Download article
Bio-data: Moisés Damián Perales-Escudero, Paula Busseniers Elsen, María del Rosario Reyes Cruz
Vol. 41 No. 4, 2017
Published: November, 2017
ISSN: 2395-9908

Variation in Pre-service EFL Teachers’ Implicit Theories of Reading: A Qualitative Study*
Moisés Damián Perales-Escudero1, Paula Busseniers Elsen2, María del Rosario Reyes Cruz1
Universidad de Quintana Roo, Chetumal, Quintana Roo, México, Universidad Veracruzana, Xalapa, Veracruz, México
* This is a refereed paper. Received: March 9, 2017; Accepted: July 28, 2017
Abstract: This study examined variation in pre-service EFL teachers’ implicit theories of reading and metatextual knowledge in connection with a transition from a previous curriculum to a new one in order to probe for variation in these theories, or reading models, according to contextual factors. The relevant changes in the curriculum were the elimination of the previous curriculum’s rigid sequences of electives in the program’s three areas of specialization (teaching, translation, and literature) and the disappearance of the restriction to take courses in only one area. We used Hernández’s (2008) implicit theories and applied his questionnaire to 60 students in two cohorts about to graduate from an ELT undergraduate program, 30 studying under the previous curriculum and 30 studying under the new curriculum. Ten participants in each cohort belonged to one of those areas. We analyzed the data using directed qualitative analysis and the constant comparative method. The data indicate that reading beliefs appear to vary with the different curricula. The three theories identified by Hernández (2008) were present in our sample, but we also found mixed models of reading, different forms of metatextual knowledge, and evidence for a new, constructive-rhetorical implicit theory of reading. Overall, most trainees under both curricula subscribe to the least advanced models, which suggests a need for further changes to the curricula and to reading instruction training.
Keywords: beliefs about reading, EFL, implicit theories of reading, metatextual knowledge, pre-service teachers***conocimiento metatextual, creencias sobre la lectura, inglés como lengua extranjera, profesores en formación, teorías implícitas de lectura

Resumen: Examinamos las teorías implícitas de lectura de un grupo de profesores en capacitación en una universidad pública mexicana en relación con diferentes planes de estudio y áreas de especialización con el fin de explorar la posible existencia de variaciones relacionadas con estos factores. El cambio relevante en el nuevo plan de estudios con respecto al anterior fue la eliminación de secuencias rígidas de materias optativas en las tres áreas de especialización del programa (docencia, traducción y literatura) y la prohibición de tomar cursos de otras áreas. Usamos las teorías implícitas de Hernández (2008) y aplicamos su cuestionario a 60 estudiantes de dos cohortes próximas a egresar. 30 de ellos estudiaron con un plan de estudios anterior y 30 con uno nuevo. Diez participantes por cohorte pertenecían a alguna de estas áreas de especialización: docencia, literatura o traducción. Analizamos los datos usando el análisis cualitativo dirigido y el método constante comparativo. Encontramos que los estudiantes se adhieren a las tres teorías implícitas postuladas por Hernández (2008), pero también encontramos modelos de lectura mixtos, otras formas de conocimiento metatextual y evidencia de una teoría implícita nueva, de tipo constructivo-retórico. Asimismo, parece que las creencias sobre lectura varían de acuerdo al plan de estudios. Una mayoría de los estudiantes de ambos planes de estudios se adhieren a los modelos menos avanzados, lo cual sugiere la necesidad de realizar nuevos cambios al currículum y a las prácticas de preparación para la enseñanza de la comprensión de lectura.


Beliefs about language learning and teaching have been an object of research since at least the 1980s (Barcelos, 2007). Research on beliefs is important as beliefs appear to mediate many of our actions, even if they do so in non-causal, non-linear ways (Navarro & Thornton, 2011). As Pajarés (1992) puts it, beliefs may have a filtering effect that “ultimately screens, redefines, distorts, or reshapes subsequent thinking and information processing” (p. 324). It follows that teachers’ beliefs may have an impact on their classroom behavior and possibly on student learning.

In line with this reasoning, an area of inquiry in the study of language learning and teaching beliefs has focused on investigating ESL/EFL teachers’ beliefs about various aspects of language teaching, such as grammar, the status of English, the learning process, or the teaching process in general. There has been less focus on ESL/EFL teachers’ beliefs about reading, especially in Mexico. Research in this area is important for a number of reasons. First, English text is a significant form of input for acquisition, perhaps the most significant one in some EFL learning situations (Grabe, 2009; Krashen 2003; Sandoval-Cruz & Perales-Escudero, 2012). Second, there is evidence that better L2 comprehenders are also better at learning L2 vocabulary (Pulido, 2009). Third, reading for understanding and learning is an important part of literacy-focused ELT approaches, such as English for Academic Purposes (EAP) or Content and Language Integrated Learning. Fourth, there is some empirical evidence that beliefs held by EFL teachers have an impact on how they treat texts for pedagogical purposes and, potentially, on students’ opportunities to learn content, language, and literacy skills (Bamanger & Gashan, 2014). This potential impact on classroom teaching and learning makes pre-service teachers’ beliefs worth investigating. However, only a handful of studies have focused on teachers’ beliefs about reading.

Chou (2008) investigated the construct of teachers’ belief systems about reading approaches among 42 university instructors and explored discrepancies or consistencies between teachers’ beliefs about reading theories and their practical teaching activities in Taiwan. The findings showed that there were no significant differences between the participants’ beliefs and their use of each reading approach.

Kuborzka (2011) conducted a study of six EAP teachers in a Lithuanian university. Through a combination of several observation and elicitation techniques that included analysis of syllabi, video stimulated recall, and interviews, she concluded that the teachers in her sample held a group of beliefs that she characterized as a skills-based approach to reading. This approach views reading as a decoding process, one whose fundamentals were learned before entering the university. Therefore, when working with texts, the teachers did not focus on discourse comprehension or metacognitive strategies, but on vocabulary, grammar, pronunciation, and translation.

