• 検索結果がありません。

及ぼす影響に関する基礎研究 及ぼす影響に関する基礎研究

N/A
N/A
Protected

Academic year: 2021

シェア "及ぼす影響に関する基礎研究 及ぼす影響に関する基礎研究 "

Copied!
290
0
0

読み込み中.... (全文を見る)

全文

(1)

博士学位論文(東京外国語大学)

Doctoral Thesis (Tokyo University of Foreign Studies)

氏 名 松崎 武志 学位の種類 博士(学術)

学位記番号 博甲第214号 学位授与の日付 2016427日 学位授与大学 東京外国語大学

博士学位論文題目 会話ダイアログ暗唱に従事させる外国語指導法がスピーキング時の定 型表現の使用と暗記学習に及ぼす影響に関する基礎研究

Name Matsuzaki, Takeshi

Name of Degree Doctor of Philosophy (Humanities) Degree Number Ko-no. 214

Date April 27, 2016

Grantor Tokyo University of Foreign Studies, JAPAN Title of Doctoral

Thesis

Research into the Role of Dialog Recitation in the Foreign Language Classroom—Its Effectiveness in Facilitating Memorization and Formulaic Speech Production

(2)

Research into the Ro le of D ia log Rec itat ion in the Fore ign Language C lassroom

—Its Effect iveness in Fac i l itat ing Memor izat ion and Formu la ic Speech Product ion

会話ダイアログ暗唱に従事させる外国語指導法が 会話ダイアログ暗唱に従事させる外国語指導法が

スピーキング時の定型表現の使用と暗記学習に スピーキング時の定型表現の使用と暗記学習に

及ぼす影響に関する基礎研究 及ぼす影響に関する基礎研究

TAKESHI MATSUZAKI

(3)

Acknowledgements

All of what I havelearned at each stage ofthis dissertation writing will serve me,inthe rest of my professional and personallife, as aninvaluable and enriching resource. The work ofthe past four years has beenthe biggest challenge of mylife, and I have no doubtthat I would never have completedthis work withoutthe support of a great many people.

I am especially fortunateto have been ableto benefit fromthe experience and sage advice ofthe members of my dissertation committee, without whose guidance and support at variouslevelsthis project could never have been realized. My supervisor, Professor Hideyuki Takashima, has been aninsightful, resourceful, and supportive mentor. I am deeplythankful to him for his guidance. Two other primary committee members, Professors Masashi Negishi and Asako Yoshitomi, were of enormous helpinthe design of this research. The remaining two members,Professors Keiko Mochizuki and Mark Petersen, have also provided

tremendous supportinthe preparation ofthis dissertation.

My specialthanks also gotothreeindividuals. Forthe statistical analyses employedin this study, I am hugelyindebtedto Yutaka Iguchi, Director atthe Laboratory of Biology. He taught me so much about statistics, especially with regardtotheimplementation of

non-parametric methods. For polishingthelanguage usedinthetext, I cannotthank enough Kevin Mark, a senior professor of Meiji University. An expertin helping othersto refine and polishtheir writing, Kev generously devoted a great deal of his personaltimetothe writing stage ofthis research. Very specialthanks are dueto mylong-standing colleague and close personal friend, Joseph Heilman. Joe was alwaysthere for me, quiteliterally, when I needed to call on his expertise or psychological support.

I must alsothankthosein “Team Takashima” who gave me a great deal of sage advice inthe preparation ofthis dissertation. Most notably, Noriko Imai and Naoyuki Kiryu

(4)

ii

generously offered me countless suggestions, many of which I followed.

While I do owe all ofthese people for completingthis dissertation, I should emphasize thatthey do not necessarily endorsethe views expressed or claims madeinthis dissertation, andit goes without sayingthat anyinaccuracies or misunderstandingsinit are entirely my responsibility.

I am also gratefultothose students at Meiji University who participatedinthe pilot and main studies ofthis research. Withouttheir generous participation,the completion ofthis dissertation would havetaken muchlonger. As for financial assistance,this research was partly funded by MEXT/JSPS Grant-in-Aid for Young Scientists (B) 24720271.

Lastly, I should express my greatestthanksto my family fortheir moral support, patience andlove. Although he was not actually a member of my family, I wouldliketo include here Philip Zitowitz, my mentor since my undergraduate days who passed awayin the summer of 2014. Phil pushed me hardto embark onthis doctoral work, and his

never-failing kindness and encouragement could only have been stopped by his untimely death. My parents, Kayoko and Fumio, my sister, Ryoko, and my aunts, Katsuko and Masae, have been a source of my motivationto undertakethis project, and I would not have hadthe courage and determinationto completethe work hadit not been for my wife, Karin, and my son, Leo. Itisto my familythatthis dissertationis dedicated.

Takeshi Matsuzaki February 2016

(5)

iii Abstract

Thereis,in SLA and cognitive psychology, strong evidencethat Formulaic Sequences (FSs) are storedin memory asindependent units. This study aimsto make a contributionto the understanding of FSsin L2learning andtothe potential effectiveness of memorization of FSs as ateaching/learning strategy. It reports on a projectin whichlearners were giventhe task, overthe course of a semester, of memorizing and reciting dialogs which had been written so astoinclude FSsthat arelikelyto be usefultolearners preparingto study abroad. A formulaic sequence can be defined as a string oflinguistic itemswherethe relation of eachitemtothe rest is relatively fixed, and wherethe substitutability of oneconstituent ofthe sequence by another ofthe same category is relatively constrained (Wray & Perkins, 2000). The phenomenon of FSs has been ofinterestto appliedlinguists for sometime (e.g., Bolinger, 1976; Fillmore, 1979; Pawley & Syder, 1983; Wray, 2002), andis atthe core of corpus linguistics (e.g., Sinclair, 1991) and various pedagogical approachesthat can be called

“lexical” (Lewis, 1993; Nattinger & DeCarrico, 1992; Willis, 1990). One reason whythey have attracted so much attention, and also have been difficultto define precisely,isthe fact thattheyaboundinlanguage use. Collectivelythey make up a substantial and vital part of a person’slexicon, and perform an essential rolein facilitatingthe understanding and

expression of messagesthat could otherwise be misinterpreted. Itistheir pervasivenessinthe languagethat makesthem animportanttarget forlanguage pedagogy. Thusthe question

“How canthey usefully betaught?”is of wide andlastinginterestinthe field ofinstructed language acquisition and pedagogy.

Chapter 1 discussesthe background ofthe present study,touched on above, and specifiesits focus astheinvestigation oftext memorization approachestotheteaching of formulaic sequencesin a foreignlanguage (FL) context, specificallythat of a Japanese university.

(6)

iv

Chapter 2 examinesthe mostimportant examples, forthis study, ofthe wide range of characteristics of FSsthat are foundintheliterature on formulaic sequences. Thisis followed by a characterization ofthe features of formulaic sequences appropriate forthe present study. The chapter alsolooks at constructs of formulaic sequences, and endeavorsto offer a more thorough account of how we process and acquire “chunks” oflanguage. In concluding,the chapter establishes connections between formulaic sequences andlanguagelearning, especiallyin an FL environment.

Chapter 3 addressesthe waysin which we process and retainlinguisticinformation. The chapter begins by examiningthe different models forthe processing and production of

language. The chapterthen moves onto discuss awareness and cognitionin relationto memory. Thisis followed by furthertreatment oflanguage processingin connectionto memory, andits related components. Particular emphasisis givento howtheseissues relate tothe experiences of L2learnersin a foreignlanguagelearning environment,takinginto accountthelimited opportunitiesto acquirethelanguagein a ‘natural’ way based on accumulated experience of authenticinteractionsinthetargetlanguage. The chapter concludes byintroducing anintegrated model forlanguage processing and acquisition in relationto memory, and outlinesthe features ofthis model.

