Carol A. Chapelle - The Concise Encyclopedia of Applied Linguistics

72 Youn, S. J. (2015). Validity argument for assessing L2 pragmatics in interaction using mixed methods. Language Testing, 32(2), 199–225.

73 Young, R. (2008). Discursive practices in language learning and teaching, Malden, MA: Wiley‐Blackwell.

KEIKO KODA

Assessing reading ability requires a theory of what constitutes reading ability. However, conceptualizations of reading ability have dramatically changed over the past several decades. In the early days, the field was dominated by two diametrically opposing views, one regarding reading as an indivisible whole and the other seeing it as a constellation of distinct capacities. The former, holistic, view held that reading, just like speech, is learned naturally as a whole through communication during the course of human development. The central tenet of the view is the belief that meaning making should be a primary focus of reading instruction. According to the “psycholinguistic guessing game” model (Goodman, 1967, 1973), the primary task of the reader is to generate a hypothesis regarding the forthcoming content of a text based on real‐life experiences and prior knowledge. The reader needs to attend to some words in the text to confirm the hypothesis, but overall, the view focuses on top‐down, conceptually driven, processing with strong emphasis on the reader's knowledge and real‐life experiences as the basis for meaning making, and assigns only a minor role to word recognition and its subskills.

In the subsequent decades, however, the top‐down claims received little support from empirical research. Contrary to the predictions from the “psycholinguistic guessing game” model, eye movement studies have repeatedly shown that the majority of content words receive direct visual fixation (Just & Carpenter, 1980, 1987; Balota, Pollasek, & Rayner, 1985). Text comprehension research has also demonstrated that the emerging text interpretation could be disrupted even by a single word (Kintsch, 1998). Collectively, these findings suggest that the reader is engaged in analyzing word forms for retrieving their meanings during text comprehension.

Emanating from these and other findings were newer conceptualizations that encapsulate the interactive nature of reading. They give equal weight to conceptually driven top‐down processes and text‐based bottom‐up operations. In interactive models, reading is described as a dynamic process through which the reader connects graphically encoded linguistic information in a text with real‐life knowledge in memory. Similarly, comprehension is regarded as a product evolving gradually from the continual reader–text interaction. The interactive models assume that reading entails several interlinked operations, including word meaning retrieval (retrieving the meaning of individual words from their visual forms), word meaning integration (integrating word meanings into larger text units), and personalization (connecting text‐based meanings with the reader's real‐life experiences) (Koda, 2016; Koda & Miller, 2018). Of the three, word meaning retrieval and word meaning integration are primarily linguistic, necessitating distinct facets of linguistic knowledge for their execution (Koda, 2007). Thus, in current models, reading ability is seen as a broad‐based competence consisting of diverse skills for constructing meanings at all levels of text units. Under this view, the centrality of linguistic knowledge is underscored as the means by which the reader collects building blocks and assembles them into a coherent whole. Without adequate knowledge of the language, sequences of graphic symbols on the page will never be linked with the reader's real‐life knowledge.

Current views of reading that recognize the close interconnection between linguistic knowledge and reading ability require a careful look at the specific ways in which diverse facets of linguistic knowledge contribute to word meaning retrieval and word meaning integration.

Well‐developed orthographic knowledge affords rapid and effortless access to a word's meaning. A good reader recognizes many words instantly and holistically. She is adept at analyzing the graphic form of an unfamiliar word, such as letters and letter clusters, to retrieve the meaning through its phonological and morphological information (e.g., Shankweiler & Liberman, 1972; Hogaboam & Perfetti, 1978; Ehri, 1998, 2014; Share, 2008). Seidenberg and McClelland (1989) define orthographic knowledge as “an elaborate matrix of correlations among letter patterns, phonemes, syllables, and morphemes” (p. 525), contending that the interletter associative network for a particular word evolves gradually through cumulative experience of decoding and encoding words during reading. The more frequently a particular pattern of letter sequences is experienced, the stronger the associations that hold them together. Familiar words are recognized effortlessly because all of the orthographic elements are well connected and represented as a whole in memory (Seidenberg & McClelland, 1989; Adams, 1990; Ehri, 1998, 2014). As Ehri (2014) puts it, “orthographic knowledge becomes a powerful mnemonic device that bond the spellings, pronunciations, and meanings of specific words in memory” (p. 5).

Word form analysis is also necessary for accessing, storing, and manipulating phonological information (Torgesen & Burgess, 1998). Studies involving English‐speaking children have consistently demonstrated that poor readers are handicapped in a variety of phonological tasks; and that their deficiencies tend to be “longitudinally predictive, and relatively unaffected by non‐phonological factors—such as general intelligence, semantic knowledge, or visual processing” (Share & Stanovich, 1995, p. 9). It is agreed that efficiency in phonological decoding is causally related to achievement in word recognition, vocabulary acquisition, and text comprehension.

The primary function of phonological decoding is to enable the child to identify the meaning of printed words through their spoken sounds (Frost, 1998). The importance of efficient decoding is also understood as the means to enhance the functioning of working memory (Kleiman, 1975; Levy, 1975). It has been shown that phonologically encoded information is more durable in working memory than visually encoded information (Gathecole & Baddeley, 1993). Because complex mental operations rely upon working memory, efficient decoding is essential for their execution and completion.

As the smallest functioning unit in the composition of words, morphemes represent both rule‐based grammatical information and item‐specific lexical information. As the child advances to higher grades, morphological analysis becomes increasingly central to word recognition, vocabulary acquisition, and text comprehension. Morphological awareness, sensitivity to the abstract structure of words, enables the child to analyze a word into its morphological constituents. Such sensitivity emerges early in spoken language development through recurrent experiences of mapping particular grammatical functions, such as plurality and tense, onto specific sounds in speech. Morphological awareness plays a significant role, particularly in later stages of reading development (Nunes & Bryant, 2006; Frost, 2012; Ehri, 2014), wherein the child learns a large number of new words. According to Nagy and Anderson (1984), roughly 60% of the new words are structurally transparent multimorphemic words, such as “fire‐fight‐er” and “un‐lady‐like.” As evident in these examples, the ability to decompose morphologically complex words makes it possible for the child to infer the meaning of an unfamiliar word based on knowledge of the component morphemes, which, in turn, contributes to incidental word learning through reading.