMain^Sub. Ratio18.62^5.93^3.1021.82^3.27^6.6018.59^6.41^2.9025.19^7.00^3.6016.46^6.95^2.4023.13^5.22^4.4012.70^1.58^8.0020.05^4.07^4.9016.49^8.06^2.0017.89^5.80^3.0017.37^3.92^4.4024.09^4.33^5.6015.58^5.35^2.9020.36^7.18^2.8019.17^5.36^3.389.58^2.68^3.38Grade Six^Main^Sub.^Ratio^16.33^5.56^2.949.09^3.89^2.3023.61^3.25^7.3013.60^9.90^1.4010.22^3.70^2.805.79^4.68^1.2019.76^1.19^16.609.52^0.00^9.5222.01^1.49^14.809.33^6.15^1.4023.70^1.37 17.2021.43^2.19^9.8015.76^0.97^16.2010.46^5.70^1.807.51^1.76^4.50ain^Sub.^Ratio21. ^5.57^3.9015.07^9.00^1.7014.09^4.68^3.2019.50^4.77^4.1014.16^8.54^1.7014.90^9.66^1.5010.56^0.00 10.5014.32^2.47^5.8012.22^2.76^4.409.96^4.76^2.1016.75^4.34^3.9017.83^2.31^7.7013.80^2.65^5.2018.29^1.56^11.707.63^2.25^3.40in^Sub. Ratio1 8 ^8.44^2.0013.27^4.30^3.098.43^4.47^1.8912.98^2.36^5.5015.32^5.21^2.9412.00^8.39^1.4317.76^4.10^4.3312.54^4.37^2.8716.07^5.51^2.9210.25^3.31^3.105.90^2.32^2.5416.39^4.17^3.939.23^6.70^1.3817.45^2.21^7.906.58^2.34^2.80#1#2#3#4#5#6#7#8#9#10#11#12#13#14Mean1 2 8Report.^Findings of the use of simple verb phrases in reportshowed the same inconsistencies in frequencies between Novemberand March and over the three years as was evident in narrative.Mean percentages of total verbs (per writing sample), however,indicated that most students had reductions in use of simple verbsover time. (See Figure 5.) Ten of 14 students used between 10 and50 percent fewer simple verbs over three years. As studentsmatured, the percentages of other forms of the verb increased intheir reports. (See Table 15.) This finding seems to indicate thatreport writing requires a wider variety of verb phrases.A comparison of main clause to subordinate clause usage (SeeTable 11) also indicated that changes take place in students' use ofsimple verb phrases in report writing. Ratios of main clause tosubordinate clause showed a decrease in the use of main clausesimple verbs and a corresponding increase in subordinate clausesimple verbs. The ratios for each successive grade are: 4.2 to 1, 3.4to 1 and 2.8 to 1. (See Figure 6.) Reasons for this reduction aredifficult to ascertain. One reason may be that students' reportsappeared to have increased numbers of subordinate clauses overtime. This increase, in turn, may have increased frequencies ofsimple verb phrases in the subordinate clause.Progressive Verb PhraseUse of the progressive verb phrase is frequent and earlyin children's writing (Perera, 1984). There is also some indicationthat use of simple progressives decreases as children mature (Hunt,1964; Perera, 1984). Results of this study indicate that students80777471686562595653505.04.54.03.53.02.52.01.51.00.50.0129FIGURE 5Percentage of Simple Verbper Total Verb PhrasesFIGURE 6Ratio of Main to Subordinate Clause1 3 0consistently used the progressive verb form from an early age andtended to reduce its use as they grew older. Results also revealeddifferences in use of the progressive verb phrase in narrative andreport writing.Narrative.^All students in Grade 4 used progressives in theirnarratives.^(See Table 12.) Twelve of 14 students used more than1.00 per 100 words and the range was .21-2.90. In Grade 4 and 5,students had higher frequencies of progressive verbs in the mainclause. Most students used this form of the verb in simple actionstatements in the past tense in their narratives. Typically, they usedphrases such as "was going," "were walking," or "was hitting."Several students had unusually high frequencies of progressiveverbs in the subordinate clause in Grade 4 due to modelling thephrase, "I could tell it was going to be a terrible, horrible, no good,very bad day" from Alexander and the Terrible, Horrible. No good.,Very Bad Day. By Grade 6, students had more even distribution ofthese verbs in the main and subordinate clauses. At the same timethere were noticeable reductions in the frequency of progressiveverbs. The range was reduced to .14-1.33 per 100 words. Whilereasons for the reduced use of progressive verbs are not readilyapparent, one postulation is that as students mature, theyincorporate a wider variety of verbs into their narratives. Table 15would seem to indicate this is true. Mean percentages of verb phraseindicate that in Grade 6 students had reduced frequencies ofprogressive verbs and increased frequencies of other forms of theverb. Also, as the students matured, they tended to use less dialoguein their narratives. As the progressive verb is often used in dialogue,Student#1 (n)(r)#2 (n)(r)#3 (n)(r)#4 (n)(r)#5 (n)(r)#6 (n)(r)#7 (n)(r)#8 (n)(r)#9 (n)(r)#10 (n)(r)#11 (n)(r)#12 (n)(r)#13 (n)(r)#14 (n)(r)^NOV^MARMAIN SUB MAIN SUB.42^-^-^1.161.22^.41••■.26.52.35^.53-^1.12^.84GM^.24^.36 2.90.28^.56^-^1.25- .95-^1.832.13^-^1.6598^.33^.66^.66^- 1.15^1.15^1.08^.54^-^.50- 1.64^.2 ^.42^-^1.30.73^.48^.90 1.46^1.09^- 1.82^1.33^.33^.24^1.461.81 -^.97- ^.57^.63 2.20^- 1.47^-^.94^N V^MARMAIN SUB MAIN SUB- 3.78^.54^- .80^-.41^.23 1.42.64^.21^.58^.23- .45. ^.08.57.18^.54^.80^.48.97^.56 1.67 .44^.48.35^.70^.38NO^- ^.74^.37^.61^-.33^.33^.25^.751.70 1.14^.19^.57.32^.64^.47^.271.35^.45^-^.421.2 ^- 1.16^- 2.57^.37.34^- 1.00Grade 6 NOV^MARMAIN SUB MAIN SUB-^.42^.42.72^.29^.71^.14.35^.35^.18^.36^- . ^.44^.44^.43^.43- ^.66^.26^.14^.57.56^.77- .70- ^.79.4 ^- .307 ^.27-^-^.703 ^.67^.39^.39.358 ^.70^.53.87^.87^.28.48^.24^.36^.1810^.17^.23^-.55^.55^.22 1.33.55 ^.68^.28^.1404 .52^.54^.54Progressive Verb PhraseGrade 4^Grade 5131TABLE 12Frequencies per 100 Words ofProgressive Verb Phrases(n) = narrative(r) = report^-null frequency132this reduction may have decreased the frequencies of progressiveverbs. Additionally, some students reduced full progressive phrasesto the present participle as they incorporated more non-finite clausesinto their writing. An example of this might be, "the sun was shiningon the table and flooded the room with light," becoming "the sunshining on the table flooded the room with light." Use of progressiveverbs was quite consistent in narrative writing over the three years.Report.^The use of progressive verb phrases in reportwriting was limited. In each year only 50 percent of the studentsused this form. While there are fewer Grade 4 students usingprogressive verbs in reports, numbers per 100 words werecomparable to those in narrative writing. Students who hadsomewhat high frequencies of progressive verb phrases in theirGrade 4 reports tended to write in the first person. Students #9 and#12 used this style in the following sentences:"I am dancing next year in ballet." (student #9)"I am playing hockey next year." (student #12)Other reports such as one written by student #14 use the progressiveto describe what people are doing presently:"People in Kelowna are cutting the apple trees because theyare becoming too old."Overall, however, use of the progressive verb phrase decreased overtime. Further, use of this form was less frequent in report writingthan in narrative writing each year.Passive Verb Phrase Several studies have noted the use of the passive verb phrase133in the research in children's writing. Use of the passive verb phrasewith omission of the agent phrase or active subject is a stylisticpreference among older writers (Hunt, 1970). Further, the increaseduse of the passive verb phrase is indicative of older, more competentwriters (Golub & Frederick, 1970a; Hunt, 1964, 1965, 1970; Lawton,1968; O'Donnell et al., 1967; Perera, 1984; San Jose, 1972). Resultsof this study seem to support the findings in previous studies withthe understanding that use of the passive in these students' writingwas limited, especially in Grades 4 and 5.Narrative.^Interestingly, all students in Grade 4 used thepassive verb phrase in at least one narrative during the year. (SeeTable 13.) This finding indicated that each student was able to usethis more complex verb form. Occurrences per 100 words weresmall and ranged from .12-.85 Use of the passive was primarily inthe main clause; 12 students had passives in the main clause whileonly four had passives in subordinate clauses. Generally, thesestudents used passives in the sentences that ended their narratives.Students #2 and #7 illustrated this use of passives in the followingstatements:"And the kids were saved by Tsongua." (student #2)"And the children were taught a lesson." (student #7)In Grade 5, only 10 students used passives in their narratives. InGrade 6 use of this form increased slightly as all students usedpassives in either a November or March writing sample. Studentscontinued to use these verbs to end narratives yet the sentencesappeared to be somewhat more complex. Student #2 wrote thisGrade 4^NOV^MARMAIN SUB MAIN SUB.84^-^.58^-.41^.82.18.28^.14.361.60^- ^.56.98^-.46.83.85.33- 1.15.24- 1.64.84Passive Verb PhraseGrade 5NOV^MARMAIN SUB MAIN SUB.23^.46^.43^.7 ^- 1.462 ^86^.35- .65^.30^.30^.16^.08.88- ^.16^.57- .97^.3 ^-^.76- ^.64.39^1.54^-- 1.14^.19.64^.32^.13Il•1.30^-^.37- .89^.78-^.17Grade 6 NOV^MARMAIN SUB MAIN SUB-^.14^.28Ile14 ^.28^.28^.141.00 .50.18 .36 - .36.28^.14- .39. ^.39-^.28^.31.39^.39 2.40-^.54.70.33 ^.50^.391.01.18 ^.18.54.50 .37 .562.50 .25 .88^- ^.54^.18- .79.17 ^.17^.92.14^.28^.14^.52 ^- .54Student#1 (n)(r)#2 (n)(r)#3 (n)(r)#4 (n)(r)#5 (n)(r)#6 (n)(r)#7 (n)(r)#8 (n)(r)#9 (n)(r)#10 (n)(r)#11 (n)(r)#12 (n)(r)#13 (n)(r)#14 (n)(r)^.241.09^.3 ^.24.29■••134TABLE 13Frequencies per 100 WordsPassive Verb Phrases(n) = Narrative(r) = report^- null frequency135sentence:"The police never found the kidnapper but suspicions weremade that now he is dead."Use of the passive and perfect forms of the verb are requiredto juxtapose events happening in the story with events that havealready happened. As students matured, they seemed to changefrom using simple past tense to using verbs such as the passive toachieve this juxtaposition in their stories. Student #13 wrote:"When he reached his castle, he saw that it was covered withvines."Thus, use of this verb form seems to indicate the beginnings of morecomplex writing.As students matured, the use of passive verbs was moreevenly distributed between the main and subordinate clause innarrative writing but frequencies did not increase. While morestudents used passives verbs each year, frequencies remained quitelow. In view of these low frequencies, this element of syntax maynot be a good measure of syntactic complexity at these grade levels.Report.^Use of the passive verb phrase in reportwriting was rare in Grade 4. Only six students used this form. All sixstudents used the passive exclusively in the main clause and therewere similar frequencies in November and March writing samples.Generally, Grade 4 students included only one or two passive verbsin their reports. The range was .82-1.64 which was higher than innarrative writing. All but one student used the passive verb phrase"are called" or "are made" in their reports. These forms appeared tobe students' beginning attempts at the use of the passive voice which1 3 6is indicative of the more impersonal style of writing used inexposition. For example, student #5 in a Grade 4 report wrote thissentence:"Teddy bears are made in lots of different places."In each year the number of students who used passivesincreased: eight in Grade 5 and nine in Grade 6. These students hadpassives verbs in the main clause predominantly. By Grade 6 therange of usage increased to .25-2.50 and was higher than innarrative writing. Use of this form, however, was inconsistent. Onlyone student at this time had passive verbs in both reports. Thisstudent had a high incidence of passives in her report on volleyball.She wrote the entire piece in the passive voice and used forms suchas "are allowed," "is scored," "is performed " and "is played."Generally, more students had passive verb phrases in theirreports as they matured. Further, these students had higherfrequencies of the passive verb in report writing than they did innarrative writing. Use of this form, as stated previously, wasinconsistent and somewhat rare compared to other forms of the verb.Thus, its use as a measure of syntactic complexity must be cautioned.Perfect Verb PhraseHunt (1964) found a statistically significant difference forgrade in the use of perfect verb phrase in his subjects' writing. Hefound students increase the frequency of perfect verb phrases insuccessive grades. While use of the perfect verb phrase is rare inthis study, results show more students are able to use this form asthey grow older.1 3 7Narrative.^In their Grade 4 narratives only five studentsused perfect verb phrases. (See Table 14.)^In Grades 5 and 6, 11and 14 students respectively had occurrences of perfect verbs in atleast one of their narratives during each year. Use of perfect verbswas more frequent in the main clause than in the subordinate clause.Again, use of this verb allows students to incorporate events whichhave happened with events currently happening in their stories. Onestudent used this technique skillfully in a Grade 6 narrative:"Trevor and Melanie Starling looked on anxiously. Trevorgripped Melanie's hand reassuringly. Newlyweds, Trevor andMelanie, had always longed for a child. Unable to give birth,they had gone to an adoption agency." (student #11)As students matured, they began to use more complex forms ofthe verb such as perfect verb phrases. Frequencies of perfect verbphrases in these subjects' writing, however, were very small: therange for each of the three years was .13-.86. Therefore, whileexamining use of this verb form is interesting, it cannot be regardedas a reliable measure of syntactic complexity in these students'narrative writing.Report.^Occurrences of perfect verb phrases were lesscommon and less consistent in report writing than in narrativewriting: students rarely used this form until Grade 6. In Grade 4,four students used perfect verb phrases whereas in Grade 5 only onestudent included this form.^Numbers per 100 words ranged from.77-1.15 which were slightly higher than in narrative. By Grade 6,eight students had perfect verb phrases in their reports. At thistime, the range increased marginally to .44-1.61. Perfect verb138TABLE 14Frequencies per 100 Words ofPerfect Verb PhrasesGrade 4Perfect Verb PhraseGrade 5 Grade 6Student NOV^MARMAIN SUB MAIN SUBNOV^MARMAIN SUB MAIN SUBNOVMAINMARSUB MAIN SUB#1 (n) - .14 - .14(r)#2 (n) -^.21 .58 .35 .28 - .14 .14(r) GO - .56#3 (n) .77^-^.18 .26 -^.21 .58 .35 .18(r) -^.77 - .50 .50 - .88 .44#4 (n) - .25 .08 - .14(r) MO#5 (n) .16 - .13 .14(r) IM .31#6 (n) - .48 .31(r) ■18#7 (n) - .46 - .38 - .79(r) -^.83 - .16 .30 .30#8 (n) .85^- .54(r) -^1.15 IN#9 (n) .50 .19 .19(r) -^1.15 de 1.01#10 (n) - .95 .18 .53 .35(r) .64 1.61 .54#11 (n) .63^.42^- .44 - .47 .27 .12 .36 .56 .28(r) • - .44#12 (n) - .85 .18(r)#13 (n) - .24 .86 .17 .46 .23(r)#14 (n) - .25 .25 - .14 .14(r) -^1.47 - .52 .54(n) = narrative(r) = report^- null frequency139phrases appear to be used in sentences which refer to facts occurringat different times. Student #10, for example, in a Grade 6 reportabout Pavel Bure wrote these sentences:"Pavel Bure is very fast on his skates and is very skilled for hisage of 19. Since Pavel has came [come] to the VancouverCanucks he has won the hearts of thousands."Generally, the majority of these students did not begin to useperfect verb phrases in reports until Grade 6 and frequencies werequite low. Thus, it is not a viable measure of syntactic complexity.Nevertheless, it does seem to indicate the beginnings of morecomplex writing by some students.Perfect Progressive/Perfect Passive Verb PhrasesUse of expanded forms of the progressive verb phrase such asthe perfect progressive increases as students mature (Hunt, 1970;Perera, 1984). Results of this study indicate that use of this form isvery rare.Narrative and Report.^Use of these forms of the verb phrasewas almost negligible in both narrative and report writing. It isimportant to note, however, that more students used these forms inGrade 6 either of the preceding years. Seven students used perfectprogressive verbs at this age level and five used perfect passives(four of whom also used perfect progressives).Percentage of Total Verb PhrasesTo provide additional information related to the use of the verb140phrase in the 14 students' narrative and report writing, meanpercentages of verb type were calculated. The percentages werecalculated by comparing the numbers of each verb type (usingcombined totals of main and subordinate clauses) to the total numberof verbs used in each writing sample. The percentages for Novemberand March were then averaged to find the mean percentage per verbtype per year for each student. These results provide a yearlycomparison of usage of the verb phrase in the students' narrativeand report writing. 12 (See Table 15.)^The mean percentages for allstudents were then averaged to find the grade means of verb typefor each of narrative and report. (See Figures 7 and 8.) As thesimple verb phrase is used with much greater frequency than theother verb phrases, separate figures excluding simple verbs areprovided (Figures 9 and 10) to permit easier comparison of meanpercentages of other verb types.Modals Over the three years the mean percentages of modals innarrative decrease slightly from 6.9 percent in Grade 4 to 5.8 percentin Grade 6. (See Figures 7 and 8.)^Mean percentages in report arenearly twice as high as those in narrative and decrease at a similarrate from 12.2 percent to 11.2 percent.^(See Figures 8 and 10.) Useof the "do" auxiliary increased marginally in narrative writing overthe years from 2.1 percent to 3.9 percent. (See Table 15.)^Thisauxiliary was used only half as much in report and the mean1 2^Without statistical tests, these percentages provide only anapproximation and are not meant to imply statistically significantdifferences.TABLE 15Mean Percentages of Total Verb PhrasesGr.4ModalGr.5 Gr.6 Gr.4DoGr.5^Gr.6CatenatedGr.4^Gr.5 Gr.6SimpleGr.4^Gr.5 Gr.6ProgressiveGr.4^Gr.5^Gr.6PassiveGr.4^Gr.5 Gr.6PerfectGr.4^Gr.5 Gr.6#1 (n) 6.5 4.5 7.0 3.5^-^3.0 5.0 8.0 3.0 75.0 74.5 82.0 5.0^13.0 3.0 5.0^- 1.0 - - 0.5(r) 7.0 15.5 5.0 -^3.0^2.0 21.5 3.0 4.0 69.0 74.0 89.0 2.5^4.5 - -^- - - -#2 (n) 4.5 5.5 5.0 - 1.0^3.5 11.5 6.0 4.0 74.0 75.5 76.5 6.0^7.5 3.0 4.0^3.5 1.5 - 0.5 4.0(r) 9.0 5.0 16.0 - 7.0^2.5 - 7.0 6.0 91.0 74.0 67.5 • - - -^7.0 6.0 - - 2.0#3 (n) 6.5 6.5 5.5 4.0 3.5^2.0 13.0 4.0 10.0 68.5 74.0 74.0 3.5^4.5 4.0 0.5^4.0 2.5 0.5 4.0 2.5(r) 5.0 11.5 16.0 2.5 -^5.0 5.0 15.0 8.5 85.0 69.0 55.0 -^1.5 7.0 -^3.0 1.0 - - 7.5#4 (n) 4.5 13.5 3.5 0.5 0.5^3.5 5.5 6.0 5.5 83.0 72.5 82.5 4.5^1.5 3.0 1.0^3.0 1.5 - 2.0 0.5(r) 26.5 4.5 18.0 4.0 -^2.0 - 14.0 16.0 69.5 71.0 60.0 -^5.0 2.0 • 2.5 - - 3.0 2.0#5 (n) 3.0 12.5 6.0 4.5 4.0^3.0 18.0 6.0 7.5 63.0 70.0 75.5 10.5^6.5 5.0 1.0^0.5 1.5 - 0.5 1.0(r) 25.5 11.0 7.5 - -^4.0 20.0 8.5 2.0 46.0 73.5 84.5 -^2.0 - 8.5^2.0 - - - 2.0#6 (n) 7.0 3.0 2.5 2.5 -^7.5 6.5 - 3.5 74.0 83.5 80.0 7.5^12.0 3.5 2.5^- 2.0 - 1.5 1.0(r) 46.0 8.0 10.5 6.0 3.0^1.5 3.0 12.0 12.5 37.5 71.0 68.5 3.0^3.0 4.5 4.5^3.0 2.5 - - -#7 (n) 11.5 - 8.5 - 1.5^3.5 9.5 3.5 7.0 69.5 82.0 61.0 6.5^6.5 3.5 1.5^5.0 13.0 1.5 1.5 3.5(r) 8.0 41.0 4.0 - -^3.0 - 9.5 10.5 84.0 47.0 77.5 -^- 2.5 4.0^2.5 - 4.0 - 2.5#8 (n) 7.0 3.5 3.5 1.0 -^- 16.0 7.0 7.0 60.0 84.5 83.5 11.0^4.0 2.0 2.5^1.0 2.0 2.5 - 2.0(r) - - 17.5 - -^- - 4.5 11.0 84.0 79.5 66.5 -^9.0 2.5 4.0^- 2.5 4.0 - -#9 (n) 13.0 11.0 9.5 1.0 6.5^6.5 9.0 7.5 5.0 68.0 73.5 66.5 8.0^5.0 7.0 1.0^- 3.0 - - 2.0(r) - 16.5 11.5 - -^- 11.0 2.5 13.0 75.0 70.0 66.5 7.0^2.5 1.0 3.5^8.5 4.0 3.5 - 4.0#10 (n) 8.0 5.0 11.0 4.5 1.5^1.5 8.5 7.0 4.5 71.5 68.0 72.5 6.0^10.5 4.5 1.5^4.0 1.0 - 2.0 3.5(r) 21.0 37.5 9.0 - -^- 8.5 - 10.0 53.5 62.5 56.0 8.5^- 2.5 8.5^- 11.5 - - 11.0#11^(n) 3.5 4.0 2.0 2.0 1.5^1.5 5.5 1.5 6.5 77.0 81.5 74.0 1.5^6.0 6.0 2.0^3.5 5.0 6.5 2.5 2.5(r) - - 16.0 - -^- - • 3.0 100.0 100.0 52.0 -^- - -^- 23.0 - - 3.0#12 (n) 9.5 7.0 3.0 1.0 -^- 15.0 12.5 7.0 65.5 71.0 80.0 8.5^6.5 5.0 0.5^- 1.0 - 3.0 -(r) 3.5 - 2.5 - 0.5^0.5 - 10.0 9.5 83.0 80.5 82.5 10.0^9.5 3.0 3.5^- 1.5 - - 0.5#13 (n) 4.5 4.5 9.5 4.5 3.0^3.0 6.0 6.0 6.0 72.0 64.5 70.0 10.0^12.0 12.0 2.0^5.5 4.0 1.0 - 5.5(r) 11.0 12.0 15.5 -^2.5^2.5 - 7.5 8.0 83.5 73.0 63.5 5.5^- 10.5 -^7.5 - - - -#14 (n) 7.0 6.0 3.0 1.0 2.0^2.0 4.5 3.5 5.0 77.5 84.0 81.5 9.5^3.5 4.5 0.5^0.5 2.0 - 1.0 1.0(r) 8.5 2.5 8.0 4.5 -^- 17.0 10.0 - 57.5 85.0 73.5 8.0^• 10.5 -^- 4.0 4.5 - 4.0Mean 6.9 6.2 5.7 2.1 1.7^3.9 9.5 5.6 5.8 71.3 75.6 75.7 7.0^7.0 4.7 1.8^2.2 3.0 1.1 1.2 2.3Mean 12.2 11.8 11.2 1.2 1.1^1.6 6.1 7.7 8.1 72.3 73.5 68.8 3.2^2.6 2.6 2.8^2.6 7.8 1.1 0.2 2.8(n) = narrative(r) r. report^- zero percent142FIGURE 7Mean Percentages of Verb Type - NarrativeFIGURE 8Mean Percentage of Verb Type - Report143FIGURE 9Mean Percentage of Verbs - NarrativeFIGURE 10Mean Percentages of Verb Type - Report144percentage remains virtually constant at 1.2 percent in Grade 4 and1.6 in Grade 6.Catenative Verb Phrase Use of catenative verb phrases decreased in narrative writingover the three years. (See Figure 9.) Mean percentages decreasedfrom 9.5 percent in Grade 4 to 5.3 percent in Grade 6. In reportwriting, however, catenative verb phrase usage increased. In Grade4 students used 6.1 percent catenative verb phrases which increasedto 7.7 percent in Grade 5 and 8.1 percent in Grade 6. (See Figure 10.)Simple Verb Phrase Mean percentages of simple verb phrases show a slight butsteady increase in usage from Grades 4 to 6 in narrative writing.(See Figure 7.) Percentages rise from 71.3 to 75.6 to 77.1. Thefindings for report writing indicate similar percentages with slightdecreases as students grow older. The percentages of simple verbphrases decreased from 72.3 percent to 68.8 percent. (See Figure 8.)Progressive Verb Phrase Use of the progressive verb phrase in both narrative andreport writing tended to decrease as students matured. (See Figures9 and 10.) In their narrative writing, students used approximately 7percent progressive verb phrases in Grades 4 and 5. This numberreduced to 3.6 percent by Grade 6. Overall, percentages are smallerin report writing. Students had approximately 5 percent progressiveverb phrases in their reports in Grade 4 which decreased to 2.6145percent by Grades 5 and 6.Passive Verb Phrase Means of percentages of verb type indicated increases in theuse of the passive in both narrative and report writing. (See Figures9 and 10.) Students had 1.8 percent passive verb phrases in Grade 4which slightly increased to 2.9 percent in their Grade 6 narratives.Percentages in report writing were marginally higher. They were 2.6percent in Grades 4 and 5 and rose to 3.4 percent in Grade 6.Perfect Verb PhrasesWhile students had only a very small percentage of perfectverb phrases in their narrative and report writing, usage increasedminimally as they grew older. (See Figures 9 and 10.) In Grade 4,students had approximately 1 percent perfect verb phrases in theirnarrative writing which was slightly increased to 2.3 percent inGrade 6. Almost identical results are found in their report writingexcept for a decrease in Grade 5 to 2.1 percent. By Grade 6, however,reports contained almost 3 percent perfect verb phrases which wasslightly higher than in narrative.Other Verb Phrases Mean percentages for the other verb forms, such as perfectprogressive and perfect passive, are too minimal for grade means tobe a useful comparison.Thus, for mean percentage per grade, the highest percentage ofverb type in narrative is simple verb phrases followed in order of1 4 6decreasing percentage rate by progressives, catenatives, modals,passives and perfect verb phrases. In report, simple verb phrasesare also the most predominant but the order of verb types changesas follows:^modals, catenatives, progressives, passives, "do" auxiliaryand perfect verb phrases.Text Forms in WritingPrevious research has indicated that students structure theirwriting in recognizably different ways as they grow older (Bereiter &Scardamalia, 1982; Gorman et al., 1988; Gundlach, 1981; Kress,1982; Langer, 1986; Wilkinson et al., 1980). Further, in order tostructure their writing students use different text forms, that is,conventional statements, forms of organization, and content indifferent genres (Crowhurst, 1983, 1987, 1990; Gorman et al., 1988;Harris and Wilkinson, 1986; Langer, 1986; Morgan, 1984;Wilkinson et al., 1980). Many methods of analysis have been used toexamine students' use of text forms in their writing. To examine thenarrative and report writing of these Grades 4, 5 and 6 students,' twomethods of analysis seemed most appropriate: an analysis ofbeginning and ending statements and an analysis of length of writingsample. Results of these analyses indicate there are observabledifferences in the way these students structured their writing asthey grew older. Similarly, students differed in the ways that theyused text forms in narratives and reports.147Beginning and Ending Statements Several researchers have examined beginning and endingstatements as a means of analyzing students' writing. They havereported that students begin and end writing pieces differently indifferent modes of discourse and that students structure beginningsand endings more competently as they grow older (Crowhurst, 1990;Gorman et al., 1988; Kress, 1982; Langer, 1986; Morgan, 1984;Wilkinson et al., 1980). In order to examine the beginning andending statements that the subjects of this study used, a system ofcategorizing the statements, modelled after Langer's (1986) work,was used.^Langer (1986) outlined three kinds of categories ofincreasing complexity to classify her subjects' beginnings andendings. In order of increasing complexity they are weak, formulaic,and structural. Within each of these categories are several sub-categories (see Appendix B for a detailed explanation of thisclassification system). Results show that not only did these studentsincrease the complexity of beginning and ending statements as theygrew older, they used differing complexities of opening and closingstatements in narrative and report.Narrative Beginnings Over the three year period the complexity of students'beginning statements increased from formulaic to structural innarrative writing. (See Figure 11.) In November of Grade 4, sixstudents had formulaic openings in their narratives, such as "onceupon a time" while eight had structural beginnings, either a1 4 8statement about setting or an 'action dialogue' opening such as "'Cut,'said the director as he stood on the set."^By March, only threestudents continued to use formulaic beginnings whereas six used'setting' statements and five used 'action dialogue' beginnings. Manyof the Grade 4 students' November narratives began with "Once upona time . . . ." Several of these students wrote an Indian legend whichseemed to require this traditional formulaic beginning. Most Grade 4writers opened with a structural statement which established thesetting of their story. Typically, they began their opening sentencewith the phrase, "One day . . . ." Student #3 used this form ofopening:"One day a girl said a magician was coming to town."Occasionally, a student at this grade level used an action/dialogueopening.^Student #11 wrote this example:"Wow, cried Claire,"In November of Grade 5, 12 students had structural openingsin their narratives: six used 'setting,' four used 'action dialogue' andtwo used 'beginning action' statements such as "Sally opened the cardoor." Two students at this time had weak beginnings which were inthe form of a dialogue between the author and the reader. Five ofthe students who used 'setting' statements in their Novembernarratives began with "one day," "one morning" or "one night."Several students in Grade 5 seemed to try different beginnings andbegan using 'beginning action' statements.^For example, student #4began, "I was having a sleepless night . . . ."^In March, two studentshad formulaic beginnings while 12 continued to have structuralbeginnings. At this grade level some students seemed to add more149FIGURE 11Narrative Beginning StatementsFIGURE 12Report Beginning Statements1 5 0detail and intrigue to their structural opening statements. Students#11 and #2, for example, wrote these sentences:"Kimberly hustled down the street watching the leaves dancearound the thick trunks of trees," (student #11)"As Penny arrived into the dark, wet, musty jail cell, the firstthing she saw was Pollywog." (student #2)In Grade 6, all but two students began with a statementreferring to setting in their November narratives. Two of thestudents began with "one day . . . ." and three wrote "It was a darkand stormy night." Many students, however, seemed to add muchmore information to their opening setting statements at this gradelevel. Student #6 wrote this sentence in her Grade 6 narrative:"It was a dark and stormy night a long time ago when in a citycalled Karapose lived a mouse and her children."The two students who did not use setting statements opened with'beginning action' and 'action dialogue' statements. In March ofGrade 6, all students began with a structural opening. Students atthis time incorporated a wider variety of structural openingstatements: seven use 'setting,' six 'action dialogue,' and one'beginning action.' Two students (#9 and #11) who used 'actiondialogue' statements began with these sentences:"Roger you're such a silly clown,' said Minnie Mouse,"(student #9)"'Ahh,' wailed the child tugging on the sleeve of a tall, primlady."^(student #11)Mean percentages of beginnings for each grade show adecrease in formulaic beginnings from 32 percent in Grade 4 to 71 5 1percent in Grade 5 and 4 percent in Grade 6. (See Table 16.) Thisdecrease is accompanied by an overall increase in structuralbeginnings from 67 percent to 92 percent from Grade 4 to Grade 6.In each of the three years, the highest percentage of structuralbeginnings were 'setting' statements followed by 'action dialogue'and 'beginning action' statements. There were only two weakbeginnings in all of the narrative writing samples.Report Beginnings Results of the analysis of report beginnings indicate onlymarginal increases in complexity for a few students over the threeyear period.^(See Figure 12.)