Acta Psychologica 159 (2015) 116–122

Contents lists available at ScienceDirect

Acta Psychologica journal homepage: www.elsevier.com/ locate/actpsy

Proficiency and sentence constraint effects on second language word learning Tengfei Ma a,b,c, Baoguo Chen a,b,⁎, Chunming Lu b, Susan Dunlap d a

Beijing Key Laboratory of Applied Experimental Psychology, School of Psychology, Beijing Normal University, Beijing 100875, China State Key Lab of Cognitive Neuroscience and Learning, Beijing Normal University, Beijing 100875, China School of Education, The Open University of China, Beijing 100039, China d Children's Learning Institute, University of Texas Health Science Center at Houston, United States b c

a r t i c l e

i n f o

Article history: Received 17 December 2014 Received in revised form 23 May 2015 Accepted 25 May 2015 Available online xxxx PsycINFO Codes: 2340 Keywords: Second language Word learning Proficiency Sentence constraint

a b s t r a c t This paper presents an experiment that investigated the effects of L2 proficiency and sentence constraint on semantic processing of unknown L2 words (pseudowords). All participants were Chinese native speakers who learned English as a second language. In the experiment, we used a whole sentence presentation paradigm with a delayed semantic relatedness judgment task. Both higher and lower-proficiency L2 learners could make use of the high-constraint sentence context to judge the meaning of novel pseudowords, and higherproficiency L2 learners outperformed lower-proficiency L2 learners in all conditions. These results demonstrate that both L2 proficiency and sentence constraint affect subsequent word learning among second language learners. We extended L2 word learning into a sentence context, replicated the sentence constraint effects previously found among native speakers, and found proficiency effects in L2 word learning. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Word learning has long been an important part of exploring human language acquisition (Lew-Williams, Pelucchi, & Saffran, 2011; Yu & Smith, 2011). Because word learning is a determinant for individual language development, it has garnered much attention from educational and developmental psychologists (Deary, Penke, & Johnson, 2010; Deary, Strand, Smith, & Fernandes, 2007; Hauser & Huang, 1997; Strenze, 2007). Many studies on word learning have used a paradigm called paired-associative learning (Bower & Winzenz, 1970; Gathercole, Hitch, & Martin, 1997; Lang et al., 1988; Nation, 1982), by presenting a picture or an object paired with a visual or auditory word, while learners build the connection between form and meaning through repetition. By including the statistical properties of language, such as manipulating the co-occurrence rate between words and objects by pairing one object with many different words, the associative learning paradigm developed into implicit associative learning, or statistical learning (Breitenstein, Kamping, Andreas, Schomacher, & Knecht, 2004; Breitenstein et al., 2007), in which correct pairs and incorrect pairs are presented in different proportions, and learners do not know the potential rule. Another paradigm, called cross-situational word learning, also adopted a similar ⁎ Corresponding author at: School of Psychology, Beijing Normal University, Beijing 100875, China. E-mail address: [email protected] (B. Chen).

http://dx.doi.org/10.1016/j.actpsy.2015.05.014 0001-6918/© 2015 Elsevier B.V. All rights reserved.

way of statistical learning by pairing a word with many objects (Medina, Snedeker, Trueswell, & Gleitman, 2011; Ramscar, Dye, & Klein, 2013; Smith & Yu, 2008; Yu & Smith, 2007, 2011). The pairedassociative learning paradigm made a great contribution for understanding the mechanisms of word learning by approximating the process of human word acquisition: seeing an object and hearing a word at the same time. At its core, the paradigm is essentially a kind of conditional reflex based on probability (Dehaene, Cohen, Sigman, & Vinckier, 2005), which relies on more general learning ability instead of language ability (Bloom, 2000; Markson & Bloom, 1997). 1.1. Language level and word learning However, like many other advanced human cognitive activities, word learning is not as simple as paired-associative learning. It cannot be separated from prior language level. Studies on native language vocabulary learning found that adults with different reading levels behave differently in subsequent word learning. More specifically, high-level readers learn novel words faster and more efficiently than low-level learners do (Balass, Nelson, & Perfetti, 2010; Perfetti, Wlotko, & Hart, 2005). In the study of Perfetti et al. (2005), adult English speakers learned rare words with definitions, and then made semantic relatedness judgments on trained words, untrained familiar words, and untrained rare words in the test. Eventrelated potentials were recorded during the word learning and testing

T. Ma et al. / Acta Psychologica 159 (2015) 116–122

stages. They found differential performances between skilled learners and less skilled learners. Skilled learners had higher accuracy and larger P600 amplitudes when recognizing trained rare words than untrained rare words. This was taken to mean that they understood the meaning of the trained rare words. Less skilled learners, on the other hand, could not distinguish trained rare words and untrained words. Balass et al. (2010) used the same method to train learners with different levels of skill to learn words in three conditions: orthography-tomeaning, orthography-to-phonology, and phonology-to-meaning. They also tested learners with a semantic relatedness judgment task to explore the differences among trained rare words, untrained familiar words, and untrained rare words. Again, event-related potentials were recorded during the testing stage. The results showed that high-skilled readers showed strong familiarity effects for trained rare words, while less-skilled readers did not. All these results demonstrate that learners with higher language levels are better at learning new words using prior knowledge and skill. 1.2. Reading and word learning In natural environments, word learning is neither limited to childhood nor limited to simple paired association. When children have a certain level of language knowledge, the process of word learning is enriched when they begin to learn to read (Nagy, Herman, & Anderson, 1985). Subsequently, they acquire most words through the context of reading (Krashen, 1989; Nagy, Anderson, & Herman, 1987; Nagy et al., 1985). For adults, the majority of new vocabulary also comes from different contexts, particularly reading (Berwick, Friederici, Chomsky, & Bolhuis, 2013). It is possible to acquire word meaning through onetime reading, in appropriate circumstances (Borovsky, Elman, & Kutas, 2012; Borovsky, Kutas, & Elman, 2010). Many previous studies about word recognition and lexical access found sentence constraint effects in L2 processing (Duyck, Assche, Drieghe, & Hartsuiker, 2007; Schwartz & Kroll, 2006; Titone, Libben, Mercier, Whitford, & Pivneva, 2011; van Hell & de Groot, 2008), and in the studies of word learning through reading, the question of whether and how readers make use of sentence context to acquire new words also gets a lot of attention. Chaffin, Morris, and Seely (2001) explored the role of informativeness of sentences in word learning using eye-tracking technology. They found that readers would gaze at a location longer if it provided more effective information for novel word meaning. Their results suggest that readers can and do make use of contextual information provided by sentences to infer word meaning. Borovsky et al. (2010) examined the effects of sentence constraint on the understanding and usage of novel words. Twenty-six native English speakers read high-constraint or low-constraint sentences with known or unknown words embedded. After each sentence, they made a plausibility judgment about the word usage. Event-related potentials were recorded during the experiment. Plausibility effects were observed in the N400 component when the novel word was acquired in a high constraint sentence, which demonstrates that native speakers rapidly acquired the novel word usage through high constraint sentences. Borovsky et al. (2012) then investigated the impact of sentence constraint on the integration of novel word meanings into semantic memory. Adult native speakers of English read high-constraint or low-constraint sentences ending with known or unknown words. Then after reading a sentence, they completed a lexical decision task in which ending words (known or unknown) served as primes for related, unrelated, and synonym target words. They found that N400 amplitudes to target words preceded by unknown word primes varied with prime-target relatedness, but only when the unknown word was embedded in high-constraint sentences previously. These results demonstrate that adult native speakers can rapidly integrate information about word meaning into their mental lexicons by reading high constraint sentences. Mestres-Missé, Rodriguez-Fornells, and Münte (2007) even directly observed the brain activity of word

