You are on page 1of 110

Masaryk University Faculty of Arts

Department of English and American Studies

Teaching English Language and Literature for Secondary Schools

Hana Tich

Cat in the Rain: The suitability of an authentic literary work for the intermediate English classroom

Masters Diploma Thesis

Supervisor: James Edward Thomas, M.A.


I declare that I have worked on this thesis independently, using only the primary and secondary sources listed in the bibliography.

Authors signature

Acknowledgement I would like to thank my supervisor, James Thomas, for his guidance and support during my research.


Invented texts

1 4


The benefits of studying authentic texts Authenticity and content Authenticity and ideology Authenticity and motivation

12 12 16 17


Challenges of literary text Benefits of using literary texts

19 22

Top-down perspective Meaning and schemata Text structure and organization Density of information Text Cohesion Bottom-up perspective Sentence length Syntactic complexity Grammar Vocabulary

25 31
31 32 41 42 43 49 50 50 55 68


93 95 104 105


Cat in the Rain is a very short story by Ernest Hemingway (an American author, journalist and the 1954 Nobel Prize winner in literature), which was first published in 1925 as a part of the short story collection In Our Time. Hemingway became famous within his own life time (18991961), particularly being known for his simple style of writing and careful structuring; thus like most of his novels, his short stories are very easy to read. Cat in the Rain is an apparently simple story about an American couple spending a holiday in Italy, however, as Taylor (1981) puts it, behind the very realistic surface there is a wealth of symbolism and possible meanings for the readers to supply for themselves (p. 62). In the pages that follow it will be argued that this simplicity of style generating multiple interpretations in the mind of the reader is what makes Cat in the Rain particularly suitable for the EFL classroom. The thesis focuses attention on the intermediate level of proficiency of English learners; it attempts to defend the view that Cat in the Rain is a text with lexical and structural difficulty that will challenge intermediate students without overwhelming them, and that it is an effective vehicle for the achievement of certain language and content goals at this level of proficiency. In order to support the hypothesis that Cat in the Rain is suitable for an intermediate student of English, the text is discussed from different points of view. The thesis touches upon linguistic, as well as methodological issues, and the overall approach applied is a whole-to-part orientation; it begins with the text as a meaningful whole and then tries to understand the various features that enable the text to function. Chapter One begins by laying out the theoretical background concerning texts and their role in the classroom. Chapter Two deals with the complex issue of authenticity of

classroom texts. Chapter Three describes the benefits and challenges of literary tests in ELT. The aim of Chapter Four is to outline some of the processes that take place during reading, as well as the ways of getting meaning from texts. Chapter Five analyses the readability of Cat in the Rain both from the bottom-up and top-down perspectives and looks at possible ways of exploiting the text in the intermediate English classroom. Before any discussion can begin, it is necessary to clarify what is meant by the intermediate level of proficiency. Collins Cobuild English Language Dictionary (Sinclair, 1987) writes that intermediate is used to refer to students and their level of work in a subject when they are not beginners, but are not yet advanced (p. 762). This is a rather broad definition, but according to Brown (2007), the intermediate level of proficiency of L2 learners is indeed richly diverse. He argues that at this stage some automatic processing takes place; phrases, structures, and conversational rules have long been practised and are increasing in number, enabling the mental processes to automatize (p.124). Brown believes that students benefit from small doses of short, simple explanations of points in English; grammar topics such as progressive verb tenses and clauses typify intermediate level teaching. Increasing complexity in terms of length, grammar, and discourse characterize reading materials; students read paragraphs and short, simple stories (p. 127). To put this concept in the context of the Czech education system, secondary students are supposed to graduate at so called threshold or intermediate level of proficiency (the B1 level). The Common European Framework of Reference for Languages (CEFR) describes what knowledge and skills Czech learners have to develop in order to communicate and interact effectively. The CEFR includes a set of Common Reference Levels dividing learners into three broad divisions (A, B, C), which can be further divided into six levels. B1 stands for an independent user of the language who

can understand the main points of clear standard input on familiar matters related to work, school, leisure, etc.

can deal with most situations likely to arise whilst travelling in an area where the language is spoken.

can produce simple connected texts on topics that are familiar or of personal interest.

can describe experiences, events, dreams, hopes and ambitions, and briefly give reasons and explanations for opinions and plans (Council of Europe, 2011, p. 24).

Chapter One: Texts in the classroom

Thornbury (2005) argues that language learning should both begin and end in texts (p. 162). But what is meant by a text? In laymans terms, it can be anything printed on a page a list of isolated words, a couple of sentences, or an excerpt from a book. However, Thornbury defines a text as a self-contained, well-formed, cohesive and coherent language event which has a clear communicative purpose, which is appropriate to the context of use, and is a recognizable text type (p. 19). Cat in the Rain (hereafter CITR) meets all the above mentioned criteria: It is a complete and independent unit with its beginning, middle and end. It is constructed in accordance with the syntactic rules of English. The parts of the text stick together and make sense. The text was written to communicate the authors message to the reader. The language was selected appropriately for the situation it describes. It can be recognized as a combination of the descriptive and narrative text types.

Thornbury (2005) identifies three reasons for including texts in textbooks: a linguistic purpose (providing contextualized examples of grammar structures, vocabulary, pragmatic functions, or as a model for text production), a skills development purpose (material for development of reading and listening skills), and a third purpose, which he calls text-as-stimulus the text is used to introduce into the classroom content to which learners can respond (p. 113). However, from my own experience in EFL teaching, full texts are generally neglected in the classroom (even though one of the expected outcomes of Czech secondary education is that students will be able to produce cohesive and coherent texts 4

in English). The following pages will look at the most common type of texts used in the classroom invented texts which will be later juxtaposed with authentic texts.

Invented texts

According to the above mentioned definition (Thornbury, 2005), a text is a language event which has a clear communicative purpose. Based on my observations, this communicative potential of texts was rarely exploited in the activities of the traditional language classroom. Students frequently read a text for the sole purpose of answering questions about it answers that the teacher already knew. Such activities communicated the students mastering of the language of the text, not its content. Nowadays it is generally agreed that concentrating on the language as a system to be learned is too narrow (Lewis & Hill, 1992, p. 23). Furthermore, especially since the advent of corpus linguistics, artificial classroom texts have been subject to extensive criticism. It is felt that invented texts provide poor models of real language use, and they are concerned with forms of the language rather than with more communicative features of a text, e.g. vocabulary and discourse features. Such texts are not used to inform the reader about the world; they are used to present language features (Thornbury, 2005, p.104). Widdowson (1978) argues that when those texts appear in structurally graded courses, they seem primarily to be used to consolidate a knowledge of structure and vocabulary that has already been introduced and to extend this knowledge by incorporating into the passages examples of elements of usage that come next in the course. In this case, the passage is intended to present selected parts of the language system, and in consequence it frequently exhibits an abnormally high occurrence of 5

particular structures. The effectiveness of passages of this kind is achieved at the expense of a normal realization of the system as use. According to Widdowson, Even when there is an attempt to introduce features to lend a verisimilitude of normality these features do not merge into the passage in a natural way but only serve to accentuate its abnormality (p. 78). Moreover, Widdowson (1978) argues that the process of simplifying vocabulary and syntax of classroom texts might actually result in a distortion of the message (p. 88). A criticism is often made of simplified texts that the simplification of vocabulary results in more difficult grammar. Another criticism is that strict control of grammar results in unnatural, awkward text (Nation, 2001, p. 171). The following example from Project 3 Students Book illustrates the situation; in a text about London, the author writes that a long time ago, in the Tower of London, many people lost their heads (Hutchinson, 2000, p. 46). The expression lost their heads is apparently an attempt to avoid the passive voice and the word executed. However, learners sometimes interpret the expression lose someones head as lose control, especially if they have never heard of the Tower of London before. Nonetheless, all the explanations above tend to overlook the fact that there are some good reasons for classroom texts to be like that. The target language is displayed in comprehensible and learnable contexts. The fact that patterns in such texts are prominently displayed increases the chances that they will be noticed. Repetition and simplification are signs of an appropriate text designed for teaching purposes because this makes them easier to understand. As Lewis and Hill (1992) write, some radical experiments have been done which suggest that it is possible to teach a language while requiring very little of the productive skills from the students during the early stages of teaching. Carefully

selected listening and reading passages are chosen so that the majority of the content will be comprehensible to the students and thus help acquisition (p. 31). Perhaps the most influential hypothesis supporting the use of simplified texts is Krashens (1981, 1985) theory of comprehensible input. This theory states that learners develop language along a natural order and by coming to understand input that is slightly beyond their current language ability level (the so-called i +1 system). This issue will be mentioned again in relation to interlanguage. However, there are arguments which go against the comprehensible input hypothesis. Stoddards (1929) and Warings (1997) experiments comparing receptive and productive learning suggest that if productive use is needed, there must be productive learning (cited in Nation, 2001, p. 32). Similarly, Swain (1985, as cited in Nation, 2001, p. 32) writes that it is still not clear if readiness for productive use can be reached by receptive over-learning, for example large quantities of reading and listening, or whether there must be pushed output with learners being made to speak or write. Although this thesis focuses more on comprehensible input, students comprehensible output will also be included in the discussion.

Chapter Two: Authenticity of classroom texts

In the pages that follow, it will be argued that CITR is suitable for the intermediate L2 classroom because it is an authentic piece of text, although there is no intention to imply that authenticity is the only prerequisite for its suitability. As authenticity is a somewhat problematic concept, the complexities surrounding this issue will be considered first. First of all, Henry Widdowson (1990) is probably the best known critic of the concept of authenticity. He argues that authenticity of language in the classroom is, to some extent, an illusion; authenticity lies in the readers response to the text and not in the text itself (p. 44). Similarly, Van Lier (1996) argues that authenticity is the result of acts of authentication, by students and their teacher, of the learning process and the language used in it (p. 128). The dichotomy of approaches advocating either authentic or simplified materials in ELT has been a much disputed subject within the field of applied linguistics. The problem is that despite the importance accorded to authenticity, there has been a marked absence of research. For example, Crossley, Louwerse, McCarthy and McNamara (2007) argue: Although advocates of both approaches cite the underlying presentation of linguistic features, syntax, and discourse structures as important elements that support their preferences, no conclusive or comprehensive study has empirically examined how these features differ between the authentic and simplified text types. (p. 15) Dissatisfied with the insufficient empirical research, Crossley et al. (2007) examined and compared cohesive devices in simplified and authentic texts. Using the computational tool Coh-Metrix (McNamara, Louwerse, Cai, & Graesser, 2005), their 8

study investigated the differences in linguistic structures between sampled simplified and authentic reading texts. Their research was meant to enable L2 reading researchers, material developers, and publishers to judge more accurately the comparative linguistic value of both text types by concentrating on the differences and similarities between them in cohesion and other language features at the lexical, intersentential, and subsentential levels (p. 16). The findings of Crossley et al. include, for example, that simplified texts provide ESL learners with more co-referential cohesion and more common connectives, and rely more on frequent and familiar words than authentic texts. Simplified texts, on the other hand, demonstrate less diversity in their parts of speech tags, display less causality, depend less on complex logical operators, and demonstrate more syntactic complexity than authentic texts (p. 27). But cohesion, for example, does not occur in normal circumstances unless it serves as a contributing factor to coherence. In other words, one does not employ cohesive markers for its own sake but as part of the business of achieving coherent discourse (Widdowson 1978, p. 78). So if an authentic text does not possess explicit linking devices such as however, moreover, nevertheless, etc., it achieves cohesion in other ways (through lexical or grammatical cohesion, for instance). Later chapters will investigate cohesion in CITR and show that although the text does not provide the reader with an access of explicit devices that link the sentences, it is tightly cohesive and thus coherent for an intermediate learner of English. It will be argued that observing cohesion in authentic texts can help intermediate L2 learners improve their own writing. Authentic texts are written with a particular group of readers in mind readers who are fluent users of the language and who approach the text with certain cultural assumptions and some background knowledge. Thus one of the most convincing arguments against authentic materials is that they can be a burden, both for the student

and the teacher. The students may find them difficult to understand as they contain language above their level of proficiency (Richards, 2006, p. 21), and the teacher may find it difficult to decide which language features should be selected for teaching purposes. Lewis and Hill (1992) warn: Nothing is more depressing as struggling word-by-word at snails pace through a piece of material so that you can do something with it or talk about it only to find that understanding the material has taken so long that the interesting follow-up activity lasts only a moment or two or disappears altogether. (p. 107) Furthermore, the matter is complicated by the fact that a written authentic text is not an isolated language event. Broughton, Brumfit, Flavel, Hill and Pincas (1978) contend that there are three kinds of relationship which concern written texts which must be taken into account when reading: 1. the relationship between the author and his text 2. the relationship between the reader and the text 3. the relationship between the text and the culture (p. 96) To elaborate on the second point, several researchers have discussed the relationship between the reader and the text. Hoey (1991), for example, touches upon the issues of reader creativity and discusses the relationship between the authors construct and the readers reconstruction of meaning from that construct (p. 152). Anderson (1999) sees reading as an active, fluent process which involves the reader and the reading material in building meaning. Meaning, he argues, does not reside on the printed page, nor is it only in the head of the reader. A synergy occurs in reading which


combines the words on the printed page with the readers background knowledge and experiences (p. 1). This has an important message for language teaching. Huangs paper (2009), for example, explains the constructivist approach to teaching reading and shows how a dialogic approach to reading empowers readers to position themselves as participants in making meaning together with the text and its authors, rather than remaining as mute outsiders to the reading process (p. 174). According to Huang (2009), Students need to understand that all readers construct meaning from texts differently, depending on their purpose for reading, their background and even their state of mind. There is usually no single, unequivocal meaning in a text. Thus reading entails constructing meaning from text through intermental and intramental dialogue. (p. 175) Huang writes that constructivists see reading, like learning, as social practice. The author states that the social constructivist approach to reading offers tools and principles for EFL teachers which can help them to draw their students into energetic participation in text events, entering into active dialogue with texts (and their authors), not as outsiders, but as active participants (p. 176). So far the discussion has indicated that using authentic texts in L2 classroom may be challenging, both for teachers and students. The following section will look at authentic materials from a more optimistic view.


The benefits of studying authentic texts

Authenticity and content

Despite the complexities surrounding the concept of authenticity, there are powerful arguments in favour of authentic materials in ELT. Lightbown and Spada (2006) argue that the path through language acquisition is rarely smooth and even. Learners have bursts of progress and then seem to reach a plateau for a while before something stimulates further progress (p. 80). Larry Selinker (1972, as cited in Lightbown and Spada, 2006, p. 80) gave the name interlanguage to learners developing second language knowledge. He described interlanguage as a separate linguistic system based on the observable output which results from a learners attempted production of a target language (Selinker, 1972, p. 214). Interlanguage differs systematically from both the native language and the target language. It has been argued at several points in this thesis that carefully selected reading material is an important source of language input. Putting this more precisely, it is carefully selected authentic material and appropriate language instruction (including corrective feedback) which bring interlanguage closer to the target language. Furthermore, Richards (2006) maintains that authentic materials provide cultural information about the target language, relate more closely to learners needs and support a more creative approach to teaching (p. 20). The second argument is somewhat controversial, as in order to state that authentic materials relate closely to learners needs, one would have to say what the needs are. Nonetheless, the argument implies that the content of the text is at least as important as the fact that it is authentic.


