Developing Orthographic Awareness Among CFL Learners:

What the Research Tells Us


 Michael E. Everson

     The University of Iowa




            The title of our conference, “How Western Learners Discover the World of Written Chinese” is apt, as it highlights the fact that foreign language learners of Chinese must come to terms with aspects of a new writing system that are very different from the alphabets that make up the writing systems of their first languages (L1).   How and what they discover about the written language has been of increased interest to scholars of second language (L2) acquisition and foreign language education, perhaps mirroring similar interests from cognitive scientists, developmental reading specialists, and early childhood educators in how Chinese children learn to read in their L1.   Whether this reflects the notion that we are living in an age of globalization where reading Chinese will be important during “China’s Century” (Newsweek, 2005), or whether we have determined that understanding the reading process in Chinese can help us understand more about the reading process in general, this conference indicates that international scholarly cooperation can go far to help us in understanding the L2 reading process.

            In trying to chart this discovery process that fascinates us as scholars and our learners as students, perhaps a good place to start is to determine what it is exactly that our students discover about the Chinese writing system when they begin to study Chinese, as well as when they progress to more enhanced proficiency levels.  More specifically, the notion that most Chinese characters provide both phonetic and semantic information has long been an anchoring point for Chinese pedagogy, but how this information is used, valued, and demonstrated by Chinese learners needs to be better understood.  This paper, then, will discuss the findings of several Chinese-as-a-foreign-language (CFL) studies that provide different types of research evidence pertinent to our understanding of this process.  In addition, it will also discuss the research methodologies that have been used to investigate it, for the methodologies have been very different in how they try to get at this problem.  This paper will then conclude with some suggestions about how we might expand our research horizons so as to better understand this important phenomenon.

Research in Chinese as L1


In spite of the fact that Chinese characters have long had the reputation of being “ideographs” or “pictographic writing” in traditional descriptions of the writing system, more and more theory development and research indicates that the semantic and phonological elements present in so-called “compound characters” are important factors in helping Chinese children access both semantic and phonological processes in reading.  This is especially significant in that it is believed that compound characters make up approximately 80-90% of modern characters (Taylor & Taylor 1995), though the precision with which sound and meaning can be retrieved is inexact.   Moreover, L1 research involving Chinese children indicates that they develop this awareness of the orthographic principles operating in Chinese characters throughout their elementary school years, with this ability predicting later reading proficiency.  Shu and Anderson (1999) describe that learning to read requires “becoming aware of the basic units of spoken language, the basic units of the writing system, and the mapping between the two,” so the development of this awareness seems paramount for Chinese children in their literacy development. 

How this orthographic awareness develops among adult L2 learners of Chinese whose L1 is alphabetic is less clear, given the fact that these learners differ from Chinese children in a number of ways.  First, L2 learners have either no command of the spoken language (or perhaps only a rudimentary one) when they begin to learn Chinese characters.  Indeed, many programs that strictly delineate the development of reading and writing skills view their development as somewhat unrelated and consequently use pedagogical strategies that separate the two instead of trying to teach the skills as being related.  Secondly, as beginning learners have no notion of how the Chinese character system operates in terms of the principles by which Chinese characters are constructed, “brute force” memorization techniques are often employed until these learners begin to understand that there are some principles that can be applied to help them learn the characters that are based on principles of how the characters are constructed.   Third, Chinese children spend a great deal of their time during elementary school learning to read, with large amounts of characters being taught to them during each successive grade.  When used with Western learners in a typical university setting where learners do not have the reinforcement of environmental print surrounding them everyday, traditional pedagogical techniques of massive character inundation often serve to overwhelm and consequently demotivate our students, thus leading to heavy student attrition.  This is especially true if young adults are taught characters without some sort of system that can help them take advantage of the structure of the writing system, leaving them to cope with memorizing Chinese on a rote-memory basis.  This last point becomes especially important because it makes a tacit assumption that our learners can “figure out” how to memorize characters on their own in an efficient and meaningful manner, an assumption that is unwarranted.  The problem, then, is that we not only expect our students to come to terms on their own with how the Chinese writing system works, but also to develop study management and organizational skills.  When viewed this way, we are saying that students should be able to develop orthographic and metacognitive abilities that, in my opinion, we woefully overestimate.