117

meaning acquisition during sentence reading by recording the brain potential. Participants read three sentences including the same novel word, while in some of the three sentences the novel word could form a congruent meaning, in some other sentences the novel word could not form a congruent meaning. She found that in sentences that a novel word could form a congruent meaning, N400 amplitude decreased across the course of three sentences, which implied meaning acquisition of novel word. All these studies so far indicate that native language learners can take advantage of immediate information provided by sentences to learn new words. Then what about L2 learners? Can they also make use of the sentence context when learning new words? And if new word learning relies on previous knowledge and language level, how important is L2 proficiency? There are some studies exploring whether L2 learners can learn new words through reading (Pitts, White, & Krashen, 1989), and if so, how many encounters do they need (Horst, Cobb, & Meara, 1998; Ferrel Tekmen & Daloğlu, 2006; Pellicer-Sánchez & Schmitt, 2010; Waring & Takaki, 2003; Webb, 2008; Zahar, Cobb, & Spada, 2001). There are also some studies that examined the role of proficiency in L2 word learning, and they found that learners with larger L2 vocabulary size had greater word learning gains through reading and needed fewer encounters (Horst et al., 1998; Ferrel Tekmen & Daloğlu, 2006; Zahar et al., 2001). Because most of these studies used published novels as reading materials, it was hard to control the familiarity and reading difficulty. Even so, there are still some studies that investigated the role of sentence context. Pulido and colleagues performed a series of studies focused on the topic familiarity of the reading materials (Pulido, 2003, 2007; Pulido & Hambrick, 2008). Pulido (2003) studied L2 vocabulary acquisition and retention through reading narratives, in which L2 learners with different proficiency levels read narratives of familiar or less familiar topics and which contained nonsense words. Then, participants completed recognition tests 2 and 28 days after reading the narratives. Topic familiarity effects were found on the initial measure of gain (2 days after), which demonstrated that sentence context could influence the gain of words. She also found that no matter how familiar the topic was, learners with high proficiency acquired more words through reading and maintained their learning better, which suggests that L2 word learning relies on existing language experience. Her following studies further confirmed that sentence context can influence word acquisition and that languageprocessing experience positively influences L2 passage comprehension (Pulido, 2007; Pulido & Hambrick, 2008). All these studies give us a preview of the role of L2 proficiency and sentence context on word learning through reading. However, none of them has investigated the role of sentence constraint or the interaction of sentence constraint and proficiency on L2 word learning. In the present study, we designed an experiment to investigate the effects of L2 proficiency and sentence constraint on subsequent L2 word learning through reading. We strictly controlled many variables of the words and sentences, and used pseudowords as the learning items and two sentence constraint contexts (high-constraint and low-constraint) for each pseudoword. We used a whole sentence presentation paradigm, and a block of sentences was presented before a block of semantic relatedness judgment tasks that measured learners' behavioral performance. L2 learners might or might not make use of the sentence context to learn new words, and their performance could vary with proficiency and sentence constraint. One possibility is that L2 word learning is similar to native language word learning such that L2 learners perform better in high-constraint sentences than in low-constraint sentences, and higher proficiency learners would outperform lower proficiency learners in L2 word learning. An alternative possibility is that L2 learners might not learn words like native speakers, thus we would fail to see effects of proficiency or sentence constraint on L2 word learning.

118

T. Ma et al. / Acta Psychologica 159 (2015) 116–122

2. Methods

a separate group of 24 college students from the same background as the participants (mean familiarity = 4.83, SD = 0.18).

2.1. Participants Participants were 48 right-handed college students from Beijing Normal University, with normal or corrected-to-normal vision. They were all native Chinese speakers learning English as a second language. They were recruited to our study and split into two groups according to their College English Test (CET) levels. The CET is a proficiency test used to estimate the English level of Chinese college students through listening comprehension, reading comprehension, cloze, error correction, writing, and translation (Zheng & Cheng, 2008). Twenty-three participants who passed CET Band 6 were categorized as higher proficiency English learners; twenty-five participants who failed CET Band 4 were categorized as lower proficiency English learners. Before the experiment, all participants completed self-ratings of their English listening, speaking, reading, and writing abilities on a 5-point scale (1 = very nonproficient, 5 = very proficient) as well as the Quick Placement Test (QPT, 2001), to provide a consistent evaluation of their English level. The QPT is a flexible test of English language proficiency developed by Oxford University Press and Cambridge ESOL to quickly find a student's level of English, including reading and structure, grammar and vocabulary. Part 1 has 40 items and is taken by all students. Part 2 has 20 items and is administered only to students who did well on Part 1 (Geranpayeh, 2003). All participants of this experiment were asked to complete the Part 1 at first, and then the Part 2 was administered according to their performance on the Part 1. In the end, all participants finished both Part 1 and Part 2. These two methods of measurements on language level were highly correlated (self-ratings on listening and QPT scores: Eta = .71, p b .05; self-ratings on speaking and QPT scores: Eta = .79, p b .01; self-ratings on reading and QPT scores: Eta = .78, p b .05; self-ratings on writing and QPT scores: Eta = .62, p b .01). For more details about participants, see Table 1. 2.2. Experiment design We adopted a mixed experimental design, 2 (sentence constraint: high, low) × 2 (word type: real word, pseudoword) × 2 (proficiency: higher, lower), with sentence constraint and word type as withinsubject factors, and proficiency as a between-subject factor. 2.3. Materials 2.3.1. Real words Real words were 120 high frequency concrete nouns. Word frequency (mean logFreq = 9.97, SD = 1.01) was rated according to HAL norms (Hyperspace Analogue to Language Frequency Norms, Balota et al., 2007; Lund & Burgess, 1996). Concreteness (M = 580.47, SD = 41.15) was rated according to the MRC database (Medical Research Council Psycholinguistic Database, Wilson, 1988). Additionally, familiarity was rated using a 5-point scale (1 = very unfamiliar, 5 = very familiar) by