Khonamri and Salimi (2012) investigated the beliefs about reading strategies of 57 high school teachers of EFL in Iran. They used Chou’s (2008) questionnaire to research these teachers’ beliefs about the importance of reading strategies, the need for reading strategies in the classroom, and the implementation of reading strategies. Their quantitative analysis showed a positive correlation between beliefs about the need for reading strategies and implementation, but no correlation between beliefs about the importance of reading and actual implementation. The authors believe this discrepancy is a result of insufficient training in reading strategies and the teachers’ desire to present themselves in a positive light. Interestingly, unlike the teachers in Kuborzka (2011), these teachers accorded the greatest importance to metacognitive strategies, although they seemed not to know how to implement them.

Bamanger and Gashan (2014) investigated Saudi EFL teachers’ beliefs concerning the significance of teaching reading strategies and the influence of those beliefs on their actual practices. Their participants considered vocabulary-focused strategies to be the most important, followed by scanning and asking questions. There was a high, statistically significant, correlation between the teachers’ beliefs and their practices as observed by the authors.

Moore and Narciso (2011), the only Mexican EFL-focused study we found, investigated Mexican undergraduate English majors’ beliefs about reading. The participants in the study were training to be either EFL teachers or translators. The researchers drew on the classification of implicit models of reading put forward by Schraw and Brunning (1996). They also investigated the connections between their participants’ implicit models of reading and their reading frequency. Through qualitative analysis of interviews, they found that a higher reading frequency is associated with a transition from a translation model to a transmission model of reading, that is, one that remains close to the text’s encoded meaning. This finding runs counter to the sequence implicit in the model (from transmission to translation to transaction). Their data also suggested the existence of transition models that merit further research.

A study not focusing on EFL teachers, but relevant to the research reported here, is Hernández (2008). Like Moore and Narciso (2011), Hernández (2008) also took an implicit models, or implicit theories, perspective to the study of beliefs about reading. Taking Schraw and Brunnings’ (1996) taxonomy of implicit models and a sociocultural perspective on beliefs as his starting point, Hernández (2008) created a qualitative questionnaire to investigate the implicit theories of reading and metatextual knowledge held by junior high school, senior high school, and college students in two different majors: literature and chemistry. From the analysis of student interviews, he proposed three implicit theories—receptive, interpretive, and constructive—that are, in his words, roughly equivalent to Schraw and Brunning’s (1996) three implicit models (“… coinciden con los trabajos realizados por R. Schraw [Schraw y Brunning, 1996]” (p. 749). He found a progression in accordance with the participants’ educational level, from receptive reading in lower levels to constructive reading in higher levels. He also found differences across majors, with literature students showing a greater tendency to subscribe to a constructive implicit theory than chemistry students. More of the latter tended to follow a receptive or interpretive implicit theory of reading.

A distinct feature of Moore and Narciso (2010) and Hernández (2008) vis-a-vis the aforementioned studies is their use of an implicit theories perspective, one where clusters of similar beliefs are thought to define an individual’s orientation towards texts. From this perspective, the focus is not on beliefs about how reading should be taught, how it should be learned, or on the uses of reading in the context of teaching the language system. Instead, the focus is on beliefs about the nature of texts and reading.

The information yielded by the approach taken by Moore and Narciso (2010) and Hernández (2008) does not provide direct data on beliefs about the teaching or learning of EFL reading. Nevertheless, it can be valuable in two ways.

First, EFL teachers’ implicit theories bear more relevance in teaching-learning situations where the primary focus is on text content and discourse rather than on the language system or reading strategies. Such situations include language arts courses in elite private schools in EFL countries, advanced EFL literature courses, and EAP courses. Second, an implicit theories approach can be especially useful when investigating the beliefs of pre-service teachers in undergraduate EFL programs who have not yet had teaching experience and who, depending on the specifics of their study program, may not have developed much knowledge about the teaching and learning of L1 or L2 reading.

This study addresses precisely that kind of situation. It uses the questionnaire developed by Hernández (2008) to investigate the implicit theories about text and reading held by a group of undergraduate students in a BA in English at a large Mexican university. A distinctive feature of the study is that it attempts to describe patterns of differences in implicit theories in connection with curricular changes (non-causally, of course). It compares the implicit theories of two different cohorts studying under different curricula due to a university-wide process of curricular reform. Within cohorts, the implicit theories held by students specializing in EFL teaching are compared to those of students specializing in literature and translation studies. We group all these students under the label “pre-service EFL teachers” because our personal experience and anecdotal evidence show that a large majority of literature and translation majors in this program go on to become EFL teachers.

This kind of study is relevant from a contextual perspective on beliefs, one from which beliefs are seen less as cognitive traits and more as situated discourses in a reciprocal, dynamic relationship with the contexts they index (Barcelos, 2006; Dufva, 2006; Kalaja, 1995). From this contextual perspective, exposure to different tasks, contents, and perspectives—or discourses—on texts and reading as a result of curricular differences may lead to the internalization and externalization (Neguerela-Azarola, 2011) of different belief systems, or implicit theories of text and reading.

Accordingly, the aim of this study is to describe and contrast the implicit theories about texts and reading verbalized by students of different cohorts and areas of specialization in a BA in English at the University of Eastern Mexico (UEM, a pseudonym). Specifically, we address the following research questions: What are the participants’ verbalized implicit theories of text and reading? What differences exist in the participants’ implicit theories as related to their cohort and their areas of specialization?

