using meaningful interpretation and chunking to enhance memory the case of chinese characters...

21
Using Meaningful Interpretation and Chunking to Enhance Memory: The Case of Chinese Character Learning Xiaoqiu Xu Pearson Knowledge Technologies Amado M. Padilla Stanford University Abstract: Learning and retaining Chinese characters are often considered to be the most challenging elements in learning Chinese as a foreign language. Applying the theory of meaningful interpretation, the chunking mnemonic technique, and the linguistic features of Chinese characters, this study examines whether the method of meaningful interpretation and chunking (MIC) can promote learnersimmediate learning and retention of Chinese characters. Mandarin Chinese learners at two high schools were randomized into a treatment group and a control group. Students in the treatment group learned Chinese characters with the MIC method, whereas their peers in the control group learned characters by the traditional method of rote repetition according to the stroke order. Four balanced character sets were introduced each day for four continuous days with three different interventions: teacherinstructed method on Day 1, teachercued method on Day 2, and studentsindependent work on Day 3 and Day 4. Studentslearning outcomes of the characters were measured with (1) immediate quizzes given each day after instruction, (2) a retention test (after one week) that integrated all the immediate quizzes, and (3) an application test administered two months after the experiment. The ndings suggest that MIC enhances learnersimmediate learning and retention of Chinese characters. In addition, the teachercued method and familiar independent work were more effective for learning and retaining Chinese characters than the teacherinstructed method and unfamiliar independent work. Furthermore, the treatment effect also varied across the measurement components (meaning vs. perception), levels of instruction, and heritage versus nonheritage groups. Key words: Chinese characters, chunking, meaningful interpretation, radical knowledge, teaching methods Xiaoqiu Xu (PhD, Stanford University) is a test development specialist at Pearson Knowledge Technologies, Sunnyvale, CA. Amado M. Padilla (PhD, University of New Mexico) is Professor of Psychological Studies in Education, Stanford University, Stanford, CA. Foreign Language Annals, Vol. 46, Iss. 3, pp. 402422. © 2013 by American Council on the Teaching of Foreign Languages. DOI: 10.1111/flan.12039 402 FALL 2013

Upload: yuxuan-liu

Post on 29-Jan-2018

524 views

Category:

Education


1 download

TRANSCRIPT

Page 1: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

Using Meaningful Interpretationand Chunking to EnhanceMemory: The Case of ChineseCharacter LearningXiaoqiu XuPearson Knowledge Technologies

Amado M. PadillaStanford University

Abstract: Learning and retaining Chinese characters are often considered to be themost challenging elements in learning Chinese as a foreign language. Applying the theoryof meaningful interpretation, the chunking mnemonic technique, and the linguisticfeatures of Chinese characters, this study examines whether the method of meaningfulinterpretation and chunking (MIC) can promote learners’ immediate learning andretention of Chinese characters. Mandarin Chinese learners at two high schools wererandomized into a treatment group and a control group. Students in the treatment grouplearned Chinese characters with the MIC method, whereas their peers in the controlgroup learned characters by the traditional method of rote repetition according to thestroke order. Four balanced character sets were introduced each day for four continuousdays with three different interventions: teacher‐instructed method on Day 1, teacher‐cuedmethod on Day 2, and students’ independent work on Day 3 and Day 4. Students’ learningoutcomes of the characters were measured with (1) immediate quizzes given each dayafter instruction, (2) a retention test (after one week) that integrated all the immediatequizzes, and (3) an application test administered two months after the experiment. Thefindings suggest that MIC enhances learners’ immediate learning and retention ofChinese characters. In addition, the teacher‐cued method and familiar independent workwere more effective for learning and retaining Chinese characters than the teacher‐instructed method and unfamiliar independent work. Furthermore, the treatment effectalso varied across the measurement components (meaning vs. perception), levels ofinstruction, and heritage versus non‐heritage groups.

Key words: Chinese characters, chunking, meaningful interpretation, radicalknowledge, teaching methods

Xiaoqiu Xu (PhD, Stanford University) is a test development specialist at PearsonKnowledge Technologies, Sunnyvale, CA.Amado M. Padilla (PhD, University of New Mexico) is Professor of PsychologicalStudies in Education, Stanford University, Stanford, CA.Foreign Language Annals, Vol. 46, Iss. 3, pp. 402–422. © 2013 by American Council on the Teaching of ForeignLanguages.DOI: 10.1111/flan.12039

402 FALL 2013

Page 2: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

IntroductionThe Foreign Service Institute (FSI) ofthe U.S. Department of State has categorizedforeign languages taught in the United Statesinto three classes based on linguisticdistance and the length of time it takesEnglish‐speaking students to achieve gener-al professional proficiency in speaking andreading (“Language Learning Difficulty forEnglish Speakers,” n.d.). Mandarin Chineseis one of just a very small number oflanguages assigned to Category III, whichare exceptionally difficult for native Englishspeakers to learn. The FSI estimates that ittakes approximately 2,200 class hours, withat least half of that time spent in immersionstudy, to reach the level of proficiencyneeded to use a Category III language in aprofessional setting (“Language LearningDifficulty for English Speakers,” n.d.).

The most challenging task in masteringMandarin is learning the Chinese characters.While English is an alphabetic languagewhose writing system roughly represents itssound system, the Chinese sound systemand writing system seem to be independentof each other. Thus, mastery of Chinesecharacters is difficult because of the largenumber of nonphonetic, visually complexsymbols that constitute the orthography ofthe language (Packard, 1990). Studentswithout sufficient knowledge of Chinesecharacters often encounter considerabledifficulty in reading (Shen, 2005), withnovice learners of Chinese claiming thatChinese characters are like “random sym-bols” that are beyond mastery and retentiondue to their large quantity and lack ofregularity (Wu, 1992).

Contrary to students’ beliefs, Chinesecharacters are not random symbols withoutpatterns and regularities. An explorationinto Chinese characters1 reveals that trace-able patterns exist that students can use tofacilitate learning characters, reading, andwriting. Linguistically, the composition ofChinese characters is categorized into sixtypes: pictograms, simple ideograms, ideo-grammic compounds, phono‐semantic com-pound characters, phonetic loan characters,

and derivative cognates (Boltz, 1994;Wang, 1993; see Appendix A, Part I).

Among the six categories of characters,phono‐semantic compounds ( in Chi-nese) form more than 90% of Chinesecharacters (Boltz, 1994). A phono‐semanticcharacter is composed of a phonetic elementand a meaning element, or so‐called pho-netic and semantic “radical” ( in Chi-nese). Theoretically, a phonetic radicalrepresents the sound of a character and asemantic radical provides clues to themeaning of the character (see examples oftypical characters with phonetic and seman-tic radicals in Appendix A, Part I). Strokesare the basic building materials for radicals.For example, the radical consists of twostrokes, and \. There are a total of 28distinguishable types of strokes, and thenumber of strokes in a particular charactermay vary from 1 to 30 (Shen, 2005).Furthermore, the way strokes combineand vary across the many Chinese radicalsand characters makes them particularlychallenging to write and remember, espe-cially for novice learners.

Adding to the difficulty of masteringknowledge of radicals and phono‐semanticcompounds is the evolution of Chinesecharacters. Because many Chinese charac-ters have evolved and changed, in modernusage, only about 26% of phono‐semanticcompound characters are transparent char-acters2 (Zhu, as cited in Everson, 1986). Inreality, the phonetic radical does not alwaysidentify with the pronunciation of thecharacter, nor does the meaning radicalalways correspond to the meaning of thecharacter. These characters are called non-transparent characters, and they make upthe majority of phono‐semantic compoundcharacters (Everson, 1986). Among thesenontransparent characters, the meaning andsound radicals only serve as a clue to themeaning and sound of a character, some-times requiring that the learner engage ineffortful imagination. For example, theChinese character is pronounced as fa(third tone3), meaning law in English.Here the left part radical is the semantic

Foreign Language Annals � VOL. 46, NO. 3 403

Page 3: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

component, which means water, and itsright part radical is the phonetic compo-nent, which is pronounced as qu (fourthtone). In this phono‐semantic compoundcharacter , the phonetic and semanticradicals are no longer consistent with thesound and meaning of the character. It is atypical example of nontransparent phono‐semantic compound characters.