Chapter 4 aimstointroducethe backgroundtothe study conducted forthis dissertation, with a discussion ofthree fundamental questionsthat were brought upinthe review and synthesis offeredinthe preceding chapters. Thethree central questions are (1) Should teaching practitionersfocus onformulaic sequencesinteachingin aforeignlanguage context withthe particulartarget population being adultlearners?, (2)Ifthey should, which formulaic sequences shouldtheyteach?, and (3)How shouldtheyteachthetargeteditems?

The first questionis addressed by providing a number of reasons fortheteaching of

formulaiclanguagetothis particulartarget population. The second andthird questions are addressed by drawing on principles and proposals discussedinthe preceding chapters. This

(7)

v

chapterthen presents a review of studies ontext memorizationthat helpto clarifythe rationale forthe present study.

Chapter 5 poses five research questions. Research Question 1 asksif ‘whole-text’ and

‘partial-text’ recitation of alarge volume of useful dialogs, preparedin advance ofinstruction, engages foreignlanguage classroomlearnersin memorization overthe course of one

semester,includingthe further, embedded question of whetherthereis a significant

difference betweenthetwointerms oftheir facilitative effect. Research Question 2isto do with whetherthe ‘wholetext’ and ‘partialtext’ dialog recitation specifiedin RQ1 facilitates formulaic speech production, and again asksifthereis a significant difference betweenthe twointheir facilitative effect. Research Question 3 asksif engaging foreignlanguage classroomlearnersinthe ‘wholetext’ and ‘partialtext’ dialog recitation specifiedin RQ1 facilitates speech fluency as measured by syllables spoken per minute. Again,the question of whetherthereis a significant difference betweenthetwotypes of memorizationis considered as part ofthe question. Research Question 4 asks whetherthese activities favorably affectthe attitude of studentstowardtext memorization as a meansto developtheir oral communication skills, and also compares whole- and partial-text approaches. Finally,the 5th Research Question, as a way of supplementing and enrichingthe data achieved withthe first four Research Questions, seeksto see what variables may have been at playthat can explainthe differencesinthe performance of high andlow achievers.

Chapter 5then goes onto set outthe rationale forthe research study andto describe how it was conducted.Atotal of 35 university studentsin Japan, dividedintothree groups

(Treatment Group 1 [TG1]: n = 12; Treatment Group 2 [TG2]: n = 12; Contrast Group [CG]: n = 11), participatedinthis study. A substantial set of model dialogs (3,182 wordsintotal) was prepared forthis study, withthe key feature beingthat each dialog contained many FSs that will be particularly useful when studying abroad. Usingthis material, TG1 and TG2, taught bythe researcher, spent athird or more of each 90-minute classtime on 1)the

(8)

vi

researcher providing formalinstruction on a set number of dialogs, 2)the students memorizing and reviewing a given number of dialogs, and 3)the students checking each other onthe dialogsthatthey had memorized. The key differencein howthetwo courses weretaught wasthat whilethe studentsin TG1 wereinstructedto memorizethe dialogs completely,thosein TG2 were onlyinstructedto memorize parts ofthe dialogs with particular focus onthose FSs. Atthe onset ofthe semester,the participantstook a speaking test containing a few quasi-interview questions and filled out a questionnaire, and atthe end ofthe semestertheytook another speakingtest with quasi-interview questions andanother questionnaire. The sametests and questionnaires were also administeredtothe CG, also taught bythe researcher.

Chapter 6 presentsthe results ofthe speakingtests and questionnaires in numerical and graphicterms.First, forthe memorization ofthe dialogs bythe TGs, both groups were found to have beeneffectively engagedinthetask. Second, for Part 1 ofthe speakingtest

(‘reading-aloud short sentences’), both TGs demonstrated significantly higher improvements than CG, and TG1 even outperformed TG2. Third, regarding Part 2 ofthetest (‘short

translations or directed responses’),while both TGs made significantimprovements

comparedto CG in‘direct application’ ofthe dialogs studied,it was TG1 alonethat showed a significantincreaseinthe‘appropriateness’ ofthe responses. Regarding Part 3 ofthetest (‘extensive oral production’), onthe other hand, it wasCGthat was foundto have displayed a significantincreaseinthe use of FSs availableinthe dialogtextbook. Interms of fluency of responses (as measured by syllables per minute) in Part 3,however, TG2 wasthe only group showing a significant advancement. As forthe attitudinalitems usedinboth Pre- and

Post-Questionnaires, no significant variance was found with any group. Lastly,in regardto thereflectiveitems usedin thePost-Questionnaire, several significant differences were found, themost notable one regarding ‘favorable changein attitudetowardtext memorization as a

(9)

vii

waytolearn a variety of features.’ Inthis case,TG1’s score was significantly higherthan those ofTG2 and CG.

Chapter 7 discussesthe findings in detail. With respectto Research Question 1,the resultsindicatethat bothtypes of classroomintervention were effectivein engagingthe learnersin memorization overthe course of one semester. The same could be said for RQ 2, but the results onthe wholesuggest that whole-text memorization facilitatesformulaic speech production morethan partial-text memorization.Whilethetest resultsindicate an advantage for partial-text memorization for ‘direct application,’whole-text memorization appears more effective except inthe case of‘modified application.’ The use of an additional n-gram analysis also shows, whilelimitations should be keptinmind, a significant

improvementin‘extensive oral production’ made only by TG1. An advantage, although a weak one, was found for TG1 onthelevel of appropriateness of production. For

improvement of pronunciation,the results stronglyindicatethat whole-text recitationis more effective. With regardto RQ 3,the partial-text memorization groupshowed a significant increaseinthe number of syllables spoken per minute. What should be bornein mind, however,isthe possibilitythatthe whole-text memorization group may have been

unintentionally invitedto pay more attentionto details atthe expense of fluency.As for RQ4, TG1’s attitudetowardtext memorization became more positive, which suggests another advantage of adopting a whole-text memorization approach.Lastly, for RQ 5, an analysis of high andlow achievers ofthe speakingtests with reference totheir responsestothe

quasi-interview questionsindicate a number of other variablespotentially affecting their performanceinthetests and responses tothe questionnaireitems.Overallthe study raised manyinteresting questions, and implications forteaching andareas for further research are discussed.

(10)

viii

Formulaic Sequences(定型的な単語[あるいは形態素]の連続:以降 FSs)は我々 の記憶内に独立して保持されている。本研究は、第二言語習得における FSsの役割、

および指導・学習ストラテジーとしての FSs暗記の効果を調査する。具体的には、

留学中に役立つことが期待される FSsを含むダイアログ集の暗記と暗唱タスクを一 学期間に渡って与えた学習者を研究対象としたプロジェクトを報告する。

FSsとは、一続きの言語情報を指すわけだが、中でも「各構成要素と残りの要素と の関係が比較的に固定化されており、かつ、連続体の一部を同一カテゴリーの別のも のに置き換える際に比較的に制限のある連続」と定義することできる(Wray & Perkins, 2000)。FSs 現象は、長期におよぶ応用言語学の関心事であり(e.g., Bolinger, 1976; Fillmore, 1979; Pawley & Syder, 1983; Wray, 2002)、コーパス言語学の核であり(e.g., Sinclair, 1991)、そして多様な“レキシカル”指導アプローチの中心である(Lewis, 1993; Nattinger & DeCarrico, 1992; Willis, 1990)。FSsが大きな関心を引き寄せる理由(そし て、その厳密な定義付けが難しい理由のひとつ)は、FSsが語彙の相当部分を占め、