Our results show the existence of sub-types of implicit theories that are slightly different from those found in previous studies. Patterns of variation appear to be associated with both cohort and area of specialization. In presenting and discussing these results, we do not wish to imply that students are firm in these beliefs in or that they guide their interaction with all kinds of texts. Rather, we think their answers are influenced by the fact that the questionnaire was administered in a school context; it is likely that asking the same questions in a different context for a different kind of text might have elicited altogether different beliefs.

Theoretical Framework

The beliefs, epistemic models, or personal epistemologies of students and teachers have been widely investigated by scholars in education and foreign language learning and teaching. As is common in the social and cognitive sciences, there have been multiple theoretical and methodological approaches to the study of beliefs. From a cognitive perspective, Hofer (2001) has defined personal epistemologies as “ideas about knowledge and knowing” (p. 355). Other terms used to describe such ideas are “epistemological beliefs” (Qian & Alvermann, 2000), “epistemic beliefs” (Bendixen, 2002), and “epistemic theories” (Hofer & Pintrich, 2002). Many important studies on beliefs about reading have drawn from this perspective (Schraw, 2000; Schraw & Brunning, 1996), including one on pre-service EFL teachers’ beliefs (Moore & Narciso, 2011). Further, at least one sociocultural study (Hernández, 2008) was inspired by Schraw and Brunning’s (1996) research on personal epistemologies. Our own study is also sociocultural and adheres to a contextual perspective on beliefs. From this perspective, beliefs are defined as:

…a way of thinking, constructions of reality, ways of seeing and perceiving the world and its phenomena, co-constructed in our experiences and resulting from an interactive process of interpretation and (re)signification. As such, beliefs are social (but also individual), dynamic, contextual and paradoxical. (Barcelos, 2006, p. 18)

…a way of thinking, constructions of reality, ways of seeing and perceiving the world and its phenomena, co-constructed in our experiences and resulting from an interactive process of interpretation and (re)signification. As such, beliefs are social (but also individual), dynamic, contextual and paradoxical. (Barcelos, 2006, p. 18)[1]

[1] The original text in Brazilian Portuguese is“uma forma de pensamento, construções da realidade, maneiras de ver e perceber o mundo e seus fenômenos, co-construídas em nossas experiências resultantes de um processo interativo de interpretação e (re)significação. Como tal, crenças são sociais (mas também individuais), dinâmicas, contextuais e paradoxais” (Barcelos, 2006, p. 18)

In other words, this is a contextual study in which beliefs are thought of as ideas which are interrelated with the contexts people inhabit and their experiences as situated in those contexts (Barcelos, 2003). We also subscribe to compatible Vygotskyan and Bakhtinian views on beliefs. From a Vygotskyan point of view, beliefs are seen as internalized from the social/communicative environment, and then externalized again in acts of enunciation; they are both socially relevant and personally meaningful (Neguerela, 2012). Following Dufva’s (2006) Bakhtinian perspective on beliefs, we posit that beliefs, because they are internalized from the social context, are also discursive and polyphonic: they consist of a network of voices, sometimes contradictory, that informants vocalize. In this perspective, beliefs may vocalize and index individual, social, or institutional discourses. Individual discourses relate to unique, idiosyncratic experiences. Social discourses pertain to attitudes in a speech community. Finally, institutional discourses concern educational systems, curricula, syllabi, and teachers. These two perspectives complement one another in that, from the Bakhtinian perspective, discourses are the manifestation from which beliefs are internalized, and they are externalized as discourses as well. A polyphonic perspective affords the possibility of exploring intertextual links across the beliefs articulated by several participants and in different discourses. From the polyphonic perspective, a belief is always evoked by dialogue with an Other. The nature of this Other has an impact on the quality of the voices that are evoked and articulated in discourse. If the Other who evokes beliefs at the moment of interaction is assigned a school identity, it is possible that the beliefs articulated by the participant may index institutional discourses. In other words, beliefs may vary contextually both as a function of the discursive participant’s constructed identities and of individual and group experiences.

Drawing on Wineburg’s (1991) seminal research on rhetorical reading in history education, Schraw and Bruning (1996) created a research-based taxonomy of three implicit models of reading. Implicit models consist of “tacit, systematic assumptions about the role of the reader” (Schraw & Bruning, 1996, p. 291). In their version, implicit models do not include beliefs about text structure or reading self-efficacy. Using quantitative methods, they found three implicit models, presented in Table 1.

Table 1. Implicit models by Schraw and Bruning (1996).

The models’ impact on specific reading tasks was investigated by Schraw (2000). He found that readers subscribing to the transactional model produced more extratextual inferences and more complex interpretations than those who adhere to other models.

Inspired by Schraw and Brunning (1996), Hernández (2008) conducted a qualitative study to characterize the evolution of implicit theories of reading across secondary and tertiary education. Hernández (2008) compared the beliefs of chemistry and literature majors with the goal of investigating sociocultural influences on beliefs. He also investigated metatextual knowledge, which was defined in terms of the participants’ responses to the questions “What is a text?” and “What is the function of texts?” Regarding the first question, he found that literature students primarily understood the nature of texts in communicative terms (related to authors’ motives, intentions, and positions), while the rest understood texts in structural terms related to the arrangement of words and information. As to the second question, students understood the functions and purposes of texts as either to educate (i.e., the sharing of information by one who knows more with others who know less) or to communicate (i.e., the desire to make meaning in a way that makes sense within a given context). Hernández (2008) interpreted these participant-reported functions in terms of the transmissive-univocal function and the generative-dialogical function posited by Bakhtin (in Werstch, 1993; Lotman, 1988).

From a qualitative analysis, Hernández (2008) created a three-level matrix of reading models, text epistemologies (derived from metatextual knowledge), and implicit theories of reading. The two former constructs are contained within the latter. That is, implicit theories are formed from a combination of reading models and text epistemologies, with the latter being derived from metatextual knowledge (see Table 2 below). 

