Reading and Listening Comprehension Subskills: The Match between Theory, Coursebooks, and Language Proficiency Tests

Different theoretical and empirical taxonomies of reading and listening comprehension (RC, LC) are available in the literature. Most of language tests and tasks in English as foreign or second language (EFL/ESL) coursebooks are based on the classifications of reading and listening subskills (micro-skills) offered in theory. However, these taxonomies have not cross-checked whether the theoretical subskills are practiced in ESL/EFL coursebooks and assessed in proficiency tests. Nor have they listed the shared and exclusive RC and LC subskills in a unified fashion. For this purpose, theoretical subskills offered in Applied Linguistics literature, nine internationally popular EFL/ESL proficiency tests and 25 widely used coursebook tasks were collected, cross-compared, repetitions were eliminated, and a final inclusive list of common and exclusive subskills was prepared. The findings suggested ten common reading and listening subskills. The subskills exclusive to reading were seven, and exclusive to listening were four. This list is hoped to be helpful for teachers in developing their own tests, and the coursebook developers in preparing the content materials.


INTRODUCTION
Language ability is manifested through reading and listening skills. These are also called receptive skills comprising different and similar processes. There are modality unspecific and modality-specific perspectives explaining the comprehension ability (Schroeders, Wilhelm & Bucholtz, 2010). According to a modality unspecific or single skill view, a unique single factor can explain the comprehension, implying that reading comprehension (RC) and listening comprehension (LC) may comprise the same cognitive processes. On the other hand, a modality-specific or multiple skill view posits that various factors can explain RC and LC with some differences and some similarities. This dichotomy also corresponds with the unitary and dual comprehension models (Lund, 1991). The unitary model posits that a single comprehension process functions for reading and listening, while the dual model holds that there are both differences and similarities between modalities.
In terms of differences, it is believed that listening is more cognitively demanding than reading (Buck, 2001) because: (a) it requires more attention to sounds and prosodic features, (b) it occurs in real time, and listeners have less control on the input, and do not have the opportunity to check, (c) speech is unplanned having pauses, false starts, hesittions and so forth (Vandergrift, 2006;Wagner, 2013), (d) it has different speeds of input, use of cognates, reductions, blending of sounds and back-channel cues, (e) speech is shorter than written units with vaguer and more colloquial language (i.e., there are more pronouns, redundancies, fillers, self-corrections, less standard grammar in speech, and conjunctions are used instead of subordination), and (f) meaning is conveyed by gestures and body language in speech (Brown, 2011). Regarding the similarities, both receptive skills involve decoding and comprehension using language and background knowledge. As Alderson (2000) suggested, reading is a cognitive problem-solving activity whereby it is also applicable to listening. Both LC and RC utilize similar cognitive processes (bottom-up, top-down, and integrative) (Hirai, 1999;Powers, 2013), and abilities (Spies, 2011). Also, it is argued that receptive skills may share common elements such as vocabulary, sentence patterns, idea organization, adjustment to the language function (Hollingsworth, 1968), or similar features utilizing the skills such as understanding, comprehending, analyzing, synthesizing, interpreting, and evaluating the input (Emiroğlu & Pınar, 2013). In other words, both RC and LC can have similarities or shared traits, as both are receptive skills. Nevertheless, to some extent, they are also different in that they have different cognitive loads and linguistics features. Reading and Listening Comprehension Subskills: The Match between Theory,Coursebooks,and Language Proficiency Tests 167 The picture becomes more complex because there are various lists of subskills, some of which could be called 'armchair' perspectives (Alderson, 2000). Different language proficiency tests and language coursebooks are built on language skills. Therefore, it is important to investigate what kind of reading and listening subskills are included in coursebooks, and what subskills are tested in language proficiency tests. It is hoped that examining the subskills in these sources would yield a more inclusive result. Moreover, it would elicit preliminary ideas on whether those subskills are shared or unique. It is hoped that the collective list would help teachers in developing their own tests, preparing lesson materials, and the coursebook designers working on the content of the materials. Motivated by this perspective, the aim of this study is to investigate reading and listening subskills across theories, language proficiency tests, and language coursebooks.

