Maximizing Vocabulary Development by Systematically Using a Depth of Lexical Processing Taxonomy, CALL Resources, and Effective Strategies

JOHN PAUL LOUCKY
Seinan Women’s University

Abstract:
How can one better use modern CALL resources to help language learners to build up a larger target language (TL) vocabulary quickly? A major need that quickly becomes apparent at the intersection of the fields of second language vocabulary acquisition (SLVA) and computer-assisted language learning (CALL) is the lack of sufficient research with respect to their most advantageous integration for language learning. The depth of lexical processing taxonomy presented here seeks to integrate the research studies that exist into a clear and effective approach that uses insights and innovations of both fields to create a user-friendly system for more rapid vocabulary acquisition and activation. Because of the proliferation of language-learning programs and websites, it becomes important to help define which most essential vocabulary and reading strategies should be included in those programs and websites. When designing programs and websites, the three major parameters of subjective enjoyment, objective effectiveness, and technological efficiency should all be considered. This article compares various studies to suggest how many of the features and functions of CALL tools can be used at the most appropriate stages of lexical processing and acquisition to make vocabulary learning smoother, quicker, and more effective. It helps to fill a noticeable gap by providing a clear, step-by-step taxonomy of 40 vocabulary learning strategies (VLSs), simplified into an eight-fold scale of major cognitive phases for ease of instruction and use in classes. These strategies are channeled through a logical depth of lexical processing (DLP) scale, which can help promote both learners receptive and productive vocabulary development, whether using traditional text-based or CALL-enhanced methods. The ultimate goal is to find practical ways to use a research-based DLP scale and VLS taxonomy to improve the teaching and monitoring of essential phases of vocabulary learning, either in traditional classes or in CALL learning environments.

INTRODUCTION

Establishing a comprehensive and effective taxonomy of vocabulary-learning strategies useful in a CALL environment has been a research area in much need of investigation (Duquette, Renié, Laurier, 1998; Loucky, 2003c). Although many foreign language learners appear to use only a limited number of language- and lexical-learning strategies, as do most of the Japanese college students surveyed in the study presented here, teachers who wish to offer more effective instruction are looking for some system that can combine as many proven strategies, good ideas, and resources as possible to help their students more quickly improve their vocabulary knowledge, reading comprehension, and language-learning skills. This article compares various vocabulary acquisition studies and suggests how many of the functions of intelligent computer-assisted language learning (ICALL) can be used at each of the appropriate stages of lexical processing and acquisition to make the process smoother, quicker, and more effective. As such, the article helps to fill a gap by providing a clear, step-by-step VLS taxonomy divided into 10 major cognitive phases. These strategies are channeled through a logical DLP scale designed to promote learners' receptive and productive vocabulary development. The specific goal of this study is to further develop Schmitt's (1997) vocabulary learning taxonomy in ways that would make it simultaneously more teachable to students, more useful in monitoring and assessing their use of specific strategies, and more applicable to either text-based or CALL-based language learning environments. Kudo (1999) and Orita (2003) completed what seems to be an adequate replication of Schmitt's (1997) taxonomy, using several distinct groups of Japanese students at different levels of proficiency, but the students' levels of proficiency were determined only by the number of years of English study.