Table 2. Hernández’s (2008) implicit theories of reading.

The students at the reproductive level assume that there is a single “correct” meaning in texts that readers must recover; the function of texts is to transmit knowledge. The students at the interpretive level believe that there is a single “correct” meaning but also think they can interpret this meaning subjectively; for these students, the function of texts is to learn from them. Finally, the students at the constructive level believe that they can go beyond the authorial/textual meaning to interrogate and critique the text. They think that the function of texts can be to apply their messages to real life situations. Hernández (2008) posits that the constructive model is more evolutionarily advanced, as it appears only in college students, and specifically in literature majors.

In agreement with Hernández’s (2008) Bakhtinian sociocultural perspective, we think beliefs are always situated in and interwoven with sociocultural contexts. Accordingly, systematic exposure to disciplinary reading practices and discourses on reading are likely to influence students’ vocalized beliefs about reading and texts. We complement Hernández’s Bakhtinian conceptualization of metatextual knowledge with a concept of contemporary rhetoric, that of the rhetorical situation (defined below) because this theoretical lens allows us to make finer distinctions between different types of metatextual knowledge found in our data. Some readers may be perplexed by this move as Bakhtin was himself dismissive of rhetoric as it was understood during his historical period, which is to say, an Aristotelian rhetoric that emphasized persuasion (Zappen, 2004). By contrast, contemporary rhetoric has largely embraced Bakhtin’s dialogism and incorporated it to its theorizations of language and discourse (Murphy, 2001).

Accordingly, contemporary rhetoricians understand the rhetorical situation as a complex arrangement of factors that both give rise to and are created by authors when writing texts. These factors include socially-constructed authorial motives and intentions (Bazerman, 1988, 2013; Bitzer, 1968), target audiences (Vatz, 1973) genre (Miller, 1994), intertextual links to other texts and genres, the activity systems the texts participate in, and the sociohistorical contexts of the text’s production and reception (Bazerman, 1988, 2013). This emphasis on audience, intertextuality, and sociohistorical context make this construct compatible with Bakhtinian dialogism.


This study approaches implicit theories of reading and metatextual knowledge from the Bakhtinian sociocultural perspective put forward by Hernández (2008). The study was conducted at the University of Eastern Mexico, a large, prestigious, public university. UEM promotes community reading programs, publishes literary and academic journals and books, and gives away such publications to the public. However, there is no specific attention to reading or writing across the curriculum beyond the introductory first-year writing courses.

The study focuses on comparing students majoring in a BA in English program who are studying under two different curricula (Planes de Estudios in Mexico). We will simply call these two curricula C-1 and C-2. At the time the data were collected, the last C-1 cohort was about to graduate and the first C-2 cohort was one term away from graduation. This was, thus, a unique moment that afforded a comparison between the two curricula. An important difference between the curricula and, consequently, the cohorts, lies in the areas of specialization. C-1 included three such areas (teaching, translation, and literature) with several mandatory courses within each one and no opportunities for taking courses in the other two areas once one had been chosen. By contrast, C-2 has less rigid areas of specialization. Students can, but are not required to, declare an area of specialization while at the same time being able to choose any combination of courses in either teaching, translation, or literature at almost any point during their BA program. Consequently, there are fewer courses in any given area in C-2 than there were in C-1, and C-2 students thus share a greater number of courses. The faculty does not vary across curricula.

We used focused sampling and saturation sampling procedures. Our goal was to have at least ten participants in each area of specialization, or area of interest for C-2 students. Because we wanted participants that had ample experience with the curricula, we chose to focus on students in their last or penultimate term before graduation. A total of 60 students participated in the study. In C-2’s first cohort, 30 were students (again, about to graduate), and 30 were students in C-1’s last cohort (also about to graduate). In each cohort, there were ten participants per area of specialization or interest (ten in teaching, ten in translation, and ten in literature, or 20 per area when both cohorts are considered). The C-2 participants had all declared an area of specialization but had also taken courses in the other areas.

Hernández’s (2008) interview questionnaire was used. However, it was administered as a written questionnaire due to the relatively large sample size, time constraints, and financial limitations. According to Victori (1999), the open-ended questionnaire is a useful instrument for comparing the beliefs of two or more large groups. The questionnaire was administered during class periods by the second author’s undergraduate assistants after obtaining consent from both the students and the teachers in charge of the classes.

The data were analyzed using directed content analysis (Hsieh & Shannon, 2005), which consists of applying pre-determined coding categories while allowing for the emergence of new codes or changes in the existing ones. Initially, we used Hernández’s (2008) categories, but we added new codes as they emerged from the data and discarded others that were not found. Because these new codes are directly connected to our findings, they are shown and explained in connection with specific results in the Results and Discussions section. When adding new codes, we followed the constant comparative method of grounded theory (Glaeser & Strauss, 1967). The coding process proceeded by consensus among members of the research group. A log was kept to record coding discussions and decisions. The coding was audited by another member of the research group with more expertise in this type of analysis.

After category generation and data coding were completed, simple counts were produced of the numbers of students that adhered to the different types of beliefs about reading and texts. However, these simple counts were not treated statistically as this is a qualitative study. We subscribe the position that using simple counts is both important and legitimate in qualitative research as it helps to discern patterns in qualitative data (Becker, 1970; Maxwell, 2010; Sandelowski, Voils & Knafl, 2009).

Results and Discussion

First, we present and discuss the reading models we found. Then, we present the kinds of metatextual knowledge. In both cases, the distribution of beliefs across cohorts and areas of interest is also presented and discussed. 

Table 3. Reading models and their core beliefs.

