Academic Portuguese is the most controlled register in the language — the place where the spoken-vs-written gap is widest and where every colloquial reflex must be suppressed. Its governing principle is impersonality: the author works hard to disappear, so that claims appear to issue from the data and the field rather than from a person with opinions. Where a casual speaker says eu acho que ("I think that"), the academic writer says observa-se que ("it is observed that") or os dados sugerem que ("the data suggest that"). This page shows how that effacement is achieved, layer by layer.
If you have written academic English, much of the instinct transfers — hedge your claims, attribute sources, prefer nouns to verbs, avoid "I think." But BR pushes the impersonality further and grammaticalizes it through the impersonal/passive se and through connectors with no everyday English equivalent. The result is denser and more agentless than typical academic English.
Layer 1: Impersonality — making the author vanish
The signature move is the impersonal/passive se construction. Instead of "I observe / we conclude," academic BR writes observa-se ("it is observed"), conclui-se ("it is concluded"), verifica-se ("it is verified"), constata-se ("it is found"), nota-se ("it is noted"), percebe-se ("one perceives").
Observa-se que a variável apresenta correlação significativa com os resultados.
It is observed that the variable shows a significant correlation with the results.
Conclui-se, portanto, que a hipótese inicial não se sustenta.
It is therefore concluded that the initial hypothesis does not hold.
A second strategy makes the text or study the grammatical subject, so the human author still doesn't appear: este trabalho analisa, o presente estudo investiga, este artigo propõe.
O presente estudo analisa o impacto da política monetária sobre o consumo das famílias.
The present study analyzes the impact of monetary policy on household consumption.
Este trabalho propõe uma nova abordagem para o problema da segmentação de imagens.
This work proposes a new approach to the image-segmentation problem.
A third option, in some fields and traditions, is the authorial nós ("we") even for a sole author — a convention of modesty, not literal plurality. Whichever you choose, never eu acho, never a gente.
Layer 2: Hedging — claims are never bald
Academic BR almost never asserts flatly. Findings are softened with hedges that mark the strength of the claim: os dados sugerem que ("the data suggest"), tende a ("tends to"), pode-se inferir que ("it may be inferred"), aparentemente ("apparently"), é possível que + subjunctive, parece indicar ("seems to indicate").
Os resultados sugerem que o tratamento tende a reduzir os sintomas, embora sejam necessários mais estudos.
The results suggest that the treatment tends to reduce symptoms, although further studies are needed.
É possível que fatores externos tenham influenciado a amostra.
It is possible that external factors influenced the sample.
Note que tenham influenciado — hedging frequently pulls in the subjunctive, because hedged claims are by definition non-factual. Precision with the subjunctive is a hallmark of the register: é possível que haja, recomenda-se que seja feito, espera-se que os resultados se confirmem.
Layer 3: Source attribution
Claims that aren't the author's own are explicitly attributed using a small set of formal markers: segundo ("according to"), conforme ("in accordance with / as per"), de acordo com ("according to"), com base em ("based on"), para + author ("for [Author], …"), and reporting verbs afirma, sustenta, argumenta, defende, aponta.
Segundo Bourdieu (1989), o capital cultural reproduz as desigualdades sociais.
According to Bourdieu (1989), cultural capital reproduces social inequalities.
Conforme apontam estudos recentes, a tendência se manteve estável na última década.
As recent studies indicate, the trend remained stable over the last decade.
Para a autora, o conceito deve ser reinterpretado à luz dos novos dados.
For the author, the concept must be reinterpreted in light of the new data.
Layer 4: Nominalization and dense phrasing
Academic BR converts verbs and adjectives into nouns (nominalization), packing more content into noun phrases and reinforcing the agentless feel. Instead of quando os preços aumentam ("when prices rise"), it writes o aumento dos preços ("the rise in prices"); instead of é importante implementar → a importância da implementação.
A implementação da política exigiu a reformulação dos procedimentos administrativos.
The implementation of the policy required the restructuring of administrative procedures.
A interpretação dos dados depende da consideração de múltiplas variáveis.
The interpretation of the data depends on the consideration of multiple variables.
Layer 5: Formal connectors
The argument is articulated with high-register connectors, several of which are essentially never spoken: por conseguinte ("consequently"), portanto / logo ("therefore"), ademais ("moreover"), outrossim ("furthermore / likewise"), não obstante ("notwithstanding"), contudo / todavia / entretanto ("however"), com efeito ("indeed"), a saber ("namely"), dessa forma / desse modo ("in this way").
Os custos aumentaram significativamente; por conseguinte, a margem de lucro foi reduzida.
Costs rose significantly; consequently, the profit margin was reduced.
A amostra é representativa; não obstante, certas limitações devem ser consideradas.
The sample is representative; notwithstanding, certain limitations must be considered.