Advocates of content-based instruction argue that an optimal way to create classroom processes that are believed to facilitate language learning is by using content as the driving force of classroom activities (Richards 2006, p. 27). With authentic materials, content is typically selected before other decisions are made. In traditional approaches to language teaching, however, grammar, texts, skills and functions are the starting point in planning the lesson or the syllabus, and after these decisions have been made, content is selected. Nunan (1988) describes various types of syllabuses. He argues that the most common syllabus is one in which the input is selected and graded according to grammatical complexity. The advocates of grammatical syllabuses assume that language consists of a finite set of rules which are combined in various ways to make meaning. The most rigid syllabuses introduce one item at a time and move on after the item has been mastered (p. 28). To illustrate this situation, three randomly selected coursebooks written for intermediate students have been compared: ProFile (Naunton, 2005), New Hotline (Hutchinson, 1999) and Headway (J. Soars & L. Soars, 1986). As Table 1 shows, grammatical structures presented in all three coursebooks appear in a similar order (e.g., present tenses are followed by past tenses, future tenses are presented after present and past tenses, conditionals, reported speech and the passive appear later in the coursebook). However, Lightbown and Spada (2006) argue that it is wrong to assume that second language development is a sort of accumulation of rules and they conclude that no matter how language is presented to learners, certain structures are acquired before others (p. 80). Moreover, they maintain, isolated presentation and practice of one structure at a time does not provide learners with an opportunity to discover how different language features compare and contrast in normal language use (p. 189).


Table 1: The order of presentation of grammatical structures compared across three intermediate coursebooks ProFile Present Simple and Continuous Past Simple and Present Perfect Countable and uncountable nouns Modals Future forms with will Making comparisons The Passive 1st and 2nd conditional Relative pronouns and clauses Reported Speech 3rd conditional New Hotline Present Tenses: Revision Present Perfect Tense Present Perfect and Past Simple Expressing the future Passive voice Question tags want someone to ... Reported Speech statements Reported speech questions Theres sb See/hear Should/Shouldnt have If clauses (3) Direct and indirect objects Passive with an indirect subject Headway Present Simple (1) Present Continuous Present Simple (2) Past simple, past continuous Modals (could, would, shall..) Future tenses (will, going to) Comparatives, superlatives Present Perfect Simple Modals (obligation) Conditionals Present Perfect Continuous Present Continuous for future Might /could (future possibility) Passive sentences Reported speech


As authentic texts are not constrained by and were not conceived with language teaching features in mind, they are therefore to be used for different purposes than inauthentic materials that are written with structures. An authentic text will often be a mixture of what intermediate students have learned and what they do not know yet at a certain point of their L2 development. As indicated above, the teachers role at the intermediate stage is to automatize (Brown, 2007) the bits and pieces of the target language available for automatic use. So if the structures met in the text are not challenging enough, the learners will get an opportunity to reinforce what they already know, and if the lexico-grammatical structures have not been mastered yet, but the input is comprehensible, students can still comprehend the general meaning of the text (which is fodder for acquisition, rather than learning from what is explicitly taught). Lewis and Hill (1992) argue that as soon as it is accepted that a clear distinction needs to be made between what students must produce and what they must understand, material which is structurally beyond the students productive level may be introduced to practise understanding (p. 27). Similarly, Lightbow and Spada (2006) maintain that learners who successfully acquire a second language outside classrooms are exposed to a great variety of forms and structures they have not mastered. Thus students need to develop strategies for dealing with authentic materials if they are eventually going to be prepared for language use outside the classroom (p. 191). Although this advocates the creative process of making decisions after content has been selected, there is no implying that coursebooks should be avoided. As suggested by Nunan (1988), the dilemma of sequencing input according to some sort of structural progression might be addressed in two ways: we can either abandon any attempts at structural grading, or we can use a list of graded structures (included in


a coursebook) to our advantage; learners are exposed to genuine texts, but they are only expected to focus on selected items from the list (p. 30). The latter approach is more acceptable in Czech state schools because, after all, teachers must meet the requirements of the curriculum.

Authenticity and ideology

Writers of authentic texts usually use language to create a point of view, which they share with the reader and thus influence the readers view. According to Talbot (2005), Readers are drawn into a kind of complicity with the texts they read. When meanings are simply obvious, that complicity and subjection are complete. Complicity is necessary for understanding; it is not inherently undesirable (p. 146). While publishers need to be particularly careful when choosing the content of classroom texts to avoid inappropriate topics informally known as the PARSNIP topics (politics, alcohol, religion, sex, narcotics, -isms, and pork), teachers are relatively free in this respect; they can adjust teaching materials to the local needs of their students, without having to overly concern themselves about the fact that the selected text touches upon one of the inappropriate topics. From an ideological point of view, CITR does not seem to exhibit such dangers. However, a careful reader can find a few hints between the lines, such as how particular characters come to embody certain values or attitudes (Lazar, 2007, p.13). First of all, the fact that the American husband is depicted as a slightly uncaring person, lacking interest in his wife, whereas the Italian padrone is the opposite, can influence the readers view of the world and can enhance stereotypical thinking. What is more, the American wife is depicted as an unsatisfied, unhappy and bored young woman who 16

desperately wants something she cannot have, while the Italian maid is her considerate and jolly counterpart. The story consists of about 1,200 words, and the word American appears five times. Judging from the high frequency of occurrence, the author apparently overuses the word deliberately as he wants the reader to notice it. The fact that the girl has no name, while her husbands name appears in the text eight times, is significant as well. This particular way of expressing (or rather omitting) social actors in the texts is never accidental. Social Actor Analysis can be applied to any discourse in which people are evaluated through the way they are named or categorized. What is absent in a text is as important as what is present (Simpson & Mayr, 2010, p. 73). There is no implying that this touch of ideology must necessarily be a drawback. On the contrary, it can widen students horizons and generate an interesting discussion in the classroom.

Authenticity and motivation

When discussing the advantages of using authentic texts in ELT, one should not forget to mention motivation. Thornbury (2005) argues that authentic texts offer the learner reliable data about the language, and they have a greater chance of capturing the learners interest and therefore attention than fake or imitation texts (p. 116). The key problem with this argument is that there is no reason to believe that inauthentic texts cannot be motivating as well. Nevertheless, for example, Lewis and Hill (1992) argue that authentic materials show students that what they are learning is useful outside the classroom. They also reassure students that what they are studying is real (p. 28). As far as motivation is concerned, it is essential to touch upon cross-curricular approach principles relevant to the Czech education system. Czech students simply need 17

to read literature with a capital L in order to pass their exams. At Czech secondary schools students studying for their Literature Maturita exam are supposed to get acquainted with literary works of selected authors, including American writers. Dealing with something in English and using the knowledge in literature lessons, or vice versa, can increase students level of motivation as it can produce a sense of achievement. Students learn and acquire the language, familiarize themselves with a style of writing, learn about the authors life, and discuss the meaning of a literary work. Working with an authentic literary text can improve cross-cultural understanding while continuing to develop linguistic skills. Moreover, consistent with a long-held view of learning styles, Marsh (n.d.), an advocate of Content and Language Integrated Learning, argues that certain types of language teaching methods may suit some students, but not others. Thus some children, for example, prefer learning about the language, while others prefer learning by using the language (p. 5). As will be shown later, the teacher can exploit CITR in numerous ways, either focusing on language features, or the content, or both.


Chapter Three: Literature in ELT

Challenges of literary text

If authenticity of classroom materials is a somewhat controversial issue, the same can be said about using literary texts in ELT. Based on my learning experience, some literature, especially of the capital L variety, may be highly de-motivating, mainly because the sophisticated use of language in literary texts will either be lost on learners, or be completely irrelevant in terms of their immediate language means. Thornbury (2005) argues, however, that there need not be any major differences between the approach to using non-literary texts and the approach to using literary ones, even though the teacher might have to work harder at the pre-text stage, providing background knowledge including cultural and biographical information, and might have to intervene more at the comprehending stage (p. 143). This will be discussed later in the section on schemata. At some point students should be encouraged to respond to the text. Lazar (2007) argues that a literary text can be an important vehicle for generating discussion, critical thinking and controversy in the classroom. However, students will often need help and guidance to discover all the multiple levels of meaning in a literary text (p. 10). Lazar (2007) argues that the teacher should not expect a student to reach a definite interpretation of the text as the meaning of a literary text can never be fixed or frozen, but manufactured by the reader, and different readers make sense of it in their own way, depending on the society in which they live and their personal psychology (p. 10). Similarly, Nuttall (2005) maintains that with controversial texts, where there is widespread disagreement about the meaning of the text, we must accept that various 19

interpretations are possible and even enriching, and we can use the opportunity to explore them all (p.18). However, Lazar (2007) warns about teachers demanding a personal response from students without providing sufficient guidance in coping with the linguistic intricacies of the text. Some texts can be so remote from students own experience that they are unable to respond meaningfully to them (p. 25). Thornbury (2005), for example, draws attention to the fact that the function of literary texts is primarily expressive. Such texts often use language imaginatively and playfully, and the authors might say one thing and mean another. Literature is full of symbolism and the meaning is often not literal, but it has to be inferred (p. 135). This can cause difficulties, especially in a less advanced language classroom. Fortunately, as mentioned earlier in the thesis, Hemingway became famous for his simple and direct style of writing. As the following examples shows, from the stylistic point of view, the overall structure of the passage its parallelism, use of repetition, and the extreme simplicity of its vocabulary certainly helps accentuate the main ideas, which makes the text (and the message), easier to understand. He stood behind his desk in the far end of the dim room. The wife liked him. She liked the deadly serious way he received any complaints. She liked his dignity. She liked the way he wanted to serve her. She liked the way he felt about being a hotel-keeper. She liked his old, heavy face and big hands. Liking him she opened the door and looked out. And I want it to be spring and I want to brush my hair out in front of a mirror and I want a kitty and I want some new clothes.


What can also complicate the matter both for teachers and students is the fact that language in literary texts is used metaphorically. Surprisingly, CITR is not particularly dense with metaphors. While a business article about increasing unemployment can have a metaphorical expression in every other sentence, here it takes some time to encounter one. But it is more complex than this the story is a metaphor itself. As Lakoff and Johnson (1980) put it: From the experientialist perspective, metaphor is a matter of imaginative rationality. It permits an understanding of one kind of experience in terms of another, creating coherences by virtue of imposing gestalts that are structured by natural dimensions of experience. New metaphors are capable of creating new understandings and, therefore, new realities. (p. 235) So what seems to be a simple tale of a couple spending their rainy afternoon in a hotel room may be a metaphor for their relationship. The description of the bad weather and an empty square in the first paragraph sets the negative mood, and the isolation of the couple is reinforced by the following sentences: There were only two Americans stopping at the hotel. They did not know any of the people they passed on the stairs. This interpretation can be inferred from what is present in the text. However, the fact that texts are not produced in a vacuum and connect to a larger discourse can complicate the situation. Hoey (1991) argues that the appreciation of much literature is dependent on the readers ability to recognize allusion to other works (p. 153). Thornbury (2005) suggests that this intertextuality will prove allusive to many second


language learners, who may lack knowledge of the shared background, both cultural and linguistic (p. 138). But it is also the authors experience that influences his or her writing. Broughton et al. (1978) point out that it may be necessary to know who wrote the text, when they wrote it and for whom, in order to understand it fully. Such information may not be present in the text itself but has to be acquired from the outside source (p. 97). To understand CITR fully, it is helpful to read what Brennen (2006) writes about Hemingways devotion to family pets, especially cats: he adopted stray cats in every town he lived, owning fifty-seven cats in Cuba and two in Idaho when he died. Students will learn that the story is also a tribute to Hemingways wife, who was dealing with her first year of marriage, the loneliness it entailed and her deep desire for motherhood (p. 16). According to Brennen, Hemingway based the story on an incident that happened in 1923. His wife was pregnant when she found a kitten hiding under a table in the rain (p. 16). The above mentioned challenges of literary texts should not be interpreted as disadvantages. On the contrary, the position taken by this thesis is that resourceful and knowledgeable teachers may use them to their advantage.

Benefits of using literary texts

Although the preceding discussion may suggest that literature with a capital L had better be avoided in ELT, there are enough arguments in favour of literary texts. Back in 1968, Rivers argued that listening and speaking skills were in the centre of attention. Gloomy prophets predict that future graduates of foreign-language classes will be fluent chatterboxes who are able to produce rapid-fire 22

utterances in a foreign language but have nothing worthwhile to say, because they have never been given the opportunity to share the thinking of the great minds of another culture, and so to widen the horizons of their knowledge and understanding. (Rivers, 1968, p. 213) This thesis does not want to argue whether the quote has become reality but it seems to support the hypothesis that CITR is a suitable text for L2 classroom because it offers students an opportunity to widen the horizons of their knowledge and peer into one of the great minds of another culture. Duff and Maley (1992) argue that genuine samples of various text types are open to multiple interpretations, and an opinion gap can lead to a genuine interaction. Moreover, literary texts are non-trivial; they deal with matters which concerned the writers enough to want to write about them. This genuine feel is a powerful motivator learners can bring a personal response from their own experience (p. 6). In addition, Kramsch (1996) maintains that arguments in favour of literature in the language classroom are that a piece of literary prose, more than any other text, appeals to the students emotions, grabs their interest, remains in their memory, and makes them partake in the memory of another speech community. In her view, the main argument for using literary texts in the language classroom is literatures ability to represent the particular voice of a writer among the many voices of his or her community and thus appeal to the particular in the reader (p. 131). Furthermore, Lazar (2007) states: Literary texts have a powerful function in raising moral and ethical concerns in the classroom. The tasks and activities we devise to exploit these texts should encourage our students to explore these concerns and connect them with the struggle for a better society. (p. 3)


Another supportive argument in favour of literary texts is, as Lazar (2007) argues, that from a linguistic point of view, literature may provide an appropriate way of stimulating acquisition as it provides meaningful and memorable contexts for processing and interpreting language (p.17). Thornbury (2005) writes that any approach that highlights the relation between a text and its context should serve the learners awareness about the way texts are produced and interpreted. Genre theory argues that language is best learned through the analysis and mastery of specific genres, since such an approach best reflects the way language is shaped by its social contexts of use (p.102). This is meant in a broader sense as the term genre originally came from literary studies but has been extended to mean any frequently occurring culturally embedded, social process which involves language. Text-based instruction, also known as genrebased approach, sees communicative competence as involving the mastery of different types of texts. Text here is used in a specific sense to refer to structured sequences of language that are used in specific contexts in specific ways. The text exists as a unified whole with a beginning, middle, and an end; it confirms to norms of organization and content, and it draws on appropriate grammar and vocabulary. Communicative competence thus involves being able to use different kinds of spoken and written texts in the specific contexts of their use (Richards, 2006, p. 36). Finally, and this is the most relevant argument to this thesis, Broughton et al. (1978) maintain that literature does not have to wait for the advanced level of the language, even though much literature is not accessible to the beginning learner. But even the most elementary learner can derive pleasure from traditional rhymes or linguistically simple poems (p.114), and an intermediate learner of English can enjoy reading a simple short story.