There are a few differences between our suggested models of reading and those in Hernández (2008). One such difference is that Hernández’s (2008) interpretive category explicitly includes the notion of respecting one “correct” meaning intended by the author, which co-exists with the belief in multiple interpretations. In our data, there were several cases of students believing only in multiple subjective interpretations and explicitly rejecting the possibility of one correct meaning, that is, the placed the locus of meaning in the reader alone. We reserved the label “interpretive” for this kind of reading model. We used the label “receptive/interpretive” for the kind of reader that Hernández labels interpretive to acknowledge their double belief on one “correct” meaning and the possibility of multiple interpretations.

Within the receptive/interpretive type, there is a subtle distinction in terms of what the participants think are valid subjective interpretations of a text. A handful of participants seem to restrict the meaning of “subjective interpretation” to the aesthetic appreciation of a text. In other words, they believe that there a unique correct message that can generate multiple appreciations or experiences. Below are some examples of what these participants answered to the relevant questions:

Un mismo texto, ¿puede generar la misma comprensión en diferentes lectores?
[Can the same text give rise to different understandings in different readers?]

Podrían tener solo la idea principal igual más si por ejemplo una persona lee un texto sobre violencia doméstica y ha sido abusada tendrá una visión diferente del de una persona que no ha sido abusada.[They may get the same main idea but, if a person reads a text about domestic violence and has been a victim, they will have a different viewpoint than a person who has not been a victim.] (I/2008/B/5/L)

Por supuesto, pero siempre habrá también controversia en las opiniones.
[Of course, but there will always be opposing opinions.]

¿Cuál es el papel del lector en la comprensión?
[What is the readers’ role in comprehension?]

El lector juega un rol de descifrar y apreciar el texto,
[The readers’ role is to decipher and appreciate the text.]

Other receptive/interpretive readers appear to believe that the construction of multiple, valid meanings are not restricted to aesthetic judgments or experience. They believe that there are multiple, valid propositional meanings, but there is one that is “better” than the others and is the author’s intended meaning:

¿Un mismo texto, puede generar la misma comprensión en diferentes lectores?[Can the same text give rise to the same understanding in different readers?]

Puede ser en el sentido de que ambos concuerden en los mensajes claves que el texto transmite. Puede ser diferente dependiendo el tipo de lector y el nivel de compresión que cada uno tenga.
[It may be the same in the sense that two readers may agree with the key messages transmitted by the text. It may be different depending on the reader and the level of understanding each one has.]

In addition, there are several students for whom the existence of multiple valid meanings depends on the text’s genre, which strikes us as a very sophisticated belief:

¿Un mismo texto, puede generar la misma comprensión en diferentes lectores?
[Can the same text give rise to the same understanding in different readers?]

Depende del tipo de texto, no es igual una novela que un reporte sobre el AND.
[This depends on the type of text, a novel is not the same as a report on DNA.]

Sería diferente si el grupo no es especialista en el tema o si el tema es muy amplio y/o subjetivo. Sería igual si se trata de un texto científico.
[Understanding would be different in non-specialized groups or if the subject is too broad and/or subjective. It’d be the same understanding if it’s a scientific text.]

We now discuss our constructive models of reading focusing first on differences and then on commonalities. In Hernández (2008), the constructive model is a capacious model with three dimensions: a dimension of application of the knowledge gleaned from texts to other areas of life, a rhetorical dimension, and a critical one. We found it useful to distinguish among these constructive models for three reasons. First, they did not overlap in our data: participants subscribing to one did not subscribe to the others. Second, as discussed by Perales-Escudero and Sandoval (2016), it can be advantageous to distinguish between critical and rhetorical forms of reading in contexts of disciplinary enculturation. This is because it may not always be appropriate or possible for college students to evaluate disciplinary texts critically as novices, but it is generally desirable and possible (with scaffolding) for students to reconstruct elements of texts’ rhetorical situations. Third, as discussed below, there appear to be enough qualitative differences in our data to warrant this three-way division.

Thus, we distinguish three types of constructive models: constructive-applied, constructive-critical, and constructive-rhetorical. This order does not imply an evolutionary hierarchy. Constructive-applied readers believe that texts are to be used practically in daily life:

¿Para qué sirve comprender los textos?
[What is the use of understanding texts?]

Sirve para incrementar y/o aplicar conocimiento, para reflexionar, para estar informado sobre algún tema en específico.
[It is useful for increasing and/or applying knowledge, to be informed about a specific topic.] (I/2008/B/4/L)

Sirve para usarlo posteriormente, o para aplicarlo lo más eficientemente posible.
[To use it afterwards or to apply it as efficiently as possible.]

Puede tener aplicaciones a la vida cotidiana, una utilidad spiritual.
[It can have uses in daily life, a spiritual usefulness.]

Constructive-critical readers believe that evaluation is central for comprehension:

¿En qué consiste comprender un texto?
[What does it mean to understand a text?]

En evaluar y examinar cada una de sus partes y luego todo en conjunto.
[It’s evaluating and examining each one of its parts and then the whole.]

Finally, constructive-rhetorical readers believe that comprehension involves examining rhetorical dimensions of the text, such as context of production, different interpretations, author’s intentions, etc.:

¿En qué consiste comprender un texto?
[What does it mean to understand a text?]

Consiste en conocer las razones, circunstancias en las que el texto fue escrito, así como las distintas interpretaciones y líneas de lectura que tiene un texto.
[It consists in knowing the reasons and circumstances in which the text was written, and also the different interpretations and lines of reading a text has.]

Es leer entrelíneas, hallar su motivo. Algunos textos tienen que entenderse con su contexto histórico.
[It’s reading between the lines, finding its motive. Some texts have to be understood within their historical context.]

En un primer momento significa tomar la idea principal del texto. Después significa cuestionarse el objetivo, por qué el autor defiende esa tesis cultural.
[At first it means grasping the text’s main idea. Then it means to question the goal, why the author defends a certain thesis. Understanding the cultural background.]

