A cross-sectional observational study was conducted ([18]), through the administration of a structured questionnaire with closed and open-ended questions, administered at a single point in time to the population of interest. This type of instrument constitutes a widely used tool for sociolinguistic data collection ([10]).

For the purposes of this study, mother tongue was operationally defined as the first language learned in childhood, at home or within the community, based on direct self-report (“What is your mother tongue: the first language you learned in childhood, at home or in your community?”), following the operational approach to mother tongue used in self-report language questionnaires such as the Language and Social Background Questionnaire ([2]). An Indigenous-language speaker was defined as any participant who reported an Indigenous language (Náhuatl, Totonac, or another Indigenous language of the region) as their mother tongue under this definition. The “gap” examined throughout this study refers specifically to the discrepancy between Indigenous ethnic self-identification (a separate, single-choice item) and this mother-tongue variable; it does not refer to general language proficiency or current language use, which were measured through separate instrument items (Section 2.3).

Undergraduate students of the Northeastern Regional Complex (CRN) of the Benemérita Universidad Autónoma de Puebla (BUAP)—comprising the campuses of Cuetzalan, Libres, Teziutlán, and Zacapoaxtla—were invited to participate through in-person outreach in classrooms, with voluntary and anonymous participation.

A total of 747 people accessed the questionnaire. Of these, 26 (3.5%) explicitly declined to participate by answering “No, I do not wish to participate” at the initial consent question. The remaining 721 people (96.5%) gave their consent and completed the questionnaire, constituting the analytic sample for this study. The distribution across campuses was uneven: Teziutlán (n = 602; 83.5%), Libres (n = 96; 13.3%), Zacapoaxtla (n = 22; 3.1%), and Cuetzalan (n = 1; 0.1%). Comparative analyses focus on the three campuses with analytically sufficient representation: Teziutlán, Libres, and Zacapoaxtla. The sample spanned 13 academic programs, concentrated mainly in the early semesters (2nd and 4th) and among students who identify as women.

The “Mapa Lingüístico CRN-BUAP” questionnaire was applied, designed for this study based on self-report language questionnaires widely used and validated in research on bilingualism and language use ([2]; [19]). The instrument focused on two reference languages, mother tongue and current language of greatest use, following the logic of instruments such as the Language and Social Background Questionnaire (LSBQ) by [2].

The instrument included, among others, the following closed-ended items analyzed in this study: Indigenous self-identification (“Do you self-identify as Indigenous?”; response options: Yes/No), mother tongue (single-choice: Spanish/Indigenous language [specify]/English/Other), current linguistic repertoire (multiple-choice checklist of languages understood, spoken, read, and written), experiences of linguistic discrimination (multiple-choice, context-specific items for schooling before university and within the CRN-BUAP), agreement with statements on institutional linguistic hierarchies (five-point Likert scale: Totally disagree to Totally agree), and, for Indigenous-language speakers, perceived replacement of the mother tongue by Spanish (single-choice: Yes, to a great extent/Yes, to some extent/No/Not sure). Participants who selected more than one language in checklist items were coded as reporting each selected language; incomplete responses to a given item were treated as missing data for that item and excluded from the corresponding analysis, without imputation. The questionnaire also included open-ended items (e.g., inviting participants to describe experiences of discrimination or to explain their perception of language replacement in their own words); these open-ended responses are not analyzed in the present study, which reports findings from the closed-ended items only, and are reserved for a separate qualitative analysis.

Outreach for the instrument was carried out in person in classrooms across the CRN’s four campuses. The questionnaire was administered exclusively online, through Microsoft Forms, during May 2026. Participation was voluntary, anonymous, and confidential.

Given the size and composition of the sample, the analysis was conducted using descriptive statistics: frequencies and percentages for categorical variables, and means/response distributions for Likert-type scales ([23]). No statistical inference tests were applied, since the subgroups of interest did not meet the minimum size recommended for reliable use.

The questionnaire included two mother-tongue items: an initial direct question (used as the primary variable throughout this study, per Section 2.1) and a closing confirmation question with identical wording. Of the 31 participants who reported an Indigenous language as their mother tongue on at least one of these two items, 12 (38.7%) did so consistently on both, while 19 (61.3%) gave a different response between the initial and confirmation items. This inconsistency is discussed as a substantive finding in Section 4 rather than treated as measurement error. All analyses in this study use the initial mother-tongue item as the single analytic denominator (n = 20 Indigenous-language mother-tongue speakers overall, corresponding to 16 self-identified Indigenous participants and 4 who did not self-identify as such); the confirmation item was used exclusively for this consistency check and not to redefine subgroup membership elsewhere in the study.