Chapter Four: Reading and deriving meaning from text

To develop some of the key issues raised above, texts (both authentic and inauthentic) are usually used in English classes for two main purposes: as a way of learning new language and as a way of developing reading comprehension. To set the scene for the content of Chapter Five, Chapter Four will present some theoretical background concerning the reading process and getting meaning from text. Firstly, Harmer (1989) suggests that reading is an exercise dominated by the eyes and brain. The eyes receive messages and the brain then has to work out the significance of these messages (p. 153). Pressley (2001) argues that Reading is often thought of as a hierarchy of skills, from processing of individual letters and their associated sounds to word recognition to textprocessing competencies. Skilled comprehension requires fluid articulation of all these processes, beginning with the sounding out and recognition of individual words to the understanding of sentences in paragraphs as part of much longer texts. (para.7) Similarly, Ehrich (2006) believes that the written text must be analysed from the interaction of two contrasting processes the decoding of individual units in conjunction with the use of context and inference to establish meaning. He maintains that the degree to which each of these two contrasting perspectives in the processing of written text is emphasized is a matter of contention among reading researchers (p. 15). This has some classroom implications. As Lewis and Hill (1992) write, natural language is structured at different levels phrases, sentences, paragraphs, and even larger units such as a whole chapter of a book. The ability to use a language implies an


understanding of the structure of such small units as phrases, and such large units as chapters. So if language teaching is to reflect what language really is, teachers need to take into consideration a wider range of structural devices, and a wider range of skills which students need in order to encode and decode the language successfully (p. 30). Moreover, as Rivers (1968) contends, there are three levels of meaning the student must learn to extract from the printed patterns: (a) lexical meaning, i.e. the semantic content of the words and expressions, (b) structural and grammatical meaning, i.e. deriving from interrelationships among words, and (c) social-cultural meaning, i.e. the evaluation which people of the students own culture attach to the words and groups of words in the text (p. 216). Similarly, Huang (2009) argues that reading in a foreign language is particularly hard, because the words and grammatical structures, the text conventions and the cultural context are all less than familiar (p. 74). The preceding discussion has implied that deriving meaning from text may be easier for a native speaker of the target language than for an intermediate L2 learner. The following passage will support this implication. In Ehrichs paper (2006), an argument is presented that so called Vygotskian inner speech (Vygotsky, 1934) acts in two main ways to process text during silent reading, as a mechanism which condenses chunks of text into compact meaning units, and as a subvocal rehearsal mechanism which elicits meanings when reading cognition becomes problematic (p. 12). Anderson (1999) argues that a fluent reader automatically chunks the language, while second language learners often read word by word striving to comprehend the text. He argues that chunking helps increasing reading speed and comprehension of what is being read (p. 57), which is due to the fact, as Nations (2001) maintains, that instead of having to give close attention to each part, the chunk is seen as a unit. So instead of having to refer to a rule or pattern to comprehend of produce the


chunk, it is treated as a basic existing unit (p. 320). Similarly, according to Favreau and Segalowitz (1983, as cited by Lightbown and Spada, 2006, p. 39), the fact that proficient language users can give their full attention to the overall meaning of a text or conversation (whereas learners use more of their attention on processing the meaning of individual words) helps to understand why second language readers need more time to understand a text, even if they eventually do fully comprehend it. Furthermore, chunking also helps students in production stages correctly understood and stored, lexical items are available for immediate use. Lewis (2000) states, [...] in addition to understanding the input the learner must notice the chunks which carry the meaning. Each chunk is a single choice of meaning; if chunks are not noticed as chunks, they cannot be stored in the way which facilitates their availability as output. If input is to become optimal intake, understanding and noticing the chunks are both necessary (though perhaps still not sufficient) conditions. (p. 180) It is unlikely that teachers will discover what exactly happens in the mind of their students during silent reading, but it is possible to draw some conclusions from the way students read aloud. Based on my observations, students often chunk incorrectly while reading aloud because they do not understand what they are reading, and conversely, they do not understand because they chunk incorrectly. It is helpful when the students are encouraged to read the text several times, or when they get an opportunity to hear the text. Phrased meaningfully, the introductory paragraph of CITR may then read as follows: There were only two Americans\ stopping at the hotel. They did not know\ any of the people\ they passed on the stairs\ on their way\ to and from their room. Their 27

room\ was on the second floor\ facing the sea. It also faced\ the public garden\ and the war monument. There were big palms\ and green benches\ in the public garden. In the good weather\ there was always an artist\ with his easel. Artists liked the way\ the palms grew\ and the bright colors\ of the hotels\ facing the gardens\ and the sea. Italians came\ from a long way off\ to look up\ at the war monument. It was made of bronze\ and glistened\ in the rain. It was raining. The rain dropped\ from the palm trees. Water stood in pools\ on the gravel paths. The sea broke\ in a long line\ in the rain\ and slipped back\ down the beach\ to come up\ and break again\ in a long line\ in the rain. The motor cars\ were gone\ from the square\ by the war monument. Across the square\ in the doorway of the cafe\ a waiter stood\ looking out\ at the empty square. The American wife\ stood at the window\ looking out. Outside\ right under their window\ a cat was crouched\ under one of the dripping green tables. The cat\ was trying\ to make herself so compact\ that she would not be dripped on.

It might be interesting to mention that more than forty years ago, the possibility of spatial separation of reading materials was considered by some researchers. However, after two minor experiments and a thorough major experiment, Carver (1969) concluded that for mature readers, the chunking of reading material did not produce faster reading rates with no loss in comprehension per passage or did not produce higher comprehension scores with no decrement in reading rate. These results suggested that the spatial separation of reading material into meaningfully related groups of words would probably not improve the reading efficiency of mature readers, reading at their normal rate, regardless of the method used to separate the material (p. 11).


Even nowadays the methodology of chunking approach is open to some criticism. Michael Swan (2006), for example, argues that it might seem disappointing, but most non-native speakers must settle for the acquisition of a variety characterized by a relatively restricted inventory of high-priority formulaic sequences, a correspondingly high proportion of non-formulaic grammatically generated material, and an imperfect mastery of collocational and selectional restrictions. We cannot do anything about it, he argues, languages are difficult and cannot be learned perfectly (section on Realism and Prioritising). It has been shown that reading and getting meaning from texts is a complex process. Nuttall (2005) argues that whenever we read, we use both top-down and bottom-up processing. These are complementary ways of processing a text. Sometimes one predominates, sometimes the other, but they are both needed. In top-down processing we draw on our own intelligence and experience the predictions we can make, based on the schemata we have acquired to understand the text (p. 16). A topdown approach makes use of all that the reader brings to the text; prior knowledge, common sense, etc. In bottom-up processing the reader builds up a meaning from recognizing the letters and words, working out sentence structure. We scrutinize the vocabulary and syntax to make sure we have grasped the plain sense correctly (p. 17). Huang (2009) explains the distinction between the top-down and bottom-up perspectives as follows: Top-down strategies include guessing meaning from context, predicting, using background knowledge, using text structure. Bottom-up strategies include looking up unknown vocabulary in a dictionary or glossary, working out sentence grammar, deciphering reference chains. (p. 175) According to Huang (2009),


Classroom reading practices which are based on bottom-up strategies may help to scaffold students development of reading skills. However, if students are not encouraged to go beyond these strategies, they may learn reading habits which are over-focused on decoding to the detriment of other reading resource. (p. 175)

Students have at their disposal a multitude of strategies which help them comprehend a written text. However, it should be stressed again that if it is a text with lexical and structural difficulty that will overwhelm the students, or if the language used is simply of little relevance in terms of their language means, they will never be able to grasp the meaning properly. The following chapter will attempt to show that there is no such danger related to CITR; the text is relevant to the intermediate L2 classroom, even though the teacher may occasionally need to provide some additional support to help the learner comprehend the text.


Chapter Five: Cat in the Rain analysis and implications

Top-down perspective

In real life, we read with a certain purpose in mind reading is always meaning based. As seen in the previous chapter, getting meaning from texts can be a challenging task, especially for an L2 learner, who may need a lot of help and support from the teacher. The following pages intend to look at some of the numerous ways of helping learners to understand CITR. This chapter outlines the top-down approaches to instructional support and, almost as a side effect, the discussion will also offer a variety of possibilities of exploiting the text for promoting language learning. Thornbury (2005) lists the top-down factors related to text difficulty as follows: topic familiarity, including background knowledge context familiarity cognitive complexity, e.g. density of information visual support length layout and signposting organization of text internal cohesion, e.g. linking sentences (p. 116)

The following sections will look at some of the factors in turn (although not necessarily in the above order).


Meaning and schemata

Lomnicka (1998) maintains that many students erroneously interpret comprehension as a process consisting of mere translation and paraphrasing. Students seem to be generally satisfied with achieving a minimal level of comprehension via quick and basic translation. If they are able to translate the words into their L2, then they assume they have understood the text (p. 49). According to Barnett (1989), foreign language reading can no longer be seen as simply the decoding of more or less unknown vocabulary and grammar. As mentioned earlier, unique reader characteristics determine the meaning that an individual reader will create for any particular text. Barnett argues that the text is still essential, but comprehension truly depends on the reader's expectations as defined by his or her linguistic proficiency, first language reading skill, reading strategies, interest and purpose in reading the text, and formal schemata (p. 111). Schemata are defined as the abstract cognitive structures which incorporate generalized knowledge about objects and events (Lazar, 2007, p. 9). According to Nuttall (2005), the kinds of assumptions we make about the world depend on what we have experienced and how our minds have organized the knowledge which we have got from our experiences. She describes a schema as a mental structure which derives from all the particular experiences we have had. It is a structure because it is organized; it includes the relationship between its component parts. Schema is a useful concept in understanding how we are able to interpret texts (p. 7). Similarly, Anderson (1999) describes a schema as the readers background knowledge which includes all experience that a reader brings to a text: life experiences, educational experiences, knowledge of how texts can be organized rhetorically, knowledge of how the first language works, knowledge how the second language works, and cultural background and knowledge, etc. (p. 11). 32

So what is the students schema of a cat in the rain? In order to discover this, students may be encouraged to visualize a cat in the rain, or asked to find different pictures of cats to describe. Do they all share the same schemata? The author expects us to share the attitude to a cat in the rain, but different readers fail to tune in to the same feeling. Some readers are indifferent, or some even hate cats, so the story may have a different impact on them. After reading, students can compare the new schema with the authors one. Is it different? What does Hemingways cat look like? The same can be done with the main characters. Students can draw or find a picture that matches their idea. However, Anderson (1999) suggests that some readers may not have prior knowledge to activate at all. In such a case, it will be necessary for the teacher to establish background prior to asking the students to read so that they have sufficient information to understand the text (p. 12). As will be shown later, the story does not exhibit any major difficulties in relation to syntactic complexity or vocabulary. It can, however, be challenging enough for an intermediate student. Here is an example: The sea broke in a long line in the rain and slipped back down the beach to come up and break again in a long line in the rain. Although the sentence above consists of high-frequency words most intermediate students know, a student who has no schema of the sea, and the way water in it moves, might face serious difficulties when deriving meaning from the sentence. It is the teachers job to help the student activate the schema (or to create it). As translation is rather problematic here, especially the part the sea broke in a long line, the teacher can help the student visualize the scene. The Internet provides teachers with useful tools


especially visual aids and students holiday snapshots taken at the seaside can also be helpful at this stage. Finally, can background knowledge hinder comprehension? As Anderson (1999) argues, if students background knowledge interferes with reading, the teacher may have to correct the background knowledge through a pre-reading activity before reading comprehension can be achieved (p. 12). This section has defined the term schema and explained why it is desirable to activate students background knowledge. The following sections will outline some techniques which help students activate appropriate schemata.

Prediction as activation of background knowledge

One of the classroom techniques which can be used to facilitate the activation of background knowledge is prediction. It is a strategy for activating both background knowledge and the most likely schema that underlies the text. We predict all the time as we read. We predict right from the beginning from the title. In our story the title is somewhat tricky because the cat and the rain are central, but also symbolic, which unfolds in time, linearly. Thus pre-reading discussions provide an opportunity for readers to see what they know about the topic. Anderson (1999) argues that if readers make predictions about what they think the text content will be, they can then read to support or reject their hypothesis. This particular activation would need to be tied into activities that are used during later phases of the reading lesson, particularly an activity to verify whether the predictions were actually realized (p. 15). Broughton et al. (1978) argue that most English native speakers faced with the beginning part of a sentence would be likely to predict how the sentence might continue 34

(p. 91). Similarly, Nuttall (2005) writes, Prediction also helps us to make sense of sentences; even the first word sets up the expectancies of what the next word will be, and as the sentence develops, our ability to predict what comes next often increases (p. 13). Nuttall illustrates the point with the following activity (which has been adjusted to the introductory sentence of CITR). Students are asked to take a card to cover the exercise. They try to guess the correct answer and move the card down the page until the whole sentence is revealed. There be|five|were There were one|only|in There were only two|any|one There were only two person|Americans|Spain There were only two Americans stop|stopping|stopped There were only two Americans stopping at|she|they There were only two Americans stopping at the|where|there There were only two Americans stopping at the morning|face|hotel There were only two Americans stopping at the hotel. This exercise demonstrates how grammatical features and meaning both influence what we expect to come next in a sentence. As the text grew, the student became able to use meaning to choose between grammatically suitable options, because they had a growing understanding of what the sentence was about. Similarly, in longer texts both structure and meaning help us to predict. This kind of longer term prediction is one of the attractions of stories; speculation about what will happen urges us to read to the end (Nuttall, 2005, p. 13). Another prediction activity is looking at the key words before reading. From a linguistic point of view, keywords are the words that are far more frequent in the text, 35

proportionally, than they are in a general reference corpus. Key words are usually content words and they are nouns (Thornbury, 2005). Complet Lexical Tutor (Cobb, 2008) a web-based tool for analysing texts from various perspectives has generated the following key words in CITR: signora monument kitty palm doorway maid gravel umbrella rain path mirror desk tight hotel wife America window square table hair Compleat Lexical Tutor cannot, however, do all the work. On the one hand, key words are always central to meaning. On the other hand, this thesis deals with a literary text and therefore, more factors must be taken into consideration; mathematical calculations are precise but they cannot reveal everything the reader must read between the lines. Anderson (1999) suggests that another way to build or establish background knowledge is through the use of semantic maps. Semantic mapping is similar to brainstorming. The readers may be given a key word or concepts that will be part of the reading material. They are then asked to generate words and concepts they associate with the key word. Semantic mapping allows students to link ideas and concepts they already know to the new concept that will be learned, thus helping to build prior knowledge before reading (p. 14). Feinberg (2009) has created a helpful resource called Wordle a web-based tool generating word clouds from a selected text. The clouds give greater prominence to words that appear more frequently in the source text. This image can be used as a prereading activity (for setting up appropriate schemata), or after-reading activity (when retelling the story or writing a summary). As Figure 1 shows, the most prominent words generated from CITR are the words cat and want. When the pre-reading activity was piloted, students came up with their ideas and predictions such as: Somebody (George or his wife) wants a cat because


he or she likes cats. George likes reading. The wife likes looking out of the window. It is raining. George stood in the square. George is American. Students ideas were recorded and later, after reading, the predictions were compared with what actually happened in the story.