Given these problems, I am happy to see that many of the papers that will be given at this conference deal with many of the problems that I have noted above.  For this paper, I would like to concentrate on the research and the various methodologies used in this research that focuses on how L2 learners develop this orthographic awareness which scholars believe is a critical component for L1 learners to develop for reading success.

Survey Research 

            One of the more popular research methodologies is that of survey research, which has as its basic premise that if you want to know how learners do things, the best way to find out is to simply ask them.  Surveys can take on many different forms, from highly analyzed questions put to rigorous statistical analysis, to open-ended questions where those surveyed can explain their answers in more depth.

            Two important studies were done during the ‘90’s that were interested in finding out what beginning students did when faced with the challenge of learning Chinese characters.  McGinnis (1999) conducted a study among learners in an intensive nine-week program, with him discovering that students used a variety of strategies when it came to learning characters.  These strategies not surprisingly included rote repetition, creating idiosyncratic stories about how the characters looked or how they were pronounced, and using the character’s semantic or phonetic information in the character’s components.  Interestingly enough, this latter strategy was not one that was favored by the students, who instead favored making up stories or memorizing the characters by rote means. 

            Another study (Ke, 1998) also sought to investigate the learner strategies of beginning Chinese students, and found that learners indeed value the use of character components in learning the characters; however, they stated that learning the characters wholistically through repeated writing of characters was an important strategy, especially practicing characters in terms of the two-character compounds that comprised much of their vocabulary, rather than practicing characters individually.  Perhaps the most important finding was the finding that there seemed to be a relationship between valuing and understanding of character components and character recognition and production. 

            Although these findings may seem contradictory, it is important to view this within the context of developing proficiency.  The McGinnis (1999) study was conducted among learners who had just been studying Chinese for four weeks, while Ke’s (1998) study involved students who had been learning Chinese for a year.  Clearly, the time-on-task variable is noticeable in that the latter group of students were able to study enough characters to where they had developed a beginning sense of orthographic awareness, a sense that seemed to be related to their proficiency in identifying and producing characters.  This is important evidence in that it may indicate that learners must be exposed to enough characters to where they see examples of different components used in different ways, thus becoming aware of a system used in Chinese orthography, as opposed to a feeling of randomness that may require them to learn characters by rote for a specific period of time.  In fact, Ke’s (1996) model of orthographic awareness speaks to this idea, thus predicating that learners perhaps develop certain thresholds of character understanding, the absence of which might make much of our teaching meaningless.

            To carry this investigation further using more precise statistical analysis and learner samples of greater proficiency, Shen (2005) designed three different survey instruments to collect strategy data over three samples of learners at the university level (1st, 2nd, and 3rd year).  Her first survey questionnaire contained open-ended questions to elicit from learners the types of strategies they used for learning characters so that a more thorough inventory of strategies could be constructed.  A second questionnaire using a Likert-scale was developed to determine how frequently the learners used the strategies in the inventory.   A third questionnaire was developed to rate the students perceptions about how useful commonly used strategies were across the different learning levels.

            Shen’s (2005) research differed somewhat from the studies mentioned previously in that her goal was to detect a commonality and set of patterns among the strategies that learners indicated they used.  Factor analysis was used to achieve this, with a total of eight factors being derived from the 30 commonly used strategies as chosen by students.  Importantly, the most heavily loaded factor was that of the use of orthographic knowledge, or using graphic structures, connecting with previously learned characters, visualization of graphic structure, and using semantic and phonetic radicals.  A total of 24.5% of the variance was explained by this factor, far more than even the second most commonly chosen factor, that of preview and review.   A second important finding of the study was that among cognitive strategies, learners considered those represented in Factor 1 to be most useful, with this perception increasing as learner proficiency increased. 

            These three studies when viewed together seem to indicate that the understanding and use of the principles that make up the Chinese writing system are not only good to know just for background information, but that they actually become part and parcel of reading development.  Moreover, it might just be that this knowledge is integral for our learners to have if their proficiency is to increase. 