As many Chinese characters are non-transparent phono‐semantic compoundcharacters, it becomes impossible to relyon a simple meaningful interpretation ofChinese characters by using the sound andmeaning radicals. Therefore, the ChineseLinguistics Bureau (2005) and Chu(2005, 2009) proposed a new method calledbujian jiaoxuefa (chunking method), or thecomponent‐oriented net‐weaving approach,which employed chunking (bujian) andconnections between chunks, or compo-nents, to promote character learning. Al-though bujian and radicals have manyoverlaps, bujian no longer represents thesemantic or phonological components ofcharacters but represents instead frequentlyappearing chunks in Chinese characters. Inthe 3,500 most frequently used Chinesecharacters, there are only 132 bujian (seeAppendix A, Part II). According to thismethod:

Chinese characters are hierarchicallyorganized into the three levels of stroke,bujian, and character. Thousands ofcharacters consist of hundreds of bujian;hundreds of bujian consist of tens ofstrokes… Strokes and bujian are repeti-tive. Characters are linked together as ahuge network by repeated bujian.(Chu, 2005, p. 250; emphasis added)

Chu (2005, 2009) proposed that, at theinitial learning stage, learners have to learnall the bujian with their accompanyingstrokes as well as the order of writing eachstroke. However, as students progress, theyacquire the ability to automatically apply thebujian knowledge to new characters withlittle difficulty. Although this new methodfor teaching Chinese characters has been

proposed and implemented in some Manda-rin classrooms (Chu, 2009), no empiricalstudies have been conducted to examine itseffectiveness in Chinese character learning.

This study examined the teaching ofChinese characters utilizing an approachcalled meaningful interpretation and chunk-ing (MIC) that integrated several linguisticfeatures of Chinese characters (i.e., origina-tion and types of Chinese characters includ-ing the radical knowledge and bujian) andemployed mnemonic strategies more com-monly found in cognitive studies ofmemory.

Literature Review

MIC in Cognitive PsychologyCognitive psychologists hold that our im-mediate recall and retention of informationis selective: We recall and retain informationthat is personally meaningful rather thanrandom symbols or structures (Anderson,2005). Evidence in support of meaningfulinterpretation has been collected withrespect to short‐term memory of meaning‐significant changes vs. detail changes in apicture (Mandler & Ritchey, 1977), imme-diate recall of meaningful vs. random chesspositions (Schneider, Gruber, Gold, &Opwis, 1993), and retention of theme andmeaning vs. details of a picture (Achor,Imoko,&Uloko, 2009; Chow,Woodford,&Maes, 2011). Furthermore, meaningful in-terpretation enhances students’ retentionof knowledge and has been widely usedby educators in different academic fields,such as biology (Cavallo, 1992), art (Cal-verley, Grafer, & Hauser, 2002), statistics(Chow et al., 2011), and math (Achoret al., 2009).

Chunking is another popular strategy(Ericsson, Chase, & Faloon, 1980) andrefers to processing small units of informa-tion (chunks) and grouping them intolarger, meaningful units (Chase & Simon,1973). The importance of chunking origi-nated withMiller’s (1956) work that showedthat short‐term memory had a capacity ofabout seven plus‐or‐minus two chunks and

404 FALL 2013

Page 4: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

that short‐term memory could be enhancedby recoding information into a small amountof high‐information‐content items. In thefield of language learning, research aboutchunking remains inconclusive and conven-tional thinking holds that chunking, ordecomposing, is effective for alphabeticword acquisition. Most research has beenconducted in the field of English wordrecognition and acquisition (Rastle, Davis,Marslen‐Wilson, & Tyler, 2000; Rubin &Becker, 1979; Taft & Forster, 1975), indi-cating that English words can be groupedinto a variety of smaller units such asindividual letters, spelling patterns, sylla-bles, and morphemes. These smaller unitscan then function as basic perceptual unitsfor processing in word recognition andacquisition (Taft & Forster, 1975).

Research has also shown that associa-tions between items or chunks can assist inshort‐term recall (Tulving & Patkau, 1962).Stuart and Hulme (2000) found that pre‐exposing pairs of low‐frequency words inorder to create associative links betweenthem had substantial beneficial effects onimmediate serial recall performance. Thesefindings indicate that associating linksbetween items or increasing availability ofprior knowledge can enhance retention. Animportant implication, then, is that teachersshould pay attention to linking items withthe same or similar chunks.

MIC in Teaching and LearningChinese CharactersDespite the evidence in support of meaning-ful interpretation and chunking in cognitivepsychology, no study has applied both ofthese cognitive processes to examine howChinese characters are taught. The majorityof related studies have focused on radicalknowledge and the essence of meaningfulinterpretation of Chinese characters andhave demonstrated a correlation betweenradical knowledge and Chinese characteracquisition. For example, Taft and Zhu’sstudies (1995, 1997) argued that there is “aradical‐transportation effect” on readers and

that all simple radicals are independentlyactivated in the process of character recog-nition, including the characters containingmore complicated radicals. Therefore, therecognition of Chinese characters is pro-moted by the activation of informationabout their component radicals. Further-more, Shen’s study (2000), which investi-gated the relationships between radicalknowledge and recognition and productionof novel phonetic‐semantic compounds,showed that students with good radicalknowledge performed significantly better inthe production of novel morphologicaltransparent characters than did studentswho lacked equivalent radical knowledge. Ina follow‐up studywith college learners, Shenand Ke (2007) found a linear relationshipbetween the development of radical knowl-edge and the ability to apply that knowledgein Chinese word acquisition.

Although a connection between radicalknowledge and Chinese character learningseems to have been established, a method toeffectively teach radical knowledge is yet tobe studied and discussed. A controversyexists as to whether Mandarin teachersshould systematically and explicitly teachradical knowledge to students (Shen, 2007;Taft & Chung, 1999; Wang, Liu, & Perfetti,2004). Recent research has supported teach-er‐assisted instruction of positional andfunctional regularities of radicals (Taft &Zhu, 1995, 1997). More important, researchhas shown that if teachers explicitly intro-duce target characters with the meaning anduse of radicals, then students are more likelyto associate the learned radical knowledgewith the pronunciation and meaning ofnewly learned characters, which expediteslearning (Taft & Chung, 1999; Wang et al.,2004).

In a recent research study, Shen (2007)showed that student‐initiated elaborationcan be as effective as teacher‐guided elabo-ration in the long term. In the study, threetypes of encoding strategies were usedduring character learning; these includedrote memorization (shallow processing),student self‐motivated elaboration (deeper

Foreign Language Annals � VOL. 46, NO. 3 405

Page 5: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

processing), and teacher‐guided elaboration(deeper processing). Her findings indicatedthat elaboration resulted in significantlybetter retention for sound and meaning ofcharacters than did rote memorization.Between student self‐motivated elaborationand teacher‐guided elaboration, retention ofsound and meaning was significantly betterwith teacher‐guided elaboration in studyintervals of 20 minutes; however, thisadvantage disappeared at a 48‐hour recallinterval. Therefore, teacher elaborationapparently only enhanced workingmemory,not the retention of character sound andmeaning. In other words, after being intro-duced to the method of deep processing,student self‐motivated elaboration can be aseffective as teacher‐guided elaboration.

