Reader variable
There is little dispute among researchers that the reader plays the central role in an act of reading. While the reader was once believed to be a passive receiver of information, he/she is now considered an active participant in a reading activity. The emphasis on the role of the reader, however, does not mean that other variables have no influence on reading comprehension. Rather, the ultimate product is the result of interaction between the reader and other variables. Based mainly on the research and arguments of Goodman (1968; 1970; 1973) and Smith (1971; 1973a, b), it is now accepted unanimously that the reader contributes more than ‘the visual symbols on the page’ do (Grabe, 1991, p. 377). However, each reader’s contribution will be different from that of others because readers are different in their ‘shared knowledge, language skills, strategies ... and “other personal characteristics” ’ (Alderson, 2000, p. 128).

Although visual input is necessary for reading to take place, it is not enough for successful comprehension. Non-visual information, or what the reader already has, plays a major role (Smith, 1973a). This non-visual information, or what the brain imposes upon the eye (Smith & Holmes, 1973, p. 50), has generally been known as background knowledge. This knowledge, also referred to as content schemata (Singhal, 1998, p. 2), refers to one’s knowledge of the world, the culture, and the language. According to Anderson and Freebody (1981, p. 84) ‘background knowledge is crucial for reading comprehension.’  Johnston (1983, p. 30) distinguishes between qualitative and quantitative background knowledge. While the former refers to inappropriate background knowledge leading to a ‘completely inappropriate model of text meaning’ (p. 31), the latter refers to a reader’s lack of related background knowledge (e.g., a rural student reading about city metros). It has been shown that differences in background knowledge ‘may indeed account for a significant portion of variance in comprehension performances in normal reading situations’ (Jenkins & Pany, 1981, p. 167), and that such differences affect the ease or difficulty with which one understands a text (Singhal, 1998, p. 2).

In discussions on background knowledge, particular attention has been given to schema in organising and activating the reader’s background knowledge. Schema theory, which comes from cognitive psychology, owes much to the work of Bartlett (Rumelhart, 1981, p. 3) and Piaget (Orasanu & Penney, 1986, p. 6). Schemata, the plural form of schema, also called ‘building blocks of cognition’ (Rumelhart, 1981, p. 3), refer to ‘abstract knowledge structure[s] stored in memory’ (Garner, 1987, p. 4). In fact, ‘The role that background information plays in comprehension has been formalised in schema theory’ (Kitao & Kitao, 2000, p. 2). ‘According to schema theories all knowledge is packed into units … [which] are the schemata. Embedded in these packets of knowledge, in addition to the knowledge itself, is information about how this knowledge is to be used’ (Rumelhart, 1981, p. 4; Rumelhart, 1984, p. 2). In schema theory, comprehension of a text involves activation of relevant schemata, which are initiated as a result of ‘bottom-up observation’ (Rumelhart, 1981, p. 18), and proceeds through a constant process of testing the activated schemata, evaluating their suitability, and refining or discarding them (Johnston, 1983, p. 17; Rumelhart, 1984, pp. 3, 6).  The schemata or old information which ‘provide general “ideational scaffolding” for new information’ (Garner, 1987, p. 7) are activated to make new incoming information comprehensible. According to Alderson (2000, p. 17) schemata act as ‘filters’ for new information. Despite the fact that schema theory ‘is not a well-defined framework for the mental representation of knowledge’ (Grabe, 1991, p. 389), ‘it has been an extremely useful notion for describing how prior knowledge is integrated in memory and used in higher-level memory processes’ (p. 390). 

A similar theory to schema, proposed mainly to account for learning, is ‘constructivism’. The theory of constructivism is generally attributed to Jean Piaget, who articulated mechanisms by which knowledge is internalized by learners. As its name implies, constructivism emphasises the building (i.e., constructing) that occurs in people's minds when they learn. Based on this theory, what a learner intakes depends more on what is already stored in that his/her brain than on what is actually read, heard or observed. In other words, every individual constructs a unique mental image by combining information in their heads with the information they receive using their senses.

Proponents of this theory contend that human beings are not passive recipients of information; namely, learners actively take knowledge, connect it to previously assimilated knowledge and make it theirs by constructing their own interpretation (Cheek, 1992). A similar phenomenon may be claimed to take place while a reader is trying to decipher written symbols.

Singhal (1998, p. 2) distinguishes between three types of schema: content schema, which refers to one’s background or world knowledge; formal schema, also known as textual schema, which refers to one’s knowledge about a text’s organisational and rhetorical structure; and language or linguistic schema, which refers to one’s knowledge of lexicon, syntax, and semantics. The implication is that while L2 readers bring their L1 content and formal schema to L2 situations, they need to develop L2 language schema from the beginning. 
   Sometimes the existence of the relevant schema may distort rather than facilitate comprehension (Rumelhart, 1981, p. 21). The reason may be that they rely too much upon what one might know. Lack of relevant schemata has also been reported to hinder comprehension.  Studying 7-year-old pupils’ story comprehension, Rosowsky (2000, p. 46) reported that while all the pupils ‘unanimously “understood” the word “heroine” in a story they were reading … What they did not understand was how and why this class A drug had turned up in a story about heroes.’