^Further, the level of complexity ofstudents' beginning sentences in report writing is substantially lessthan in narrative writing.In November of Grade 4, eight students had weak and two hadformulaic opening statements in their reports, whereas two usedstructural beginnings: one used a form of 'dialogue' ("I have a dognamed Sandy") and one an 'evaluative' statement ("Pekinese arecute dogs"). Weak 'generalized' beginning statements whichpredominated in the students' writing were usually opening facts.An example of this kind of opening was written by student #2:"Owls can turn their heads all the way around."These opening statements appear to be dependent upon the title ofthe report. Student #1, for example, wrote a report entitled"Nintendo-Legendary Wings." His opening sentence, "In the gameyou are a man with wings," could not stand independently of the titlebecause the referent for the game is otherwise missing. Only fourTABLE 16Percentages of Beginnings and EndingsBEGINNINGSSTRUCTURALNARRATIVEGrade 4^Grade 5 Grade 6REPORTGrade 4 Grade 5 Grade 6Setting 46.0 46.0 57.0 0.0 0.0 0.0Action/Dialogue 21.0 21.0 28.0 0.0 0.0 0.0Thesis Statement 0.0 0.0 0.0 4.0 0.0 4.0Beginning Action 0.0 21.0 7.0 0.0 0.0 7.0Total 67.0 88.0 92.0 4.0 0.0 11.0FORMULAICThis is 0.0 0.0 0.0 8.0 0.0 0.0Formulaic 32.0 7.0 4.0 0.0 7.0 0.0Topic Starter 0.0 0.0 0.0 19.0 0.0 14.0Total 32.0 7.0 4.0 27.0 7.0 14.0WEAKDialogue 0.0 0.0 4.0 4.0 11.0 14.0Evaluation 0.0 0.0 0.0 8.0 11.0 7.0Hello 0.0 0.0 0.0 0.0 0.0 0.0General 0.0 0.0 0.0 65.0 71.0 50.0Total 0.0 0.0 4.0 77.0 93.0 71.0ENDINGSSTRUCTURALSolution/Res. 0.0 11.0 14.0 0.0 0.0 0.0Natural End 11.0 29.0 21.0 0.0 0.0 0.0Evaluation 4.0 0.0 0.0 4.0 0.0 18.0Moral 0.0 8.0 7.0 0.0 4.0 7.0Summary 25.0 29.0 25.0 0.0 0.0 18.0Total 40.0 77.0 67.0 4.0 4.0 43.0FORMULAICThat's all I know 0.0 0.0 0.0 12.0 7.0 4.0Formulaic 14.0 4.0 4.0 0.0 0.0 0.0The End 35.0 11.0 25.0 4.0 0.0 4.0Total 49.0 15.0 29.0 16.0 7.0 8.0WEAKLast Fact 7.0 0.0 4.0 73.0 89.0 45.0No Clear Ending 0.0 8.0 0.0 7.0 4.0 4.0End & Begin Again 4.0 0.0 0.0 0.0 0.0 0.0Total 11.0 8.0 4.0 80.0 93.0 49.0152153Grade 4 students used a formulaic report beginning sentence such as"This is a report about dogs," or a 'topic starter' like "My topic isabout video games." No students had a structural beginning. InMarch there was little change in the results: 10 students openedwith weak beginning statements of which nine had 'generalized'sentences and one had an 'evaluative' statement. Three students hadformulaic beginnings while only one student began with the morecomplex structural 'thesis' statement:"The F-16 Fighting Falcon is the ultimate fighting aircraft."(student #7).In Grade 5, the number of students who used weak openingsincreased to 13 and one person had a formulaic beginning. Nostudent used a structural beginning at this age level. Results inMarch were identical: 13 weak openings and one formulaic opening.Students continued to open their reports with weak 'generalized'facts as they appeared to lack strategies for beginning their writingin more appropriate ways.^Classroom observations at this timeindicated that the students had a great deal of difficulty finding atopic and getting started. One student even used a form of dialogueto open a report which is a strategy that has been noted in previousstudies examining text forms such as beginning statements(Crowhurst, 1983, 1987, 1990; Wilkinson et al., 1980; Langer,1986).^She wrote the following sentence:"If you buy a dog always buy a puppy because they stay longerwith you." (student #5)Grade 6 November reports had few changes in types of154beginning statements. In this grade there appears to be little changein the types of beginnings that students use: the number of studentswho used weak beginning sentences decreased to 10, two studentsused formulaic statements and two used structural beginningstatements. Results in March did not change. It is important to notethat some sentences at this time became difficult to categorizeaccording to Langer's (1986) classification system. Student #2, forexample, used this opening statement:"McDonald's is a good place to eat."This sentence could be categorized as a 'topic starter' (a formulaicbeginning), a 'thesis' statement (a structural opening), or an'evaluation' statement (weak opening). Several of the students'opening statements seemed to fit more than one category, whichmade classification difficult. Also, in spite of the students' continuingto use opening facts predominantly to begin their reports, thecontent of the facts appeared to change. Although the sentences didnot introduce the topic or make a proposition (thesis statement) theyoften contained more information as students grew older. Student#3, for example, wrote:"In volleyball you need certain equipment to play like avolleyball, a volleyball net and a team."Mean percentages for each grade indicate little change in thecomplexity of beginning statements in reports over the three years.In Grades 4 and 5, students used weak beginning statementspredominantly, 77 and 93 percent respectively. By Grade 6 thesepercentages were reduced slightly to 71 percent. The next highestpercentage was in formulaic beginnings, particularly topic starting155statements. Percentages of formulaic sentences decreased from 27percent in Grade 4 to 14 percent in Grade 6. This decrease wasaccompanied by a corresponding increase in structural beginningswhich increased from 4 to 11 per cent from Grades 4 to 6. Thus,whereas report beginning sentences generally were 'weak' and showlittle change over the three years, narrative beginnings were mostly'structural' and in each successive year show increased complexity.As these categories are highly interpretative and classificationsometimes difficult, these results must be viewed with somereservation.Narrative Endings Use of structural endings increased over the three yearperiod. In November of Grade 4 just under half of the students usedformulaic endings to close their narratives: three wrote "the end"and three wrote "they lived happily ever after." (See Figure 13.)Eight students had structural endings: four wrote a 'summary'statement such as "So they took all the money and went to live intheir new house with family and friends," and four wrote a 'naturalending' such as "By the end of the day he had many friends." ByMarch of the same year, 11 students had structural endings: ninewrote 'summary' statements and two wrote 'natural endings.' Thenumber of students who used a formulaic ending decreased to twowhile no student had a weak ending statement.During the analysis it became apparent that some of thestudents' ending statements were very difficult to classify. Fivestudents, for example, wrote, "the end" in addition to including an1 5 6ending sentence which could have been classified as a structuralending. Thus, it was difficult to determine whether these studentshad used a formulaic ending (the end) or a structural one. Similarly,some students' statements appeared to fit more than one category.Student #10 wrote:"The green man told Penny that the guns were fake and heldPenny and Pollywog prisoners."This could be considered a 'summary statement,' a 'natural ending'statement or a 'no clear ending' statement.In Grade 5, 11 students had structural endings in theirNovember narratives. Further, use of a wider variety of categories ofstructural endings was apparent. For example, four students had'summaries,' five had 'natural ends,' one a 'moral' and one a 'solutionresolution' statement. Only two students used formulaic endings atthis time and one student had a weak ending statement. Numbers ofstructural, formulaic and weak endings in March of Grade 5remained identical to those in November.Another difficulty encountered with ending statements is thatthe last statement in a narrative may not be the author's actualending. In a Grade 5 narrative, student #12 wrote "It waswonderful!" which seemed to need the 'natural ending' statementfrom his previous sentence, "Then I smelled the delicious odor offresh pork." to provide a full summation. Generally, however, thestudents' use of ending statements in Grade 5 indicated they have astrong sense of the types of endings that are needed to successfullyclose a narrative. Student #11 demonstrated this capability with the157following sentence:"Kimberly took care of Molly's dog and often took the collie (toMolly's grave) to leave flowers and reflect on their specialfriendship."In November of Grade 6, 10 students had structural endingsand four students had formulaic endings. Similar results occurred inMarch: 10 students had structural closings, three had formulaic andone had a weak closing. Of the 10 students, four concluded with a'solution resolution', two a 'natural end' and four a 'summary'statement. In Grade 6, a much wider variety of ending statementswere used than in previous years. Some students seemed to considerthe purpose for which they were writing and used ending statementswhich seemed appropriate to their type of writing. Several students,for example, use a formulaic ending ("and they lived happily everafter") at this time. It appeared they considered this conventionalstatement as the most appropriate end to their fairy tales. Students#5, #7 #9 and #11 used 'solution resolution' statements to end theirnarratives. Use of these endings indicated the students had toclearly define a problem within the story which needed resolving.Student #5 wrote this example of a 'solution resolution' statement:"Kim and Diane walked onto the stage, did their routine, and atthe end everyone clapped and Diane and Kim bowed andwalked off the stage smiling."In Grade 4, 49 percent of the students ended their narrativeswith formulaic statements, while 40 percent concluded withstructural endings and 7 percent had weak concluding statements.In Grade 5, the percentage of formulaic endings was reduced to 15percent. Correspondingly, structural endings increased to 77 percent.158FIGURE 13Narrative Ending StatementsFIGURE 14Report Ending Statements159Only 8 percent of the students continued to have weak endings. ByGrade 6, formulaic endings increased again to 29 percent. Thisincrease, in turn, reduced structural endings to 67 percent and weakendings to 4 percent. Due to some difficulty with the interpretationof ending statements, however, these results must be viewed withcaution.Report Endings Generally these students' report writing samples showed anoverall decrease in the use of weak endings followed by an increasein structural endings over the three year period. (See Figure 14.)The use of 'weak' endings dominated the writing of students inGrades 4 and 5.^In November of Grade 4, 11 students ended on alast fact' and one had 'no clear ending.' Use of the formulaic ending"the end" or "that's how you . . . ." was limited at this time to twostudents. Typically, students in Grade 4 concluded with statementswhich were last facts such as these two sentences:"It also has two flaps as air brakes" (student # 10)"You need a field to play on." (student #12)Some reports (student #9) end like a personal narrative or concludewith an imperative (student #14). These students wrote:"When I grow up I want to be a professional." (student #9)"Never copy someone's stories but you can get an idea from abook." (student #14)Results in November and March of this year were identical with theexception of one student who closed with the following structural,'evaluative' sentence:1 60"Pekinese are very loving dogs and if you want a dog I wouldsuggest you get a Pekinese." (student #3)In Grade 5 results were similar except more students endedwith weak closing statements: 14 in November and 12 in March.Two students had formulaic endings in their March reports.Generally, the students seemed to have difficulty ending theirreports and continued to rely on ending with a last bit ofinformation. One student's ending statement reflects the difficultyhe had concluding his report:"And I do not want to tell anymore." (student #1)Some students' ending statements showed an increased amount ofinformation at this time. Students #11 and #5 end with thesedetailed statements:"Their nests can be near lakes, rivers, marshes, ponds, sloughsand grasslands." (student #11)"At your last term you get a report card that tells you if youpassed the grade you're in." (student #5)By Grade 6, some students began to have more complexendings in their reports. Five students had structural endings andthe number of students who had weak endings decreased to eight.In March, over half of the students concluded with structural endingswhile six had weak endings and two had formulaic closings. Many ofthe Grade 6 structural endings were 'evaluative' statements. Often'evaluative' closings accompanied reports which were written in firstor second person. Student #9, for example, wrote:"I have to say Michael Jordan is one of the greatest basketballplayers in the N. B. A."1 6 1Again, some of the students' ending statements did not seem tofit the categories in this classification scheme. Students #13 and #9used these statements:"I think all people should try it [basketball] at least once intheir life." (student #13)"The end and remember, be friendly to animals or else theywill go endangered." (student #9)It is difficult to classify statements of this kind which do not appearto fit Langer's (1986) categories. Student #13 appeared to concludewith an opinion which was not a choice provided in thesubcategories. Further, the statement used by #9 best fits the 'moral'ending category which is an atypical ending for a report. Thisstudent appears to have adapted a narrative ending to end herreport. Use of narrative endings to conclude non-narrative writingtasks is a strategy which has been noted in previous research(Crowhurst, 1983, 1987, 1990; Gorman et al, 1988; Wilkinson et al.,1980).By Grade 6, the students were just beginning to explorealternative endings from 'last fact' sentences. Clearly, they continuedto struggle to find interesting and appropriate ways to end theirreports. Even the 'structural' endings that they used did not yetresemble those that might be found in skillful adult writing.An examination of mean percentages of ending statementsused in each grade demonstrates the students' increasing ability touse more complex endings, particularly in Grade 6. (See Table 16.)In Grades 4 and 5 weak, 'last fact' endings dominated the students'report writing. Students used 80 percent of these endings in Grade 4162and over 90 percent in Grade 5. Over the two years, only 4 percentof the students' ending statements were structural while 16 percentwere formulaic. By Grade 6, use of weak endings decreased to 49percent and formulaic endings to 8 percent. Correspondingly, therewas a rise in structural endings to 43 percent. Again, ascategorization in this way is subjective, some caution must be usedwhen interpreting these results.Length of Writing SampleResearchers have traditionally used length of writing sample asan indicator of maturity in writing. As students grow older, length ofwriting tends to increase (Crowhurst, 1978; Harpin, 1973, 1976;Kress, 1982; Langer, 1986; O'Donnell et al., 1967). Furthermore,length of writing has been found to be affected by mode of discourse(Crowhurst, 1978, 1983, 1987; Harpin, 1976; Hidi & Hildyard, 1983;Langer, 1986; San Jose, 1972).^Indeed, results of this study supportthe previous research.^The students' writing samples increased inlength as they became older and there were differences in the lengthof writing samples in the students' narrative and report writing.Results also indicate that students' writing did not increase in asteady and continuous pattern over the years.NarrativeResults of the examination of length in narrative writingsamples show individual fluctuations within the three years butoverall increases in length from Grades 4 to 6. (See Table 17.)1 6 3Students in Grade 4 had a median 13 length of 307 words in theirNovember narratives. (See p. 171, Figure 15.) Individual student'swriting tended to be within approximately 150 words of the grademedian as illustrated by the range of 185-474 words. In March, thelengths increased and ranged from 152-712 words. The median wasslightly lower at 276 words. Within the years, students tended tohave fluctuations in length of writing. In Grade 4, students hadvariations of up to 500 words in the lengths of their narratives.Reasons for individual fluctuations in length of writing samples aredifficult to ascertain. One reason for longer narratives appeared tobe related to students' use of chapters in their stories. Two students,#3 and #4, increased the length of their narratives by over 300words from November to March when they wrote several chapters.In Grade 5 only 50 per cent of the students began the yearwith longer narratives than in Grade 4. In November, the range(from 176-585 words) was actually smaller than in the previousMarch. The grade median, however, was higher at 332 words. InMarch of Grade 5, 12 students wrote narratives which were longerthan in November. The range of number of words increasedconsiderably: from 185-1505 words. The median increased to 402words. Three students wrote well over 1000 words in theirnarratives. Reasons for these three students' large increases inlength are not clear. One student used several chapters which1 3^As the lengths of some students' narratives are either exceptionallylong or short, grade means are not a valid means of comparison.Therefore, median lengths are used to compare students' narratives andreports. The median length is the length that is in the middle of theranked order of lengths.164TABLE 17Length of Writing SamplesGrade 4Length of Writing SamplesGrade 5 Grade 6Nov. Mar. Mean Nov. Mar. Mean Nov. Mar. Mean#1 (n) 239.0 172.0 205.5 175.0 185.0 180.0 349.0 888.0 918.5(r) 108.0 132.0 120.0 111.0 85.0 98.5 183.0 213.0 198.0#2 (n) 246.0 385.0 315.5 433.0 704.0 568.5 698.0 1013.0 855.5(r) * 77.0 77.0 130.0 137.0 133.5 201.0 178.0 189.5#3 (n) 388.0 683.0 535.5 468.0 1178.0 823.0 569.0 753.0 661.0(r) 123.0 130.0 126.5 220.0 308.0 264.0 362.0 228.0 295.0#4 (n) 214.0 712.0 463.0 331.0 1216.0 773.5 413.0 1186.0 799.5(r) 78.0 165.0 121.5 113.0 200.0 156.5 255.0 152.0 203.5#5 (n) 411.0 276.0 343.5 540.0 627.0 583.5 762.0 1185.0 973.5(r) 125.0 58.0 91.5 142.0 175.0 158.5 228.0 318.0 273.0#6 (n) 363.0 240.0 301.5 180.0 209.0 194.5 360.0 925.0 642.5(r) 105.0 101.0 103.0 70.0 103.0 86.5 366.0 146.0 254.5#7 (n) 240.0 219.0 229.5 285.0 264.0 274.5 1043.0 326.0 684.5(r) 84.0 120.0 102.0 88.0 157.0 122.5 254.0 125.0 189.5#8 (n) 235.0 256.0 299.0 256.0 269.0 262.5 371.0 367.0 369.0(r) * 42.0 42.0 * 67.0 67.0 142.0 129.0 135.5#9 (n) 307.0 152.0 229.5 308.0 398.0 353.0 600.0 517.0 555.5(r) 43.0 82.0 62.5 65.0 164.0 114.5 289.0 99.0 194.0#10 (n) 185.0 200.0 192.5 176.0 528.0 352.0 571.0 895.0 733.0(r) 65.0 61.0 62.5 77.0 126.0 101.5 314.0 186.0 250.0#11 (n) 474.0 230.0 352.0 315.0 1505.0 910.0 807.0 721.0 764.0(r) 73.0 30.0 51.5 110.0 140.0 125.0 395.0 227.0 311.0#12 (n) 414.0 233.0 323.5 222.0 236.0 229.0 414.0 555.0 484.5(r) 92.0 55.0 73.5 83.0 173.0 128.0 144.0 127.0 135.5#13 (n) 301.0 411.0 356.0 384.0 272.0 328.0 582.0 433.0 507.5(r) 55.0 103.0 79.0 113.0 128.0 120.5 183.0 453.0 318.0#14 (n) 349.0 318.0 333.5 585.0 402.0 493.5 731.0 716.0 723.5(r) 204.0 107.0 155.5 108.0 158.0 133.0 185.0 193.0 189.0Mean (n) 319.5 320.5 320.0 332.7 570.9 451.8 590.7 748.6 690.9Mean (r) 96.3 90.2 90.6 102.1 151.5 129.3 250.1 198.1 224.0(n) = narrative ' not acceptable for analysis(r) = report1 6 5appeared to extend her story. Another student had several largesections of dialogue in her story which seemed to be responsible forits increased length. The third student simply seemed well-motivated and intrigued by her work or perhaps her choice of topicwhich may have encouraged her to write more.Eleven students' narratives were longer in Grade 6 than inGrade 5. Some students showed marked fluctuations in their writingsample lengths at this time. In particular, the three students whohad written over 1000 words in Grade 5 wrote just over half thatlength in November of Grade 6. The length of most students' writing,however, continued to increase. The range was 349-1043 with amedian of 582 words in November and of 753 words in March.Students continued to have much variation in the lengths of theirwriting samples throughout the year.Reasons for the individual fluctuations in length of writingsample are unclear.^Classroom observations suggest that motivationwas a contributing factor. Some students appeared to choose topics,for example, which did not seem to motivate them. Why certaintopics do not appear to motivate students is a difficult question toanswer. Perhaps these students chose topics about which they hadlittle experience or perhaps the topic did not lend itself to muchelaboration. Further, other motivational factors seem to contribute tothe amount of writing that students produce within one year.Student #7, for example, had shorter story lengths in March of eachyear. In Grade 6, his March story was one third the length of hisNovember story. It would appear that the time of year affected hiswriting performance. Observations during the writing sessions1 66seemed to confirm that he lacked motivation toward his school workgenerally and particularly toward his writing at this time. Otherstudents, at one time or another, displayed a general lack ofmotivation toward their schoolwork which, in turn, affected theirwriting. It is impossible to speculate about the numerous personalreasons which could have affected these students' motivation.Thus, often length of writing sample is affected by the kind ofwriting (e.g. chapters and dialogue), topic, and motivation. Resultsshould be viewed with these factors in mind. It is also generallyapparent that students increased the lengths of their narratives eachyear. As they mature, students seemed to add many more events totheir stories. Similarly, they are apparently much more attentive toimportant details in describing story events which, in turn, increasesthe lengths of their narratives.ReportResults of the study of length of report writing indicate thereare the same fluctuations in growth of individuals' writing butoverall increases for the three year period. (See Table 17.) The dataalso show students wrote considerably less in their reports than intheir narratives. A grade by grade comparison of the ratio of themean length of narrative writing to the mean length of reportwriting (for each individual) shows narratives are consistently threeto four times longer than reports. (See Table 18.) In November ofGrade 4, students wrote reports that were up to seven times shorterthan narratives written at the same time of year. Observations ofthe students during the writing sessions indicated most students had1 6 7a great deal of difficulty writing reports. Even though types of topicswere discussed and the students seemed to clearly understand thepurpose of report writing, they displayed much frustration withwriting in this genre. The students asked many questions and haddifficulty finding topics and getting started. As a result of thisdifficulty, many students' reports were very short. Over 50 percentof the students wrote fewer than 100 words at this time.Interestingly, two students had difficulty with writing in this genreand wrote reports in narrative form (these were excluded fromanalysis). As stated previously, researchers have found that theirsubjects also wrote narratives or partial narratives in response tonon-narrative task demands (Crowhurst, 1983, 1987, 1990; Gormanet al., 1988; Harris & Wilkinson, 1986; Wilkinson et al., 1980).They note that students of this age often rely on their narrativeschema, and lapse into narrative form, when they are uncertain as tothe demands of the non-narrative writing task.The range of lengths for Grade 4 reports was between 43 and204 words with most students writing approximately 80 words perreport. The median of the reports in November was 84 words. InMarch of the same year, eight students increased the length of theirreports but six wrote shorter reports. The range narrowed slightlyfrom 55-165 words but the median increased to 101 words.In Grade 5, only 50 per cent of the students' reports werelonger in November than in Grade 4. The median length at this timewas 110 words.^By March, 13 out of 14 students had the increasedlength of their reports. The median length for the grade increased to157 words. In November of Grade 6, 13 students wrote longerTABLE 18Ratio of Mean Length of Narrative toMean Length of ReportStudentRatio ofGrade 4Narrative to ReportOverall lengthGrade 5 Grade 6#1 1.7 1.8 4.6#2 4.1 4.2 4.5#3 4.2 3.1 2.2#4 3.8 4.9 3.9#5 3.8 3.7 3.5#6 3.0 2.2 2.5#7 2.3 2.2 3.6#8 7.1 3.9 2.7#9 3.7 3.1 2.8#10 3.1 3.5 2.9#11 6.8 7.3 2.5#12 4.4 1.8 3.6#13 4.5 2.7 1.6#14 2.4 3.7 3.8Grade Mean^3.9 3.4 3.2168169reports than in the previous year. Overall, the grade medianincreased to 254 words with a range of 142-395 words.Interestingly, in March most students' reports decreased in length;only four students had longer reports. This reduction in length bymost students decreases the median to 186 words, over 60 wordsless than in November of the same year.Fluctuations in report writing seem to be caused by many ofthe same factors which affect narrative writing. For example, choiceof topic appears to have an influence on the length of reports.Students often chose topics which interested them but about whichthey had little general knowledge. Even the topics about which theyhad information seemed to create difficulty for them. They seemedunable or unwilling to write down information that was apparentlyavailable to them during conversation. Further, they appeared tolack the ability to speculate about information on a topic.^Evidently,they were unwilling to put down information unless they knew it tobe completely accurate. Additionally, students' attitudes towardreport writing seemed different from their attitudes towardnarrative writing. On the whole, the students were less interestedand found report writing difficult which, in turn, appeared tonegatively influence their motivation.Figure 15 compares the grade median lengths in narrative andreport over the three year period. While students' writing in bothgenres showed overall increases each year, this figure also highlightssome interesting comparisons of student growth in each of thegenres. The diagram accentuates the difference in median lengths innarrative and report. Length in narrative writing is consistently170three times longer than length in report writing over the three yearperiod. The differing patterns of growth in the two genres are alsoapparent. In narrative writing there is little growth until March ofGrade 5. Lengths then sharply rise in Grades 5 and 6. In report,however, the growth is gradual and much less dramatic. Further,whereas length of narrative continues to rise in Grade 6, length inreport writing decreases in year end writing samples.In summary, the lengths of the students' reports show gradualgrowth over the three year period. However, in each year students'reports were on average at least three times shorter than theirnarratives. The shorter lengths of reports seem attributable to thedifficulty the students had with report writing. Classroomobservations both at the beginning of the study and over the threeyears indicate that the students had difficulty choosing topics, gettingstarted, and generating enough content to sustain their writing.Also, their attitudes toward writing in this genre appeared much lessfavourable than their attitudes toward narrative writing.Apparently, many of the factors which affect length of writing innarrative such as choice of topic and motivation also affect lengths inreport writing.171FIGURE 15Median Lengths of Narrative and Report172CHAPTER VPROFILES OF THREE WRITERSTo provide additional information about the use of syntax andtext form in the narrative and report writing of students in theirupper elementary school years, case studies of three studentsconstitute this chapter. The students were chosen because of thedifferences in their abilities as writers at an early age and thegrowth and changes in their writing over the three years.^Further,they were chosen because of interesting and atypical syntax and textforms that were found in their writing samples during analysis. Forthe purpose of describing the students and their work, threepseudonyms have been adopted: John, Tina and Ian. The followingis a description of the three students, a brief synopsis of observationsnoted during the writing sessions, and a comparative analysis of thestudents' writing. The purpose of the comparison is to highlight thediffering syntactical elements and text forms that the students usedin their narrative and report writing over the three year period.Description of the Students and SessionsThe following is a description of the students' academic abilitiesas described by school records and teachers and some generalobservations about the students made by this researcher during thewriting sessions.173A Brief Description of the Students This study is a description of linguistic structure and text form;therefore the observations that were recorded during the three yearsare limited. Nevertheless, they do provide some insight into thestudents' academic abilities, interest incapabilities.Tina (Student #11)Tina is described by her teachersHer report cards and standardized testswriting, and writingas an "exceptional" student.indicate she is an "A" studentand has been very capable academically from an early age. Tina is awarm, friendly and cooperative student. She is reserved and quiet,yet shows intense interest in all areas of her schoolwork. She setshigh goals for herself and rarely fails to achieve them. Tina readsavidly and particularly enjoys novels. From the beginning of thestudy it was very apparent that Tina enjoyed writing and had littledifficulty with the writing process. She confidently and carefullydrafted, edited and proofread her work. Tina proved to be anexceptionally interesting subject because of her apparent ability towrite well from a very early age. As compared to most otherstudents, her writing is longer, appears more syntactically complexand receives high assessments from her teachers. Tina's writing isreviewed here in order to address the question: What linguisticstructures and text forms does she use to produce such sophisticatedwriting?