When foreign language learners are confronted with an unfamiliar word through intensive or extensive L2 reading, they must decide how to handle the word.Â This can be accomplished in a variety of ways, which include ignoring the word, checking a dictionary for the word's definition, and attempting to infer the meaning.Â Specifically, thisÂ will focus on inferencing strategies in L2 reading.Â More specifically, it will identify different inferencing strategies, how effective these inferencing strategies are, what factors can influence the success of inferencing, and what this means for teaching inferencing strategies in the L2 classroom.
Almost all literature on inferencing strategies distinguishes between strategies centered on the unfamiliar word itself and strategies that are contextually based.Â When referring to the word based strategies, Hamada (2009) used the term local strategies.Â In contrast, the phrase global strategies was used to describe the contextually based inferencing strategies.Â However, these terms only represent the broadest definition of inferencing strategies.
From their study on the role of linguistic knowledge, Kaivanpanah and Alavi (2008) identified seven different inferencing strategies.Â The first strategy they found was sentence level grammatical knowledge, which involves looking for relationships between words to determine meaning.Â This strategy would fall under Hamada's idea of local strategies, where the focus is more on the word than contextual clues within the text.Â Word morphology and considering class membership was another strategy, and this involved deconstructing word parts and examining the meaning of each part.Â The third strategy Kaivanpanah and Alavi found was analyzing the compound words into their constituents.Â This strategy is similar to the second in that it involves deconstructing words and examining the meaning of each part. Both of these strategies would be classified as a local strategy.Â The next strategy identified was sentence level semantic clues which is a local strategy, but one that is more sophisticated and shows a deeper word knowledge than the previously mentioned strategies.Â The next strategy involving discourse/text most resembles Hamada's global strategies as it attempts to look beyond the word, or even the sentence, level for clues to determine the meaning.Â The last two strategies Kaivanpanah and Alavi identified were homonymy/phonetic similiarity and collocation.Â The former compares words that sound similar in the second language and the latter uses knowledge of how words are usually paired together to infer meaning.
Nassaji (2006) identifies similar strategies but categorizes these strategies into three types: identifying, evaluating, and monitoring.Â Under the category identifying, there are three subtypes, two of which overlap with the Kaivanpanah and Alavi (2008) findings.Â The first is word analysis, and this aligns with word morphology and considering class membership, as well as analyzing the compound words into their constituents.Â The second is word-form analogy, and it overlaps with phonetic similarity.Â The final subtype of the identifying category is repeating, where the learning attempts to infer the meaning of a word by repeating it or any of the words near the unfamiliar word.Â Evaluating is the second type and it has two subtypes: verifying and self-inquiry.Â Monitoring is Nassaji's final type, which he defines as, "the learner shows a conscious awareness of the problem by judging its ease or difficulty" (p. 392).
While inferencing strategies are often known by different names, the underlying concept remains the same.Â Learners can utilize the type of strategies that Hamada (2009) identified as local strategies, where an attempt is made to infer the meaning of a word by analyzing the unfamiliar word with a narrow focus on the word itself, and global strategies that encourage the learner to utilize their knowledge of the outside world and their understanding of the text for clues as to meaning of the unfamiliar word.
In order for inferencing to be a valid option for learning new words in L2 reading, it needs to be effective.Â This paper will now turn to examining the effectiveness of inferencing in L2 reading.
In Hamada's (2009) study of the development of L2 word meaning inferences, over a period of four weeks, she studied five English learners and their ability to infer word meanings in weekly readings.Â For each participant at the end of the study, Hamada reported a mean success rate range of 50% to 32% and that multiple strategies were used when attempting to infer word meaning.Â There is no indication whether these figures take into account the number of words the participants guessed correctly and partially correctly, or if only correct guesses were included.Â Regardless, these results suggest that inferencing may not be very effective, though Hamada offers implications for the classroom, which will be addressed later in this paper.Â Although Nassaji (2006) distinguished between lexically skilled and lexically less skilled participants in his study of the connection between depth of knowledge and inferencing strategies, even the lexically skilled participants had a similar success rate to the participants in the Hamada (2009) study.
However, there are many factors that can affect the success of inferencing.Â Kaivanpanah and Alavi (2008), Hamada (2009), and Nassaji (2006) argue that the level of language proficiency plays a role in inferencing effectiveness.Â As previously mentioned, Nassaji (2006) made a distinction between learners who were lexically skilled and those who were lexically less skilled.Â He accomplished this by sorting the participants according to their scores on a depth of vocabulary knowledge test.Â After the participants were assigned to the lexically skilled or lexically less skilled group, they were tested on their inferencing skills.Â The lexically skilled group outperformed the lexically less skilled group.
Nassaji (2006) found that the lexically less skilled participants used more local strategies than the lexically skilled participants, but even when the lexically skilled participants used local strategies, they were more effective than their less skilled coparticipants.Â Kaivanpanah and Alavi (2008) had similar findings in their study with higher proficiency participants using more global strategies and lesser proficient participants using local strategies.Â Hamada's (2009) results echoed this by finding that high proficiency learners used more contextual clues, and that many participants in her study shifted from using local strategies to global ones.Â She further added that these contextual clues were more closely related to successful inference of unknown words.
Walters' (2006) findings indicate that with those who have been trained in inferencing strategies, the learners' level of proficiency does not play a large role in inferencing effectiveness.Â She adds to that by saying it is possible that learners at various proficiency levels may respond differently to inferencing training.Â This could imply that lower proficiency language learners benefit more from inferencing strategy training than more proficient learners.Â Hamada (2009) alluded to something similar in her study postulating that there may be a threshold of strategy use, and that success depends more on how the inferencing strategies are utilized.
Implications for the classroom
The Walters (2006) study perhaps offers the best data for classroom instruction.Â That is, her findings suggest that strategy training as well as teaching learners to use context clues can aid in inferencing success for learners of most levels of proficiency.Â She further goes on to say that more practice helps the learner to use his inferencing strategies more effectively.Â Hamada (2009) likewise stressed the importance of effective strategy use.Â Knowing at what time to use what strategy may even be more important than knowing the different inferencing strategies.Â This effective strategy use may further be an integral part of classroom instruction and interaction if there is indeed a threshold of strategy use.
Kaivanpanah and Alavi (2008) stress that more grammar instruction is needed to raise linguisitic knowledge because greater linguistic knowledge translates to more effective inferencing.Â They further warn teachers to pay attention during the text selection process, as texts that lack enough viable context clues increase the likelihood a learner's ability to infer the meaning of a new word will be severely limited or, in the worst case, fail.Â Kaivanpanah and Alavi conclude their study by stressing the relative inaccuracy of inferencing and that learners must also be taught how to check their inferences.
In his study, Nassaji (2006) demonstrated that lexically skilled learners were more effective inferencers due to their greater depth of vocabulary knowledge.Â These findings indicated that a greater focus on developing word knowledge needs to be implemented rather than simply increasing the size of a learner's vocabulary.
Inferencing holds a special place in second language reading due to the vast number of words that are unknown to the L2 learner.Â The ability of an L2 reader to effectively handle an unfamiliar word can affect not only reading comprehension, but also how the reader approaches unfamiliar words within a text in the future.Â To promote success in L2 reading, it is therefore critical that teachers be acquainted with inferencing strategies as well as how they can be utilized by learners to aid better reading comprehension.