A component of background knowledge, called cultural knowledge in this study, is responsible for distortions in reading comprehension as a result of a mismatch in L1 and L2 cultural schemata. Cultural knowledge gains importance when a reader reads a text with cultural elements with which he/she is less familiar. Steffenson, Joag-Dev and Anderson (1979) and Steffenson and Joag-Dev (1984) report that cultural knowledge differences between American and Indian readers led to different interpretations of texts about marriage in these cultures. Comparing Iranians and Americans on their comprehension of culturally-biased folklore stories, Johnson (1981, p. 169) found that ‘the cultural origin of the story had more effect on the comprehension of the ESL students than the level of syntactic and semantic complexity.’ Rosowsky’s (2000, p. 50) study with Asian bilingual students studying in the UK also revealed that reading comprehension was much influenced by ‘cultural bias’.
   Accordingly, a beginning Iranian EFL learner may become puzzled on reading that Johnny goes to school on Fridays and that the school is closed on Saturdays and Sundays. The simple reason for this is that he/she uses his/her L1 cultural knowledge in understanding the text but since there is a mismatch between L1 and L2 situations in this regard, the attempt for comprehension fails. (In Iran, and many other Islamic counties, weekend days are Thursday and Friday, rather than Saturday and Sunday). Likewise the notion that the Iranian calendar week starts with the first weekday (Saturday), whereas in L2 situation (American culture) it starts with a weekend day (Sunday) rather than the first weekday, may cause some comprehension problems. Also an Iranian EFL learner reading texts about traffic regulations may find it hard to understand that drivers should keep left rather than right while driving, as is the normal practice in the UK and some other countries. Similarly, Iranian EFL readers may have problems in understanding why plugs and sockets are three- rather than two-pinned; how it is possible to hear a public phone ringing in the street; how one can have his/her photo taken using a machine with nobody else to take the photo; and so on.  No doubt, some of these cultural differences are related to differences in progress made in technology.

Another element of reader variable is knowledge of language, or linguistic knowledge. According to Alderson (2000, p. 80), linguistic knowledge includes ‘phonological, orthographic, morphological, syntactic, and semantic information’ and ‘discourse-level knowledge, including … text organisation and cohesion, text types and associated conventions, as well as metalinguistic knowledge’. An example of phonological-orthographic language knowledge is the knowledge that not all types of consonant cluster in English can occur at the initial or final word positions (Zakaluk, 2000, p. 8). The observation that there are obvious differences in the phonological and orthographic pattern of Iranian EFL readers’ L1 and L2 may introduce some problems at least for beginning readers. For example, no consonant clusters can occur at the initial word position and none with more than two phonemes anywhere else in a word in Farsi, the official language of Iran.
   Along the same lines, one’s knowledge of morphological and syntactic patterns of the language will facilitate the comprehension and reading speed (Anderson & Freebody, 1981, p. 80). In the case of L2 learning, similar patterns brought from the L1 situation will help at all linguistic levels (i.e., phonological, orthographic, etc.). Semantic knowledge, or the knowledge of words and relationships between them, plays the most important role compared to other linguistic knowledge types. There is no dispute among researchers that vocabulary knowledge is crucial to reading comprehension (Anderson & Freebody, 1981, p. 77; Jenkins & Pany, 1981, p. 176; Johnston, 1983, p. 14; Grabe, 1991, p. 392; Schoonen, Hulstijn & Bossers, 1998, p. 98; Khaldieh, 2001). Coleman (1971) reports that lexical complexity of texts account for 80% of the variance in reading comprehension (in Anderson & Freebody, 1981, p. 80), implying that the higher one’s degree of lexical knowledge, the higher the degree of comprehension. Schoonen et al. (1998, p. 87) bring evidence that in the case of L1, 60% of the variance in reading comprehension is related to vocabulary knowledge, and that ‘FL [Foreign Language] vocabulary is the best predictor of FL reading comprehension’ (p. 89).
   One’s knowledge of discourse and text structure has also been reported to affect reading comprehension. For example, understanding coherence, or superficially non-existent relations between sentences, is a determining factor in reading comprehension (Taylor, 1985, p. 2). A difference in text-organisation in L1 and L2 is a potential source of misunderstanding. An instance of this is the system of writing addresses in Farsi and in English: while Farsi speakers move from more general to more specific in writing addresses and do not begin each heading on a separate line, the reverse is true in English.
   Readers from a different linguistic environment and speaking a non-standard dialect may face problems in reading not only because of sub-culture differences but also because of differences in linguistic knowledge. As Zakaluk (2000, p. 8) points out, in some non-standard dialects there are differences in pronunciation of special items between standard and non-standard codes, such as ‘test’ pronounced as ‘tess’, ‘mend’ pronounced as ‘men’, ‘find’ pronounced as ‘fine’, and ‘cold’ pronounced as ‘coal’. However, whether such differences in pronunciation pose any reading difficulties has not been shown (p. 9). In a nutshell, it seems that even though cultural and background knowledge play a great role in reading comprehension, without linguistic knowledge, no reading may exist at all.