^Despite her facility with written production, Tina appearedto find report writing very challenging. However, substantialchanges occurred which seemed to improve her report writing overtime. It therefore seemed pertinent to consider which elements of174syntax and text forms accompanied these changes.John (student #7)John is also a very capable student. His report cards andstandardized tests indicate he is above average in most subject areas.John is a confident, outgoing student who displays a keen interest incertain areas of the curriculum, particularly Science, Math and SocialStudies. John approached his writing with less enthusiasm than Tinabut was by no means a reluctant writer. He wrote quickly and withconfidence, taking time to edit and proofread his work. John alsoreads extensively but rarely reads novels. He has a strong interest innon-fiction books, especially those which provide information aboutfast cars and fighter planes. John's writing was especially interestingbecause of his apparent ability (at an early age) to perform morecompetently and confidently in report writing than the otherstudents. It was interesting to compare the syntax and text formsused by John in his report writing with those of the other students.While his strength is report writing, John is also quite competent innarrative writing.Ian (student #8)Ian is described by his teachers as an "underachiever." Testresults and report card marks indicate he is an average student. Ianis a fairly quiet and cooperative student who readily admits hewould rather be playing any sport than be in school. While Ian readswith no difficulty, he does not spend time reading books outside ofthe classroom. He does, however, read sports articles and the sportssection of the newspaper. Ian is not a well-motivated writer. Overthe three years he showed a growing interest in writing but found175the process somewhat tedious. Ian's work was interesting to studybecause, despite some reluctance on his part, his writing exhibitedsome sophisticated features, particularly in syntax, early in Grade 4.Also, Ian's writing showed the most change in the use of morecomplex syntax and text forms over the three years and thusprovided an interesting comparison with the other two writers.A Brief Description of the Writing SessionsThe following is a synopsis of some very general observationsabout the 14 students as they participated in the writing sessions.As stated previously, all students were familiar with the writingprocess. At the outset of the study, in Grade 4, the students werebriefed on the procedures of the study. After this introduction, theywere asked to write a narrative. All students had little difficultysetting about this task; they had a good idea of what was expected ofthem, had no difficulty generating ideas and were obviouslycomfortable with the task. With the task of report writing, however,the students were obviously much less comfortable. After anintroduction and a request to initiate the writing, "write aboutsomething you know a lot about," the students readily understoodthat the task was different than story writing and more importantly,unfamiliar. The students perceived report writing to be verydifficult. Characteristically, questions such as the followingpermeated the writing sessions:^"What will I write about?" "How doI begin?" "I don't know how to do this can you help me?" Even1 7 6though the students were able to brainstorm about lists of topics,understood that they needed to write factually, and were given help(in the form of questioning) to generate ideas and get started, theywere noticeably agitated and frustrated with the task. Almost all ofthe 14 students found writing reports very difficult, especially thefirst report. Even after three years, some students continued toexperience feelings of frustration and anxiety when writing reports.Another interesting observation was the amount of time takento write in each genre. Students persevered with their narratives.They edited, rewrote and expended considerable time, sometimesthree to four weeks or more to complete one piece of writing. Withreport writing, however, the students took much less time to writeand even after three years, rarely spent much time editing theirreports. Some students had to be reminded to rewrite rough draftsbefore submitting them to the researcher or teacher. Often thestudents rewrote their good copies exactly as their rough drafts hadbeen written, correcting few or none of the drafting errors. On theother hand, no students had to be reminded about rewritingnarratives.Observations over the three years indicate that even thosestudents who were able to write reports with some confidence didnot always enjoy the process. For almost all students, writing reportswas considered a duty, or something done because it was requiredby the teacher, whereas narrative writing was generally anenjoyable experience and not always considered to be a requirement.177Tina. John and Ian - Writing Sessions In Grade 4, Tina, John and Ian each handled the two writingtasks differently. Tina approached writing her narrative withobvious enthusiasm, John wrote dutifully, and Ian persevered withthe task at hand. When asked to write a report, however, thestudents' attitudes changed. Tina was visibly upset. Her confidenceas a writer seemed to diminish and she had a great deal of difficultybeginning to write and sustaining her efforts. In contrast, Johnbecame excited at the prospect of writing a report. He knew just thecar he wished to write about and his only apparent difficulty wasinitiating the writing exercise. Ian seemed to show little or nodifference in the way he approached the report writing task ascompared to the narrative task.Over the years, John continued to approach most of his reportwriting with enthusiasm and confidence but lacked enthusiasm whenwriting narratives, particularly in the latter part of Grade 6. Heseemed to be inconsistent in his level of motivation in each year.Although a very capable student, his enthusiasm for schoolwork ingeneral, and writing in particular, varied over time. At times duringthe years, John was easily distracted from his work and seemed toprefer to socialize rather than attend to the task at hand. He almostappeared to question the value of attending to his schoolwork. Tinabecame more confident in report writing but still much preferred towrite in the narrative form. Her level of enthusiasm, however,remained relatively constant.^Over time, Ian seemed to mature andbecome more interested in his writing. He displayed much lessreluctance toward writing in both genres.178The following account is a description of the use of syntax andtext forms in the writing of the three students.Syntactic ComplexityThe following is a comparison of the students' use of syntax intheir narrative and report writing. The comparison highlights theindividual differences in use of syntax over the three years.T-UnitOver the three year period it is interesting to see the growth incomplexity, as measured by mean T-unit length, in the threestudents' narrative writing. Two of the three students demonstratedgrowth in the complexity of their writing as measured by T-unitlength. Fluctuation in mean lengths is apparent in each grade foreach student.Tina's mean T-unit length in narrative writing increased fromjust over seven words per T-unit to 11 words per T-unit from Grades4 to 6. This increase in T-unit length seemed to be attributable toher increasing ability to extend noun and verb phrases and addclauses to her writing. As she matured, Tina incorporated a greatdeal more descriptive detail into her narratives. Often she seemed tobe able to create a mood or tension in her stories as she apparentlybegan to understand function of the language and her repertoire oflinguistic elements expanded. This overall competence in languageusage increased her mean T-unit length. This change in language isreadily apparent as one compares several T-units from a Grade 4179narrative to those in a Grade 6 narrative:"Once they arrived at the mansion they shrieked with joy.//The mansion had over 100 rooms // and there were 11gardens.// They even had servants.// Claire met her maid,Lorraine."//(6.2 words per T-unit)Grade 4"The dark, looming gates of Musselwaite Manor clashedtogether.// Across the street from the Duff House, the gates ofMusselwaite Manor creaked slowly open.// Tanya leanedtoward the window.// Todd, who had come from the livingroom into the kitchen, pointed out a dark figure prowling out ofthe gates.//(12.2 words per T-unit)Grade 6Admittedly, the intent of the two paragraphs is slightly different;the second one is meant to create a mysterious mood whereas thefirst one simply introduces a setting. This difference makes strictcomparisons difficult. Nevertheless, the content of the language Tinaused is more complex in the Grade 6 narrative, resulting in increasedT-unit length.^(Some of the elements of syntax that Tina uses toachieve longer T-units are discussed in the following sections.)In each year Tina's writing showed fluctuations in mean T-unitlength in narratives. Often T-unit length varied from two to threewords per T-unit. As stated previously, dialogue appears to affectmean T-unit length. A short example of dialogue from her Grade 6narrative demonstrates this effect:"Wake up, Todd said cheerfully.// "I made my breakfastmyself.// It's ready now."// Tanya turned over.//"Uh huh don't bother me now.// I'll come down later," shemumbled through her pillow drowsily.//1 8 0"Come on," whined Todd impatiently.// "Hurry up."// I'mgoing downstairs now," Todd called back from the doorway.//(5.8 words per T-unit)Over the years the students did not appear to change thelanguage that they used in their dialogues. The informal languageused in these conversations results in shorter T-units. The stylistictendency to use dialogue, which is so common in students' writing atthis age level, calls into question the use of T-unit length as anaccurate measure of the complexity of the language. None of theprevious studies which used T-unit length to measure syntacticcomplexity in narrative writing of children of these ages mention theuse of dialogue (Crowhurst, 1978; 1980a; 1983; Harpin, 1973, 1976;Hunt, 1965; Langer, 1986; O'Donnell et al., 1967; Perron, 1977;Richardson et al., 1976; San Jose, 1972). Perhaps researchers needto account for the types of language subjects use, such as dialogue,(and other kinds of complexity that it entails, such as voice andperspective-taking, which a simple index such as T-unit fails tocapture) when they conduct studies which measure T-unit length.In both Grade 4 narratives, John had the longest mean T-unitlengths of all 14 students (12.2 words per T-unit). Curiously, thelanguage (the content and complexity of linguistic elements) in hisstories seemed to change slightly over the three years but his meanT-unit length did not. Over the years John's mean T-unit lengthdecreased slightly indicating that, on the average, his T-units wereshorter. Yet, his work contained some substantially longer T-units inGrade 6 than in Grade 4. Here is a comparison of the first part of twoof his narratives:1 8 1"One day a man thought I'll play a little game with the policetoday.// So he jumped in his red Ferrari Testarosa and speddown the road.// When he approached the highway he wentover 200 mph.// Just when he passed on a ramp he heard asiren."//(10.9 words per T-unit)Grade 4On March 26, 1977, in the city of Manaus in the Amazon junglein Brazil, the famous explorer, Indiana Jones, was about to starton an expedition to look for the lost temple of doom and itssolid gold statue in the mighty Amazon jungle.// His firstadventure started to happen when he was packing his whip,gun, food, matches, torch, knife boomerang canteens hikingstraps and backpack.// When he opened his backpack hefound the warning note in his pocket.//(26 words per T-unit)Grade 6John showed his early capability as a Grade 4 writer by using clauses("When he approached the highway") and an extended noun phrase("his red Ferrari Testarosa") in this piece. As he grew older, hebecame even more proficient at providing descriptive detail, oftenusing complex linguistic features, as demonstrated by these few T-units. It is clear that while his mean T-unit length (that is, theaverage length of his T-units) may not have changed, his ability touse some much longer T-units and more complex language did.Mean T-unit length seemed to obscure the fact that his languagebecame more complex.John's work also had fluctuations in mean T-unit length withinany one year which may be attributable, at least in part, to his levelof motivation. Some of his stories (usually written in November)appeared to have much more detail and better reflected hiscapability as a writer. It may be the type of story (adventure,1 8 2fantasy, mystery etc.), or his choice of topic (whether it relates to hisexperience), attitude, or simply the time of year that causes John notto always approach various writing tasks with equal enthusiasm. Attimes he seemed to question the importance of the writing task.John's fluctuating levels of motivation demonstrate an interestingproblem for the researcher in studies of writing; that is, to ensurethat sampling procedures attempt to accommodate the variable ofmotivation of the writer. Clearly, basing analyses on a widersampling of writing over a period of time is important to adequatelyreflect the capabilities of a writer.By contrast, Ian's mean T-unit length steadily increased in eachyear in narrative writing. Generally, Ian's stories were not asdetailed or had as complex language as those written by John andTina. As he matured, however, his writing changed and developed ashe began to incorporate the more descriptive detail and complexlanguage that John and Tina used in earlier years. His T-unit lengthincreased accordingly. These two narratives provide an example ofthis change:"Once a garbage truck was coming// and they threw nails outin front of the garbage truck// and the tires popped. 'Nowwhat happened,' said the garbage truck driver."//(7.2 words per T-unit)Grade 4"It was about three o'clock in the morning when I was awokenby four gunshots.// Don Karnish had killed Fred Banks.// Icould hear their planes roar passed my house.// I ran outsideto get in my plane to follow them!! but I couldn't find mykeys."//(9.5 words per T-unit)Grade 61 8 3Fluctuations also occurred in Ian's mean T-unit length. Onereason Ian may have used shorter T-units in his stories is his choiceof topic. It appeared that by choosing a topic which enabled him todraw upon information from his recent experience, his writingbecame much more detailed. Correspondingly, his T-unit lengthincreased. Soon after seeing the movie, Robin Hood Prince ofThieves, Ian wrote a narrative which he based on the movie. Hecombined events from the movie with events of his own. This partialre-telling of a story is a strategy which appeared to help himincrease syntactic complexity in his narrative. His mean T-unitlength increased to 11.5 words per T-unit which was one of thelongest in Grade 6. He was able to provide much more specific detailas he described the scenes of his story. This example providesevidence of the effect of writing from vicarious experience and theimportance of students having the freedom to choose their own topic.T-unit length also increased in the three students' reportwriting over the three years. Further, T-units are consistently longerin report writing than in narrative writing. Over the years Tina'smean T-unit length increased from 8.32 to 15.0 words per T-unit inher reports. In each year but Grade 4, she had longer T-units in herreport writing than in her narrative writing.In Grade 4 it was readily apparent that Tina found reportwriting very difficult. She obviously lacked experience with thisgenre as she asked several questions as to how to write reports. Shefailed at several attempts to get started. Her behaviour at this timewas surprisingly different from her behaviour when she was askedto write a story. She seemed frustrated as opposed to confident. The1 8 4difference in her use of language and T-unit length was mostapparent in her early reports as compared to her early narratives.This is exemplified in the following excerpt:In school you learn.// Teachers help us learn.// There aresubjects like Spelling, Math, Social Studies, Whole Language,Writing, Reading, Music, Science and P. E.// School starts from9:00 to 3:00. // There are many Grades.// Every term thereare report cards to show your grades.//(7.1 words per T-unit)This reports consists of simple statements with little elaboration ordetail. Also, discounting the second T-unit , the mean T-unit lengthdropped from 7.1 to 5.6 words per T-unit. At this time, Tina's onlyapparent strategy was to write short, factual statements. She alsoincluded a list, which is a common strategy at this age. Tina'sdifficulty with reports seemed to correspond with a reduced T-unitlength in her report writing. Perhaps her lack of experience withreports or her struggle with writing them interrupted the flow oflanguage she readily used in narrative. Over time, however, Tinahad less difficulty with report writing. In Grade 5, she still includeda long list in her report but also had longer T-units such as:"When one of the 88 keys of the piano is struck, one hammerhits one of the strings causing it to vibrate and make a sound."(26 words)By Grade 6 Tina did not have any lists in her reports. Thetypes of reports she wrote also changed from simple descriptionsabout items to instructions as to how to play volleyball and collecthockey cards. These reports contained longer T-units. An example ofa T-unit from her report on volleyball demonstrates the complexity185of many of her T-units:"The side arm swing serve is performed by standing sideways,holding the ball at arm's length and hitting the ball with anarm, swinging straight at the ball."(27 words)This long T-unit indicates the complex thinking needed to organizeand present such a complicated series of instructions. At this age,Tina seemed to be able to handle such an intricate cognitiveoperation.Ian also struggled with report writing, particularly in Grades 4and 5, yet his T-unit length increased over time. Discussions withIan in Grade 4 indicated he believed he had a clear idea of thefunction of a report. However, when he began his composing areport on Sea Otters, he wrote "Attack of the Sea Otter," in narrativeform. By March, after several attempts at getting started, Iancompleted the following report. His T-unit length was comparable tohis narrative writing at that time:"At Disneyland there is a ride called the Pirates of theCarribean.// You go in a boat// and you see pirates chasingpeople and lighting towns on fire.// You go down hills// andthere are people shooting people and putting people in jail."//(8.2 words per T-unit)Ian appears to be telling or 'reporting' about the ride. The use of thesecond person appears to give the report a conversational flavour.Bereiter and Scardamalia (1982) found their subjects often relied on"open discourse schemata" (conversation) when writing exposition;samples of writing resembled one part of a conversational turn.Perhaps Ian was relying on conversation as a strategy for writingthis report. While it is very short, Ian used some complex syntax in1 8 6this report (i. e., Stage II relative clauses). His use of these clausesalso reflects the fact that his report reads like part of a conversation.O'Donnell et al. (1967) observe that Stage II relative clauses (thosewhich use the present participle of the verb, "ing") occur early andare common in children's speech yet they do not appear until later inwriting. Ian's use of these clauses may be a result of theconversation-like style of his report or an indication that he wasbeginning to use more complex syntax.Ian continued to have difficulty with writing a report inNovember of Grade 5. He wrote a report about the Stanley Cup andlapsed into narrative form part way through. As stated previously,many researchers have noted that students of this age will writenarratives or partial narratives in response to non-narrative tasks(Bereiter & Scardamalia, 1982; Crowhurst, 1983, 1987, 1990; Kress,1982, Harris & Wilkinson, 1986; Gorman et al., 1988; Wilkinson etal., 1980). In March, he wrote a report on hockey which had a meanT-unit length of 11.6 words but he included several lists whichappear to have inflated his T-unit length. In November of Grade 6,Ian continued to write simple reports with little elaboration of detail.By March that year, the language in Ian's report changed which, inturn, seemed to create longer T-units. He wrote a report whichprovided instructions about how to play basketball, again adopting asecond person stance. This report also resembled conversation. Ianappeared to relate this sequence of instructions as if teachingsomeone to play the game. The following are examples of T-unitsfrom this report:1 8 7"There are many rules you must learn before you playBasketball.// One is you must bounce the ball every two stepsor else its called travelling// and the other team gets to bringthe ball in from the side.// Also, if you touch someone whenthere [sic] shooting they get two foul shots// and if they sinkone there [sic] worth one point."//(12.2 words per T-unit)Even though Ian's T-unit lengths were longer, the flow of language inthis report compared to his narratives seems to reflect that he stillhas some difficulty with writing in this genre. Some of the lengthyT-units he created also presented difficulty for him as in thefollowing example:"You play basketball by both teams try to get the ball fromeach other and get it in the other team's basket."This example indicates he has some problems holding the first partof the sentence in his short term memory. As a result, he makes agrammatical error (specifically the relationship of the adverbial "by"clause to the matrix sentence).John's mean T-unit length in report writing seemed very highin comparison with the other students. Further, he also hadfluctuations of up to seven words per T-unit in his Grade 4 and 5reports. In Grade 4, John was the only student who apparently hadlittle difficulty writing a report. He had no reference materialavailable to him yet he produced a detailed descriptive report with amean T-unit length of 15.0 words. The following are T-units takenfrom his report:"The F-16 Fighting Falcon is the ultimate flying aircraft.// Ithas one Pratt and Whitney rocket engine and is capable ofreaching Mach 2 (2160 mph).// The F-14 Tomcat has two Prattand Whitney rocket engines and reaches just over Mach 2(2500) mph - just a tad faster than the F-16.// The F-16 has1 8 8awesome menorbility [sic] and can skim around in circles whileturning.!!(12.2 words per T-unit)Some of the fluctuations in John's mean T-unit length seemed to becaused by the type of report. In Grade 5 he wrote a lengthy reportdescribing Indy 500 cars which had a mean T-unit length of 10.46.Although comparable to his mean T-unit length in narrative writing,this mean was less than in previous reports. By contrast, a reportwritten in the same year giving instructions on collecting cards had amean T-unit length of 17.6 words. John wrote this report in secondperson adopting a similar conversational strategy as Ian's. John,however, used a series of imperatives to explain card collecting asdemonstrated by this example. His T-units are longer in this style ofwriting:"To start collecting you need to buy some pages and an albumto hold your cards in.!, Then you go out and buy somecards.// You can stick with one type of card or collect otherones like Score, Tips, Leaf, Ob and Upper Deck."//(15.0 words per T-unit)As with narrative writing, John appeared less motivated to writereports at certain times, which appeared to create fluctuations in T-unit length. In Grade 6, his November report had a mean T-unitlength of 14.0 words, yet in March, he wrote a report which had amean T-unit length of 9.6. His March report was considerablyshorter and lacked sufficient detail to provide much interest for thereader.1 8 9Extended Noun Phrases - Noun Phrase with Modifiers All three students consistently used modifiers to extend thenoun phrase in their narrative writing. The frequency of nounphrases with modifiers increased in each year. Both Tina and Johnused higher frequencies of this form in their narratives than moststudents, particularly those in Grade 4.In this study, some elements of syntax seem more prevalent incertain types of stories. These students, for example, tended to havemore noun phrases with modifiers in fantasy stories as opposed toother forms of narrative. Tina had the highest incidence of anystudent in her Grade 5 fantasy story, "I Am Doll," which was richwith descriptive phrases. An example of this description is containedin the following sentence (extended noun phrases "with modifiers"are italicized):"I also had a pale white face with rosy cheeks, big blue eyes, asmall red nose and a small pink mouth."John also had high frequencies of noun phrases with modifiersin a fantasy about a "Secret Ninja." He had phrases such as, "anexperienced, Ninja craftsman " and "two, medium-sized diamonds."When these two students wrote fantasies, their frequencies ofmodifiers increased. While their increased usage may be acoincidence, they may also have increased modification to morevividly • create the imaginative scenes that would draw the readerinto the fantasy.Ian did not have modifiers in his narratives until March ofGrade 4. In each year he had fewer modifiers in his narratives thaneither John or Tina. His sentences were quite simple and included1 9 0little description. By Grade 6 Ian's writing began to change. Someelaboration of noun phrases was used at this time. Ian includedseveral modifiers in an adventure story about a treasure map. Atthis time Ian began to use more precise description as the followingexample illustrates:"A little way ahead of me, I saw a worn down trail."John is the only student who had modifiers in a report in Grade4. He had a high incidence of this form in his March report about theF-14 Falcon. In each year his use of modifiers increased and he hadthe highest incidence of any student in March of Grade 5. Johnseemed to have more modifiers in his reports when he was veryfamiliar with his topic (usually because he had recently read aboutit). 14 This seemed to enable him to include many specific details. Hisreport on the Indy 500 cars had several noun phrases with modifierssuch as, "the magnificent European Formula 1 race car." John's useof many descriptive phrases in his reports, especially from an earlyage, seems to illustrate the powerful effect of other literacy activitieson his writing such as reading extensively on a topic.Tina did not have modifiers in her reports until Grade 5. Atthat time, she had several modifiers in her report on "Ducks." Sheused phrases that provided specific detail about her topic such as: a1 4^Throughout the study these students' teachers used a whole languageapproach to reading. Student reading materials were chosen to supportthe theme currently being studied in the program. Materials weretaken from a variety of sources including reference materials, novels,and basal readers. Further, the children were given time andencouraged to read independently using their own reading materials.As stated previously, John often chose non-fiction material forindependent reading. He wrote reports about many of the items in hisindependent reading materials.1 9 1bright yellow bill," "a glossy green head" and "a mallard female'sbody." Tina's use of modifiers increased in her Grade 6 reports. Herincreased use of this form resulted in better description whichenables her readers to have a better understanding of the object thatshe is describing.Ian rarely used modifiers in his reports; one report in Grade 5contained very low frequencies of this form. By Grade 6, he hadslightly higher frequencies of this form in a report about hockey. Heused phrases such as "the other team's goal." Even in Grade 6,however, Ian did not have much description of the items in hisreports. It appeared he was only beginning to produce some of thedepth of descriptive detail needed to competently write a report.Noun Phrase with Prepositional Phrase Use of prepositional phrases to modify the noun phraseincreased in each year in all three students' narrative writing.John had the highest incidence of noun phrases withprepositional phrase of all students, particularly in Grades 5 and 6.Reasons for John's higher incidence of this form are unclear. Itappears that his use of prepositional phrases was almost a stylisticpreference for, unlike most students, he used this form frequentlyand consistently in all of his narratives. Often John had a series ofprepositional phrases in one T-unit. He wrote this phrase in a Grade5 narrative, "Suddenly, there was a space at the front of the V." Ineach year John's use of prepositional phrases increased. Unlike moststudents he also began his phrases with a variety of prepositionssuch as "with," "about," "in," "around" and "through," particularly in192his Grade 6 narratives.Ian also used prepositional phrases very consistently in hisnarratives. He had slightly lower frequencies of this form than Johnbut his use increased in each year.