Finally, Everson and Ke (1997) foundthat highly proficient learners have a moreadvanced understanding of Chinese orthog-raphy and morphology. In a sight‐readingtask employing think‐aloud protocols, theselearners more easily applied their radicalknowledge to identify unknown charactersand made fewer random decisions to figureout the pronunciation and meaning of thesecharacters.With regard to language learners’family background, although no specificstudy has been identified investigating thedifference on Chinese character learningbetween heritage and non‐heritage learnersof Chinese, research has demonstrated thatthe two groups differ significantly inlanguage learning in terms of their motiva-tion, resources, and general learning strate-gies (Liu, 2012; Scrimgeour, 2012; Wen,2011). Therefore, in the study, both heritageand non‐heritage participants were re-cruited, and their learning outcomes usingthe MIC method were analyzed andcompared.

Research QuestionsTo examine the effectiveness of the MICmethod as well as the instructional setting inwhich the method was delivered, thisempirical study compared an experimentalgroup and a control group’s learning and

retention of Chinese characters and ad-dressed the following research questions:

1. Is the MIC method more effective thanthe traditional stroke‐order method instudents’ immediate learning of Chinesecharacters?

2. Does the MIC method produce greaterretention of Chinese characters than thetraditional stroke‐order method?

3. Does the MIC method have a long‐termeffect on students’ learning of Chinesecharacters?

4. Which treatment is more effective inlearning Chinese characters: teacher totalinstruction, teacher‐cued instruction, orstudents’ independent work?

5. How does level of instruction (2nd yearvs. 3rd year students) and heritage vs.non‐heritage learner status affect instruc-tional treatment for learning Chinesecharacters?

Methods

DesignThe study was conducted in the Palo AltoUnified School District in northern California.Due to limited time and resources, a con-venient sampling strategy (Alreck & Settle,1994) was used to select the study partici-pants from two local high schools. Theparticipants included 124 students enrolledin Chinese classes. Given that the mainexperiment lasted for only four days, the16 students (seven students in the controlgroup and nine students in the treatmentgroup) who were absent on any of the fourdays were excluded from the analyses. Ttest comparisons showed that the pretestresults of the absent students in the controland treatment groups were not signifi-cantly different (p > 0.05), and thereforetheir absence would not bias the study.The participants ranged in age from 14 to17 years. Among the 108 participants, 70were male (65%) and 38 were female(35%). Thirty‐five students (32%) were

406 FALL 2013

Page 6: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

heritage speakers, and 73 students (68%)were non‐heritage speakers.

This study employed a randomizedexperimental design. Students were ran-domized into a treatment group and acontrol group within the same class asfollows. First, the original 124 students werehomogeneously grouped by pretest scores.The pretest was a recognition test adminis-tered one day before the experiment andconsisted of all 32 characters to be taughtin the study. During the pretest, studentswere asked to write down the pinyin andmeaning of the characters. Test papers werecollected immediately afterward, and resultswere never shared with students so thatstudents did not receive any instructionalexposure to those characters except in theexperiment. After ranking students by theirpretest performance, matched pairs withsuccessive ranks were formed. Second,matched pairs were randomly assignedsuch that one student participated in thetreatment group and the other in the controlgroup. Third, a few participants within amatched pair were switched to balance thegender distribution between the treatmentand control groups. Table 1 presentsinformation about participants’ group,school, and level.

Students at both schools used the sametextbook series, Nihao, and had completedapproximately the same amount of instruc-tional time and content based on their level of

Chinese instruction. A standardized Chineseassessment instrument, the Standards‐basedAssessment and Measurement of Proficiency(STAMP), was administered to studentsapproximately two weeks before the study.The results indicated that the majority ofChinese II students fell into the novice‐low4

to novice‐mid categories, the majority ofChinese III students fell into the novice‐midto novice‐high categories, and themajority ofChinese IV students fell into the novice‐highto intermediate‐low categories.

MaterialsBoth the control and treatment groups weretaught eight Chinese characters daily forfour consecutive days for a total of 32different characters (see Table 2). Thecharacters were all selected from the wordlists of Lessons 4, 5, and 6 in Nihao IV. Thehighest‐level students (Chinese IV studentsat School A) in the study were learningLesson 2 in Nihao IV when the study began.Therefore, the characters used in the studyhad never been taught formally in class tostudents. Five criteria were used to select thecharacters for the study: human memorycapacity,5 number of strokes (i.e., density),frequency of characters, transparency ofphono‐semantic compounds, and learningin context (see Appendix B). By followingthese criteria, the daily level of difficulty wascontrolled to enable within‐subject compar-isons across the four instructional days.

TABLE 1

Distribution of Experimental and Control Group

ExperimentalGroup

Control Group

Level II (about 120 instructionalhours)

11 (School A) þ11 (School B)

12 (School A) þ11(School B)

Level III (about 240 instructionalhours)

8 (School A) þ15 (School B)

9 (School A) þ15 (School B)

Level IV (about 360 instructionalhours)

8 (School A) 8 (School A)

Total 53 students 55 students

Foreign Language Annals � VOL. 46, NO. 3 407

Page 7: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

For both the treatment and controlgroups, handouts on which characters andactivities would be covered were provided tostudents. A substitute Chinese teacher (see“Procedure” below for details) was trainedin using the handouts for the study lessons.The treatment group also received an extrahandout illustrating types of Chinese char-acters based on their origin and the MICmethod (see Appendix A). In addition tothe handouts, Chinese‐English dictionarieswere available to all participants in thestudy’s follow‐up activities.

ProcedureTo prevent research outcomes from beinginfluenced by subjective bias, the partici-pants and the experimenter were blind towho was assigned to the experimental andcontrol groups. A credentialed substituteMandarin teacher was employed as theexperimenter to teach all participants inboth groups. The teacher was trained in thesteps that should be followed in deliveringthe study instruction to the participants.

The experiment was conducted duringthe regular Chinese class periods in bothschools. For the control group, the teacher

used a traditional method (i.e., stroke‐orderrote memorization) to teach the Chinesecharacters. On Day 1, students followed theteacher’s model, writing the character strokeby stroke (teacher‐instructed traditionalmethod). On Day 2, the teacher providedthe character’s stroke order on the board,and students practiced writing by them-selves (teacher‐cued traditional method).OnDay 3, students were asked to study eightcharacters by themselves and practice eachcharacter’s stroke order (student indepen-dent traditional method). The teacher wasnot in the classroom. On Day 4, the studentindependent traditional method from Day 3was repeated for a new character set. Thispractice lasted for approximately 15minuteson each day of the experiment. Afterward,students participated in “making up wordsand phrases,” an activity in which they hadto look up characters in the dictionary anduse those characters to make two words orphrases. This activity lasted approximately10 minutes on each day of the experiment.

For the treatment group, on the daybefore the experiment, the teacher intro-duced theMICmethod with three steps. Thefirst step in this method was to introduce theorigin and types of Chinese characters (see

TABLE 2

Character Sets Selected for Four Days

Strokes Day 1 Day 2 Day 3 Day 4

�5 strokes 90�

129 1018 589213 166 320 565

6–7 strokes 360 988 127 73827 531 110 1652

8–9 strokes 219 846 586 1701149 504 499 362

10–15 strokes 326 134 824 474822 55 41 295

Words

�Note: Numbers are the frequency ranking of Chinese characters based on the ModernChinese Character Frequency List (http://lingua.mtsu.edu/chinesecomputing/statistics/char/list.php?Which¼MO).

408 FALL 2013

Page 8: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

Appendix A, Part I). Because introducing allsix linguistic breakdowns can be over-whelming for new Chinese as a foreignlanguage learners, teachers can simplify theknowledge by leaving out phonetic loancharacters and derivative cognates becausethey are themost difficult to deconstruct andmake up less than 1% of all Chinesecharacters (Boltz, 1994; Wang, 1993).Therefore, in this study, only the first fourtypes of Chinese characters (pictograms,simple ideograms, ideogrammic com-pounds, and phono‐semantic compoundcharacters) were introduced to students.The teacher emphasized radical knowledgewhen teaching the phono‐semantic com-pound characters. For each type of charac-ter, she gave abundant examples of Chinesecharacters and demonstrated how theycould be meaningfully interpreted.

