Context

The state where the university is located is characterized by contact between Spanish, Yucatec Maya, and English. The former is the dominant language. However, an undetermined number of native Spanish speakers are direct descendants of Yucatec Maya speakers and use Spanish in non-standard ways (Perales-Escudero et al., 2022). An SLA course is taught in the fifth semester.

The university’s BA in ELT has followed a traditional approach to grammar. Prescriptivism, negative attitudes toward errors, and corrective feedback prevail (Perales-Escudero, 2017). Nevertheless, through interviews and class observation, I determined that the SLA course professor follows a critical, conceptual change approach.

Using an ODF (ODF 1), the professor attempted to make preservice teachers aware of prescriptive, received views of error in the L1 (e.g., the stigmatization of the non-standard subjunctive form of the verb “haber/to exist:” “haiga,” and other non-standard forms in local Spanish that the preservice teachers themselves use) so that they come to see them as products of a legitimate linguistic system that happens to differ from standard Spanish. To the same end, the professor taught the preservice teachers about native varieties of English where the present simple third person -s is not used.

He then recruited this awareness of native non-standard forms to engage preservice teachers in ODFs 2 and 3, which dealt with interlanguage and how L2 non-standard forms (errors and mistakes) evince development rather than deficit. To the professor, this is a descriptive, decolonial, and complex approach to interlanguage development. He tied this approach explicitly to the concept of descriptivism. The prompts for the discussions required teams of preservice teachers to integrate concepts such as interlanguage, error, mistake, nontarget-like forms, prescriptivism, descriptivism, and power asymmetry, and then apply them to analyze their own and others’ experiences with corrective feedback and language learning.

Next, the professor implemented a fourth ODF focused on Larsen-Freeman’s (2014) self-referential assessment. In this ODF, preservice teachers discussed the grading of a hypothetical learner who has made morphological mistakes in aspect marking in a hypothetical test but nonetheless shows development of tense-temporality connections in the inclusion of auxiliary verbs and standard word order. In the fifth ODF, the preservice teachers reflected on their grading and how it contrasted with previous learning about interlanguage and related concepts. In all the ODFs, participants were required to write their individual views first, then negotiate agreements and disagreements, and finally collectively write a concluding paragraph defining and connecting the target concepts. The professor gave the participants the chance to write in Spanish in the ODFs. All of them but one native speaker of Belizean English did so.

The professor was aware of threshold concepts and explicitly followed this approach to conceptual change in that he sought preservice teachers to discard their previously received views and embrace the deep learning of interlanguage and related concepts that he advocates.

Participants

The participants were all preservice teachers enrolled in the target course, taught in the fifth semester of the university’s BA in ELT. By this point in their degree program, they were expected to have reached a B1 level of English competence. All but one were EFL learners who began their BA with a pre-A1 or A1 level of English. I recruited them by physically visiting the classroom, explaining the study, and obtaining written informed consent. Sixteen students agreed to participate, which meets the minimum recommended by Trigwell (2000) for phenomegraphic studies. Following Hajar’s (2021) advice to contextualize conceptions in the participants’ lifeworld, information was gathered on the following topics pertaining to their prior, contextualized experiences through a semi-structured interview conducted before the onset of the study (Interview 0):

1. Relevant apprenticeship of observation (i.e., experiences of error treatment and corrective feedback during their prior EFL learning)

2. Attitudes toward error and corrective feedback

3. Bi- or multilingual status

4. Previous knowledge of the construct of interlanguage and its implications for language learning/teaching

This interview, like all others, was designed and piloted using the Interview Protocol Refinement Framework (IPRF; Castillo-Montoya, 2016). Through content analysis of these interviews, the participant profiles in Table 1 were constructed. “Deficit” means a participant views errors and corrective feedback from a deficit perspective due to their prior experience. “Developmental” means that their views and prior experience are more aligned with recommended practices regarding interlanguage development and corrective feedback. “Mixed” means that the participant reported both types of views and experiences. “Maya” or “Belizean English” means the participant is bilingual in Spanish and one of these. “No” means they speak only Spanish. None of them knew anything about interlanguage or related constructs.