^His highest frequencies of thisform were in each of his Grade 6 narratives. Most of his prepositionalphrases were introduced by the preposition "of." He wrote thesephrases in Grade 6: "hundreds of empty barrels of gas," "the city ofCape Sunset," and "little pieces of paper."For some reason Tina did not have as many prepositionalphrases in her narratives as most other students. These phraseswere not included, perhaps for stylistic reasons, particularly inGrades 4 and 5. She did, however, have some of these phrases in herGrade 6 narratives. Like John, she had a variety of prepositions thatbegan the phrase. Examples are: "draft from the open window,""brilliant lights like stars" and "some concerns about Archie."All three students had higher numbers of prepositional phrasesto modify the noun phrase in their reports than they did in theirnarratives. While Tina and Ian did not use these phrases in thelatter part of Grade 4 and 5, John used them early in Grade 4.In his Grade 4 reports John had several "of" prepositionalphrases such as "shape of a bubble" and "the canopy of the jet." Useof these phrases added specificity to the attributes of the planes hewas describing. John had high frequencies of prepositional phrasesin every report. Further, the variety of prepositions which beganthese phrases increased in each year. It appeared the type of reportaffected John's frequency of prepositional phrases. For example, heused more of this form in reports that gave instructions. His report,1 9 3"Baseball Card Collecting," had high frequencies of these phraseswhich seemed to reduce ambiguity in his writing. For example heused, "one type of Upper Deck card," and "numbers on the back."Tina began to use prepositional phrases in March of Grade 4.In each of the three years, frequencies of this form increased in herreport writing and she had the highest incidence of this form of anywriter in Grade 6 (6.16 per 100 words). Unlike John, Tina did nothave a variety of prepositional phrases in her reports; in each reportshe used "of" to begin the phrases. Nevertheless, these phrasesserved to add specificity to the description of her topic. Like John,Tina had more of these phrases in reports which providedinstructions. She had many of these phrases in a report whichdescribed how to play volleyball such as: "the object of the game,""one of my favourite sports" and "side of your wrists."Ian did not use prepositional phrases in his reports until Marchof Grade 5. He had few of these phrases in his reports and in everycase he used the preposition "of." In his Grade 5 report on hockey,he wrote, "There are lots of good hockey players." Ian oftenrepeated the phrase "lots of " in his reports. This is a phrase whichstudents of this age commonly use in speech. Again, it seems Ianincluded many phrases from spoken language to assist him in writinghis reports. As stated previously, this is a common strategy thatstudents of this age employ when writing exposition. Frequencies ofprepositional phrases in Ian's Grade 6 reports actually decreased.One reason for the lack of this form in Ian's reports might have beenhis inability or unwillingness to provide the same level of elaborationand detail readily found in Tina's and John's reports. Curiously, it194seems unlikely that Ian did not have the information needed toinclude more descriptive detail in his reports. One discussion withhim quickly revealed the depth of knowledge he has about the gameof hockey. Further, by Grade 6, Ian used prepositional phrasesconsistently in narrative writing. This leads the researcher toquestion why he was unable to provide detail that he apparently haslittle difficulty articulating in conversation and why he did not usesyntax in report writing that he readily used in narrative writing.Perhaps writing reports requires a certain level of cognitivematurity. Ian may not have been ready to handle the cognitiverequirements such as the abstract, more global thinking and planningneeded to organize and present the information in report form.Noun Phrase with Appositive Phrase All three students used very few appositive phrases to extendthe noun phrase in both narrative and report writing. None of thethree students had a noun phrase with an appositive in their Grade 4narratives or reports. Tina and John are the only two of all 14students who used appositives with any frequency in Grades 5 and 6.Low frequencies of appositive phrases do not permit this form to beused as a meaningful measure of syntactic complexity. Yet, the factthat these two very competent writers use this form to some extentmay indicate increasing syntactic complexity in their writing. Use ofappositive phrases seems to enable the students to be moreeconomical in their use of the language. Appositive phrases permitthe writer to reduce an entire clause into a simple phrase. Forexample, Tina reduced the words to describe Trevor and Melanie as1 9 5being newly married to, "Newlyweds, Trevor and Melanie . . . . "Previous research in the field (Hunt, 1970) has indicated a moreeconomical use of writing is more indicative of skilled adult writing.Relative Clauses Use of relative clauses in all three students' narrative andreport writing was limited and sporadic over the three years.Generally, use of these clauses is too limited for individual or groupcomparison. It is important to note, however, that the students usedmore relative clauses in Grade 6 than in previous years. Further,examination of the use of these clauses provides some insights intothe initiation of these more complex linguistic elements.Tina was the only student who had Stage I relative clauses(finite modifying object) in her Grades 4, 5, and 6 narratives. Use ofthis form in Tina's writing was consistent but frequencies remainedquite low. Examples of her Grade 4 and 6 sentences that include thisclause are:"I was "the gardener that built this garden.""They were waiting worriedly for their agent to show them theboy that suited Trevor and Melanies' description."It was uncommon for most of these students to use a finite relativeclause to modify the object of their sentences. Use of the Stage Irelative clause seems to be a more formal construction as it includesa marker of subordination and a finite verb phrase unlike the StageII relative clause which does not have a marker of subordination anduses a non-finite clause. Use of the full subordinate clause seemed tosuit Tina's more formal style of writing. While Tina had many196incidences of these clauses in her narratives, her use of them wasnegligible in report writing. She had more incidences of the non-finite Stage II relative clause than Stage I relatives in her reports.Ian had incidences of Stage I relative clauses but used themrarely in his narrative writing in Grade 4 and 5. He also had limited,although increased, frequencies of these clauses in his Grade 6narratives. His slightly increased use of these clauses in Grade 6demonstrated his increasing ability to use more complex syntax.These clauses were used in sentences which provided descriptivedetail. Examples of his relative clauses are: "their planes that wereparked behind the mine" and "back up to the top where I fell in."Ian did not use Stage I relative clauses in his reports until Grade 6.At this time he had small numbers of these phrases in his reportabout playing hockey. An example of one of his relative clauses is:"There are twelve people that go on the ice at once."John had sporadic occurrences of Stage I relative clauses in hisnarrative writing. He had low frequencies of these clauses in oneGrade 5 and one Grade 6 narrative. Reasons why John did not userelative clauses are not clear. It seems his style of narrative writingdid not suit this more formal construction. However, John did havesmall numbers of Stage I relatives in his reports. He used this formin one report in each of Grades 4, 5 and 6. Fluctuation in his use ofthese clauses was apparent. It may be that the type of report affectsfrequencies of these clauses as John had higher incidences of Stage Irelatives in descriptive reports. He did not include these clauses inreports which gave instructions.Frequencies of Stage II relative clauses (non-finite modifying197object) are also too low for meaningful comparisons among thestudents. Yet, it is interesting that both Tina and John began to useStage II clauses in Grade 4. Generally, they used clauses which beginwith the present participle. John, for example, wrote:"Then he glanced over his shoulder and saw a police carpursuing him."In subsequent years, both John and Tina continued to use relativeclauses of this type. By Grade 6, Tina also had occurrences of StageII clauses that do not start with a present participle. She used theinfinitive form of the clause in this sentence:"Melanie worried about Archie's ability to almost read minds."Ian had a minimal number of Stage II clauses in a Grade 6 narrative.He also used the present participle form of the clause.Tina was the only student who used Stage II relative clauseswith any consistency in report writing. She began to use this form inthe latter part of Grade 5 and 6. Her highest frequencies of this formwere in her reports which provided instructions. Clauses whichbegin with the present and past participle were found in thesereports. Examples of these clauses are: ". . . . the overhand servedriving over the net" and ". . . . rectangular cards constructed out ofcardboard."Ian only had Stage II clauses in one Grade 4 report. It is verycurious that he used these clauses only at that time. He wasdescribing a ride at Disneyland when he wrote:"You go in a boat and you see pirates chasing people andlighting towns on fire."As mentioned previously, his use of this form may be the result of1 9 8using speech-like language in his early reports. There were noincidences of Stage II relative clauses in John's report writing.Relative III clauses (which are introduced by whom, whose ora preposition plus a relative pronoun) are virtually nonexistent inthese students' writing. This is a very formal written construction.Surprisingly, Tina has a form of one of these complex clauses in herGrade 6 narrative. She wrote:"Archie was without guardians whom [sic] had abandonedArchie at the orphanage mysteriously without notice, since hewas only an infant."The use of a relative IV clause (non-finite modifying subject) isextremely rare. John had one such clause in each of a narrative anda report in Grade 5 and Tina used this form in a report and narrativein Grade 6. The fact that these two students used these clauses at allsuggests they are within their linguistic capability; yet the studentswere just beginning to use them in their work. Previous work in thefield indicates that use of non- finite relative clauses to modify thesubject is more prevalent in adolescent and skilled adult writing(Hunt, 1970; O'Donnell et al., 1967; Perera, 1984). It would beinteresting to see if these students had increased use of these clausesin subsequent years. It might also be useful to study whetherstudents might be taught to recognize mature forms of syntax suchas the use of relative clauses within their writing and whether theycould be encouraged to use more. Also, a question which arises fromthis analysis is: How often does a mature form of syntax need to beused before it becomes a consistent element in a student's writing?199Subject and Complement ClausesSubject nominals were virtually non-existent in the writing ofthe three students. John is the only one to use a subject nominal inhis narrative writing and he and Tina used subject nominals in onlyone report in each of the three years. This form of syntax isapparently not yet a consistent part of their working repertoire oflinguistic structures.Similarly, use of complement clauses is also rare in the writingof students this age. While other subjects in the study usedcomplement clauses, John, Ian and Tina did not. Perhaps use ofcomplement clauses, in addition to being a very mature form ofsyntax, is a stylistic preference of some writers.The Verb Phrase - Modal AuxiliariesIn Grade 4, all three students had fairly consistent frequenciesof modal auxiliaries in their narrative writing. Ian had a highincidence of modals in his narrative writing, particularly when heused dialogue. He used modals several times to construct futuretense; for example, "I will take," and "I will do." These studentsused the modals "could" and "would" predominantly in theirnarratives.^They often used these modals when they wrote in firstperson.^Further, higher frequencies of modals were often due to theuse of the contractions "couldn't" and "wouldn't." All three studentsused contractions in early narratives. Frequencies of modalsdecreased over the years in each student's narratives. Meanpercentage per total verbs indicate use was reduced fromapproximately 7 per cent to just over 2 percent in each students'2 0 0work. Reasons for this reduction are not clear. It seems that thestudents did not use contractions as frequently in Grade 6 which mayaccount for the decreasing use of modals. Also, they tended toreduce dialogue in their stories at this time. This may also havecontributed to the reduced numbers of modal auxiliaries. There islittle consistency or pattern of use between the main andsubordinate clauses with modal auxiliaries in the three students'narratives.John had several modal auxiliaries in his Grade 4 reports.Modals were used in sentences which stated the capability of the F-14. Thus, use of the modal "can" was frequent, particularly in themain clause. An example of this use is:"The F-14 Tomcat has awesome menorbility [sic] and can skimaround in circles while turning."John had the highest incidence of modals of any student in Grade 5.He continued to use the modal "can" to describe Indy 500 cars. Hewrote several sentences which began with "they can." Reports whichare written in second person also seem to have high incidences ofthis form. Generally, John wrote "you can . . . . [do this or that]." Johnhad substantially fewer modals in his Grade 6 reports. Meanpercentages of total verb phrases indicate that from Grade 5 to 6, hisuse of modals decreased from 41 per cent to 4 percent. At this timehe changed from using the tentative description (you can do . . . .) ofhow to do something to a less tentative description (you do ^ ) .This change in intent in the writing appeared to result in reducednumbers of modals.Ian did not use modals in his report writing until Grade 6. At201that time his use of modals increased from 0 to 17 per cent of histotal verb phrases. Ian used the modal "can" exclusively. Anexample of his use of modals is:"In basketball you can only have 5 guys on per team."Again, this sentence is indicative of the informal speech-like style ofIan's reports, as illustrated by his use of second person and the word"guys."Tina rarely used modals in her report writing. Her highestfrequency of modals (which was 16 percent of her total verbphrases) occurred in a Grade 6 report. In this report, she used theobligatory modal "must" extensively. Using this modal appears to bea stylistic preference as Tina's style was not tentative (as John's andIan's) but forthright about what was needed to play volleyball.All three students used modals primarily in the main clause inreport writing. This may be due to their style of writing in first andsecond person which seems to increase frequencies of the modals'can' and 'should' in the main clause.Do AuxiliaryNone of the students use the "do" auxiliary with any frequency.Results, therefore, are not provided.Catenative Verb Phrase Use of the catenative verb form in the three students' narrativewriting was quite inconsistent, especially in Tina and John's work.There do not seem to be any patterns among main and subordinate202clause usage.Of the three students, Ian uses catenative verb phrases mostfrequently. Ian's higher incidence of catenative verb phrases in hisGrade 4 narrative are due to repeating a phrase beginning with "got."Examples of this phrase are evident in this sentence:"He almost got killed because he got crushed."Ian continued to use catenative verb phrases extensively in Grades 5and 6. His use of this form tended to change, however, from the lesssophisticated, two word phrases using "got" to three word phrases.Several examples from his Grade 6 narrative include, "decided towalk," "want to kill," and "decided to split up."Generally, Tina and John had fewer numbers of catenativeverb phrases than Ian. Further, use of catenatives decreased in theirwriting over time. John had one high incidence of this form in aGrade 4 narrative due to modelling a phrase but generally had smallnumbers of catenative verb phrases in all of his stories. In Grade 6,he used three and four word phrases such as "was about to start,"and "started to stop." Tina had low frequencies of catenative verbphrases in her narrative writing in each grade. She used three andfour word phrases in her Grade 4 narratives; for example, "wouldlike to take" and "had to go." Tina had wider variety of catenativeverb phrases in her Grade 6 narratives than did the other students.She included less common forms of three word catenatives such as"seemed to open" and "try to worry."Generally over time the students change the forms ofcatenative verbs and reduce usage of them in their narrative writing.2 0 3Reasons for the reduction appear to be that students begin toincorporate a wider variety of verb forms into their writing (asindicated by Table 15). Also, they tended to eliminate two wordphrases beginning with "got" and "started" such as "got killed" and"started going" which may have reduced overall frequencies of thisform.The three students had lower frequencies of catenative verbphrases in their reports than in their narratives. John had a smallnumber of catenative verb phrases in his Grade 5 reports. He oftenused three word catenatives such as "they want to make" or "youneed to buy." Frequencies of catenative verbs increased onlymarginally in his Grade 6 reports. Ian also used this formsporadically in Grades 5 and 6. Often he repeated the same phrasewhich increased the frequency of this form in certain reports.Examples of these repetitive phrases are "try to get" and "try toscore."^(Both John and Ian increased their proportion of catenativeverbs in Grade 6 to 11 per cent of their total verb phrases.) Tina hadone low incidence of catenative verb phrases in a report in Grade 6.Generally, Tina did not use catenative verbs in report writing.Simple Verb Phrase Use of simple verb phrases dominated the three students'narrative and report writing.^Numbers per 100 words of this formand its use in the main and subordinate clause were extremelyvaried among the students; patterns are not discernible. Thus, inorder to provide a more meaningful comparison of the three writers'use of simple verb phrases, mean percentages of total verb phrases204are discussed.Group means of percentages of simple verb phrases indicatelittle change in the proportion of simple verb phrases in narrativewriting over the three years. John and Tina showed a slightreduction in their use of simples verbs from Grade 4 to Grade 6. Ian,on the other hand, generally had an increased use of simple verbs inhis narrative writing, from 60 to 84.5 to 83.5 percent. Reasons forIan's increase in use of simple verb phrases are not clear. It seems,however that they occur in inverse proportion to his use ofcatenative verb phrases which decreased in Grades 5 and 6. Johnhad reductions in simple verb phrases and slight increases inpercentages of passive verbs (from 5 to 13 percent), perfect verbs(from 1.5 to 3 percent) and the "do" auxiliary (from zero to 3.5percent). Similarly, Tina's use of simple verb phrases was reducedas she had increased proportions of passives (from 2 to 5 percent),more progressives (from 1.5 to 6 percent) and slightly morecatenative verbs (from 5.5 to 6.5 percent).Thus, even though grade means indicate little change in the useof simple verb phrases, these students did vary their rate of use ofthis form over time in their narrative writing. It seems John andTina, who are considered more capable writers, had reductions insimple verb phrases as percentages of other forms of the verbincreased whereas Ian did not. Variation in the use of the verbphrase may make an important contribution to the complexity oftheir narrative writing.Mean percentages of verb phrase of report writing indicatemuch more noticeable changes in the proportion of verb phrases2 0 5used over the three years. All three students had reducedproportions of simple verb phrases in their report writing. John'suse of simple verbs decreased from 84.5 to 47 (this large reductionwas due to an increase in modals of 41 percent in Grade 5) to 77.5percent. At the same time his use of catenative and progressiveverbs increased (from zero to 10 percent and zero to 2.5 percentrespectively). Ian also had reduced percentages of simple verbphrases from 84 to 79.5 to 66.5 percent while other verb formsincreased: modals (from zero to 17.5 percent) and catenative verbforms (from 4.5 to 11 percent). Tina had the most substantialreduction in simple verb phrases (from 100 to 98 to 52 percent).Correspondingly, use of other forms of the verb increased in Tina'swriting: passives (from zero to 23 percent), modals (from 2 to 16percent) and perfects (from zero to 3 percent). As they maturedthese students had a wider and more complex variety of verbphrases in their reports. Further, they had higher percentages ofmore complex verb forms in report writing than in narrative writing.Thus, it would seem that report writing elicits more varied andcomplex use of the verb phrase.Progressive Verb Phrase Each student used progressives in narrative writing. Therewere no discernable differences in usage between the main andsubordinate clauses. Tina had small numbers of progressives in hernarratives. Numbers were consistent and increased slightly in eachyear. Typically, Tina used the simple past progressive in hernarratives. Examples of these are: "was going," "were talking," and206" was testing."^John also had small frequencies of progressive verbphrases in his Grade 4 and 5 narratives but usage was quiteirregular. In the latter part of Grade 4, John's use of these verbsincreased as he modelled a phrase from a story. Generally, however,frequencies decreased over time and by Grade 6, John rarely usedthis form. Ian also had irregular frequencies of progressive verbs inhis narrative writing. He had particularly high frequencies of thisform in Grade 4. This is due, in part, to his use of dialogue. Thefollowing is an example of one his dialogues which has progressiveverbs:"So I'm going to die with it""ARE YOU CRAZY?' screamed the soldier."All right I am going."Ian also had reduced frequencies of progressive verbs in each year.Use of the progressive verb phrase is very rare in all three students'report writing. Use of this form was non-existent in Tina's reportwriting and negligible in Ian's writing. John had minimal frequenciesof progressives in a Grade 6 report which gave instructions inpainting with latex. An example of his use of progressives is, "Whenyou are brushing you need to . . . ." It seems that use ofprogressives is not common in report writing, probably because it isoften used to refer to action currently happening in the writing.Logically, past or present action verbs which are used to depictaction scenes would seem more suited to narrative writing.Passive Verb PhraseUnlike most students of her age, Tina used passive verb207phrases as early as Grade 4 and quite consistently in her narrativewriting. Generally, she used passives in the main clause in hernarratives. The passive form of the verb is used in her stories tocombine events currently happening with events that havehappened. This type of construction is exemplified in a Grade 5narrative:"'Ouch,' I yelped as the factory worker twisted on my head.Today was Friday, the busiest day of the week and everyonewas bustling about finishing their assigned doll at the factory.After I was put together and finished I was set in a shiningwindow case where I saw my sparkling reflection."Few students are capable of making this transition from active topassive voice at an early age. Tina occasionally had a passive verb ina subordinate clause which was also not common in these students'writing. An example of this use is ". . . . Tanya sighed with relief nowthat the Phantom was arrested."^Tina had difficulty with this rathercomplex sentence construction, however, as she failed to use thecorrect form. It may be that she did not yet have the capability touse the correct verb which should have been "had been arrested."Nevertheless, the use of this construction in the subordinate clause isnotable as few students used passives in the subordinate clause.Tina also had the highest frequencies of passive verbs in herreport writing of any of the 14 students but not until Grade 6. Herreport on volleyball was written in the passive voice, a formal,impersonal style uncommon at this age. (Generally, these studentsused the active voice such as first or second person in their reports.)Examples of the many sentences in Tina's report which containpassives verbs are:208"Only three hits are allowed before getting the ball over thenet. A point can be scored by the serving team otherwise it isa side out meaning the other team gains the servingadvantage."John rarely used the passive verb phrase in his narrative orreport writing. He only had one high incidence of passive verbs in aGrade 6 narrative.Curiously, Ian uses several passive verbs in a story written inGrade 4, indicating the passive verb is within his repertoire oflinguistic structures. 15 Yet, over time, he has lower frequencies ofthis form in both narrative and report writing. When and whystudents begin to incorporate mature forms of syntax regularly intotheir writing is puzzling. Perhaps over time both Ian and John willbegin to use passives on a more regular basis as they adopt a moreformal style of writing.Perfect Verb PhraseUse of the perfect verb phrase highlights Tina's capability as awriter. There are examples of the more complex perfect verbphrases in her narratives in each of the three years. She usedseveral perfect verb phrases in November of Grade 4. At that time,however, she did not always use the form correctly. In this exampleshe may have omitted one necessary perfect verb:"One day your father went on a business trip and that day thegardener found your mother [had] dropped dead, she had15^Ian's use of the verb phrase using "got" as in "got killed" was notincluded in the frequency counts of passive verbs. Had it been countedhe would have had higher frequencies of passives in his Grade 4narratives.2 0 9fallen off a tree."Tina continued to use this form in Grade 6 although her frequenciesare low compared to other forms of the verb. This time, however, noerrors in her use of perfect verbs are apparent. Tina's use of perfectverb phrases in report writing is negligible.Both John and Ian rarely used perfect verb phrases. John onlyused a slight number of these phrases in his Grade 6 reports and Ianhad a small number in a Grade 4 and 6 narrative and a Grade 4report.Text Forms in WritingA comparison of the students use of text forms is also provided.The comparison points out the variations in their use of beginningand ending statements and overall length in their narrative andreport writing over the three years.Beginning and Ending Statements Categorizing the three students' beginning and endingstatements indicated they had differences in the way they openedand closed both their narratives and reports.Narrative Beginnings.^Almost all the students' openingstatements in their narratives were characteristic of Langer's (1986)most complex category, structural beginning statements. Thestudents' beginning statements fit into several different sub-categories. Tina and John had primarily 'beginning action,''action/dialogue' and the occasional 'setting' statement in their210narratives while Ian had 'setting' statements and rarely an 'action/dialogue' opening. Over the years, the types of beginningstatements in the students' narrative writing showed little change.Yet, while the types of beginnings remained constant, the content ofthe statements often changed although classifying the statementsinto sub-categories did not reflect the changes. Ian's settingstatements, for example, had similar content in Grades 4 and 5. Thenin Grade 6, the content appeared to change as he added many moredetails to his beginning statements. Notice the difference in his styleof writing over the three years:"Once in a garbage dump this man glued beer cans and winebottles and jumped in someone's garbage can." (Grade 4)"Once on Hallowe'en two kids, Peter and Ian, went out trick ortreating and they saw a haunted house." (Grade 5)"It was a cold night in the dark cave just off the coast ofIceland where John Little was about to get his head cut of bythe evil Sheriff of Nottingham until Robin Hood came along andpulled out his bow and arrow and killed all the soldiers."(Grade 6)Perhaps Ian could be criticized for too many details in his Grade 6opening statement but the increased attention to detail andintroduction of mood to his narrative is apparent. In each year hisstatements became more specific: from "this man," to "two boys,Peter and John," to "John Little the evil Sheriff and Robin Hood."Further, Ian's statement contained implicit detail. As an example, weknow from Ian's last sentence that Robin protects John Little. Ianalso had progressively more modifiers in his opening sentences, suchas "cold night, dark cave," and "the evil Sheriff of Nottingham."2 1 1Whereas Ian's narratives contained only 'setting' statements ineach grade, John had both 'setting' and 'beginning action' statements.As he matured, his opening 'setting' statements changed to providemuch more detail and intrigue for the reader. Here are two 'setting'statements from John's Grade 4 and 6 narratives:"One day a man thought I'll play a little game with the policetoday." (Grade 4)"On March 26, 1977, in the city of Manaus in the AmazonJungle in Brazil, the famous explorer, Indiana Jones wasabout to start on an expedition to look for the lost temple ofdoom and its solid gold statue in the mighty Amazon Jungle."(Grade 6)In fact, John adds so much detail in his Grade 6 example that heseems to lose his train of thought as he repeats information about theAmazon Jungle. While this sentence is not perfect, it clearlydemonstrates John's increasing ability to draw the reader into hisstory by describing the situation in great detail. John also hasseveral 'beginning action' and 'action/dialogue' statements whichengage the reader. He wrote:"I woke up early and couldn't go back to sleep." (Grade 4)"I am Goose." (Grade 5)" Ahh," down falls another foe!" (Grade 5)"Bang, and they're off!" (Grade 6)In each case, these statements motivate the reader to read on asoften they provide implicit narrative information. In John's Grade 5opening, for example, the reader is aware that someone has theadvantage in a fight or battle. Clearly, the structural beginnings that212John used represent a series of competent ways to open a narrative.Tina also had variety of structural beginning statements in hernarratives. She rarely began with a 'setting statement' but used'action/dialogue' openings most often even in Grade 4. Examples ofher beginning statements are:"'Wow,' cried Claire." (Grade 4)"'Ouch,' I yelled as the factory worker started to twist on myhead." (Grade 5)"'Ahh!' wailed the child tugging at the sleeve of the tall, primlady." (Grade 6)Tina's 'action/dialogue' statements became more complex by Grade 5and 6 as they included intriguing information. They seem muchmore sophisticated than this Grade 4 statement for there is implicitinformation in the sentence which might engage a reader. Forexample, the reader might be intrigued to find out why a factoryworker would twist someone's head. Tina's few 'setting' statementscontain the same intrigue as exemplified by this Grade 5 openingsentence:"Kimberly Gibson hustled down the street, watching the leavesdance around the thick trunks of trees."This sentence has a great deal of subtle information. The readerknows, for example, that it is autumn, it is windy, that Kimberly is ina hurry and that she lives somewhere where there are large,deciduous trees. Few writers in Grade 5 could write such a complex'setting' statement.All three students demonstrated that they are capable ofbeginning a narrative competently. Further, Tina and John began in213a variety of ways and had sentences which provided much interestfor the reader of their stories. All three students show growth andincreasing maturity in the way they added both explicit and implicitdetail in their narrative beginnings. It seems, however, that growthof this nature is not accommodated by Langer's (1986) classificationsystem.Report Beginnings.^By contrast, report beginnings show littleof this change and increasing complexity. Most of the statementsused to open reports are classified as weak beginnings and are'generalized' facts.In all of his reports Ian began with statements which were'generalized' facts. It seemed that Ian depended on the title of thepiece to provide its introduction. According to Langer (1986), this isa common phenomenon in report writing with students of this age.Further, there appeared to be little change in, not only the category,but in the content of his beginning statements. Examples of hisbeginning sentences from three of his reports titled (in order)"Disneyland," "Hockey," and "Hockey" are:"At Disneyland there is a ride called the Pirates of theCaribbean." (Grade 4)"Hockey is a sport that lots of people enjoy." (Grade 5)"In hockey there are twelve people who go on the ice."