Then, the teacher told students thatphono‐semantic compounds (xingsheng zi)comprised about 90% of Chinese charactersand that only about 26% of the phono‐semantic compounds were transparent orclose to transparent. Definitions and exam-ples of transparent and nontransparentcharacters were given to students. Tomemorize those nontransparent characters,the teacher suggested that students usebujian (chunks) for memorization. Sheintroduced the method of bujian andindicated that only 132 of the most commonbujian could be found in a list of the 3,500most frequently used Chinese characters(see Appendix A, Part II). She also gaveexamples on how to chunk several Chinesecharacters.

In the third step, the teacher listedseveral characters that consisted of the sameradicals or chunks; by doing so, shefacilitated students’ identification and asso-ciation of those radicals, chunks, andcharacters (see Appendix A, Part III). Theteacher then stated that, although theelement of meaningful interpretation fo-cused on the types of Chinese characters andradical knowledge and while the element ofchunking focused on the structures ofChinese characters, students should com-

bine both elements as there were overlapsbetween chunks and radicals that frequentlyconnect the characters consisting of the samechunks or radicals. This three‐step intro-duction lasted approximately 90 minutesand did not involve the characters selectedin the study.

After students participated in the abovethree‐step orientation to the MIC method,on Day 1 of the experiment, the teacherintroduced eight characters to students byproviding her own interpretation and mem-ory tips based on the MIC method (teacher‐instructed MIC). For example, when intro-ducing the character (ren, recognize), theteacher mentioned that the left radicalrepresented the meaning (talk, express out)and the right radical represented the sound(ren, similar pronunciation of ). Whenintroducing the character (shi, recognize),the teacher mentioned that the left radical ,again, represented the meaning (talk, ex-press out) and the right radical repre-sented the sound (zhi), emphasizing that thephonetic radical might not exactlymatch thesound of the compound character. Anotherexample was the character of . The teachertold students that chunking could be used inmemorizing this character and wrote the tipon the board: ¼ þ þ þ . OnDay 2, the activities repeated Day 1’sstructure; however, the teacher changedthe character set and encouraged students’own interpretation and character chunkingby providing cues (teacher‐cued MIC). Forexample, when she taught , she said, “Thisis a phono‐semantic compound, which meansone radical represents pinyin and the otherrepresents the meaning. Who can share hisor her memory tips?” When teaching , theteacher told students that they could chunkthe character into familiar bujian and askedstudents which two bujian they couldidentify and in what other characters theyhad seen them before. On Day 3, studentswere asked to self‐study the eight charactersusing the MIC method. The teacher was notpresent (student independent MIC). Thestudent independent MIC method wasrepeated on Day 4, except that the

Foreign Language Annals � VOL. 46, NO. 3 409

Page 9: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

participants studied a different characterset. Similar to the control group, time wascontrolled for learning the eight characterseach day (about 15 minutes). Afterlearning the character set, the treatmentgroup joined the control group for thewords‐and‐phrases activity, which lastedapproximately 10 minutes, in the sameclassroom.

A quiz was administered in the sameclassroom each day to both groups after thewords‐and‐phrases activity (see AppendixC). In the first section of the quiz, after theteacher pronounced the Chinese charac-ters one by one, students wrote thecharacters and their meanings in English.In the second section, students completeda character‐recognition task in which theysaw the printed characters and were askedto provide their pinyin spelling and mean-ings in English.

On Day 5, a week later, a retention testwas administered to both the treatmentand control groups. Items on this cumula-tive test were presented in the same formatas the previous quizzes (administered onDay 1 through Day 4) and includedcharacters taught on Day 1 through Day 4.

Finally, to examine whether the MICmethod enhanced long‐term student learn-ing, an application test was given twomonths after the experiment. Rather thanassess students on the Chinese charactersthey learned in the experiment, theapplication test followed the same formatas the retention test but examined studentson the characters they learned from theirclassroom teachers in the two‐monthperiod after the experiment finished. Theintention of this test was to determinewhether students could apply the MICmethod to their own Chinese characterlearning. The experimental design isshown in Table 3.

MeasuresThe immediate, retention, and applicationtests applied the same four types ofmeasurement to test students’ memory of

TABLE3

DesignoftheStudy

Day

1Day

2Days3an

d4

Oneweeklater

Twomon

thslater

Con

tent

Characterset1

Characterset2

Charactersets

3and4

Retention

test

(Quiz

4)App

licationtest

onChinesecharacters

taugh

tin

betw

een

Step

1Con

trol

grou

pTeacher‐instructed

traditional

method

Teacher‐cued

traditional

method

Students’independent

traditional

method

Treatmentgrou

pTeacher‐instructed

MIC

Teacher‐cued

MIC

Students’independent

MIC

Step

2(Both)

Related

activity

Related

activity

Related

activity

Step

3(Both)

Quiz

1Quiz

2Quiz

3,4

410 FALL 2013

Page 10: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

characters: (1) reading the character andwriting down its pinyin, (2) reading thecharacter and writing down its meaning,(3) listening to the character and writingdown the character, and (4) listening to thecharacter and writing down its meaning. Forthe two reading tasks, characters werepresented on a handout and students wererequired to write down their pinyin spellingand meanings. For the two listening tasks,characters were dictated and students wererequired to write down the characters andtheir meanings.

The number and percentage of correctpinyin spellings, written characters, andcharacter meanings were measured. Stu-dents earned one point if they wrote thepinyin correctly (including tones), onepoint if they wrote the character correctly,and one point if they wrote the correctmeaning. When students provided the rightpronunciation with an incorrect tone, theyreceived a score of 0.75. When studentsprovided an intelligible pronunciation thatwas almost correct, they received a score of0.5. When students made a small mistake inthe character writing, such as missing oradding a stroke, but the character was stillrecognizable, they received a score of 0.5. Ifthe meaning provided was close, but notexactly correct, they also received a scoreof 0.5.

AnalysesFirst, to make the analysis more concise andefficient, the two most relevant measure-ments—reading the character and writingdown its pinyin spelling and listening to thepinyin spelling and writing down thecharacter—were combined and renamed asthe perception component. The other twotasks—reading the character and writingdown its meaning and listening to thecharacter and writing down its meaning—were also combined and renamed as themeaning component.

Multi‐factor ANOVA was performed toanalyze three main variables in the study:instructional differences as indicated by the

variable day (four levels6), the treatment orcontrol condition as indicated by thevariable group (two levels), and differentmeasurements as indicated by the variablecomponent (two levels). To investigate themain effects of the MIC method anddifferent instructional strategies on the testcomponents as well as their interactionsbetween each other, a 2 (group) � 4 (day)� 2 (component) multi‐factor ANOVA wasconducted.

In addition, two additional variables,language level as indicated by the variablelevel (three levels) and heritage or non‐heritage students as indicated by the variableheritage (two levels) were also of highinterest and were included in the results.To analyze the impact of the MIC methodincluding the level and heritage variables, a 2(group) � 3 (level) / 2 (heritage) � 4(day) � 2 (component) multi‐factor AN-OVA was performed on the dependentvariables.

