BRAZILIAN TEACHERS OF PORTUGUESE AS A FOREIGN PERCEIVE THEMSELVES AS LESS PREPARED TO TEACH GRAMMAR THAN THEIR EUROPEAN COUNTERPARTS

: Brazilian teachers profess to enjoy teaching and studying grammar and rank the importance of explicit grammar instruction similarly to their European counterparts or teachers from other nationalities. However, Brazilian teachers differ from other instructors when it comes to their perception of their own preparedness to teach grammar. They consistently rank themselves below their peers when it comes to evaluating their background formation, and their ability to find grammatical information when they don’t k now it. They also rate the didactical materials at their disposal as less adequate than both foreign and Portuguese teachers do. This may be explained by the academic culture in Brazilian universities, especially when it comes to educating new teachers. There has been a consistent dismissal of the importance of grammar in Brazilian Applied Linguistic fields and the results of this survey may indicate that teachers could benefit and acquire security from a more comprehensive formational process.


Introduction
In this paper, we present the results of an exploratory survey aiming to compare the perception of the role of grammar in the foreign language classroom among Brazilian, Portuguese and non-native teachers.More specifically, we try to explore how different academic traditions are reflected in their perceptions of their own professional skills and how confident they are when dealing with grammatical aspects of instruction.
The motivation for the study was a clear difference in practices and materials contrasting Brazilian teachers from colleagues of other origins that we observe in our own practice.Therefore, our goal was to explore if this perceived difference can be confirmed via a quantitative study of teachers of different origins perceptions (for a review on the study of perception in foreign language teaching research, see Jean & Simard, 2011).
In the last decades, we witnessed a heated debate about the role of grammar in foreign language teaching (Schultz, 1996;Ellis, 2006).Traditionally, the study of foreign languages was rooted in grammatical descriptions and a descriptivist approach to language knowledge.During the first half of the 20 th century, under the influence of structuralism and behaviorism, language learning was seen as a process of habit formation.Thus, the teaching was strongly focused on drill exercises that tried to ensure knowledge of the structural aspects of languages.As a result, classes used to follow this dynamic: first, repetitions of dialogue, sentences, and structure, both in groups and individually, followed by exercises such as substitution, memorization, and transformation.This method also took care of correcting immediately any mistakes, to avoid the acquisition of the so-called "bad habits".There was not much space for a metalinguistic description of languages and an explicit focus on grammatical aspects.
From the 1960s, and especially, 1970s, the behaviorist methods received a lot of critiques.Instrumental for this was the perception (first introduced by Chomsky, 1967) that language has a strong creative component.Due to its creative nature, it was also recognized that it was much more difficult to predict what was going to happen in actual linguistic interaction them previously imagined by behaviorist and structuralists alike.Hymes (1972) had a very impactful perspective on language (and specifically, grammar) teaching.According to Hymes, grammatical rules are all but a small set of the choices a speaker must make in real communicative situations.Besides the choice of grammatical forms, speakers must take into consideration the ample gamut of dimensions of actual linguistic interaction, for example, the communication goals, interlocutors, place, and moment of communication all impact the selections a speaker must in practice make.For this reason, Hymes proposed that teaching a language should focus on communicative competence, the capacity of speakers to deal with real situations and react adequately to them.Krashen (1982), for instance, proposes that grammar does not ever need to be explicitly taught, especially in contexts when one wants to develop communicative competence and not metalinguistic awareness.
Another example of ostracization of grammar practices in the classroom is Widdowson (1991).The author proposes that teaching formal aspects does not lead to or guarantee communicative competence.In contrast, according to Widdowson, the knowledge of usage leads to the acquisition of the form.Therefore, all teaching focusing on grammar would be a sort of transgression to the communicative method.
This theoretical debate, however, had an uneven effect on actual teacher practices in the classroom.This effect is even more uneven depending on how much the different positions influenced different academic environments and different teacher training traditions.
The emphasis on communicative competence and on "real" usage situations led to the development of a very influential communicative method, that, albeit with a lot of transformation, is still very influential to this day.It is important to stress that the communicative method does not summarily condemn grammar systematization (Thompson, 1996) but rather proposes an inductive approach to its teaching.The learner should first be exposed to language structure and its possibilities and afterwards, be directed to the examination of the underlying grammatical rules.However, in practice, sometimes, especially as we will see, in Brazilian Academic circles, there was resistance to grammar in the foreign language classroom.
Studies show that, to the present day (Schulz 1996;Loewen et al. 2009) as well as teachers (Burgess & Etherington, 2002) value grammar study and rule learning.Jean & Simard (2013) declare that, despite efforts on the contrary, grammatical traditional teaching is still prevalent in foreign language classrooms.Despite the attention in research and theoretical debate, the role of grammar in the classroom has remained relatively stable for the last decades (Garth, 2024).Traditionally, instructors use explicit grammar explanations and route exercises to drill into the rules being acquired.
Ortale & Ferroni ( 2015) published an ethnography-based study on Brazilian teachers of foreign language anxieties about teaching grammar.According to the authors, their participants revealed that they felt unprepared for the task and had difficulties managing students' demands on the subject matter.In this work, teachers reveal insecurities based on their perceived lack in their formation.Gatti (2014), investigated in large-scale teacher training programs in Brazilian universities, focusing on their syllable and pedagogical components.The author claims that there is a schism between theoretical and practical aspects of the training and that, in a lot of cases, teacher training programs fail to properly connect classroom practices to the theoretical debates.
In contrast, Almeida & Yang (2021), talk about the Portuguese approach to grammar and what they call "focus on form".They also examine the influence of this tradition in the classroom practices commonly found in the European Portuguese classes in China.In this environment, grammar teaching takes a central role in the classroom.Many authors, such as Gonçalves (2016) and Jiang (2019) endorse grammar-based methods, based on a deductive way of presenting grammar (e.g.Presentation-Practice-Production models).According to Almeida & Yang, this approach is predominant in the great majority of European Portuguese classes and attributes this tendency both to a more traditionalist view on teaching and method and also to teachers' limitations, namely, that their training mostly prepared them predominately this class style.
In our study, we aimed to explore if these different academic and training traditions impact how teachers from different origins perceive the task of teaching grammar.More pointedly, we tried to shed light on how confident and prepared those different groups feel to teach grammar.