(Grade 6)In looking at these examples, one might argue that Ian's Grade 5statement might be a 'thesis' statement rather than an opening fact.Unfortunately, Ian does not support this statement with anyinformation that is related to it; therefore, the sentence must be2 1 4viewed as an opening fact. This example illustrates the fact thatclassifying beginning sentences according to specific categories maybe problematic. The rater has to make decisions as to the probableintent of the author and classify the sentence accordingly.Tina opened many of her reports with a general fact in thesame manner as Ian. Like Ian, she seemed dependent upon the titleof the piece to provide sufficient introduction to the topic. Sheapparently lacked strategies to open her reports in other moreappropriate ways. In Grades 4, 5, and early Grade 6 her openingstatements showed little change. The following are examples of heropening sentences from her reports entitled "School," "Ducks," and"Volleyball":"In school you learn." (Grade 4)"Mallard ducks have a glossy green head." (Grade 5)"Volleyball is one of my favourite sports." (Grade 6)In March of Grade 6, however, Tina began one of her reportsdifferently:"In this report I plan to inform you what a professional hockeycard is and the value of hockey cards."This 'topic starter' sentence was her first attempt to provideintroduction to her topic rather than beginning with a fact. It issurprising that a writer of this capability appears to make littleattempt to find other appropriate ways to open her reports. Thereseems to be little transfer of ability from narrative writing to reportwriting. It is highly unlikely that Tina (or any of these students) haddirect instruction as to how to begin a report. Thus, it would beinteresting as a topic for further study to examine the effects of2 1 5direct instruction on students' ability to write beginning sentences.John was the only writer of the 14 who opened some of hisreports with 'structural' opening statements. Some of his reports hadsentences that introduced the topic as well as included an opinion.He then supported that opinion with several statements of fact whichwere related to the topic. Examples from his Grade 4 and 5 reportsdemonstrate this type of opening thesis statement:"The F-14 Flighting Falcon is the ultimate fighting aircraft."(Grade 4)"Indy cars are special race cars." (Grade 5)John did not always open with a structural opening. In Grade 6, hewrote a report which gave instructions. He did not appear to have astrategy for opening this kind of report. He had this 'weak'generalized fact for his opening:"First to start painting you need a paint brush."Overall, opening reports presented much more difficulty thanopening narratives for the students as indicated by the lack ofvariety in the sentences they used. As stated previously, it seemsimportant to address the question of why students, who socompetently begin their narratives, have such difficulty beginningreports.Narrative Endings.^The three students used a variety ofending statements in their narrative writing. In almost every casethe endings are either 'formulaic' or 'structural.'Ian's narratives generally closed with the formulaic ending "theend" early in Grade 4. His ending sentences provided little closureand this phrase was his only signal to the reader that the story was2 1 6finished. Later in Grade 4 and over the three years, Ian's narrativeshad structural endings, particularly 'natural ending' statements, thatis statements which seem to wind down the action of the narrative.Ian brings his stories to a close by adding a very general closing idea.These statements showed little change over time as is shown by thefollowing examples:"I hope tomorrow is a better day than today." (Grade 4)"Two weeks later they were both billionaires." (Grade 5)"Months later John and Robin built a castle right behind Robin'sfather's old one." (Grade 6)John often had structural endings in the form of 'summary'statements that closed his narratives. These ending statementsprovided closure through reference to a key point in the story and itssubsequent change or finish. John includes these two summarystatements:"Orki Sako used to be good but the love of the sword made himevil." (Grade 5)"Later on he married again and finished living a happy andnon-betting life." (Grade 6)Over the years John's 'summary' endings show little change in thecontent or type of sentence. Like many of the students in the study,John often added the formulaic ending "the end" after his summarystatement, almost as a precautionary measure. He continued to addthis conventional ending after his structural closing statements evenin Grade 6. John had several other forms of structural endings in hisnarratives. He wrote this 'natural ending' in one Grade 6 narrative:"Then he got a free trip back to Canada."2 1 7Also he concluded one story with a 'solution resolution' statement.In this case John had to resolve a problem which he had outlined inhis narrative. This sentence is from his Grade 5 narrative:"The police at the other States caught the hunters and gavethem all a super big fine."John used a variety of structural ending statements to close hisnarratives yet there seemed to be some sentences which provided a"better" summation than others. For example, some of his 'solutionresolution' and 'summary' statements provide better closure than his'natural ending' statements. Furthermore, some of his 'naturalending' statements seemed more complex than Ian's 'natural ending'statements. This apparent finding serves to question Langer's (1986)categories used to classify these ending statements. It appears thatthere are levels of complexity within the sub-categories. Perhaps amore refined system of classification needs to be developed.Tina also had a variety of structural ending statements in hernarratives. Many of her stories clearly define a problem that neededto be resolved. Consequently, she used solution resolutionstatements to end at least one narrative in each grade.^Examples ofher endings are:"So the secret of the golden tree remained unknown forgenerations until now." (Grade 4)"Kimberly took care of Molly's dog and often took her collie (toMolly's grave) to leave flowers and reflect on their specialfriendship."^(Grade 5)"Archie Smith was a boy wonder!" (Grade 6)Tina also used summary and natural ending statements to218conclude her narratives.^She used these types of statements in twoGrade 5 narratives. Examples are:"I'm going to name him phantom." (natural ending)"I seemed older my hair was mostly gone my colours faded yetI did seem older wiser and gentler." (summary statement)Examining Tina's ending statements highlights the difficulty ofcategorizing some statements. Many of the students' endings appearto fit in more than one category. The above Grade 5 solutionresolution statement ("Kimberly took care of . . . ."), for example,could be a summary statement or perhaps even a natural ending.These categories tend to be very subjective and there are oftendecisions that have to be inferred as to the author's intent. Thus,caution is needed when interpreting these results.Report Endings.^Over the years all three students appearedto have difficulty ending their reports. None of the students wasable to vary the types of endings they used. In every case theyended with a 'weak,' last fact ending.Ian's report endings show little change over the three years.Every report ended on a last fact. It appeared that most of theseendings simply reported the last bit of knowledge that he was able toproduce. Examples of his ending statements are:"You go down hills and there are people shooting guns andputting people in jail." (Grade 4)"The positions in hockey are left wing, centre, left defense andright defense." (Grade 5)"In basketball you can only have 5 per team." (Grade 6)In each case, Ian simply provided the reader with more information2 1 9but no sense of closure.John's report endings resembled Ian's, featuring simple lastfact sentences. While John seemed to have some strategies forbeginning reports, he had few for ending them. This series ofendings exemplifies the lack of change in his concluding statements:"Their top speed ranges from 160-201 mph." (Grade 4)"Stock cars drive on the Indianapolis 500, too." (Grade 5)"When you're finished, you hang your paint brush in a safeplace." (Grade 6)Only John's Grade 6 sentence hinted at closure as it mentions theword "finish."Tina evidently also lacked strategies for ending her reports.Each of her reports ended with a last fact which often included a list.She wrote:"Some great composers are: J. S. Back, Wolfgang ArmadausMozart, and Beethovan." (Grade 4)"The nests can be near lakes, rivers, marshes, ponds, sloughsand grasslands." (Grade 5)"The volleyball game is played until a score of 15, however, theteam with 15 points must win by 2 points." (Grade 6)By the end of Grade 6, Tina was only one of a few students whoattempted a different kind of report ending. She concluded with thissummary statement:"Throughout this report I have told you about what aprofessional hockey card is."Tina's ending indicates she was beginning to understand the idea ofproviding closure for her topic. This statement seems to represent a2 2 0change in her thinking as she apparently consciously alerted herreader that the report was finished. It is interesting, however, thatshe used the first person. She seems to be relying on an "opendiscourse schemata" (Bereiter & Scardamalia, 1982) or conversationalstyle to conclude her report. Perhaps using this type of ending isalmost an intermediary step for Tina from writing weak 'generalized'endings to using more appropriate structural endings.In summary, these students clearly found ending reports verydifficult. They appeared to lack strategies or perhaps were not evenaware of how to competently end expository writing. Only Tina, inlate Grade 6, purposefully wrote a sentence which was designed toprovide her reader with a sense of closure. By contrast, the threestudents appeared to have little difficulty ending their narratives.Length of Writing Sample Within each year, the students' length of writing samples inboth narrative and report varied substantially. Similarly, there weremarked increases in the lengths of samples that were written in thetwo genres.Narrative.^Ian's narratives were approximately the samelength in Grades 4 and 5 (250 words) varying by only about 50words. In Grade 6, the length of his stories increased, becoming over100 words longer than his previous narratives. Reasons for thisincrease are not clear. Classroom observations, however, suggest thatIan was much more interested in his writing in Grade 6. He tookmore time with his narratives and appeared to find the writing taskless tedious. Further, as stated previously, Ian's Grade 6 narratives2 2 1contained more events and descriptive detail as well as morecomplex syntax, all of which contributed to the length of his writing.John's narratives also had little fluctuation or change in lengthin Grades 4 and 5. His narratives were approximately 250 wordslong in these years. In Grade 6, however, the length of John'snarratives increased. In November, he wrote well over 1000 wordsin a detailed narrative entitled "The Expedition." This increase inlength may be due to the fact that John modelled his story after an'Indiana Jones' movie. His story was filled with descriptive detaildepicting exciting events and the syntax, as indicated in previousdiscussion, was more complex. John's recent viewing experienceapparently enabled him to generate a great deal of content as herecreated some of the movie events and intertwined them with hisown story line. Also, classroom observations indicated John enjoyedwriting this story as evidenced by his animated discussion of thetopic with his classmates and increased attentiveness to his writing.By contrast, John wrote a story in March of the same year which wasonly one third as long. This story did not appear to interest him andhe was not well motivated in the writing sessions as notedpreviously.Tina's narratives increased in length in each year. Further, thelengths of her narratives varied between November and March ineach year. In Grades 4 and 5 the length of her narratives variedbetween 250 and 1200 words. One of the longest stories Tina wrotealso seemed to be modelled after a recent experience, in this caseafter an idea from a book. In Grade 4 she wrote a narrative about asecret garden. She changed the events in the story but it closely2 2 2resembled the novel The Secret Garden. This book seems to haveinspired Tina to write her own mini-novel. Tina's Grade 5 narrative,which was also very lengthy (over 1500 words) is a very sad storyabout a young girl who dies of leukemia. It is difficult to saywhether this story had been modelled after an existing novel, but itclearly is a story which deeply involved its writer. Tina includedmuch information about her protagonist's feelings and theexperiences she had as she lost her closest friend, the effect of whichcontributed to the length of the narrative.As stated previously, Tina's stories have much descriptivedetail which enables her reader to have a clearer idea of the eventsin the piece. Further, it seems Tina understands the requirements ofgood story writing. She creates a problem, builds suspense to aclimax, and provides a resolution to her stories. The use of all thesetechniques added length to her writing.Overall, these students continued to increase the length of theirnarratives over the years. All of them apparently wrote longernarratives when they drew upon their recent experiences, especiallythose which involved powerful media such as exciting movies orexcellent novels. This finding underlies the importance of allowingstudents to choose their own topics. Choice of topic seems toempower most writers as it enables them to build upon their recentexperiences and background knowledge.Report.^In each of the three years the students' reports areconsiderably shorter than their narratives, yet the lengths of theirreports increased over the years. Compared to narrative writing,there were fewer fluctuations in lengths among the students' reports.2 2 3Ian's reports are very short in both Grades 4 and 5; he wroteunder 100 words in each.^Even though the topics Ian chose clearlyinterested him, he seemed unable or unwilling to write more than afew sentences. Classroom observations indicated that Ian not onlystruggled to begin his report, but had much difficulty sustaininginterest in his writing. He did not edit or redraft these reports inspite of having been asked to follow the same procedures as innarrative writing. Ian's attitude toward his writing at that time wasquite indifferent, perhaps because he found the task difficult. ByGrade 6, the length of Ian's reports tripled. He appeared to put muchmore effort into his work but continued to experience some difficultyas evidenced by the type of information in his reports. Part of hisdifficulty may have been due to the topic about which he chose towrite. While Ian knows a great deal about hockey and basketball,for example, the two topics are extremely broad. It appears Ian'sdifficulty might have resulted from trying to find a way to organizeand pare down the vast quantity of information.^Traditionally,written exposition (in this case, report writing) does not have thetime-related organizational pattern of narration. 16 Harris (1986)notes that when writing is not time-related some overall plan isrequired. Further, he states, "that the non-time-related categories[such as exposition] will be inherently more difficult in terms oforganization for the young writer" (p. 52). Perhaps all Ian's thinkingwas taken up with the organizational task of report writing whichleft him little time for including a great deal of information.16^Unless exposition takes the form of 'reporting' such as reporting a newsevent which has events occurring in a time-related sequence.224John's reports are among the longest of the students in Grade 4.He did not appear to struggle with report writing as did most of theother students. Generally, the reports written about a specific planeor car were his longest.^Often John had read a great deal about thetopic he chose for his reports. Thus, he was often writing fromrecent experience and seemed to have many details stored in hismemory. Further, John's choice of a specific topic (or perhaps thecoincidence of reading about a more specific topic) seemed to enablehim to remember very precise details and to present them in alogical sequence in his report. Perhaps the narrower topic was lessdifficult to organize which afforded John more time for attending tothe content of his reports which in turn, increased their length.John's reports increased marginally in length in each of Grades 5 and6. One of his longest reports explained how to paint with latex. Thisreport was very well organized and presented a complex set ofinstructions. It may be that this report was longer because the topichad an inherent sequence of steps to be explained. Perhaps, becausethe topic was organized, John was freer to focus on including thenecessary information.Tina's reports are very short in Grade 4. Observations indicateshe was extremely frustrated in her early attempts at report writing.She struggled to find a topic and had difficulty getting started. Herfirst and second reports are under 100 words long. In Grade 5, Tinaapparently began to feel more confident with report writing. Alongwith the other students in the study, she likely benefitted frompractise in writing reports which increased her confidence andenabled her to write longer reports. The lengths of her reports2 2 5increased to over 100 words as she wrote about "Ducks" and "ThePiano." Tina's longest reports were in Grade 6. Her report onVolleyball was just under 400 words. Perhaps her topic, playingvolleyball, also provided an organizing framework for the report,that is, the sequence of playing the game. This sequence may haveallowed her to concentrate on articulating the complex instructions ofthe game rather than on how she would organize the information.Thus, the report became longer as she discussed how to play thegame from the beginning to the end.All students clearly write less in their reports than in theirnarratives. The students seemed to have less interest and lessconfidence in writing reports. The students who are able to writelonger reports appear to choose topics wisely. Students, for example,may have chosen a topic which was very broad and provided littleorganizing frame for presenting the information. Consequently,much of the students' thinking may have been consumed with tryingto begin and organize a report, thus leaving less time for generatingcontent. Another reason students' wrote shorter reports may havebeen due to the type of report. Perhaps the sequential nature ofproviding instructions allowed the students more time to concentrateon providing more detail. It seems important to acknowledge thatthere are many types of report writing and that some may be morechallenging to write than others. Perhaps report or expositorywriting should be further subcategorized to allow for the apparentlydiffering demands that different types of writing place upon thewriters.226CHAPTER VISUMMARY AND DISCUSSIONThe purpose of this study was to examine the use of syntax andtext form in the narrative and report writing of a select group ofstudents over a three year period. Results indicate there areobservable differences in both the use of syntax and text form in thestudents' writing in the two genres. The following is a summary ofthe results and factors which may have affected them, in addition toa discussion of the findings including pedagogical implications andsuggestions for interesting further research.General Findings It might be argued that the descriptive analysis provided inthis study lacks the needed statistical tests to adequately compareresults among the writers or with previous works in the field.However, there are several reasons why statistical tests were notperformed on the data. The design of the study uses a descriptivecase study approach with a relatively small number of participants.It is intended to provide information about children's growing use ofsyntax and text forms in narrative and report writing. Due to thesmall sample size, which is not representative of a normaldistribution and types of data obtained, parametric statisticalanalysis is not recommended (Borg & Gall, 1989; McMillan &Schumacher, 1989; Wiersma, 1986). Further, there are severalreasons why the data do not permit the application of non-227parametric statistical tests for significant differences in results.Reasons for the exclusion of statistical treatment of the data arediscussed in the following paragraphs.Many of the elements of syntax that are used to describesyntactic complexity are mature forms of syntax according toprevious research. This research has often been conducted witholder children and skilled adults. Predictably, these forms are usedinconsistently or sporadically by these younger writers. Further, theother less complex forms of syntax are also used inconsistently; thatis, the students vary their usage according to a number of factors.The infrequent or inconsistent use of most of the elements of syntaxoften created skewed data which had outliers of very high, or oftennull frequencies. This type of data precludes the use of measures ofcentral tendency such as means. As many^parametric statisticaltests require calculations based on means, these measures did notseem appropriate to apply to the data. Therefore statisticaldifferences in results are not provided. Additionally, data such asmean T-unit length and overall length of writing sample have verylarge ranges in the writing of individual students within and acrossthe years. The analyses suggest that measures of central tendencysuch as means are not a reliable method of describing this unevenlydistributed data.While comparisons are made between results in existingstudies and among individual writers, it is with the understandingthat the differences have not been demonstrated to be statisticallysignificant. As such, the results are considered to be observationsintended to provide a springboard for discussion and speculation as2 2 8to explanations for observed behaviours and indications ofproductive areas for further study. Further, as the study wasconducted over three years with the same subjects by aknowledgeable and skilled practitioner, observations and datacollection are thought to be consistent with expectations inqualitative research design. Merriam (1985) in her review of casestudy and qualitative research suggests there are methods ofimproving the credibility of case study findings. These methodsinclude prolonged data-gathering on sites and triangulation (using avariety of data sources). Several types of data were gathered overan extended period of time in this study (use of syntax, use of textform, discussions with teachers and classroom observations). Theyprovide triangulation thereby increasing the study's credibility(McMillan & Schumacher, 1989; Merriam, 1985). It is with theseconsiderations in mind that the following summary and discussionare presented.Syntactic ComplexityThe following is a synopsis of the use of syntactic complexity inthe writing of the 14 subjects over the three year period. Thewritten data were analyzed on the basis of the calculation of mean T-unit length and frequencies per 100 words of forms of extendednoun phrases, including nominal clauses and types of verb phrases.The syntactic analyses provided data for the purposes of comparingthe students' use of syntax in their narrative and report writing. Asummary of the findings follows:229T-unit Length In general, grade means for T-unit length show increased meanT-unit length in successive grades, particularly in report writing. Inexamining individual student writing, however, several additionalfindings are apparent:1. Grade means for T-unit length do not reflect the substantialfluctuation in T-unit length of the students' narrative and reportwriting within and across the years, nor do they reveal the wideranges in T-unit lengths or the individual capabilities of thesestudents in each year.2. While grade means do indicate much higher T-unit length inreport writing over the three years, it is important to note that inevery year except Grade 4, almost every student had longer T-unitsin report writing than in narrative writing.3. Measures of mean T-unit length appear to obscuredeveloping features of language such as changes in the content of thelanguage and elements of syntax.Extended Noun Phrases Several forms of extended noun phrases and subordinateclauses were examined to determine differences in the use of syntaxin the two genres: frequencies per 100 words of noun phrases thatwere extended with either modifiers, prepositional phrases,appositive phrases or relative clauses and subject and complementnominal clauses. Results show extended noun phrases occur withincreasing frequency in students' writing as they grow older.230Further, there are differences in the use of these phrases in narrativeand report as indicated by the following findings:Noun Phrases with Modifiers.1. Students' use of this form was somewhat inconsistentbetween November and March, yet in each successive year, morestudents used modifiers to extend the noun phrase in both narrativeand report writing.2. While all students had occurrences of noun phrasesextended with modifiers in their narratives early in Grade 4, moststudents did not use this form in their reports until Grade 5 or 6.3. The students' rate of use of this form remained fairlyconstant in narrative whereas it increased rapidly in report; byGrade 6 students had four times as many modifiers in their reportsas in their narratives.Noun Phrase with Prepositional Phrase. 1. Students' use of prepositional phrases was early-forming.All students used this form in Grade 4 in both narrative and reportwriting. Usage is somewhat inconsistent between November andMarch, yet in each successive year, use of this form increased in bothgenres.2. The students' rate of use of prepositional phrases was two toseven times higher in report writing than in narrative writing ineach grade.3. The types of prepositional phrases that some students usedchanged over time. Many students had a wider range of prepositionsto begin these phrases.231Noun Phrase with Appositive Phrase. 1. Use of appositive phrases is inconsistent and not as frequentas other forms of modification (it appears not to be a reliable meansof describing syntactic complexity), yet the number of students whoused this form increased in each year in both narrative and report.2. Students used appositive phrases earlier in narrativewriting than in report writing; most students had occurrences of thisform in their narratives in Grade 5 but not until the latter part ofGrade 6 in their reports.Noun Phrase with Relative Clause. 1. The use of these clauses is infrequent and therefore notconsidered a reliable means of describing syntactic complexity.However, the fact that students used these forms suggests they arewithin their linguistic capability. It is therefore interesting tocompare the students' initial uses of these complex clauses.2. Use of the least complex Stage I relative clause increases ineach year in both narrative and report writing. Students wereinconsistent in their use of these clauses, particularly in Grades 4 and5. Usage became more consistent in Grade 6. More students hadStage I relatives in their narratives than in their reports butnumbers per 100 words are higher in reports.3. Use of Stage II relatives is also inconsistent in Grades 4 and5, yet more students used these clauses in successive years. Twice asmany students used these clauses in their narratives than in theirreports but in Grade 6, numbers per 100 words are two to four timeshigher in report writing than in narrative writing.2324. Only two students in all three years had an example of thecomplex Stage III relative clause in their narrative writing.5. Use of the Stage IV relative clause is infrequent andsporadic within and across the years but the number of students whoused this form increased each year in both narrative and reportwriting. Although frequencies were low, twice as many studentsused these relative clauses in their narratives as in their reports.Subject and Complement Clauses. 1. Subject and complement nominal clauses are usedinconsistently and infrequently and therefore results areinconclusive. Nonetheless, more students used them in both genresin successive years.2. Subject nominal clauses are used by more students andmore often in report than in narrative.3. In Grades 4 and 5, twice as many students used complementclauses in narrative than in report writing but by Grade 6, equalnumbers of students used these clauses in both genres. Frequencies,while very small, are also consistently higher in report writing inGrade 6.The Verb Phrase Several types of finite verb phrases were examined in thestudy: use of modal auxiliaries, do auxiliary, catenative, simple,progressive, passive, perfect, perfect progressive and perfect passive.In general, the verb phrase becomes more complicated in successivegrades; several changes are apparent in the types and quantity of233verb phrases used in narrative and report writing over the threeyears.Modal Auxiliaries. 1. Students had consistent frequencies of modals in theirnarrative writing from an early age. Use of modals also appeared tobe consistent in main and subordinate clauses. Numbers per 100words of modals decreased slightly in successive years. Grade meanpercentages of total verb type also indicate students used fewermodals in narrative writing over time, particularly in Grade 6.2. Most students did not have modals in their reports until thelatter part of Grade 5. In each year more students used modals andhad increased rates of this form in report writing. Modals were usedprimarily in the main clause in Grade 4 and 5 reports but were moreevenly distributed in the main and subordinate clause in Grade 6.Frequency rates indicate the students had between two and fourtimes as many modals in their reports as in their narratives eventhough the rates were reduced slightly in Grade 6.3. More students used the "do" auxiliary over time but usagewas limited and sporadic. More students used this form in narrativethan report writing. Use is higher in the main clause in Grades 4 and5 yet became more balanced in Grade 6.Catenative Verb Phrase.1. Catenative verbs were used early and frequently by moststudents in narrative writing. Numbers per 100 words decreasedover the years and were higher in the main clause. Meanpercentages of total verbs show a reduction from 9.5 to 5.3 percentfrom Grades 4 to 6.23 42. Use of catenatives was inconsistent and varied in reportwriting between November and March in most students' writingsamples. Numbers per 100 words were higher in report writing thanin narrative writing but decreased in each year. Frequencies wereup to three times higher in the main clause in Grades 4 and 5 butbecame more balanced between main and subordinate clauses inGrade 6. Mean percentages of total verbs show slight increases from6.1 to 8.1 percent over time, a result of more students having usedcatenatives in report writing in each successive grade.Simple Verb Phrase. 