Results

Immediate TestsAll means and standard errors are reportedin Table 4. The results of the analysisshowed that treatment was significant,F(1,106) ¼ 4.73, p < 0.05, indicating thatthe MICmethod was more effective than thetraditional method for short‐term memoryof Chinese characters. Day was significantfor both groups, F(3,104) ¼ 49.42,p < 0.001. Generally speaking, studentperformance was ranked from high to lowin the following order: Day 4 (studentfamiliar independent work), Day 2 (teach-er‐cued instruction), Day 1 (teacher totalinstruction), and Day 3 (student unfamiliarindependent work). Component was signif-icant, F(1,106) ¼ 45.23, p < 0.001, show-ing that students performed significantlybetter on the meaning tasks than on theperception tasks. There was an inter-action between group and component,F(1,106) ¼ 7.59 p < 0.01, which demon-strated that the treatment group did espe-cially well on the meaning‐related tasks as

Foreign Language Annals � VOL. 46, NO. 3 411

Page 11: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

compared to the control group. Figure 1displays the means of the immediate tests bycomponent and group. The interactionbetween day and group was not significant,F(3, 104) ¼ 1.17, p > 0.05. This indicatesthat the MIC method was equally effectiveacross the four instructional settings.Figure 2 displays the means of the immedi-ate tests by day and group.

Retention TestsAll means and standard errors are reportedin Table 5. Analyses of the retention testsproduced results that were similar to theanalyses for the immediate tests. Thetreatment was significant, F(1, 106) ¼6.33, p < 0.05, indicating that the MICmethod was also effective for retention ofChinese characters.

Day was significant for both groups, F(3,104) ¼ 33,26, p < 0.001. Generally speak-ing, student performance was ranked fromhigh to low in the following order: Day 2(teacher‐cued instruction), Day 4 (studentfamiliar independent work), Day 1 (teachertotal instruction), and Day 3 (student unfa-miliar independent work). Component wasfound to be significant, F(1,106) ¼ 158.09,p < 0.001, with students performing signifi-cantly better on themeaning tasks than on theperception tasks. Similar to the immediatetest, the retention analysis revealed a signifi-cant interaction between group and compo-nent, F(3, 104) ¼ 6.74, p < 0.01, whichdemonstrated that the treatment group didespecially well on meaning‐related tasks ofthe retention test when compared to thecontrol group. The interaction between day

TABLE 4

Means (Standard Errors) of Immediate Tests

Day 1 Day 2 Day 3 Day 4 AverageMeans

Control Perception 9.59 (0.58) 12.48 (0.41) 10.55 (0.55) 13.15 (0.42) 11.44 (0.44)Meaning 10.40 (0.58) 12.88 (0.42) 10.9 (0.56) 13.92 (0.33) 12.31 (0.45)EM Means 9.99 (0.56) 12.68 (0.40) 10.72 (0.53) 13.53 (0.34) 11.73 (0.41)

Treatment Perception 10.67 (0.59) 13.27 (0.42) 11.82 (0.56) 13.47 (0.43) 12.03 (0.41)Meaning 11.83 (0.59) 14.36 (0.43) 13.30 (0.57) 15.35 (0.33) 13.71 (0.42)EM Means 11.26 (0.57) 13.82 (0.40) 12.56 (0.54) 14.41 (0.35) 13.01 (0.42)

Both Perception 10.13 (0.42) 12.88 (0.29) 11.18 (0.39) 13.31 (0.30) 11.88 (0.31)Meaning 11.12 (0.42) 13.62 (0.30) 12.10 (0.40) 14.63 (0.23) 12.87 (0.29)Avg. Means 10.62 (0.40) 13.25 (0.28) 11.64 (0.38) 13.97 (0.24) 12.43 (0.37)

FIGURE 1

Means of Immediate Tests, by Component and Group

412 FALL 2013

Page 12: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

and group was not significant, F(3,104) ¼ .26, p > 0.01, indicating that theMIC method was equally effective across thefour instructional settings.

Application TestTwomonths later, studentswere tested on thenew characters they had learned from theirteachers since the completion of the experi-ment. Chinese II students learned 29 charac-ters, and Chinese III and Chinese IV studentseach learned 44 characters. A similar applica-tion test consisting of the perception andmeaning components of those characters wasgiven to students in both the control andtreatment groups. To make their scorescomparable, the percentage of correct scoresfor data analysis was used.

The means of both the perception(M ¼ 50.72%, SE ¼ 1.62%) and meaningcomponents (M ¼ 59.22%, SE ¼ 1.07%) forthe treatment group were higher than themeans of the control group (for the perceptioncomponent M ¼ 49.37%, SE ¼ 1.06%; forthe meaning component M ¼ 57.49%, SE¼ 1.02%), but these differences were notsignificant. For the perception component,F(1, 102) ¼ .13, p > 0.05, and for the mean-ing component, F(1, 102) ¼ .09, p > 0.05.This result indicates that the treatment effectof the MIC method disappeared in theapplication test given two months later.

Level and Heritage EffectLevel was significant in both the immediateand the retention tests, F(2, 106) ¼ 7.03,

FIGURE 2

Means of Immediate Tests, by Day and Group

TABLE 5

Means (Standard Errors) of Retention Tests

Day 1 Day 2 Day 3 Day 4 AverageMeans

Control Perception 4.77 (0.55) 5.56 (0.55) 3.67 (0.39) 4.98 (0.57) 4.74 (0.48)Meaning 6.15 (0.65) 6.97 (0.60) 4.94 (0.52) 7.11 (0.64) 6.29 (0.55)EM Means 5.46 (0.54) 6.26 (0.56) 4.30 (0.44) 6.04 (0.58) 5.52 (0.51)

Treatment Perception 6.37 (0.57) 7.07 (0.57) 4.98 (0.40) 6.26 (0.59) 6.17 (0.50)Meaning 8.68 (0.67) 9.21 (0.62) 6.93 (0.54) 9.27 (0.67) 8.52 (0.57)EM Means 7.53 (0.60) 8.14 (0.58) 5.96 (0.45) 7.76 (0.60) 7.35 (0.52)

Both Perception 5.57 (0.40) 6.31 (0.40) 4.33 (0.28) 5.62 (0.41) 5.46 (0.35)Meaning 7.42 (0.47) 8.09 (0.43) 6.39 (0.37) 8.19 (0.46) 7.41 (0.40)Avg. Means 6.49 (0.42) 7.20 (0.40) 5.13 (0.31) 6.90 (0.42) 6.50 (0.38)

Foreign Language Annals � VOL. 46, NO. 3 413

Page 13: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

p< 0.01, and F(2, 106)¼ 10.63, p < 0.001,respectively. Higher‐level students per-formed better than lower‐level students onboth the immediate and retention tests.Furthermore, the analyses revealed that,compared to Chinese III and Chinese IVstudents,Chinese II students laggedbehindtoa greater extent on perception tasks whencompared to meaning tasks in the tests.

Heritage was a significant factor in boththe immediate and retention tests, F(1,106) ¼ 22.73, p < 0.001, and F(1, 106) ¼45.31, p < 0.001, respectively. Heritagestudents not only performed better thannon‐heritage students on both tests, but theformer also showed less of a performancedecrement in the retention tests. Similar tothe level variable, compared to heritagestudents, non‐heritage students lagged be-hind more on perception tasks than onmeaning tasks in the tests.

Discussion

Immediate Learning and RetentionThe results of the study support a cognitiveinterpretation of the role of MIC in immedi-ate recall and retention of information(Anderson, 2005; Chase & Simon, 1973;Gobet, Retschitzki, & de Voogt, 2004; Gobet& Simon, 1998; Mandler & Ritchey, 1977).In converting Chinese characters into moremeaningful material, the MIC method firstinformed students that Chinese charactersare not random symbols but are groundedupon historical stories and possess mean-ingful interpretation. In this sense, mean-ingful interpretation provided students witha framework and contextual clues to learnand retain Chinese characters. The otherelement of the MIC method, chunking andassociation between chunks, was also sup-ported by the results of the study. Whenlearning compound Chinese characters,students in the treatment group first decom-posed the characters into familiar chunks(bujian) and then proceeded to unite thecharacters by connecting the writing andmeaning. As Taft and Forster (1975)claimed, the obvious advantage of this

kind of decomposition procedure is aneconomy of memory storage. Instead ofmemorizing more than 10 strokes, studentsonly need to retain two or three chunks.Although the initial process may be moretedious for students who are still in theprocess of learning the chunks, the taskbecomes easier as students become exposedto and remember chunks.