Another important dimension of the constructive model is that all the participants who subscribe to it also show evidence of adhering to a receptive/interpretive model. In other words, they simultaneously believe that: a) texts must be interrogated/applied/evaluated, and b) there is one propositional knowledge that the author intends to convey. Therefore, in our sample, students with a constructive model also always hold a receptive/interpretive model, and our counts in the relevant tables reflect this (Constructive students were only counted as such, not as receptive/interpretive). For example, the same student produced both answers to the questions below:

Un mismo texto, ¿puede generar diferentes interpretaciones?
[Can the same text give rise to different interpretations?]

No creo en una comprensión mediana. Se capta o no un mensaje. Una persona no confiesa haber comprendido a medias, puesto que un buen lector indaga hasta la comprensión.
[I don’t believe in partial comprehension. You either understand a message or you don’t. A person won’t avow to partial understanding because good readers keep asking questions until they have understood.]

¿Qué factores pueden hacer fácil o difícil la comprensión?
[What factors can help or hinder comprehension?]

En consideración con la época, la cultura, etc. Cada lectura es subjetiva y susceptible al contexto en que se da.
[We have to consider the historical period, the culture, etc. Each reading is subjective and influence by the context where it happens.]

These findings support those of Moore and Narciso (2011) that readers with a transmission model display more critical reading strategies than those with a translation model. In other words, the belief that texts are to be interrogated and/or evaluated seems to go hand in hand with the belief in one single propositional meaning conveyed by the author.

The Figures below show the counts of the number of participants that adhered to each reading model. Figure 1 compares the number of C-1 (N=28)[1] and C-2 (N=30) students.

[1] It was not possible to identify a reading model for two participants because their answers were quite incoherent.

Figure 1.Proportion of students holding each reading model per curriculum.

As Figure 1 shows, the largest differences lie in the number of students vocalizing a receptive model (R). While these students make up 18.5 percent of C-1 participants, they account for 36.7% of C-2 participants. C-2 also features a smaller number of students vocalizing a constructive-applied model (C). However, C-2 includes more students vocalizing a constructive-critical model (CCrit) and a constructive-rhetorical one (CRet). No C-2 students were found to hold the latter.

Once the data are disaggregated per area, it becomes apparent that the numbers in Figure 1 are related to a greater presence of the receptive and receptive/interpretive models in C-2’s teaching and literature majors. This is shown in Figures 2 and 3.

Figure 2.Distribution of C-1 students’ reading models per area.

Figure 3.Distribution of C-2 students’ reading models per area.

As shown in Figure 2, only C-1 translation majors held a receptive model (R). In contrast, Figure 3 shows that students of the three majors hold this theory, although only one is a literature major.

C-2 also includes a smaller number of translation students holding a receptive/interpretative model (R/I) and a rise in the number of translation students holding one of the constructive models. Most C-1 literature students held an interpretive model (I), but most C2 literature students are receptive-interpretive (R/I). These results suggest that the teaching of literature and translation courses changed under C-2, promoting the vocalization of beliefs related to greater closeness to textual meaning in literature and greater distance in translation.

The presence of the constructive-rhetorical model (C-R) in C-2 contrasts with its total absence in C-1. The fact that it has adherents in the three areas might be an indication that general reading and writing courses included an attention to rhetorical aspects. However, the fact that only three students hold this C-R model means that it could be the product of extracurricular experiences or prior high school teaching.

We now present results related to metatextual knowledge. Our codes and types are inspired in Hernandez’s (2008) distinction between the transmissive-univocal and the generative-dialogical types of metatextual knowledge. However, we felt the need to develop new codes/types since we found a distinction between two types of metatextual knowledge previously subsumed under the generative-dialogical type. These are the functional type and the rhetorical type of metatextual knowledge. The former was used to code segments where students commented on different purposes of texts, such as to inform or to educate readers. The latter was used for segments where students talked about aspects of a text’s rhetorical situation, such as authorial intentions, the text’s socio-historical context, intended audiences, and/or intertextual links. These codes are summarized in Table 4. 

Table 4. Metatextual knowledge types.

In our sample, we found that all students who wrote about texts’ social functions also vocalized an understanding of texts in terms of formal features. All students in our sample displayed this type of functional metatextual knowledge, as in the example below:

Conocer los textos es conocer sus tipos y funciones, lo más importante es la introducción y lo menos importante es la conclusión.
[Knowing texts is knowing their types and functions, introductions are the most important and conclusions are the least important.]

However, there was a third group of students who, in addition to those two ways of understanding, also understood texts in rhetorical terms.

¿Para qué crees que sirven los textos?
[What do you think is the use of texts?]

Depende del tipo de texto y su intención, incluso existen textos que desinforman deliberadamente porque ése es su objetivo.
[It depends on the type of text and its intention. There are even texts that spread false information because that is their goal.]

El texto es el producto, que a menudo, cobra vida propia y permanece por siglos, a veces, con diversas interpretaciones dependiendo del contexto, en consideración con la época, la cultura.
[The text is the product that often takes on a life of its own and remains for centuries, sometimes with several interpretations according to the context, taking into consideration the historical period and the culture.]

Cada texto está hecho específicamente para cierto tipo de lectores; entonces cada texto tiene un diferente nivel.
[Each text is written specifically for a certain type of readers; then each reader has a different level.]

All these students are aware that texts are written with specific intentions and are aimed at concrete audiences. The second one shows a sophisticated awareness of the contingency of textual interpretation as related to sociohistorical contexts. As stated above, not all students showed rhetorical metatextual knowledge, but all students who showed it also showed functional knowledge. It seems then, that rhetorical knowledge is a more advanced form of the generative-dialogical type of metatextual knowledge as it both includes and transcends an awareness that texts and their features serve different communicative purposes. One mention of a rhetorical element was enough for us to count a student as holding rhetorical knowledge of texts. Tables 5 and 6 show our numbers of students that vocalized each type of metatextual knowledge in each curriculum and area.