1. Simple verb phrases predominated students' narrative andreport writing.^Variable frequencies and inconsistent use of thesephrases within the year prohibit useful comparisons. Ratios ofsimple verbs used in the main clause to those used in thesubordinate clause, however, reveal differences between narrativeand report writing. Ratios of main to subordinate clause usage innarrative writing remained similar in Grade 4 and 6 and rose slightlyin Grade 5; students used approximately three to five times as manysimple verbs in the main clause as they did in the subordinateclause. Ratios in report writing decreased steadily in each year; asthey grew older, students had a more balanced distribution of simpleverbs in the main and subordinate clauses.2. Mean percentages of type of verb per total verbs showvariable use of simple verbs in narrative and decreasing use inreport over the three years.235Progressive Verb Phrase. 1. Most students used progressive verbs early and consistentlyin their narratives. There are no patterns of usage between mainand subordinate clauses. Numbers of progressive verb phrasesdecreased in students' narrative writing over the years. Grade meanpercentages of type of verb also indicate reduction in use from 7 to3.6 per cent from Grades 4 to 6.2. Use of progressive verbs in report writing was rare.Further, frequencies of this form decreased in students' reports overtime. Grade mean percentages indicate decreases of 5.5 to 2.6percent from Grade 4 to Grade 6.Passive Verb Phrase. 1. Use of passive verbs was infrequent and inconsistent inboth genres (usage is too infrequent for reliable analysis), yet in eachyear more students used this form of the verb.2. Frequencies of passive verbs increased slightly in thestudents' narrative writing in each year. Mean percentages of verbtype show limited increases from 1.8 to 2.9 percent. Students usedpassives predominantly in the main clause.3. Fewer students used these verbs in report writing than innarrative writing and frequencies decreased over time. Further,most students only began to use passives in Grade 5 or Grade 6 anduse was almost exclusively in the main clause.^Numbers per 100words indicate students had slightly higher frequencies of passivesin their Grade 6 reports than in their narratives. Grade mean236percentages of verb type show increases from 2.8 to 7.8 (3.4)percent. 17Perfect Verb Phrase. 1. Use of perfect verb phrases was very inconsistent andsporadic (usage is too infrequent for reliable analysis). Only twostudents used this form with any consistency over the three years.However, the number of students using perfects increased fromGrade 4 to 6. No patterns are discernible between use in the mainand subordinate clauses.2. More students used perfect verbs in their narrative writingthan in their report writing. Numbers per 100 words, however, werehigher in report writing.Perfect Progressive and Perfect Passive Verb Phrases. 1. Both of these complex verb phrases were used very rarelyand inconsistently, thus results are inconclusive. Some students,however, began to use these verbs in Grade 6.^Low frequencies ofthis form were found in their narrative writing in the main clause.Mean Percentages of Total Verb Phrases1. Percentages of type of verb per total number of verbs usedindicate students decreased use of modals, catenatives andprogressives in their narrative writing over the three years whilethey increased use of simple, passive perfect, perfect progressive and17^This percentage rate is slightly skewed by one writer's high incidenceof passive verbs in a report. Without this outlier the rate is 3.4 asindicated in the parentheses.237perfect passives. (The underlying data base is small and thereforeunqualified conclusions cannot be stated.)2. Percentages calculated for report writing show decreaseduse of simple and progressive verb phrases and increased use ofmodals, catenatives and passives. Students also had increased use ofperfects, perfect progressives and perfect passives in their reportsalthough percentages are very low.Text Forms in WritingOn the basis of the writing samples, two forms of structuralanalysis were undertaken: first, the beginning and endingstatements and second, the length of writing samples. Results ofthese analyses reveal substantial differences between the students'narrative and report writing over the three years.Beginning and Ending Statements Results show that, from an early age, students structured theirnarrative and report writing differently. The students used morecomplex beginning and ending statements in their narratives than intheir reports, especially in Grades 4 and 5.Beginning Statements. 1. Students used formulaic and structural opening statementsearly in Grade 4 in their narrative writing. Over time the use offormulaic beginnings decreased as the use of structural beginningsincreased. Grade mean percentages of types of beginnings showstudents have increased structural beginnings from 67 to 92 percent2 3 8from Grades 4 to 6. Weak beginnings were extremely rare innarrative writing.2. Weak beginning statements, usually in the form of a'generalized' opening fact, dominated students' report writing. Meanpercentages show students used 77 and 93 percent weak openings inGrade 4 and 5 respectively and reduced them to 71 percent in Grade6. This reduction in weak beginnings was due to the increase offormulaic beginnings (14 percent) and a marginal increase instructural beginnings (11 percent) in that year.Ending Statements. 1. Formulaic endings decreased while structural endingsincreased in narrative writing over the three year period. Meanpercentages indicate reductions in formulaic endings from 40 to 29percent. The highest percentages of structural endings in narrativewriting are in Grades 5 and 6: 77 and 67 percent respectively. Overhalf of the students used "the end" in addition to their structuralendings in these grades. Weak endings were reduced from 7 to 4percent.2. Most students had weak 'last fact' endings in their reports,particularly in Grades 4 and 5. Grade mean percentages indicate upto 90 percent of endings were weak in these grades. The students'weak endings decreased to 49 percent in Grade 6 while structuralendings increased from 4 to 43 percent. Eight percent of thestudents' closing statements were formulaic endings at that time.3. Many ending (and beginning) statements are difficult toclassify due to what appear to be overlapping categories in theclassification system. This has the effect of reducing reliability of the239data and suggests the classification system (Langer, 1986) may needto be tested more in the field.Length of Writing Sample1. There is considerable fluctuation in length of narrativeswithin each year for every student. Overall, most students'narratives increased in length in successive years. Mean lengths ofnarratives in each grade rose from approximately 300 words to 700words from Grade 4 to 6.2. Generally most students wrote longer reports each year.Grade mean lengths show report lengths increased fromapproximately 90 to 220 words from Grades 4 to 6. Most of theincreases in length occurred in November of Grade 6.3. Ratios of mean length of narrative to mean length of reportshow narratives were approximately three times longer than reports.DiscussionWhile these results are not statistically significant, they doindicate general patterns in the subjects' narrative and reportwriting. The study clearly suggests that the 14 students used morecomplex syntax and text forms in both their narrative and reportwriting over the three years. Further, the use of more complexsyntax and text forms varies substantially in the two genres.240Syntactic ComplexityThe first part of the study examined the students' writing usingfrequency counts of various forms of syntax. Previous researchindicates that increased use of many of these forms demonstratescomplexity, hence maturity, in writing. The results of the analyses ofthe data demonstrate the students' growing ability to use morecomplex syntax in their writing over the three year period.Generally, despite individual inconsistencies and fluctuations inusage within each year, students had increased complexity of syntaxin both genres over time. More students had longer T-units andincreased frequencies of extended noun phrases with modifiers,prepositional and appositive phrases and relative clauses in theirwriting in each year. Students also began to use clauses in place ofsimple noun phrases and, in particular, began to use complexlinguistic forms such as complement and subject clauses. Further,these students' writing showed increased use of complex forms of theverb phrase, more modals, passives, perfects, perfect progressivesand perfect passives yet decreased use of less complex verbs such assimple phrases and progressive forms. While these students usedincreasingly more complex elements of syntax over the years, it isalso apparent that use of these elements differed in narrative andreport writing. The students' use of syntax appears to differ in thetwo genres in three ways: some elements of syntax are later-forming in one genre than in another; some elements appear to bemore prevalent in one genre; some elements are used with higherfrequency in one genre than another.241Most of the more complex elements of syntax are used early inGrade 4 in the students' narrative writing, yet do not appear untilthe later grades in their report writing. Many forms of extendednoun phrases such as the use of modifiers, appositive phrases andrelative clauses are not used until the latter part of Grade 5 or earlyGrade 6 in students' reports but appeared early in their narratives.Similarly, many forms of the verb phrase are used much later inreport writing; students did not begin to use modals until Grade 5 intheir reports while modals were used frequently in Grade 4narratives. Use of the passive and perfect forms of the verb wererarely found until Grade 6 in report writing yet appeared muchearlier in narrative writing. Furthermore, analysis of the ratios ofuse of verbs in the main clause to subordinate clause indicates thatuse of complex verbs in the subordinate clause did not occur untilmuch later in report writing. The students used modal, catenative,simple, passive and perfect forms of the verb predominantly in themain clause in their reports in Grades 4 and 5 and did not have amore balanced distribution between main and subordinate clauses ofthese verbs until Grade 6. These more complex verbs were used inboth the main and subordinate clauses more frequently and from anearlier age in the students' narratives.Closer examination of the writing in the case studies of thethree students provides further evidence that some forms of syntaxare later-forming. Ian, for example, did not use modals orappositives in his narrative or report writing until Grade 6 whereasTina and John used these forms much earlier. Similarly, the threestudents (particularly Tina and Ian) did not use complex forms of242relative clauses or many elements of the verb phrase such as modals,passives or perfects until Grades 5 and 6. Further, the three studentsused the most complex verbs such as passives and all forms of theperfect verb phrase almost exclusively in the main clauses in theirreport writing before they began to use these forms in both the mainand subordinate clauses. Forms of these verbs were found muchearlier in both main and subordinate clauses in their narrativewriting. Thus, it appears from these results that the students' use ofsyntax is later-forming in one genre than in another. Many of themore complex forms of syntax that were found early in narrativewriting are not found until later their reports.There are also differences in the use of syntax in the twogenres in that all students appeared to use certain elements ofsyntax more often in one genre than another. For example, forms ofextended noun phrases were used by more students in theirnarrative writing than in their report writing. While frequencieswere very low, the students had higher frequencies of appositivephrases, relative clauses and complement clauses in their narratives.Tina, for example, used twice as many relative clauses in hernarratives as in her reports. Similarly, in Grade 6, John had morecomplement clauses in his narratives than in his reports. On theother hand, he used subject nominal clauses more often in reportwriting than in narrative writing.Forms of the verb phrase are also more prevalent in one genrethan another. Students used catenatives, simple verb phrases,progressives, passives and all forms of the perfect verb phrase moreoften in narrative than report writing. Yet, they used modals more2 4 3often in report writing. In the three students' writing, use of formsof the verb were different in the two genres. Tina used moreprogressives in her narratives than almost any other student whileIan used more catenatives. John and Tina also had more passiveverbs in their narrative writing than in their report writing. Further,all three students, particularly John, used modals more often in theirreport writing. From this data it appears that genre affects the useof certain forms of syntax. Also, students appeared to use a widervariety of more complex forms of syntax in the narrative writing.Interestingly, there are also differences in the frequency ratesof certain elements of syntax in the two genres. While more studentsover the years used forms of syntax earlier and had a wider varietyof these forms in one genre than another, frequency counts of morecomplex syntactic elements were often higher in report writing thannarrative writing. Some forms of extended noun phrases and verbphrases, for example, were used with higher frequency in thestudents' reports than in their narratives. Prepositional phrases,which were used early in Grade 4 in both genres, were two to seventimes more frequent in reports than in narratives. Further, numbersper 100 words of Stage I and II relative clauses (while infrequent)were higher in report writing as were forms of the verb such ascatenatives. Similarly, forms of syntax which were not used bymany students until the later grades were also found in highernumbers in report writing. Noun phrases extended with modifiers,for example, were four times higher in Grade 6 reports than in Grade6 narratives. Also, more students began to use appositive phrasesand nominal clauses with higher frequencies in their Grade 6 reports.2 4 4John and Tina had higher rates of many forms of extendednoun phrases in their report writing than in their narrative writingas early as Grade 4. They had higher frequencies of modifiers,appositives, and nominal clauses. Ian also had higher frequencies ofsome of these forms, but only in Grade 6. In particular, he usedmore modifiers in his reports than in his narratives. Students alsohad higher frequencies of more complex forms of the verb in reportwriting. By Grade 6, students had two to four times as many modalsin their reports as in their narratives in addition to higherfrequencies of passives and perfect forms of the verb phrase. Tina,John and Ian also had higher numbers per 100 words of these forms:Tina and John in Grade 5 and Ian in Grade 6.Overall, more students had a wider variety of complex forms ofsyntax at an earlier age in their narrative writing. Yet, resultsindicate that, by Grade 6, most students had the same variety ofcomplex forms of syntax but with much higher frequencies in theirreports. This finding indicates that students' report writing is moresyntactically complex than their narrative writing especially at thesegrade levels. Further, an examination of mean T-unit lengthsupports the notion that students used more complex syntax in theirreport writing. Individual comparisons of T-unit length indicate thatstudents produced longer T-units in their report writing than in theirnarrative writing from Grades 4 to 6. Individual data show thatmost students did not have longer T-units in their reports until earlyGrade 5 but there were some students whose mean T-unit lengthwas longer in the early part of Grade 4. John, for example, hadsubstantially longer mean T-unit length in his Grade 4 reports than245in his Grade 4 narratives. These data combined with the finding thatmany students had higher frequencies of some early-formingelements of syntax such as the use of prepositional phrases andcatenative verb phrases, provide evidence that some students usedmore complex syntax in their report writing from as early an age asGrade 4. Certainly, John is an example of a student who had morecomplex syntax in his report writing than in his narrative writing inGrade 4 and he continued to use more complex syntax in successiveyears.Factors Affecting the Analysis and Production of Syntactic Complexity In examining the writing of the subjects and, in particular, thethree case studies, it becomes apparent that there are several factorswhich seem to affect the use of syntax in these subjects' writing.These factors are evident in the following general areas: use ofcertain forms of language, stylistic preferences, types of writing,maturation, and motivation.The use of certain forms of language in narrative and reportwriting apparently affects mean T-unit length and decreases orincreases frequencies of syntactic elements. In narrative writing, forexample, the use of dialogue reduces mean T-unit length. Further,this reduction in T-unit length due to dialogue persists even aschildren mature. A section of dialogue in Tina's Grade 6 narrativehad a mean T-unit length of less than half of the majority of the T-units in the writing sample. Dialogue often includes short simplesentences in which the frequencies of certain forms of the verb such246as progressives and modals are increased while forms such asextended noun phrases with modifiers decrease. A commonlanguage feature in the report writing of students of this age is theuse of lists. In every grade students use long lists to explain a topic.These lists inflate mean T-unit length, particularly in the earliergrades, when reports can be less than 100 words long. One T-unitwith a lengthy list such as in Tina's report about school can inflatemean T-unit length by almost two words per T-unit. Thus mean T-unit length did not seem to be an adequate measure to describe thesyntactic complexity of these students' writing.Another form of language that affects the analysis of syntax inboth narrative and report writing is the use of repetitive phrases.The inclusion of phrases which are typically found in students'speech, for example, can increase frequencies of certain forms ofsyntax. Ian demonstrates this phenomenon as he frequently used"got" with a past participle form of the verb phrase in a narrative.This repetition greatly increased his frequency of catenative verbphrases. Similarly, students may also include phrases which aremodelled from language heard or read in books. It appears studentsuse modelling as a strategy to help them write. Several studentsrepeated a phrase modelled from a novel in this study which createdhigh frequencies of several elements of syntax and inflated T-unitlength.Style of writing is also a factor which appears to createdifferences in the frequencies of certain elements of syntax amongindividual writers. Stylistic preferences such as John's use ofprepositional phrases and Tina's use of relative clauses and247progressive verb phrases create higher frequencies of these elementsin their narratives than in those of most students. Other stylisticpreferences such as writing reports from a first or second personstance increase frequencies of modal auxiliary verbs, particularly inthe main clause.Clearly, students' use of syntax is affected by genre. Reportwriting requires the use of more complex syntax than narrativewriting according to previous research (Langer, 1986). As theymature, these students appear to have more complex forms of syntaxin their report writing as compared to their narrative writing. It alsoseems that type of writing within a genre affects the use of syntax.In narrative writing, for example, some students (particularly Tinaand John) had high frequencies of modifiers, prepositional phrasesand relative clauses in fantasies. The frequencies of these forms arenoticeably higher than in adventure or mystery stories written bythe same students in the same year. These forms of modificationmay be used with higher frequency to provide the descriptive detailthat is necessary to enable the reader to visualize the imaginaryscenes in the story. Similarly, some students used moreprepositional phrases, modals and passives in reports which giveinstructions than in those which are simply descriptive. Reportswhich present a sequence of instructions seem to elicit more complexforms of syntax.In analyzing the writing of the three case study subjects itappears that maturation rate plays a substantial role in the use ofsyntax. Whereas Tina and John used many more complex forms ofsyntax in Grades 4 and 5, Ian did not begin to use the same forms248until Grade 6. Tina and John seemed to mature earlier than Ian asthey had increased descriptive detail in their narratives which betterportrayed characters and events. Similarly, they seemed morecapable linguistically from an early age as they had a widerrepertoire of complex verb phrases and extended noun phrases intheir narrative writing. Further, Tina and especially John had morecomplex syntax in their reports at an early age whereas Ian did not.They had higher frequencies of modals, prepositional phrases, andcatenative forms of the verb. Also, Tina's use of the passive in herGrade 6 reports indicated a more mature style of writing than eitherJohn or Ian. It is interesting, however, that by Grade 6 Ian began touse most of the forms that Tina and John had been using from anearlier age. His writing showed much more maturity at that time.Other students also demonstrated varying levels of maturity intheir writing as described by the use of many of the elements ofsyntax. Many wrote more complex and detailed narratives from anearlier age than others. Similarly, while some students were stillproviding simple descriptive reports with little elaboration of detailby Grade 6, others began to increase detail as well as vary the typesof reports that they wrote. Some appeared to develop the ability, forexample, to handle the complex cognitive operations that arenecessary to organize and explain an intricate set of instructions.One other factor which affects use of syntax is one thateducators cope with every day but researchers seldom mention -motivation. As these subjects were studied over a long period oftime with the researcher as participant observer, it became apparentthat students' attitudes and levels of motivation affected their249writing which, in turn, affected their use of syntax. Classroomobservations indicate several aspects of the writing situation canincrease or decrease motivation. One is choice of topic. If studentschose a topic which enabled them to draw upon recent, first handexperiences, their writing tended to contain much more detail, hencemore complex syntax. The ability to draw upon recent experienceseemed to enabled the students to incorporate more modifiers and awider variety of verbs and clauses as they described the events in anarrative or an object in a report. Further, they were moremotivated to write. By contrast, students often experiencedfrustration and a lack of motivation if they chose a topic which mighthave been of interest to them but about which they had littleknowledge. This often happened as these students wrote reports. Astudent, for example, might choose to write about a famous person.Failure to have much information about the person often resulted inless complex syntax as the student wrote simple, choppy sentenceswith little elaboration of detail.Another motivating factor can be difficulty or perceiveddifficulty in accomplishing the writing task. One reason studentsmay not use complex forms of syntax in report writing is that thetask appears to be more difficult than narrative writing. Forexample, even though they often had pre-writing discussions, bothTina and Ian struggled with report writing in Grades 4 and 5. Thisdifficulty, in turn, appeared to affect the students' flow and apparentcompetency in the use of syntax; elements of complex syntax thatwere used often in their early narratives disappeared in theirreports. Other motivational factors include general attitude toward250the writing task (whether it is perceived as important), attitudetoward school in general, personal problems, or even time of year.Text Forms in WritingThe second part of the study examined the text forms used bystudents in their writing. Analyses of the beginning and endingstatements and overall length of writing demonstrate the students'increasing ability to structure their writing in more complex ways.Further, these analyses indicate there are clear differences in thestudents' use of these forms in the two genres.Beginning and Ending Statements Generally, students used much more complex beginning andending statements in their narratives than in their reports. In eachyear, all students began with either a formulaic or structuralbeginning in their narratives. These beginnings clearly establish thegenre and orient the reader to the form and content of the narrative.There are very few weak beginning statements in the students'narrative writing. Over time, the number of formulaic beginnings inthe students' writing decreased while the number of structuralbeginnings increased. The students also began to write a widervariety of structural openings as they grew older. Further, thecontent of their opening sentences often changed as they added moredescriptive detail. The students' use of the more complex openingsin their narratives is one indication of their well-developed sense ofschema for narrative writing. By contrast, beginning statements in251report writing rarely established the genre successfully. Most oftenstudents opened their reports with weak beginnings statements.These weak openings were in the form of a generalized fact. Oncloser examination of the writing, it appears most studentsconsidered the title to be all that was needed to successfully open areport. Both Ian and Tina opened their early reports with anopening fact framed around the title. Structural and formulaicbeginnings were rare in report; John was the only student who had astructural beginning in Grades 4 or 5. Only four students used thesemore complex beginnings in Grade 6. Most students, even by the endof the study, seemed unable to use a structural opening statement tobegin their reports.The majority of students had formulaic endings, in particular"the end," in their earlier narratives. In each successive year asstudents matured, they used more structural endings. These endingsbring the writing to a structural close, leaving the reader with asense of completeness. As the students grew older, they also had awider variety of structural endings. Tina and John, for example, hada selection of 'summary', 'moral', 'solution resolution' and 'naturalending' statements in their Grade 5 and 6 narratives. Throughoutthe years many students seemed to be influenced by theconventional storytelling ending "the end," in their narrative writing.Over half the students in Grade 6 who concluded their narrativeswith a structural ending statement also added "the end." John, inparticular, showed this tendency as he added "the end" to both of hisGrade 6 narratives even though he also provided a complexstructural ending statement. Weak endings were rare in narrative252writing. Ian is one of a very few students who had a weak ending ina narrative.Ending statements used in report writing are much lesscomplex than those used in narrative writing. Weak 'last fact'endings predominated in students' reports in all grades. Theseendings do not provide the reader with any sense of closure. Only bythe end of Grade 6 were students able to use more complex endingstatements which successfully closed their reports. At this time,slightly less than half the students began to use structural orformulaic ending statements. Of the three case study students, onlyTina used a structural ending in her Grade 6 report. Neither John norIan used a structural ending in any of his reports.Length of WritingThe examination of mean length of writing reveals that therewere wide variations in the lengths of students writing sampleswithin each year. Tina and John, for example, had variations oflength of 600 and 1000 words respectively within one year in theirnarratives. Notwithstanding the varying lengths of writing samplesin each year, every student had increased lengths of narratives andreports through the years. Also, in each year, ratios of mean lengthof narrative to mean length of report indicate that studentsconsistently wrote narratives which were at least three times longerthan their reports.253Factors Affecting Text FormsThus, it appears that students have an ability to use morecomplex beginning and ending statements in narrative than inreport. Further, they were able to use a wider variety of morecomplex beginnings and endings in narrative from an earlier age. Inreport writing, by contrast, the students used opening facts and lastfacts as beginning and ending statements which did not increase incomplexity or variety until Grade 6 and even then in a limited way.Reasons for the students' ability to use more complex beginning andendings statements are not clear. It seems, however, that task orperception of task and development of schema are important factorswhich contribute to the differences in ability to open and closenarratives and reports competently. Whereas students had littledifficulty writing opening and closing sentences in their narrativewriting, they had a great deal of difficulty completing the same taskin report writing, especially in the earlier grades. They seemedfrustrated with the task and were unable to create any variety intheir opening and closing statements. Apparently the students wereable to depend upon their knowledge of story schema to assist themin beginning and closing their narratives. Their wide variety ofcompetent opening and closing statements demonstrated thisknowledge. By contrast, students seemed to lack any schema forreport writing. Their only strategy was to frame their reportsaround the title of the piece and end on a last fact. As statedpreviously, this is a common strategy in report writing with studentsof this age (Langer, 1986).254Reasons for the differences in length of writing appear to beidentical to those that affect the use of syntax: use of language, styleof writing, maturation and motivation. The main reason fordifferences between narrative and report appear to be attributableto the difficulty students have in generating content for their reports.This difficulty is demonstrated by the three case study students whocomplained that they had trouble generating enough information