The effectiveness of the MIC method isalso consistent with studies that investigatedthe relationship between radical knowledgeand Chinese character acquisition (Shen,2000; Shen & Ke, 2007; Taft & Chung,1999; Wang et al., 2004). Although there isno literature on the MIC method inparticular, teaching radical knowledge mir-rors the MIC method because a radical is atype of Chinese character chunk withmeaning or sound indications. The findingsof the MIC method are not only consistentwith radical knowledge theory but alsoextend its benefits to retaining the pronun-ciation and meaning of Chinese characters.

The findings reported here also demon-strate that students are better at learning andretaining the meaning components of char-acters as compared to the perceptioncomponent. This is consistent with Shen(2010), who reported that students hadmore difficulty learning the sounds andshapes of characters than they did learningthe character’s meanings. One conclusionthat can be drawn from this finding is thatlanguage instruction should aim to havestudents attend to strategies for internalizingthe sound andwriting of Chinese characters.

Long‐Term EffectsA puzzling problem to emerge from thestudy was the absence of a significant long‐term effect of the MIC method whenstudents were given an application testtwo months following the experimentalmanipulation. It is not clear why studentsdid not retain and transfer the benefits of theMIC method when learning new Chinesecharacters. It is possible that, for long‐termmemory to be maintained, some level ofrehearsal of thematerial is needed (Atkinson

414 FALL 2013

Page 14: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

& Shiffrin, 1968; Baddeley, 1986; Bower,1982). During the two‐month intervalbetween the experiment and the applicationtest, the Chinese classroom teachers contin-ued to use the traditional stroke‐order rotememorization method to teach Chinesecharacters and thus no overt rehearsal hadtaken place. Thus it is conceivable that thewaning of the MIC method in the applica-tion test could be attributable to both a lackof rehearsal and, possibly, to interferencefrom the rehearsal of an opposing method oflearning. The other possibility could be thatthe short duration of the experiment was notsufficient to effect a change in students’strategic learning orientation, somethingthat would have to be nurtured over timethrough principled pedagogical interven-tions. Therefore, it is important that theteacher consistently act as a facilitator toencourage students to incorporate the MICmethod in their Chinese character learning.Support for this can be found in Everson(2011), who stated in his synthesis of theliterature on learning character languages:

As well, given the variable nature ofstudents’ ability to actually use thesecomponents successfully, it appears thisis no longer nice‐to‐know informationto be taught from a cultural or historicalperspective, but information that mustbe woven systematically into the fabricof Chinese reading classroom pedagogy.(pp. 263–264)

Day EffectGenerally speaking, Day 2 and Day 4 wereranked at the top in both the immediate‐recall and retention tests, which indicatesthat teacher‐cued instruction and familiarindependent learning were more effective forlearning Chinese characters in this study.This conclusion is consistent with Bransford,Franks, Vye, and Sherwood’s theory (1989)of “Wisdom can’t be told.” According toBransford et al. (1989), when studentswere instructed on knowledge and prob-lem‐solving models for learning, studentscould think of the knowledge and themodels

mechanically butwere not able to transfer theknowledge or model creatively to newsituations. The more effective method, asdiscussed by Bransford et al. (1989), was theproblem‐oriented acquisition procedure, inwhich students, under a teacher’s tutelage,receive extensive opportunities to manipu-late objects and problems by themselves.Shen’s study (2007) on learning Chinesecharacters also reinforced this result, show-ing that, after being introduced to themethodof deep processing, student self‐motivatedelaboration could be as effective as teacher‐guided elaboration in the long run.

Subgroup EffectsNo significant interaction was found be-tween level and heritage and treatment. Inaddition, the results indicated that the MICmethod was equally effective both for lower‐level and higher‐level learners as well as forheritage and non‐heritage learners. On theother hand, the results of the subgroupanalyses also demonstrated that lower‐leveland non‐heritage students showed thelargest performance decrement on theimmediate test of Day 3, when they wererequired to conduct independent learning ofthe characters for the first time. Anotherfinding was that lower‐level and non‐heri-tage students lagged behind on the percep-tion tasks rather than on the meaning taskswhen compared to the higher‐level andheritage students. Thus, when using theMIC method to teach Chinese characters,teachers should provide more scaffolding tolower‐level and non‐heritage students andpay particular attention to improving stu-dents’ abilities on perception tasks (pro-nouncing and writing the characters).

Implications for Teaching PracticesThis study provides a number of sugges-tions for teaching Chinese characters andpossibly has implications for teaching otherlanguages that use logographic characters(e.g., Japanese). First of all, the MIC methoddiffers from other methods that use radicalknowledge in teaching Chinese characters

Foreign Language Annals � VOL. 46, NO. 3 415

Page 15: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

—although it integrates the knowledge ofradicals as a way to ensure meaningfulinterpretation, it embeds the following threekey elements: (1) meaningful backgroundknowledge (origination and types ofChinese characters, radical knowledge);(2) chunking (bujian); and (3) associationamong characters consisting of the sameradicals or chunks.

Second, when using the MIC method inteaching Chinese characters, the teachertakes on the role of facilitator. To be morespecific, the following suggestions emerge:

1. Teachers need to clearly introduce theMIC method and demonstrate to stu-dents through abundant examples howthe method can aid in memorizingChinese characters.

2. When a radical or a bujian first appears ina character, teachers should highlight itso that students can store this informa-tion as nodes in a memory net and lateractivate these memory nodes whenencountering characters with the sameradical or bujian.

3. In daily lessons, teachers can instructstudents on how to connect Chinesecharacters with similar radicals or bujiantogether, thus enabling the activation ofprior knowledge, the radical or bujianthat students have learned and stored in amemory net.

4. Teachers should not impose their owninterpretation of every character onstudents as this inhibits students’ criticalthinking skills and deep processing ofcharacters. In this sense, teachers shouldencourage students to independently usethe MIC method to learn characters andprovide scaffolding to students whenneeded.

5. Students need constant exposure to andrehearsal in using the MIC method.Teachers need to provide ample oppor-tunities for students to use this method inclass as well as in self‐study at home.

Third, with regard to when to introducethe MIC method and how to provide

differentiated instruction among students,results of the study indicated that thismethod could be introduced to studentsby their second year of Chinese instruction.It is presumed that, by learning the MICmethod at an earlier stage of their acquisi-tion of Chinese characters, students couldalso benefit in their overall learning ofcharacters.

Limitations and Implications for

Future ResearchThis study has several limitations. First,from the first day of the study, studentsknew that they were participating in anexperiment. Although they remained intheir original classrooms, they had a newteacher, new material to learn, and a newstructure of lessons that only focused oncharacter learning. Second, due to thepossible interference of the method forlearning characters that was used by thestudents’ regular teacher, MIC’s long‐termeffects could not be demonstrated.

To overcome the above limitations,instead of randomly assigning students todifferent conditions, future research couldassign teachers to different treatment andcontrol conditions. Teachers in the treat-ment group could be trained on the MICmethod, and teachers in the control groupcould be directed to only use a traditionalmethod to teach Chinese characters. Stu-dents in both groups could then be trackedlongitudinally to measure their performancein learning Chinese characters. This re-search design simulates a real‐life classroomsetting and could reveal whether the MICmethod has long‐term effects on Chinesecharacter learning.