Figure 1: Key words in CITR generated by Wordle

Asking questions to help learners activate background knowledge and understand the text

As mentioned above, activation of prior knowledge of text organization can facilitate reading comprehension. A pre-reading discussion on the type of text structure, and what expectations a reader may have about the organization of the material, is very valuable for learners (Anderson, 1999, p.14). According to Nelson (2011), Asking questions before reading can provide valuable insight as to how a learner will process new information. Not only can background knowledge be quickly assessed by a few pre-reading questions, they can 37

also assist learners in making early connections between previous instruction and new information. Further, the right pre-reading questions can also incite interest in a story or text, gathering the attention of the learner and focusing it where it needs to be. (n. p.) But questions can also be asked at other stages of reading. According to Doff (1991), there are two main aims in asking questions on a text; to check understanding and to help students read the text (p. 175). Doff argues that it is best to ask short, simple questions which help to break down the meaning of the text and make it easier to understand. Students may only be required to give short answers as the aim is to check comprehension, not to get students to reproduce the text. Students may be allowed to keep their books open, so that they can refer to the text to answer the questions. Even if the textbook contains good comprehension questions, it is often good to ask the teachers own questions first. The appended questions can be answered later in pairs or for homework (p. 63). As written by Tsai and Wilkinson (n.d.) in their paper, the effect of text-based discussion as a means of promoting reading comprehension has been demonstrated in a variety of empirical studies, which have shown that features of whole-class discussion were positively related to the students reading comprehension and literary response (n. p.). Haynes (19982009) argues that it is advisable, especially with reading comprehension, for teachers to ask critical thinking questions from all levels of Blooms Taxonomy. Questions at higher levels of Blooms taxonomy can be modified so that the language is simplified but the task remains the same (n. p.). Nelson (2011) adds to this topic: Set aside the fact that comprehension is the second classification level of Blooms Taxonomy, assessing true reading compression skills must reach across the


levels of the cognitive domain (n. p.). Here is a list of suggested questions and tasks across the six levels of Blooms taxonomy:

1. Do you remember the characters in the story? (Knowledge) 2. Describe the characters as fully as you can. (Comprehension) 3. Dramatize the dialogues. (Application) 4. Compare the two male (female) characters in the story. (Analysis) 5. How does Hemingway address George in the first half of the story, and then in the second half? Explain why. (Synthesis) 6. Can you interpret the significance of rain in the story? (Evaluation)

Widdowson (1978) argues, however, that we are not, in normal circumstances, required to submit ourselves to interrogation after having read something, knowing at the same time that the person asking the questions already knows the answers. Although Widdowson maintains that this practice emphasizes the artificiality of the learning process (p. 96), the above examples show that well-designed questions and tasks may have no definite answers (at least those from the higher levels of Blooms Taxonomy). In fact, the only definite answer could be the one to question number one. While answering question two, the learners can be quite creative, although they have to stick to the details given in the text. Some learners may consider the wife a negative character, while others may feel sorry for her, which will be reflected in the way they describe her. Nevertheless, the most interesting open questions will inevitably emerge at the end of the story: What kind of cat did the maid actually bring? Why did she do it? Did this make the girl happy? At this point, students may also talk about their own experience related to the topic, and they can discuss their wishes, hopes, expectations, desires, dreams, and ambitions.


Another way to eliminate the artificiality of the learning process is encouraging students to ask their own questions (while reading or after reading). Based on my experience, this is what small children do while being read to; they jump in and ask questions when they do not understand, or when something captures their attention. The parent usually provides a gloss or answers the question, or sometimes the childs reaction does not require any response at all. As Nuttall (2005) suggests, one of the most important things students have to learn is to interrogate texts; to continually stop and ask themselves questions. Students attention should be focused on the text, especially on pieces of the text that are important and problematic; things that can give rise to misunderstanding texts. In the early days, the teacher may like to talk through some texts in this way with the class, thinking aloud about the questions the text raises (p.37). However, the teacher will first need to create the kind of atmosphere that motivates students to ask questions. As there are not many opportunities to ask genuine questions in the classroom, this will be a valuable exercise, particularly for intermediate students, who still make basic errors (due to their interlanguage and L1 transfer). The teacher can ask students to read the first sentence and answer the following question: There were only two Americans stopping at the hotel. > Were there only two people at the hotel? > We dont know. Maybe. At this point, the sentence looks ambiguous but while reading on, the same question can be asked again. They did not know any of the people they passed on the stairs on their way to and from their room. > Were there only two people at the hotel? > No. Was the hotel in America? > We dont know. Maybe. We cannot say at this point. Italians came from a long way off to look up at the war monument. > Was the hotel in America? > Perhaps in Italy.


Then the students can go on asking their own questions. From my reading experience, this is what happens when reading. The meaning is constantly reconstructed after new information is obtained.

Text structure and organization

As discussed earlier in the thesis, CITR is organized in a particular way to convey the authors message. Barnett (1989) argues that, as research on formal schemata shows, readers' familiarity with standard text structures (description of the place, characters, chronological sequence, etc.) and their ability to recognize and follow structure help them comprehend. Teachers may need at first to identify text structure for beginning or weaker readers; later, students should learn to recognize it on their own. The teacher will not always be there to lead them; moreover, too much direction can destroy reader interest (p. 118). Once students perceive that a text belongs to a certain familiar genre, many unfamiliar words and even sentence structures may become less difficult. Since first language readers generally know (from source, title, or illustrations) the genre of a text they are about to read, it is fair to enlighten foreign language students before they begin reading and to ask what they know about that genre (p. 119). Similarly, according to Barnett (1989), some texts closely follow a recognizable genre pattern or familiar story line; some have a particularly appropriate title or helpful illustrations; some require more introduction and background explanation than others; some have an absurd or ironic twist. Previewing a text with students should arouse their interest and help them approach reading more meaningfully and purposefully as the discussion compels them to think about the situation or points raised in a text (p. 123).


Every text is organized in a specific way, and it is not incidental. What principle is used to organize CITR as a whole? Is it organized chronologically? When skimming the text, the students can spot at first sight that there is an introductory paragraph, and the rest of the text consists of stretches of direct and indirect speech. These direct speech interruptions may promise that the reading is not too challenging as the sentences are usually shorter, sometimes consisting of only two words. And why did Hemingway give a description of the place at first? There were only two Americans stopping at the hotel? Why did he start like this? A possible answer was offered in the earlier chapter discussing metaphors in literary texts.

Density of information

As was implied in the section on authentic texts and ideology, when a careful reader tries to decipher the meaning and the writers point of view, he or she does so by looking closely at what is in the text, but more importantly, especially with literary texts, at what is not in the text. Good writing gets the most from the least, and it often leaves a lot unsaid. As Thornbury (2005) writes, one of the features of good literature is its economy. It is the unsaid that is as important as what is said. He argues, Only at hinting at something can there be resonance (p. 137). So in a literary text, the problem may not be that there is too much information, but that there is too little. Lazar (2007) maintains that if the readers possess the schemata assumed by the writer, they easily understand what is said in the text and also make the necessary inferences about what is implicit, rather than stated (p. 9). This can be well exploited in the classroom; when discussing CITR, the following questions can be asked: Who are the people from the story? How did they


arrive at the hotel? Do they have children? How long have they been married? What kind of life do they lead in their own country? What class are they? None of this is explicitly said, and the students will have to use various hints and their imagination to answer the questions. Hemingway's title to his story carries more meaning than the literal cat in the rain. The story talks about a cat stuck in the rain. There are many interpretations available on the Internet (e.g. at or, so as a post-reading activity, the teacher can ask students to read those interpretations and compare them with their own ideas. At this point, students will get beyond the text and will be exposed to additional language input related to the topic.

Text Cohesion

If speakers of English hear or read a passage of the language which is more than one sentence in length, they can normally decide without difficulty whether it forms a unified whole or is just a collection of unrelated sentences (Halliday & Hasan, 1976, p. 1). Cohesion is one of the features which distinguish a text from a random collection of sentences (Thornbury, 2005, p. 34). Thornbury maintains that cohesion is what makes texts hang together (and thus makes them coherent). If a text is not cohesive, it is not coherent it does not make sense to the reader. Writers intentionally use cohesive devices with the aim to make their texts easier to follow (p. 36). In other words, cohesion helps the reader to get meaning out of the text.


In CITR there are not many explicit linking words which connect the sentences. In the first paragraph, only one can be found in the fourth sentence, and it refers back to the previous sentence. (It also faced the public garden and war monument). However, the text is tightly cohesive as the author uses a great deal of grammatical reference. In the story there are precisely thirty-five instances of the additive conjunct and, seven instances of the causal so, three examples of the temporal then, and only one example of the adversative but. Furthermore, pronouns and the definite article refer back or forward to someone (or something), making the text cohesive. They can also refer outside the text to their referents. Table 2 shows the number of occurrences of pronouns and the definite article in the text.

Table 2 Reference pronouns and the number of their occurrences in CITR Reference the she her he his their they him Occurrence 118 34 17 13 12 6 3 2

The teacher can ask the students to identify these referents. Their room was on the second floor facing the sea. It also faced the public garden and war monument.


What does it refer to? Identifying the referent correctly can help the students comprehend the passage.

Table 3 Lexical cohesion repetition of lexical items in the introductory paragraph of CITR and the number of occurrences in the text rain/raining war monument square palms/palm trees face/faced/facing sea (public) garden(s) dripping/dripped window room hotel(s) cat artist(s) American(s) in a long line broke/break stood looking out 5 3 3 3 3 3 3 2 2 2 2 2 2 2 2 2 2


But what is even more striking than the amount of grammatical cohesion in the introductory passage of CITR is the use of lexical cohesion, especially through repetition and lexical chains (see Table 3). The passage, consisting of 212 words, almost feels like an excerpt from a childrens book. At the top of Table 3 there is rain and raining with 5 occurrences. However, the words dripping and dripped with 2 occurrences are also related to rain, which would make 7 occurrences altogether. The analysis of lexical cohesion is closely related to the concept of key words the repetition is apparently an attempt to enhance the message of the story, and a careful reader may notice that the words related to water (rain, raining, dripping, dripped) are crucial to meaning. This can also be ascertained via cross-reference to the authors own life experience. Our understanding of texts in part depends on our ability to make connections across text on the basis of multiple repetitions. In the above mentioned set of sentences (Their room was on the second floor facing the sea. It also faced the public garden and war monument.), the verb face appears twice, although in different forms, which is not only useful for comprehension, but also for vocabulary acquisition. The more often a word is met, the better the chances are it will be remembered. In his volume on patterns of lexis in text, Hoey (1991) argues that words should be learned in the morphological forms in which they are encountered because they may be the most frequent forms in use (p. 240). Drawing on the evidence from British National Corpus (accessed via Sketch Engine, Kilgariff, n.d, retrieved from, Table 4 shows that the verb face will be encountered most frequently in the form face, but the forms faced and facing are also frequent.


Table 4 Different forms of the verb face and their appearances in BNC Part of speech + its subcategory VVI the infinitive form of lexical verbs (e.g. face) VVG the -ing form of lexical verbs (e.g. facing) VVN the past participle form of lexical verbs (e.g. faced) VVB the finite base form of lexical verbs, including the imperative and present subjunctive (e.g. face) VVD the past tense form of lexical verbs (e.g. faced) 1,579 2,468 Number of hits in BNC 4,018 3,674 2,838

But learners can also benefit from observing cohesion in authentic texts at the productive level when producing their own texts. Hoey (1991) argues, for example, that the traditional advice to avoid repetition in writing needs to be couched with special care. Learners should not be encouraged to say the same thing over and over again, but they should be advised to make connection not only between what they are saying and what they said before, but also with what they later intend to say (p. 243). In order to make language a lot more expressive and interesting, students should be encouraged to vary the words they use as often as possible. Synonyms (and also hypernyms) are useful for making ones writing cohesive and more attractive. Apart from printed dictionaries, various web-based tools, such as (, are valuable and quick sources of information for learners. Furthermore, students attention should be drawn to the fact that the linking devices do not necessarily have to be used to make reference to the sentence immediately preceding. In fact, as Hoey (1991) argues, native speakers


characteristically use repetition devices to connect over a considerable distance in a text, whereas L2 learners tend to repeat at shorter distances, typically within the paragraph boundaries (p. 242). In the following passage, the whole clause is an attempt to make the story cohesive. It refers back to the sentence, which appears earlier in the text. As she stood in the doorway an umbrella opened behind her. It was the maid who looked after their room. You must not get wet, she smiled, speaking Italian. Of course, the hotelkeeper had sent her. With the maid holding the umbrella over her, she walked along the gravel path until she was under their window. Here, the repetition occurs within one paragraph for obvious reasons the sequencing of the events. However, there are instances where the author connects what he is saying to something he has said earlier (through lexical repetition again). In the same passage above, the gravel path refers to the same gravel path mentioned in the introductory paragraph. The first mention of gravel path appears on the 130th and 131st position in the text, and the second mention is 333 words later. Here is another example to illustrate this point: There were only two Americans stopping at the hotel. It is the very first sentence of the story. The use of the definite article indicates that some additional information will follow, since the reader does not and cannot know what hotel the author has in mind. At this point, Hemingway is already making connection to what he is going to say later; he already has the knowledge of the hotel, which he is going to share with the reader. This way of sequencing is typical for the narrative literary text type (starting a story with a piece of information of which the reader has no knowledge and elaborating on it later).


The most obvious conclusion to emerge from this section is that if students observe how cohesion is achieved in authentic texts, their own writing improves significantly. And as mentioned earlier, improving students writing performance should become one of the priorities at the intermediate level of proficiency in Czech state schools.

Bottom-up perspective

As indicated in the preceding pages, text selection is absolutely essential to successful reading instruction. Based on my observations, most intermediate students have a very limited experience with reading in L2, let alone with authentic literary texts. Thus for individual readings, teachers should find a balance between texts that are too long or difficult and texts that do not challenge students adequately. A variety of bottom-up factors should be considered before students are exposed to a specific text. However, it would be wrong to assume that top-down and bottom-up factors affecting text difficulty can be strictly divided into two separate categories; in fact, some factors fall into both categories, depending on the researchers point of view. Here is, for example, Thornburys (2005) list of bottom-up factors affecting text difficulty: sentence length and complexity grammatical familiarity lexical familiarity and idiomaticity lexical density (p. 116)

As will be shown later, these factors do not work separately; they interact and interweave. So if one of the factors appears problematic, it is not a reason for rejecting


the text, but for providing further instructional support. When looking at CITR from the bottom-up perspective, this thesis focuses on the sentence length, syntactic complexity, grammar, and vocabulary.

Sentence length

CITR is not a lengthy text; it is a very short story, with about 1,200 words and 117 sentences. In this thesis the term sentence will be used to refer to a grammatical unit consisting of one or more words that bear minimal syntactic relation to the words that precede or follow it, a string of words that are not arbitrary in their sequence (Plag, Braun, Lappe & Schramm, 2007), or orthographic units that are contained between full stops (Halliday, 1985, p. 193). Halliday distinguishes between a sentence (a constituent of writing) and a clause complex (a constituent of grammar). In CITR there are instances of sentences which include all the usual sentence constituents subject, verb, object and adverbial (I wanted it so much, she said.). However, not all the sentences are full sentences, which is not surprising in literary fiction, which usually contains fabricated utterances simulating genuine conversation (A cat in the rain?). Based on my experience as an L2 reader, reading feels much easier when the text includes short exchanges which interrupt long stretches of language. This point will be discussed in the following section, where the text and sentence length will come into play again.