Participants were presented with informed consent, following principles of respectful research that is culturally appropriate for Indigenous peoples and ethnic groups ([8]). Participation was voluntary, anonymous, and confidential.

Of all participants (n = 721), 74 (10.3%) self-identified as Indigenous. Cross-referencing this variable with reported mother tongue showed that, of these 74 cases, 58 (78.4%) indicated Spanish as their mother tongue, while only 16 (21.6%) reported an Indigenous language (Table 1). Among those who did not self-identify as Indigenous (n = 647), 642 (99.2%) reported Spanish as their mother tongue and 4 (0.6%) an Indigenous language.

Table 1. Indigenous self-identification by reported mother tongue (n = 721).

Source: the authors.

Regarding the Indigenous people to which participants belonged, among those who self-identified as Indigenous (n = 74), 75.7% (56/74) identified as Nahua, 18.9% (14/74) as Totonac, 4.1% (3/74) as Mixtec, and 1.4% (1/74) as Triqui. Among those who reported an Indigenous language as their mother tongue (n = 20, using the initial mother-tongue item as the analytic denominator; see Section 2.5), 65.0% (13/20) corresponded to Náhuatl and 35.0% (7/20) to Totonac (Table 2). Regarding gender, the distribution within both subgroups was similar to that of the overall sample: 71.6% (53/74) of those self-identified as Indigenous and 75.0% (15/20) of those reporting an Indigenous mother tongue identified as women.

Table 2.Ethnic and gender profile of the Indigenous subpopulation.

Source: the authors.

Overall, 66.0% (476/721) of participants reported speaking a single language, 30.4% (219/721) two languages, and 3.6% (26/721) three or more languages (Table 3). Regarding the specific languages mentioned by participants, Spanish was reported by 87.5% (631/721) and English by 32.7% (236/721), making the latter the second most frequent language in the sample’s repertoire, well above any Indigenous language. Among Indigenous languages, Náhuatl was the most frequently reported (6.0%; 43/721), followed by Totonac (1.5%; 11/721) and, in a single case, Triqui (0.1%; 1/721). Other foreign languages were reported less frequently: Portuguese (1.0%; 7/721), Italian (0.8%; 6/721), Korean (0.7%; 5/721), French (0.6%; 4/721), German (0.4%; 3/721), Japanese (0.3%; 2/721), and, in one case each, Chinese and Thai (0.1%; 1/721 each).

Table 3.Self-reported linguistic repertoire (n = 721).

Source: the authors.

The proportion of Indigenous self-identification varied notably across the three campuses analyzed: Teziutlán (54/602; 9.0%), Libres (10/96; 10.4%), and Zacapoaxtla (10/22; 45.5%). Examining, within each campus, the proportion of self-identified Indigenous participants who reported an Indigenous mother tongue, differentiated patterns were found: in Teziutlán, 11 of 54 (20.4%); in Libres, 0 of 10 (0.0%); and in Zacapoaxtla, 5 of 10 (50.0%). In this context, the proportion of students who reported having experienced or observed some form of linguistic discrimination within the CRN-BUAP also varied by campus: 52 of 602 (8.6%) in Teziutlán, 6 of 96 (6.2%) in Libres, and 3 of 22 (13.6%) in Zacapoaxtla (Table 4).

Table 4.Indigenous self-identification, Indigenous mother tongue, and linguistic discrimination, by campus.

Source: the authors.

Participants (n = 721) were asked about their level of agreement with four statements related to the status of Indigenous languages at the CRN-BUAP. 80.9% (583/721) agreed or strongly agreed that the institution has respected the cultural and linguistic diversity of the student body. However, 47.0% (339/721) also agreed or strongly agreed that Spanish is perceived as the only “appropriate” language in most academic and administrative spaces and processes. Likewise, 80.9% (583/721) considered that valuing and making Indigenous languages visible should be normalized, and 79.8% (575/721) supported the institution offering specific support for Indigenous-language-speaking students (Table 5).

Table 5.Level of agreement with statements on institutional linguistic ideologies and hierarchies (n = 721).

Source: the authors.

While the previous section presented linguistic discrimination broken down by campus, this section describes the same variable for the sample as a whole (n = 721), distinguishing when it occurred and the type of experience reported. 5.7% (41/721) of participants reported having experienced discrimination related to language during their schooling prior to university. Within the CRN-BUAP, 1.7% (12/721) reported having experienced it personally and 6.8% (49/721) having observed it in others; 16.1% (116/721) indicated they were not sure, and 75.4% (544/721) reported not having experienced or observed it (Table 6).

Table 6.Reported experiences of linguistic discrimination (n = 721).

Source: the authors.