2.3.2. Pseudowords We made 120 pronounceable pseudowords using Wuggy, a multilingual pseudoword generator developed by Keuleers and Brysbaert (2010). This generator uses a specific algorithm to generate pseudowords that can match the subsyllabic structure and transition frequencies with real English words. (It is also currently available for Dutch, German, French, Spanish, Serbian, and Basque.) Pseudowords were matched to real words in length in letters and syllables. 2.3.3. Semantically related/unrelated words Semantically related or unrelated words were selected to pair with the 120 real words, which would be used in the semantic relatedness judgment task after sentence reading. The degree of semantic relatedness was rated by the same separate group of 24 college students who rated the familiarity of real words, using a 5-point scale (1 = absolutely unrelated, 5 = closely related). The average score of semantically related words was 4.43 (SD = .42), for example, “agriculture” was rated as highly semantically related to “farm”; the average score of the semantically unrelated words was 1.18 (SD = .22), for example, “bottle” was rated as semantically unrelated to “mountain.” 2.3.4. Sentences Two sentences (high-constraint and low-constraint) were constructed for each real word. The pseudoword replaced the real word in the same sentences. All sentences were made up of 9 words, with the key word (real word or pseudoword) always appearing at the end of the sentence (see Table 2). The constraint (high or low) of sentences was determined with the help of a separate group of 40 college students from the same school as the participants. Using a cloze test, these students completed the sentences with the first noun that came to mind. The cloze probability was defined as the percentage of times the same word was provided by this group of students. The mean cloze probability of high-constraint sentences (87.75%, SD = .11) was significantly different from low-constraint sentences (9.81%, SD = .14), t (119) = 47.35, p b .001, Cohen's d = 4.35. To ensure that all sentences were easily understood by our participants, the difficulty degree was rated by the 40 college students who rated the constraint of sentences. On a 5-point scale (1 = very easy, 5 = very difficult), the overall score was 1.27 (SD = .29), and no statistical differences were found between high-constraint sentences (M = 1.29, SD = .29) and low-constraint sentences (M =1.26, SD = .28), t (119) = .63, p = .53, Cohen's d = .06. Sentences were split pseudorandomly into four lists to ensure that no items (both real words/ pseudowords and sentences) were repeatedly presented in one list, and the real words and their corresponding pseudowords (along with sentences the real words/pseudowords embedded in) were never presented in the same list. Each list included 120 sentences, 30 sentences per condition. Each participant received only one of the four lists. 2.4. Procedure

Table 1 Mean (SD) age, age of English acquisition (AoA), English L2 proficiency ratings, and QPT test scores of the participants by proficiency level. Participants

Higher proficiency Lower proficiency t-test Note: ⁎ p b .01. ⁎⁎ p b .001.

Age

21.91 (1.41) 22.28 (1.90) −.75

AoA

10.96 (1.26) 11.16 (1.14) −.59

Self-rating English level Listening

Speaking

Reading

Writing

3.22 (.60) 2.44 (.71) 4.07⁎⁎

3.04 (.21) 2.28 (.61) 5.67⁎⁎

4.00 (.52) 3.28 (.54) 4.68⁎⁎

3.48 (.51) 3.04 (.68) 2.52⁎

QPT score 47.52 (2.98) 38.64 (3.79) 8.96⁎⁎

Stimuli were presented on a computer using E-prime software version 1.1. Participants were seated in front of the computer and were provided with instructions and practice trials prior to the experiment.

Table 2 Example of experimental materials by condition.

Real word Pseudoword

High-constraint

Low-constraint

Let's go to the cinema to see a movie Let's go to the cinema to see a speath

I am not very interested in that new movie I am not very interested in that new speath

T. Ma et al. / Acta Psychologica 159 (2015) 116–122

Twenty blocks of sentences were presented in random order. In each block there were six sentences, and each sentence was presented on the screen as a whole. Participants read a block of six sentences one by one by pressing the space bar on the computer keyboard. When they finished a block of sentences, a question mark prompt would be presented on the screen for 2000 ms to indicate the subsequent semantic relatedness judgment task. In this task, learners read six word pairs corresponding to the six sentences just presented and judged whether the words were semantically related. Within one block, six sentences and corresponding word pairs were presented in pseudo-random order. “Yes” or “No” responses were made by pressing “F” or “J” on the keyboard. Half of the participants pressed “F” for “Yes,” and the other half pressed “J” for “Yes.” The last word of the preceding sentence was always presented on the left, with another semantically related or unrelated word presented on the right. These two words were presented at the same time and would disappear if there was no response detected within 5000 ms. Because we focused on whether L2 learners could make use of sentence context to learn novel word meanings, the accuracy of the semantic relatedness judgment task was our dependent variable. We also recorded response times. To confirm that participants had no difficulty in reading the sentences, after the semantic relatedness judgment task was complete, all participants were given a checklist of all the sentences they had just read, with all the pseudowords in the materials replaced with the corresponding real words. They were asked to mark the words or sentences they felt were hard to understand. No marks were made on these checklists, which we take to mean as an indication that the materials were easily processed by all participants.