fortheir reports in all years, especially in Grades 4 and 5. In fact, moststudents found that producing enough information for their reportswas difficult. Their difficulty may have been caused by a poor choiceof topic, lack of knowledge about a topic, or an overall lack ofunderstanding of the form of report writing.ConclusionsThe study indicates that there were substantial differences inthe use of syntax and text forms in these students' narrative andreport writing over the three years. These differences in use ofsyntax and text forms have important meaning for educators.Further, it was also apparent that the examination of the students'syntax and text forms appeared to be affected by the methodologyused in the study. Some methods of analysis did not seem to providean accurate assessment of these forms in the children's writing. Thisfinding suggests some of these analyses may need carefulconsideration when used in children's writing research.Over the three years, the students appeared to improve writtencompetency in both narrative and report writing as measured by the255use of more complex syntax. It was also apparent that there weremany differences in students' levels of maturation. The studentsbegan to use more complex forms of syntax at varying timesindicating differences in maturity in their writing. These beginninguses of complex syntax are important cues for educators. Forexample, if teachers are able to recognize the beginning uses of morecomplex syntax which indicate maturity in writing, they may bebetter able to assess their students' linguistic capabilities. Thisassessment is important in describing what children "can do."Further, teachers may wish to help students assess their own writingthrough recognition of mature forms of syntax which, in turn, mightencourage continued and more intensive use of these forms.The students' increasing competency in the use of syntax,however, appeared to be affected by genre. Generally, moststudents' use of syntax was more complex in narrative writing thanin report writing, particularly in Grades 4 and 5. Over time,however, the complexity of syntax increased in report writing untilthe use of some forms superseded those in narrative writing. Thestudents' well-developed sense of story appeared to enable them totap their full linguistic capabilities and use complex syntaxconfidently at an early age in their narrative writing. By contrast,the students' lack of knowledge of report writing seemed to affectthe use of syntax as many of the complex linguistic structures thatwere readily apparent in narrative writing were not evident inreport writing. Students' complexity of syntax seemed to decrease asthey apparently struggled with writing in this genre. Over time,however, complex elements of syntax were used with rapidly256increasing frequency in report writing indicating this genre elicitsmore complex syntax. Most of the students in this study werebeginning to use more complex syntax in report writing than innarrative writing by the end of Grade 6. This finding suggests reportwriting appears to place increased linguistic demands upon writersof this age. Teachers will need to be cognizant of the increaseddemands that exposition places upon their students, especially interms of planning instructional activities and assessing writtencompetency.Students' use of text forms was also different in narrative thanin report writing. From an early age the students' knowledge ofform, indicating strong cognitive schema for narrative, wasdemonstrated as they began and ended their narratives competentlywith a wide variety of complex statements and generated stories ofincreasing length in each year. On the other hand, students' apparentlack of schema for report writing created difficulty in that they wereunable to begin or end reports successfully. Even after three years,many of the students were still unable to use the more complexstructural beginnings and endings that were so readily apparent intheir narrative writing. Furthermore, students were unwilling orunable to generate sufficient content in their reports. Most studentswere only beginning to grasp the amount of detail needed to write areport by Grade 6. The students' lack of knowledge of the genre isapparent in that most students' only strategy for writing reports wasto begin with a title and list a number of facts. This strategy appearsto be identical to Bereiter and Scardamalia's "knowledge telling"257which they describe as "a simple serviceable strategy that . . . . lacksplanful pursuit of rhetorical goals" (1982, p. 34).Thus, report writing appeared much more difficult for thestudents than narrative writing as evidenced by the increasedlinguistic demands placed upon them and their apparent lack ofschema for the genre. Accordingly, teachers need to recognize thedemands of certain tasks upon their writers and find ways to assistthem. The study indicates there may be ways teachers can assisttheir students with report writing. One way to assist students maybe to increase their experience with report writing. Levels ofexperience seemed to affect these students' ability to write innarrative and especially, report writing. John, for example, appearedto be successful writing reports, perhaps because he read non-fictionextensively and wrote from his reading experiences. Most of theother students seemed to lack experience with reading non-fictionand therefore had little background knowledge from which to write.This finding seems to demonstrate that in order for students to writereports successfully, they have to gain experience with the genre.Generally, these students' experiences with narrative were plentiful.They had stories read to them, studied novels, and wrote storiesextensively, all of which contributed to their schema for narrativewriting. Similar activities for report writing were not evident.Therefore, it would seem that educators must provide the sameopportunities for students to experience and receive exposure toreports, that is, reading reports aloud, analyzing reports andpractising report writing. This exposure may not only assist indeveloping cognitive schema for the genre but might also improve258students' attitudes toward report writing (which seemed quitenegative in this study).The difficulty students have writing reports suggests educatorsmust also provide students with strategies for report writing. Onestrategy is to assist students in choosing topics. Many of the studentsin this study appeared to write much more competently (morecomplex syntax and increased content) when they wrote from recentexperiences or with sufficient background knowledge. Students maybenefit from instruction in how to choose topics which best representtheir knowledge and experiences. Interestingly, many studentsseemed to have less difficulty when they wrote reports which gaveinstructions. These reports were logically ordered and containedmore complex syntax and content. The inherent sequence of steps inthe format of instructions may have provided the students with theorganizational framework for the report which freed them toconcentrate on generating content. Students cannot always dependon a time-related sequence to organize their reports as they can innarrative writing. This makes organization difficult. Perhapsteachers could assist by providing report topics which have anorganized framework, such as the sequence to a simple game. Thiskind of report writing may provide an intermediate step to writingmore sophisticated reports which would require organizing largeamounts of information.Many of the students' reports resembled parts of conversationor had speech-like phrasing in them. Teachers may wish to buildupon students' oral language ability that they so naturally transfer totheir report writing. Perhaps a "talk/write" strategy might lessen2 5 9the difficulty of finding content for writing reports. Students couldgenerate content through conversation, writing down the informationas they conversed. Teachers could then provide instruction as tohow to organize the information.The study also indicates that students' use of syntax and textforms is affected by several other factors such as motivation, use ofcertain forms of language and type of writing within narrative andreport writing. Many of these factors contribute to the differences intheir use of these forms in the two genres. Teachers will need to beaware that growth in written competency does not occur in a gradualsteady incline but rather in an irregular pattern of progression anddigression that can apparently be affected by such factors asdiscussed in this study.In addition to the factors which affected the students' use ofsyntax and text forms in this study, there were also severalmethodological factors which appeared to affect the results. Thesefactors provide interesting information for teachers and researcherswho might conduct studies of this kind. The writing processapproach, for example, appeared to assist the student in some waysin writing both narratives and reports. The students in this studywrote much longer narratives and reports than subjects in otherstudies (Harpin, 1976; Langer, 1986). Perhaps having the time towrite without restrictions enabled the students to generate morecontent in their writing. If longer writing is indicative of moremature writing, perhaps teachers need to allow students to writewithout time restrictions (although this may not always be practical).Freedom to choose topics appeared to benefit some students who260chose to write from previous experiences, yet this freedom may havecreated difficulties for others. Teachers will have to make decisionsas to whether to assist students with topic choice which is acomponent of process writing.Another methodological factor which affected the study wasthe analysis of writing. Many of the measures used in the studycreated difficulty for the researcher. T-unit length, for example,appeared to obscure important changes in language use and wasaffected by the use of dialogue and lists. Perhaps T-unit lengthshould be used in conjunction with other measure of syntax instudies such as this. Further, forms of extended noun phrases suchas appositives, relative clauses, both types of nominal clauses andforms of the verb phrase did not always seem to be suitablemeasures of syntactic complexity in the students' writing due to lowfrequencies. Using these forms, however, seemed important for theyhighlighted students' beginning use of more complex syntax which,as indicated previously, may be an important indicator of students'linguistic capabilities. Also, the classification system used tocategorize beginning and ending statements could not adequatelyreflect the changes in the language used over time. Sentences fromthe same sub-category in Grade 4 and 6 were often much moredetailed and complex in Grade 6. It would appear from this findingthat the system either needs supplementing or adapting. It shouldbe augmented by the use of qualitative data in order to adequatelydescribe changes in the beginning and ending statements of studentwriters. Finally, the study demonstrates a need for a wide samplingof writing to adequately reflect the capabilities of writers. Many of261the factors which affected the use of syntax and text forms, forexample, motivation of the students as well as the fluctuations andinconsistencies that were revealed in the data suggest one samplingof writing may not represent a writer's abilities.Thus, the finding that there are many differences in the use ofsyntax and text forms in students' narrative and report writing andmany factors that affect the use of these forms has importance forboth educators and researchers. This finding has several general andpedagogical implications and presents some interesting directions forfurther study.Implications of the StudyThis study is descriptive in nature and thus findings cannot begeneralized to the population from which the sample was drawn orto other populations. Nevertheless, several interesting pedagogicaland methodological implications arise from the findings, discussionand conclusions.1. In analyzing student writing samples it becomes apparentthat, due to many factors, the students have substantial fluctuationsin frequencies of syntactic elements and inconsistencies in the use ofsyntax within each year. Also, there are inconsistencies in mean T-unit length and the lengths of writing samples. The fact thatstudents are inconsistent in their use of many of the elements ofsyntax and inconsistent in the amount of writing they producedemonstrates the importance of sampling. Not only is a wide262sampling of writing important in terms of research but it would seemto be essential for instructional and assessment purposes.2. Mean T-unit length appears to be an unclear measure ofsyntactic complexity in the writing of students of these ages. It isaffected by certain uses of language such as dialogue and lists, itobscures fluctuations in an individual's writing and changes inlanguage that occur within the T-units, and it is affected not only bygenre but types of writing within the genre. Thus it should be usedin conjunction with other measures of syntactic complexity tosupplement its reliability.3. The study also appears to demonstrate that using complexsyntax is a developmental process. Generally, in each successiveyear, the students use more of the mature forms of syntax withhigher frequency. Individual developmental differences are alsoapparent. Both Tina and John, for example, used many more complexforms of syntax before Ian used them. Recognition of thesebeginning uses of syntax is an important element in assessment ofwritten competency. Teachers will need to be aware of increasingcomplexity in their students' writing. Further, they also need to beaware of the apparent effects of different genres and the types ofwriting and topics within a genre and their effect on the use ofsyntax. Teachers will need to consider these differences as theydesign instructional programs and evaluation procedures.4. The results seem to suggest that students use differentforms of syntax in different types of narratives. This has importantresearch implications. Studies which examine students' writingshould be cautioned against considering only mode in the analysis2 6 3and not types of writing within the mode. Perhaps researchersshould only compare data where students have written the sametypes of narrative; that is, compare fantasy to fantasy and notfantasy to mystery.5. The results of this study indicate that from an early age,students understand that report writing is different from narrativewriting. Even as early as Grade 4, they began to use different syntaxand text forms in their reports. Report writing also appeared topresent them with difficulty. The students clearly lacked strategiesfor writing reports. Traditionally, teachers do not begin to teachstrategies for writing exposition (such as reports) until laterelementary or early high school. It may be that students wouldbenefit from receiving both exposure to, and instruction in, reportwriting at an earlier age, especially in strategies for writingexposition such as how to begin and end reports.6. These students experienced difficulty generating content intheir report writing, especially at an early age. Previous researchhas found a positive correlation between the reading and writingconnection: experience with reading exposition enhances the writingof exposition (Belanger, 1984, 1987; Hammil & McNutt, 1980;Shanahan, 1984; Taylor & Beach 1984). As stated previously, thestudents in this study had had little exposure to exposition at anearly age and this genre created much difficulty for them. Ifeducators are to expect students to be able to write in a specificgenre they must provide experience with that genre.7. Over the three years, it was also apparent that students'attitudes toward report writing were much less favourable than their264attitudes toward narrative writing. These less favourable attitudesmay be a result of less exposure to this genre. Perhaps teachersneed to provide students with more exposure to intriguing andinteresting reports. They might read reports to their students asoften as they read narratives. This exposure may, in turn, familiarizethe students with the genre of report writing, thereby enabling themto be more effective and enthusiastic toward writing their ownreports.Areas for Further ResearchThere are many questions which arise out of this study whichwould provide interesting considerations for further research.1. The results of this study suggest that these students used amore complex and varied syntax in their report writing than in theirnarrative writing, particularly as they grew older. Would these sameresults be found in similar studies with a larger subject population?Similarly, how might the results change in an investigation whichexamined younger or older students' writing?2. The study provides some evidence that the use of syntax isaffected by types of writing within a genre. Would there besignificant differences in the use of syntax in two types of narrativewriting or two types of report writing?3. The profiles of three writers indicate that students havecomplex forms of syntax in their narratives that do not occur in theirreports until they are older. Why, for example, does Ian useprepositional phrases extensively in his narratives but not in his265reports? How does report writing appear to influence his linguisticcapabilities? Further, why is Ian apparently unable to write muchon a topic about which he obviously has a great deal of information?Is this related to motivation or maturation or an inability to useorganizational strategies which would help to access information?4. Many of these students used a complex linguistic form suchas a complement or subject clause and then do not use it again.Would it be possible to teach students to recognize mature forms ofsyntax in their writing and encourage them to continue toincorporate these forms into subsequent pieces of writing? Would itbe possible to measure the effects of such instruction?5. All students in each year had difficulty generating contentfor their reports. The students were not permitted to consultresearch books to aid them in producing content for their reports. Itmight be interesting to study the effects of reading about a giventopic just prior to writing about the topic. Would access toinformation, without plagiarizing form or content, enable students touse more complex syntax and text forms from an earlier age?6. In terms of generating content, what influence does thestudents' choice of topic have on the length of reports they write?Could students benefit from having instruction in how to selecttopics; for example, learning to select topics that would be narrowenough to reduce organization problems and best represent theirgeneral knowledge and interest?7. The case studies of the three writers indicated that somewriters appeared to have stylistic preferences in the use of syntax;for example, Tina used many more relative clauses and progressive266and passive verb phrases in her narrative writing than any of theother students. It may be useful to investigate whether students'use of certain forms of syntax are indeed stylistic preferences andwhat appeared to influence or cause such preferences.8. Fluctuations and inconsistencies in the data suggest thatstatistical tests are not always appropriate for studying andanalyzing writing. Also, classification systems such as the one usedto describe beginnings and endings cannot adequately account forthe changes that take place in a student's writing. There is a needfor more detailed qualitative research in this area in order todescribe the changes that take place in writing over time. It wouldhave been interesting, for example, to interview the students toobtain their perceptions. How might they interpret the changestaking place in their writing or the differences in their narrative andreport writing?These are but a few of the questions that are generated by thisstudy. It is by testing of questions such as these that researcherswill continue to acquire new understandings of students' use ofsyntax and text forms in writing, thus providing teachers and otherprofessionals with new insights into the development of writing ingeneral.267BIBLIOGRAPHYAnderson, R.^(1937). An evaluation of various indices of linguisticdevelopment. Child Development, 8 62-68.Anderson, R. (1977). The notion of schemata and the educationalenterprise. In R. Anderson, R. Spiro, & W. Montague (Eds.)Schooling and the acquisition of knowledge (pp.415-431).Hillsdale, New Jersey: Lawrence Erlbaum Associates.Anderson, R., Spiro, R. & Montague, W. (1976). Instantiation ofgeneral terms. Journal of Verbal Learning and Verbal Behaviour, 15, 667-679.Anderson, R. Spiro, R. & Montague, W. (Eds.) (1977). Schooling andthe acquisition of knowledge. Hillsdale, New Jersey: LawrenceErlbaum Associates.Anderson, R. C., Spiro, R. J. & Anderson, M. C. (1978). Schemata asscaffolding for the representation of information in connecteddiscourse. American Educational Research Journal, 15, 433-440.Applebee, A. N. (1982). Writing and learning in school settings. InM. Nystrand, What writers know: the language and structure ofwritten discourse (pp. 365-381). New York, New York:Academic Press Ltd.Applebee, A. N. (1984). Contexts for learning to write: studies of secondary school instruction. Norwood, New Jersey: AblexPublishing Corporation.Barritt, L. S. & Kroll, B. M. (1978). Some implications of cognitive-developmental psychology for research in composing. In C. R.Cooper and L. Odell, Research on composing points of departure(pp. 49-59). Urbana Illinois: National Council of Teachers ofEnglish.Bateman, D. R. & Zidonis, F. J. (1966). The effect of a study of transformational grammar on the writing of 9th and 10th graders(Research Report No. 6). Champaign, Illinois: National Council ofTeachers of English.268Bauers, A. & Nicholls, J. (1986). Early writing. In A. Wilkinson (Ed.)The writing of writing (pp. 134-157). Milton Keynes: OpenUniversity Press.Beach, R. & Taylor, B. (1981). The relationships between reading recall ability and writing of expository texts. UnpublishedManuscript, University of Minnesota.Beach, R. & Hynds, S. (1990). Developing Discourse Practise in Adolescence and Adulthood. Norwood, New Jersey: AblexPublishing Corp.Bear, M. V. (1939). Children's growth in the use of written language.Elementary English Review, 16, 312-319.Belanger, J. (1987). Theory and research into reading and writingconnections: a critical review. Reading - Canada - Lecture, 1_10- 21.Belanger, J. (1989). The influence of a reading treatment on thewritten compositions of Grade 11 Students: a research update.The Alberta Journal of Education, 35 34 - 46.Bereiter, C. (1980). Development in writing. In L. W. Gregg & E. R.Steinberg, Cognitive processes in writing (pp. 73-93). Hillsdale,New Jersey: Lawrence Eribaum Associates Inc.Bereiter, C. & Scardamalia, M. (1982). From conversation tocomposition: the role of instruction in a developmental process.In R. Glaser (Ed.) Advances in instructional psychology (pp. 1-48). Hillsdale, New Jersey: Lawrence Eribaum Associates, Inc.Black, J. B., Wilkes-Gibbs, D. & Gibbs, R. W. (1982). What writersneed to know that they don't know they need to know. In M.Nystrand What writers know: the language and structure of written discourse (pp. 365-381). New York: Academic PressLtd.Borg, W. K. and Gall, M. D. (1989). Educational Research. WhitePlains, New York: Longman Inc.269Bortz, D. R. (1969). The written language patterns of intermediategrade children when writing compositions in three forms:descriptive expositive and narrative. Dissertation Abstracts International, 30, 5332A. (University Microfilm No. 70-10, 586)Boyd, W. (1927). Development of sentence structure in childhood.British Journal of Psychology, 17, 77-111.Braddock, R., Lloyd-Jones, R. & Schoer, L. (1963). Research in writtencomposition. Champaign, Illinois: National Council of Teachers ofEnglish.Braun, C. & Klassan, B. (1976). A transformational analysis of writtenstructures of children representing varying ethno-linguisticcommunities. Research in the Teaching of English, 7, 312-323.Britton, J. et al. (1975). The development of writing abilities (11-18).London: Macmillan Education.Britton, J. (1978). The composing processes and the functions ofwriting. In C. R. Cooper & L. Odell (Eds.), Research in composing points of departure (pp. 13-29). Urbana, Illinois: National Councilof Teachers of English.Brooks, L. W. & Dansereau, D. F. (1983). Effects of structural schematraining and text organization on expository prose processing.Journal of Educational Psychology, 75, 811-820.Burgess, et al. (1973). Understanding children writing.Harmondsworth, Middlesex: Penguin Books Inc.Burke, G. et al., (1980). The Richmond Writing Project: Richmond,B.C. Richmond School District.Carlin, E. (1986). Writing development - theory and practise. In A.Wilkinson (Ed.), The writing of writing. (pp. 177-197). MiltonKeynes: Open University Press.Chomsky, C. (1969). The acquisition of syntax in children from 5 to 10. Cambridge (Research Monograph No. 57). Cambridge: The M.I. T. Press.2 7 0Chomsky, N. (1957). Syntactic Structure. 's-Gravenhage: Moutin andCo.Chomsky, N. (1965). Aspects of a theory of syntax. CambridgeMass.: M.I.T. Press.Christie, F. (1986). Writing in schools: generic structures as ways ofmeaning. In B. Couture (Ed.), Functional approaches to writing:research perspectives. (pp. 221-240). London: Frances Pinter(Publishers) Ltd.Christensen F. (1967). Notes toward a new rhetoric: six essays forteachers. New York: Harper & Row.Clay, M. (1983). Getting a theory of writing. In B. Kroll & G. Wells(Eds.) Explorations in the development of writing (pp. 259-284).Chichester: John Wiley& Sons.Collerson, J. (1983). One child and one genre: developments in letterwriting. In B. Kroll & G. Wells (Eds.), Explorations in the development of writing. (pp. 239-259). Chichester: John Wiley &Sons.Collins, A. & Gentner, D. (1980). A framework for a cognitive theoryof writing. In L. W. Gregg & E. R. Steinberg, (Eds.), Cognitive processes in writing (pp. 51-71). Hillsdale, New Jersey: LawrenceErlbaum Associates Inc.Cooper, C. R. & Odell, L. (1978). Research on composing: points of departure. Urbana, Illinois: National Council of Teachers ofEnglish.Cooper, C. R. & Matsuhashi, A. (1983). A theory of the writingprocess. In M. Martlew, (Ed.) The psychology of written language: developmental and educational perspectives. (pp. 3-39).Chichester: John Wiley & Sons.Couture, B. (Ed.) (1986). Functional approaches to writing: researchperspectives. London: Frances Pinter (Publishers) Ltd.Cowie, H. (Ed.) (1984). The development of children's imaginative writing. Beckenham, Kent: Croom Helm Ltd.2 7 1Crossley, M. & Vulliamy, G. (1984). Case study research methods andcomparative education. Comparative Education. 20. 193-207.Crowhurst, M. (1978). The effect of audience and mode of discourseon the syntactic complexity of the writing of sixth, and tenthgraders. ^Dissertation Abstracts International, 38 7300A -733110A.Crowhurst, M. (1980a). Syntactic complexity in narration andargument at three grade levels. Canadian Journal of Education, 5,7 -1 5 .Crowhurst, M. (1980b). Syntactic complexity and teacher's qualityratings of narrations and arguments. Research in the Teaching of English, 14, 223-231.Crowhurst, M. (1983). Persuasive writing at grades 5. 7. and 11: a cognitive developmental perspective. Paper presented at theannual meeting of the American Educational ResearchersAssociation. (ERIC Document Reproduction Service No. ED 230977).Crowhurst, M. (1987). Cohesion in argument and narration at threegrade levels. Research in the Teaching of English, 21, 185-201.Crowhurst, M. (1990). The development of persuasive/argumentative writing. In R. Beach & S. Hynds (Eds.) Developing discourse in adolescence and adulthood volume XXXIX. (pp. 200-223). Norwood, New Jersey: Ablex Publishing Co.Crowhurst, M. & Piche, G. (1979). Audience and mode of discourseeffects on syntactic complexity in writing at two grade levels.Research in the Teaching of English, 13, 101-109.Daiker, D. A., Kerek, A., & Morenberg, M. (Eds.) (1978). Sentence combining and the teaching of writing: selected papers from the Miami University Conference, Oxford, Ohio. Conway, Arkansas:Department of English, University of Central Arkansas.Davis, E. J. (1941). The location of the subordinate clause in oral andwritten language. Child Development, 12 333-338.2 7 2D.E.S. (Department of Education and Science) (1975) A language forlife. [The Bullock Report.] London: H.M.S.O.Dillon, G. L. (1981). Constructing texts: elements of a theory of composition and style. Bloomington, Indiana: Indiana UniversityPress.Durkin, K. (1986). Language development in the school years.London: Croom Helm Ltd.Durst, K. (1984). The development of analytic writing. In A.N.Applebee, Contexts for learning to write (pp. 9 - 102). Norwood,N.J. Ablex Publishing Corp.Durst, R. K. (1987). Cognitive and linguistic demands of analyticwriting. Research in the Teaching of English, 21, 347-376.Endicott, A. (1973). A proposed scale for syntactic complexity.Research in the Teaching of English, 7, 5-12.Englert, C. S., Raphael, T. E. & Kirschner, B. W. (1985). The impact oftext structure instruction and process writing on students' comprehension and composition of expository text. Paperpresented at the National Reading Conference, San Diego, California.Faigley, L. (1979). Names in search of a concept: maturity fluency,complexity and growth in written syntax. College Composition andCommunication. 31, 291-300.Fawcett, R. P. & Perkins, M. R. (1980). Child language transcripts 6-12, vols. I-IV Pontypridd: Ploytechnic of Wales.Flavell, J. H. (1963). The developmental psychology of Jean Piaget.Princeton, New Jersey: Van Norstrand.Fredericksen, C. H. & Dominic, J. F. (Eds.) (1981). Writing the nature.development and teaching of written communication, volume 2:process. development and communication. Hillsdale: LawrenceErlbaum Associates.Freedle, R. 0. (Ed.) (1979). New directions in discourse processing:volume 2. Norwood, New Jersey: Ablex Publishing Co.2 7 3Frogner, E. (1933). Problems of sentence structure in pupil's themes.English Journal, 22, 742-749.Glaser, R. (1982). Advances in instructional psychology: volume 2.Hillsdale, New Jersey: Lawrence Erlbaum Associates Inc.Golub, L. S. & Fredrick, W. C. (1970a). Written language. Paperspresented at the annual meeting of the American EducationalResearch Association, Minneapolis. (ERIC Document ReproductionService No. ED 042 750)Golub, L. S. & Fredrick, W. C. (1970b). An analysis of children'swriting under different stimulus. Research in the Teaching of English, 30, 168-180.Golub, L. S. & Kidder, C. (1974). Syntactic density and the computer.Elementary English. 51. 