Webb: The effects of synonymy on second language vocabulary learning 121 Reading in a Foreign Language 19(2) The present study was designed to determine whether learning synonyms for known second language (L2) words is easier than learning non synonyms. The reason for this was not to advocate the teaching of synonyms but rather to improve our understanding of the vocabulary acquisition process. If words with known synonyms are learned more easily than words without known synonyms, learning new words would become easier as a learner's vocabulary size increases because more and more words are likely to have known synonyms. Vocabulary difficulty would thus be affected by the amount of prior learning. Learners with larger vocabularies would be able to learn words more easily than learners with smaller vocabularies because increased vocabulary knowledge would help to facilitate learning. Literature Review In the only study that investigated the effects of synonymy on vocabulary learning, Higa (1963) found that pairs of synonyms took longer to learn than pairs of unrelated words. The results suggested that learners are more likely to confuse words that are similar in meaning than words that do not have close semantic links. Higa (1963, 1965) reported that the closer the semantic relationship between words, the more difficult it may be to learn the words in a set. Studies by Tinkham (1993) and Waring (1997) also indicate that learning sets of semantically related words is more difficult than learning words that are not linked by meaning. Together, these studies suggest that learning synonyms together may reduce the likelihood of acquisition. While this is a very useful finding, it may not apply to the way in which synonyms are usually learned. Teachers are unlikely to teach pairs of unknown synonyms together for the simple reason that it is confusing for learners, as Higa's research suggests, and learners are unlikely to learn synonyms together because they may lack the motivation to learn two words that convey similar information. Instead, a synonym for a known word may be easier to learn in the later stages of vocabulary learning. At present, no research appears to have investigated the difficulty of learning the synonyms of known words versus learning non synonyms. Laufer (1990) reported that synonymy is one of seven interlexical factors that can reduce the chances of vocabulary acquisition. She gave two reasons that synonyms may be more difficult to learn than other words. First, learners often make mistakes using synonyms because some of them may be substituted effectively in some contexts but not in others. For example, strong and powerful have similar meanings, but usually tea is only strong, and engines may be powerful but are rarely strong. This is an important point and demonstrates that some synonyms might be more difficult to learn than others because synonyms with similar meanings do not always have the same collocates. Certainly, some words such as good and nice are synonymous in many contexts, while others such as powerful and strong are synonymous to a lesser degree. The second reason she gave is that less advanced learners are unlikely to try to learn words with similar meanings when they have a greater need to learn unknown L2 meanings. Learning words with known synonyms may be easier than learning words without known synonyms because learners may be able to transfer their knowledge of syntax and collocation from known synonyms to less frequent synonyms. Typically, when learning a non synonym, L2 learners use their first language (L1) knowledge of that item and information from the context in which it was encountered to help learn that word. While on some occasions, this combination of
Webb: The effects of synonymy on second language vocabulary learning 122 Reading in a Foreign Language 19(2) L1 and L2 knowledge may be sufficient to use or understand the word quickly, more often, learning the word is likely to be a slow process that involves repeated encounters of the word in context. Learning the synonyms of known words may be faster than learning non synonyms because learners may gain substantial L2 vocabulary knowledge from known synonyms. Learners may gain L2 knowledge of syntagmatic association and collocation, paradigmatic association, and grammar from known synonyms. A transfer of L2 knowledge from known words to their less frequent synonyms could make using and understanding the synonyms relatively easy. Nation's (2001) discussion of "learning burden" also suggests that learning a synonym for a word that is already known may be easier than learning a non synonym. He argued that the amount of effort required to learn a word is different for different words and for different learners. He referred to the effort needed to learn a word as its learning burden and stated The general principle of learning burden is that the more a word represents patterns and knowledge that the learners are already familiar with, the lighter its learning burden. These patterns and knowledge can be from the first language, from knowledge of other languages, and from previous knowledge of the second language. So, if a word uses sounds that are in the first language, follows regular spelling patterns, is a loan word in the first language with roughly the same meaning, fits into roughly similar grammatical patterns as in the first language with similar collocations and constraints, then the learning burden will be very