LITERATURE REVIEW
The reading literature has suggested different reading models including the componential and process models (Urquhart & Weir, 1998). These models also correspond with the 'reading processes and 'reading product' identified by Alderson (2000). The componential model describes reading in terms of certain factors, while the process model may explain how these factors operate. The process model deals with the reader-text interaction involving top-down, bottom-up and interactional models, while the componential model focuses on the reading product and skills or knowledge involved in that product.
The literature predominantly suggests that reading is at least two-dimensional (Urquhart & Weir, 1998) although, in the current era, this premise is superseded by multidimensional models. The review of the literature showed that applied linguists explained reading in terms of its (a) components, (b) types, and (c) subskills. As summarized in Koda (2005), different componential models have been put forth, if not all empirically validated, including the two-component model (i.e., lower level decoding and higher-level linguistic comprehension); the three-component model (i.e., conceptual abilities, process strategies and background knowledge), and another three-componential model (i.e., language, literacy, and background knowledge). To explain how learners understand large amounts of texts, different reading types, including careful and expeditious reading at global and local levels have been suggested (Unaldi, 2004;Urquhart & Weir, 1998;Weir, 2005). Finally, there is a growing number of studies explaining reading in terms of its underlying subskills. In this research, reading was explained by splitting this process into component skills (Grabe, 1991), or interrelated and interdependent subskills, which can also be applied to listening, speaking and writing (Cummins, 2014).
The listening literature shows that LC has been distinguished under the 'general' and 'comprehension' models by Aryadoust (2013). The general model is mainly based on L1 cognitive psychology studies representing listening processes, including bottom-up, top-down, and interactive processes. Like RC, the role of both approaches in listening input processing is emphasized since LC is an outcome of communication between linguistic and general knowledge (Buck, 2011). Moreover, a combination of these approaches is important to validate listening ability tests (Rost, 2011). This is because it fits individual learning styles (Flowerdew & Miller, 2005), and is necessary for listening skill curriculum for effective teaching (Hinkel, 2006). Similar to RC, LC is also accepted as a multidimensional trait containing multiple divisible constituents (Aryadoust, 2013). Consequently, multiple dimensions lead to a varying number of subskills reflected in the aforementioned speculated and research-based taxonomies.
Although there is no consensus on the number of subskills, the multi-divisibility of reading and listening has most intrigued the researchers leading to a varying number of subskills. Subskills have been proposed either as a result of qualitative and quantitative empirical investigations (e.g., Buck, Tatsuoka & Kostin, 1997;Eom, 2006;Goh & Aryadoust, 2015;Jang, 2005;Kim, 2011;Powers, 1985;Song, 2008), or theories (Brown, 2004;Hughes, 2003;Richards, 1983;Weir, 1993). Literature shows that Northern American researchers (Eom, 2006;Jang, 2005;Kim, 2011;Song, 2008; and others) explained the multidimensionality in terms of particular subskills or language attributes, while British researchers and followers (Khalifa & Weir, 2009;Shiotsu, 2010;Weir, Huizhong & Yang, 2000) have mostly premised the subskills on classification of reading types classified by Urquhart and Weir (1998). Regardless of the definitions and techniques, it can be concluded that the types, abilities, skills, micro-skills explain the construct of reading (Weir, Hawkey, Green, Unaldi & Devi, 2009) and listening. Therefore, it is important to arrive at a complex picture of the subskills from pedagogic and theoretical perspectives. Considering that the majority of language tests and tasks in language coursebooks are based on the classification of language skills offered in applied linguistics, it is necessary to arrive at a unifying picture by identifying the match between these skills mentioned in different sources. The study aims to answer: What are the subskills shared between and exclusive to reading and listening comprehension mentioned in applied linguistics theory and applied to language tests and coursebooks?