メッセージの理解と発信において最重要な役割を担っていることにある。このため、

「FSsをいかにして教えるべきか?」という問いは言語教授法において大きな関心事 となっている。

本論文の第1章では、論文内で報告する研究の背景を議論する。その際、研究フォ ーカスが(特に日本の大学という)外国語環境における文章暗記(text memorization)

による FSs指導であることを述べる。

第2章では、まず、関連文献で議論されている多様な FSsの特徴を検討する。本章 はまた、FSsの概念を検討し、人がいかに言語チャンクを処理し習得するのかについ て詳細に述べる。本章の最後では、FSsと言語学習の関係性を説明し、外国語環境で の学習におけるその関係性の示唆にも触れる。

第3章は、言語情報の処理と保持について議論する。本章はまず、言語理解と発信 に関する理論モデルを検討する。次に、認知科学の観点から見た情報処理、学習につ いて議論する。その後、第二言語における FSsの習得を困難にする言語特性、そして、

(11)

ix

思春期以降の学習者が外国語環境で学習する際の制約について議論を展開する。本章 の最後では、それまでの議論を踏まえた言語処理・習得の統合的モデルを提示し、こ のモデルの含意を説明する。

第4章は、次章以降で報告していく本研究の導入として、まず、前章までの議論を 踏まえた3つの問題点について議論する。3つの問題とは、すなわち、【1】「特に大 人を対象とした外国語環境での指導において FSsにフォーカスした指導をすべき か?」、【2】「フォーカスすべきである場合、どのような FSsを教えるべきか?」、そ して【3】「どうやって FSsを教えるべきか?」である。これら3つの問題に対して、

前章までに議論してきた様々な理論モデルに鑑みながら、それぞれ、答えを多角的に 検討していく。とりわけ問題3については、多様な指導アプローチの効果や注意点を 議論していく。本章の最後では、本研究のテーマとなっている文章暗記の指導効果に 関する主要な先行研究のレビューを行う。このレビューでは、以下の点を指摘する。

まず、FSs使用はオーラル面において最重要であり FSsの即時的な使用を適切に計測 するためにはスピーキングによる能力測定が不可欠と思われるにも関わらず、これま での実証研究にはスピーキングテストによる効果検証が欠如している。第二に、特定 の暗記指導法の認知負荷に差を設定した複数処置群についての有意差研究、そして

(一学期間という)比較的長期の暗記指導を施した場合の効果研究も不足している。

さらに、特定の暗記指導アプローチを実施することで学習者の暗記への態度がどう変 わるかについての検証がなされてきていない。

第5章では、まず、第4章の最後で述べた文章暗記に関する先行研究の不足点を踏 まえ、本研究における研究課題を次のとおり挙げる。【研究課題1】指導開始前に準 備しておいたダイアログ集の全文暗唱あるいは部分暗唱を促すことにより、一学期間、

暗記作業に従事させることができるか、また両方法の間に有意差は見られるか。【研 究課題2】課題1で示した全文暗唱と部分暗唱指導により、FSsを用いたスピーチ産 出を促すことができるか、そして両方法の間に有意差は見られるか。【研究課題3】

課題1で示した両方法により1分間当たりの産出シラブル数で計測するスピーチ流

(12)

x

暢さを向上させることができるか、そして両方法の間に有意差は見られるか。【研究 課題4】両方法により、コミュニケーション力を伸ばすために文章暗記をすることに 対する学習者の態度を(より)肯定的に変化させることができるか、また両方法の間 に有意差は見られるか。【研究課題5】両方法がどの程度機能するかについて、どう いった個人差が影響を及ぼすか。

第5章では、次に、研究方法の詳細を記述する。まず、研究協力者については、処 置群1(TG1)が12名、処置群2(TG2)も12名、対照群(CG)が11名の計3 5名が本研究に参加した。本研究で用いる FSsについては、事前に、語数合計が 3,182 となるダイアログ集を準備しておいた。この教材を用い、TG1そして TG2の指導を 研究者が行った。各回の授業は、90分授業の最低3分の1以上の時間を次の活動に 充てた。まず、研究者がダイアログ集から順番にいくつか、文法、語彙、発音等の指 導を行った。続いて、各学生には、導入済のダイアログの中からいくつか選択したも のをその場で覚えるか、あるいは予習で覚えてきたものの再確認をさせた。次に、学 生同士で暗唱をさせた。相手がいない学生には研究者が暗唱パートナーとなった。

TG1と TG2の指導法における最大の違いは、前者にはダイアログの全文暗唱を課し、

後者には特定の FSs箇所のみ覚えれば済む部分暗唱を課したことである。このため、

部分暗唱群は、全文暗唱群と比較して、約3分の1に相当するテキストを暗唱するこ とになった。指導開始時と終了時には、英語のスピーキングテストを実施し、日本語 による疑似インタビューと選択肢形式のアンケート回答による調査も行った。CGに 対しても、同じテストを実施し、同じインタビュー・アンケートへの回答をさせた。

第6章は、処置群によるダイアログ暗記、および全群のスピーキングテスト・アン ケート回答の統計データを提示する。まず、処置群によるダイアログ暗記については、

TG1、TG2いずれも効果的になされたことがわかった。スピーキングテスト Part 1(短 文音読タスク)では、両処置群が CGに対して有意なスコア向上を見せ、TG1はさら に TG2に対しても有意な伸びを示した。スピーキングテスト Part 2(短文翻訳・回答 タスク)では、学習したダイアログ表現をそのまま再現するプロンプトでは両処置群

(13)

xi

が CGに対して有意なスコア向上を見せたのに対して、回答の適切さについては TG1 にのみ有意なスコア向上が見られた。スピーキングテスト Part 3(長文回答タスク)

では、ダイアログ集にある FSsの使用について CGのみ有意なスコア向上が見られ、

1分当たりの産出シラブル数で計測する流暢さについては TG2のみ有意なパフォー マンスの向上が見られた。指導開始時と終了時に設定したアンケート項目については、

いずれも有意な変化は見られなかった。最後に、指導終了時にのみ設定したアンケー ト項目については、「様々な言語項目を学習するためにテキスト暗記学習をすること への好意的な態度変化」に関する項目において TG1のスコアが TG2、CGを有意に上 回るなど、いくつかの項目において群間の有意なスコア差が見られた。

第7章では、各研究課題に関する発見について議論する。まず、課題1については、

全文暗唱、部分暗唱いずれも一学期間に及んで暗記作業に従事させられることがわか った。課題2については、覚えた FSsをそのままスピーキングで用いることについて は部分暗唱群の方が全文暗唱群よりも有意な伸びを見せたが、他のテスト結果も含め て総合的に解釈すると、全文暗唱の方が部分暗唱よりもさらに効果的に定型表現によ るスピーチ産出を促すと考えられる。追加で実施した n-gram分析では、解釈におけ る注意点はあるものの、全文暗唱群のみ「長文回答タスク」において有意なスコア向 上が見られた。スピーキング回答内容の適切さにおいても、限定的ではあるものの、

全文暗唱群にのみ有意なスコア向上が見られた。そして、発音問題のスコアに関して は、両処置群ともに有意な向上が見られたが、全文暗唱は部分暗唱と比べても有意な 伸びが見られたのである。課題3に関しては、1分当たりの産出シラブル数で測る流 暢さについて部分暗唱群に有意なスコア向上が見られた。ただし、全文暗唱群は、流 暢さを犠牲にして産出内容に注力を傾けていた可能性がある。課題4については、全 文暗唱のみ、アンケート回答において、文章暗記に対する肯定的な意識変化が有意に 見られた。このことは、全文暗唱アプローチを採用する動機となりうる。最後に課題 5については、スピーキングテストの高得点者と低得点者を各群から3名ずつ選び、