Material and Methods
We prepared a survey (appendix I, translated into English) with two sections, totaling 28 questions.The first section probed on the profile of the respondents, their academic background, professional experience, and demographic characteristics such as age, nationality, and gender.
The second part of the questionnaire, containing 14 assertions, was a Likert scale survey designed to investigate how the instructors perceive the role of grammar in foreign language teaching and their own ability and confidence in teaching grammar.For each assertion, participants were required to express their agreement by choosing from a 7-point scale, from 1, corresponding to "strongly disagree" to 7, "strongly agree", as illustrated in Figure 1 below: Our survey received a high degree of participation.For statistical reasons and to keep the sample sizes comparable, we closed our data collection when the sample reached 42 participants from each of our groups of interest: Brazilian instructors, Portuguese instructors and instructors that do not have Portuguese as their native language, that we labeled in our analysis, foreign instructors.Therefore, our final sample was composed of 126 respondents, equally distributed in the three groups.
The first part of the analysis was to see if the groups were homogenous and comparable in their professional experiences, academic backgrounds and demographic profiles.The results showed that the three groups do not differ significantly in any of those criteria, meaning that a comparison between them is possible and sound.
Due to the ordinal nature of Likert scale surveys, a statistical test that relies on a central tendency measure, for instance, an analysis of variance -ANOVA, is not adequate because it is nonsensical to ask what is the average between strongly agree and strongly disagree.The Likert scale points are ordinally organized.For this reason, a non-parametric test is widely considered a more suitable approach to this kind of data (Sullivan and Artino, 2013).A test that is frequently used in cases like this is the Mann-Whitney U test (Corder and Foreman, 2014), however, this test has a limitation because it can compare only two groups as the independent variable, and we had three groups: Brazilian, Foreign and Portuguese teachers.
The presence of three groups in our samples led us to use the Kruskal-Wallis H test (Wayne, 1990), a non-parametric method for assessing if two or more samples stem from the same distribution.This is a good hypothesis test for Likert scale type of data because it does not presuppose that the data sample has a normal-shaped distribution, meaning that it does not follow the expected parameters for a hypothesis test.
We proceeded to compare the Likert scale answers comparing the three populations and adopting a 0.05% significance level.All significance values have been adjusted by the Bonferroni correction for multiple tests.