Finally, this study did not investigatethe relationship between types of Chinesecharacters and the intervention effect. Forexample, because the MIC method empha-sizes meaningful interpretation, the methodmay be more effective for transparentChinese characters, which can be inter-preted more easily than nontransparentChinese characters. Other characteristics

416 FALL 2013

Page 16: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

of Chinese characters that may influence theeffect of the intervention include number ofstrokes, frequency, and whether they areintroduced in context. Furthermore, onlysimplified characters were selected in thisstudy. As simplified and traditional charac-ters bear strong resemblance to each other,our hypothesis is that the MIC method canalso promote the learning of traditionalcharacters. However, this should be studiedto ensure that there is no differential effect ofMIC instruction between simplified andtraditional character learning.

Notes1. There are two types of Chinese charac-

ters, simplified and traditional, with theformer used mainly in Mainland Chinaand Singapore, and the latter used inTaiwan, Hong Kong, and Macau. Simpli-fied characters originated from thetraditional characters, and they bear astrong resemblance to each other. As themajority of the Chinese‐speaking popu-lation uses simplified characters as thewritten language, this study also used thesimplified characters in every aspect ofthe experiment.

2. Transparent characters refer to charac-ters where meaning and sound radicalsmatch with the meanings and sounds ofthe characters (see Appendix A, Part I,for examples of phono‐semantic com-pound characters).

3. Mandarin Chinese has four pitched tonesplus a “toneless” tone. Identical pronun-ciations that carry different tones inChinese represent different charactersand convey differences in meaning.

4. Definitions of the language levels areprovided at http://www.languagetesting.com/scale.htm#novice_low.

5. Humans are able to hold 7 � 2 items inmemory (Miller, 1956). Therefore, wechose eight characters to teach per day.

6. There were three different instructionaldifferences used across four days: teachertotal instruction (Day 1), teacher‐cuedinstruction (Day 2), and student self‐

manipulation (Days 3 and 4). Althoughon Days 3 and 4 the instructionalstrategies were both student self‐manip-ulation, the effects could be different, ason Day 4 students might feel morecomfortable and familiar with indepen-dent work after Day 3’s exposure.Therefore, Day 3 and Day 4 were treatedseparately as two levels of the variable.

References

Achor, E. E., Imoko, B. I., & Uloko, E. S.(2009). Effect of ethnomathematics teachingapproach on senior secondary students’achievement and retention in locus. Educa-tional Research and Reviews, 4, 385–390.

Alreck, P. L., & Settle, R. B. (1994). The surveyresearch handbook: Guidelines and strategies forconducting a survey. Chicago: Irwin.

Anderson, J. R. (2005). Cognitive psychologyand its implications. New York: WorthPublishers.

Atkinson, R. C., & Shiffrin, R. M. (1968).Human memory: A proposed system and itscontrol processes. In K. W. Spence & J. T.Spence (Eds.), The psychology of learning andmotivation: Advances in research and theory(Vol. 2, pp. 89–195). New York: AcademicPress.

Baddeley, A. (1986).Workingmemory. Oxford:Oxford University Press.

Boltz, W. G. (1994). The origin and earlydevelopment of the Chinese writing system.American Oriental Series, vol. 78. New Haven,CT: American Oriental Society.

Bower, G. H. (1982).The psychology of learningand motivation: Advances in research andtheory. New York: Academic Press.

Bransford, J. D., Franks, J. J., Vye, N. J., &Sherwood, R. D. (1989). New approaches toinstruction: Because wisdom can’t be told. In S.Vosniadou & A. Ortony (Eds.), Similarity andanalogical reasoning (pp. 470–497). Cam-bridge: Cambridge University Press.

Calverley, A., Grafer, B., & Hauser, M. (2002).Using instructional and motivational techniquesin the art classroom to increase memoryretention (Unpublished doctoral dissertation).Retrieved June 24, 2013, from http://www.eric.ed.gov/PDFS/ED465698.pdf

Foreign Language Annals � VOL. 46, NO. 3 417

Page 17: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

Cavallo, A. L. (1992, March). The retention ofmeaningful understanding of meiosis and genet-ics. Paper presented at the Annual Conferenceof the National Association for Research inScience Teaching, Boston.

Chase, W. G., & Simon, H. A. (1973). Themind’s eye in chess. In W. G. Chase (Ed.),Visual information processing (pp. 215–281).New York: Academic Press.

Chinese Linguistics Bureau. (2005). Moderndictionary of frequently used Chinese characters.Beijing: The Commercial Press.

Chow, A. F., Woodford, K. C., & Maes, J.(2011). Deal or no deal: Using games toimprove student learning, retention and deci-sion‐making. International Journal of Mathe-matical Education in Science and Technology,42, 259–264.

Chu, C. (2005). A reflection on traditionalapproaches to Chinese character teaching andlearning. In D. Yao (Ed.), Chinese languageinstructional materials and pedagogy (pp. 240–279). Beijing: Beijing Language and CultureUniversity Press.

Chu, C. (2009, August). The component‐oriented net‐weaving approach to characterteaching: Rationale and application. Paperpresented at the 2nd International Conferenceon Chinese Language Pedagogy, University ofCalifornia‐Berkeley.

Dictionary of Modern Chinese Characters.(2005). Shanghai: Science Press.

Ericsson, K. A., Chase, W. G., & Faloon, S.(1980). Acquisition of a memory skill. Science,208, 1181–1182.

Everson, M. E. (1986). The effect of word‐unit spacing upon the reading strategies ofnative and non‐native readers of Chinese: Aneye‐tracking study (Unpublished doctoraldissertation). The Ohio State University,Columbus.

Everson, M. E. (2011). Best practices inteaching logographic and non‐Roman writingsystems to L2 learners. Annual Review ofApplied Linguistics, 31, 249–274.

Everson, M. E., & Ke, C. (1997). An inquiryinto the reading strategies of intermediate andadvanced learners of Chinese as a foreignlanguage. Journal of the Chinese LanguageTeachers Association, 3, 1–20.

Gobet, F., Retschitzki, J., & de Voogt, A. J.(2004).Moves in mind: The psychology of boardgames. Hove, UK: Psychology Press.

Gobet, F., & Simon, H. A. (1998). Expert chessmemory: Revisiting the chunking hypothesis.Memory, 6, 225–255.

Language Learning Difficulty for EnglishSpeakers. (n.d.). In Wikipedia. Retrieved June24, 2013, from http://en.wikibooks.org/wiki/Wikibooks:Language_Learning_Difficulty_for_English_Speakers

Liu, Y. (2012). Heritage and non‐heritageChinese students’ perceptions of language learn-ing strategy use [Unpublished doctoral disser-tation]. Retrieved June 24, 2013, from https://theses.ncl.ac.uk/dspace/handle/10443/1532

Mandler, J. M., & Ritchey, G. H. (1977). Long‐term memory for pictures. Journal of Experi-mental Psychology: Human Learning and Mem-ory, 3, 386–396.

Miller, G. A. (1956). The magical numberseven, plus or minus two: Some limits on ourcapacity for processing information. Psycho-logical Review, 63, 81–97.

Packard, J. L. (1990). Effects of time lag in theintroduction of characters into the Chineselanguage curriculum. Modern Language Jour-nal, 74, 167–175.

Rastle, K., Davis, M. H., Marslen‐Wilson, W.D., & Tyler, L. K. (2000). Morphological andsemantic effects in visual word recognition: Atime course study. Language and CognitiveProcesses, 15, 507–538.

Rubin, G. S., & Becker, G. A. (1979).Morphological structure and its effect onvisual word recognition. Journal of VerbalLearning and Verbal Behavior, 18, 757–767.

Schneider, W., Gruber, H., Gold, A., & Opwis,K. (1993). Chess expertise and memory forchess positions in children and adults. Journalof Experimental Child Psychology, 56, 328–349.

Scrimgeour, A. (2012). Understanding thenature of performance: The influence oflearner background on school‐age learnerachievement in Chinese. Australian Review ofApplied Linguistics, 35, 312–338.

