Counter-arguments and replies

Summary of anticipated objections to the premise that CAPES-7 programs across fields should show comparable levels of international journal placement, and why each fails.

1. Object-language coupling

Counter: Linguistics/Letras research is inseparable from Portuguese — you can’t study Brazilian literature or Portuguese phonology in English without losing something essential.

Reply: By that logic, a study of a German author has no reason to be published in English. But Germanists publish in English routinely, because the audience is not limited to speakers of the object language. A syntactician studying clitic placement in Brazilian Portuguese is contributing to universal grammar. A literary scholar working on Machado de Assis is contributing to narrative theory, postcolonial studies, world literature. The language of the object does not constrain the language of the paper.

Status: Does not hold.

2. Local relevance

Counter: Some research is inherently regional — a historian studying a municipality in Piauí, a sociolinguist documenting a local dialect. Publishing in Portuguese is the natural choice because the audience is local.

Reply: This concedes that the work is only relevant to scholars in Brazil, which is a self-destructive characterization of one’s own research. A historian of colonial Piauí is (or should be) contributing to comparative colonialism, economic history, demographic patterns. Claiming local-only relevance isn’t a defense of publishing in Portuguese — it’s an inadvertent argument that the research is parochial. CAPES-7 is supposed to certify the opposite.

Status: Does not hold.

3. Domestic prestige economy

Counter: The high-prestige venues in Linguistics/Letras are Portuguese-language journals. Publishing in English means accepting a lower-status venue for a smaller readership within the actual peer group.

Reply: No field has its highest-status journals in Brazil. Everyone — physicists, economists, computer scientists — knows the top journals are international. This is not specific to any area. The prestige hierarchy is global and the same career incentives apply everywhere.

Status: Does not hold.

4. Reviewer competence

Counter: English-language journals lack reviewers who can competently evaluate work on Brazilian topics. Researchers publish in Portuguese because that’s where the review process works.

Reply: Top journals routinely invite specialist reviewers across countries and languages — it is the editor’s job. In Linguistics specifically, Brazilian researchers are well-represented on editorial boards and reviewer lists of leading international journals (Language, Phonology, Natural Language & Linguistic Theory, etc.). Anyone with submission experience in the field knows this.

Status: Does not hold.

5. SJR/Scopus coverage bias

Counter: The study uses SJR indexing as a proxy for “international placement,” but SJR reflects Scopus coverage decisions (curated by Elsevier), which systematically undercover humanities and non-English journals. The 12% indexing rate for Linguistics might measure Elsevier’s coverage gaps, not the field’s actual global visibility.

Reply: This is a real limitation of the metric. Partially addressable by replicating with alternative indexing sources (Web of Science, DOAJ) to show the pattern holds across databases. Also mitigated by the fact that the gap is so large (12% vs. 73%) that even substantial coverage bias cannot fully account for it.

Status: Survives as a limitation to acknowledge. Does not invalidate the finding.

6. CAPES-7 is field-relative

Counter: CAPES-7 is awarded by area-specific committees using area-specific criteria. It certifies excellence relative to a field’s own standards, not cross-field equivalence on any particular output dimension. Comparing international placement across fields assumes the rating promises something it doesn’t.

Reply: CAPES-7 is the highest possible score and is meant to signal programs of international excellence. If the rating is purely field-relative and carries no cross-field implications for international visibility, then its value as a signal to funding agencies, international partners, and prospective students is diminished. The gap documented here is relevant precisely because it tests what “international excellence” means in practice across evaluation areas.

Status: Survives as a framing issue. Strengthens rather than undermines the paper’s contribution — the finding reveals that “excellence” means structurally different things across CAPES areas.

7. Comparison group selection

Counter: The 12 study areas include mostly STEM and quantitative social science fields where >90% of output is in English. Including only “journal-heavy” CAPES-7 fields creates a comparison group engineered to make Linguistics look anomalous. If other humanities fields (Philosophy, Education, Arts) were included, the gap would narrow.

Reply: Addressable methodologically — the comparison group can be expanded with --all-areas and the analysis re-run. However, the current selection is defensible: it restricts to fields where journal articles are the primary output, making the comparison fair on the dimension being measured.

Status: Addressable. Not a fundamental objection.