Background knowledge and second language reading comprehension theories

¶ … Knowledge and Learning and Teaching a Second Language:

Researchers have divided the skills necessary for the acquisition of second language comprehension, particularly in the reading area, into two general theories: bottom-up, text-based, psycholinguistic approaches or top-down, socially-oriented conceptual approaches. In each case, lack of second language comprehension is attributed to misunderstanding of some key variable of the approach. For example, bottom-up studies tend to trace miscomprehension to misunderstanding of grammar (syntax), vocabulary (semantics), or other textual aspects. Accordingly, comprehension from the bottom-up is a data-driven process (Carrell and Eisterhold, 1983).

In contrast, top-down studies primarily attribute miscomprehension to the lack of specific background knowledge or cultural familiarity that is necessary to understand the text. Top-down understanding is seen as a process that is driven by concepts (Carrell and Eisterhold, 1983). Goodman (1967) is credited with first recognizing this additional aspect to reading comprehension, although he did not use the term "top-down" (Adamson, 1993, p. 45). Another early researcher in this area, Steffenson, Joag-dev, and Anderson (1979), focused on the cultural barriers to reading comprehension. Later work solidified this type of approach into the Schema Theory Model, where understanding involves an interaction between background knowledge of the reader, described as the reader's "schema," and the text. This model has been applied to second language reading comprehension by Carrell (1983a) and Johnson (1982), among many others.

More recently, it has become obvious that the division between these two approaches is somewhat artificial and true comprehension involves a combination of both types of these skills. Thus, more current research, such as the work of Bernhardt (2001), attempts to provide models that take into account the contribution of both approaches. However, investigating of true role of background knowledge, specifically determining the correct amount of emphasis that should be placed on these types of variables, remains a central area of study in the acquisition of comprehension within a second language.

What is Background Knowledge?

Investigators in this area use the terms "background knowledge" or "prior knowledge" to mean a variety of contributors to the comprehension of a reader's second language (L2). Specifically, background knowledge relates to the reader's knowledge of their first language (L1). There are two levels of interest in this particular background knowledge. First, there is the effect of merely knowing L1 and this can alter the acquisition process of L2. This type of background knowledge brings up issues of transfer, avoidance, language loss, and rate of learning, which can collectively be termed cross-linguistic influence (Gass and Selinker, 2001).

The second interest in background knowledge in this area involves the proficiency of the student in L1 and how this alters the acquisition of L2. To study this variable, testing of the comprehension of text in L1 are done followed by attempts to find parallels between this proficiency and the ability to understand text in L2. The work of Lee and Schallert (1997) is an example of this type of study, and will be discussed more extensively below.

Background knowledge can also mean general knowledge of the subject matter of the text, and has been broken down into three characteristic components: familiarity, context, and transparency (Carrell, 1983b). Studies have found significant relationships between these providing or not providing these aspects of background knowledge and the overall percentage recall of passages in L2. One study relatively recent study that used this classification of background knowledge is Roller and Matambo (1992).

A third type of background knowledge that is discussed in the literature is one of cultural familiarity. Although this may be a subset of Carrell's "familiarity" component discussed above, the studies looking that this aspect of background knowledge focus upon the effect of the experience of living in the culture expressed by the writer of the passage being tested. A well-known example of this definition of "background knowledge" is the work by Johnson (1982). There, students who had or had not personally participated in the celebration of Halloween read a text related to this holiday. The effect of the actual cultural experience on reading comprehension was tested.

Adamson (1993) reports a fourth possible component of background knowledge that appears to be a subset of the cultural aspects -- scripts for school. In general, when students understood how they should act in the classroom this was accompanied by an increase in success. Conflicts between scripts for school learned in other cultures and those in play in the United States were numerous. These included misinterpretation of group study as being able to do whatever the student wanted in the classroom, misunderstanding that girls need not be overly respectful to boys in the U.S., and misinterpretation of what is appropriate subject matter for study where highly moralistic and patriotic information was expected (p. 50-51). So these school scripts form a fundamental component of the cultural aspects of background knowledge that are important to success for English as a Second Language (ESL) students. Whether examining the knowledge related to L1 to L2 transfer, L1 proficiency, knowledge about the text subject, a cultural experience, or school scripts, each researcher attempts to find a connection between second language comprehension and having particular background information.

Background Knowledge and its Effect on L2 Acquisition

Cross-Linguistic Influence. In its early stages, the study of language transfer from L1 to L2 dismissed the effect of the native language as background knowledge on the acquisition of L2 (Gass and Selinker, 2001). This approach has been termed a behavioristic, in that the acquisition of various English morphemes (meaningful linguistic units) seemed to follow a set behavior pattern no matter what the L1 of the student (Gass and Selinker, 2001, p. 112). However, more recent research has indicated that if the right research questions are asked, significant differences can be detected, and L1 knowledge has been found to possibly have both positive effects and negative effects on acquisition of L2 (Noor, 1994). This new approach to L1 to L2 transfer has been termed the "mentalist" approach.

A good example of a negative effect of the L1 background knowledge is the phenomena of avoidance, that is the selective use of particular language structures in L2. In 1993, Laufer and Eliasson attempted to determine whether the amount of difference between L1 and L2 or the complexity of the new L2 structures was responsible. Although the exact source of avoidance remains in controversy, the only variable clearly supported by the research is the amount of difference between L1 and L2, with greater difference correlating with greater avoidance (Laufer and Eliasson, 1993).