彼らのテスト結果およびアンケート回答と疑似インタビュー回答に言及しながら1

(14)

xii

名ずつ分析を試みたが、様々な個人差がテスト結果、アンケート回答に影響を及ぼし ていた可能性がある。例えば、本研究中に授業外で海外から来ている留学生と交流の 多かった研究協力者は、顕著にダイアログ集の FSs使用のテストスコアが伸びていた。

別の例としては、コミュニケーションでよく使われる表現が便利であることはわかっ ていつつも、それらを覚えることに対して様々な理由で抵抗感を感じている学生は、

暗唱はしても、事後テストにおける FSs使用の大きな増加は見られなかった。

第7章の結論部分では、本研究におけるデザイン上の問題をいくつか指摘し、それ らを考慮しても、本研究は先行研究で扱われていない研究課題を扱っており、したが って、FSs指導に関心のある言語教師、そして FSsを研究テーマとしている研究者に とって新たな知見を提示することが期待される旨を述べる。最後に、本研究でも未着 手の FSs研究領域、そして、本研究のデータ分析過程で新たに見つかった FSs指導の 研究領域について指摘して終える。

(15)

xiii

Table of Contents 1

Acknowledgements ...i

Abstract ...iii List of Tables ...xix List of Figures ...xxii Chapter 1: Introduction ...1

1.1 Background ...1

1.2 Research aims andthe organization ofthis dissertation ...5

Chapter 2: Characteristics of formulaic sequences ...8

2.1 Introduction ...8

2.2 Defining formulaic sequences ...8

2.3 Characteristics of formulaic sequences ...10

2.3.1 Formal mutability of formulaic sequences ...10

2.3.2 Semantictransparency of formulaic sequences ...13

2.3.3 Pragmatic functions of formulaic sequences ...14

2.3.4 Summary ...16

2.4 Knowledge and processing oflanguage ...17

2.4.1 Formulaic nature of ourlanguage knowledge ...17

2.4.2 Predominant reliance on formulaiclanguage...18

2.4.3 Dual system and processing model ...19

2.5 Acquisition of formulaic sequences ...21

2.6 Pattern-basedlanguage acquisition ...23

Chapter 3: Language processing andlearning ...25

(16)

xiv

3.1 Introduction ...25

3.2 Three knowledge sources forlanguage comprehension and production ...25

3.3 Noticing, attention, and awareness ...27

3.4 Working memory ...28

3.5 Transfer-appropriate processing ...30

3.6 Inherent difficulties of alinguistic feature ...31

3.6.1 Frequency ...31

3.6.2 Perceptual saliency ...32

3.6.3 Communicativeload ...32

3.6.4 Form-meaning-function complexity ...33

3.6.5 Grammatical reliability...33

3.6.6 Complexity onthelevel ofindividual words ...34

3.7 The foreignlanguage context ...35

3.8 Language knowledge, processing andlearning: anintegrated model ...36

Chapter 4: Teaching of formulaic sequences ...40

4.1 Introduction ...40

4.2 Rationale for focusing on FSsinteaching ...40

4.3 What formulaic sequences should betaught? ...41

4.4 How should formulaic sequences betaught? ...43

4.4.1 Input enhancement ...43

4.4.2 Chunkingintext comprehension ...44

4.4.3 Exercise of schematic and contextual knowledge ...45

4.4.4 Teaching grammarto facilitate formulaiclearning ...46

4.4.5 Use ofinteractionally non-demandingtasks ...47

4.4.6 Text memorization...48

4.4.7 Final comments on howtoteach formulaic sequences ...50

(17)

xv

4.5 Review of studies ontext memorization ...51

4.5.1 Wray’s (2004) study ...52

4.5.2 Wray and Fitzpatrick’s (2008, 2010; Fitzpatrick & Wray, 2006) study ...52

4.5.3 Yu’s (2009) study ...53

4.5.4 Dai and Ding’s (2010) study ...54

Chapter 5: Methodology ...56

5.1 Introduction ...56

5.2 Research questions ...57

5.3 Pilottesting ...58

5.4 The main study ...59

5.4.1 Participant characteristics ...59

5.4.2 Instruments and procedures ...60

5.4.2.1 Dialog materials ...61

5.4.2.2 Instructionalinterventions ...64

5.4.2.3 Measures ...69

5.4.2.3.1 Speakingtests ...69

5.4.2.3.2 Questionnaires ...80

5.4.2.3.3 Reliability ofthe quantitative measures ...86

5.4.2.3.4 Quasi-interviews...87

5.5 Statistical analyses ...88

Chapter 6: Results ...91

6.1 Introduction ...91

6.2 Check achievement ...91

6.3 Part 1 ofthe speakingtest ...93

6.4 Use of formulaic sequencesin Part 2 ofthe speakingtest ...95

(18)

xvi

6.5 Appropriateness of responsesin Part 2 ofthe speakingtest ...105

6.6 Part 3 ofthe speakingtest ...114

6.7 Attitudinalitems usedin both Pre- and Post-Questionnaires ...118

6.8 Reflectiveitems on engagementin memorization given onlyto TGs ...122

6.9 Reflectiveitems onimprovementin output production (except for articulatory aspects) ...124

6.10 Reflectiveitems onimprovementin articulation ...127

6.11 Reflectiveitems onimprovementinlanguage processing ...130

6.12 Reflectiveitems on changesin attitudetoward memorization, especially of FSs ...131

Chapter 7: Discussion and conclusions ...134

7.1 Introduction ...134

7.2 Differential effects of whole-text and partial-text recitation on engagementin memorization ...134

7.2.1 First Check of dialog recitation ...135

7.2.2 Second Check of dialog recitation ...136

7.2.3 Questionnaireitems pertinentto RQ1 ...139

7.2.3.1 A gradeincentive ...139

7.2.3.2 Motivational effects ofin-class recitation ...139

7.2.3.3 Motivational effects ofin-class memorizationtime ...140

7.2.3.4 Practicing until fast ...140

7.2.4 Conclusions for RQ 1 ...141

7.3 Differential effects of whole-text and partial-text recitation on holistic processing...141

7.3.1 Use by students of formulaic sequences ...142

7.3.1.1 Direct application ...142

7.3.1.2 Modified application ...143

7.3.1.3 Use of FSsin non-restrictive conditions ...144

(19)

xvii

7.3.1.4 Questionnaireitems pertinenttothelearning of FSs ...150

7.3.2 Appropriateness of responsestotime-pressed cued prompts ...150

7.3.3 Pronunciation ...151

7.3.4 Abilityto processlanguage data ...152

7.3.5 Conclusions for RQ2 ...152

7.4 Differential effects of whole-text and partial-text recitation on fluent speech production ...153

7.5 Differential effects of whole-text and partial-text recitation on attitudinal change towardtext memorization...155

7.6 Analysis of high andlow achievers ...156

7.6.1 Three high achieversin each group ...159

7.6.2 Threelow achieversin each group ...167

7.7 Summary of major findings and pedagogicalimplications ...174

7.7.1 Significant effects and pedagogicalimplications of whole-text memorization and recitation ...174

7.7.2 Significant effects and pedagogicalimplications of partial-text memorization and recitation ...176

7.7.3 Common effects of whole-text and partial-text memorization and recitation...177

7.7.4 Other variables that may affectthe effectiveness ofthetext memorization approach ...177

7.8 Limitations ofthis study ...179

7.9 Conclusion and future directions ...181

References ...183

Appendix A: Dialogs ...196

Appendix B: Check Sheet ...215

Appendix C: The original speaking Pre-Test ...217

(20)

xviii

Appendix D: The original speaking Post-Test ...230

Appendix E: A summary of all promptsin Part 1 of the speaking tests and notes ...243