Another reader characteristic that influences reading comprehension is one’s cognitive abilities. Readers with normal cognitive abilities vary slightly in their performances and this is perhaps why some scholars relate reading to thinking. Goodman (1970, p. 108), for example, notes that efficient reading results from ‘an interaction between thought and language.’ Others have emphasised readers’ cognitive strategies and meta-cognitive knowledge in affecting reading comprehension (Baker & Brown, 1984, p. 34; Aslanian, 1985, p. 20; Johnson, 1998, p. 23; Schoonen et al., 1998, p. 75).
   The point is not, however, having a certain type of knowledge; rather the reader’s ability and success in activating and using that knowledge and relating new information to old information seem to be vital for understanding. The reader’s (cognitive) ability to link all sources of information to one another in order to construct the writer-intended meaning has been recognised by Trabasso (1981, p. 56), Lorch and van den Broek (1997, p. 214), and Johnson (1998, p. 22). Lorch and van den Broek (1997, p. 244), for instance, assert that ‘Individual differences in working memory capacity have been demonstrated to be good predictor of variance in both overall reading ability … and specific reading skills.’  Similarly, Rumelhart (1984, p. 19) compares reading comprehension to ‘a detective act’ where the reader has to use his/her cognitive ability to connect all the relevant information in the situation to solve the problem. This implies that one’s degree of comprehension is not only affected by his/her background, cultural, and language knowledge but also by his/her cognitive abilities including intelligence. The resulting comprehension is the product of the interaction among these and other factors elaborated in the following pages.
   An often-neglected aspect of the reader is his/her affective state. The affective state refers, on the one hand, to a reader’s purposes, perspectives, motivation, emotional mood, etc., prior to reading. On the other hand, it refers to the affective responses brought about by the text while reading. A reader may read the same text at different times for different purposes and with different kinds of attention paid to information on the page, and therefore, with different degrees of comprehension. The influence of one’s goals, views, and emotional states on reading comprehension has been documented by Garner (1987, p. 7), Lorch and van den Broek (1997, p. 237), Pressley (1997, p. 248), and Alderson (2000, p. 80). The way affective responses during reading may affect the direction of reading and the extent of the reader’s cognitive involvement have been discussed by Carroll (1970, p. 30), Lorch and van den Broek (1997, p. 233), and Alderson (2000, p. 83).

Text variable
The second important variable which considerably affects reading comprehension is text variable, since without written material, there will be no reading at all. Nuttall (1982, p. 15) believes that text is ‘the core of the reading process’. Text, or written discourse, is the product of the writer’s thought expressed through some visible shapes, whether alphabetic or ideographic, printed or hand-written, something kinesthetic or written in Braille, written on a piece of paper, carved on a stone, or displayed on a computer screen. The factors subsumed under the text heading include graphic, para-linguistic, linguistic, and organisational characteristics of a text, and text types. What follows is a brief look at these text-related elements.
   No doubt the first text characteristic is whether the symbols and the code used are familiar to the reader or not, i.e., whether it uses Latin symbols, Chinese symbols, Arabic symbols, etc., and whether the code or language is familiar.  Although there should be a complete match between the symbols and the language used in the text and those possessed by the reader, this is by no means a guarantee for the reader’s success. Goodman (1973, p. 26) believes that because readers read for meaning and with minimal use of graphic cues, the differences in L1 and L2 in terms of direction of reading (right-to-left in Farsi, top-to-bottom in Japanese, left to right in English) will not affect ‘the basic reading process’ considerably. While Grabe (1991, p. 387) agrees that direction of reading will not influence reading comprehension, Kitao and Kitao (2000, p. 4) contend that Japanese English readers may face problems for this reason. It seems that while such differences may be problematic in early stages of reading, they will become less important as readers get more exposure to and fluency in L2 reading (Grabe, 1991, p. 387).
   Similarly there has also been some attention paid to whether para-linguistic features of a text, i.e., capitalisation, underlining, paragraph headings, and subtitles have any influence on reading comprehension. According to Johnston (1972), these graphic cues signal meaning, but ESL students do not give such cues due attention; instead, they focus on ‘phonologically related graphic aspects of individual words’ (in McLeod & McLaughlin, 1986, p. 115). Lorch, Lorch and Klusewitz (1995, p. 51) report that underlining, or ‘light signalling’, helps the reader’s recall in comparison with no underlining and ‘heavy signalling’ in which important information is underlined along with some non-important information. Their findings also indicate that while capitalisation reduces reading speed, it improves the memory for recalling capitalised information.
   Jenkins and Pany (1981) cite studies by several researchers, in some of which paragraph and title headings helped readers to ‘answer more comprehension questions correctly’ (p. 169), and in others, carried out with children, adolescents and adults separately, ‘no advantage was found for passages with paragraph headings’ (p. 170). Pictures in text have been recognised as having a positive effect on reading comprehension for the reason that they give the reader ‘broader and deeper background knowledge’ (Jenkins & Pany, 1981, p. 171). On the other hand, Maxwell (1994, p. 69) brings evidence from a study done by Grant and Davey (1991) in which paragraph or title headings did not seem to influence ‘overall comprehension’.