1128-1131.Gorman, T.P., White, J., Brooks, G., Maclure, M. & Kispal, A. (1988).Language performance in schools: a review of APU language monitoring. London: Assessment of Performance Unit, HMSO.Graves, (1984). Writing: teachers and children at work. Portsmouth,New Hamshire, Heinamann Educational Books.Gregg, L. W. & Steinberg, E. R. (Eds.) (1980). Cognitive processes in writing. Hillsdale, New Jersey: Lawrence Erlbaum Associates Inc.Gundlach, R. A. (1981). On the development of children's writing. InC. H. Fredericksen & Dominic, J. F. (Eds.) Writing: the developmentand teaching of written communication, volume 2: process, development and communication (pp. 133-151). Hillsdale:Lawrence Erlbaum Associates.Halliday, M. A. K. & Martin, J. R. (Eds.) (1981). Readings in systemic linguistics, London: Botsford Academic and Educational Ltd.Hammill, C., & McNutt, G. (1980). Language abilities and reading: areview of the literature of their relationship. The Elementary School Journal, 80, 269-277.274Harpin, W. S. (1973). Social and educational influences on children'sacquisition of grammar. (Unpublished Report). Notttingham:School of Education University of Nottingham.Harpin, W. S. (1986). Writing counts. In Wilkinson, A. (Ed.) Thewriting of writing (pp. 158-176). Milton Keynes: Open UniversityPress.Harrel, L. E. (1957). A comparison of the development of oral andwritten language in school-age children. Monographs of the Society for Research in Child Development, 22, 3-78.Harris, J. (1986) Organization in children's writing. In J. Harris & J.Wilkinson. (Eds.). Reading children's writing (pp. 47-61). London:Allen and Unwin Ltd.Harris, J. & Wilkinson, J. (Eds.) (1986). Reading children's writing. London: Allen and Unwin Ltd.Hayes, J. R. & Flower, L. S. (1980). Identifying the organization ofwriting processes. In L. W. Gregg & E. R. Steinberg (Eds.) Cognitiveprocesses in writing (pp. 3-30). New Jersey: Lawrence ErlbaumAssociates.Heider, F. K. & Heider, G. M. (1940). A comparison of sentencestructure of deaf and hearing children. Psychological Monographs,52, 42-103.Hennings D. G. (1984). A writing approach to reading. In J. Jensen(Ed.) Composing and comprehending (pp. 191-200). Urbana, Illinois: ERIC Clearinghouse.Hidi, S. E. & Hildyard, A. (1983). The comparison of oral and writtenproduction in two discourse types. Discourse Processes, 6, 91-105.Hiebert, E. H., Englert, C. S., & Brennan, S. (1983). Awareness of textstructure in recognition and production of expository discourse.Journal of Reading Behaviour, 15, 63-79.Hillocks, G. Jr. (1986). Research On written composition. Urbana,Illinois: ERIC Clearinghouse on Reading and Communication Skills,National Institute of Education.275Hunt, K. W. (1964). Differences in grammatical structures written at three grade levels the structures to be analyzed by transformational methods (Cooperative Research Project No.1990). Florida State University.Hunt, K. W. (1965).^Grammatical structures written at three grade levels (Research Report No. 3). Urbana, Illinois: National Council ofTeachers of English.Hunt, K. W. (1970). Syntactic maturity in schoolchildren and adults.(Society for Research in Child Monographs, No. 134). Chicago:Chicago University Press.Hunt, K. W. (1983). Sentence combining and the teaching of writing.In M. Martlew (Ed.), The psychology of written language developmental and educational perspectives (pp. 99-126).Chichester: John Wiley & and Sons.Jacobs, Suzanne E. (1989). From syntax to genre: writers' use ofdefinite constructions in fifteen editorials. Written Communication,6, 528-554.Jensen J. (Ed.) (1984). Composing and comprehending. Urbana,Illinois: ERIC Clearinghouse.King, M. & Rentel, M. (1983). Present at the beginning. In P.Mosenthal, L. Tamor, & S. A. Walmsley. Research on writing: principles and methods (pp. 139-176). New York: Longman Inc.Kintsch, W. (1982). Text representations. In W. Otto & S. White(Eds.) Reading expository material. (pp. 33-68). New York:Academic Press,Klein, M. L. (1985). The development of writing in children pre-K through Grade 8. New Jersey: Prentice Hall Inc.Kress, G. R. (Ed.) (1976). Halliday: system and function in language.London: Oxford University Press.Kress, G. (1982). Learning to write. London: Routledge and KeganPaul Ltd.276Kroll, B. M. (1978). Cognitive egocentricism and the problem ofaudience awareness in written discourse. Research in the Teachingof English, 12, 269-281.Kroll, B. M., Kroll, D. L. & Wells, C. G. (1980). Researching children'swriting development: the Children Learning to Write' Project.Language for Learning, 2, 53-81Kroll, B. M. (1983). Antecedents of individual differences inchildren's writing attainment. In B. M. Kroll & G. Wells (Eds.)Explorations in the development of writing. (pp. 95-118).Chichester: John Wiley & Sons.Kroll, B. M. & Wells, G. (1983). Explorations in the development of writing. Chichester: John Wiley & Sons.Kroll, B. M. & Anson, C. (1984). Analysing structure in children'sfictional narratives. In H. Cowie (Ed.) The development of children's imaginative writing (pp.153-183).^Beckenham, Kent:Croom Helm Ltd.La Brant, L. (1933). A study of certain language development ofchildren in Grades 4-12 inclusive. Genetic Psychology Monographs,14, 387-491.Langer, J. A. (1986). Children reading and writing: structures and strategies. Norwood, New Jersey: Ablex Publishing Corp.Langer, J. A. (1992). Reading, writing, and genre development. InJ.W. Irwing & M. A. Doyle (Eds.) Reading/Writing Connections: Learning from Research (pp. 32-53). Newark: InternationalReading Association.Langer, J. A. & Applebee, A. N. (1987). How writing shapes thinking: A study of teaching and learning. Urbana, Illinois: NationalCouncil of Teachers of English.Lawton, D. (1968). Social class, language and education. New York:Schochen Books.277Lloyd-Jones, R. (1981). Rhetorical choices in writing. In C. H.Frederiksen & J. H. Dominic (Eds.) Writing: the nature. development and teaching of written communication, volume 2:process. development and communication (pp. 169-176).Hillsdale: Lawrence Erlbaum Associates.Loban, W. (1963). The language of elementary school children.(Research Report No. 1), Champaign, Illinois: National Councilof Teachers of English.Loban, W. (1976). Language development: Kindergarten through Grade Twelve. (Research Report No. 18), Urbana Illinois: NationalCouncil of Teachers of English.McCarthy, E. (1954). Language development in children. InCarmichael, L. (Ed.) Manual of child psychology. New York: Wiley.McMillan, J. H & Schumacher, S. (1989). Research in education a conceptual introduction. Glenview, Illinois: Scott Foresman andCompany.Mandler, J. (1979). Categorical and schematic organization ofmemory. In C. R. Puff (Ed.), Memory organization and structure. (pp. 37-52). New York: Academic Press, 1979.Mandler, J. & Johnson, N. (1977). Remembrance of things parsed andstory structure and recall. Cognitive Psychology, 9, 111-151.Martlew, M. (Ed.) (1983). The psychology of written language: developmental and educational perspectives. Chichester: JohnWiley & Sons.Martlew, M. (1984). The development of written language. In K.Durkin, Language development in the school years. London:Croom Helm Ltd.Martin, J. R. & Rothery, J. (1981).^The otogenesis of written genre. InJ. R. Martin & J. Rothery (Eds.), Working papers in linguistics no. 2.Sydney: University of Sydney.2 7 8Martin, J. R. & Rothery, J. (1986). What a functional approach to thewriting task can show teachers about good writing. In B. Couture,Functional approaches to writing: research perspectives. London:Frances Pinter Publishers Ltd.Mellon, J. (1969). Transformational sentence-combining: a method for enhancing the development of syntactic fluency in English composition (Research Report No. 10). Urbana, Illinois: NationalCouncil of Teachers of English.Mellon, J. (1978). Issues, theory and practice of sentence combining:a twenty year perspective. In D. A. Daiker, A. Kerek & M.Morenberg (Eds.), Sentence combining and the teaching of writing: selected papers from the Miami University Conference. Oxford. Ohio. Conway, Arkansas: Department of English, Universityof Central Arkansas.Meyer, B. J. (1982). Reading research and the composition teacher:the importance of plans. College Composition and Communication,33, 37-49.Ministry of Education. (1989). Primary program document. Victoria:Province of B. C. Educational Programs.Ministry of Education. (1990). Intermediate Program: Learning in British Columbia. Response Draft. Victoria: Province of B. C.Educational Programs.Morgan, P. (1983). Responding to children's narratives. In Harris, J.and Wilkinson, J. (Eds.) Reading children's writing (pp. 32-45).London: Allen and Unwin Ltd.Mosenthal, P., Tamor, L. & Walmsley, S. A. (1983). Research on writing: principles and methods. New York: Longman Inc.Myklebust, H. R. (1973). Development and disorders of written language: volume two studies of normal and exceptional children.New York: Grune and Stratton.Newkirk, T. (1987). The non-narrative writing of young children.Research in the Teaching of English, 21, 121-145.279Nystrand, Martin. (1982). What writers know: the language and structure of written discourse. New York, New York: AcademicPress Ltd.Nystrand, Martin. (1986). The structure of written communication: studies in reciprocity between writers and readers. Orlando,Florida: Academic Press Ltd.O'Donnell R. C., Griffin, W. J. & Norris, R. C. (1967). Syntax of Kindergarten and elementary school children: a transformational analysis (Research Report No 8). Champaign, Illinois: NationalCouncil of Teachers of English.0' Donnell, R. C. (1976). A critique of some indices of syntacticmaturity. Research in the Teaching of English, LQ, 31-38.O'Hare, F. (1973). Sentence combining: improving student writing without formal grammar instruction (Research Report No 15).Urbana, Illinois: National Council of Teachers of English.Oliver, E. I (1989). Effects of assignment on writing quality at fourgrade levels. English Quarterly, 21, 224-232.Olson D. R. & Torrence, N. Learning to meet the requirements ofwritten text: language development in the school years. In C. H.Fredericksen & J. F. Dominic (Eds.), Writing the nature. development and teaching of written communication, volume 2: process. development and communication. (pp. 233-255).Hillsdale: Lawrence Erlbaum Associates.Otto, W. & White, S. (Eds.) (1982). Reading expository material. NewYork: Academic Press.Pellegrini, A. and Yawkey, T. (1984). The development of oral and written language in social contexts. Norwood, New Jersey: AblexPublishing Co.Perera, K. (1984). Children's writing and reading: analysing classroom language. Oxford: Basil Blackwell Inc.280Perera, K. (1986). Grammatical differentiation between speech andwriting in children aged 8-12. In A. Wilkinson (Ed.), The writing ofwriting (pp. 91-108). Milton Keynes: Open University Press,1986, 91-108.Perron J. D. (1976). The impact of mode on written syntactic complexity. part III - fifth grade studies in language education,Report No. 27. Department of Language Education, University ofGeorgia.Perron, J. D. (1977). Written syntactic complexity and the modes of discourse. Paper presented at the Annual Meeting of the AmericanEducational Research Association, New York.Petrosky, A. R. (1982). From story to essay: reading and writing.College Composition and Communication, 33. 19-36.Petty, W. T. (1978). The writing of young children. In C. R. Cooper &L. Odell, Research on composing: points of departure (pp. 73-85).Urbana, Illinois: National Council of Teachers of English.Phelps, L. W. (1988). Composition as a human science. New York:Oxford University Press Inc.Propp, V. (1978). Morphology of the folktale (Translated by L. Scott).Austin, Texas: University of Texas Press.Puharich, K. C. (1989). The influence of story schema on reading response and writing process. Unpublished Masters Thesis,University of British Columbia.Quirk, R., Greenbaum, S., Leech, G. & Svartvik, J. (1972). A grammarof contemporary English. London: Longman, 1972.Raphael, T. E., Englert, C. S., & Kirschner, B. W. (1989). Students'metacognitive knowledge about writing. Research in the Teachingof English, 23, 343-379.Richardson, K., Calman, M., Essen, J. & Lambert, L. (1976). Thelinguistic maturity of 11 year olds: some analysis of the writtencompositions of children in the national child development study.Journal of Child Language, 3, 99-15.2 8 1Riling, M. E. (1965). Oral and written language of children in Grades 4 and 6 compared with the language of their textbooks (Cooperative Research Project No. 2410). U. S. Office of Education:Durant, Oklahoma.Rosen H. (1969). An investigation of the effects of differentiated writing assignment o the performance in English composition ofa select group of 15-16 year old pupils, Unpublished doctoraldissertation, University of London.Rosen, H. (1973). Written language and sense of audience.Education Research, 15, 177-187.Rosen, C. & Rosen, H. (1973). The language of primary schoolchildren. Harmondsworth: Penguin.Rumelhart, D. E. (1975). Notes on a schema for stories. In D. Bobrow& A. Collins (Eds.), Representation and understanding: studies in cognitive science (pp. 211-236). New York: Academic Press.Rumelhart, D. E. & Ortony, A. (1977). The representation ofknowledge in memory. In R. Anderson, R. Spiro, & W. Montague(Eds.), Schooling and the acquisition of knowledge (pp. 415-431).Hillsdale, New Jersey: Lawrence Erlbaum Associates.Sam, N. H. & Stine, E. S. (1965). Structural analysis of the written composition of intermediate grade children (Cooperative ResearchProject No. S-057). U. S. Office of Education.Sampson, 0. (1964). A linguistic study of the written compositions of10 year old children. Language and Speech, 7, 176-182.San Jose, C. P. M. (1972). Grammatical structures in four modes ofwriting at fourth grade level. Dissertation Abstracts International, 33, 5411A (University Microfilms No. 73-9563).San Jose, C. P. M. (1978). Notes and comments. Research in theTeaching of English, 12, 91-92.2 8 2Scardamalia M. & Bereiter, C. (1986). How children cope with thecognitive demands of writing. In C. F. Frederiksen & J. F. Dominic(Eds.), Writing the nature, development and teaching of written communication volume 2 process, development and communication (pp. 88-103). Hillsdale: Lawrence ErlbaumAssociates.Schank, R. C. & Abelson, R. P. (1977). Scripts. plans. goals and understanding. Hillsdale, New Jersey: Lawrence ErlbaumAssociates.Schallert, D. L. (1982). The significance of knowledge: a synthesis ofresearch related to schema theory. In W. Otto & S. White.Reading expository material (pp. 14-47). New York: AcademicPress Inc.Seegers, J. C. (1933). The form of discourse and sentence structure.Elementary English, 10, 51-54.Smith, W. L. & Coombs, W. E. (1980). The effects of overt and covertcues on written syntax. Research in the Teaching of English, 14, 1 9 -3 8 .Stein, N. C. & Glenn, C. G. (1979). An analysis of story comprehensionin elementary school children. In R. 0. Freedle (Ed.) , New directions in discourse processing: volume 2. Norwood, NewJersey: Ablex Publishing Co.Stein, N. L., & Trabasso, T. (1982). What's in a story: an approach tocomprehension and construction. In Glaser, R. Advances in instructional psychology: volume 2. (pp. 213-247). Hillsdale, NewJersey: Lawrence Erlbaum Associates Inc.Stotsky, S. (1975). Sentence-combining as a curricular activity: itseffect on written language development. Research in the Teachingof English, 9, 30-71.Stotsky, S. (1983). Research on reading/writing relationships: asynthesis and suggested directions. Language Arts, 60, 627-642.Stormzand, M. J. & O'Shea, M. V. (1924). How much English grammar? Baltimore, Maryland: Warwache and York.283Strickland, R. G. (1962). The language of elementary school children:its relationship to the language of reading textbooks and the quality of reading of selected children. Bloomingdale, Indiana:School of Education, University of Indiana.Taylor, R. M. & Beach, R. W. (1984). Aspects of text instruction onmiddle grade students' comprehension and production ofexpository texts. Reading Research Quarterly, 19. 134-146.Tierney, R. J. & Pearson D. P. (1984). Toward a composing model ofreading. In J. Jensen (Ed.) Composing and comprehending (pp.33-45). Urbana, Illinois: ERIC Clearinghouse.Veal, L. R. (1984). Syntactic measures and noted quality of young children: studies in language education (Report No. 8) Athens:University of Georgia. ERIC Document Reproduction Services No.ED. 090555.Veal, L. R. & Tillman, M. (1971). Mode of discourse variation in theevaluation of children's writing. Research in the Teaching ofEnglish, 5, 37-45.Vygostsky, L. S. (1962). Thought and language (E. Haufman & G.Vabor, Eds. and Trans.) Cambridge, Mass., M. I. T. Press, 1962.(Originally published 1934.)Watson, C. (1983). Syntactic change: writing development and therhetorical context. In M. Martlew (Ed.), The psychology of writtenlanguage: developmental and educational perspectives (pp.127-140). Chichester: John Wiley& Sons.Watts, A. F. (1944). The language and mental development ofchildren: an essay in educational psychology. London: George G.Harrap.Wilkinson, A., Barnsley, G., Hanna, P. & Swan, M. (1980). Assessing language development. London: Oxford University Press.Wilkinson, A., Barnsley, G., Hanna, P. & Swan, M. (1983). Towards acomprehensive model of writing development. In Kroll, B. & Wells,G. Explorations in the development of writing (pp. 43-70).Chichester: John Wiley and Sons.284Witte, S. & Davis, A. S. (1980). The Stability of T-Unit Length: APreliminary Investigation. Research in the Teaching of English, 14,5-17.White, J. (1989) Children's argumentative writing: a reappraisal ofdifficulties. In F. Christie (Ed.). Writing in schools: Reader(ECT418) (pp. 9-23). Geelong, Victoria: Deakin University Press.Yau, M. S. S. & Belanger, J. (1985). Syntactic development in thewriting of EFL students. English Quarterly, 18, 107-118.APPENDICES285286APPENDIX ASEGMENTATION RULES 181. A T-unit consists of one independent clause with all thesubordinate clauses attached to it.2. Mark the end of a T-unit with a double stroke (//); mark witha single stroke (/) any subordinate clause which falls withinthe T-unit.e.g. As Tom got dressed for the show that afternoon/ he feltnervous.// The feeling was not unusual / because there'ssomething about working with a killer whale / thatmakes you nervous.2 T-units, 5 clauses3. Ignore mispunctuation in analyzing into T-units.e.g. Just think/ how much fun we could have stayingback here. Begin obnoxious and having a ball.//4. Eliminate (i.e., strike out):a. garbles, i.e., unattached sentence fragments andunintelligibe word strings;b. interjections (Hey! Hi, Jane!) c. tag questions (won't you, isn't he) ;d. parenthetical expressions (I guess, I think, You see) ,e.g. It's a kind of skinny tree, I guess.And there's a bush, I think, about fiveor ten feet from the tree. You see, I like him.5.^Retain:a. fillers like now, well;18^Source: Crowhurst, M. The effect of audience and mode of discourse onthe syntactic complexity of the writing of sixth and tenth graders(Doctoral dissertation, University of Minnesota, 1977).^Dissertation Abstracts International, 1978, 387300A -7301A, pp. 133-135.287b. exclamatory words that introduce a longerexpression,e.g. Boy, was he surprised.6. Consider contractions (she's, they're) as two words.7. Count compound words as one word if normally written as oneword (bedroom, breakfast), as two words if hyphenated (fox-catcher).8. Count as one word dates (October 1) and proper namesconsisting of more than one word (New York, Clear Lake National Park); but Tuesday, October L 1975 = 3 words; St.Paul, Minnesota = 2 words.9. Count numbers as one word whether written in digital form(171) or in words (one hundred and seventy-one).10. Treat so as either a coordinate or a subordinate conjunctionaccording to context. If so is equivalent to "in order that", treatit as a subordinate conjuction; otherwise it will be equivalentto "and so" or "and therefore" and is to be treated as acoordinate conjunction.11. Treat for as a coordinate conjunction whether it occurs at thebeginning of a sentence or between two clauses within asentence.e.g. Tom took one step at a time going up the platform,//for he was still a bit afraid of heights.//^12.^Analyze direct discourse as follows:a. Discard syntactically incomplete expressions (e.g.,answers to questions which lack the repetition of thequestion elements), and one- or two-word answers toquestions (Yes, all right) unless they occurbefore/after he said or unless they introduce a longexpression (e.g., All Right, let's go).Treat as a direct object the first expressionbefore/after he said; this is to be done whether thatexpression is a sentence or some smaller fragment.2 8 8b. Analyze subsequent words in the direct discourse intoT -units according to regular rules.e.g. John said,/ "I really like Minneapolis.// ButChicago is my home// and most of myfriends are there." //"All right," I said happily.//"Well Sara," George drawled, / "you take thecanoe."//13. Supply any single word (or two words contracted) accidentallyomitted, and count in the total.14. Count the number of words in each T-unit and write thenumber above each double stroke.e.g. Tom walked into the store and bought an icecream.//APPENDIX B289290APPENDIX BBEGINNING AND ENDING STATEMENTSCLASSIFICATION SYSTEMBeginnings - Narrative and Report1. Structural Beginnings - structural beginnings provide both aclear indication of the genre of the piece and are integratedinto the overall structure. They are the most mature formof beginning statements. The following are definitions andexamples of structural beginnings in narrative and report:a. Setting - a setting statement introduces a narrative bymentioning one or more aspects of the place andtime in which the story takes place. Settingstatements are not found in report writing.e.g. "The stars shone brightly over the littletown.""It was a dark and cloudy night."b. Action/Dialogue - a statement of direct discoursewhich opens the action in a story. These types ofstatements are not found in report writing.e.g. "AHH, down falls another foe!""Cut," said the director as he turnedtoward me.c. Thesis statement - a statement which introduces thetopic and is integrated into the entire piece ofwriting. There were no thesis statements found innarrative and few in report writing.e.g. "The F-6 Fighting Falcon is the ultimatefighting aircraft."d. Beginning of Action - used to begin a narrative ratherthan a report; a statement about a character doingsomething to begin the action but without dialogue.291e.g. "Lisa looked longingly out the carwindow."2. Formulaic Beginnings - statements used in narrative andreport which are based on traditional beginnings and useconventional phrasings.a. Formulaic - used in narrative writing rather thanreport writing.e.g. "Once upon a time..."b. This is a story/report about - a statement whichsimply tells the reader about what is to come; rarelyused in narrative.e.g. "My report is about Nintendo.""This is a report about dogs."c. Topic starters - statements which introduce the topicbut are not a generalized fact about the topic; used inreport writing.e.g. "At Disneyland there is a ride called thePirates of the Carribean."3. Weak Beginnings - statements which require the reader toread further to determine the genre. Weak beginnings donot establish an introduction to the story or topic.a. Dialogue - statements which are directed to thereader in the form of a conversation. Statementssuch as these are found in both narrative and report.e.g. "My name is Susan and I'm here to tellyou about my move to a new school.""Well, what do you know about MichealJordan?"b. Evaluation - statements which express an opinion inthe form of an evaluation of the topic; rarely used innarrative.292e.g. "Pekinese are cute dogs.""Indy cars are special race cars."c. Hello - direct conversation addressed to the reader,usually in the form of an introduction to the author;used in both narrative and report writing.e.g. "Hi, my name is Kevin."d. Generalized Opening - a statement used in reportwriting which uses a general fact as an openingstatement; the fact is unrelated to the statementsthat follow it.e.g.^"Bassets are dogs.""In volleyball you need a net to play.""The piano is a large percussioninstrument."Endings - Narrative and Report1. Structural Endings - provide a sense of closure and areintegrated into the whole piece of writing.a. Solution/Resolution - provides a solution to theproblem presented in a narrative and resolves andconcludes the action; not found in report writing.e.g. "The police in the other states caught thehunters and gave them all a super bigFine."b. Natural End - a statement which winds down theaction in a narrative.e.g. "By the end of the day he had mademany friends.""The only thing that was left was thehead of a man."c. Evaluation - a statement which evaluates the narrativeor some character in the narrative or evaluates the293topic in report writing. No evaluative statementswere found in the narratives of this subjectpopulation.e.g. "I really like dancing.""And I personally give it thumbs up!"d. Moral - a statement which provides some lesson ormessage to the reader; used in both narrative andreport writing.e.g. "Remember never give up on a dream.""Remember be friendly to animals orelse they will go endangered.""Orako Saki used to be good but thelove of the sword made him evil."2. Formulaic Endings - statements which use conventionalphrasings to provide closure to narrative and reports.a. Formulaic - used in narrative writing onlye.g. "And they lived happily ever after."b. The End - these two words often accompanied awinding down of action in narrative writing or wereused after a last fact in report writing.e.g. "Horses have long tails." "The End."c. That's how you do it - a statement which restates theproblem presented in a report. There were none ofthese statements used in this subject population'swriting.3. Weak Endings - statements which do not signal closing; thereader is left without any sense of closure.a. Last fact - statements which add an extra piece ofinformation to the piece but do not relate to previousinformation; last facts are found in report writing.294e.g. "When I grow up I want to be aprofessional.""And whenever she sees little kids shewill bite them."b. No clear ending - statements which seem to carry onthe topice.g. "I like the Flames, Canucks, Kings, .. ."I like working and playing on thecomputer."c. End to begin again - statements which are used towrite chapter such as in narratives; rarely found inreport writing.e.g. "Where will Fluffy go next?"d. That's all I know - a statement which tells the readerthat the author has told all the information heknows; used inreport writing.e.g. "And that's how I take care of her.""That's about it."APPENDIX C295296APPENDIX CSTUDENT TEST SCORES AND LETI E^R GRADESStandardized test scores and letter grades for the students.Writing^Writing^WritingI.D. # Gates McGinitie C.T.B.S. letter grade letter grade letter gradeMay '88 (G.E.) May '89^May '90^May '91^May '92#^1 3.3 C C C+# 2 3.6 B A B# 3 5.3 B B A# 4 - 4.8 A B C+# 5 3.8 C+ B B# 6 4.9 C+ B C+# 7 6.0 A B B# 8 2.7 - C+ C C+# 9 3.7 - B B C+# 10 3.5 - C C+ C+#1 1 6.0 - A A A#1 2 3.5 C+ C+ B#13 - 3.4 B B C# 14 3.4 C+ B C+APPENDIX D297Shirley A. Tb6mpsonCoordinator,Ethical Review of ResearchInvolving Human SubjectsTHE UNIVERSITY OF BRITISH COLUMBIARoom 331, IRC Building2194 Health Sciences MallVANCOUVER, B.C., CANADAV6T 1W5OFFICE OF RESEARCH SERVICESOUR FILE: B89-350Dec 4,1989Dr. M. CrowhurstJ. MartinLanguage EducCampusRE: Your proposed study: A longitudinal study of the development ofsyntax and structure in childrens narrative and report writingDear Dr. Crowhurst/J. Martin,The Behavioural Sciences Screening Committee for Research & Other StudiesInvolving Human Subjects has reviewed the protocol for your proposed researchproject. The Committee found the procedures to be ethically acceptable and a' Certificate of Approval will be issued upon the Committee's receipt of writtenagency consent from the Richmond School District.If you have any questions, please call me at 224-8584 or Dr. R. Johnston(Chairman of the Behavioural Sciences Screening Committee) at 228-5456.298299SCHOOL DISTRICT No.38 RICHMOND)7811 GRANVILLE AVENUE / RICHMOND / B.C. / V6Y 3E3 / (604) 668-6000Office of the Superintendent of Schools1990.02.01Ms. Jennifer Martin2250 W. 35th AvenueVancouver, B.C. V6M 1J5Dear Ms. Martin:Thank you for your letter of January 29th outlining your desire to conduct a longitudinal study onthe development of syntax and structure in children's narrative and report writing.All research activities in the Richmond School District must be approved at the Board level beforethey can proceed. There is a standard application and review process in place for such requests.However, since you are a member of our staff and since you have already received the consent ofMr. Holman and the parents involved I do not think that it will be necessary for you to completethis form. In fact, based upon a review of the materials which you have submitted I am pleased toprovide you with permission to proceed. The involvement of any particular school or child in thisresearch in either the current or future years is, of course, entirely at their own discretion. Districtpermission simply authorizes you to make the request.Next year and in the final year of your study I would appreciate receiving a brief letter outliningyour intent to continue and explaining any changes in the research program which you may havemade. If a different parental consent form is used I would also appreciate receiving a copy of thatform. When your study is complete, you should submit to my attention a copy of the final reporton your findings.This letter should be a satisfactory indication of consent for the Ethical Review Board.If I can be of any further assistance in the conduct of your research please do not hesitate to contactme.Yours truly,J. .B. BeairstoSupervisor of CurriculumJABB/swc.c. Bob Holman, Principal, Thompson Elementary SchoolOUR FOCUS IS ON THE LEARNER"Best of luck in your work.^Jud' h A. Langeressor300School of Education Education B9Albany, NY 12222518/442-5026Fax: 518/442-5933UNIVERSITY AT ALBANY STATE UNIVERSITY OF NEW YORKNATIONAL RESEARCH CENTER ON LITERATURE TEACHING & LEARNINGApril 30, 1993Mrs. Jennifer Martin2258 West 35th AvenueVancouver, B.C.V6M 1J5Dear Mrs. Martin,I enjoyed receiving your letter and learning of your research.Of course I think you are studying an important topic.Unfortunately, I've moved across the United States since ChildrenReading & Writing was written, and the original oral and writtenlanguage samples and coding sheets you've requested were discardedat some point during the move. I'm sorry. However, negotiatingthe category placements of the beginnings and ends with a colleagueshould be a perfectly viable alternative for you.I have enclosed a more recent paper that was an offshoot of myearlier work--my way of begging your forgiveness.enc.