Appendix F: A summary of the promptsin Part 2 of the speaking tests and scoring criteria ...244

Appendix G: The original Pre-Questionnaire to all groups ...247

Appendix H: The original Post-Questionnaire to TGs ...250

Appendix I: The original Post-Questionnaire to CG ...254

Appendix J: A comprehensive summary of the questionnaireitems usedin this dissertation along with their English translation ...257

Appendix K: The original Pre-Interview ...259

Appendix L: The original Post-Interview ...261

(21)

xix List of Tables

Table 2.1 Characteristics of Formulaic Sequences 1 ...17 Table 5.1 Attitudinal Items Usedin Both Pre-Questionnaire and Post-Questionnaire 2 ...82 Table 5.2 Reflective Items on Memorization Given Onlyto TGs ……...………..83 Table 5.3 Reflective Items on Improvementin Output Production (Except for Articulatory

Aspects) 4 ...84 Table 5.4 Reflective Items on Improvementin Articulation 5 ...85 Table 5.5 Reflective Items on Improvementin Language Processing 6 ...85 Table 5.6 Reflective Items on Changesin Attitude Toward Memorization (Esp. of FSs) 7 86 Table 5.7 Reliability Estimates forthe Speaking Test Components (Cronbach Alpha)8 ...87 Table 6.1 ‘Check’ Achievement of Dialogs by TGs 9 ...92 Table 6.2 Improvementin Articulatory Appropriatenessin Part 1 (Reading-Aloud Short

Sentences) of Speaking Test 10 ...94 Table 6.3 Improvementin Number of Formulaic Sequences Used from Dialogs for

‘Repeated & Direct Application’ Promptsin Part 2 (Short Translations or

Directed Responses) of Speaking Test 11 ...96 Table 6.4 Improvementin Number of Formulaic Sequences Used from Dialogs for

‘Repeated & Modified Application’ Promptsin Part 2 (Short Translations or Directed Responses) of Speaking Test 12 ...98 Table 6.5 Improvementin Z-Score for Formulaic Sequences Used from Dialogs for

‘Non-Repeated & Direct Application’ Promptsin Part 2 (Short Translations or Directed Responses) of Speaking Test 13 ...100 Table 6.6 Improvementin Z-Score for Formulaic Sequences Used from Dialogs for

‘Non-Repeated & Modified Application’ Promptsin Part 2 (Short Translations

(22)

xx

or Directed Responses) of Speaking Test 14 ...102 Table 6.7 Improvementin Z-Score for All Formulaic Sequences Used from Dialogs for

Part 2 (Short Translations or Directed Responses) of Speaking Test 15 ...104 Table 6.8 Improvementin Appropriateness of Responsesto ‘Repeated & Direct

Application’ Promptsin Part 2 (Short Translations or Directed Responses) of Speaking Test 16 ...106 Table 6.9 Improvementin Appropriateness of Responsesto ‘Repeated & Modified

Application’ Promptsin Part 2 (Short Translations or Directed Responses) of Speaking Test 17 ...108 Table 6.10 Improvementin Z-Score for Appropriateness of Responsesto ‘Non-Repeated &

Direct Application’ Promptsin Part 2 (Short Translations or Directed

Responses) of Speaking Test 18 ...109 Table 6.11 Improvementin Z-Score for Appropriateness of Responsesto ‘Non-Repeated &

Modified Application’ Promptsin Part 2 (Short Translations or Directed

Responses) of Speaking Test 19 ...111 Table 6.12 Improvementin Z-Score for Overall Appropriateness of Responsesin Part 2

(Short Translations or Directed Responses) of Speaking Test 20 ...113 Table 6.13 Improvementin Number of Formulaic Sequences Used from Dialogs for Part 3

(Extensive Oral Productionin English) of Speaking Test 21 ...115 Table 6.14 Improvementin Number of Syllables Spoken per Minute for Part 3 (Extensive

Oral Productionin English) of Speaking Test 22 ...117 Table 6.15 Likert-Scale Score Changesin Attitudinal Survey Items 23 ...119 Table 6.16 Reflections on Engagementin Memorization (Askedto TGs Only) 24 ...123 Table 6.17 Reflections on Improvementin Output Production (Except for Articulatory

Aspects) 25 ...126 Table 6.18Reflections on Improvementin Articulation 26 ...128

(23)

xxi

Table 6.19 Reflections on Improvementin Language Processing 27 ...130 Table 6.20 Reflections on Favorable Changesin Attitude Toward Memorization 28 ...132 Table 7.1 Trigramsthat Were Produced by Four or More Participantsin Any Group

during Post-Test and Also Occurred More Than Once during Pre-Test 29 ...146 Table 7.2 Trigramsthat Were Produced by Four or More Participantsin Any Group

during Post-Test but did Not Appear during Pre-Test 30 ...147 Table 7.3 Total Number of Trigram Types and Tokensin Part 3 (Extensive Oral

Productionin English) of Speaking Test 31 ...147 Table 7.4 Improvementin Number of Targeted Trigrams for Part 3 (Extensive Oral

Productionin English) of Speaking Test 32 ...147 Table 7.5 The Three Highest and Three Lowest Achievers Based on Their Within-Group

Pre-Post Z-Score Difference Rankings 33 ...158

(24)

xxii List of Figures

Figure 3.1. A unified model oflanguage knowledge, processing, andlearning. 1 ...36 Figure 5.1. Detailed schedules for all three groups. ………71 Figure 5.2. Photos usedin Part 3 of the speaking tests. 3 ...79 Figure 6.1.1. Mean distribution of percentages of ‘Check’ achievement by TGs. …...92 Figure 6.1.2. Boxplots showing variations of ‘Check’ achievement by TGs. 5 ...93 Figure 6.2.1. Mean distribution of scores for articulatory appropriatenessin Part 1

(reading-aloud short sentences) of speaking test. ……….94 Figure 6.2.2. Group-by-group boxplots showing improvementin articulatory

appropriatenessin Part 1 (reading-aloud short sentences) of speaking test. 7 ...95 Figure 6.3.1. Mean distribution of number of formulaic sequences used from dialogs for

‘repeated & direct application’ promptsin Part 2 (short translations or

directed responses) of speaking test. ………...96 Figure 6.3.2. Group-by-group boxplots showing improvementinnumber of formulaic

sequences used from dialogs for ‘repeated & direct application’ promptsin Part 2 (short translations or directed responses) of speaking test.9 ...97 Figure 6.4.1. Mean distribution of number of formulaic sequences used from dialogs for

‘repeated & modified application’ promptsin Part 2 (short translations or directed responses) of speaking test. ………...98 Figure 6.4.2. Group-by-group boxplots showing improvementinnumber of formulaic

sequences used from dialogs for ‘repeated & modified application’ prompts in Part 2 (short translations or directed responses) of speaking test.11 ...99 Figure 6.5.1. Mean distribution of z-score for formulaic sequences used from dialogs for

‘non-repeated & direct application’ promptsin Part 2 (short translations or

(25)

xxiii

directed responses) of speaking test. ……….101 Figure 6.5.2. Group-by-group boxplots showing improvementin z-score for formulaic

sequences used from dialogs for ‘non-repeated & direct application’ prompts in Part 2 (short translations or directed responses) of speaking test. ……101 Figure 6.6.1. Mean distribution of z-score for formulaic sequences used from dialogs for

‘non-repeated & modified application’ promptsin Part 2 (short translations or directed responses) of speaking test. 14 ...103 Figure 6.6.2. Group-by-group boxplots showing improvementinz-score for formulaic

sequences used from dialogs for ‘non-repeated & modified application’

promptsin Part 2 (short translations or directed responses) of speaking test.