More important text characteristics in affecting comprehension include lexical density, syntactic complexity, and semantic abstractness. The influence of new information and lexical density as well as passage length on reading comprehension has been emphasised by Johnston (1983, p. 21) and Alderson (2000, p. 127).  Complexity of structure accompanied by sentence length has also been reported to reduce text comprehensibility (Pearson & Camperell, 1981, p. 33). An issue in structural complexity has been the transformational distance between a sentence’s surface and deep structures, i.e., the shorter the distance, the easier the comprehension. Therefore, kernel sentences (simple, active, and declaratives) are expected to be understood more easily and rapidly than non-kernel transformed sentences such as interrogatives and passives (Pearson & Camperell, 1981, p. 32). Differences in syntactic structures between L1 and L2 have also been reported as a source of reading problems for L2 learners. For example, German EFL readers might face problems because they ‘attend more to function words’ in their language, while English readers ‘attend more to content words’ (Grabe, 1991, p. 388).
   To reduce the degree of lexical density and syntactic complexity, some have suggested simplification, arguing that linguistic simplification increases readers’ comprehension (Yano, Long & Ross, 1994, p. 189). ‘Elaborate modification’ has been used as another alternative to reduce text difficulty, in which texts are elaborated with the hope that ‘redundancy and explicitness compensate for unknown linguistic items.’ Simplification has, however, proved to be a better tool than elaboration (Yano et al., 1994, p. 211). Alderson (2000, p. 82), however, questions the former on the grounds that simplification ‘not only does … disauthenticate the text, it also risks making the texts harder to understand.’ As an alternative, he suggests developing easier tasks or test questions.
   Structural complexity, also called micro-structure variation, is different from macro-structure variation in that the former operates on the intra-sentential level, but the latter deals with organisational variation on the inter-sentential and discourse level.  The changes in a text’s macro-structure, e.g., changing the place of main and subordinating ideas, have been reported to disturb comprehension (Pearson & Camperell, 1981, p. 32). Johnson (1981, p. 169) notes that organisation of ideas in a text affects reading comprehension more than its language complexity. Johnston (1983, p. 25) cites a study by Freebody (1980) in which ‘the order in which subjects read passages affected their comprehension.’ Grabe (1991, p. 338) mentions a study done by Carrell (1984) with readers from Spanish, Asian, and Arabic language backgrounds in which different cultures favoured different organisational structures, which in turn affected readers’ recall and comprehension. Similarly, Maxwell (1994, p. 68) reports a study by Dee-Lucas and Larkin (1990) in which the change of rhetorical organisation affected the degree of reading comprehension. Singhal (1998, p. 4) confirms that ‘differences in text structure can lead to differences in reading.’