Among participants who reported an Indigenous language as their mother tongue (n = 20), 15 (75.0%) responded to this item. Of these, 53.3% (8/15) perceived that Spanish has replaced their mother tongue “to a great extent”, and 13.3% (2/15) “to some extent”; overall, 66.7% (10/15) reported some degree of replacement. 33.3% (5/15) indicated they perceive no replacement, and none selected “not sure” (Table 7).

Table 7.Perceived replacement of mother tongue by Spanish, among Indigenous-language speakers who responded to this item (n = 15 of 20).

Source: the authors.

This finding can be considered within the broader national context, with an important conceptual caveat. The national-level INEGI (2020) figure of 69.2% refers to the proportion of self-identified Indigenous people who do not speak an Indigenous language—a measure of current language ability—whereas the 78.4% figure reported here refers specifically to mother tongue, a related but conceptually distinct construct (see Section 2.1). These two indicators are therefore not directly comparable, and this study’s finding should not be interpreted as indicating a larger gap than the national one; rather, both figures independently point in the same direction—a substantial disconnect between Indigenous self-identification and the ancestral language, whether measured as current speaking ability (national level) or as mother tongue (this study). One possible explanation for the gap observed in the present sample may lie in the university context: entering higher education entails, for many Indigenous students, going through schooling trajectories conducted mostly in Spanish, which may have intensified language shift relative to the general population of the country. This finding aligns with the idea that ethnic identity can be maintained and sustained through pathways other than linguistic competence ([20]; [22]; [28]), and with the evidence that Indigenous self-identification in Mexico is a trait that changes over time and across different contexts ([14]). From this perspective, this gap could be regarded as evidence that, among the student community of the CRN-BUAP, ethnic belonging is defined predominantly through bonds that are not limited to language and may be associated with family ancestry, community memory, or identification with an ethnic group.

The inconsistency observed between the initial and confirmation mother-tongue items, in which 19 of 31 participants (61.3%) who reported an Indigenous language on at least one of the two items answered differently between them (Section 2.5), is consistent with this broader pattern of fluidity rather than representing a simple measurement artifact. If Indigenous self-ascription itself is a dynamic, context-dependent trait ([14]), it is plausible that reporting an Indigenous language as one’s mother tongue is similarly sensitive to how the question is framed, to the point in the questionnaire at which it is asked, or to momentary shifts in how participants position themselves linguistically. This possibility does not undermine the validity of the present findings, which rely on a single, consistently applied item as the analytic denominator (Section 2.5), but it does suggest that mother-tongue self-report among this population may be less stable than typically assumed, and that this instability is itself a phenomenon worth investigating in future research, ideally through repeated measurement over time.

That English is far more present than any Indigenous language in the self-reported linguistic repertoire of the participating student community is a relevant finding. It is important to distinguish this repertoire measure from the institutional-norm items reported in Section 3.4, which reflect students’ perceptions of language expectations at the CRN-BUAP rather than their own language repertoire. Considered together, these two distinct measures suggest, though do not by themselves constitute direct proof of, a linguistic hierarchy consistent with that documented in higher-education institutions elsewhere in Latin America ([27]) and specifically in Mexico ([9]). In this regard, Despagne notes that in Mexico it is easier and more common to find English courses than Indigenous-language courses within educational institutions themselves, which is consistent with the repertoire data collected in this study. This does not imply that English is directly displacing Indigenous languages—both occupy different positions within the pattern observed—but the combined repertoire and institutional-norm findings are consistent with symbolic prestige and the availability of formal instruction systematically favoring English over Náhuatl, Totonac, or any other language native to the region. However, it is worth noting that the Medicina General y Comunitaria program is the only one at the CRN-BUAP that offers Náhuatl (two courses) and Totonac (two courses) courses as part of students’ basic training, with the aim that they can communicate adequately with the communities where they might practice.

Among the three campuses, Zacapoaxtla concentrated both the highest proportion of Indigenous self-identification (45.5%) and the highest proportion of Indigenous mother tongue (50.0%), while Libres, with lower self-identification (10.4%), recorded no Indigenous mother-tongue retention whatsoever (0.0%). Teziutlán, however, introduced a relevant particularity: despite having the lowest Indigenous self-identification of the three campuses (9.0%), its mother-tongue retention (20.4%) is higher than that of Libres, indicating that both indicators reflect different community contexts in each locality and do not move proportionally across campuses, despite belonging to the same region. This heterogeneity among nearby localities has been documented in other contexts in Mexico, in which intergenerational transmission at the family level explains why nearby communities can diverge in the vitality of their Indigenous language beyond mere geographic proximity ([6]). The case of Santiago Tlaxco, in the state of Puebla, offers a relevant parallel: there, Náhuatl remains vigorous among older generations while younger people increasingly use Spanish ([11]), a pattern of generational shift that could also be occurring in Zacapoaxtla, although the data from this study do not allow this to be confirmed directly. It is worth emphasizing that, given the small size of the sample subgroups, these findings should be treated as initial and exploratory, and can serve as a starting point for in-depth research with greater coverage across the three campuses.