light. (pp. 23 24) The concept of learning burden provides a reasonable explanation as to why L2 words that have known L2 synonyms may be easier to learn than those that do not. Synonyms may represent different aspects of knowledge that have already been acquired. For example, if learners are faced with learning the low frequency words revolver and spear, learning revolver may be easier because it has a high frequency synonym (gun) that represents vocabulary knowledge that can be used to learn revolver. The word revolver can be substituted for gun in many sentences, allowing learners to produce sentences such as he held the revolver tightly in his hand, he pointed the revolver and fired, and she took out her revolver and shot him three times in the head. If a newly learned word can be substituted in a sentence for a known synonym, then collocations and syntagmatic associates might be acquired when meaning and form are learned. Because spear does not have a high frequency synonym, a greater amount of vocabulary knowledge may need to be acquired, and therefore, it may be more difficult to produce in a sentence. Although the degree of overlap of vocabulary knowledge varies from synonym to synonym, at least partial overlap may help facilitate acquisition. Further research examining the difficulty of learning words that have known L2 synonyms versus learning words that do not would be valuable for three reasons. First, it would help researchers better understand the vocabulary acquisition process. If vocabulary knowledge gained from knowing words may facilitate the future learning of their synonyms, unknown words may become easier to learn as a person's vocabulary develops. Second, it would provide some support for L2 incidental vocabulary learning through reading. L2 research has shown that learners may incidentally acquire vocabulary through reading and suggests that it may play a large role in L2 vocabulary development (Day, Omura, & Hiramatsu, 1991; Dupuy & Krashen,
Webb: The effects of synonymy on second language vocabulary learning 123 Reading in a Foreign Language 19(2) 1993; Pitts, White, & Krashen, 1989). However, little evidence supports the position that incidental learning is responsible for the huge number of words acquired in L1 learning. L2 studies examining incidental learning have shown that vocabulary gains are extremely small and that incidental learning is a gradual process in which repeated encounters in context are needed to learn words (Horst, Cobb, & Meara, 1998; Rott, 1999; Waring & Takaki, 2003). As a learner's vocabulary increases, the number of synonyms for known words that they learn is also likely to increase. If learning the synonyms of known words is easier than learning words without known synonyms, it would help to explain how some L2 learners may incidentally learn great numbers of words in the later stages of language learning through reading and listening. Third, determining whether synonyms are acquired more easily than other words would be very useful for teachers and learners. Knowing which words are easier or more difficult to learn would allow them to develop more efficient vocabulary teaching and learning strategies. The present study was designed to investigate the effects of synonymy on vocabulary learning. Two sets of target words 10 low frequency words with high frequency synonyms and 10 low frequency words without high frequency synonyms were used to determine whether words with high frequency synonyms are more easily learned than non synonyms. Disguised forms replaced the L2 forms of the target words to eliminate the possibility that the learners knew the target words. The target words were learned in two tasks: learning word pairs and learning glossed sentences. The effects of the two tasks on the different aspects of vocabulary knowledge are discussed in detail in Webb (2007). Using L2 target words with high frequency synonyms allowed the participants the possibility of gaining L2 knowledge from their L1 translations, their L2 synonyms, and the contexts in which they were presented in the glossed learning task. In contrast, using target words without high frequency synonyms ensured that participants could only gain knowledge of those words from their L1 translations or the glossed sentences. Because synonymy may affect several aspects of vocabulary knowledge, each target word was tested in 10 different ways. The tests were designed to measure receptive and productive knowledge of (a) orthography, (b) syntagmatic association, (c) paradigmatic association, (d) grammatical functions, and (e) meaning and form. Researchers tend to agree that knowing a word is much more than simply knowing its meaning (Aitchison, 1994; Laufer, 1997; McCarthy, 1990; Miller, 1999; Nation, 1990, 2001; Richards, 1976), and measuring multiple aspects of vocabulary knowledge may provide a more accurate assessment of the effects of synonymy on vocabulary knowledge than simply testing meaning. Specifically, the present study examined the following questions: 1. Are synonyms for known words easier to learn than words that do not have known synonyms? 2. How are different aspects of vocabulary knowledge affected by synonymy? If learners can substitute a synonym for a word in a text, the learners could be expected to score higher on tests measuring knowledge of syntagmatic association, paradigmatic association, and grammatical functions for target words with known synonyms than target words without them.