In earlier studies, this author has suggested a taxonomy of vocabulary acquisition which postulates that effective lexical processing of new vocabulary may be done by using a cyclical series of open-ended phases for assessing, accessing, archiving, analyzing, anchoring, associating, activating, anticipating, reassessing, and relearning/remeeting new terms (Loucky, 2003c). As new TL terms are processed and recycled through these various phases, automaticity of word recognition is developed, and, eventually, the item is retained. In itself, this cyclical process does not prove anything, but it does illustrate current theory on lexical processing strategies and provides a simple, research-based framework for more effective CALL program design and classroom instruction. The author has compared the use of many types of media and methods for enhancing the vocabulary learning of Japanese college students, investigating the relative benefits of using monolingual, bilingual, and fully bilingualized (both L1 and L2 glosses) dictionaries (Loucky, 1996, 2004a, 2005b). He has also researched the use of computerized dictionaries versus more traditional textbook or paper dictionaries for improving English lexical development and reading comprehension (Loucky, 2002a, 2002b, 2003a, 2003b, 2004a, 2005b). Finally, he compared the effectiveness of the so-called “mnemonic keyword” versus the “semantic field keyword method” (Loucky, 2004b) since the latter approach had never been developed either bilingually or tested to its full potential when used in a CALL environment. As part of that on-going study, the author undertook a study of the relationships between various patterns of strategy use, electronic dictionary use, and subsequent vocabulary growth, culminating in the DLP scale and VLS taxonomy discussed here.

LITERATURE REVIEW

Few online vocabulary development or reading courses have been created that take into account current CALL practice, SLA theory, and the practical needs and proficiency levels of language learners. The interplay among learners' language proficiency, computer proficiency, and academic needs based on their majors must be more carefully considered.

Chapelle (2001), who outlined the foundations for electronically enhanced teaching, testing, and research, noted six major criteria to consider in determining whether a CALL task is appropriate for a given language-learning situation:

1. language learning potential

2. learner fit,

3. meaning focus,

4. authenticity,

5. impact, and

6. practicality.

Chapelle provided several tables with excellent guiding questions for logically evaluating CALL tasks (see especially her Tables 3.4-3.7). The practical questions that Chapelle raised can help CALL practitioners and developers to better design and evaluate programs by focusing their attention on the six major criteria above. Both vocabulary and comprehension components of reading should be evaluated, as well as opportunities for more integrated language skill development. Well balanced and holistic language development can be encouraged either by giving students more individual computerized interaction in multimedia formats or by providing socially oriented language-learning experiences in which students are asked to apply new learning more productively. Joe (1995, 1998) showed that one of the types of learning behaviors strongly associated with promoting L2 vocabulary acquisition is generation, requiring students to generate language forms as in a “pushed output production” condition (see Swain, 1985, 1995; de la Fuente, 2002). Vocabulary learning activities that promote attention, such as the use of a vocabulary-knowledge-scale assessment (see below), recall or retrieval, and generative use in original, creative ways by students can surely help to foster more rapid second language acquisition. Chen (2004) stated that many

CALL researchers worldwide are interested in developing and using reading materials/environments to enhance second language learners' language competence … in the effect of multimedia annotation modes on reading comprehension and vocabulary acquisition (e.g. Chun & Plass, 1997; Seghayer, 2001) … in screening and arranging of authentic texts by controlling vocabulary items (Ghardirian, 2002) … in the use of electronic dictionaries or online glossing in reading processes (e.g. Lomicka, 1997; Roby, 1998; Laufer & Hill, 1999). (p. 51)

368

However, few CALL reading specialists like Chen have concentrated on the practical task of providing a supportive reading environment for both language teachers and language learners, especially online environments (however, see http://www.CALL4ALL.us, http://CALL4ALL.us///home//_all.php?fi=r) While many CALL-related articles have mainly focused on technological issues, Shield and Kukulska-Hume (2004) have affirmed that “Pedagogical and intercultural aspects of usability are often overlooked, and as a result, the contribution that content providers [teachers, researchers and web designers] could make towards the evaluation of their own web sites is not even considered” (p. 27). The arguments presented here take these factors into consideration since it is well known that higher levels of student enjoyment and engagement improve learning motivation and results.