Syntactic complexity

As the text consists of 117 sentences and about 1,200 words, the average sentence length is ten words. This indicates that from a syntactic point of view, the text is relatively easy to read. The longest sentence consists of 31 words (And I want it to be 50

spring and I want to brush my hair out in front of a mirror and I want a kitty and I want some new clothes.), whereas the shortest one has only two words. Forty-five sentences are in direct speech (or parts of direct speech). The majority of the non-direct-speech sentences (42 out 72) are sentences containing only one clause (She opened the door of the room.). Six sentences are clause complexes, in Hallidayan terms (1985), containing one independent clause and at least one dependent clause (She liked the way he wanted to serve her). Finally, there are 24 clause complexes composed of at least two independent clauses joined by a coordinator (She went over and sat in front of the mirror of the dressing table.). Even one of the most complex sentences is not too daunting, but definitely challenging enough for intermediate students ( With the maid holding the umbrella over her, she walked along the gravel path until she was under their window.). However, there are some problematic passages, such as the following sentence containing 29 words: The sea broke in a long line in the rain and slipped back down the beach to come up and break again in a long line in the rain. As mentioned earlier, this sentence may be confusing not just because it is so long and complex it also needs to be chunked correctly, and the reader must have an appropriate schema of the sea. Syntactic complexity and readability of whole texts can be measured by quantifiable linguistic and textual factors. The following version of the Flesch-Kincaid formula, for example, simplified by Carrell (1987), focuses on easily measurable features of word and sentence length. The original Flesch-Kincaid Grade Level formula (OFKF) and the simplified Flesch-Kincaid (SFKF) read as follows (Hedgcock & Ferris, 2009): OFKF: (0.39 ASL) + (11.8 ASW) 15.59 SEKF: 0.4 (ASL) + 12 (ASW) 16


ASL refers to the average sentence length (the number of words divided by the number of sentences), and ASW refers to the average number of syllables per word (the number of syllables divided by the number of words). Applying the Flesch-Kincaid formula is particularly useful for teachers, parents, and librarians to use because the value converts to a U.S. grade school level (the Flesch-Kincaid Level). The readability score of the first paragraph of CITR has been calculated for illustration. As the passage contains 15 sentences, 194 words, and 246 syllables, the formula reads as follows: SEKF: 0.4 (194/15) + 12 (246/194) 16 = 4.3 The result 4.3 is the equivalent of grade 4 text in U.S. grade school system. However, it should be stressed that the overall score is lower, since the sentences in the introductory paragraph are much longer than in the rest of the story. Calculating readability scores manually is extremely time-consuming, but the Internet provides teachers with various tools for quick text analysis. Readability can be measured easily at, for example, where free readability tests for texts are available. The Text Consensus Readability Tool will analyse the selected text and calculate the number of sentences, words, syllables, and characters in the sample. The program takes the output of these numbers and plugs them into seven readability formulas. These readability formulas will output the reading level and grade level of the text and help to judge if the audience can read the selected material. The results can be read in several ways. Firstly, according to seven reading formulas provided by the website (Flesch-Kincaid being one of them), the Readability Consensus score of CITR is 5, and the text is marked as very easy to read (see Figure 2).


Figure 2. CITR Readability Consensus retrieved from Part 1

Figure 3. CITR Readability Consensus retrieved from Part 2

However, as Figure 3 shows, another sample of the text, including mainly direct speech utterances, displays a lower score (when measuring the score, the text had to be divided into two parts as the pasted sample must be between 150600 words).

Figure 4. U.S. Grade Level Conversion Table graph statistics retrieved from 53

Furthermore, the graph statistic (see Figure 4) shows that CITR could be easily read by American students at the age of 1011. Unfortunately, it is hard to draw any conclusions here, as comparing native speakers of the target language with L2 learners is rather subjective. However, the Word Statistics (see Figure 5) can provide the teacher with a valuable source of information. It shows, for instance, all the unique words as opposed to repeated words. This can be useful for teachers and students when identifying key words, or when analysing text cohesion. It should be stressed, though, that one of the problems with lexical statistics is that they do not take account of other uses of words which typically manifest in different patterns: break is a common word someone breaks something or something breaks into pieces or stops working. But the sea breaks is quite different. It is clear that the meanings of words are related to word grammar. Grammar will be the subject of discussion in the following chapter.

Figure 5. CITR Word Statistics. It can be concluded that, as far as syntactic complexity is concerned, the text is very easy to read. However, the fact that CITR is easy to read does not necessarily mean it will be fully comprehended.



Halliday (1985) sees a text as a semantic unit, not a grammatical one, but without grammar, there is no way of ones explicit interpretation of the meaning of a text. Language, he argues, is a system for making meanings; a semantic system, with other systems for encoding the meanings it produces. He sees grammar as naturally related to meaning (p. xvii). Similarly, Hoey (2005) argues that lexis is systematically structured and that grammar is an outcome of this lexical structure (p. 1). According to Rivers (1992), readers decode print in two ways: semantically (they identify the lexical meaning of words, which exist out of context, but they also create a broader meaning for these words within the context of phrase, sentence, and discourse), and syntactically (they recognize the meaningful structural relationships within a sentence). Fluent readers rely more on semantic than syntactic information, except when meaning is not clear (p. 70). Similarly, Broughton et al. (1978) maintain that an efficient reader must have a clear understanding of the grammatical relationships which hold between the lexical items. Good readers need to grasp the semantics of a particular grammatical item in a particular context. They also need to be familiar with the precise meaning of the particular grammatical devices. The consequences of subordinating one clause to another, choosing one tense over another, or the way of relating sentences are the proper subject of the good readers attention (p. 95). The skilled reader makes use of the information, the signals, passed to him by the lexical and grammatical patterns to discover the architecture of the passage, the framework upon which it is built. (Broughton et al., 1978, p. 96) This point can be illustrated by the following example from CITR: The sea slipped back down the beach to come up and break again in a long line in the rain. Why did Hemingway use infinitive here? The meaning would not change significantly if a 55

different structure was used (The sea slipped back down the beach and came up and broke again in a long line in the rain), but authors choose their words and grammar carefully. In the classroom, the teacher can juxtapose the two sentences and ask the students if they can feel any difference in meaning. To make the task easier, the students can, for example, discuss the difference between I came home to have a rest and I came home and had a rest. Thornbury (2005) argues that the grammar that is embedded in texts is bound to be fairly representative of English grammar as a whole. In this sense, language shares a feature of other complex systems: its smallest self-standing components (its texts) are miniature representations of the system as a whole, i.e. lexico-grammar (p. 14). Thornbury (2001, 2005) argues that the teacher should develop students ability to unlock certain features of English grammar in texts because noticing is a prerequisite for learning. It is not enough to throw the chunks of language at learners; their attention must be drawn to them (Thornbury, 2012). The Noticing Hypothesis (Schmidt, 1990), proposing that input does not become intake for language learning unless it is consciously registered, has been known for some time. On the other hand, as Rothman (as cited in Schmidt, 2010, p. 5) maintains, awareness of abstract rules of grammar cannot be a prerequisite for learning, since native speakers have vast intuitive understanding of grammatical nuances that they cannot verbalize. Moreover, some advanced L2 learners also have vast intuitive knowledge that is closer to native speaker intuitions than to the simplified rules that are taught in language classes. Through exposure, they have acquired more than they have been taught. Although CITR, as an authentic literary text, was not primarily intended to display language features, it can be used as an effective vehicle for emphasizing textual features such as grammar, vocabulary use, cohesion, and discourse organization.


Corpora show that some words and phrases are predominantly used in literary fiction (e.g. suddenly, darkness, astonished, immediately, to name a few). Moreover, Nation (2001) maintains that the particular nature of the text will determine, for example, how frequent certain grammatical words are and their frequencies relative to other function words (p. 206). Thus the conjunction whereas or adverbial particle hence will be fairly frequent in an academic thesis, while they may not be encountered so often in a short story, where words such as well, oh, yes will be more common. Otherwise, it is difficult to distinguish a literary text from a non-literary one, because any seemingly non-literary utterance can be a part of literary discourse. The following pages of this chapter shall argue that CITR is suitable for intermediate L2 learners because it mainly contains language appropriate for this level of proficiency, which can be well exploited in the classroom.

Patterns in texts

The traditional approach to teaching grammar assumes that there is something the learners do not know and the teachers role is to provide them with that knowledge. Thornbury (2001) looks at grammar from a different point of view; he believes that there is something that the learners can already do, and that the teachers role is to help them to do it more effectively (p.78). Thornburys emergent view of grammar has the following implications: Work from texts and topics rather than from a structural syllabus. Generate language and then look for items and patterns. Talk to the learners and scaffold their emergent language (p. 78).


The question which needs to be addressed now is what constitutes a pattern. Thornbury (2001) sees a pattern as any regularity that has productive potential, i.e. which can act as a template for the creation of novel utterances. In Thornburys words, patterns in language cut across the traditional boundaries between vocabulary and grammar, and they are everywhere they ripple through whole texts (p.67). Thornbury (20092013) argues that if grammar is defined as something like generative multi-morpheme patterns, and if it is understood as any sequence that recurs with more than chance frequency, a quick Google search will show many more patterns in a text than the standard grammar syllabus accounts for. This gives the teacher an idea of how the text is intensely and intricately patterned (blog comment C is for Construction). Another useful tool StringNet ( an English lexico-grammatical knowledge base consisting of multiword patterns of word behaviour will provide a list of patterns in which the query word is conventionally used. One of the patterns recurring in the story is [pronoun/noun] liked the way [pronoun/noun] [verb]. Artists liked the way the palms grew. She liked the deadly serious way he received any complaints. She liked the way he wanted to serve her. She liked the way he felt about being a hotel-keeper. StringNet then provides the teacher with other patterns in which the word way frequently occurs. 1. the way in which 2. in the same way 3. the only way 4. on the way


For an intermediate Czech learner of English the word way is quite interesting, as the Czech equivalent for examples 1, 2, and 3 is different from the one in example 4. Another interesting word is face as a verb: [pronoun/noun] face the [noun] and the bright colors of the hotels facing the gardens and the sea. Their room was on the second floor facing the sea. It also faced the public garden and the war monument. And here are more patterns in which the verb frequently appears: 1. lets face it 2. be faced with 3. problem faced by [adj] [noun] Again, the verb in the chunk taken from CITR has a different meaning and a different Czech equivalent than the examples taken from StringNet. Finally, CITR is full of sentences including the existential structure there is/are/was/were, which is another problematic area for Czech learners. This structure is already included in elementary syllabuses, yet, based on my observations, students are often penalized at their Maturita exam for not using it properly. There were only two Americans stopping at the hotel. There were big palms and green benches in the public garden. In the good weather there was always an artist with his easel. There was a cat, said the American girl. The existential structure is present in some frequent collocations, which deserve attention in the intermediate classroom. 1. There is no doubt 2. There is no need


3. There is no reason 4. There is no way 5. There is no evidence This way of dealing with grammar takes the teachers (and the students) beyond the text. And the teachers can go as far as they wish. Thus one text consisting of only 1,200 words becomes a great source of language input.

Small words

Not being a highly inflected language, English makes use almost entirely of grammatical words (or function words), such as auxiliary verbs, determiners, and prepositions, in order to convey all manner of grammatical relations. According to Francis & Kuera (1982, as cited in Nation, 2001, p. 206), the approximately 270 function word types (176 word families) account for 4344% of the running words in most texts. These small words (Thornbury, 20092013) serve to make connections across stretches of text (e.g. and, so, but), to connect utterances to their context (here, now, this), and to manage speaker turns (well, oh, yes). They have the highest degree of connectivity with other words; the most frequent patterns that are formed by these connections are what we know as the grammar of the language. Thus learning about the behaviour of these words, including their constructional properties, is the key to learning the structure of English. Learners often do not notice small words and the teachers job is to make them salient (Thornbury, 20092013, blog comment S is for Small words). The Compleat Lexical Tutor (Cobb, 2008) has extracted the following list of function words from CITR (see the Figure 6). A simple cloze test (for example an on-


line one created at can draw students attention to these words.

Figure 6. The list of function words in CITR extracted by Compleat Lexical Tutor.

The reason why small words remain unnoticed may be related to the fact that shorter words have a lower semantic load. While Zipf (1936) showed that the length of a word correlated with how often it was used, and that short words were used more frequently than long ones, according to Piantadosi, Tily, and Gibson (as cited in Ball, 2011), to convey a given amount of information, it is more efficient to shorten the least informative and therefore the most predictable words, rather than the most frequent ones. After analysing word use in 11 different European languages (Czech and English being two of them), Piantadosi and colleagues found that word length more closely correlated with their information content than with how often they were used. Piantadosi et al. make the assumption that the more predictable a word is, the less informative it is, and they suggest that the relationship of word length to information content might not only make it more efficient to convey information linguistically, but also make language cognition a smoother ride for the reader or listener (cited in Ball, 2011, para 3). This would have some important pedagogical implications. Firstly, what can be a highly predictable word for a native speaker of English may be an unpredictable one for an L2 learner, who has not had so much exposure to the target language. And if English makes use almost entirely of small words, and one of the priorities of ELT is to 61

help an intermediate student of English become a fluent user of the target language, the importance of small (and short) words is evident. Obviously, a distinction must be made between a small word (i.e. a function word) and a short word. In this respect, the words pub (a content word, semantically loaded) and the (a function word, carrying little meaning) cannot be considered of one kind, even though they both consist of three letters. Nonetheless, the findings of Piantadosi et al. are relatively new, so a lot of research and interdisciplinary consideration will be needed to discover how their findings apply to ELT.


This passage will focus on two small words that are among the five most frequent words in English the articles. In CITR, there are 118 instances of the definite article, 29 instances of the indefinite article a, and three instances of an. Articles occupy more than 12 per cent of the entire text. The advantage of using an authentic text and its content as the starting point of a lesson is that the teacher does not have to be restricted by the grammar sections of the textbook. Moreover, as the examples below show, textbooks and grammar books do not always provide complete information about grammar rules. Thornbury (20092013) maintains that the definite and indefinite article must win the prize for having the most nonsense written about them in coursebooks and student grammars (blog comment A is for Articles). He gives the following example: We use the indefinite article when we are talking about a single countable noun in a general non-specific way or when we introduce it for the first time: There was a car outside the bank.


Looking at an example from CTIR, nouns such as sea, public garden, war monument, and square are mentioned for the first time, even though the definite article was used: Their room was on the second floor facing the sea. It also faced the public garden and the war monument. The motor cars were gone from the square by the war monument. It is essential to raise this problem in the class, mainly because articles are not found in Czech. Thornbury (20092013) argues that the definite article can only be taught, explained and practised in contexts that are normally larger than a sentence. It is less a grammatical item than a feature of discourse (and also of pragmatics). Similarly, Lightbow and Spada (2006) write that as the article system in English is both complex and abstract and notoriously difficult to teach, it may be better to learn about articles via exposure in the input (p. 179). From my perspective as an L2 teacher, it sometimes helps the learners if a simple rule of thumb is provided. Thus the definite article can be contrasted, for example, with determiners this and that, which do have their equivalents in Czech. The indefinite article can be explained as meaning one. However, articles remain a persistent problem for learners and require plenty of patience, exposure to the target language and corrective feedback.

Phrasal verbs

From my teaching and learning experience, phrasal verbs belong to the most problematic areas of English lexico-grammar. They are notoriously difficult for Czech learners to learn, and they are often avoided at the intermediate level of proficiency, even though they usually consist of high-frequency words.


Grammars typically claim that phrasal verbs consist of a verb and a particle (a preposition or adverb) or a verb and two particles (an adverb and a preposition). This particle adds a special meaning to the verb. They often but not always have a oneword equivalent. Some phrasal verbs take an object (transitive); others do not take an object (intransitive). They are important because they are frequent. Phrasal verbs are identified by their grammar, but it is best to think of them as individual vocabulary items, to be learned in phrases and chunks (McCarthy & ODell, 2007, p. 6). When dealing with CITR, teachers clearly do not need to overwhelm students with too much metalanguage concerning phrasal verbs. Instead, students attention could be, for example, drawn to the differences between the following examples: Italians came from a long way off to look up at the war monument. (up = adverb) vs. look up a word in a dictionary = consult a dictionary The American wife stood at the window looking out (out = adverb) vs. look out of the window (out = preposition) vs. look out = be careful The examples below are listed among the regular phrasal verbs at ( The sea slipped back down the beach The husband went on reading. lying propped up The hotel owner stood up the maid who looked after their room get back inside She sat down She went over to let my hair grow out


shut up the light had come on Whether they are all regular phrasal verbs is disputable and my teaching experience shows that finding the correct answer is not relevant to the needs of the intermediate classroom; it is more useful to encourage learners to notice them and use them in context. Another interesting feature of phrasal verbs is illustrated by the following example: She laid the mirror down. Sometimes the separable phrasal verbs only allow one choice of word order if the object is a pronoun (She laid it down). It is so because the pronoun which usually has backreference seldom ever encodes new information, hence it looks unnatural at the end of a sentence. The following rule can be found in traditional grammar books and textbooks: Anns asleep. Dont wake her up. (not wake up her) (Murphy, 1985, p. 262). She laid it down but not She laid down it. However, there are some examples when the rule is not so strict but this is not taught in traditional courses: Kim can pick up Chris, but who will pick me up? Kim can pick up Chris, but who will pick up me? Because of the principle of putting the newest piece of information at the end of the sentence, the first example sounds just as natural as the second one. (Thornbury, 2009 2013, blog comment P is for Phrasal Verb).