Results
A Kruskal-Wallis test was performed on the final scores of the three teacher populations (foreigners, Brazilian and Portuguese) for each of the 14 Likert scales from the questionnaire propositions.In Table 1 below, we show the statistical results of the propositions that did not show any significative difference between teacher populations.We follow the table with a general discussion of those results.As stated earlier, the sample size for all the assertions is 126.For the assertions above, there was no significant difference between groups.Those findings confirm Burgess & Etherington (2002) finding that most language teachers see value in teaching grammar, as well as their perception that students require explicit grammar instruction (either due to attachment to tradition or actual intellectual demand).
All groups either slightly agree or strongly agree with assertion number 5: "Communicative settings are enough for students to acquire the grammatical rules, with no need for explicit grammatical work", with median responses of 6 (as a reminder, 7 is equivalent to strongly agree).This curiously somewhat contradicts participant agreement with the assertion that "I believe that explicit teaching of grammar is important".This contradiction, common to all groups, perhaps reflects the actual dilemma teachers encounter in their practice and the literature.In other words, there seems to be some cognitive dissonance going on that allows teachers to simultaneously agree that is necessary to teach grammar explicitly and that students can deduce the underlying grammar rules without explicit instruction.
The first interesting contrast is about how teachers from different populations project beliefs about the student's perceptions of grammar in the classroom.When asked to agree or disagree with the sentence "Students think that grammar is boring or tedious a", a pairwise comparison of the populations shows that Brazilian and Portuguese teachers had a more optimistic view, tending to disagree more with the statement (p = 0.33) than their foreign counterparts.This perception is perhaps due to the particular position of teaching the grammar of one's own mother tongue instead of one that is a second language to the instructor.
Graph 1: Likert scales notes according to teachers' origins This is the only contrast we found that group native speakers (Portuguese and Brazilian on one side) and foreign teachers on the other.All other significant differences oppose Brazilians to their counterparts.Specifically, Brazilian teachers differ from their colleagues when it comes to their perception of their own preparedness and training for tasks related to grammar teaching.
First, a post hoc pairwise comparison showed that Brazilian teachers declare to feel significantly less prepared to answer and discuss grammatical aspects of language than their Portuguese counterparts (p=0.045),albeit not different from their foreign' colleagues (p=0.558).In a similar fashion, Portuguese and foreigners did not appear to have a difference in their confidence level (p=0.156).
Graph 2: Likert scales notes according to teachers' origins On the same line, both Portuguese teachers feel significantly more confident about their ability to find and research unknown grammatical information (p=0.43)than their counterparts.Post hoc pair-wise show Brazilian teachers feel significantly less confident than Portuguese (p=0.038) and that foreign teachers feel less confident than Portuguese (p=0.023)but, interestingly, they feel as confident as Brazilian teachers (p=0.88).
Graph 3: Likert scales notes according to teachers' origins Together with previous findings, there seems to be a trend of Portuguese teachers perceiving their grammatical knowledge with more confidence than other teachers of Portuguese as a foreign language.Another finding that seems to support that Brazilian teachers feel less confident than the other groups is their answer to the assertion: The published materials available to teach Portuguese are adequate.The answers reveal a sharp difference between the three groups (p<0,001), with Brazilian teachers declaring that they do not perceive available materials as adequate.In contrast, both Portuguese and Foreigners seem to be satisfied with the quality of available materials.It is important to emphasize that Portuguese and Brazilian teachers are forced to use different materials due to important differences between the European and Brazilian varieties of the language.Whereas Portuguese and foreign teachers feel that the available materials are good enough to suit their needs, Brazilians feel comparatively less supported by the Brazilian Portuguese published books and manuals.An interesting data point is that no Brazilian teachers strongly agreed with this assertion, showing that the dissatisfaction is generalized.This points to an important gap that Brazilian researchers and educational planners must overcome to better Brazilian Portuguese teaching.
It is not only the materials that are perceived as insufficient.Brazilian professors tend to disagree much more than members of other groups with the assertion: During my training, I received enough preparation to teach grammar.Both Portuguese and foreign teacher tend to generally agree with that assertion at the same rate (p=0.521),whereas Brazilians tend to feel that their training was lacking and left them less prepared to face the classroom challenges.Pair-wise comparison confirmed that Brazilians perceive more inadequacy in their training than Portuguese (p = 0.045) and other non-native teachers (p=0.017).
Graph 5: Likert scales notes according to teachers' origins