2.5. Results In the semantic relatedness judgment task, stimuli that elicited response times beyond ±3 SD in each condition were excluded (1.23%). A mixed-effects logistic model of accuracy and a mixed-effects model of response time in the semantic relatedness judgment task were separately built to analyze the behavior of participants (Baayen, Davidson, & Bates, 2008; Jaeger, 2008). All statistical analyses were carried out using R 2.15.2 (R Core Team, 2014), implemented with package lme4 (Bates, Maechler, Bolker, & Walker, 2013), lmerTest (Kuznetsova, Brockhoff, & Christensen, 2013), and language R (Baayen, 2011). For each of the two proficiency groups, accuracy and reaction times for the different conditions, along with predicted values of mixed-effects models are shown in Table 3.

Table 3 Mean (SD) accuracy and response times by condition for each proficiency group and the predicted values from mixed-effect models. Real words

Higher proficiency Accuracy (%) RT (ms) Predicted value of accuracy (%) Predicted value of RT (ms) Lower proficiency Accuracy (%) RT (ms) Predicted value of accuracy (%) Predicted value of RT (ms)

Pseudowords

High constraint

Low constraint

High constraint

Low constraint

90 (6.83) 1970 (934) 90 (6.23)

91 (6.26) 1931 (823) 92 (4.68)

74 (10.71) 2608 (1205) 74 (9.80)

61 (8.74) 2589 (1247) 61 (12.61)

1970 (511)

1932 (500)

2607 (515)

2589 (521)

84 (11.08) 2114 (1006) 85 (8.34)

85 (10.23) 2017 (881) 86 (7.70)

67 (12.22) 2546 (1454) 67 (12.25)

55 (7.64) 2572 (1423) 55 (13.71)

2114 (611)

2017 (595)

2546 (606)

2572 (617)

119

Table 4 Mixed-effects logistic model of accuracy in the semantic relatedness judgment task. Predictor

Estimate

SE

z value

p (N|z|)

(Intercept) WordType Constraint Proficiency WordType × constraint WordType × proficiency Constraint × proficiency WordType × constraint × proficiency

1.18 1.20 −.66 −.39 .93 −.07 .11 −.30

.13 .18 .14 .16 .23 .21 .17 .30

.13 6.84 −4.85 −2.50 4.05 −.32 .62 −1.00

.0000 .0000 .0000 .0125 .0000 .7483 .5332 .3194

For accuracy data, a mixed-effects logistic model was conducted with word type, sentence constraint, and proficiency as fixed factors, and subject and item (combination of sentences and key words) as random factors. Results are summarized in Table 4. There was a significant main effect of word type (z = 6.84, p b .001), such that accuracy was higher in the real word condition. There was a significant main effect of sentence constraint (z = − 4.85, p b .001), such that accuracy was higher in the high constraint condition. And there was a significant main effect of proficiency (z = −2.50, p b .05), such that accuracy was higher in higher proficient learners. We also found a significant interaction between word type and constraint (z = 4.05, p b .001). And a further simple effect analysis showed that constraint effects were only exhibited in the pseudoword condition (pseudoword: z = −4.72, p b .001; real word: z = 1.52, p = .13). No other significant interactions were found. We also examined whether the accuracy of different conditions was significantly higher than chance. For higher-proficiency learners, there was no significant difference between chance level and the accuracy in low constraint sentences with pseudowords (z = 1.05, p N .05); whereas the accuracy was above chance level in the other three conditions (high constraint sentences with real words: z = 3.84, p b .01; high constraint sentences with pseudowords: z = 2.30, p b .05; low constraint sentences with real words: z = 3.93, p b .01). For lowerproficiency learners, the pattern was the same. No significant differences were found between chance level and accuracy in low constraint sentences with pseudowords (z = .5, p N .05), whereas mean accuracy was above chance level in the other three conditions (high constraint sentences with real words: z = 3.4, p b .01; high constraint sentences with pseudowords: z = 1.7, p b .05; low constraint sentences with real words: z = 3.5, p b .01). For response time data, a mixed-effects model was conducted with word type, sentence constraint, and proficiency as fixed factors, and subject and item (combination of sentences and key words) as random factors. The model also included a by-participant random slope for the trial order (Baayen & Milin, 2010). Results are summarized in Table 5. The analysis revealed that the main effect of word type was significant, t = − 10.67, p b .001, such that response times were longer in the pseudoword condition. The interaction of word type and proficiency was significant, t = 2.29, p b .05. A further simple effect analysis showed that proficiency effects were both nonsignificant, but exhibited opposite directions in the pseudowords and real words (pseudoword: t = − .43,

Table 5 Mixed-effects model of response time in the semantic relatedness judgment task. Predictor

Estimate

SE

t value

p (N|t|)

(Intercept) WordType Constraint Proficiency WordType × constraint WordType × proficiency Constraint × proficiency WordType × constraint × proficiency

2470.46 −617.48 −20.05 −58.26 −32.41 173.76 47.14 −82.87

109.64 57.88 58.57 148.95 76.68 76.01 74.63 107.15

22.53 −10.67 −.34 −.39 −.42 2.29 .63 −.77

.0000 .0000 .7322 .6972 .6725 .0223 .5276 .4394

120

T. Ma et al. / Acta Psychologica 159 (2015) 116–122

p = .67; real word: t = 1.14, p = .26). No other main effects or interactions were significant.