Gerund is another problematic feature of English lexico-grammar. Some authors, such as Tom Hutchinson, use the term gerund in their textbooks but others, for example Murphy and Hewings prefer the term -ing form. Thornbury (20092013) writes that there seems to be a whole range of -ing form uses that cover a spectrum from total verby-ness to total nouny-ness (not to mention total adjectivey-ness), and with lots of instances which are somewhere in between (blog comment G is for Gerund) . My teaching experience shows that intermediate Czech learners of English can usually use the -ing form correctly when they are encouraged to use it. When they are not pressured to use it, they will tend to avoid it and use what they assume to be more natural (for example two clauses separated by a coordinating or subordinating conjunctions and/that/which/when/while, or a noun phrase). From a grammatical point of view, the -ing form is one of the most prominent features in the story. It is advisable to draw the students attention to this grammatical area, especially because the English ing form is expressed in a different way in Czech. Various mime games, role plays and visuals (magazine pictures, photos, or Google images) can be used to present and practise the structure in the classroom. There were only two Americans stopping at the hotel. Their room was on the second floor facing the sea. Across the square in the doorway of the cafe a waiter stood looking out at the empty square. Outside right under their window a cat was crouched under one of the dripping green tables. The poor kitty is out trying to keep dry under a table.


The husband went on reading, lying propped up with the two pillows at the foot of the bed. Liking him she opened the door and looked out. You must not get wet, she smiled, speaking Italian. With the maid holding the umbrella over her George was on the bed, reading. Did you get the cat? he asked, putting the book down. Wonder where it went to, he said, resting his eyes from reading. She went over and sat in front of the mirror of the dressing table looking at herself with the hand glass. I get so tired of looking like a boy. There is one item which deserves special attention: She liked his old, heavy face and big hands. Liking him she opened the door and looked out. First, it should be stressed that in this sense, liking him is a rather unusual construction; it primarily serves cohesion, i.e. it links the previous passage to the next one, and it might have been used due to Hemingways economy of language. It is almost impossible to translate this chunk literally into Czech, but students may attempt to substitute the expression with a more common English alternative, or discover a suitable Czech equivalent, which can be carried out as a competition. Nation (2001) argues that translation has the advantage of being quick, simple, and easily understood (p. 86). However, a word of warning is necessary here; Nation maintains that the major drawback of translation is that its use may encourage other use of the first language that reduces the time available for use of the second language. Moreover, translating literary texts is an art that requires a sophisticated knowledge of the native language as well as English. According to Rivers and Temperley (1978), Translation of literary works


requires a sensitivity to nuances and subtleties of meaning, speech registers, and levels of style in English and a perceptive awareness of the flexibility and potentialities of the original language (p. 201). There is another point worth mentioning: it is at the intermediate level of English when students are constantly being told by their teachers and grammar books that the verb like is an example of a stative verb, never used in the progressive form. According to BNC, however, the form liking is not so unusual (618 hits/5.5 per million). So the grammar book can only be a generalized representation of the language system, just an approximation of the vast knowledge that speakers actually have. As a natural consequence, many rules that one can find in grammar books are inadequate, or even wrong, since they do not represent the reality of language; and this is especially true for rules in grammar books for less advanced students (Plag et al., 2007, p. 107).


As discussed earlier, reading involves matching elements of language with socially constructed meaning. According to Broughton et al. (1978), the most salient of all elements of language are words and part of what is involved in understanding a text is understanding the meanings of individual words in the text (p. 94). Laufer (1992, as cited in Nation, 2001, p. 146) was interested in determining the minimal language proficiency level where a teacher can usefully switch from concentrating on language development to the development and transfer of reading skills. Her studies have shown that the 3,000 word family level is a minimum for the reading of unsimplified text. In this thesis the term word family is used to describe the


base form of a word plus its inflected forms and derived forms made from affixes (Hirsh & Nation, 1992). Hu and Nations study (as cited in Nation, 2001, p. 147) found that for largely unassisted reading for pleasure, learners would need to know around 98% of the running words in a fiction text. With this coverage almost all learners have a chance of gaining adequate comprehension. In CITR with approximately 1,200 words, 98% coverage means about 22 unknown words. Betts (1957, as cited in Anderson, 1999, p. 117) outlined some general guidelines for teachers to use in determining if the correct books were being selected for classroom. In order to read independently, the reader must have 90% comprehension. If they do not have that 90%, they cannot actually understand all that they are reading or they cannot do it independently. If the comprehension level drops to about 75%, they have reached the instructional level or teaching level. This is the level which children use in schools if they are being taught reading. If the reader drops below that, particularly if they come down to 50% or less, they become frustrated. It is impossible for anyone to read if they understand only 50% or less, or maybe even 60%. If a person can understand when being read or spoken to, they have reached the so-called capacity level, which is also called the listening level. Based on Bettss framework, Anderson (1999) suggests a strategy for selecting appropriate reading material. When reading, students are asked to raise a finger for each new vocabulary they encounter. Depending on the length of the page, if they have raised 10 fingers, perhaps the level of difficulty is above their current level of reading and they should select something else to read (p. 117). For the sake of illustration, a small group of intermediate students aged 1415 were asked to read CITR once and highlight words they assume they do not know. The


results in Table 5 show that although the students are in the same class, their vocabulary knowledge is richly diverse. Not many objective conclusions can be drawn, though, as the results are based on students subjective feelings. Nonetheless, there are some interesting observations to be made, subjectivity being one of them. For the sake of simplicity, the participants will from now on be labelled S1, S2, S3, S4 and S5.

Table 5 Words reported by students as unknown Student 1 Student 2 Student 3 Student 4 Student 5 bowed brutto cape clipped compact crouched darn dignity dim dripped dripping easel eaves glistened hand glass knot X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X


Student 1 Student 2 Student 3 Student 4 Student 5 laid lap maid padrone piove propped up purr rubber serve shifted smooth stood (up) stroke supreme swung tempo tightened tortoise-shell waiter total number of unknown words 19 15 25 X X X X X X X X X X X X X X X X X X X 20 5 X X X X X X X X X X X X X X X X X X X X

Firstly, based on Andersons (1999) conclusion, it is apparent that S3s level of difficulty is above his current level of reading, as he raised his finger 12.5 times. This 71

calculation is, however, highly relative and speculative, as the original text in the collection is printed on three pages, while the copy with which students were working only had two pages. The original printed version consists of 114 lines, though of unequal length, which means that on average, S3 encountered an unknown word in approximately every fifth line. Secondly, S3 (with 25 unknown words) took a totally different approach to completing the task than S5 (with 5 unknown words). First of all, it should be stressed that having marked five words as unknown does not necessarily mean that these are the only words S5 does not know. This particular student is a very ambitious teenager who, even though told that she should be absolutely honest, wanted to achieve the best score. She cannot be blamed for the fact that she strived to guess the words from context to avoid having to mark them. On the other hand, S4, one of the best students in the class, highlighted words such as serve and waiter, which he knows and normally uses. Furthermore, although some of the words appear in the story more than once, it is interesting that two participants could not guess their meaning and marked them as unknown. The word maid, for example, appears in the story seven times in various contexts, and at least two sentences offer clear contextual cues: It was the maid who looked after their room. With the maid holding the umbrella over her, The students might have failed to notice the cues because they had explicitly been asked to search for unknown words. Or it may be that although they eventually could guess the meaning, they did not take their initial choice back. The case of dripping versus dripped is also worth attention. Three students marked dripped as unknown and dripping as known, or vice versa, even though the words are two forms of the same verb. But Nation (2001) points out that arriving at a


correct guess from word part analysis can be tricky, and it is less sure, for example, than using context clues (p. 257). The experiment can be taken two steps further. First, in order to minimize the discrepancies, especially in relation to the suspiciously low number of unknown words in Student 5s case, the students could be asked to translate all the unknown words collected from the five participants, and the words translated correctly would then be added to the total number of unknown words. Second, in order to present absolutely precise results, it would be necessary to ask the students to translate all the text, because there is no guarantee that the words in Table 5 are all the words the students do not know. However, this would not guarantee precise results either, as this thesis deals with a literary text, and the fact that the students cannot translate an utterance does not necessarily mean that they do not know the individual words. It is also important to mention that the story includes several Italian utterances which the students treated inconsistently. It should be stressed that none of them claims to be able to speak Italian. 1) "Ha perduto qualque cosa, Signora?" 2) "Si, il gatto" 3) "Avanti!" 4) II piove 5) Si, si Signora, brutto tempo. 6) padrone Table 5 shows that all students highlighted the word padrone, presumably under the impression that it is an English word which they do not know. S3 marked brutto tempo as one expression, while S4 only marked brutto. The same student also underlined "Ha perduto qualque cosa, Signora?"and il gatto, while the other students 73

perhaps finally arrived at the conclusion that there is no point in underlining words other than English, or they could guess the meaning. Hemingways occasional codeswitching (in this case between the native language and the foreign language) can also be well exploited in the EFL classroom. The following sets of exchanges between the Italian speaking characters and the Americans demonstrate that Hemingway gives many clues, helping the readers guess the meaning, even though they do not speak the language. 1. .the hotel owner stood up and bowed to her. Il piove, the wife said. 2. Ha perduto qualque cosa, Signora? There was a cat, said the American girl. 3. A cat? Si, il gatto. 4. Someone knocked at the door. Avanti, George said. 5. Si, si, Signora, brutto tempo. It is very bad weather. The first exchange proves to be rather problematic, as four out of five students marked bowed as unknown. That is why bowed could not be a clue for guessing the meaning of Il piove. The question mark in the second exchange indicates that the speaker is asking about something, and she gets the answer a cat. The same thing happens in exchange number 3, where the American wife speaking Italian confirms what the Italian maid speaking English is asking about it is the cat, again. When someone knocks at the door, the person inside the room usually invites him or her in. So avanti could also be guessed easily from the context and co-text. As Czech students are familiar with both tempo from Italian tempo (Online Etymology Dictionary) they can also deduce what brutto means, provided they take into consideration the co-text (the bad weather).


Vocabulary frequencies

There is a more objective way of finding out how difficult a text will be for students. This involves two steps: (1) finding out how many of the most frequent words in English the students know and (2) analysing the frequency ranges of vocabulary used in the text. Both steps will be elaborated on in the following passage. Before the students were exposed to reading CITR, they were asked to complete Nations Vocabulary Level Test (Nation, 1990). The test includes isolated words tested with multiple choice meanings and words in non-defining contexts. The tests can be used to see if learners know enough vocabulary to read (but it is important not to confuse having an adequate reading vocabulary with skill in reading). As Table 6 shows, all the students did very well in the 1,000 word level test, S2 and S5 scoring the best (39 out of 39). S2 achieved the maximum score in the 2,000 word level test, and S1 had a fairly good result in the 3,000 word level test. In comparison with S3, S2s results declined gradually throughout the test levels (with one exception in the 2,000 word level test, where he scored 100%). S3 did quite well in the University Level test, but achieved a low score in the 5,000 word level test. This might be due to the fact that the University Word Level Test includes vocabulary with which Czech students are familiar their cognates appear in school subjects such as biology or science. According to COCA The Corpus of Contemporary American English (Davies, 2008), 87% of the words in CITR are within the 3,000 frequency range. The rest of the words are outside the 3,000 range (see Figure 7). Nation (2001) writes that to gain adequate comprehension of the text, most readers would need to know 98% of the running words in the text (p. 165). The results of Nations Vocabulary Level Test imply that most of the participants could not understand CITR adequately. 75

Table 6 Nations Vocabulary Test results word level 1,000 2,000 3,000 5,000 University 10,000 S1 38 15 16 9 9 8 S2 39 18 15 10 9 5 S3 37 12 13 3 8 3 S4 34 13 12 4 7 7 S5 39 8 11 10 4 1 maximum 39 18 18 18 18 18

To help the students achieve better comprehension, the low-frequency words may be handled through pre-teaching, guessing from context or dictionary use. As dictionary use is rather time-consuming, unknown words can be glossed in texts for second language learners. Guessing from context and glossing will be discussed later in detail.

Figure 7. A screenshot of Range 3 COCA list > 3000 words


As far as vocabulary frequencies are concerned, there are some further pedagogical implications. Nation (2001) argues that when deciding which words should be taught and learned, and which should be ignored, it is important to make a distinction between high-frequency and low-frequency words. The distinction is important because teachers need to deal with these two kinds of words in different ways (p. 21). Firstly, English, like any language, has the habit of recycling a relatively small number of words over and over again, and if the learners know these high-frequency words, their reading power can be enhanced dramatically, for a relatively modest learning investment (Cobb, 2008). Secondly, teachers may need to give most attention to less frequent words in a particular text that are high-frequency words across a range of texts (Nation, 2001, p. 209). The question that needs to be addressed now is how teachers find out what the most frequent words in English are. The Longman Communication 3000 ( is a list of the 3,000 most frequent words in both spoken and written English, based on statistical analysis of the 390 million words contained in the Longman Corpus Network a group of corpora, i.e. databases of authentic English language. The Longman Communication 3000 represents the core of the English language and shows students of English which words are the most important for them to learn and study in order to communicate effectively in both speech and writing. Analysis of the Longman Corpus Network shows that these 3,000 most frequent words in spoken and written English account for 86% of the language. This means that by knowing this list of words, a learner of English is in a position to understand 86% or more of what he or she reads (which is, however, still insufficient for adequate comprehension). There is a distinction between words that are


in the top 1,000, 2,000 and 3,000 most frequent words in written English, and the top 1,000, 2,000 and 3,000 most frequent words in spoken English. The words are accompanied by special symbols S1, S2, S3 for spoken English and W1, W2, W3 for written English.

Table 7 Examples of words marked as unknown and their frequencies in Longman Communication 3000 and COCA lay v serve v smooth adj. swing v rubber n shift v stand v S1, W2/ 803 S1, W1/ 401 W3/ 2,438 W3/ 2,469 S3/ 4,836 S3, W3/ 1,686 S1, W1/ 281

Note: S stands for spoken English and W for written English. Numbers 1, 2, and 3 refer to the frequency ranges. The number following the slash refers to COCA frequencies. Table 7 shows some of the words which the students marked as unknown and their occurrences in Longman Communication 3000 and, for the sake of comparison, the frequency ranks in COCA. The lower the rank, the more frequent the word is. The words that are not listed here are outside the 3,000 frequency range in Longman Communication 3000.


The results show, for example, that the word rubber is not a very useful word for students to learn, although it is one of the first words young learners of English encounter in Get Set Go 1 (Lawday, 1995).