Gallo-Crail and Zerwekh (2002) examined how 20 students of Tagalog at Northern Illinois University “used different learning strategies with different Web-based tools as they studied new vocabulary words and how this affected their success in learning and mastering the new vocabulary. … This case study describes student VLS use of five different learning strategies that were supported by Internet-based activities” (p. 56). Gallo-Crail and Zerwekh measured student achievement on quizzes and collected data on learners' VLS preferences in five kinds of vocabulary-learning activities: (a) association by word matching: a memory strategy, (b) translation by flashcard exercise: a cognitive strategy, (c) using linguistic clues via a movable online dictionary browser with sound: a compensatory strategy, (d) cultural background knowledge via weekly culture themes online: an affective strategy, and (e) reviewing and linking known to new vocabulary: a metacognitive strategy via online preview cards using sound and charts to introduce new words coming in the next weekly theme. Students made weekly study reports on accountability charts, participated in interviews and online strategy surveys, and completed online tests. Among Gallo-Crail and Zerwekh's findings most relevant to this study are the following. Learning efforts that combine “a semantic processing strategy and a keyword strategy … promote more vocabulary acquisition than a keyword or semantic mapping alone” (p. 56). The researchers also determined that learners familiar with and using a “wider variety of strategies achieved greater success in learning new vocabulary words than those who did not” (p. 65). The present study found this same pattern of relationships between higher VLS usage and higher vocabulary retention and use.

THE BASIS FOR THE PROPOSED TAXONOMY IN THE LITERATURE

It should be noted that depth of processing does not exclude the use of translation pairs as one of the initial methods of accessing and confirming new TL meanings. A translation pair has a solid memory connection and the benefit of immediate recognition, so that at early stages of language learning word-pair learning, even in decontextualized lists, can be an effective way to increase receptive vocabulary. Nevertheless, something must be done with new TL terms in order for them to be adequately assimilated into the learner's L2 mental lexicon, as well as transferred from merely passive recognition to more active recall and productive use.

Contextualized use of language by learners generating their own productive expressions seems to be required for more thorough assimilation and retention. For new TL vocabulary to move into productive use and long-term memory, it seems that some form of negotiation of meaning must take place using actual language in tasks requiring productive output of some kind. Swain (1985) and de la Fuente (2002) have called these teacher-initiated production tasks “pushed output.” Two recent studies (Loucky 2004a, 2004b) have shown very good results for both lower level and higher level proficiency language learners when pushed output tasks are blended with CALL-enhanced bilingual glosses added to arrays of related word groups using Crow's semantic field keyword approach (Crow & Quigley, 1985; Crow, 1986a, 1986b).

Different vocabulary-learning strategies seem to be more appropriate at different levels of learning and proficiency (e.g., initial learning strategies vs. subsequent consolidating strategies). Learners at beginning levels seem to benefit more from the use of bilingual dictionaries and word pairs. More advanced language learners, however, benefit from the use of both L1 and L2 tools (see Laufer & Hadar, 1997) because they are ready to process L2 explanations and use the L2 expressively. Both conventional wisdom and research studies have suggested that there are important links between language learners' vocabulary knowledge and literacy skills and their academic and career achievements. For example, Wells (1985) found that the foundation of good literacy skills is laid in the home and that good learning experiences at school can help to reinforce those skills. He concluded that the “major determinant of education achievement is the extent of a child's mastery of literacy” (p. 193).

Nation (2001) also provided a system for analyzing vocabulary-learning strategies, one that the current study has adopted in a modified form. Nation's three major categories of vocabulary strategies are

1. planning and choosing what to focus on,

(This is interpreted here as meaning what teachers should focus on and help learners to focus on because many lower proficiency language learners do not know what their initial focus or processing steps in vocabulary learning should be.)

2. finding sources of information about vocabulary, and

(This is interpreted here as accessing and elaborating new word meanings.)

3. processes for establishing vocabulary knowledge.

(This is interpreted here as consolidating vocabulary knowledge.)