3. Discussion The present study investigated the effects of L2 proficiency and sentence constraint on L2 word learning. We predicted significant effects of proficiency and sentence constraint in the pseudoword condition, similar to native speakers' performance. Higher proficiency learners outperformed lower proficiency learners in all conditions, and L2 word learning benefited from high-constraint sentences. In the present study, we strictly controlled participants' familiarity of materials. To verify this, all words and sentences were assessed by separate groups of raters from the same population as our sample. After the experiment, we also used a material checklist to confirm that each participant had no difficulties in reading the items. Therefore, any differences between the two groups could not be attributed to difficulty of the materials. Based on previous studies about bilingualism, lower L2 proficiency results in weaker semantic processing (Ojima, Nakata, & Kakigi, 2005; Sagarra & Herschensohn, 2011) and syntactic processing (Ojima et al., 2005; Osterhout et al., 2008; Rossi, Gugler, Friederici, & Hahne, 2006; Sagarra & Herschensohn, 2010, 2011; Steinhauer, White, & Drury, 2009). The efficiency of working memory is partially dependent on proficiency (van den Noort, Bosch, & Hugdahl, 2006). When bilinguals process L2 online, two languages could potentially be activated at the same time, causing interference (Baten, Hofman, & Loeys, 2011; Dijkstra & van Heuven, 2002; van Hell & Tanner, 2012). At lower proficiencies, L2 processing is more resource intensive. The less automatic, more effortful processing puts a larger burden on working memory of lower proficiency L2 learners (van den Noort et al., 2006). Thus, lower proficiency learners perform worse when they process L2, and are poorer in making use of the immediate information of sentence contexts. With increasing proficiency, language processing does not require as many cognitive resources, so it becomes highly automatic and qualitatively comparable to native speakers in both online and offline tasks (Coughlin & Tremblay, 2012; Sagarra & Herschensohn, 2010, 2011). Thus, higher proficiency L2 learners could take advantage of high-constraint sentences to learn novel words more efficiently. Results in this study also revealed the role of sentence constraint in L2 novel word learning. High-constraint sentences were supportive for L2 word learning, which is consistent with previous studies on native speakers. Chaffin et al. (2001) used eye-tracking technology to examine the role of informativeness of sentences in word learning. They found that native language learners could make use of the information provided by sentences to infer word meaning. Borovsky et al. (2010) embedded novel words into sentences with different levels of constraint, and recorded the event-related brain potentials during learning and testing. They found that in the high-constraint sentences condition, native speakers could acquire the meanings of words through only one exposure (Borovsky et al., 2010, 2012). Sentence constraint effects were found both in native language learners and in L2 learners, which may indicate that semantic constraint works similarly in the first and second language. Moreover, both higher and lower proficiency L2 learners benefit from highconstraint sentences, which indicates that the difference between higher and lower proficiency L2 learners on L2 word learning is quantitative not qualitative. L2 word learning through reading showed the same pattern as native language word learning. Sentence constraint and current language level both influenced L2 word learning. All these results suggested that word learning relies not only on domain-general abilities, but also on language ability and previous language level. Word learning

is more than a simple association; it is a combination of domain-general and language-dependent processes. Previous studies on word learning focused more on the role of shortterm memory and word knowledge and were mostly based on situations without sentential context (Chen & Cowan, 2009; Gathercole & Masoura, 2003; Gray, 2006; Hu, 2003; Maury & Luotoniemi, 2007; Storkel, Armbruster, & Hogan, 2006). A large number of studies adopted training tasks with clearly defined learning objectives, requiring participants to retain and later retrieve what they've learned (Jeong et al., 2010; Laufer & Rozovski-Roitblat, 2011; McLaughlin, Osterhout, & Kim, 2004; Raboyeau et al., 2004; Shtyrov, 2012; Yu & Smith, 2011; Yu, Zhong, & Fricker, 2012). This is closer to word learning in classroom settings, with explicit learning goals. Because word learning in reading context is so common in real life (Berwick et al., 2013; Krashen, 1989; Mestres-Missé et al., 2007; Mestres-Missé, Càmara, RodriguezFornells, Rotte, & Münte, 2008; Nagy et al., 1987; Onnis & Thiessen, 2013), more emphasis in future research should be placed on the mechanism of incidental word learning in meaningful contexts. In the current study, we examined word learning in a more natural way, by simulating word learning in everyday life through a reading task without an obvious learning requirement. All participants needed to do was read sentences then make semantic relatedness judgments. Thus, we extended the area of research to word learning in context and got some instructive results on how sentence context and proficiency affect L2 word learning. In this study, each novel word (pseudoword) was presented in only one sentence context. Although many studies have found that one-time acquisition occurs during word learning (Borovsky et al., 2010, 2012; Shtyrov, 2011; Shtyrov, Nikulin, & Pulvermüller, 2010), more evidence suggests that we still need practice and repetition to consolidate and refine word learning (Medina et al., 2011; Mestres-Missé et al., 2007, 2008; Munro, Baker, McGregor, Docking, & Arciuli, 2012; Ramscar et al., 2013). Studies on children's word learning found that repetition of context (stories or non-target words) is better in facilitating word learning than an equal number of varied contexts (Axelsson & Horst, 2014; Williams & Horst, 2014). However, a study on adult L1 learners found the exact opposite pattern: that varied sentences are better than repeated sentences in facilitating native language word learning (Bolger, Balass, Landen, & Perfetti, 2008). In the future, novel words should be presented multiple times to explore the dynamic process of L2 word learning through sentence reading and how repetition and varied contexts influence novel word learning. In conclusion, the present study extended L2 word learning into sentence contexts, replicated the sentence constraint effects previously found among native speakers, and found proficiency effects in L2 word learning. Our results provide strong evidence that both L2 proficiency and sentence constraint affect subsequent word learning among second language learners. Acknowledgments This research was supported by the Open Fund of State Key Lab of Cognitive Neuroscience and Learning (CNLYB1309) to Baoguo Chen. Appendix A. Words, pseudowords and semantic related/unrelated words

No.

L2 Word

Pseudoword

Related/unrelated Words

Relatedness

1 2 3 4 5 6 7 8

Farm Shape Ocean Storm Map Mountain Camera Hat

Arram Banble Bliat Bopple Dusin Lectode Jetter Coddin

Agriculture Hotel Sea Arm Chart Bottle Expert Cap

Related Unrelated Related Unrelated Related Unrelated Unrelated Related

T. Ma et al. / Acta Psychologica 159 (2015) 116–122 Appendix A. (continued)

121

Appendix A. (continued)

No.

L2 Word

Pseudoword

Related/unrelated Words

Relatedness

No.