15 ...103 Figure 6.7.1. Mean distribution of z-score for all formulaic sequences used from dialogs

for Part 2 (short translations or directed responses) of speaking test. 16 ...105 Figure 6.7.2. Group-by-group boxplots showing improvementin z-score for all

formulaic sequences used from dialogs for Part 2 (short translations or

directed responses) of speaking test.17 ...105 Figure 6.8.1. Mean distribution of score for appropriateness of responses to ‘repeated &

direct application’ promptsin Part 2 (short translations or directed

responses) of speaking test. ………106 Figure 6.8.2. Group-by-group boxplots showing improvementin score for

appropriateness of responses to ‘repeated & direct application’ promptsin Part 2 (short translations or directed responses) of speaking test.19 ...107 Figure 6.9.1. Mean distribution of score for appropriateness of responses to ‘repeated &

modified application’ promptsin Part 2 (short translations or directed

responses) of speaking test. ………108 Figure 6.9.2. Group-by-group boxplots showing improvementin score for

(26)

xxiv

appropriateness of responses to ‘repeated & modified application’ promptsin Part 2 (short translations or directed responses) of speaking test.21 ...109 Figure 6.10.1. Mean distribution of z-score for appropriateness of responses to

‘non-repeated & direct application’ promptsin Part 2 (short translations or directed responses) of speaking test. 22 ...110 Figure 6.10.2. Group-by-group boxplots showing improvementinz-score for

appropriateness of responses to ‘non-repeated & direct application’ prompts in Part 2 (short translations or directed responses) of speaking test.23 ...110 Figure 6.11.1. Mean distribution of z-score for appropriateness of responses to

‘non-repeated & modified application’ promptsin Part 2 (short translations or directed responses) of speaking test. ………111 Figure 6.11.2. Group-by-group boxplots showing improvementinz-score for

appropriateness of responses to ‘non-repeated & modified application’

promptsin Part 2 (short translations or directed responses) of speaking test.

25 ...112 Figure 6.12.1. Mean distribution of z-score for overall appropriateness of responses in

Part 2 (short translations or directed responses) of speaking test. ………113 Figure 6.12.2. Group-by-group boxplots showing improvementinz-score for overall

appropriateness of responses in Part 2 (short translations or directed

responses) of speaking test.27 ...114 Figure 6.13.1. Mean distribution of number of formulaic sequences used from dialogs

for Part 3 (extensive oral productionin English) of speaking test. 28 ...116 Figure 6.13.2. Group-by-group boxplots showing improvementin number offormulaic

sequences used from dialogs for Part 3 (extensive oral productionin English) of speaking test. 29 ...116 Figure 6.14.1. Mean distribution of number of syllables spoken per minute for Part 3

(27)

xxv

(extensive oral productionin English) of speaking test. ………..117 Figure 6.14.2. Group-by-group boxplots showing improvementin number ofsyllables

spoken per minute for Part 3 (extensive oral production in English) of

speaking test. 31 ...118 Figure 6.15.1. Item-by-item meandistribution of Likert-scale scores for questionnaire

items regarding attitudes that were surveyedin both Pre-Questionnaire and Post-Questionnaire. 32 ...120 Figure 6.15.2. Item-by-item boxplots showingLikert-scale scores for questionnaire

items regarding attitudes that were surveyedin both Pre-Questionnaire and Post-Questionnaire. 33 ...121 Figure 6.16.1.Mean distribution of Likert-scale score for questionnaireitems regarding

engagementin memorization. ………123 Figure 6.16.2. Item-by-item boxplots showingLikert-scale score for questionnaireitems

regarding engagementin memorization. 35 ...124 Figure 6.17.1. Mean distribution of Likert-scale score for questionnaireitems regarding

improvementin output production (except for articulatory aspects). …..126 Figure 6.17.2. Item-by-item boxplots showingLikert-scale score for questionnaireitems

regardingimprovementin output production (except for articulatory aspects). 37 ...127 Figure 6.18.1. Mean distribution of Likert-scale score for questionnaireitems regarding

improvementin articulation. 38 ...129 Figure 6.18.2. Item-by-item boxplots showingLikert-scale score for questionnaireitems

regardingimprovementin articulation..39 ...129 Figure 6.19.1. Mean distribution of Likert-scale score for questionnaireitems regarding

improvementinlanguage processing.40 ...131 Figure 6.19.2. Item-by-item boxplots showingLikert-scale score for questionnaireitems

(28)

xxvi

regardingimprovementinlanguage processing. 41 ...131 Figure 6.20.1. Mean distribution of Likert-scale score for questionnaireitems regarding

favorable changesin attitude toward memorization. 42 ...133 Figure 6.20.2. Item-by-item boxplots showingLikert-scale score for questionnaireitems

regarding favorable changesin attitude toward memorization. 43 ...133 Figure 7.1. The varying achievement results for each dialog group by group. ……..138 Figure 7.2.1. Mean distribution of number of targeted trigrams for Part 3 (extensive

oral productionin English) of speaking test. 45 ...148 Figure 7.2.2. Group-by-group boxplots showingimprovementin number of targeted

trigrams for Part 3 (extensive oral productionin English) of speaking test. 46 ...148

(29)

CHAPTER 1 Introduction

1.1 Background

Formulaic language abounds in language use, and a number of studies have shown its pervasiveness. For example, in Foster’s investigation (2001), 32.3% of the unplanned native speech analyzed was judged to consist of formulaic language. As another example, Erman and Warren (2000) categorized 58.6% of the spoken and 52.3% of the written English discourse that they examined as formulaic word strings of various kinds. In regard to the percentages, Altenberg (1990) further suggests that if the enormous set of simple lexical collocations, not possible to be elegantly categorized from a formal grammatical point of view, is regarded as part of formulaic language, then, possibly as much as 70% of adult native language may be formulaic. A range of corpus studies (e.g. Altenberg, 1993; Baayen

& Lieber, 1991; Barkema, 1993; Kjellmer, 1984) have also demonstrated that most ordinary language production, written or spoken, appears to be composed largely of collocational sets or frameworks, manifesting far less variability than could be possible on the basis of

grammar and lexicon alone. Taken together, as Sinclair (1991) puts it, “all the evidence points to an underlying rigidity of phraseology, despite a rich superficial variation” (p. 121).

According to Wray’s (2002) thorough search for past observations of formulaic language, the existence of this linguistic phenomenon was recognized as early as the mid-nineteenth century. Narrowing down the scope of her search to the past half century, among the first to discuss the significance of formulaic language are Bolinger (1976), Fillmore (1979), and Pawley and Syder (1983).1 Their critiques were then followed up by Sinclair (1991; see also Section 2.4.3) with his well-known ‘idiom principle,’ while the exploration of the relationship between lexical phrases and functional language was

(30)

2

commenced by Nattinger and DeCarrico (1992). Wray (2002, p. 9) also offers a list of over fifty terms denoting the phenomenon of formulaic language. The list includes some everyday words used by ordinary people (e.g., idioms, formulae, clichés), terms that we do not see fully addressed in current literature, such as amalgams, fossilized forms, frozen phrases, gambits, gestalt, holophrases, and those that seem to be preferred by present linguistic specialists, which include the following:2

chunks / constructions / collocations / conventionalized forms / fixed expressions /

formulaic language / lexical phrases / lexicalized sentence stems / multiword items (units) / non-compositional / prefabricated routines and patterns / ready-made expressions /

recurring utterances / sentence builders

In many ways, formulaic sequences accomplish the same functions as single words.