Shen, H. H. (2000). Radical knowledge andcharacter learning among learners of Chineseas a foreign language. Yuyan Yanjiu [LinguisticStudies], 6, 85–93.

Shen, H. H. (2005). Linguistic complexity andbeginning‐level L2Chinese reading. Journal of theChinese Language Teachers Association, 40, 1–28.

Shen, H. H. (2007). Learning vocabularythrough independent reading: Evidence from

418 FALL 2013

Page 18: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

advanced Chinese learners. In G. Andreas, X.Jiang, & Y. Wan (Eds.), The cognition, learningand teaching of Chinese characters (pp. 227–250). Beijing: Beijing Language and CultureUniversity Press.

Shen, H. H. (2010). Analysis of radicalknowledge development among beginningCFL learners. In M. E. Everson & H. H.Shen (Eds.), Research among learners ofChinese as a foreign language. Chinese Lan-guage Teachers AssociationMonograph Series,Vol. 4 (pp. 45–65). Honolulu: University ofHawai’i, National Foreign Language ResourceCenter.

Shen, H. H., & Ke, C. (2007). Radicalawareness and word acquisition among non-native learners of Chinese. Modern LanguageJournal, 91, 97–111.

Stuart, G., & Hulme, C. (2000). The effects ofword co‐occurrence on short‐term memory:Associative links in long‐term memory affectshort‐term memory performance. Journal ofExperimental Psychology: Learning, Memory,and Cognition, 26, 796–802.

Taft, M., & Forster, K. I. (1975). Lexicalstorage and retrieval for pre.xed words. Journalof Verbal Learning and Verbal Behavior, 14,638–647.

Taft, M.,&Chung, K. (1999). Using radicals inteaching Chinese characters to second lan-guage learners. Psychologia, 42, 243–251.

Taft, M., & Zhu, X. P. (1995). The representa-tion of bound morphemes in the lexicon: A

Chinese study. In L. B. Feldman (Ed.),Morphological aspects of language processing(pp. 109–129). Hillsdale, NJ: Erlbaum.

Taft, M., & Zhu, X. (1997). Submorphemicprocessing in reading Chinese. Journal ofExperimental Psychology: Learning, Memory,and Cognition, 23, 761–775.

Tulving, E., & Patkau, J. E. (1962). Concur-rent effects of contextual constraint and wordfrequency on immediate recall and learning ofverbal material. Canadian Journal of Psycholo-gy, 16, 83–95.

Wang, H. (1993). The origins of Chinesecharacters. Beijing: Sinolingua.

Wang,M., Liu, Y., & Perfetti, C. A. (2004). Theimplicit and explicit learning of orthographicstructure and function of a new writingsystem. Scientific Studies of Reading, 8, 357–379.

Wen, X. (2011). Chinese language learningmotivation: A comparative study of heritageand non‐heritage learners. Heritage LanguageJournal, 8, 41–66.

Wu, W. (1992, April). Where is linguistics inthe CFL classroom? Paper presented at theAnnual Meeting of the Association of AsianStudies Southeast Council, Athens, GA.

Submitted April 14, 2013

Accepted June 16, 2013

APPENDIX A

Meaningful Interpretation and Chunking (MIC)

Part I: Composition of Chinese CharactersNote: As the phonetic loan characters and derivative cognates make up less than 1% ofChinese characters, they were excluded from instruction and handout for the experimentalgroup.

1. Pictograms ( xiàng xíng “form imitation”)Thought to be the oldest types of characters, pictographs were originally pictures of things.During the past 5,000 years or so they have become simplified and stylized.e.g.,

2. Simple ideograms ( zhı̌ shì “indication”)Ideograms express an abstract idea through an iconic form, including iconic modificationof pictographic characters.e.g.,b�en, “root”—a tree ( mù) with the base indicated by an extra stroke.mò, “apex”—the reverse of (b�en), a tree with the top highlighted by an extra stroke.

Foreign Language Annals � VOL. 46, NO. 3 419

Page 19: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

3. Ideogrammic compounds ( huì yì “joined meaning”)In ideogrammic compounds, two or more pictographic or ideographic characters arecombined to suggest a third meaning.

4. Phono‐semantic compound characters xíng sh�eng “form and sound”)These are often called radical‐phonetic characters. A phono‐semantic character iscomposed of a phonetic and a meaning radical.

�Note: The pronunciation of does not exactly match the pronunciation of the phoneticradical , but they share the same final ai.

5. Phonetic loan characters ( jia9jiè “borrowing; making use of”)Phonetic loan characters are characters that are “borrowed” to write anotherhomophonous or near‐homophonous morpheme.

6. Derivative cognates ( zhua9n zhù “reciprocal meaning”)It may refer to characters that have similar meanings and often the same etymological rootbut have diverged in pronunciation and meaning. For example, the characters la9o “old”and ka9o “a test” derive from a common etymological root and the characters differ only inthe modification of one part.

Part II: ChunkingThe 132 most common bujian (chunks) used in 3,500 frequently used Chinese characters(Chinese Linguistics Bureau, 2005):

� 2 ¼ � 3 ¼ þ ¼lín s�en xi�u

two trees ! grove three trees ! forest a man leaning against a tree ! rest

Meaning Pronunciation Compound

stone zhu�an zhu�an “brick”vehicle zhu�an zhua9n “turn”hand b�a ba9 “hold, grasp”hand bái � p�ai “clap, hit”

Pictograph or

ideograph

Original

word Secondary word

New character for

original word

b�ei “north” bèi “back (of the body)”yào “to want” y�ao “waist”sha9o “few” sh�a “sand”

420 FALL 2013

Page 20: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

Part III: Associations Between Radicals and ChunksRead the following Chinese character sets and circle the common radicals or chunks shared bythe characters. Can you add another character sharing the same radical or chunk?

TipThe big secret here is: Meaningful Interpretation þ Chunking! Also, there is no right orwrong in using this method. Whatever works for you works best. Sometimes, you can createyour own way to memorize the character. And associating characters consisting of the sameradicals or chunks can further improve your memory.

APPENDIX B

Character Selection Criteria

Characters learned each day of the experiment were selected based on the following criteria:

1. Human Memory CapacityRationale: Humans are able to hold 7 � 2 items in memory (Miller, 1956). Therefore, wechose eight characters to teach per day.

2. Number of Strokes/Density

Rationale: According to the Dictionary of Modern Chinese Characters (2005, p. 155), thenumber of average strokes of the 1,000 most frequent characters is 7.958. Therefore,for each character set, I selected four characters at seven strokes and below, and fourcharacters at eight strokes and above. There are two characters for each category: fivestrokes and below, six to seven strokes, eight to nine strokes, and 10 strokes and above.

3. FrequencyRationale: Modern Chinese Character Frequency List (http://lingua.mtsu.edu/chinese‐computing/statistics/char/list.php?Which¼MO)Characters are all selected from Nihao IV, and most are among the 1,000 most frequentcharacters (more than 60% are among the 500 most frequent characters). In each characterset,�

Frequency top 100 and below: one characterFrequency top 101–500: four charactersFrequency top 501–1000: two charactersFrequency top 1001–2000: one character

�Note: There is an exception on day 2. On day 2, there is no character ranking beyond 1,000,but , ranking 988th, is very close to 1000th.

4. Transparency of Phonetic‐Semantic CompoundsTwo to three characters in each set are not transparent and hard to analyze (highlighted)

Foreign Language Annals � VOL. 46, NO. 3 421

Page 21: Using meaningful interpretation and chunking to enhance memory the case of chinese characters writing 6

5. Learning in contextIn each character set, there are four characters that canmake up two words (each word oftwo characters). The other four characters are introduced not in a word set, butindividually.

APPENDIX C

Quiz

Section 1: Please write down the characters you hear and their meanings in English.

Section 2: Please write down the pinyin and meaning of each character

Character Pinyin Meaning

422 FALL 2013