Multiple Choice Selection

            Another way to discover whether or not learners understand the cues provided by the semantic and phonetic elements of a character is to ask them to guess the meaning or sound of characters that they have never studied before.  In one such study, Jackson, Everson, and Ke (2003) used multiple choice questions to determine this, telling the students to circle the choice in English that best represented the meaning of the character.  The subjects were also told that if they did not know the character, they should do their best to guess.   Among the items in the vocabulary test were characters that were free-standing radicals—it was important to include these so that we could see if the subject was able to identify certain radicals as free-standing characters.  Thus, a question such as the following was included:


The character      means ...


a.  sun              b. *fire             c. moon   d. water



In the test, aside from giving the subjects characters that were present in their curriculum, we gave them questions with characters they did not know, such as:



The character    means to...


a. discolor        b. scratch         c. *burn             d. shred


In this way, we could go back through the individual tests to infer “transfer,” that is, the ability to correctly identify the meaning of the free standing radical, then use it to infer the meaning of an unknown character that contained that radical. 

            Similarly, we were interested to see whether learners could use the phonetic elements of characters to guess the pronunciation of characters they had never studied before.   In a similar fashion, we would test an item to make sure that the subject knew how a free standing phonetic could be pronounced, such as


The character   is pronounced...


(write in correct pinyin)__________________________



and then later in the test among items testing characters that the subjects had already studied, we would place a question such as


The character     is pronounced...

(write in correct pinyin)__________________________


As in the items containing semantic radicals, the purpose of this testing strategy was to determine whether the learners could infer the pronunciation of a novel character from a phonetic component that they had already demonstrated they knew how to pronounce.

            The results of the study indicated that as a group, the learners were able to transfer their knowledge of both the semantic and phonological elements of characters they knew to characters they did not, though they performed better with the semantic elements.  It should also be noted that there was a high degree of variability in both tasks among the learners, indicating that these abilities are neither straightforward nor uniform among beginning learners of Chinese.  The data also indicated that when the familiar radical was used in a way to change its position from the left side of the novel character, or when the radical appeared changed in shape or proportion in the novel character, the subjects sometimes were not able to apply this transfer ability. 

            Another finding from the study that was extremely revealing had to do with our classroom observations of the first year Chinese class from where these subjects were drawn.  Data collected by two independent observers indicated that the instructor over time spent more time explaining and dealing with the semantic radicals than with the phonetic elements of characters.  This brings to mind, then, the issue of whether the use of the semantic and phonetic elements of the characters can be taught in such a way during class time as to enhance the student’s ability to remember them.

A Classroom Intervention Study

               One study that attempted to do that was Shen’s (2004) where she tested whether using character presentation strategies would have an effect on short and long term retention of new two-character words.  In this study, Shen would introduce 10 new two-character words in three different sessions over three days under one of three conditions to three separate groups of second-year Chinese language learners.  In all conditions, the ten words were shown on an overhead projector with their English meanings and pinyin pronunciations. In one condition, termed rote memory, the students were repeatedly introduced to the words by the instructor through presentation and pronunciation of the pinyin and English meaning through the use of an overhead projector, with the student asked to name the sound and meaning.  Under the condition of self-generated elaboration, students were told to use whatever strategies they wanted to learn the words, and to write down in English what these strategies were; in the last condition, instructor-guided elaboration, the researcher introduced the words using etymological aids, use of radicals, and used the words in context.   After a 20-minute learning period, the students handed in their lists of words.  This activity was followed by a 20-minute grammar review period, where the newly acquired target words were not reviewed or studied.  After this period, the students were given the list of new two character words they had learned during the first 20 minute learning period and asked to write their pronunciation in pinyin and their meaning in English.  The students were also required to perform this same task with these same two-character words 48 hours later.