Many (e.g., collocations: tie your shoes, still waters; and complex verbs: run over, break it down) have mainly a referential or ideational purpose and thus operate as content words do.

Others (e.g., exclamations: Are you serious, no way; and idioms: back to the drawing board, far cry from) are particularly effective for portraying an evaluative stance. Some ensure effortless social interaction (pragmatic formulae such as Good to see you and I’m really happy for you), while others are similar to function words in that they act, for example, to unify discourse (e.g., as a side note, to offer a different perspective). Collectively they make up a substantial and vital part of one’s lexicon, performing an essential role in facilitating the understanding and expression of messages that could otherwise be misinterpreted.

Pervasive and integral as it is, formulaicity remains an area where L2 learners only very slowly approximate to or will never reach the proficiency of native speakers. Results of a

1 For other publications on formulaic language prior to Bolinger, see Wray (2002, pp. 7-8).

2 Wray (2000) cautions against the assumption that researchers are dealing with very much the same phenomenon with various terms (Wray, 2002). As observed by Wray and Perkins (2000), “it seems that there are genuinely deep-seated and significant differences, which have become obscured by the tolerance of terminological variation on the one hand, and, on the other, the indiscriminate

appropriation of certain favoured terms across data types” (p. 3).

(31)

3

number of studies (Altenberg & Granger, 2001; Bardovi-Harlig, 2009; Durrant & Schmitt, 2009; Howarth, 1996; Kaszubski, 2000; Laufer & Waldman, 2011; Li & Schmitt, 2010; Nekrasova, 2009; Qi & Ding, 2011; Siyanova & Schmitt, 2007; Siyanova & Schmitt, 2008; Yamashita & Jiang, 2010) revealthat L2learners even at advancedlevels of proficiency are unableto produce formulaic sequencesinthe L2that are comparabletothose used by native speakers. Pawley and Syder (1983) observethatitis oftenthe failureto utilize nativelike formulaic sequencesthat ultimately distinguishes the advanced L2learner as non-native. Notwithstanding such a disappointing reality, secondlanguageteaching specialists have strived for effective approachesto helplearnerstodevelop formulaicity, notable examples being Lewis (1993), Nattinger and DeCarrico (1992), and Willis (1990). Such attempts have been made precisely because ofthe ubiquity and essential centrality of formulaiclanguage. It has also been suggestedthat deviant use of formulaic sequences by L2learnersis associated with anincreased and sustained processing burden by native speakers (Millar, 2010).

Accordingto data reported by Stengers et al. (2011),inaccuraciesinlearners’ use of

formulaic sequences exert a non-negligible influence ontheir oral proficiency scores.Since certain sequences are strongly linkedto particularlanguage functions orinformation, our interlocutorsanticipate them, andthey arethe preferred choice;thus,formulaic sequences are not merely useful for proficient language usage, but also vital for appropriatelanguage use (Schmitt & Carter, 2004).

Another example of statistical support for an emphasis on formulaic sequencesin teaching comes fromrecent studiesthat found significant correlations between L2learners’

knowledge of multiwordlexis andtheir proficiency ratings (e.g., Boers, et al., 2006; Dai &

Ding, 2010; Hsu & Chiu, 2008; Keshavarz & Salimi, 2007; Stengers, et al., 2011).3 While

3 Althoughthere seemsto be a moderate connection between vocabulary size and formulaic sequence knowledge, Schmitt, Dornyei, Adolphs, and Durow’s (2004) study suggeststhatthe relationship between thesize ofthe ‘individual wordlexiconandthe ‘formulaic sequencelexicon’ is not straightforward.

(32)

4

correlations are notthe same as causal relationships, given all thebenefitsthat knowledge of language patterns and collocations purportedly conferstolanguage users (see Chapter 2 for details),it is,nevertheless,very reasonable for languageteachersto be inclinedto help learners developtheir proficiency byteachingthem formulaiclanguage.

Whenthe choiceto focus on formulaic sequencesis madeinteaching adultlearners, then, several considerations needto betakeninto account for effectiveteaching (see Chapter 4 for details). First and foremost, processing burdens onlanguage users during realtime communicativetasks seriouslyinterfere with thesuccessful processing required forthe learning of formulaic sequences. It willalso be suggested inthis dissertationthat adult learners’ existing knowledge about howthings workinthe world will not only help comprehension and production but also hinderlanguagelearning. Additionally, as pointed outlater,thereare theinherent difficulties of formulaiclanguage. Whentargeting learnersin a foreignlanguage context, wherethereis paucity ofinput andthe needto usethetarget language outsidethe classroom,then,it will be arguedthat engaginglearnersintext memorizationis one promising course of actiontotake.

The role oftext memorizationin L2learningis controversial (Cook, 1994; Dai & Ding, 2010);as a casein point,Wray and Fitzpatrick (2010) expresstheir general observationthat planned memorizationinlanguagelearningis neither standard practicenor fashionable,and thatlanguageteachers whoutilize this approachhave atendency tobelieve thatthey are out of tunewith contemporary methods of languageteaching, while often admitting privatelythat they favor some memorization and findit effective. Cook (2001) also observesthat

memorization remains widely used and relied on by teachers andlearners alike. Good languagelearner studies such as Ding (2007) and Stevick (1989)lend supporttothe

argumentthat memorization oflinguistic materialis a keyto high achievement. Nesselhauf (2003) states that although rotelearningseemsto havelost credit, along with behaviorism,it iscritically importantthat asignificant amount of collocations betaught andlearnt explicitly.

(33)

5

SLAliteratureto date does not offer muchanalysis of text memorization, andthereis an unquestionable scarcity of studies onthe effects oftext memorization (Boers &

Lindstromberg, 2012; Dai & Ding, 2010). The few studies available (e.g., Ding, 2007) are mainlybased on qualitative analysis of student work and reflection. Giventhe controversies overthe practice oftext memorization,then, as Dai and Ding (2010) argue, studiesthat can identifythe effects oftextmemorization on L2learning are called for to addresstheir validity.

1.2 Research aims and the organization of this dissertation

Based onthe need for further researchinto the effects oftext memorizationin classroom teaching,the study presentedinthis dissertationinvestigatedthe effectiveness of engaging classroom FL adultlearnersintext memorization. More specifically,it aimedto seethe extentto which memorization oflengthytexts over an extended period oftime,prepared priortoinstruction and consisting of conversationalturns,would effectivelyleadto formulaic learning.

The discussioninthis chapter hasintroduced the background ofthepresent study and specified itsfocus astheinvestigation oftext memorization approachestotheteaching of formulaic sequencesin a foreignlanguage (FL) context. Chapter 2 discusses various definitions intheliteraturefor formulaic sequences,and provides the particular

characterization of formulaic sequences appropriate for the present study. Italso looks at constructs of formulaic sequences,endeavoringtooffer a morethorough account of how we process and acquire “chunks” oflanguage. In concluding,thechapter establishes connections between formulaic sequences andlanguagelearning, specifically secondlanguagelearning in a foreignlanguage environment.

Chapter 3 addresses the waysin whichwe process and retainlinguisticinformation. The chapter begins by examining thedifferent models concerning the processing and production

(34)

6

oflanguage. The chapterthen moves onto discuss awareness and cognition in relationto memory. Thisis followed by further treatmentof language processingin connectionto memory,and itsrelated components. The dissertationlooksto address issues L2learners experiencein a foreignlanguagelearningenvironment where possibilitiesto acquire an L2 are ratherlimited. The chapter concludes byintroducing anintegrated model forlanguage processing and acquisition in relationtomemory,and outlinesthe features ofthis model.

