Electronic Thesis and Dissertation Repository

Thesis Format



Master of Arts




Dr. Frank Boers


Adding glosses to a reading text is expected to be helpful for second (L2) language learners, especially when inferring the meaning of words or expressions is challenging. This project examines the use of glosses to foster comprehension and retention of L2 idioms (e.g., go against the grain and stick to your guns). More specifically, it compares the benefits of different types of information in glosses: simply clarifying the idiomatic meaning, clarifying the literal meaning from which the idiomatic meaning is derived, or clarifying both. The participants were 37 Chinese ESL learners who read texts with one of the three types of glosses, and then sat immediate and one-week delayed post-tests. Fifteen of them also participated in a stimulated recall interview with the researcher. The post-tests showed no significant difference in the overall effectiveness of the three gloss types for idiom learning. The interview data, however, revealed substantial variation in the ways participants approached the glossed texts, regardless of the reading condition they had been assigned to. The interview data also suggested that the effectiveness of providing information about the literal underpinning of an idiom depends on how easy it is for the individual learner to appreciate the connection between this literal use and the idiomatic meaning.

Keywords: Second language acquisition; idioms; reading with glosses; incidental vocabulary acquisition; mixed-methods research.

Summary for Lay Audience

Second language learners often fail to understand idiomatic expressions (e.g., follow suit, throw in the towel, it goes against the grain, a wet blanket) in discourse, including reading texts. One solution is to add glosses to the texts to clarify the meaning of such expressions. Apart from assisting text comprehension, glosses have the potential to help learners remember idiomatic expressions. However, the effectiveness of glossing for the latter purpose may depend on the kind of glosses that are added to the text. An interesting possibility with regards to idioms is to include information not just about the current figurative meaning of the expression but also about the context in which it was originally used literally. For example, the idiom learn the ropes (learn how to do a task) has its origin in seafaring, where a novice sailor had to learn how to handle the ropes on a sailing vessel. It is hypothesized that this kind of information can make the expressions more memorable. This study, therefore, compares the effectiveness of glosses with and without notes about the original use of such expressions for learners' comprehension and retention of idioms. Moreover, it is hypothesized that providing glosses only with information about the literal use of the expressions will pique the learners’ curiosity about the actual meaning of the idioms and engage them in efforts to infer the latter, which may be expected to be beneficial for retention in memory as well.

37 Chinese ESL students from Western University in Canada were divided into three groups. They read two articles that each contained five idiomatic expressions and answered several content-related questions. The texts were accompanied by marginal glosses about the idioms, but these glosses differed depending on the group the participants were assigned to. One group read the texts with glosses that simply explained the contemporary, figurative meaning of the idioms—corresponding to their use in the texts); another group was given glosses that only explained the original, literal use of the expressions; and the third group was given glosses which presented both types of information. They were asked to complete a number of tests shortly after the reading activity and again one week later to compare how much idiom learning happened under the three reading conditions. After the final tests, 15 participants (5 from each group) were invited for an interview in which they were asked to recall how they approached the reading texts, the glosses, and the idioms. The test results showed no relationship between reading conditions and learning gains. However, the interviews revealed substantial variation in how the participants approached the reading tasks and the information about the idioms, often regardless of the type of glosses they were given. The nature of the idioms also helped to explain why a certain type of gloss (e.g., literal-origin only) was more suitable for some items than for others. The study thus illustrates the usefulness of using different sources of information (in this case test and interview data).