Layer 6: Genre conventions — resumo / abstract
A Brazilian academic paper opens with a resumo (the Portuguese abstract) and usually an English abstract, followed by palavras-chave ("keywords"). The resumo is a single dense paragraph that compresses objective, method, results, and conclusion — written impersonally and in a fixed rhetorical order.
Este artigo tem como objetivo analisar o impacto da automação sobre o emprego industrial. Utilizou-se uma abordagem quantitativa, com base em dados de painel. Os resultados indicam uma relação negativa significativa. Conclui-se que políticas de requalificação são necessárias.
This article aims to analyze the impact of automation on industrial employment. A quantitative approach was used, based on panel data. The results indicate a significant negative relationship. It is concluded that requalification policies are necessary.
Notice the moves stacked together: tem como objetivo (objective), utilizou-se (impersonal method), os resultados indicam (hedged result), conclui-se que (impersonal conclusion). That four-beat structure is the academic resumo in miniature.
Layer 7: Lexicon — technical and erudite, zero colloquialism
The vocabulary is technical and Latinate, and any colloquial element breaks the spell instantly: no a gente, no tem for há, no pra, no coisa where a precise term exists, no slang, no exclamation marks, no rhetorical né. Prefer utilizar over usar, evidenciar over mostrar, constatar over ver, elevado over alto, reduzido over pequeno.
Verifica-se uma elevada incidência do fenômeno entre os grupos analisados.
A high incidence of the phenomenon is found among the groups analyzed.
Common Mistakes
❌ Eu acho que os dados mostram que tem um problema aí.
Three register breaks — 'eu acho', colloquial 'tem', and 'aí'.
✅ Os dados indicam a existência de um problema. / Observa-se um problema nos dados.
The data indicate the existence of a problem. / A problem is observed in the data.
❌ A gente concluiu que a hipótese tá correta.
'A gente' and 'tá' are colloquial — both forbidden in academic prose.
✅ Conclui-se que a hipótese está correta. / Concluímos que a hipótese está correta.
It is concluded that the hypothesis is correct. / We conclude that the hypothesis is correct.
❌ Os resultados provam definitivamente que o método é o melhor.
Over-claiming — academic style hedges; 'provam definitivamente' is too absolute.
✅ Os resultados sugerem que o método tende a apresentar melhor desempenho.
The results suggest that the method tends to show better performance.
❌ Como disse o autor, isso é muito importante!
Vague attribution, colloquial intensifier, and an exclamation mark — all off-register.
✅ Conforme afirma o autor (2020), tal aspecto é de fundamental importância.
As the author (2020) states, this aspect is of fundamental importance.
❌ Nesse trabalho a gente vai mostrar como funciona.
Colloquial 'a gente' and the bare future 'vai mostrar'; informal throughout.
✅ Neste trabalho, demonstra-se o funcionamento do processo.
In this work, the workings of the process are demonstrated.
Key Takeaways
- Academic BR maximizes impersonality: the author vanishes behind observa-se que, conclui-se que, este estudo analisa — never eu acho, never a gente.
- Claims are hedged (os dados sugerem, tende a, é possível que
- subjunctive) and sourced (segundo, conforme, de acordo com).
- The prose is dense with nominalization, passive/impersonal se, and formal connectors (por conseguinte, não obstante, ademais).
- The resumo follows a fixed impersonal four-beat: objective → method → results → conclusion.
- A single colloquialism (a gente, tem, pra, an exclamation mark) breaks the register — academic style tolerates none.
Now practice Portuguese
Reading grammar gets you part of the way. The exercises are where it sticks — free, no signup needed.
Start learning Portuguese→Related Topics
- Academic and Educational ExpressionsB2 — The impersonal, hedged formulas that signal scholarly register in Brazilian Portuguese essays, abstracts, and academic writing.
- Formal RegisterB2 — How Brazilian Portuguese stacks up formality — o senhor/a senhora address, enclisis, erudite vocabulary, impersonal constructions, and set formulas for contracts, courtrooms, and ceremony.
- Literary StyleC1 — The devices of high literary Brazilian Portuguese — stylistic inversion, the synthetic pluperfect, mesoclisis, the atmospheric imperfect, participial reduction, and elevated lexicon.
- Se-ImpersonalB1 — The impersonal se for generic 'one/people' — trabalha-se muito, como se diz — and how it differs from the se-passive.
- Se-Passive (Sintética Passive)A2 — The passive with se plus a third-person verb that agrees with the logical object — vende-se, alugam-se — and why Brazilians often skip the agreement.
- Written vs Spoken BR PortugueseB1 — Brazil's central register axis — how spoken norms (a gente, cê/tá/pra, proclisis, invariable tem) diverge so far from formal writing (nós, full forms, há, enclisis) that learners must master both, plus the hybrid texting register.