L2 Word

Pseudoword

Related/unrelated Words

Relatedness

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84

Prison Church Beer Ice Window Song Medicine Football Sugar Bell Island Stone Oil Bridge River Patient Gift Lawyer Letter Dog Bone Skin Plane Customer Fish Teeth Egg Novel Smile Kiss Bank Flag Toy Guard Rice Teacher Artist Forest Leg Chair Bird Hospital Sky Kid Audience Author Tower Horse Gun Moon Army Floor Cloud Banana Dance Key Bear City Coffee Newspaper Flower Cat Eye Water Box Pen Clock Nose Umbrella FIRE Ring Restaurant Bee Carrot Dress Ant

Callian Hustack Mubbon Naswin Capret Lumic Bribod Poltan Punpet Notid Pramon Rallen Ratid Redim Lurple Hildet Pestler Rasium Pectute Wellop Mucket Tuser Spazz Warliz Musin Wartet Vavier Purder Barpit Brinny Sevie Zamper Borty Lowen Candan Anspar Victap Geadle Candot Bindom Ricket Romber Edlay Desuce Tonnas Revasm Feson Devart Rossage Empock Refite Hoddle Pibbit Plare Scrool Buite Sneam Peapt Drane Swood Plave Rictor Drist Swink Tince Gumph Hetper Unane Ceague Nacoan Owfan Snamp Narein Gurrs Peague Brire

Crime God Cold Village Curtain Cell Drug Battery Candy Tree Ocean Model Fat Fight Boat Wheel Present Brain Mail Fox Meat Ball Train Beach Net Police Chicken Blood Laugh Lips Bike Wave Lake Safety Car Tutor Office Foot Trip Seat Chicken Illness Baby Wind Son Reader Menu Lab Bullet Sun Heart Dollar Queen Pear Form Lock Rain Town Grass Magazine Café Kitty Bottle Liquid Tears Write Raincoat Ear Time Water Finger Storage Sand Vegetable Link Insect

Related Related Unrelated Unrelated Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Related Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Related Unrelated Related Unrelated Related Unrelated Related Unrelated Unrelated Unrelated Related Related Related Unrelated Unrelated Unrelated Related Unrelated Unrelated Related Related Unrelated Unrelated Unrelated Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Related Unrelated Unrelated Related Unrelated Related

85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120

Chocolate Light World Friend Servant Passenger Zoo Sheep Piano Leader Glasses Mirror Garden Dust Game Phone Knife Ticket Snow Library Sister Paper Socks Rat Factory College Boat Doctor Milk Bread Book Movie Apple Money Purse Door

Nusks Phosh Quing Thimp Vattey Fover Jalleb Loler Wheen Ptoter Adpelt Smider Oribod Swock Unigen Cipter Isefy Swoin Gulct Creath Gussr Quord Athits Punger Wobtin Goftar Guilor Mocher Clore Cucket Jirys Speath Viboon Becieu Thire Gutlin

Earth Bright Candy Companion Fiction Driver Battle Sun Guitar Follower Corn Fur Park Dirty Roof Call Fork Animal Chicken Study Waiter Note Court Mouse Baseball University Gene Nurse Access Hot dog Orange Film Ship Cash Bag White

Unrelated Related Unrelated Related Unrelated Related Unrelated Unrelated Related Related Unrelated Unrelated Related Related Unrelated Related Related Unrelated Unrelated Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Unrelated Related Related Unrelated

References Axelsson, E. L., & Horst, J. S. (2014). Contextual repetition facilitates word learning via fast mapping. Acta Psychologica, 152, 95–99. Baayen, R. H. (2011). languageR: Data sets and functions with “Analyzing Linguistic Data: A practical introduction to statistics”. R package version 1.4. from http://CRAN.Rproject.org/package=languageR Baayen, R. H., Davidson, D. J., & Bates, D. M. (2008). Mixed-effects modeling with crossed random effects for subjects and items. Journal of Memory and Language, 59(4), 390–412. Baayen, R. H., & Milin, P. (2010). Analyzing reaction times. International Journal of Psychological Research, 3(2), 12–28. Balass, M., Nelson, J. R., & Perfetti, C. A. (2010). Word learning: An ERP investigation of word experience effects on recognition and word processing. Contemporary Educational Psychology, 35(2), 126–140. Balota, D., Yap, M., Cortese, M., Hutchison, K., Kessler, B., Loftis, B., et al. (2007). The English Lexicon Project. Behavior Research Methods, 39(3), 445–459. Baten, K., Hofman, F., & Loeys, T. (2011). Cross-linguistic activation in bilingual sentence processing: The role of word class meaning. Bilingualism: Language and Cognition, 14(3), 351–359. Bates, D., Maechler, M., Bolker, B., & Walker, S. (2013). lme4: Linear mixed-effects models using Eigen and S4. R package version 1.0–4. from http://CRAN.R-project.org/ package=lme4 Berwick, R. C., Friederici, A. D., Chomsky, N., & Bolhuis, J. J. (2013). Evolution, brain, and the nature of language. Trends in Cognitive Sciences, 17(2), 89–98. Bloom, P. (2000). How children learn the meanings of words. Cambridge, MA: MIT Press. Bolger, D. J., Balass, M., Landen, E., & Perfetti, C. A. (2008). Context variation and definitions in learning the meanings of words: An instance-based learning approach. Discourse Processes, 45(2), 122–159. Borovsky, A., Elman, J. L., & Kutas, M. (2012). Once is enough: N400 indexes semantic integration of novel word meanings from a single exposure in context. Language Learning and Development, 8(3), 278–302. Borovsky, A., Kutas, M., & Elman, J. (2010). Learning to use words: Event-related potentials index single-shot contextual word learning. Cognition, 116(2), 289–296. Bower, G. H., & Winzenz, D. (1970). Comparison of associative learning strategies. Psychonomic Science, 20(2), 119–120. Breitenstein, C., Kamping, S., Andreas, J., Schomacher, M., & Knecht, S. (2004). Word learning can be achieved without feedback: Implications for aphasia therapy. Restorative Neurology and Neuroscience, 22(6), 445–458. Breitenstein, C., Zwitserlood, P., de Vries, M. H., Feldhues, C., Knecht, S., & Dobel, C. (2007). Five days versus a lifetime: Intense associative vocabulary training generates lexically integrated words. Restorative Neurology and Neuroscience, 25(5), 493–500.