Discussion
Our results show a stark contrast between Brazilian teachers of Portuguese as a foreign language and their Portuguese and non-native colleagues.Brazilian teachers seem to feel less prepared to tackle grammatical issues.This perception, which needs much more investigation, may be due, as we initially hypothesized, by different academic environments and the difference in teacher training programs.Specifically, the Brazilian academy is resistant to straying away from a more orthodox and strict communicative method.
This hypothesis cannot, though, exclude other factors that we would like to bring attention to.A grammatical description of Brazilian Portuguese is notoriously behind the changes that the variant has undergone for the last century or so, staying closely connected to the European Portuguese norm as the standard.In other words, when teachers look at grammar books to find answers to their doubts, they often find it difficult to relate the grammatical rule to their own linguistic experience and dialect.
It is common for teachers to say, as Ortale & Ferroni (2015) put it, that they frequently do not recognize the language they are trying to teach as their native language and as the "real" Brazilian Portuguese.
This aspect, which demands a much more careful look and consideration, can create a difficult dilemma for teachers when confronted with the difference between what people really say and what the published grammars prescribe.This dilemma can lead to many of those insecurities and the perception of themselves as less prepared.
In this sense, we would like to welcome more research in this area and advocate for teacher training programs to include more awareness of the specificity of the Brazilian Portuguese variety.It is also very important that Brazilian academies work to prepare materials able to support teachers when they must research grammatical information.The current available materials are not enough as our results emphatically show.
Finally, we would like to add that we do not take a specific stance on how exactly to teach grammar in the classroom in this paper.We just want to point out that grammar is an important component of language teaching and the teacher should feel confident and ready to make the appropriate choices when necessary.

Conclusion
In this study, we surveyed teachers of Portuguese as a foreign language from three different populations -Brazilian, Portuguese and teachers that do not have Portuguese as their native language.In the survey, we are comparing teachers' perceptions of grammar teaching.Our results show that Brazilians perceive themselves as less prepared to teach grammar, less prepared to find unknown grammar information and less supported by published teaching materials than both their Portuguese and generic colleagues.Teachers' training and academic traditions seem to impact those groups differently.The examination of how those traditions and training programs impact teachers is just beginning and we think is a very promising avenue for research that would have a significant impact in rethinking appropriate ways to implement teacher instruction.

Conflict of Interest Statement
The author declares no conflicts of interest.

About the Author
Maria Luiza Cunha Lima is a Full Professor of Portuguese at the Portuguese Department of Hankuk University of Foreign Studies in Seoul, South Korea.Her research interests include grammar teaching in foreign languages, the concept of linguistic distance and how it impacts or should impact pedagogical practices.She has also published many articles and book chapters about Portuguese grammar, particularly researching the phenomena of definiteness in Portuguese.The professor's publications also include a series of handbooks for teaching Portuguese to Korean learners, called Conversando em Português.

Figure 1 :
Figure 1: Example of the Likert scale survey

Graph 4 :
Likert scales notes according to teachers' origins

Table 1 :
Survey questions in which no significant statistical difference was found