Semantic abstractness refers to the notion of whether ideas expressed in the text are easy enough for the reader to understand, an issue also connected to the reader’s background knowledge. Apart from semantic abstractness, the ‘density of arguments in propositions’ also affects reading comprehension (Johnston, 1983, p. 22).
   Text type, topic, genre and writer’s style have also been recognised as factors affecting reading comprehension (Alderson, 2000, pp 83, 127). Welrich (1998) distinguishes five text types: descriptive, narrative, expository, argumentative, and instructive (ibid.). To be added to this list is poetry. Clearly, the kind of reading and the degree of attention each type requires are somehow different from others.  As Johnson (1998, p. 22) points out, while narrative texts are mainly read for enjoyment, expository texts have the goal of giving information, and therefore, ‘narrative texts can never be comprehended; they can only be experienced.’
   Another text-related factor important for comprehension is coherence and cohesion. Text-related context, or co-text, has been shown to facilitate comprehension, recall, and reading speed (Smith, 1973a, p. 75; Smith & Holmes, 1973, p. 60; Zakaluk, 2000, p. 15). Text meaningfulness is usually conveyed through local coherence, or cohesion, and global coherence, simply called coherence (Colley, 1987, p. 129). The existence of logical relation markers or discourse markers, such as conjunctions has been reported to facilitate reading comprehension (Geva, 1992, p. 731). Coherence has been recognised as more crucial than cohesion because, as Nuttall (1982, p. 16) points out, a text which is not coherent will be nonsense, although it may be cohesive.
   The way new and old information is organised in the text has also been shown to affect reading comprehension. For example, Colley (1987, p. 130) observed that if ‘the given information and its antecedent are separated by intervening material,’ cohesion is reduced or lost, making it difficult to comprehend the text. Therefore, foregrounding the given information rather than backgrounding it, i.e., keeping it active in the mind of the reader by keeping it as close as possible to its antecedent, makes the text more cohesive and easier to understand.
   One of the text characteristics is how far it presumes the reader to have the necessary background knowledge. Some texts have been reported to lead to incomprehension or misunderstanding because they do not provide the reader with effective bottom-up clues to activate the related schemata (Rumelhart, 1981, p. 22, & 1984, pp. 7, 19; Kitao & Kitao, 2000, p. 2). Another text-related factor is the nature and the number of inferences the text invites the readers to make (Lorch & van den Broek, 1997, p. 219). The time it takes a reader to make inferences, such as referential and causal inferences, and the ease with which these inferences are made may be potential factors affecting the speed and the degree of reading comprehension. In L2 reading, inferences can be made using L1 or L2 cues or a combination of these (Ringbom, 1992, p. 100).

Context variable
Although researchers have mainly emphasised the role of reader and text and also the interaction between the two (Alderson & Urquhart, 1984, p. xvi, for example), the role of context has largely remained unnoticed. Context refers to something beyond the text itself, and therefore, is not to be mistaken with textual context or co-text. Generally, context variable refers to all reader-, writer-, and text-external factors, such as environmental and situational elements, and the larger socio-economic context. As Walberg, Hare and Pulliam (1981, p. 154) rightly emphasise, comprehension may not be blocked due to the absence of linguistic or background knowledge but ‘due to environmental distractions’. Comprehension may, therefore, be affected by the time of reading, i.e., early in the morning, after a day’s work, etc.; and also by the place of reading, i.e., in the library, in the classroom, in an exam session, in a car, etc. Although such environmental elements may be considered trivial, their potentiality for affecting comprehension, small as it may be, should not be neglected.
   There seems to be little research addressing the above issues. Studying 368 Nigerian primary students, Williams (1981, p. 31) found that environmental elements like school type (whether private or public, for example), degree of exposure to mass media, and subjects’ language environment contributed significantly to their reading performance in ESL. Walczyk, Kelly, Meche and Braud’s (1999) study on the effect of time-limit on reading comprehension reveals that with no time-limit, ‘readers are not probably being optimally challenged’ (p. 158), and with severe time pressure, readers do not think critically (p. 157), and that ‘the best reading comprehension was observed under mild time pressure’ (p. 156) because of the readers’ being probably more mindful (p. 164).
   Another contextual factor is the social setting in which the act of reading takes place. Reading does not take place in a vacuum (Alderson, 2000, p. 25) and the situation in which it occurs may have an impact on how it is comprehended. For instance, the sign ‘No entry’ means ‘Please, do not enter.’ However, the situation in which a reader reads it may change its meaning. If one notices the sign on a private room, it tell ‘outsiders’ not to enter the room, but not to the occupier himself/herself. However, if one notices the same sign at the entrance to a street, it means different things depending on whether one is driving or walking. To use Widdowson’s (1978) terms, while the sign has a single significance, it has different values depending on settings in which it occurs. The kind of comprehension gained through the interaction between the text and the context, which may be called pragmatic comprehension, indicates how differences in situational contexts can result in different understandings, and are, therefore, a potential source of miscomprehension if the reader does not attend to contextual elements. Similarly, the meaning of ‘prayer’ depends on whether it is said in a church or in a mosque. Here, the difference both in situation and in culture contributes to the value of the word.

The influence of pictures, charts, etc. was examined in the discussion under text variable as para-linguistic elements. Goodman (1968, p. 21), however, believes that such cues are external to language and reader, thus forcing them to come under our category of contextual elements. However, they are, in essence, different from our definition of context which covers environmental, situational, and social elements.