Sanaoui (1992, 1995) compared both ESL and EFL students and found clear distinctions between structured versus unstructured learners, noting that these two groups of learners differed in five important ways:

1. learners' opportunities for vocabulary learning (i.e., independent study vs. reliance on course [and teacher] direction),

2. learners' range of self-initiated vocabulary learning (i.e., extensive vs. restricted ranges),

3. learners' records of lexical items they were learning (i.e., extensive/systematic vs. minimal/ad hoc),

4. how much learners reviewed their lexical records (i.e., extensive vs. little or none), and

5. whether learners practiced lexical items (creating opportunities for use in and out of class vs. relying on class instruction only).

Other studies have also concluded that electronic dictionaries were superior to print dictionaries mainly based on reduced search time (e.g., Taylor & Chan, 1994; Koga, 1995; Inami, Nishikata, Nakayama, & Shimizu, 1997; Loucky, 2002a, 2002b, 2003b, 2003c).

Clearly, the strategic choice between dictionary use and other VLS options is an important skill for L2 readers to develop. So far, however, few studies have considered how CALL could help to raise awareness and practice in using a much broader range of effective and essential VLSs. The present study addresses this question.

THE STUDY

There are as yet few studies examining what specific kinds of vocabulary learning strategies L2 students employ when using CALL materials and how the use of these strategies may differ in quality or quantity from those used when reading print materials. To contribute towards an understanding of online reading and vocabulary development, the author developed a taxonomy of vocabulary learning strategies based on Schmitt's (1997) useful distinction between discovery versus consolidating strategies. This 40-item VLS taxonomy gives brief descriptions of strategies used at the discovery and consolidating stages of vocabulary development (see the modified version of Schmitt's survey in Appendix A and the results of its use in Table 3 below). Besides studies investigating the effects of hypermedia on vocabulary development, (e.g. Akbulut, 2004a, 2004b; Coll, 2002), few detailed CALL or electronic-dictionary-based studies of VLS use have been done, with the exception of the studies by Hill and Laufer (2003) and Kobayashi (2005).

Optimal conditions governing strategy use in traditional versus CALL environments need to be further developed using various computer functions and online features to encourage learners to include essential phases of lexical processing. The author undertook the present study to better assess Japanese students' use of VLSs based on the modified version of Schmitt's survey and a DLP checklist derived from a combination of these VLSs and an eight-phase DLP scale.

METHOD

Procedures and Participants

The project involved 112 students in five groups distributed over a 5-year period of time. The students in groups 1-4 were Japanese freshmen engineering students at a national university in Kyushu between 2000 and 2004. Group 5 consisted of engineering graduate students enrolled in a one-semester-only course at the same university. While most Japanese students have completed 6 years of English study at the secondary level before entering the university, the students in this project averaged 7.32 years of English study, including after school conversation and cram school English training. Table 1 lists the number students in each group and their vocabulary grade level, as assessed by a US-normed standardized reading test (Gates McGinite). The author has found consistent results in using these levels to estimate Japanese high school and college students' reading and vocabulary levels over two decades (Loucky, 1996, 1997a, 2002c, 2003a).

Based on the results of this pilot study, Schmitt's original survey, the modified 40-item VLS taxonomy form, and the short VLS checklist (Part C in Tables 2 and 3) were administered to students in groups 2-5 in the following years. Students in each group were given instructions in a teacher-made workbook on how to monitor

377

their use of VLSs and were encouraged to use as many of them as possible in the lexical processing stages listed in the tables above whenever they encountered new words during their readings. They completed the VLS taxonomy during the academic year and the short VLS checklist at the end of the year. Students were asked to total and compute the percentage of their VLS use (2.5 points per VLS used X 40 = 100%) and then to take the shorter DLP checklist in Part C in Table 2 and summarize their overall use of its eight phases (x/8 actively used versus x/8 thought to be helpful).