Guessing words from context

When reading a text, students will always need to make intelligent guesses as to the meaning of unknown words. While Harmer (1989) argues that it is often difficult to convince students that texts in English can be understood even though there are vocabulary items and structures the learner has never seen before (p. 153), my teaching experience shows that this is highly dependent on the students age; the younger the students are, the less concerned they are about unfamiliar words they encounter in texts. Thornbury (2002) argues that guessing from context is probably one of the most useful skills learners can acquire and apply both inside and outside the classroom (p. 148). Nation (2001) sees guessing from context as a sub-skill of reading which seems to draw heavily on other reading skills (p. 250). Clarke and Nation (1980, as cited in Nation, 2001, p. 257) recommend a five-step inductive procedure for guessing from context: Step 1: Decide on the part of speech of the unknown word. Step 2: Look at the immediate context of the word, simplifying it grammatically if necessary. Step 3: Look at the wider context of the word (including the adjoining clauses and sentences). Step 4: Guess.


Step 5: Check the guess. Is the guess the same part of speech as the unknown word? Does it fit into the context? Break the unknown words into parts. Does the meaning of the parts support the guess? Look up the word in the dictionary.

For illustration, here is a sentence from the story, including a word the students found problematic. In the good weather there was always an artist with his easel. Following the proposed steps, the student can see that the word easel is a noun because its immediate preceding collocate is the pronoun his and there is nothing after easel. The surrounding context can reveal that easel has something to do with the artist. Also, the expression in the good weather may be helpful. Artists usually paint outdoors when the weather is good. But there are still more alternatives as artists have brushes, palettes, canvases, etc. As the word is not repeated in the text, the student will probably have to consult a dictionary or Google images. When she talked English the maids face tightened. Here the -ed ending indicates that it is a verb. Its immediate collocate is face, so the students could guess that the verb is related to facial expressions. The word tightened is made up of tight + en +ed. Students might be familiar with tight, although they might find the meaning of tightened somewhat elusive. Also the preceding clause When she talked English can help because she refers to the Italian maid who might find speaking English rather difficult. This procedure is strongly based on language clues and does not draw on background content knowledge. Nuttall (2005) suggests, however, that we stop and go back, using a bottom-up approach, only if we discover inconsistencies. When using a top-down approach, we can make imaginative leaps from one bit to another (p. 22).


To conclude, learners need not to follow a rigid procedure when guessing, but they should be aware of the range of possible clues and should have the skill to draw on them (Nation, 2001, p. 259).


As indicated above, teachers must take care that texts do not present difficulties for the learners which would inhibit comprehending to the extent necessary to read the text effectively as discourse. If the text consists of syntactic structures and lexical items which the learner just has not got enough competence to comprehend, the teachers and learners will fail to realize its full potential. As discussed earlier in this chapter, one of the possibilities of dealing with unknown words is guessing from context. However, this can be particularly challenging for an intermediate learner of English. As Lightbow and Spada (2006) write, the research gives evidence that in order to guess successfully the meanings of new words in a text, a reader usually needs to know 90 per cent or more of the words in that text (p. 188). According to Nation (2001), it is 95%, or even 98% (p. 233). One of the things the teacher can do to help learners with comprehension, and thus make the text more suitable, is to provide glossaries. However, findings related to the relationship between glossing and reading comprehension have been inconsistent. Nation (2001) argues that the effects of glossing on comprehension and on vocabulary acquisition are mixed, but overall, research on glossing shows that it has certain attractions. Firstly, it allows texts to be used that may be too difficult for learners to read without glosses. Secondly, glossing provides accurate meanings for words that might not be guessed correctly. Thirdly, glossing provides minimal interruption of the reading


process, especially if the words appear near the words being glossed (p. 175). Nation concludes that glosses are best in the margins of the text and can be in the L1 or L2, as long as they are easily understood (p. 177). As Azari, Abdullah and Hoon (2012) suggest in their study aimed to explore the effects of different glosses on reading comprehension of low proficiency postgraduate students, and which employed marginal glosses, second language instructors should provide L2 learners with glossed texts. In this way, the readers attention is drawn to glosses, and it will result in reading comprehension. Furthermore, the provision of textual glosses reduces the burden of looking up words in a dictionary and prevents L2 learners from choosing wrong meanings for unknown words in a particular context (p. 49). Similarly, Thornbury (2005) writes that providing a glossary of difficult words in the text can help reduce the vocabulary load and save time spent consulting dictionaries (p.117). The research evidence shows that, particularly in the first meeting with a word, any explanation should not be complicated or elaborate (Nation, 2001, p. 90). Widdowson (1978) describes two kinds of glossaries: priming glossaries (explanations which precede the reading passage and their purpose is to prepare the learner beforehand for his encounter with possible problems in the text), and prompting glossaries (explanations which are linked to particular problems as the reader actually encounters them in context; they usually appear after the text and it is assumed that the readers will refer to them whenever they encounter a difficulty (p. 82). CITR is a part of the More Modern Short Stories collection selected and edited by Taylor (1981), which offers a prompting glossary appended to each short story. The following glossary is provided at the end of CITR (p. 66): Crouched lying close to the ground, protecting itself from the rain compact fitting into a small space


Il piove Italian for Its raining. Ha perduto qualque cosa, Signora? Italian for Have you lost something, Madam? padrone hotel-keeper to grow out to grow longer and fuller to get tired of to become bored with the same thing all the time pretty darn nice an American way of saying very nice Avanti Italian for Come in. tortoise-shell yellow and brown coloured According to Widdowson (1978), glosses can divide into two categories: some of them give the signification of the lexical item, its definition as a linguistic element in the language code, other explanations provide the value which the lexical items take on in this particular context (p. 84). With the signification glosses, the learner has to perform a further operation on the explanation so that it fits into the context. The learner must adjust it in some way. This adjustment may be principally syntactic. What the learner has to do is to extract the basic meaning from the explanation and apply it appropriately when the word which is glossed appears in the text. This may involve semantic as well as syntactic adjustment. Widdowson (1978) argues that if teachers decide on the signification type of gloss, they leave the learner with the problem of working out meanings from context or adjusting the explanation so that it makes appropriate sense. Also, by giving the learner a meaning in advance the teacher can give the learner the impression that reading is simply a matter of recognizing of given and fixed meanings. So glosses of this kind may encourage a mistaken attitude to the reading task (p. 85).


Widdowson (1978) goes on to argue the other side of the coin, pointing out that a value gloss leaves the learner with too little to do and deprives him of the opportunity of discovering the values of different expressions in the process of interpreting the text (p. 86). As the example below shows, most of the glosses provided are value glosses which can be transferred directly to the context with very little adjustment. Outside right under their window a cat was [crouched] = [lying close to the ground, protecting itself from the rain] under one of the dripping green tables. The cat was trying to make herself so [compact] = [fitting into a small space] that she would not be dripped on. Dont you think it would be a good idea if I let my hair [grow out] = [grow longer and fuller]? I [get tired of] it = [become so bored with (the same hair) all the time]. You look [pretty darn nice] = [very nice]. She held a big [tortoise-shell] = [yellow and brown coloured] cat So far paper glosses have been discussed. Lyman-Hager (1999) maintains that traditionally, text glosses (i.e. margin notes or footnotes that explain obscure words or phrases in a particular literary passage) are restricted to space and resources. Multimedia annotations avoid these restrictions and allow for students to easily access and identify words or phrases. This type of teaching tool expands the amount of information available to students, individualizes the learning experience by hiding the gloss until the student feels the need to access it, and permits instructors and students to have more class time to devote to the content of literary texts (p.3). Davis (as cited in Lomnicka, 1998) emphasizes that hypertext is invisible and unobtrusive, allowing the users to consult as much or as little information as they desire (p. 42). Lomnicka (1998) maintains that to achieve a more global understanding of the text, other multimedia


annotations such as images, sounds, cultural, historical and geographical references, and guiding questions could enhance comprehension (p. 42). She adds that although students are often accustomed to single word or phrase decoding, experience with multimedia annotations can help to convey the necessity of moving beyond a simple text-base comprehension (p. 49). Below there are two examples of multimedia glossing. Figure 8 shows CITR glossed with selected items. The text was created at ( After clicking the selected link, it will immediately direct the user to a visual (a picture, or video) or verbal gloss, which the user has inserted.

Figure 8. A screenshot of CITR glossed with multi-media annotations, retrieved from

Figure 9 shows a screenshot from Help For English website ( After clicking any word in the text, the user is taken to a dictionary entry including a Czech translation, an explanation in English, and a phonetic transcription. The advantage of the first type of glossing is that the teacher can choose what to gloss and how to gloss (based on the needs of the students); this type of glossing keeps different learning styles in mind. Visuals are often seen as the most valid way of 85

communicating the meaning of a word learners see an instance of the meaning, and this is likely to be remembered as the meaning is stored both linguistically and visually (Nation, 2001, p. 85). Moreover, although the underlined words are rather distracting, glossing words increases the chances of them being noticed, and thus learned (Nation, 2001, p. 72)

Figure 9. A screenshot of a glossed text retrieved from

In the other text (see Figure 9), the words are not underlined, and learners can choose which glosses they want to see. The glosses are only verbal (translation, a simple definition, and a phonetic transcription), and the user is dependent on what the website offers (at this moment, unfortunately, it does not offer CITR; it only provides the users with authentic texts at the upper-intermediate and advanced level). Translation is often being criticized as being indirect, taking time away from the second language, and encouraging the idea that there is an exact equivalence between words in L1 and L2. But Nation (1978, as cited in Nation, 2001, p. 85) points out that this criticism applies to most other ways of communicating meaning which involve the changing of an idea into some observable form; they are indirect, they are likely to be 86

misinterpreted, and they may not convey the exact underlying concept of the word. Nation maintains that there is no exact equivalence between, for example, a second language word and its second language definition (p. 86). From an L2 learner point of view, reading glosses do save time and, if designed and used effectively, they do not diminish the pleasure of reading. From a teachers perspective, glosses (especially in the form of clear, simple, and brief definitions) are a source of additional comprehensible language input which brings the intermediate learners interlanguage closer to the target language.


So far the section on vocabulary has dealt with isolated lexical items (i.e. individual words). Broughton et al. (1978) argue, however, that understanding the meanings of individual words is not the end itself. The efficient reader needs to be able to understand the patterns of relationships between words (p. 94). Lewis (2000) argues that intermediate students may know quite a lot of individual words which they struggle to use, in addition to their grammatical knowledge, but they lack the ability to use those words in a range of collocations which pack more meaning into what they say or write. Lewis maintains that most intermediate students would improve dramatically if they spent less time trying to perfect their grammar and learn new, rare words, and instead simply learning to use the words they already know in the huge number of collocations of which these words are parts (p. 14). The term collocation is used to refer to a group of words that belong together either because they commonly occur together, or because the meaning of the group is not obvious from the meaning of the parts (Nation, 2001, p. 317). According to Hoey


(1991), collocation is a psychological association between words up to four words apart and is evidenced by their co-occurrence in corpora more often than is explicable in terms of random distribution (p. 5). Ellis (as cited in Nation, 2001, p.319) sees the learning of collocation as one level of chunking. Nation (2001) argues, however, that just because collocation exists does not mean that it deserves attention. In order to decide if classroom time and effort should be spent on an item, two factors should be taken into consideration: whether the frequency of the collocation is high, and whether it occurs in many different uses of the language. Frequent collocations of frequent words also deserve attention (p. 325). Partington (1998) argues that what is frequent and normal in one kind of text may be quite unusual in another. In other words, collocational normality is dependent on genre, register and style (p. 17). Lewis (2000) examines written texts from different genres and argues that different kinds of texts exhibit different collocational characteristics, making some texts more suitable than others for the EFL classroom (p. 56). Lewis believes that there is the temptation to think that good writers do not use ready-made chunks of language: A novelist, by definition, is free to make their own word combinations in other words, to break our expectations. And it is in the breaking of the conventional that the greatness of great literature partly resides. (p. 57). CITR is generally considered a literary masterpiece and, as was shown in previous chapters, Hemingway use of the language is truly unique. In addition, CITR was first published in 1925, almost a century ago, which may substantially change the view of the authors choices of collocations and decrease the suitability of the text for the classroom. The following procedure, inspired by the process that Professor Hoey (University of Liverpool) used in his famous book Lexical Priming (2005),


demonstrates that one wording of a sentence sounds more natural, or more idiomatic, than another. James Thomas (Masaryk University) calls this activity the Hoey Procedure. By comparing several chunks from the text with corpus data, Table 8 shows that Hemingway made most collocational decisions with respect to what nowadays normally occurs in the language.

Table 8 Hoey Procedure BNC there were there were only only two two Americans stopping at the hotel stopping at the at the hotel at the they did not know did not know did not know any any of the people they passed passed on the stairs on the stairs on their way 21,246 446 2,214 30 0 15 347 151,585 272 5,160 42 24 353 1 239 665 COCA 73,791 1,532 8,311 263 1 182 1,367 652,060 444 4,902 31 126 1,682 1 656 2,903


BNC their way to and from from their room their room on the second floor facing the sea facing the room facing the sea the public garden the war monument big palms in the public garden in the good weather there was always artist with his easel 2,995 110 5 178 26 15 877 1 8 0 0 2 0 394 0

COCA 12,742 1,650 16 582 946 35 4,557 0 62 1 0 11 4 1,166 0

On the other hand, it also shows that Hemingway used some words in unique collocations. For example, as BNC shows, the collocation stay at the hotel is more typical than stop at the hotel, while the verb stop is used more frequently with collocates such as overnight in this context. In addition, the expressions the war monument and big palms are not found in BNC at all. Furthermore, the phrase in the good weather is not found in BNC, while in good weather shows five hits. Although Hemingway was an American writer, even in COCA, the collocation in good weather is more frequent (93 hits) than in the good weather (4 hits).


Hunston (2002) maintains that corpora cannot be used to determine what is impossible in a language, as they do not offer negative evidence, and they cannot even be used to determine what is possible, as a corpus may well contain utterances which the native speaker of a language would reject as incorrect. However, corpora can provide teachers and students with what is typical, such as the most frequent collocates (p. 42). Just as a lexical item may tend to co-occur with another lexical item, so it may tend to occur in or with a particular grammatical function. This phenomenon colligation is defined by Hoey (2005, p. 43) as follows: the grammatical company a word or word sequence keeps (or avoids keeping) either within its own group or at a higher rank; the grammatical functions preferred or avoided by the group in which the word or word sequence participates; the place in a sequence that a word or word sequence prefers (or avoids).

Czech intermediate learners of English may find it interesting to discuss the difference between bowed slightly and slightly bowed. Both orders are possible but XY bowed slightly (past tense form 18 hits in BNC and 61 hits in COCA) is more frequent than XY slightly bowed (past participle 5 hits in BNC and 35 hits in COCA). The same applies to tightened mouth (0 hits in BNC and 0 hits in COCA) and mouth tightened (with 50 hits in BNC and 80 hits in COCA). The words bow and tightened were marked as unknown by some of the participants of the above mentioned experiment. Their collocates slightly and mouth do not appear in CITR; they have been chosen because they frequently co-occurs with bow and tightened in other texts. As shown earlier, bow and tightened are not frequent words 91

(they are outside the 3,000 frequency range in COCA and Longman Communication 3000), but if the teacher still needs to teach a low-frequency word, it may be better done with its collocates. And as teachers wish to ease the vocabulary learning burden, it is advisable to draw attention to frequent collocates of low-frequency words. To conclude this section and return to the argument posed at the beginning of the thesis, collocation particularly increases in importance at the intermediate level of proficiency, when some automatic processing takes place. Students need more than just separate words to produce simple connected texts on topics that are of personal interest, to describe experiences, events, dreams, hopes and ambitions, and to briefly give reasons and explanations for opinions and plans.