Data Collection

Data were gathered during the 5-week instructional unit focusing on interlanguage, using three semi-structured interviews designed according to phenomenographic guidelines (Åkerlind, 2005a, 2005b, 2024; González-Ugalde, 2014; Hajar, 2021) and the IPRF (Castillo-Montoya, 2016). Non-participant online ODF observations were also used, and the five written ODFs were collected by downloading them from Moodle. The interviews took place at three different moments: one during the second week, another during the fourth week, and the last one immediately after the fifth week. They were conducted face-to-face in Spanish either in the classroom or in the professor’s office. This yielded 48 interviews. The written ODFs were used as prompts to bring participants’ conceptions to consciousness and to build relational understandings of experience between the interviewer and the interviewee (Åkerlind, 2024).

Data Analysis

I conducted the analysis, and my supervisor audited it. The categories emerged from the data using a discovery approach first, followed by a more constructive approach when naming the categories (i.e., using theoretical labels; Walsh, 2000). Below are the analytic steps:

1. I transcribed the interviews and read them entirely, segmented the ODFs into messages, and assigned codes to each one.

2. I randomly selected four participants’ interviews (12 interviews) and their messages in the five ODFs. I read them, manually noting similarities and differences in the ways they talked about the same concepts and keeping handwritten memos in a notebook. This resulted in a preliminary book code, which was then applied to the same interview transcripts in MAXQDA© v. 2018, yielding a first set of codes. These codes were presented at a doctoral seminar. Upon receiving feedback, changes were made accordingly. This is a type of phenomenographic validity (Åkerlind, 2005b).

3. The supervisor applied the resulting codes to two unanalyzed interviews and randomly selected ODF segments. This auditing is the main type of validity in phenomenographic research (Walsh, 2000). She found almost the same codes, but some new ones were negotiated in dialogue with me, which conferred dialogic reliability (Hajar, 2021).

4. I applied the new coding scheme to the remaining interviews and ODFs. Through constant comparison and reflection, conceptions were grouped into categories in MAXQDA© v. 2018, and an initial outcome space was produced as a Microsoft Word© document. I used the threshold-concepts literature to name the categories.

5. The supervisor read 12 randomly selected interviews and segments of ODFs to verify the applicability of the outcome space. Refinements to the outcome space were negotiated between the supervisor and the author through dialogue until it stabilized.

6. The outcome was presented at a doctoral seminar, where it received positive feedback (communicative validity, Åkerlind, 2005b).

7. Embedded units consisting of information-rich cases (Patton, 2015) that illustrate conceptual change and the outcome space were identified and extracted by re-reading the whole data set.

The next section presents the results. First, the outcome space is shown, followed by a chronological analysis of the evolving conceptions of selected, information-rich cases.

Team 1

The participants in this team run the gamut of achieved conceptions, with P6 and P8 in no crossing, P4 in threshold crossing, and P9 in liminality. There were important contrasts between their apprenticeship of observation and initial stances. P4 experienced a developmental apprenticeship of observation in high school, as shown in the following excerpt. Due to space limitations, only English translations of excerpts are presented:

Since I was in high school, the teacher would tell me that sometimes we make grammar mistakes that aren’t as important in communication; some are more serious than others, of course, and this idea stuck with me: that what’s important is communication. (P4, Interview 0)

After ODF 1, “Feelings about correctness,” P4 also reflected on the power imbalances involved in imposing standard varieties and the need to adopt a descriptivist stance toward speakers of contact Spanish. Although he is not bilingual, he grew up in a rural community with Maya speakers:

I think corrective feedback is okay sometimes, and sometimes it is not. The Maya people were forced to learn Spanish, and this involved discrimination. I think we have to respect the way they speak Spanish; in a way, it’s about taking a humanistic approach. (P4, Interview 1)

P4’s experiences and initial stance contrasted with those of his teammates. For example, P6 experienced a deficit-oriented apprenticeship of observation that involved punishment for mistakes: “In high school, we answered English exercises on computers, and if we made a mistake, they forced us to start all over again from the beginning, and we couldn’t leave for lunch break until everything was completed perfectly” (Interview 0).

During Interview 1, P6 also resisted the professor’s descriptivist stance toward non-standard varieties of English, which is evidence of the conflicts that arise with grappling with threshold concepts that clash with the apprenticeship of observation:

We need to speak correctly. When I saw that the professor wrote examples without the third-person s, I thought, “How can that be right?” I mean, we’ve been taught all the time that it’s wrong, even our grammar teachers say it’s wrong.