Writer variable
Undoubtedly, the essence of a text owes much to its producer, i.e., the writer. Although the writer does not influence the reading act as directly as other variables do, he/she contributes the most to the reading act, indirectly though. The kind of interaction that the reader has with the writer may not be similar to the kind of interaction he/she has with the text, or the kind of communication a listener has with a speaker, as in both these latter cases both of the communication or interaction parties are present. However, the reader and the writer can be supposed to have some abstract form of interaction/communication, because when producing the text, the writer most certainly takes the characteristics of his/her readers into account (Nuttall, 1982, p. 14; Johnston, 1983, p. 14).
   Similarly, while reading a text, the reader intuitively constructs a picture of who the writer had in mind; and the more the reader is familiar with the writer’s style and purposes, the more successful he/she will be in ‘getting’ his/her message, which in turn leads to a successful communication between the two. In reading, the reader cannot consult the writer to clarify ambiguities, which is why readers ultimately get less or more than the intended message, and that the ‘intended meaning is ultimately unknowable’ (Johnston, 1983, p. 11). Taylor (1985) argues that there is an active co-operation between the reader and the writer in that the reader tries to get the meaning out of the text by bringing meaning to the written material.
   Obviously, the writer’s assumptions about the reader will not always come true, and because no two people may have exactly the same background knowledge, ‘there always is a mismatch of some kind’ between the writer’s and the reader’s background and expectations (Nuttall, 1982, p. 7). Nonetheless, the writer should still be as helpful as possible, adhering to what Grice (1975) called ‘co-operative principle’ for oral discourse (Pearson & Camperell, 1981, p. 30; Johnston, 1983, p. 27). Therefore, the degree of reading comprehension depends on the ‘active collaboration between writer and reader’ (Taylor, 1985, p. 5) on the one hand, and the interaction of other factors previously described on the other. 

L1-L2 relationship
All the variables discussed so far influence reading comprehension in any language. When a reader can already read in a language and tries to read in a second language, another element, i.e., the relationship between L1 and L2, is added to the previous factors. The following is a short review of research and arguments about the relationship between L1 and L2 reading.
   Most of our understanding about reading in a second language comes from research in first language (Alderson & Urquhart, 1984, p. xv; Grabe, 1991, p. 378; Chun & Plass, 1997, p. 61), and therefore, the majority of findings from L1 situations have been readily generalised for L2 situations with little consideration given to potential differences. However, such generalisations may not always be true and, as Grabe (1991, p. 389) puts it, L1 research findings ‘cannot always be applied directly to L2 contexts’.
   The relationship between L1 and L2 reading has been studied from two different perspectives. While some researchers have focused on the differences and similarities between reading in the first language and in a second or foreign language, other researchers have studied the role of L1 reading ability in L2 reading comprehension. Studies in the former group have pointed out that although there are many similarities in the process of reading in L1 and in L2 and virtually in any other language (Singhal, 1998, p. 3), they have also discovered that there are certain dimensions to reading which are unique to a second language (Wade-Woolley, 1999, p. 448). Researchers in the latter group have spoken up for a balance between linguistic interdependence and threshold hypotheses (Ulijn & Salager-Meyer, 1998, p. 83; Taillefer & Pugh, 1998, p. 98; Yamashita, 2002, p. 92).
   In terms of similarities between L1 and L2 reading, Goodman (1973, p. 27) argues that reading is ‘much the same for all languages’ implying that a good reader in L1 will also be a good reader in L2. Such a similarity between L1 and L2 not only in terms of reading but also other language skills and mainly spoken skills has been proposed by second-language acquisition researchers (e.g., Krashen, 1981). MacLean (1984) reports another similarity between L1 and L2 reading in terms of the speed of reading. His longitudinal study with one subject who read a variety of texts in L1 and L2 showed that she read all passages at nearly the same speed and that ‘she exhibited minimal flexibility in her reading rate across the passages’ (p. 61).
   Jonz’s (1994, p. 305) study showed that non-native speakers were less sensitive than natives to textual sequence and he concluded that ‘the non-natives depended more on bottom up processing … than on top down processing’. It should be mentioned, however, that the comprehension processing Jonz is talking about is based on reading cloze-format texts rather than normal texts. Singhal (1998, p. 1) argues that reading in either language requires the possession of certain types of schema (content, formal, linguistic); in both cases reading is done for the construction of meaning; and both L1 and L2 readers use similar reading skills and strategies.
   While Singhal’s last assertion seems to have been confirmed by studies done by Block (1986; 1992), who found that L1 and L2 readers used similar reading strategies and comprehension-monitoring processes (Li & Munby, 1996, p. 201), his claim that effective readers in L1 and L2 use both top-down and bottom-up processes (p. 1), does not seem to have been substantiated by Jonz’s study above. In either case, however, reading comprehension involves the interaction of many variables.
   While Alderson and Urquhart (1984, p. xv) suggest that the differences between L1 and L2 reading are less clear, many other researchers have pointed out differences and explained the reasons for these differences. For example, Chun and Plass (1997, p. 63) note that while L1 readers are mainly involved in higher-level processing such as generating inferences, L2 readers ‘pay more attention to lower-level processes’. They divide differences between L1 and L2 reading into three groups: background knowledge differences; language processing differences; and social context differences, i.e., ‘expectations about reading and how texts can be used’ (p. 62).
   One of the differences between L1 and L2 reading is that while L1 readers are already proficient enough in the spoken language, having acquired 5,000 to 7,000 words and a working knowledge of the grammar by the time they start reading (Grabe, 1991, p. 387), L2 readers only start learning these aspects of the target language at the time they start reading. However, many L2 learners already do have the experience of reading in their own language from which they will have built up additional ‘world knowledge’ (Chun & Plass, 1997, p. 62). The case is different for L2 readers who do not know reading in L1 (Alderson, 2000, p. 24). An L2 reader’s L1 can help or disturb his/her reading depending on the similarities or differences in orthographic, lexical, syntactic and discourse systems between the two languages (Ringbom, 1992, pp. 88, 90; Chun & Plass, 1997, p. 62).
   Cziko (1994, p. 151) cites a study done by Hatch, Polin and Part (1974), who found that while L1 readers mainly relied on contentives, L2 readers relied both on contentives and syntactically redundant elements in comprehending cloze passages. His own study showed that only advanced non-native readers as well as L1 readers could use semantic constraints in cloze tests while less advanced L2 students were able to use syntactic constraints only (p. 154). According to Ulijn and Salager-Meyer (1998, p. 81), while L2 reading problems are attributed to L2 language problems, L1 reading problems are attributed to problems within reading itself. According to Li and Munby (1996), translation, which may take the form of mental conversion, is a reading strategy unique to L2.
   To summarise the research findings discussed so far, several reasons have been put forward by different researchers for the differences between L1 and L2 reading (Grabe, 1991, p. 387; Singhal, 1998, p. 3; Mitchell & Myles, 1998, p. 61). The above researchers have identified the following reasons for the differences between L1 and L2 reading:

• L2 learners are cognitively mature;
• L2 learners already know at least one other language;
• L2 learners have different kinds of motivation for reading in L2 than L1 learners have for L1 (L2 learners are motivated both instrumentally and integratively) (Mitchell & Myles, 1998);
• L1 readers already have a big vocabulary repertoire and know grammar when beginning to read, while L2 readers begin from scratch;
• Older L2 readers have a more well-developed conceptual sense of the world;
• L2 readers make elaborate logical inferences from the text;
• L2 readers make more use of meta-cognitive strategies (Grabe, 1991);
• L2 learners draw upon strategies of first-language learning, knowledge of likely language systems, and knowledge of how language operates socially;     
• L2 readers operate in a different linguistic context (use different vocabulary and grammar);
• L2 readers may not have the relevant cultural/background knowledge (Singhal, 1998).

The above differences between L1 and L2 makes the task of the L2 reader different from that of the L1 reader; and comprehension failure may occur for an L2 reader if he/she is either not fully competent in the language and/or there is a mismatch between the cultural knowledge he/she processes and that required to function successfully in L2 context (Wallwork: 1978, p. 83; Montgomery, 1986, pp. 174-75).
   A controversial issue regarding the relationship between L1 and L2 reading is how far reading ability in L2 is affected by L1 reading ability. The linguistic inter-dependence hypothesis or reading universals, assumes that the process of reading in all languages is the same (Goodman, 1973, p. 27) and once the related ability is acquired in one language, it is easily transferable to all other subsequent languages; that ‘we really only learn to read once’ (Taylor, 1985, p. 4); and that L2 readers do not need to gain proficiency in L2 reading ability if they have acquired such an ability in L1 (Chun & Plass, 1997, p. 63).
   The alternative hypothesis called threshold hypothesis, ‘short circuit hypothesis’ (Chun & Plass, 1997, p. 63) and ‘lexical threshold’ (Ulijn & Salager-Meyer, 1998, p. 81), was introduced by Alderson (1984; 2000). In this hypothesis, before L1 reading ability can be successfully transferred to L2 situation, a certain language threshold level needs to be attained by L2 reader (Alderson, 2000, p. 23). In this view, both L1 reading ability and L2 linguistic knowledge contribute to the reading ability in L2 (Alderson, 2000, p. 122). Both hypotheses presuppose that reading ability in any language is different from linguistic ability or language-specific knowledge, and therefore, a good reader in L1 would be a good reader in L2 (Schoonen et al., 1998, p. 73). 
   Empirical studies, however, have not found a high degree of relationship between reading comprehension in L1 and L2, indicating that not all L1 reading ability is transferable to L2 situations. For example, Groebel (1980) conducted a study with 454 university students (L1 was Hebrew and L2 was English for most subject in the study). The correlation between reading comprehension in L1 and L2 ‘which was statistically significant’ was 0.315 (p. 57). She concludes that when reading involves higher levels of comprehension, there is a correlation between L1 and L2 reading comprehension (p. 59). Studying Dutch EFL learners, Schoonen et al. (1998, p. 96) found that L1 and L2 reading comprehension ‘correlate substantially, but not highly; about 38% … of the variance is common.’ They also report other studies in which a low degree of correlation has been found between reading ability in L1 and L2 (p. 73). Schoonen et al. (1998, p. 74) rightly emphasise that even if L1 reading ability transfers to L2 reading, not all that ability is transferable because the reading ability is composed of both L1 linguistic knowledge and language-independent general reading skills, of which only the latter are transferable.