Chapter 4 aimstointroduce the backgroundtothe study conducted forthis dissertation, with a discussion ofthree fundamental questionsthat were brought up inthe review and synthesis offeredin theprevious chapters.Thethree central questions are (1) Shouldteaching practitionersfocus onformulaic sequencesinteachingin aforeignlanguage context withthe particulartarget population being adultlearners?,(2) Ifthey should, whichformulaic sequences shouldtheyteach?, and (3)How shouldtheyteachthetargeteditems?The first questionis addressed by providing a number of reasons fortheteaching of formulaic

languagetothis particulartarget population.The second andthird questions are addressed by drawing on principles and proposals discussed inthepreceding chapters.This chapterthen presents a review of studies ontext memorizationthat helpto clarifythe rationale forthe present study.

Chapter 5 begins with the research questions, showing howtheylogically follow from thereview oftheliterature givenin Chapters 2through 4. Chapter 5then provides details of the methodology ofthe present study. These detailsinclude pilottesting, participant

characteristics, materials and procedures, as well as the statistical analysesused. In Chapter 6, study results and preliminary findings are reported.

Chapter 7 revisitsthe research questions, discussesthe results and offers conclusions based onthem. The majorfindings are restated together with pedagogicalimplications. This chapterthen identifiesmethodological limitations ofthis study, andfinally specifies future

(35)

7

directions for research onthe roles oftext memorization as a wayto develop formulaic,as well as analytic,knowledge. References and appendices appearlast.

(36)

8 CHAPTER 2

Characteristics of formulaic sequences

2.1 Introduction

This chapter begins with a consideration ofthe various definitions given for formulaic sequences (henceforth, FSs),in orderto provide a comprehensive characterization of FSsthat caninformthis research. It also aimsto examinethe different constructs of formulaic

sequences forthe purpose of better understanding their functionalityin communication. Later sections ofthe chapter endeavorto further clarify how we process and acquire “chunks” of language. The chapter concludes with an attemptto connectthese componentsto second language acquisition,illuminatingin so doingthe waysin which FSs further complicate learning, particularly for adultlearnersin aforeign as opposedto secondlanguage environment.

2.2 Defining formulaic sequences

Formulaiclanguage has been studied from diverse perspectives, resultingin a variety not only of criteria or definitionsto describethe phenomenon but also ofterminology (as introduced atthe beginning of Chapter 1) (Schmitt & Carter, 2004). Corpuslinguisticsis generally concerned withtheidentification and description of formulaic sequences asthey are foundin various kinds of corpus data (Schmitt & Carter, 2004). Criteriathat are

commonly usedinthis field ofinquiryinclude institutionalization,fixedness, non-compositionality (see Section 2.3 for details), and frequency of occurrence.4

4The need for cautionin using frequencytoidentify a formulaic sequence has been recognized (Hickey, 1993; Wray & Perkins, 2000). Whilethereis no doubtthatthereis some sort of connectedness between a sequence being frequentin a corpus andthe conventionalized status accordedtoit by a given speech community,this connectedness may actually beincidental: “[i]t has yetto be establishedthat commonness of occurrenceis morethan a circumstantial associate” (Wray

& Perkins, 2000, p. 7).

(37)

9

Psycholinguists andlanguage acquisition specialists,in contrast, employ criteria such as whether a word stringis used morethan once by a participant (suggestingthatthe useis not so much a single, one-time-onlyimitation as a manifestation ofthe participant’s

proceduralized knowledge) and whetherthe productionis accompanied by anintact

intonation contour (indicatingthatthe sequenceis stored and retrieved as a whole) (Schmitt

& Carter, 2004). Allison Wray, author ofthe seminal book “Formulaic Language andthe Lexicon” (2002),providesthe followinginclusive, umbrella definition of a formulaic sequence:

a sequence, continuous or discontinuous, of words or other elements, whichis, or appearsto be, prefabricated:thatis, stored and retrieved whole from memory atthe time of use, ratherthan being subjectto generation or analysis bythelanguage grammar. (p. 9)

Thisis a definition thatis very extensiveinits coverage, applicabletothe entire spectrum of differenttypes of word strings. Theseinclude, for example,tightlyidiomatic andimmutable strings (e.g., by andlarge) at one end ofthe spectrum, and range totransparent and flexible strings with slots for open classitems such as NP be-TENSE sorryto keep-TENSE you waiting atthe other. Wray (2002, p. 10 and Chapters 11 and 13) further arguesthat even single words and morphemes can be seen as formulaic sequences (e.g., un-believe-able; see also Boers & Lindstromberg, 2012; Wray, 2008).5 Broadly speaking, formulaic sequences can usefully be defined as strings oflinguistic itemswherethe relation of eachitem tothe rest is relatively fixed, and wherethe substitutability of oneconstituent ofthe sequence by another ofthe same category is relatively constrained (Perkins, 1999). An additional, and essential, component ofthis definitionincludes “[w]ords and word strings which appearto be processed without recoursetotheirlowestlevel of composition” (Wray, 2002, p. 4). Since

5 Conceptualizingthe boundary of formulaic sequencesthis wayis convenient,though, especially whenthelanguage underinvestigationis an agglutinative one such as Japanese. Wray (2008) introducesthemorpheme equivalent unit as an alternativeto formulaic sequence.

(38)

10

the focus ofthe present studyis onteaching FSs, not definingthem, I will employthis most comprehensive definitionto date, acknowledging atthe sametimethat, while usefulinthe sensethatit virtually subsumes everything potentially classified as formulaic,the

comprehensive wording and fundamental complexity come atthe cost of makingits

comprehensibility and applicability a colossal obstacle for L2 educatorstryingto work from and build on such a definition.

2.3 Characteristics of formulaic sequences

Becausethereis so much diversityinthe use of formulaic sequences,itis difficultto agree upon absolute criteriato definethem. The definition of a formulaic sequence

introducedin Section 2.2isthus deliberately comprehensive, andincludesvarioustypes of patternedlanguage (Schmitt & Carter, 2004). Takinginto accountthe proposal by Schmitt and Carter (2004)that even though each particular example may not manifest all

characteristics,it seems helpfulto discuss thetypical characteristics of formulaic sequences, the following sub-sections overviewtheir distinctive features from formal, semantic, and functional perspectives.

2.3.1 Formal mutability of formulaic sequences

One ofthe most obvious formal characteristics of FSsis, perhaps,their varyinglengths. Very short sequences can be composed oftwo words (e.g., Come on!) or even one word as introducedin Section2.2 (e.g., Unbelievable!). FSs can be verylongtoo, asinlengthy proverbs (e.g., The grassis always greener onthe other side ofthefence), andtheoretically thelongest varieties can be of such alengththatit could seemimplausibleto assertthatthey are formulaic structures at all. Thelyrics of popular songs could betaken as one such

example.

参照

関連したドキュメント

挿し木苗生産システムの開発を行った。2種のフタバガキ科樹種、S/to剛Sc伽jca

小田25)は「デトラヨ■一ドフエノールフタレンナ

[r]

Key Words : CIM(Construction Information Modeling),River Project,Model Building Method, Construction Life Cycle Management.

Under small amount of pressure reduction, a fine yarn has a little enlarged lashing—end radius, on the contrary, a thick yarn largely increases its lashing—end radius, the power

Taichi ISHIZAWA, Satoshi WATANABE, Shingo YANO, Masaki ABURADA , Ken-ichi MIYAMOTO, Toshiyuki OJIMA, Shinya HAYASAKA:Relationship between Bathing Habits and Physical and

而してCocaine導流開始後5分より10分に至る 迄の期間に現はれる房室伝導系の不完全遮断は

[r]