METHOD
Language pedagogy, tests, and theories are interrelated; therefore, it is necessary to identify the subskills included in these sources to understand the complexity of them. First, the subskills mentioned in the literature of applied linguistics were checked. These skills were mentioned in theoretical and empirical papers (for details, see Richards, 1983;Buck & Tatsuoka, 1998;Hughes, 2003;Brown, 2004;Eom, 2006;Kim, 2011 Academic English Language Assessment (CAEL), MELAB (Michigan English Language Assessment Battery) and Examination for the Certificate of Competency in English (ECCE)) were analyzed, and subskills attempted in these tests were listed (Appendix 1). These tests were consulted because they are popular international tests.
Third, the tasks in five mostly used EFL/ESL coursebooks published by the UK and US publishers, including "New English File", "Face2Face, "New Inside Out", "Outcomes", "Language Leader" with all available levels were listed (Appendix 2 & 3). These books were chosen considering that they are the most widely used series at universities to train the students in English. Only the tasks (exercises) under reading and listening sections were checked, and the ones under different sections, such as pronunciation or grammar were not included in the list.
Fourth, the most frequent RC and LC subskills documented in applied linguistics literature, EFL/ESL proficiency tests, and coursebook tasks were also tallied. Finally, the items in these taxonomies were cross-checked, repetitions were eliminated, and a list of subskills shared between RC and LC and those exclusive to RC and LC were prepared ( Table 1). These lists were also cross-checked and agreed by other 2 experienced EFL teachers with a Master's degree in ELT.

RESULTS AND DISCUSSION
This study attempted to prepare a taxonomy of the subskills shared between and exclusive to RC and LC. For this purpose, the subskills documented in Applied Linguistics theories, applied to different tests, and practiced in coursebooks were listed, and a final conclusive taxonomy was prepared. As shown in Table 1, the agreed number of common subskills was ten, while subskills unique to reading were seven. The subskills unique to listening were four. This study is in line with previous empirical findings in that RC and LC are divisible containing several subskills or micro-skills (Eom, 2006;Goh & Aryadoust, 2015;Kim, 2011 and others). Yet, this study can be considered unique as it attempted to offer common and exclusive RC and LC subskills.
Research shows that EFL teachers, test developers, syllabus designers have benefited from multidimensional employment of reading ability to design exercises and test tasks (Grabe, 1991;Jordan, 1997;Urquhart & Weir, 1998;Weir, Huizhong, & Yan, 2000;Khalifa & Weir, 2009). Therefore, this taxonomy could be helpful for teachers and material developers because this inclusive list offers operationalized subskills. Moreover, because subskills are helpful for diagnostic purposes (Alderson, 2005;Field, 1998;Goh & Aryadoust, 2015), this taxonomy may help teachers or teacher-researchers to diagnose the learners' strengths and weaknesses on subskills. To add further, since students themselves view their academic reading as multidimensional (Weir et al., 2009), this taxonomy may help researchers or teachers to design self-assessment tests or tools based on the subskills provided here.
To follow up this study, it is necessary to design RC and LC tests with the offered subskills. It is also necessary to observe the students while answering test questions on the measured subskills. The test scores, students' think-aloud protocols and the interviews could further explain the dimensionality of the subskills. This would also explain learners' performance in subskills common to both RC and LC.

LIMITATIONS
Nevertheless, this study has some limitations. Subskills in diagnostic tests could have been investigated as well. The study's purpose was to investigate the subskills in reading and listening comprehension. For this purpose, subskills in theory, language proficiency tests and coursebooks were listed. After rigorous analyses of the data, a list of conclusive subskills was prepared. In the next stage, the taxonomy was refined further to investigate the common and exclusive subskills. The results showed that RC and LC are not unitary and both consist of similar and different subskills. This taxonomy offers a list of operational subskills; therefore, it may be helpful for teachers and content developers to design tests and prepare course materials.