The main purpose of this thesis was to assess the readability and exploitability of Cat in the Rain (CITR), an authentic literary text, and to support the hypothesis that it deserves attention in the intermediate English classroom. This thesis has given an account of advantages and disadvantages of authentic texts in the classroom. Although authenticity is a somewhat controversial concept, it can be concluded that authentic texts have enough advantages to justify their use in the intermediate L2 classroom. This thesis has also argued that the main disadvantage of authentic texts is their potential difficulty. The most obvious conclusion to emerge from this thesis is that teachers cannot simply choose any authentic text without considering the challenges of such a text; various top-down and bottom-up factors affecting difficulty must be taken into account, including the students background knowledge, syntactic complexity of the text, vocabulary load, to name a few. Various readability calculations support the idea that CITR is not a very difficult text to read. However, CITR is a literary text and, therefore, these calculations are only one aspect of assessing the text difficulty. Literary texts often use language imaginatively and playfully, and the authors may say one thing and mean another. Thus students will often need help and guidance to discover the multiple levels of meaning in a literary text. Research shows, for example, that students can better comprehend a text if they have appropriate schemata underlying the text. Although CITR was not written for EFL learners, there is no reason to assume that the text cannot be used for highlighting certain language features. On the contrary, since CITR is an example of a cohesive piece of text as opposed to a random


collection of sentences the teacher can develop students ability to focus on certain representative features of English vocabulary use, cohesion, and grammar. Returning to the hypothesis posed at the beginning of this thesis, it is now possible to state that CITR is a suitable text for an intermediate L2 classroom because it is comprehensible and exploitable. The thesis has gone some way towards enhancing our understanding of what an important role full texts, especially authentic ones, play in the classroom. The thesis has only examined a literary genre, but the findings are transferable to other text types as well.



Anderson, N. J. (1999). Exploring second language reading: Issues and strategies. Boston: Heinle & Heinle. Azari F., Abdullah, F.-S., & Hoon, T. B. (2012). Effects of textual glosses on reading comprehension of low proficiency EFL postgraduates. The English Teacher, XLI (1). Retrieved from Ball, P. (2011). How words get the message across. Nature. Retrieved from: Barnett, M. A. (1989). More than meets the eye: Foreign language reading. Language and education: Theory and practice. Engelwood Cliffs : Prentice Hall Regents. Betts, E.A. (1957). Foundation of Reading Instruction. New York: American Book. Brennen, C. F. (2006). Hemingways cats. Sarasota, FL: Pineapple Press. British National Corpus. (n.d.). Retrieved from Broughton, G., Brumfit, C., Flavel, R., Hill, P. & Pincas, A. (1978). Teaching English as a foreign language. London: Routledge & Kegan Paul. Brown, Douglas H. (2007). Teaching by principles: An interactive approach to language pedagogy. White Plains: Pearson Education. Carrell, P. L. (1987). Readability in ESL. Reading in a foreign language, 4 (1), 21 40. Retrieved from Carver, R. P. (1969). Effect of a "chunked" typography upon reading rate and comprehension. Washington, D.C.: American Psychological Association. Clarke, D. F. & Nation, I. S. P. (1980). Guessing the meanings of words from context: Strategy and techniques. System, 8, 211220.


Cobb, T. (2008). The Compleat Lexical Tutor. Retrieved from Council of Europe (2011). Common European framework of reference for language. Council of Europe. Retrieved from Crossley, S., Louwerse M., McCarthy, P., & McNamara, D. (2007). A linguistic analysis of simplified and authentic texts. The Modern Language Journal, 91 (1), 1530. Retrieved from lified_and_authentic_texts.pdf Davies, M. (2008). The Corpus of American English. Retrieved from Davis, J. N. (1989). Facilitating effects of marginal glosses on foreign language reading. The Modern Language Journal, 73 (1), 4148. Doff, A. (1991). Teach English: A training course for teachers. Oxford: Oxford University Press. Duff, A. & Maley, A. (1992). Literature. Oxford: Oxford University Press. Ehrich, J. F. (2006). Vygotskian inner speech and the reading process. Australian Journal of Educational & Developmental Psychology, 6, 1225. Retrieved from AJEDP/Vol%206/v6-ehrich.pdf Ellis, N. C. (2001). Memory for language. In P. Robinson (Ed.), Cognition and second language instruction. Cambridge: Cambridge University Press. Favreau, M. & Segalowitz, N. S. (1983). Automatic and controlled processes in the first and second language reading of fluent bilinguals. Memory and Cognition, 11 (1), 5674.


Feinberg, J. (2009). Wordle. Retrieved from Francis, W. N. & Kuera, H. (1982). Frequency analysis of English usage. Boston: Houghton Mifflin Company. Halliday, M. A. K. & Hasan, R. (1976). Cohesion in English. London: Longman. Halliday, M. A. K. (1985). An introduction to functional grammar. London: Edward Arnold. Harmer, J. (1989). The practice of English language teaching. London: Longman. Hasbn, L. H. (2005). The effect of explicit vocabulary teaching on vocabulary acquisition and attitude towards reading. Actualidades Investigativas En Educacion, 5 (2), 14094703, Retrieved from Haynes, J. (19982009). Bloom's taxonomy and English language learners. Retrieved from 02.php Hedgcock, J. S. & Ferris, D. R. (2009). Teaching readers of English: Students, texts, and contexts. New York: Routledge. Hirsh, D. & Nation, P. (1992). What vocabulary size is needed to read unsimplified texts for pleasure. Reading in a Foreign Language, 8, 68996. Retrieved from Hoey, M. (1991). Pattrens of lexis in texts. Oxford: Oxford University Press. Hoey, M. (2005). Lexical priming : A new theory of words and language. New York: Routledge. Hu, M. & Nation, P. (in press). Vocabulary density and reading comprehension. Reading in a Foreign Language.


Huang, Q. (2009). English reading base on social constructivist approach. Canadian Center of Science and Education, 174176. Retrieved from Hunston, S. (2002). Corpora in applied linguistics. Cambridge: Cambridge University Press. Hutchinson, T. (1999). New Hotline. Oxford: Oxford University Press. Hutchinson, T. (2000). Project 3. Oxford: Oxford University Press. Kilgariff, A. (n.d.). Sketch Engine. Retrieved from, accessible from Kramsch, C. (1996). Context and culture in language teaching. Oxford: Oxford University Press. Krashen, S. (1981). Second language acquisition and second language learning. Oxford: Pergamon Press. Krashen, S. (1985). The input hypothesis: Issues and implications. London: Longman. Lakoff, G. & Johnson, M. (1980). Metaphors we live by. Chicago: University of Chicago Press. Laufer, B. (1992). How much lexis is needed for reading comprehension. In Arnaud and Bjoint. Lawday, C. (1995). Get Set Go 1. Oxford: Oxford University Press. Lazar, G. (2007). Literature and language teaching: A guide for teachers and trainers. Cambridge: Oxford University Press. Lewis, M. & Hill, J. (1992). Practical techiques for language teaching. Hove: Language Teaching Publications. Lewis, M. (Ed.). (2000). Teaching collocation further developments in the lexical approach. Hove: LTP Language teaching publications. 98

Lightbown, P. M. & Spada, N. (2006). How languages are learned. Oxford: Oxford University Press. Lomnicka, L. L. (1998). To gloss or not to gloss: An investigation of reading comprehension online. Language Learning & Technology, 1 (2), 4150. Retrieved from Lyman-Hager, M. A. (1999). GALT: Glossing authentic language texts. Language Acquisition. Resource Center (LARC), San Diego State University. Marsh, D. (n.d.). Using languages to learn and learning to use languages. Retrieved from McCarthy, M. and ODell, F. (2007). English phrasal verbs in use Advanced. Cambridge: Cambridge University Press. McNamara, D.S., Louwerse, M.M., Cai, Z., & Graesser, A. (2005, January 1). CohMetrix version 1.4. Retrieved from http// Murphy, R. (1985). English grammar in use. Cambridge: Cambridge University Press. Nation, I. S. P. (1978). Translation and the teaching of meaning: some techniques. ELT Journal, 32, 171175. Nation, I. S. P. (1990). Teaching and Learning Vocabulary. New York: Newbury House. Nation, I. S. P. (2001). Learning vocabulary in another language. Cambridge: Cambridge University Press. Naunton, J. (2005). ProFile intermediate. Oxford: Oxford University Press. Neslon, M. A. (2011). Using Bloom's taxonomy for reading comprehension. Retrieved from Nunan, D. (1988). Syllabus design. Oxford: Oxford University Press.


Nuttall, C. (2005). Teaching reading skills in a foreign language. Oxford: Macmillan. Partington, A. (1998). Patterns and meanings: Using corpora for English language research and teaching. John Benjamins Publishing. Piantadosi, S. T., Tily, H. & Gibson, E. (2011). Proc. Natl Acad. Sci. USA doi:10.1073/pnas.1012551108 Plag, I., Braun, M., Lappe, S., & Schramm, M. (2007). Introduction to English linguistics. Berlin: Mouton de Gruyter. Pressley, M. (2001). Comprehension instruction: What makes sense now, what might make sense soon. Reading Online, 5 (2). Retrieved from Readability Formulas. (n. d.). [Graph illustration Average Sentence Length]. Retreived from Readability Formulas. (n. d.). [Graph illustration Readability Consensus]. Retrieved from Readability Formulas. (n. d.). [Graph illustration U.S. Grade Level Conversion Table]. Retreived from Richards, J. C. (2006). Communicative language teaching today. New York: Cambridge University Press. Rivers, W. M. (1968). Teaching foreign-language skills. The University of Chicago Press. Rivers, W. M. (1992). Interactive language teaching. Cambridge: Cambridge University Press. Rivers, W. M. & Temperley, M. S. (1978). A practical guide to the teaching of English as a second or foreign language. Oxford: Oxford University Press.


Rothman, J. (2008). Aspect selection in adult L2 Spanish and the competing systems hypothesis. John Benjamins Publishing Company. Retrieved from 20in%20Contrast.pdf Schmidt, R. (1990). The role of consciousness in second language learning. Oxford: Oxford University Press. Schmidt, R. (2010). Attention, awareness, and individual differences in language learning. In W.M. Chan, S, Chi, K.N. Cin, J. Istanto, M. Nagami, J. W. Sew, T. Suthiwan, & I. Walker (Eds.), Proceedings of CLaSIC 2010 (pp. 721737), Singapore: Centre for language studies. Selinker, L. (1972). Interlanguage. International Review of Applied Linguistics in Language Teaching 10(14), pp. 209232. Simpson, P. & Mayr, A. (2010). Language and power: A resource book for students. London: Routledge. Sinclair, J.M. (ed.) (1987). Collins Cobuild English Language Dictionary. London: Harper Collins. Soars, J. & Soars, L. (1986). Headway intermediate. Oxford: Oxford University Press. Stoddard, G. D. (1929). An experiment in verbal learning. Journal of Educational Psychology, 20, 452457. Swain, M. (1985). Communicative competence: some roles of comprehensible input and comprehensible output in its development. In Glass and Madden, 235253. Swan, M. (2006). Chunks in the classroom: Lets not go overboard. Michael Swans Language and Poetry Website. Retrieved from


Talbot, M. (1995). A synthetic sisterhood: False friends in a teenage magazine. In: K. Hall and M. Bucholtz (eds) Gender Articulated: Language and the Socially Constructed Self. New York: Routledge, 14365. Taylor, P. J. W. (Ed.). (1981). More modern short stories. Oxford: Oxford University Press. Tempo (n.d.). In Online Etymology Dictionary. Retrieved from archmode=none Thornbury, S. (2001). Uncovering grammar. Oxford: Macmillan. Thornbury, S. (2002). How to teach vocabulary. Essex: Longman. Thornbury, S. (2005). Beyond the sentence. Oxford: Macmillan. Thornbury, S. (20092013). An A-Z of ELT. [Blog comments]. Retrieved from Thornbury, S. (2012). The ABC of SLA. Pecha Kucha KOTESOL. Retrieved from Tsai, H.-F. & Wilkinson, I. A. G. (n.d.). Why should discussion affect reading comprehension? An analysis of theoretical frameworks. Retrieved from rence_0614/Hsiao-Feng%20Tsai.pdf Van Lier, L. (1996). Interaction in the language curriculum: Awareness, autonomy, and authenticity. New York: Longman. Vygotsky, L. S. (1986). (Original work published 1934). Thought and language (A. Kozulin, Trans.). Cambridge, MA: The MIT Press.


Waring, R. (1997). A comparison of the receptive and productive vocabulary sizes of some second language learners. Immaculata (Notre Dame Seishin University, Okayama), 1, 5368. Wible, D. & Tsao, N.-L. (2012, May 16). StringNet 3.0. Retrieved from Widdowson, H. G. (1978). Teaching language as communication. Oxford: Oxford University Press. Widdowson, H. G. (1990). Aspects of language teaching. Oxford: Oxford University Press. Zipf, G. (1936). The psychobiology of language. London: Routledge.



Hlavnm clem diplomov prce je posoudit vhodnost autentickho literrnho textu povdky Cat in the Rain od Ernesta Hemingwaye pro vuku anglitiny na stedn pokroil rovni znalosti jazyka. Vhodnost m autorka na mysli to, e textu je mono na tto jazykov rovni dobe porozumt a e ho lze vyut k rznm pedagogickm elm. Autorce jde tak o to poukzat na dleitost souvislho jazykovho celku ve vuce cizho jazyka. Pestoe je autentinost vukovch materil v hodinch anglitiny ponkud kontroverzn tma, autorka dochz k zvru, e existuje dostaten mnotv dvod tyto materily ve vuce pouvat. Za hlavn nevhodu autentickch text je povaovna jejich nronost, proto nelze pracovat s jakmkoli textem; pi vbru je poteba zvit mnoho hledisek, kter jsou zde podrobn popsna. V prci je provedeno tak nkolik matematickch vpot, kter dokazuj, e pro stedn pokroilho studenta anglitiny nen vybran text ze syntaktickho hlediska pli nron. Ponvad jde ale o literrn text, jsou tyto vpoty pouze jednm z faktor ovlivujcch jeho nronost. Studenti budou potebovat pedagogovu pomoc a veden, aby byli nsledn schopni odhalit mnohoznanost tohoto dla. Zvrem lze ci, e akoliv povdka Cat in the Rain nebyla pvodn napsna pro studenty cizho jazyka, nen dvod pedpokldat, e ji nelze vyut pro jazykov ely. Prv naopak, jeliko se jedn o ukzku souvislho jazykovho celku, lze tento text vyut pro rozvjen schopnosti student vmat si uritch jazykovch aspekt, jako napklad pouit slovn zsoby, prvk jazykov nvaznosti i gramatickch struktur.



The main purpose of this thesis is to assess the readability and exploitability of Cat in the Rain (CITR) a short story written by Ernest Hemingway and to support the hypothesis that it deserves attention in the intermediate English classroom. It also demonstrates the importance of including full authentic texts in ELT. This thesis gives an account of advantages and disadvantages of authentic texts in the classroom. Although authenticity is a somewhat controversial concept, it can be concluded that authentic texts have enough advantages to justify their use in the intermediate L2 classroom. It is argued that the main disadvantage of authentic texts is their potential difficulty. That is why the most obvious conclusion to emerge from this thesis is that teachers cannot simply choose any authentic text without considering the challenges of such a text; various top-down and bottom-up factors affecting difficulty must be taken into account. Various readability calculations support the idea that CITR is not a very difficult text to read. However, CITR is a literary text and therefore, these calculations are only one aspect of assessing the text difficulty; students will often need help and guidance to discover the multiple levels of meaning in a literary text. In conclusion, although CITR was not written for EFL learners, there is no reason to assume that the text cannot be used for highlighting certain language features. On the contrary, since CITR is an example of a cohesive piece of text, the teacher can develop students ability to focus on certain representative features of English vocabulary use, cohesion, and grammar.