            Among the many findings of this study were that rote memory learning resulted in consistently less retention of the sound and meaning of characters than the two elaboration conditions.  Regarding the two elaboration conditions, retention of sound and meaning of the characters was superior for the instructor-guided elaboration condition than for the student self-generated condition for the first 20-minute interval of retention, but the advantage dissipated for retention at the 48-hour period.  Shen (2004) theorized that the processing of the characters was deeper under the instructor-guided condition because it draws more heavily on concept-driven processing.  The elaboration required in concept-driven processing requires students to associate prior knowledge of orthographic structure with the new items in the vocabulary list, thus promoting the creation of new and unique information based upon knowledge already stored in memory.  This theory holds that the students often are unable to access this old information by themselves, thus depriving themselves of its access.  Through instructor-guided elaboration, however, students are able to receive consistent instruction about the various orthographic principles of Chinese characters, as well as how this knowledge can be linked and elaborated when learning new characters.

Think-Aloud Protocols

            Up until now, the research methodologies used to determine our learners’ strategies have relied largely on some basic survey methodology, as well as the creation of strategy inventories to determine the factors that these strategies seem to comprise.  Another method that has been used is the use of think-aloud protocols, a type of method whereby the learner talks about what he/she is actually doing while they are reading, so that the researcher can obtain a window into their problem solving strategies as they make their way through a text. 

            In applying this methodology to Chinese learners, Everson and Ke (1997) were interested in seeing what Chinese learners verbalized when they were reading a portion of a Chinese newspaper text.  They used two samples in their study, one composed of learners that had completed two years of university Chinese and one group that comprised two learners of Chinese that were advanced graduate students, each having studied Chinese for about five years.  Among the many finding in the study were that the learners who had studied Chinese for two years still experienced problems dealing with their lack of vocabulary, as well as issues dealing with word knowledge and word formation.  The advanced learners, on the other hand, were particularly skillful in figuring out words and characters that they did not initially know.  For the purposes of this study, it is important to note that the data clearly show that the advanced learners consistently attempted to pronounce unknown characters based on the phonetic element that was present, and in some cases, used the semantic element to identify certain classes of characters.  Given this impressive use of breaking down characters into their constituent elements to try to solve word recognition problems, advanced students were able to rely on the different elements of the characters as trusty sources of information.

 Initial Summary Conclusions

In conclusion, different types of research methodologies have given us some unique insights into the development, valuing, and use of orthographic awareness among CFL students.  The more prominent conclusions that can be reached from the research are that orthographic awareness:




Future Directions and Needs


            Given the number of pressing issues in research that need to be investigated, I’ll just speak to a few that come to mind that impinge upon the topic of orthographic awareness, and hope that our conference will further solidify ideas for a focused research agenda in the area of CFL reading.  First, I would like to say that we need a better understanding of vocabulary acquisition that develops among our learners.  All of us at this conference, and all the CFL teachers who teach Western learners know that our students go through a process of vocabulary acquisition that differs markedly from students who are learning French, German, and Spanish.   It is good to see conferences such as this with papers being delivered on the role of romanization in CFL acquisition, the role of the spoken language in learning to read, and ways that advanced ideas can be put into the classroom for enhanced pedagogy.  Both L1 and L2 reading have been very upfront about the huge contribution that vocabulary makes to the reading process, so orthographic awareness and the many other variables that contribute to vocabulary knowledge need to be understood and documented in a more thorough way.  This would be an excellent start to attempt to model the many variables that are responsible for reading in CFL.

            Secondly, I think it is important that we broaden our theoretical perspectives to see what other traditions of research have to offer our knowledge base.  Many of us come from traditions of research that emphasize quantitative data analysis techniques.  While this is my background in training and advocation, I must say that I have learned a great deal from research traditions that describe the importance of the setting in which Chinese is learned, as well as exploring and describing the process of learning that is going on in the classroom among the students and the teacher.  Exploration, therefore, using theoretical approaches that attempt to capture the co-construction of meaning are worthwhile to pursue.