P8 also showed a strong adherence to prescribed standard norms coming from her apprenticeship of observation, which also caused conflict: “In all my English courses, we have always been taught that what is in the books is the way things must be done, like writing or conjugating verbs. How come native speakers decide to speak the wrong way?” (P8, Interview 1).

Interestingly, during ODF 2, “First concept connection activity,” where participants had to define interlanguage and connect it to several other terms, P4 showed an inaccurate, partial understanding of interlanguage as “a system that helps us learn an L2 by realizing what’s right and what’s wrong” (P4, ODF 2, Message 2). This was corrected by P6, who wrote the following complex and integrated understanding of interlanguage:

What I’m going to write is my opinion about how these terms can be defined and related. Interlanguage is the linguistic system of an L2 learner. I think it is a very personal version of the language based on the rules in each learner’s mind. Many factors are involved in the process of interlanguage, like the input you get, whether you learn the L2 as a second or foreign language, and the result of this process, which can be target-like or non-target-like forms.

During Interview 2, P6 expanded on this definition:

In my opinion, interlanguage is like a new version of the language, one’s own individual version, and for that reason, we don’t all have the same level of the language. And this is very stressful for me because there are many related concepts in interlanguage, but they all have their own meanings, and it’s difficult to learn them all.

Despite her difficulties, the contribution of P6 was instrumental in helping her teammates understand interlanguage: “At the beginning, I had a lot of questions about what interlanguage was. But thanks to P6, who wrote an accurate definition, I was able to write the other concepts and relate them to one another” (P4, Interview 2). This illustrates the role of collaboration in conceptual change.

In ODF 3, these participants did not show an accurate understanding of descriptivism and how it may relate to interlanguage development:

About interlanguage development, we may say there are two branches to this process: target-like forms and non-target-like forms. The former are correct ways of speaking, taking established rules into account, and the latter are deviations from the norm, incorrect ways of using the rules of a language. This whole process of interlanguage happens when a person is learning a new language and draws on their L1 to speak the L2. We can define two schools of thought when it comes to language learning. The first is prescriptivism, which is the way of teaching in schools, that is, following the grammar rules in books. For example, in a test, students must write verb tenses and words correctly, or they are penalized. The focus is on the correct way of speaking. Descriptivism is the opposite, as it allows us to express ourselves freely without following book rules. (ODF 3, Collective concluding paragraph)

My interview with the professor and my class observations showed that he was actually trying to get preservice teachers to distance themselves from the “correct vs. incorrect” dichotomy. He also did not define descriptivism as simply “speaking without rules” but rather as an inquiry-based attitude to non-standard forms that seeks to understand how and why they are produced for didactic and critical purposes, that is, to subvert the imposition of deficit views and the power imbalances and oppression inherent to native speakerism. Nevertheless, P6 and P8 did not understand this and, instead, adhered to prescriptivism: “Prescriptivism tries to follow the rules of a language, and it tries for people who speak this language to follow the rules. And prescriptivism is easier to understand because it’s the opposite of descriptivism” (P6, Interview 2).

By contrast, P4 did reflect on the connections between prescriptivism and oppression:

Prescriptivism is only about saying that something is right or wrong, without understanding why people speak that way. Prescriptive rules are based on the ways of speaking of those with more power or more social and economic prestige. But the fact that someone thinks the way I speak is wrong doesn’t make it wrong. (Interview 2)

The prescriptivist stances and received deficit conceptions of P6 and P8 influenced their participation in ODF 4, an assessment task intended to be self-referential (Larsen-Freeman, 2014). In this task, the professor provided a simulated vocabulary test on past and present tenses that a hypothetical learner answered using target-like semantics and syntax, combined with some non-target-like morphological marking of tense, aspect, and/or subject-verb agreement. The test involved selecting the right verb and tense/aspect to complete sentences. The prompt asked participants to grade the test based on what they had learned about interlanguage development. The prompt suggested using decimals (i.e., .25 or .5 or .75) instead of a full point to reward the learner for what they could do. P6 and P8 were unanimous in awarding 0 points to all the answers: “To the answer to Question 1, ‘I am walk home today because my car broke down’ I give 0 points because the verb should have been in the present continuous with -ing” (P6, ODF 4, Message 2).