RESULTS

The responses of students in Groups 2-4 (N = 58) are reported in aggregate form in Table 4 representing the VLSs that the students used most often or found most helpful. In the cells in Table 4, the first number (e.g., 44 Ls) refers to the number of learners who reported using that strategy or finding it helpful. The second number (e.g., SS VLS 8) refers to the strategy number in Schmitt's survey. The final number (e.g., DT 7, meaning item 7 on the author's DLP taxonomy questionnaire) refers to the comparable strategy found in the author's VLS taxonomy, based on its DLP scale grid number. The results presented in Table 4 mirror many of findings of Schmitt (1997), Kudo (1999), and Orita (2003), among others. Schmitt and Orita found that the most frequently used VLSs for discovery were guessing from context, using a bilingual dictionary, and asking classmates. Likewise in this study, these two VLSs were the most highly preferred strategies for learning new words. The preferred VLSs were, in rank order, (a) using bilingual dictionaries (VLS 8 used by 44 learners), (b) guessing from context (VLS 13 used by 36 learners), and (c) using monolingual dictionaries (VLS 7 used by 35 learners). For consolidation, the preferred VLSs were (a) studying new words often over time (VLS 27 used by 22 learners) and writing new words out (VLS 26 used by 19 learners).

DISCUSSION

In this study, most of these Japanese college students showed a low average use of VLSs, except for students in the most advanced class (the graduate students in Group 5). Despite having had an average of 7.32 years of previous study in English, only about half of the students (52%) reported using even the first three of the eight stages of vocabulary learning. Although three fourths of the students (74%) thought that each of the eight types of VLSs would be useful in their language learning, their average use of all eight steps was only 37%.

Students in the lower proficiency level classes used, on average, only 26% of the 40 VLSs. The most widely used strategies were accessing unknown words in Japanese (62%, compared to only 20% for accessing unknown words in English), asking classmates for help or using an English-Japanese book dictionary (50%), archiving/recording meanings of unknown words or simply skipping over unknown words (45%), trying to guess meanings from context (40%), using monolingual dictionaries (37%), and using English to Japanese electronic dictionaries (35%, compared with 25% for using Japanese to English electronic dictionaries). The other VLSs were used by under 30% of the lower proficiency level students.

The graduate students in group 5 used a slightly higher percentage of VLSs than the lower proficiency undergraduate students, 32% versus 26%. However, the most proficient undergraduate students (Group 3), who had the highest average use of electronic dictionaries, also had the highest average use of VLSs (41%). The students in this group also had an average vocabulary gain of 1.07 over the academic year, the highest vocabulary gain observed by the researcher in the past decade at this institution. Students who used electronic dictionaries consistently and diligently generally reached higher levels of English proficiency than those who did not (Loucky, 2005c). A more detailed analysis of dictionary use by the students in Group 3 revealed that 11 out of 21 learners surveyed (52%) used electronic dictionaries (2 bilingual only and 9 fully bilingualized—with both bilingual and monolingual lexicons), whereas 6 students (29%) used only bilingual print dictionaries. Four other students (19%) did not bring a dictionary to class; they each borrowed a monolingual learner's dictionary of their choice from a class collection. Some students carried both print and electronic dictionaries to class but tended to use their electronic devices for easier, more rapid access to unknown words. These results seem to indicate that systematic use of electronic dictionaries can help foreign language learners to focus more time and attention on essential VLSs, especially regularly reassessing, recycling and reusing new terms. This is because quickly accessing new word information and automatically saving it for review are two of the major advantages of portable electronic dictionaries.

Schmitt (1997), Kudo (1999), Orita (2003), and the findings of this study show that most Japanese students, regardless of the number of years of English language learning, use relatively few strategies when learning new TL vocabulary, suggesting that these students have not had effective training in VLSs. Few secondary English teachers in Japan seem to be aware of VLS taxonomies to help students learn more effective strategies for discovering the meaning of unknown TL vocabulary or for consolidating their knowledge of TL vocabulary.