   Taillefer and Pugh (1998) studied the L1-L2 reading relationship with 39 subjects for whom French was L1 and English was L2. Their findings suggested that strong L1 readers who were weak in L2 linguistic knowledge were poor L2 readers, but they were better readers than those who were poor L1 readers and also weak in L2. They also found that those readers who were strong in L1 reading and in L2 linguistic knowledge were equally efficient readers in either L1 or L2 (p. 96). They conclude that good L1 readers ‘who are good “linguists” in L2’ have no difficulty in transferring their L1 reading strategies to L2. They also assert, based on their findings, that ‘reading is not “just” reading’ for those L2 readers who are weak in either L1 reading or L2 linguistic knowledge or both (p. 105).
   Reviewing a few studies on the relationship between L1 and L2 reading, Yamashita (2002, p. 82) concludes that ‘the contribution of L1 reading ability increases when learners’ L2 proficiency level becomes higher’ and that ‘L1 reading ability is more likely to be transferred when the [L2] reading task is less demanding.’ She cites a study carried out by Carrell (1991) with English Spanish readers and Spanish English readers, in which it was found that Spanish language proficiency was more important than English reading ability for the former group, but the Spanish reading ability was a stronger predictor for the latter group. Such a finding seems to support Taillefer and Pugh’s (1998, p. 98) claims that the relative importance of each predictor variable varies depending on the ‘directionality of learning’ (i.e., L1 English to L2 Spanish vs. L1 Spanish to L2 English), and also on ‘the nature and cognitive demands of the task itself’.
   Yamashita’s (2002, p. 81) own study with 241 Japanese EFL students, in which the compensatory relationship between L1 reading ability and L2 linguistic knowledge in L2 reading comprehension was studied, showed that there was a mutual compensation between the two independent variables in arriving at the highest comprehension in L2 reading. Yamashita (2002, p. 92) concludes that there is a complex interaction between L2 linguistic knowledge and L1 reading proficiency or cognitive ability. The same idea is supported by Ulijn and Salager-Meyer (1998, pp. 82-83) who state that researchers now call for a ‘balanced blend’ of the threshold hypothesis and linguistic interdependence hypothesis. They assert that ‘empirical research has largely resolved the debate about whether L2 reading is a linguistic or a cognitive problem. It is neither one, but both of these’ (p. 83).
   In brief, the general consensus seems to be that while L1 reading ability affects L2 reading comprehension, the role of L2 linguistic knowledge seems to be much more significant at least for low-proficient L2 readers (Chun & Plass, 1997, p. 62; Taillefer & Pugh, 1998, p. 98; Ulijn & Salager-Meyer, 1998, p. 82; Alderson, 2000, p. 23; Yamashita, 2002, p. 92). A somewhat different finding has been reported for the relationship between L1 and L2 writing by Schoonen, van Gelderan, de Glopper, Hulstijn, Simis, Snellings and Stevenson (2003, p. 166), in which L2 writing proficiency correlated highly with L1 writing proficiency, which was more than its correlation with L2 linguistic knowledge.

Conclusion

To reiterate, four broad variables were discussed as factors affecting the process and the product of the act of reading: the reader, the text, the context, and the writer, each with some other inter-related elements. It was also made clear that none of these variables or their elements act separately; rather, the resulting outcome, be it comprehension or miscomprehension, is the product of the interaction among all the variables on the one hand, and all the factors within each variable on the other hand. To arrive at comprehension, therefore, all variables and elements inside them have a share with different degrees of contribution. In the case of L2 learning, in addition to the above variables, another variable enters the scene, i.e., the relationship between L1 and L2 on the one hand (similarities or differences), and the degree of L2 readers’ literacy in L1 on the other. The success of the communicative activity in question here owes much to a proper interplay of all the factors discussed above. The implications for both reading/literacy teachers and researchers are far-reaching: they need to regard their students’ or subjects’ reading process not as a simple activity but a complex multi-facetted process, and when analysing their comprehension failure or success, teachers and researchers should not neglect the potential interactive role of so many factors mentioned above.

References 

See PDF or SWF File