Webb: The effects of synonymy on second language vocabulary learning 124 Reading in a Foreign Language 19(2) Method Participants The participants were 84 Japanese students of English as a foreign language from two first year classes at a university in Fukuoka, Japan. All of the students had scored 80% or higher on the 2000 word level of Version 1 of the Vocabulary Levels Test (Schmitt, 2000). Their average raw score was 27.1/30, indicating that they were well in control of that level and that they had receptive knowledge of almost all of the 2,000 most frequent words (Schmitt, Schmitt, & Clapham, 2001). Design Two sets of words were used: 10 low frequency words with high frequency synonyms and 10 low frequency words without high frequency synonyms. Disguised forms replaced the L2 forms of the low frequency words and were matched with their L1 meanings to create the target words used in the experiment. Using the disguised forms ensured that the participants would have no prior knowledge of the target words. The use of the disguised forms also eliminated the need to use pre tests. Pre tests may not be sensitive enough to measure for partial knowledge of target words and may also tip students off as to the nature of a study. The participants were randomly assigned to two learning conditions. In one condition, the participants encountered the target words in glossed sentences, and in the other, the participants encountered the target words in word pairs. In each condition, the participants attempted to learn all 20 target words. The participants were given 8 minutes to learn the target words in both conditions. After the conclusion of the treatment, a surprise vocabulary test was administered. Five types of word knowledge orthography, paradigmatic association, meaning and form, syntagmatic association, and grammatical functions were isolated and measured in a series of 10 tests. The experiment was conducted within one 90 minute class period. Materials Target words. Twenty low frequency words were selected from the fifth frequency band in the COBUILD dictionary. It accounts for the 6,601 st to 14,700 th most frequent words. Ten were low frequency words with high frequency synonyms, and the other 10 were low frequency words without high frequency synonyms. Both sets of target words comprised six nouns and four verbs. This ratio of nouns to verbs was used because nouns and verbs are the most common parts of speech found in natural text, and the 6:4 ratio approximates their proportional frequency of occurrence (Kucera & Francis, 1967). The target words with high frequency synonyms were locomotive, visage, lane, abode, boulder, crave, doze, sob, abhor, and dagger. The target words without high frequency synonyms were lick, spear, recluse, pawn, landfill, mourn, convent, pier, reef, and marinate. Two sets of 10 disguised forms were created and replaced the L2 forms of the 20 low frequency words. All of the disguised forms were two syllables and resembled English words phonetically and orthographically. Because the participants could have mistaken the disguised forms for real
Webb: The effects of synonymy on second language vocabulary learning 125 Reading in a Foreign Language 19(2) words that were orthographically similar, the spellings of the disguised forms did not always conform to common spellings. However, in pilot tests, Japanese learners were able to pronounce all of the words correctly and reported that they believed the disguised forms to be authentic English words. Because the participants in the experiments did not know that disguised forms were used, learning the disguised forms simulated for the learners the experience of learning actual words. In each frequency group, seven of the disguised forms were five letters long, and three were six letters long. The 20 disguised forms were ancon, cader, dangy, denent, faddam, hodet, masco, pacon, sagod, tasper, copac, gishom, hattaw, ictay, mesut, nasin, nuggy, tagon, toncop, and dapew. Two versions of each learning condition were prepared to ensure that the disguised forms paired with one set of target words were not easier to learn than the other set of disguised forms. The first version of the learning tasks presented the first 10 disguised forms (ancon, cader, dangy, denent, faddam, hodet, masco, pacon, sagod, and tasper) paired with the L1 meanings of the words with high frequency synonyms and the second 10 disguised forms (copac, gishom, hattaw, ictay, mesut, nasin, nuggy, tagon, toncop, and dapew) paired with the L1 meanings of the words without high frequency synonyms. In the second version, the pairings were reversed. Because the disguised forms used