While about half of all students reported using the first three types of VLSs, discovery strategies following Schmitt's model, less than one third of the students (28%) made an effort to use any of the other five types VLSs. This means that 72% of the students failed to use more cognitively or communicatively demanding processes, namely any of those VLSs characterized by lexical phases 5-8 in the DLP scale. Only the most advanced students, who routinely used electronic dictionaries, reported using the greatest number of VLSs.

More research should also be done on the beneficial language-learning effects of using various CALL reading programs (see Chen, 2004; Loucky, 2003d, 2003e) along with consistent and well guided use of computerized bilingual and/or monolingual dictionaries (see Loucky, 2002a, 2002b, 2003b, 2003c), especially when these dictionaries are systematically used to enhance the most essential lexical-processing steps elucidated by the DLP scale.

PEDAGOGICAL IMPLICATIONS

Even though a taxonomy of essential vocabulary processing strategies can be easily taught, as the author did with the students in Group 1, effective vocabulary learning strategies must be practiced and actively used to become regular, independent language-learning habits. One can apply Sanaoui's (1992, 1995) distinction between “structured” versus “unstructured” to evaluate learners' approach to vocabulary learning since their approach clearly affects the resulting degree of lexical learning.

PEDAGOGICAL IMPLICATIONS

Even though a taxonomy of essential vocabulary processing strategies can be easily taught, as the author did with the students in Group 1, effective vocabulary learning strategies must be practiced and actively used to become regular, independent language-learning habits.

Constructing an Ideal CALL Environment

Based on the considerations discussed above, the specific learning factors should be clearly distinguished for better analysis of their respective roles in second-language acquisition (SLA) in general and second-language-vocabulary acquisition (SLVA) in particular when developers construct an electronic text or CALL program. The following factors can serve as useful guidelines for constructing an ideal CALL environment to maximize SLA and SLVA:

1. inclusion of learning style preferences (often due to the how individuals have learned their first language and its linguistic characteristics, e.g., Kanji versus alphabetic orthography),

2. vocabulary learning strategies (e.g., those required by the teacher/text or imposed by the CALL program vs. those chosen by learners),

3. means of glossing (e.g., L1 translations, L2 definitions only, or fully bilingualized combinations of both L1 and L2 information, perhaps including full sentence examples of major meanings [Laufer & Hadar, 1997; Loucky 2004b] and still picture or video annotations),

4. language-learning tasks or factors included in language-learning tasks (e.g., quantity, richness, and degree of organization of new TL vocabulary),

5. depth of lexical processing or more elaborative cognitive processing (see Craik & Tulving, 1975),

6. attractiveness (e.g., quality of learner interface and ease of use),

7. degree and type of language interaction (e.g., interaction required or prompted by the program),

8. availability of technical functions (e.g., pacing of voice, text screens, types of semantic and syntactic elaboration, and multimedia links),

9. degree of learner motivation, interest, and purpose (i.e., how these are encouraged in the program), and

10. cost effectiveness (e.g., students' preference for free online or downloadable programs, see examples at http://www.CALL4ALL.us).

Strategies Instruction

The use of both a DLP scale and a VLS taxonomy acquisition model would enable teachers to focus on training their students to use effective strategies to develop fluency in the TL. These instructional strategies should clearly include teaching students how to

1. assess their vocabulary knowledge and learning by means of some kind of vocabulary knowledge scale such as the one proposed by Loucky (2002a, 2005a) (see also Loucky, 2003b, 2004d; Wesche & Paribakht, 1996),

2. use intelligent, integrated CALL tools such as computerized bilingual dictionaries of various sorts, to enhance each phase of lexical processing (e.g., rapidly accessing new word meanings and recording lexical information in an organized way),

3. use a DLP scale such as that proposed here to guide them to greater variety and depth of both word and text processing (Craik & Tulving, 1975),

4. use CALL software and online programs that include a greater number of multimedia features as well as more lexical-processing steps (see Takefuta, 1999),

5. identify and focus on high-frequency, common-core vocabulary (Nation & Newton, 1997; Loucky, 2002c) and academic words (Coxhead, 2000),

6. maximize active practice of TL terms in productive development and communicative interactions in computer-mediated or face-to-face sessions (Loucky, 1997b, 1998; Hatch & Brown, 1995; Nation, 2001; Thornbury, 2002), and

7. systematically organize new TL vocabulary to build up a broad web of associative networks by means of a semantic field keyword approach (Crow, 1986; Quigley, 1986; Loucky, 2004a, 2004b, see also http://www.CALL4ALL.us/SFKAINDEX.htm).