Table 5.Kinds of metatextual knowledge in C-1 students.

Table 6. Kinds of metatextual knowledge in C-2 students.

As suggested by the results related to reading models, there is a greater number of readers with rhetorical metatextual knowledge in C-2, and most are in the translation and literature areas.


Our first research question inquired about the implicit theories vocalized by the participating students, and our second research question was intended to explore variation in implicit theories as related to curriculum and area of specialization. Our data confirm the three implicit theories found by Hernández (2008). However, there are three important differences. The first is that our data show an interpretive model of reading that does not include a belief in one correct meaning. The second is that our data show three sub-types of the constructive model of reading: an applied one that is identical to the one in Hernández (2008), plus a critical one, and a rhetorical one. The third difference is that the data bore a distinction between two types of generative-dialogical metatextual knowledge: one functional and one rhetorical. The latter is associated with a constructive-rhetorical and constructive-critical model of reading in that the students holding those models also hold rhetorical metatextual knowledge. This association suggests the existence of a rhetorical-critical implicit theory, which considers both an understanding of rhetorical elements of texts and an evaluative reading stance to be intrinsic dimensions of reading comprehension. In our sample, this theory occurs only in C-2 students. Although some C-1 students vocalized rhetorical metatextual knowledge, they did not vocalize a rhetorical model of reading. This is thus an important area of difference in implicit theories between the two curricula.

Another important difference is the distribution of students holding the different theories across areas of specialization. This distribution was more clearly distinct per area in C-1 students, where the interpretive model was vocalized only by literature students and the receptive model was vocalized only by translation students. In C-2, by contrast, the three models were vocalized by students in all three areas, with one exception: the interpretive theory was vocalized only by one C-2 teaching student and was absent in the other two areas. Together with the finding that the rhetorical model occurs only in C-2 students, these results suggest that students in each curriculum have been exposed to different discourses about reading. In C-2, this exposure appears to be both more homogenous (i.e., more students exposed to more discourses than in C-1), and more focused on close readings of texts and rhetorical dimensions of texts. This is suggested by the higher number of C-2 students adhering to one of the constructive models (11 in C-2 vs. 7 in C-1) and the fact that the rhetorical model occurred only in C2. At first glance, these findings point to the possibility that C-2 students are better-equipped to treat texts in pedagogically savvier ways in those conditions where it is appropriate to focus more on discourse and content. However, the fact that more C2 students than C1 students hold a receptive model and that the C2 receptive students belong to the three areas suggests just the opposite. Further, the fact that a majority of students in both curricular subscribe to the less advanced receptive, interpretive and receptive/interpretive models should give curriculum designers and teachers pause. If one goal of an undergraduate degree in ELT is to prepare future teachers to teach reading in sophisticated ways, it is clear that more steps need to be taken by policy makers and faculty to make this happen at UEM.

Some of our participants clearly understand comprehension in highly sophisticated ways as they distinguish between the possibility of interpreting the propositional content differently (i.e., two readers may understand the content of the same text differently) and the possibility of different appreciations of the text based on the same propositional understanding (i.e., two diverging opinions or aesthetic judgments based on the same understanding of the propositional content). At the same time, the fact that many participants hold more than one model of reading simultaneously (receptive/interpretative and the combination of constructive with receptive/interpretative happening in all constructive readers) demonstrates the polyphonic and sometimes contradictory nature of personal beliefs: the same individual may vocalize different discourses as a function of their sociocultural exposure. The fact that the questionnaire was administered during class time by teachers and students may have contributed to the participants’ vocalization of school-based beliefs.

Our findings appear to confirm those in Moore and Narciso (2011) in the sense that believing in one “correct” propositional meaning and struggling to get to it is a feature of more advanced readers. In our study, constructive readers, who would hold the most advanced theory, also believe in one propositional meaning encoded in the text by the author. To them, a faithful reconstruction of this intended meaning seems to be a pre-requisite for further application, interrogation or evaluation of the text. Our finding of mixed models also aligns with Moore and Narciso’s (2011) finding that there are transition models of reading.

This study is limited because its findings are not triangulated with an exploration of the actual curricula and syllabi, teachers’ beliefs, or reading textbooks. Future research should attempt to obtain and analyze data from such sources in order to gain a better understanding of the discursive processes involved in the development and vocalization of pre-service EFL teachers’ beliefs about reading. Further studies with larger samples are also needed to confirm the existence of a critical-rhetorical implicit theory of reading in both pre-service EFL teachers and other populations, as well as the factors that may promote its emergence.


Bamanger, E. M. & Gashan, A. K. (2014) In-service EFL teachers’ beliefs about teaching reading strategies. English Language Teaching, 7(8), 14-22. doi:

Barcelos, A. M. F. (2006). Researching beliefs about SLA: A critical review. In P. Kalaja & A. M. F. Barcelos (Eds.), Beliefs about SLA: New research approaches (pp. 7-34). Amsterdam, Netherlands: Kluwer Academic Publishers.

Bazerman, C. (1988). Shaping written knowledge: The genre and activity of the experimental article in science. Madison, WI: The University of Wisconsin Press. Retrieved from

Bazerman, C. (2013). A rhetoric of literate action. Fort Collins, CO: The WAC Clearinghouse and Parlor Press. Retrieved from   

Becker, H. S. (1970). Sociological work: Method and substance. New Brunswick, NJ: Transaction Books.

Bendixen, L. (2002). A process model of epistemic belief change. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (pp. 191-208). Mahwah, NJ: Lawrence Erlbaum.

Bitzer, L. (1968). The rhetorical situation. In W.A. Covino (Ed.), Rhetoric: concepts, definitions, boundaries (pp. 1-14). Boston MA: Allyn and Bacon.