in the experiment were the same for both sets of target words (using two versions of each task), the factors affecting word difficulty pronounceability, orthography, morphology, and synformy could not affect any differences found between the variables. Two other factors affecting word difficulty, specificity and register constraints, are more likely to affect the later stages of acquisition and were therefore unlikely to be a factor in this study. To ensure that idiomaticity and multiple meanings did not influence vocabulary learning, none of the disguised forms encountered in sentence contexts involved idioms and only one sense of a target word's L1 and L2 meaning was presented in the contexts. Tasks Word pairs. In this task the 20 disguised forms were presented on the right of their L1 translations on one page. In the following example, the target word locomotive was replaced with the disguised form masco and paired with the Japanese translation of locomotive: masco The order of the word pairs, those with and without high frequency synonyms, was controlled to ensure that it did not affect learning. The target words with high frequency synonyms were alternated with the target words without high frequency synonyms. For half of the participants, a disguised form with a high frequency synonym was presented first. For the others, a disguised form without a high frequency synonym was first. Sentence contexts. In the task of learning from glossed sentences, the 20 target words were presented with their L1 meanings and one sentence containing each target word followed each word pair (see Appendix). The following example is for the target word masco, which was matched with the L1 meaning of locomotive: masco The driver got off the masco .
Webb: The effects of synonymy on second language vocabulary learning 126 Reading in a Foreign Language 19(2) The sentences were selected from the British National Corpus. Pilot studies and the participants' scores on the Vocabulary Levels Test indicated that the participants would know all of the running words in the sentences and would be able to quickly understand the sentences. The order of the contexts was controlled in the task. Half of the participants received contexts containing target words with high frequency synonyms first, and half of the participants received contexts without high frequency synonyms first. Dependent measures Ten tests measuring knowledge of orthography, paradigmatic association, syntagmatic association, grammatical functions, and meaning and form were administered after the treatment. Each test was created to isolate a specific aspect of knowledge. Five of the tests measured productive knowledge, and five measured receptive knowledge. The tests were carefully sequenced to avoid earlier tests affecting the participants' answers on later tests. Each test was on one page. The first test, measuring productive knowledge of orthography, was administered to all of the participants at one time because it involved aural cues spaced 12 seconds apart. The learners were given as much time as they needed to complete the rest of the tests. When a participant finished a test, it was collected, and the next test was handed out. Webb (2005) described the tests as follows: Test 1: Productive knowledge of orthography. On the first test the learners heard each disguised form pronounced twice and then had 10 seconds to write it correctly. Any spelling mistakes caused an answer to be marked incorrect. Because the learners were at the intermediate level and were likely to have learned most if not all of the rules of spelling, aural cues would be enough to lead them to write at least a close approximation of the target words. If responses with minor spelling mistakes were marked as correct, then whether the spelling was due to the learning task or the aural prompt could not be determined. On all of the other productive tests, spelling was not a determining factor in the scoring if the response could be clearly understood. Test 2: Receptive knowledge of orthography. On the second test the learners had to circle the correctly spelled target words, which appeared with three distracters. The distracters were created to resemble the target words both phonetically and orthographically. The following examples are for the target words dangy and hodet. (a) dengie (b) dengy (c) dungie (d) dangy (a) hodet (b) holat (c) halet (d) hedet Usually, all productive tests need to be completed before receptive tests to avoid a learning effect. However, the receptive test of orthography was highly unlikely to contribute to a learning effect in this study for two reasons: the target words were used as cues on three of the four remaining productive tests, and the determining factor on the other test measuring productive knowledge of meaning and form was whether the learners could link the L2 form with its L1 meaning rather than spelling the L2 form correctly. Test 3: Productive knowledge of meaning and form. On the third test, productive knowledge of
