2.3 Collocation Competence
2.3.2 Collocational Problems and Errors
Many linguists, lexicographers, and translators have recognized collocation competence as one of the major problems for L2 learning. For instance, Bahns and Eldaw (1993) pointed out that learners’ knowledge of collocations was inferior to general vocabulary knowledge and collocation use was a major problem for advanced learners. Waller (1993) compared collocation use of L1 and L2 learners in writing and concluded that of all the features in the non-native speakers’ writings, collocational expression appeared to be the most tangible marker, as “a foreign accent in writing”
(p. 224). Farghal and Obiedat (1995) examined the productive knowledge of English collocations in written tasks and found that L2 learners (Arabic students) could not cope with collocations due to unawareness of collocations. Strategies of lexical
simplification such as synonymy, paraphrasing, avoidance, and transfer, as well as
literal translation from L1, were employed in the erroneous production process.Howarth (1998) investigated collocation errors of non-native advanced learners and remarked that collocation errors reduced precision and obscured clarity in academic writing.
Similarly, Källkvist (1998) analyzed the types of collocational errors made by advanced Swedish learners of English and found that most errors in verb use involved
high-utility transitive verbs such as “make”, “put”, and “get”. She posited that the
most common type of collocation errors was overextension which resulted in awkward collocations, indicating unawareness of usage restrictions. DeCock et al.(1998) noted that L2 advanced learners used the idiom principle, the prefabs, as native speakers did, but the chunks they used were different in frequencies, syntactic uses and pragmatic functions. Nesselhauf (2003) analyzed the verb-noun collocations in texts written by German learners of English and found that the highest rate of mistakes occurred in the word combinations with a medium degree of restriction.
These erroneous combinations usually involved a verb, such as “exert”, “perform”, or
“reach”, collocated with a wide, but restricted, range of nouns. The least problematic verb-noun collocations turned out to be with a high degree of restriction, but not fixed as idiomatic expressions. Nesselhauf therefore concluded that “collocations with a low degree of restriction are the most difficult kind of combinations for the learners”
(p.234).
Dechert and Lennon (1989) and Lennon (1996) conducted error analysis in learners’ essay and explained the reasons of learners’ errors in terms of the two factors, extra-casual and intra-casual blends. The extra-casual blends resulted from an incoherent composition of their thoughts, improperly differentiated, ordered and linked to form a coherent thought because the ability to express their thought appropriately had priority over their ability to organize and relate thoughts coherently.
The intra-casual blends emerged because learners were not taught in details how
lexical items in English might conjunct at the multiword level. Therefore, it caused learners’ collocational deficiency. Nasselhauf (2003) suggested two sources of L2 learners’ difficulties in verb-noun collocations: (a) under the great influence of L1, (b) under the linguistic situations in which congruent expressions between L1 and L2 could not be found.
In the context of Taiwan, L2 collocation errors have also received much research attention. For instance, Liu (1999), and later Li (2005), both examined miscollocations made by Taiwanese undergraduates and found that collocation errors in verb-noun/pronoun and verb- propositional phrases occurred most frequently. Shei (2000) investigated collocation errors of high-frequency verbs based on Taiwan
Learner Corpus of English (TLCE) and concluded that synonyms with different
collocation behaviors had a strong influence on both grammatical and lexical collocations. Chen (2002) examined collocation errors in Taiwanese high school students’ writings and found that adjective-noun and verb-noun miscollocations were the most frequently-made errors. Liu (2002) conducted a corpus-based study on written text of Taiwanese EFL learners at high school and college levels. Liu attributed verb-noun miscollocations to three main reasons: (a) L1 interference, (b) false concepts hypothesized (misuse of de-lexicalized verbs), and (c) lack of knowledge about collocational restrictions in semantic-related lexemes such as synonyms, hypernyms, and troponyms. Four common types of verb-noun miscollocations made by Chinese learners of English were also summarized: (a) synonyms verbs with different collocational restrictions, e.g., “carry out the goal”instead of “reach the goal”; (b) hypernym and troponym verbs involving different semantic properties and collocation behaviors, e.g., “break the foundation” instead of
“damage the foundation” (“break” is the troponym of “damage”); (c) verb-noun collocations without translational equivalents in Chinese and English, e.g., “open the
computer” instead of “turn on the computer”; (d) de-lexicalized verbs, such as “make”
and “take”. The results showed that verb-noun miscollocations occupied the main portion of learners’ miscollocations. In addition, verb-based errors were produced more frequently than noun-based errors, indicating that learners had greater difficulty in selecting and using felicitous verbs to combine with nouns.
Shei (2000) observed that varied content words in lexical collocations made it more difficult for users to distinguish nuances among lexical choices and appropriateness. Chen (2002) pinpointed L1 influence as one of the common causes of miscollocations and remarked that students constantly used the literal translation strategy to produce collocations. Chang and Yang (2009) reported that learners had difficulties with the use of verbs and delexical verbs, and L1 transfer and induced errors were the two most common causes contributing to Chinese learners’
miscollocations. Liao (2011) also showed that Chinese learners’ English collocation errors were derived from both Mandarin negative transfer and word usage problems.
Based on the previous research findings, it is evident that collocation is an area of particular difficulty for L2 learners, warranting further attention and a more prominent role within L2 learning. It is generally accepted that L2 learners’
collocation errors are higher than other errors they committed (Ellis, 2001).
Learners’
L2 productions revealed mistaken examples of collocation transfer from L1, poor use of
phraseology, and the creation of atypical combinations.
The collocation error typology distinguished between three parallel dimensions. The first dimension captured whether the error concerned the collocation as a whole or one of its two elements,base and collocate. As in the multilevel-annotation of Falko-corpus (Lüdeling et al.,
2005) and in accordance with Tono’s (2003) study, the second dimension conducted error analysis, and the third explanatory analysis. Each dimension was captured by a typology tree the intermediate nodes of which were error classes, indicating concreteerror types served as annotation labels in the corpus.
While it was acknowledged that most empirical researchers were uncertain with the idea of measuring avoidance, it has been shown that focusing on error alone was an inadequate method of identifying problems in L2 collocation acquisition.
Linguistic errors should be considered as part of combinations which included avoidance and calquing from the L1 to match L2 patterns (a fortuitous accuracy).
Both of them could be measured contrastively: avoidance was determined by its low frequency in relation to equivalent native speaker texts in the L2 (Martelli, 2006), while fortuitous accuracy could be identified through the comparison of back-translated forms in a general reference corpus in learners’ L2. The combination of errors, avoidance and fortuitous accuracy, captures a more detailed picture of the strategies that advanced L2 learners use in their language production.