Chou, C. (2008). Exploring elementary English teachers’ practical knowledge: A case study of EFL teachers in Taiwan. Asia Pacific Education Review, 9(4), 529–541.

Dufva, H. (2006). Beliefs in dialogue: A Bakhtinian review. In P. Kalaja & A. M. F. Barcelos (Eds.), Beliefs about SLA: New research approaches (pp. 131-152). Amsterdam, Netherlands: Kluwer Academic Publishers.

Glaser, B. G. & Strauss, A. (1967). The discovery of grounded theory: Strategies for qualitative research. Chicago, IL: Aldine Press

Grabe, W. (2009). Reading in a second language: Moving from theory to practice. Cambridge, UK: Cambridge University Press.

Hernández, G. (2008). Teorías implícitas de lectura y conocimiento metatextual en estudiantes de secundaria, bachillerato y educación superior. Revista Mexicana de Investigación Educativa, 13(38), 737-771. Retrieved from

Hsieh, H. & Shannon, S. E. (2005). Three approaches to qualitative content analysis. Qualitative Health Research, 15(9), 277-1288. doi:

Hofer, B. K. (2001). Personal epistemology research: Implications for learning and teaching. Journal of Educational Psychology Review, 13(4), 353-383.

Hofer, B. & Pintrich, P. R. (Eds.) (2002). Personal epistemology: The psychology of beliefs about knowledge and knowing, Mahwah, NJ: Lawrence Erlbaum.

Kalaja, P. (1995). Student beliefs (or metacognitive knowledge) about SLA reconsidered. International Journal of Applied Linguistics, 5(2), 191−204. Doi: 10.1111/j.1473-4192.1995.tb00080.x

Khonamri, F. & Salimi, M. (2012). The interplay between EFL high school teachers’ beliefs and their instructional practices regarding reading strategies. Novitas-Royal (Research on Youth and Language), 4(1), 96-107.

Krashen, S.D. (2003). Explorations in language acquisition and use: The Taipei lectures. Portsmouth, NH: Heinemann.

Kuborzka, I. (2011). Links between teachers’ beliefs and practices and research on reading. Reading in a Foreign Language, 23(1), 102-128.

Lotman, Y. M. (1988). Text within a text. Soviet Psychology, 26(3), 32-51.

Maxwell, J. (2010). Using numbers in qualitative research. Qualitative Inquiry, 16(6), 475-482. doi: 10.1177/1077800410364740

Miller, C. R. (1994). Genre as social action. In A. Freedman & P. Medway (Eds.), Genre and the new rhetoric (pp. 20–36). London, UK: Taylor & Francis.

Moore Hanna, P. & Narciso Sánchez, E. (2011). Modelos epistémicos de la lectura en estudiantes mexicanos. Revista Mexicana de Investigación Educativa, 16(51), 1197-1225. Retrieved from

Murphy, J.M. (2001). Mikhail Bakhtin and the rhetorical tradition. Quarterly Journal of Speech, 87(3), 259-277. doi

Navarro, D. & Thornton, K. (2011). Investigating the relationship between belief and action in self-directed language learning. System, 39(3), 290-301. doi:

Neguerela-Azarola, E. (2011). Beliefs as conceptualizing activity: A dialectical approach for the second language classroom. System, 39(3), 359-369. doi:

Pajarés, M. F. (1992). Teachers’ beliefs and education research: Cleaning up a messy construct. Review of Educational Research, 62(3), 307-332. doi:

Perales-Escudero, M. D. & Sandoval-Cruz, R. I. (2016). La lectura retórica desde la enseñanza discursivo-cognitiva: una alternativa para la alfabetización académica. In Bañales, M. Castelló & N. Vega (Eds.), Enseñar a leer y escribir en la educación superior: Manual de buenas prácticas basadas en la investigación (pp. 55-76). México, D.F.: Editorial SM.

Pulido, D. (2009). Vocabulary processing and acquisition through reading: Evidence for the rich getting richer. In Z. Han & N.J. Anderson (Eds.), Second language reading research and instruction: Crossing the boundaries (pp. 65-82). Ann Arbor, MI: The University of Michigan Press.

Qian, G. & Alvermann, D. E. (2000). Relationship between epistemological beliefs and conceptual change learning. Reading & Writing Quarterly, 16(1), 59–74. doi:

Sandelowski, M., Voils, C. I. & Knafl, G. (2009). On quantitizing. Journal of Mixed Method Research, 3(3), 208-222. doi:

Sandoval Cruz, R. I. & Perales Escudero, M. D. (2012). Models of L2 reading comprehension and related pedagogical practices: A discussion of the evidence. MEXTESOL Journal, 36(2), 1-18. Retrieved from

Schraw, G. (2000). Reader beliefs and meaning construction in narrative text. Journal of Educational Psychology, 92(1), 96-106.

Schraw, G. & Bruning, R. (1996). Readers’ implicit models of reading. Reading Research Quarterly, 31(3), 290-305. doi: 10.1598/RRQ.31.3.4

Vatz, R. E. (1973). The myth of the rhetorical situation. Philosophy and Rhetoric, 6(3), 154-161. Stable URL:

Victori, M. (1999). Methodological issues in research on learners’ beliefs about language learning. Paper presented in the 12th Conference of the Association Internationale de Linguistique Appliquée (AILA 99). Tokyo, Japan.

Wertsch, J. V. (1993). Voices of the mind: A sociocultural approach to mediated action. Cambridge, MA: Harvard University Press.

Wineburg, S. S. (1991). On the reading of historical texts: Notes on the breach between school and academy. American Educational Research Journal, 28(3), 495-520. doi:

Zappen, J. P. (2004). The rebirth of dialogue: Bakhtin, Socrates and the rhetorical tradition. Albany, NY: State University of New York Press.