Webb: The effects of synonymy on second language vocabulary learning 127 Reading in a Foreign Language 19(2) meaning and form was measured using a translation test. The learners were given the L1 meanings and asked to write the words that the meanings had been paired with in the treatment. The aim of this test was to determine whether the learners could link the L2 forms of the target words with their L1 meanings. For example, to get the following answer correct, the learners had to write the target disguised form masco beside the L1 translation with which it was paired. Because the aim of the first two tests was to determine whether learners could write the target words correctly and recognize the correct spellings of the target words, spelling was not the determining factor for a correct answer on this test. Therefore, spellings that demonstrated that the learners could link an L2 form with its L1 meaning were marked correct. In the example above, close approximations of the target word masco (e.g., mosco, masko, or mascoe) were acceptable responses. Close approximations that were real English words were marked as incorrect because whether the participants had intended to write the target words could not be ascertained. Test 4: Productive knowledge of grammatical functions. This was essentially a sentence construction test. The learners were cued with the target words and had to write each one in a sentence. The instructions made clear that the determining factor for a correct response was using the target words with grammatical accuracy. For example, the target word masco (locomotive) would have been scored as incorrect in The girl mascoed to school and correct in both The masco left the station early and It is a masco. Test 5: Productive knowledge of syntagmatic association. In word association research, responses are often classified as either paradigmatic (demonstrating productive knowledge of the semantic relationships between words) or syntagmatic (demonstrating productive knowledge of syntax; Soderman, 1993). On this test the learners had to produce L2 syntagmatic associates beside the cues, which were the target words. For example, for the target word masco, which had been paired with the Japanese translation of locomotive, acceptable responses were words commonly encountered in context with locomotive, such as station, tracks, left, and arrived. Words less frequently found in context with locomotive, such as clock, ate, and hard, were marked as incorrect. Two native speakers of English evaluated the responses. Their inter rater reliability was 94%. Because the following test measured productive knowledge of paradigmatic association, paradigmatic associates were marked as incorrect. This was carefully explained in the instructions. Test 6: Productive knowledge of paradigmatic association. In this test learners were presented with the target words and asked to write an associate beside each item. Coordinates, superordinates, subordinates, antonyms, and synonyms were all scored as correct. Because the previous test measured productive knowledge of syntagmatic association, syntagmatic associates were marked as incorrect, as was carefully explained in the instructions. Examples of acceptable responses for masco (locomotive) are train, airplane, and vehicle. Two native speakers of English evaluated the responses, and their inter rater reliability on this test was 98%. Test 7: Receptive knowledge of grammatical functions. A multiple choice test was used to
Webb: The effects of synonymy on second language vocabulary learning 128 Reading in a Foreign Language 19(2) measure receptive knowledge of grammatical functions. The learners were presented with three sentences containing each target word and had to choose the correct one. Knowledge of the target word's part of speech enabled the learners to select the correct answer. Because the productive test of grammatical functions was perhaps the most demanding test, the receptive test of grammatical functions needed to be sensitive to smaller gains. If the learners knew that masco (locomotive) was a noun, they would be able to choose (a) as the correct answer in the following example. (a) It is a masco. (b) It mascoed. (c) It is very masco. Test 8: Receptive knowledge of syntagmatic association. On