           Thirdly, I have always thought it interesting that there is not more research ongoing into what our textbooks are teaching our students in terms of characters, as well as approaches to learning them.  I think more research of this nature needs to be carried out.  All too often, our claims about Chinese reading use as a basis the notion of “Chinese,” meaning all that is possible in the language from a native speaker standpoint.  In reality, however, our learners are getting only a very thin slice of this reality, bound by the characters they learn from their textbooks.  I therefore was happy to see that there are papers that deal with this notion of not only Chinese characters, but which ones seem to be more important to learn, and why.  In looking at our textbooks, we therefore need to investigate the characters we include that are students actually see, as these are the ones that they will begin to draw their conclusions from that develop (or hinder) their orthographic awareness.  Consequently, topics such as character frequency, number of compound characters introduced, etc., need to be further analyzed.

Lastly, as researchers we need to make sure that our presence is known in CFL efforts in our regional, national, and international organizations, as well as in projects that will contribute significantly to our research base.  For instance, proposals are now being taken in the United States for a Chinese K-16 Pipeline Project (for more information, download from the ACTFL website at  The objectives of the project are to: (1) establish a Chinese Flagship program that addresses the needs of students already at the advanced proficiency level and (2) work closely with one or more geographically proximate elementary/middle/high school systems to establish an articulated Chinese language program that progresses from the elementary grades into advanced Chinese at the university level.  The final outcome of the program is expected to be students with superior (3) level proficiency in Chinese.  This project is exciting for a number of reasons.  First, it gives all participants in this project, from curriculum designers to language researchers, a unique opportunity to evaluate longer sequences of language instruction, something that typical programs in American schools have traditionally been unable to accomplish.  Secondly, it provides a unique opportunity for a variety of stakeholders to participate jointly, thus bridging some of the fault lines that have appeared in the articulation between elementary schools and high schools, high schools and higher education, as well as the teachers, researchers, and administrators who run these programs.  Lastly, it emphasizes the need for the program to be based on solid second language research evidence, and thus provides a laboratory for more of this research evidence to be gathered and reported.  Other U.S. initiatives such as the introduction of the Chinese AP curriculum and testing for high school students, as well as the strategy building ongoing to have 5% of U.S. high school students studying Chinese are all endeavors that must include a rigorous research component.  By taking advantage of these opportunities for further research and sharing the findings at conferences such as this, we will definitely become more informed about how Western learners discover the wonder of the written world of Chinese.




Everson, M. E. & Ke, C. (1997). An inquiry into the reading strategies of intermediate and advanced learners of Chinese as a foreign language. Journal of the Chinese Language Teachers Association 32[1], 1-20.

Jackson, N. E., Everson, M. E., & Ke, C. (2003). Beginning readers' awareness of the orthographic structure of semantic-phonetic compounds: Lessons from a study of learners of Chinese as a foreign language. In C.McBride-Chang & Chen Hsuan-chih (Eds.), Reading development in Chinese children (pp.141-153).  Praeger Publishers.

Ke, C. (1996). A model for Chinese orthographic awareness. Unpublished manuscript.

Ke, C. (1998). Effects of strategies on the learning of Chinese characters among foreign language students. Journal of the Chinese Language Teachers Association 33[2], 93-112.

McGinnis, S. (1999). Student goals and approaches. In M.Chu (Ed.), Mapping the Course of the Chinese Language Field (pp. 151-188). Kalamazoo, MI: The Chinese Language Teachers Association, Inc.

Shen, H. H. (2004). Level of cognitive processing: Effects on character learning among non-native learners of Chinese as a foreign language. Language and Education 18[2], 167-183.

Shen, H. H. (2005). An investigation of  Chinese character learning strategies among non-native speakers of Chinese. System 33, 49-68.

Shu, H. & Anderson, R.C. (1999). Learning to read Chinese: The development of metalinguistic awareness.  In Wang, J., Inhoff, A.W., & Chen, H.C. (Eds.), Reading Chinese Script: A Cognitive Analysis.  Mawah, NJ: Lawrence Erlbaum Associates, Publishers.

Taylor, I., & Taylor, M.M. (1995). Writing and literacy in Chinese, Korean, and Japanese.  Philadelphia: John Benjamins.
            Zakaria, F. (2005, May). Does the future belong to China? Newsweek, CXLV, 19, 28-47.