This result reproduces, almost literally, the tension between discourse and practice that [27] identify in Colombian language policies: a tension in which a widely endorsed discourse of respect for diversity coexists, within the same institutional space, with a practical expectation of exclusive use of the dominant language, even though this reality is not consistent with the discourse of formal policies. The fact that students perceive that the CRN-BUAP respects cultural diversity while, at the same time, Spanish is expected as the only “appropriate” language in most academic and administrative spaces and processes does not necessarily indicate a contradiction, but rather that different dimensions of the phenomenon may coexist: a declarative dimension, visible in official discourse and in general attitudes toward linguistic diversity, and a pragmatic dimension, present in day-to-day experiences of which language is functional and valid in academic and administrative settings. This pattern takes on particular significance when one considers that the CRN-BUAP is an educational institution without a specific intercultural origin, whose academic offerings and structure were not designed around Indigenous linguistic diversity—that is, it is not an intercultural university per se ([24]). In this context, the gap between attitude and practice identified in this study could reflect a structural condition of Mexican universities that do not identify as intercultural yet also receive and serve Indigenous populations without having been conceived specifically for that purpose. Nonetheless, making this phenomenon visible can be a starting point for institutional discourse on cultural diversity to align more closely with the reality experienced day to day in academic and administrative spaces and procedures.

The group that responded “not sure” (16.1%) deserves its own analysis, distinct from a mere information gap. [25] described how linguistic discrimination can become normalized and embedded to the point of being woven into everyday life such that, at some point, it stops being perceived as a clear or explicit discriminatory act, even when it has a real effect on Indigenous-language speakers. From this perspective, the uncertainty reported by one-sixth of the sample’s participants may reflect difficulty recognizing discriminatory experiences when they occur in subtle forms, rather than their absence, since such experiences can be very subtle and may take the form of a simple comment about accent, an instinctive correction of pronunciation, or a joke that becomes normalized among peers.

This type of subtle linguistic discrimination can lead to the disuse of an Indigenous language in common spaces such as the university, causing speakers of these languages to be regarded as “language deniers” on the assumption that they lack interest or feel ashamed of their origin, without recognizing that this disuse could be related to prior social pressure, consistent with [17] findings on how silence or disuse of an Indigenous language in the classroom can be misinterpreted. If some of this dynamic is also at play at the CRN-BUAP, the 16.1% who reported uncertainty could represent, in practice, a rate of normalized discrimination higher than the 8.5% who report it explicitly.

That 66.7% (10 of the 15 respondents who answered this item) of those who speak an Indigenous language as their mother tongue perceive that Spanish has replaced their language is, perhaps, the study’s most eloquent finding, because it captures language shift as a conscious experience of the speakers themselves rather than as an external figure calculated by researchers. [7] distinguishes between “language panic” and “language pride” as competing subjective forces within a shift process; the proportion found here suggests that, for the majority of Indigenous-language speakers at the CRN-BUAP who responded to this item, the former outweighs the latter. This pattern coincides with what [12] found in Santiago Tlaxco, Puebla, applying the “ecology of pressures” model developed specifically for the shift of Mexican Indigenous languages toward Spanish ([29]): Náhuatl remains strongly used among older generations but loses ground to Spanish among young people. That a similar pattern emerges at the CRN-BUAP, with a different instrument and population, supports the idea that this reflects a broader regional dynamic that also runs through the Sierra Norte de Puebla, rather than an isolated particularity of a single community.

Taken together, these findings describe the following linguistic landscape for Indigenous languages at the CRN-BUAP: Indigenous identity persists strongly among the student body even as the corresponding mother tongue has begun to fall into disuse or has already been lost; this loss is unevenly concentrated across campuses and is not uniform across the region; there is an institutional linguistic hierarchy that favors Spanish and, increasingly, English; forms of linguistic discrimination exist that are not always recognized as such, yet are consciously experienced by those who still speak the Indigenous language, and which could lead to a process of active replacement. This pattern responds directly to the research gap identified in the literature on Indigenous university students in Mexico, concentrated so far in a limited number of institutions and regions ([26]), and contributes empirical evidence situated in a region of the Sierra Nororiental of Puebla that had until now remained absent from that body of knowledge.