this test the learners circled the responses that were most likely to appear in context with the target words. All distracters were words that the learners were likely to know and were the same parts of speech as the correct answers. The following examples are for the target words dangy (boulder) and hodet (lane). dangy (a) fall (b) wash (c) walk (d) catch hodet (a) drive (b) sit (c) take (d) know Test 9: Receptive knowledge of paradigmatic association. On this test, the learners had to circle the responses that were paradigmatic associates of the target words. All distracters were words that the learners were likely to know and were the same parts of speech as the correct answers. This is illustrated for the target words dangy (boulder) and hodet (lane) in the following examples. dangy (a) stone (b) plant (c) tree (d) person hodet (a) park (b) highway (c) garden (d) building Test 10: Receptive knowledge of meaning and form. This was a receptive translation test in which the target words cued responses in the L1. An argument can be made that a receptive translation test involves both receptive and productive processes (Waring, 1999); however, the same argument can be made for most receptive tests. Because the learners had learned the answers in the treatments, a recognition test would have been extremely easy. The receptive translation test was better suited for this experiment because it was more demanding, requiring learners to recall rather than recognize meanings. In the following example, the learners were required to write the Japanese translation of locomotive beside the target disguised form masco. masco ____________ Results The descriptive statistics (means, standard deviations, and numbers of participants) of the vocabulary knowledge scores on the 10 dependent measures for words with and without high frequency synonyms are reported in Table 1. A repeated measures MANOVA was performed
Webb: The effects of synonymy on second language vocabulary learning 129 Reading in a Foreign Language 19(2) using the scores on the 10 dependent measures to analyze the within subjects factor (synonymy words with and without known synonyms), and the two between subjects factors (test version and learning condition). The MANOVA showed an overall statistically significant difference for synonymy, F(10, 71) = 9.74, p < .001, 2 = .58. However, the different versions of the treatment showed no significant difference, F(10, 71) = 1.50, p = .16, 2 = .17, nor did the learning conditions, F(10, 71) = 1.50, p = .59, 2 = .11. Table 1. Means and standard deviations of learning conditions on dependent measures for words with and without high frequency synonyms Word pairs (N = 43) Context (N = 41) Word pairs and context combined (N = 84) M SD M SD M SD PO Synonyms 5.56 (2.24) 5.37 (2.61) 5.46 (2.41) Non synonyms 5.42 (2.27) 4.95 (2.52) 5.19 (2.39) RO Synonyms 7.53 (2.02) 7.71 (1.82) 7.62 (1.91) Non synonyms 7.16 (2.16) 7.12 (2.01) 7.14 (2.08) PM Synonyms 5.51 (2.85) 5.41 (2.59) 5.46 (2.71) Non synonyms 5.70 (2.73) 5.05 (2.86) 5.38 (2.80) RM Synonyms 5.60 (2.93) 6.05 (2.77) 5.82 (2.85) Non synonyms 5.56 (2.81) 5.78 (3.11) 5.67 (2.94) PA Synonyms 4.88 (2.58) 5.37 (2.77) 5.12 (2.67) Non synonyms 3.47 (2.21) 3.51 (2.66) 3.49 (2.42) RA Synonyms 5.95 (2.50) 6.66 (2.09) 6.30 (2.32) Non synonyms 6.05 (2.54) 6.27 (2.31) 6.15 (2.42) PS Synonyms 4.37 (2.64) 4.27 (2.76) 4.32 (2.68) Non synonyms 3.74 (2.51) 3.39 (2.31) 3.57 (2.41) RS Synonyms 5.98 (2.61) 6.56 (2.26) 6.26 (2.45) Non synonyms 5.72 (2.58) 6.78 (2.13) 6.24 (2.41) PG Synonyms 4.88 (2.76) 5.12 (2.64) 5.00 (2.69) Non synonyms 4.91 (2.78) 4.71 (3.04) 4.81 (2.90) RG Synonyms 6.86 (2.60) 7.66 (1.62) 7.25 (2.20) Non synonyms 6.95 (2.49) 7.15 (2.07) 7.05 (2.28) Note. Maximum score = 10. PO = productive knowledge of orthography, RO = receptive knowledge of orthography, PM = productive knowledge of meaning and form, RM = receptive knowledge of meaning and form, PA = productive knowledge of paradigmatic association, RA = receptive knowledge of paradigmatic association, PS = productive knowledge of syntagmatic association, RS = receptive knowledge of syntagmatic association, PG = productive knowledge of grammar, RG = receptive knowledge of grammar. The results of the MANOVA indicate that synonymy may facilitate an increase in vocabulary knowledge. Significant differences were found on the productive tests of paradigmatic association, F(1, 80) = 75.30, p < .01, 2 = .49, and syntagmatic association, F(1, 80) = 13.06, p < .01, 2 = .14, and the receptive test of orthography, F(1, 80) = 5.15, p < .05, 2 = .06. A summary of the findings on the repeated measures MANOVA is shown in Table 2.