122

T. Ma et al. / Acta Psychologica 159 (2015) 116–122

Chaffin, R., Morris, R. K., & Seely, R. E. (2001). Learning new word meanings from context: A study of eye movements. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27(1), 225–235. Chen, Z., & Cowan, N. (2009). Core verbal working-memory capacity: The limit in words retained without covert articulation. The Quarterly Journal of Experimental Psychology, 62(7), 1420–1429. Coughlin, C. E., & Tremblay, A. (2012). Proficiency and working memory based explanations for nonnative speakers' sensitivity to agreement in sentence processing. Applied Psycholinguistics, 1–32. Deary, I. J., Penke, L., & Johnson, W. (2010). The neuroscience of human intelligence differences. Nature Reviews Neuroscience, 11(3), 201–211. Deary, I. J., Strand, S., Smith, P., & Fernandes, C. (2007). Intelligence and educational achievement. Intelligence, 35(1), 13–21. Dehaene, S., Cohen, L., Sigman, M., & Vinckier, F. (2005). The neural code for written words: A proposal. Trends in Cognitive Sciences, 9(7), 335–341. Dijkstra, T., & van Heuven, W. J. B. (2002). The architecture of the bilingual word recognition system: From identification to decision. Bilingualism: Language and Cognition, 5(3), 175–197. http://dx.doi.org/10.1017/S1366728902003012. Duyck, W., Assche, E. V., Drieghe, D., & Hartsuiker, R. J. (2007). Visual word recognition by bilinguals in a sentence context: Evidence for nonselective lexical access. Journal of Experimental Psychology: Learning, Memory, and Cognition, 33(4), 663–679. Ferrel Tekmen, E. A., & Daloğlu, A. (2006). An investigation of incidental vocabulary acquisition in relation to learner proficiency level and word frequency. Foreign Language Annals, 39(2), 220–243. Gathercole, S. E., Hitch, G. J., & Martin, A. J. (1997). Phonological short-term memory and new word learning in children. Developmental Psychology, 33(6), 966–979. Gathercole, S. E., & Masoura, E. V. (2003). Contrasting contributions of phonological shortterm memory and long-term knowledge to vocabulary learning in a foreign language. Memory, 13(3–4), 422–429. Geranpayeh, A. (2003). A quick review of the English Quick Placement Test. Research Notes, 12, 8–10. Gray, S. (2006). The relationship between phonological memory, receptive vocabulary, and fast mapping in young children with specific language impairment. Journal of Speech, Language, and Hearing Research, 49(5), 955–969. Hauser, R. M., & Huang, M. -H. (1997). Verbal ability and socioeconomic success: A trend analysis. Social Science Research, 26(3), 331–376. Horst, M., Cobb, T., & Meara, P. (1998). Beyond a clockwork orange: Acquiring second language vocabulary through reading. Reading in a Foreign language, 11(2), 207–233. Hu, C. F. (2003). Phonological memory, phonological awareness, and foreign language word learning. Language Learning, 53(3), 429–462. Jaeger, T. F. (2008). Categorical data analysis: Away from ANOVAs (transformation or not) and towards logit mixed models. Journal of Memory and Language, 59(4), 434–446. Jeong, H., Sugiura, M., Sassa, Y., Wakusawa, K., Horie, K., Sato, S., et al. (2010). Learning second language vocabulary: Neural dissociation of situation-based learning and text-based learning. NeuroImage, 50(2), 802–809. Keuleers, E., & Brysbaert, M. (2010). Wuggy: A multilingual pseudoword generator. Behavior Research Methods, 42(3), 627–633. Krashen, S. (1989). We acquire vocabulary and spelling by reading: Additional evidence for the input hypothesis. The Modern Language Journal, 73(4), 440–464. Kuznetsova, A., Brockhoff, P. B., & Christensen, R. H. B. (2013). lmerTest: Tests for random and fixed effects for linear mixed effect models (lmer objects of lme4 package). R package version 2.0-0. from http://CRAN.R-project.org/package=lmerTest Lang, W., Lang, M., Uhl, F., Kornhuber, A., Deecke, L., & Kornhuber, H. H. (1988). Left frontal lobe in verbal associative learning: A slow potential study. Experimental Brain Research, 70(1), 99–108. Laufer, B., & Rozovski-Roitblat, B. (2011). Incidental vocabulary acquisition: The effects of task type, word occurrence and their combination. Language Teaching Research, 15(4), 391–411. Lew-Williams, C., Pelucchi, B., & Saffran, J. R. (2011). Isolated words enhance statistical language learning in infancy. Developmental Science, 14(6), 1323–1329. Lund, K., & Burgess, C. (1996). Producing high-dimensional semantic spaces from lexical co-occurrence. Behavior Research Methods, Instruments, & Computers, 28(2), 203–208. Markson, L., & Bloom, P. (1997). Evidence against a dedicated system for word learning in children. Nature, 385(6619), 813–815. Maury, S., & Luotoniemi, E. (2007). Individual differences in phonological learning and verbal STM span. Memory & Cognition, 35(5), 1122–1135. McLaughlin, J., Osterhout, L., & Kim, A. (2004). Neural correlates of second-language word learning: minimal instruction produces rapid change. Nature Neuroscience, 7(7), 703–704. http://dx.doi.org/10.1017/S1366728902003012. Medina, T. N., Snedeker, J., Trueswell, J. C., & Gleitman, L. R. (2011). How words can and cannot be learned by observation. Proceedings of the National Academy of Sciences, 108(22), 9014–9019. Mestres-Missé, A., Càmara, E., Rodriguez-Fornells, A., Rotte, M., & Münte, T. F. (2008). Functional neuroanatomy of meaning acquisition from context. Journal of Cognitive Neuroscience, 20(12), 2153–2166. Mestres-Missé, A., Rodriguez-Fornells, A., & Münte, T. F. (2007). Watching the brain during meaning acquisition. Cerebral Cortex, 17(8), 1858–1866. Munro, N., Baker, E., McGregor, K., Docking, K., & Arciuli, J. (2012). Why word learning is not fast. Frontiers in Psychology, 3(41). http://dx.doi.org/10.3389/fpsyg.2012. 00041. Nagy, W. E., Anderson, R. C., & Herman, P. A. (1987). Learning word meanings from context during normal reading. American Educational Research Journal, 24(2), 237–270. Nagy, W. E., Herman, P. A., & Anderson, R. C. (1985). Learning words from context. Reading Research Quarterly, 20(2), 233–253. Nation, I. (1982). Beginning to learn foreign vocabulary: A review of the research. RELC Journal, 13(1), 14–36.