By contrast, P4 considered awarding decimals because the participants’ answers made sense: “In Answer 4, I would consider giving .5 because the main verb is in the past participle. They just didn’t conjugate the auxiliary verb correctly” (ODF 4, Message 13).

Well, guys, it looks to me that we’ve only evaluated grammar, but what about semantics? I think it’s important to evaluate this as well. As we can see, he has used each verb well. He knows what each one means. I would give him a 10 in semantics (P4, ODF 4, Message 20).

The other participants replied that the answers had to be perfect and, since they were not, it was inappropriate to award decimals: “For Number 8, I also thought about awarding .5, but I believe that, as in 5, both the main and the auxiliary verb must be correct, so the verb tense is correct. My grade is 0” (P6, ODF 4, Message 33, italics in the original text). These conceptions aligned with their deficit-oriented apprenticeship of observation and views of their current grammar teacher.

After ODF 4, the professor asked students to engage in a new discussion, ODF 5, where they had to reflect on their grading. P4 gave an extended, well-argued explanation that shows he crossed the threshold:

The student knew the meanings of the verbs very well, and he knew word order because he placed the auxiliary verb before the main verb. He couldn’t conjugate them well, but he clearly has a structure in his brain about how to speak. That’s why, to me, it was unfair to grade him with 0 points because clearly there is a structure in his brain about how to use the tenses. (ODF 5, Message 4).

I see evidence of crossing the threshold because the answer displays an awareness of interlanguage as a developmental process and, consequently, an embracement of self-referential assessment. In ODF 5 and Interview 3, P8 remained adhered to prescriptivist views and showed no conceptual change, adhering to his apprenticeship of observation:

I was able to notice all the grammar mistakes he made, and this is basic to me. I don’t think it’s right to make these mistakes, but I liked the activity because I could show my knowledge, what I have always been taught.

By contrast, P6 expressed openness to considering P4’s perspective, which is evidence of the liminality category. P6 shows incipient reflection thanks to collaboration:

Something that helped me learn from the ODF was P4’s last message, where he said that we were only taking grammar into account, but we should consider that the student’s sentences communicated something meaningful, so we should assess that too. We didn’t think that way before, but it’s true. (Interview 3)

Team 2

This section focuses on P2, who most clearly illustrates conceptual change. During Interview 1, following ODF 1, P2 expressed a strong prescriptivist stance against non-standard varieties: “To me, saying “haiga” is incorrect. My whole life, I have been taught what’s correct, and that is completely incorrect.”

However, by the second interview, after ODFs 2 and 3, P2’s conceptions had begun to change in connection with a more accurate understanding of descriptivism:

I used to say that being prescriptivist was better, but now I understand descriptivism better. I used to think that all the rules I’ve been taught during my studies are always better because they are established in books, and they must be followed. But now I realize that when teachers are like that, they tend not to explain anything when someone makes a mistake. They just say, “It’s wrong,” and that’s it. That happened to me in middle school and high school when I was learning English. The teachers would just say “that’s wrong,” but didn’t explain why. And maybe that’s why I began to hold this belief that, well, this is how it’s done, this is well done. But now I’ve realized it isn’t. And if I’m going to be a teacher in the future, it’s better to be a descriptivist. Because that’s how people understand and learn more.

P2 experienced oscillation during ODF4, where she refused to reward the hypothetical learner’s developmental gains. However, during ODF 5 and Interview 3, she distanced herself from that stance and restated her change in views:

Thanks to the class, we realized we had to see things from the learner’s perspective, that he did show some learning, and that we had to acknowledge it instead of just saying no, no, no. I remembered that something similar happened to me on a test, with a grade. So, I changed my perspective on things.

P2 became very critical of her deficit-oriented apprenticeship in observation, reflecting on how it had affected her as a learner and how she had come to accept, irreflexively, that it was the way things were done in English teaching (a deficit-oriented dimension of local professional identities). Her reference to negative feedback and testing experiences evinces reflection on prescriptivism and power imbalances. It seems that these class-prompted reflections promoted conceptual change and led her to cross the threshold.