Metalinguistic Awareness and the Vocabulary–Comprehension Connection

To a great extent, our interest in vocabulary growth is motivated by the correlation between vocabulary knowledge and reading comprehension: people with bigger vocabularies also tend to be better readers. However, before we can draw any implications for instruction from that correlation, we need to understand what sorts of causal relationships lie behind it. It turns out that these relationship are rather complex (Anderson & Freebody, 1981; RAND Reading Study Group, 2002). In this section, I explore one particular aspect of the relationship between vocabulary knowledge and comprehension: the hypothesis that a significant portion of the variance underlying the correlation between tests of vocabulary knowledge and tests of reading comprehension can be accounted for by metalinguistic awareness. At the end of the chapter, I discuss some implications of this hypothesis for literacy instruction.

Anderson and Freebody (1981) suggested three possible types of causal links between vocabulary knowledge and reading comprehension. The first they labeled the instrumentalist hypothesis. This is the common-sense idea that knowing more words per se makes you a better reader. A second hypothesis is the knowledge hypothesis: that it is one’s store of concepts and the relationships among them that drives comprehension, with vocabulary knowledge simply being the visible tip of the conceptual iceberg. Their third hypothesis is the aptitude hypothesis. According to this hypothesis, vocabulary knowledge and reading comprehension are correlated with each other because both are impacted by a common set of aptitudes or abilities. What makes a person a good comprehender also makes a person a good word learner. This could be true even if knowing more words did not have a direct impact on reading comprehension. The implications of the aptitude hypothesis depend on which aptitudes or abilities one has in mind. If the primary factor contributing to both word learning and reading comprehension is simply general intelligence, for example, there might be little that could be done in terms of instruction. On the other hand, the abilities that are hypothesized to underlie the correlation between vocabulary knowledge and reading comprehension might be amenable to instruction. Sternberg and Powell (1983) suggested that the abilities linking word learning and reading comprehension had to do with making inferences, and that instruction could improve students’ abilities to make inferences about the meanings of unfamiliar words (Fukkink & de Glopper, 1998; Sternberg, 1987). In this chapter, I explore the hypothesis that the abilities shared by word learning and reading comprehension are primarily metalinguistic in nature. Like Sternberg and Powell’s (1983) version of the aptitude hypothesis, mine is an optimistic one, given that metalinguistic abilities are demonstrably teachable (e.g., National Reading Panel, 2000). For convenience, I refer to this as the metalinguistic hypothesis.

William Nagy