Ojima, S., Nakata, H., & Kakigi, R. (2005). An ERP study of second language learning after childhood: Effects of proficiency. Journal of Cognitive Neuroscience, 17(8), 1212–1228. Onnis, L., & Thiessen, E. (2013). Language experience changes subsequent learning. Cognition, 126(2), 268–284. Osterhout, L., Poliakov, A., Inoue, K., McLaughlin, J., Valentine, G., Pitkanen, I., et al. (2008). Secondlanguage learning and changes in the brain. Journal of Neurolinguistics, 21(6), 509–521. Pellicer-Sánchez, A., & Schmitt, N. (2010). Incidental vocabulary acquisition from an authentic novel: Do “Things Fall Apart”? Reading in a Foreign language, 22(1), 31–55. Perfetti, C. A., Wlotko, E. W., & Hart, L. A. (2005). Word learning and individual differences in word learning reflected in event-related potentials. Journal of Experimental Psychology: Learning, Memory, and Cognition, 31(6), 1281–1292. Pitts, M., White, H., & Krashen, S. (1989). Acquiring second language vocabulary through reading: A replication of the Clockwork Orange study using second language acquirers. Reading in a Foreign language, 5(2), 271–275. Pulido, D. (2003). Modeling the role of second language proficiency and topic familiarity in second language incidental vocabulary acquisition through reading. Language Learning, 53(2), 233–284. Pulido, D. (2007). The effects of topic familiarity and passage sight vocabulary on L2 lexical inferencing and retention through reading. Applied Linguistics, 28(1), 66–86. Pulido, D., & Hambrick, D. Z. (2008). The “virtuous” circle: Modeling individual differences in L2 reading and vocabulary development. Reading in a Foreign language, 20(2), 164–190. Quick Placement Test: Paper and Pen Test: User Manual (2001). University of Cambridge local examinations syndicate. R Core Team (2014). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing from http://www.R-project.org/. Raboyeau, G., Marie, N., Balduyck, S., Gros, H., Démonet, J. -F., & Cardebat, D. (2004). Lexical learning of the English language: A PET study in healthy French subjects. NeuroImage, 22(4), 1808–1818. http://dx.doi.org/10.1016/j.neuroimage.2004.05.011. Ramscar, M., Dye, M., & Klein, J. (2013). Children value informativity over logic in word learning. Psychological Science, 24(6), 1017–1023. Rossi, S., Gugler, M. F., Friederici, A. D., & Hahne, A. (2006). The impact of proficiency on syntactic second-language processing of German and Italian: Evidence from eventrelated potentials. Journal of Cognitive Neuroscience, 18(12), 2030–2048. Sagarra, N., & Herschensohn, J. (2010). The role of proficiency and working memory in gender and number agreement processing in L1 and L2 Spanish. Lingua, 120(8), 2022–2039. Sagarra, N., & Herschensohn, J. (2011). Proficiency and animacy effects on L2 gender agreement processes during comprehension. Language Learning, 61(1), 80–116. Schwartz, A. I., & Kroll, J. F. (2006). Bilingual lexical activation in sentence context. Journal of Memory and Language, 55(2), 197–212. Shtyrov, Y. (2011). Fast mapping of novel word forms traced neurophysiologically. Frontiers in Psychology, 2(340). http://dx.doi.org/10.3389/fpsyg.2011.00340. Shtyrov, Y. (2012). Neural bases of rapid word learning. The Neuroscientist, 18(4), 312–319. Shtyrov, Y., Nikulin, V. V., & Pulvermüller, F. (2010). Rapid cortical plasticity underlying novel word learning. The Journal of Neuroscience, 30(50), 16864–16867. Smith, L., & Yu, C. (2008). Infants rapidly learn word-referent mappings via crosssituational statistics. Cognition, 106(3), 1558–1568. Steinhauer, K., White, E., & Drury, J. (2009). Temporal dynamics of late second language acquisition: Evidence from event-related brain potentials. Second Language Research, 25(1), 13–41. Storkel, H. L., Armbruster, J., & Hogan, T. P. (2006). Differentiating phonotactic probability and neighborhood density in adult word learning. Journal of Speech, Language, and Hearing Research, 49(6), 1175–1192. Strenze, T. (2007). Intelligence and socioeconomic success: A meta-analytic review of longitudinal research. Intelligence, 35(5), 401–426. Titone, D., Libben, M., Mercier, J., Whitford, V., & Pivneva, I. (2011). Bilingual lexical access during L1 sentence reading: The effects of L2 knowledge, semantic constraint, and L1–L2 intermixing. Journal of Experimental Psychology: Learning, Memory, and Cognition, 37(6), 1412–1431. van den Noort, M. W., Bosch, P., & Hugdahl, K. (2006). Foreign language proficiency and working memory capacity. European Psychologist, 11(4), 289–296. van Hell, J. G., & de Groot, A. M. B. (2008). Sentence context modulates visual word recognition and translation in bilinguals. Acta Psychologica, 128(3), 431–451. van Hell, J. G., & Tanner, D. (2012). Second language proficiency and cross-language lexical activation. Language Learning, 62, 148–171. Waring, R., & Takaki, M. (2003). At what rate do learners learn and retain new vocabulary from reading a graded reader. Reading in a Foreign language, 15(2), 130–163. Webb, S. (2008). The effects of context on incidental vocabulary learning. Reading in a Foreign language, 20(2), 232–245. Williams, S. E., & Horst, J. S. (2014). Goodnight book: Sleep consolidation improves word learning via storybooks. Frontiers in Psychology, 5, 184. Wilson, M. (1988). MRC psycholinguistic database: Machine-usable dictionary, version 2. 00. Behavior Research Methods, 20(1), 6–10. Yu, C., & Smith, L. B. (2007). Rapid word learning under uncertainty via cross-situational statistics. Psychological Science, 18(5), 414–420. Yu, C., & Smith, L. B. (2011). What you learn is what you see: Using eye movements to study infant cross-situational word learning. Developmental Science, 14(2), 165–180. Yu, C., Zhong, Y., & Fricker, D. (2012). Selective attention in cross-situational statistical learning: Evidence from eye tracking. Frontiers in Psychology, 3(148). http://dx.doi. org/10.3389/fpsyg.2012.00148. Zahar, R., Cobb, T., & Spada, N. (2001). Acquiring vocabulary through reading: Effects of frequency and contextual richness. Canadian Modern Language Review/La Revue canadienne des langues vivantes, 57(4), 541–572. Zheng, Y., & Cheng, L. (2008). Test review: College English Test (CET) in China. Language Testing, 25(3), 408–417.