In sum, language learners at an early acquisition level should focus on phonetic-word decoding and sight-word learning and, at later intermediate and advanced stages, on developing those strategies most needed for discovering new meanings from context based on predictions from known words. A consistent finding from observing the students in this project is that the higher a class's average vocabulary level is, the greater the number of students who use electronic dictionaries and CALL resources to improve their learning (Loucky, 2002a, 2002b, 2003b, 2003c, 2004a, 2004b). From the early intermediate level, language learners should be trained in how to effectively use more rapid-access computerized tools and instant online glossing, starting at word and phrase levels with L1 translations and moving on to sentence and discourse levels of complexity with L2 glosses. Consistent use of VLS taxonomies like those reviewed here and systems that encourage (a) greater depth of lexical processing, (b) wider breadth of syntactic complexity, and (c) repeated encounters with new TL forms and meanings in as many different contexts as possible can often be significantly facilitated by the rapid access provided by portable and online CALL dictionaries and translation software (Saunders, 2002; Coll, 2002; see also http://www.call4all.us///home/_all.php?fi=d).

CONCLUSIONS

The 40-item VLS taxonomy organized by phases in the DLP scale presented here consolidates Schmitt's (1997) original vocabulary-learning-strategies taxonomy, while, at the same time, adding both the semantic field keyword approach and the use of portable electronic or online dictionaries (distinguishing bilingual from fully bilingualized dictionaries). The DLP scale not only preserves the helpful distinction between initial discovery and subsequent consolidating strategies, but also adds a useful analytical grid of the major phases of lexical processing to help clarify SLVA processes.

The author's review of the limited number of VLS classification systems available—each with different parameters and means of organization—led him to the conclusion that a descriptive, learner-centered approach has not helped to provide a sufficiently clear and complete system that can be easily grasped in order to be taught, learned, and used for actual language learning and research. Thus, this VLS taxonomy and DLP scale sought to combine insights from previous studies to help language teachers and language learners maximize vocabulary development. Koyama and Takeuchi (2004, p. 41) are correct in stating that “With regard to foreign language acquisition … task-induced involvement might be more important than look-up frequency (Laufer & Hulstijn, 2001). Look-up frequency, thus, does not necessarily result in better learning.” Although Koyama and Takeuchi claimed that “words searched in a longer process in PD [print dictionary] condition could be retained better than those in ED condition” (p. 42)—supposedly supporting the depth of processing hypothesis—merely increasing learners' time on task does not necessarily ensure deeper or more elaborative cognitive processing.

We should examine not only the amount of time on task and number of words accessed, but also the quality of lexical processing being employed by the systematic use of more thorough elaborative vocabulary learning strategies. A comparison of such studies leads us to conclude that dictionaries, whether print or electronic, are merely tools whose features and functions require training for effective use. Further, language learners have different profiles, needs, and preferences, so that various kinds of dictionaries and various kinds of information presented in dictionaries have different benefits for learners at different levels of proficiency. While print picture dictionaries are often best for younger, beginning-level learners, older learners with computer skills may be far more motivated to use CALL-based lexical resources. Deeper lexical processing seems to result in better retention, but only if other essential learning phases and strategies are included such as archiving, anchoring, and activating. These findings show where CALL can help

learners by providing more specific focus and elaboration on new word forms, meanings, and appropriate usage in each phase of vocabulary development