Overall, as noted by Gass and Selinker, the mentalist approach can be summarized as including three interacting factors to determine how language transfer will occur: "a learner's psychotypology, perception of L1 to L2 distance, and actual knowledge of L2"(2001, 131). It is important to note that each of these factors has a significant background knowledge component. The exact relationship between background knowledge of L1 and transfer to L2 is still be elucidated, but there is now strong support for an existing relationship. Researchers have urged the need for more investigation in this area, particular as it relates to initial word recognition (Durgunoglu and Hancin-Bhatt, 1992).

L1 Proficiency. A second way to interpret the knowledge of L1 as background knowledge is by measuring the level of proficiency in the native language and see if this has an effect on L2 comprehension. In general, language proficiency can be divided into five components: communication, conceptualization, critical thinking, context, and culture (Diaz-Rico and Weed, 2002). Lee and Schallert examined the effect of L1 proficiency in their 1997 study, focusing on measure of the first three aspects. Specifically, this study questioned whether there was a stronger correlation between L2 proficiency and L2 reading comprehension or between L1 reading ability and L2 reading comprehension (p. 730). Although the study concluded that the L2 proficiency was more closely related to L2 reading ability than the L1 reading ability (background knowledge), their work did support the existence of a relationship between L1 and L2 reading abilities as well (p. 732). This study was also significant in that it suggested that attaining a threshold level of language proficiency in L2 was necessary for the reader to take advantage of the "reading skills, strategies, and prior knowledge" that had been developed in L1 (p. 733).

Subject Matter of the Text. A third type of study addressing the effect of background knowledge on L2 reading comprehension focuses upon supplying students with varying amounts of background information about the text before reading. The Roller and Matambo (1992) study is an example of this approach. This study utilized the breakdown of background knowledge into familiarity, context, and transparency first set forth by Carrell (1983b) to determine whether the interactions between these components could be replicated.

According to the Carrell (1983b) study, familiarity means the amount of prior exposure that the reader has had to the content area of the text and involves a subjective classification by the researchers of the text as either familiar or unfamiliar (p. 184). Context is most closely related to the presentation of the materials and in many experimental designs involves either the use or nonuse of a picture to accompany the text (p. 185). Finally, transparency involves the use of lexical words, where lexical means that the words used reveal the content area of the text (p. 186). The Roller and Matambo study looked for significant relationships between better recall the provision of these particular components of background information (1997).

Roller and Matambo's conclusions were slightly unusual in that the students did better in English, their second language, than their native Shona and they recalled a passage classified as unfamiliar more consistently than one classified as familiar. The authors hypothesized that these effects may be related to the nature of the passages used and the cultural traditions of a British education with the students, rather than a real difference between contributions of the background knowledge (p. 137).

Chen and Graves (1995) also examined the effects of providing background knowledge to readers before they begin the test. Specifically, they provided a preview of the short story to be read (really a type of story-specific background knowledge enhancement) and/or what the researcher's termed "background knowledge" (Chen and Graves, 1995, p. 665). In this study, previewing included "text information important to the story, introduced the characters, described the plot... And gave direction for how to read the story" (pp. 668-69). In contrast, background knowledge included "historical background at the time of the story... necessary background knowledge about the story, and culture-specific information that is needed to fully understand it" (p. 669). Thus, this study contrasted the effects of providing highly story specific information with providing more general background.

Through various measures of comprehension Chen and Graves found that previewing and a combination of previewing and background knowledge all increased the students apparent understanding of the story (p. 678). However, background knowledge alone, as defined in this study, did not have a significant effect on the student's test scores (p. 679). Additionally, the increase in test scores seen with the addition of the background knowledge did not appear to justify the additional preparation time needed to include that information (p. 679). Thus, the researchers concluded that the most time-efficient method was the previewing only procedure (p. 680).

Cultural Considerations. A final aspect of that has been examined by a number of researchers is the effect of the background knowledge embodied in the reader or listener's culture. Rather than merely provide written enhancement of a cultural nature, Johnson (1982) ensured that the study participants actually lived through the cultural experience of celebrating Halloween before testing them on passages dealing with this subject matter. She concluded that prior experience in the American culture improved comprehension for ESL students as measured by improved recall (p. 508). The effect of this background knowledge experience appeared to be greater than other supply of information, including raw vocabulary knowledge (p. 513).

A second study examining the effect of the cultural of background knowledge is Malik (1990). This study examined the reading behavior of students using culturally familiar and unfamiliar texts. Using a technique that determined comprehension by the ability to rank the importance of idea units, Malik found that students did much better with familiar texts than the unfamiliar. He attributed this difference to the lack or presence of "well-developed" schemata for the different subject areas (p. 218). The study also concluded that the proficient-ESL students utilized both syntactic and semantic information (learning strategies) more efficiently when the text was culturally familiar (p. 220).

Conclusion

It is now well established that both top-down and bottom-up approaches are necessary for proficiency in a second language. A central component of these approaches is the input of background knowledge. Background knowledge is believed to provide a schema, or source of reference, necessary for attributing a textually consistent meaning to the passage read or spoken by the student. Background knowledge can be defined in a variety of ways, including L1 knowledge or proficiency, information related to the subject matter of the text, information related to the cultural background of the text creator, or a combination of these factors. Despite some minor inconsistencies likely due to text selection issues, the research does support an important relationship between L2 language acquisition and the array of background knowledge that can be supplied by the reader or listener.