How to Choose the Right Reporting Verb in Academic Writing (2026)

A reporting verb is the verb you use to introduce someone else’s work: Smith (2023) argues that…, Chen and Okoro (2024) found that…, Reyes (2022) claims that…. It looks like a small stylistic choice. It is not. The verb tells your examiner three things at once — what the source did, how strong its evidence was, and whether you agree with it. Choose it carelessly and you will attribute an argument to someone who ran an experiment, endorse a finding you meant to question, or make a tentative pilot study sound like settled fact.

This guide gives you a five-step procedure and a stance-graded verb bank you can use while you are writing, not after.

Step 1: Decide what the source actually did

Before you reach for a verb, name the act. Most sources in a thesis are doing one of four things, and each act licenses a different family of verbs.

• Reporting an empirical result. The authors collected data and observed something. Verbs: found, observed, measured, reported, recorded, identified.
• Making an argument. The authors reason from premises to a position. Verbs: argues, contends, maintains, posits, proposes, holds.
• Reviewing or summarising others. The authors synthesise a field. Verbs: reviews, surveys, summarises, traces, catalogues.
• Recommending or prescribing. The authors tell practitioners what to do. Verbs: recommends, advocates, urges, calls for.

The commonest error in a literature review is the category mistake: writing Nguyen (2023) argues that adherence fell by 12% when Nguyen measured adherence and reported a number. Nobody argues a percentage. Use found or reported. Conversely, Foucault found that power is productive attributes an empirical claim to a theoretical argument.

Step 2: Pick the stance band

Once you know the act, the verb chooses your position. Reporting verbs sit on a scale from purely neutral through supportive to openly critical, and readers hear that scale whether you intended it or not.

Two of these bands cause almost all the damage. Claims is not a neutral synonym for says. In academic English it carries an implicature of doubt: Weber (2021) claims that… tells the reader you think Weber may be wrong. If you did not mean that, use states or argues. And demonstrates and shows endorse. Writing Patel (2024) demonstrates that mindfulness reduces burnout means you have accepted the finding as established. If the study was a single sample of 42 undergraduates, found or reported is the honest verb.

Step 3: Match the verb to the strength of the evidence

The verb should be no stronger than the design behind it. This is the point where reporting-verb choice becomes a research-integrity question rather than a stylistic one.

• One small correlational study → found, reported, observed an association between. Never demonstrated that X causes Y.
• A randomised trial or well-powered experiment → showed, demonstrated are defensible.
• A meta-analysis or systematic review → established, confirmed, provides consistent evidence that.
• A theoretical or conceptual paper → argues, proposes, theorises. Not found.
• A single qualitative case → describes, documents, illustrates. Not proves.

The verb proves deserves its own warning. Outside formal logic and mathematics, empirical work does not prove things; it supports, corroborates or fails to reject. An examiner who sees Ahmed (2023) proves that… will read it as a signal that the writer does not understand inference. The same instinct governs how you describe your own non-significant results — absence of evidence is not evidence of absence, and the verb has to say so.

Step 4: Get the grammar around the verb right

Three mechanical points cause most of the red pen.

Tense

English academic writing uses three patterns, and style authorities differ:

• Past simple — a specific study at a specific time: Chen (2022) found that… APA 7 requires past or present perfect for reporting the results of prior research.
• Present perfect — a body of work or a continuing line of enquiry: Several authors have argued that…
• Present simple — the source’s enduring position, or a generally accepted fact: Bourdieu argues that taste is class-marked. Humanities conventions, including MLA, favour the present for texts and arguments.

Mixing these inside a paragraph without reason is one of the most visible surface faults in a draft. The full chapter-by-chapter breakdown is in our guide to which tense to use in each part of your thesis.

The that-clause

Most reporting verbs take that: argues that, found that, notes that. A few take a noun phrase instead: discusses the relationship, examines the effect, criticises the assumption. A handful take both. Writing discusses that… or examines that… is ungrammatical and is one of the tells that a sentence was translated rather than composed in English.

Human versus non-human subjects

Studies do not argue; authors do. The study argues that… is tolerated in some journals but reads loosely. Prefer The authors argue… or Smith and Ito (2024) argue…. Conversely, data and findings can suggest, indicate or point to — those verbs take non-human subjects comfortably.

Step 5: Vary the verb without drifting from your stance

A literature review that says states forty times reads mechanically; one that reaches for a thesaurus reads dishonestly, because each synonym changes the stance. The fix is to vary within a band, not across bands.

If you are relaying three neutral summaries in a row, rotate notes / reports / observes — all neutral. If you are building a critical paragraph, rotate claims / assumes / overlooks — all distancing. What you must not do is swap suggests for demonstrates because you used suggests in the previous sentence. That is not variation; it is an accidental promotion of the evidence.

Ten overused verbs and precise replacements

Discipline conventions you should check before you commit

There is no single network-wide rule, and departments notice when you use the wrong one.

• Psychology, education, health sciences (APA 7). Past or present perfect for prior findings. Heavy use of found, reported, examined. Authors as subjects.
• Literature, history, philosophy (MLA, Chicago). Present simple for what a text or thinker does: Arendt argues…, The novel dramatises… Past tense reserved for historical events.
• Engineering and computer science (IEEE, numbered styles). Frequently the citation number is the subject: In [7], the authors propose… Verb choice is sparse and factual — propose, present, evaluate, report.
• Medicine and nursing (Vancouver, AMA). Strongly evidence-graded. Demonstrated is reserved for trials; associated with is the standard hedge for observational work.

If you are unsure, take three recent articles from your target journal or three recent theses from your department, highlight every reporting verb in the literature review, and count the bands. That fifteen-minute exercise settles the question better than any general rule. It pairs well with the wider academic voice conventions covering hedging, signposting and paragraph structure, which govern the sentences around the citation.

Worked example: the same source, four stances

One study — a cross-sectional survey of 310 postgraduates reporting a correlation between supervisor contact and completion confidence. Four legitimate ways to introduce it, each committing you to something different:

• Neutral relay. Osei (2024) reports a positive association between supervisory contact and completion confidence.
• Tentative, building your case. Osei’s (2024) findings suggest that supervisory contact may support completion confidence.
• Attributing a position you will test. Osei (2024) contends that supervisory contact is the principal driver of completion confidence.
• Critical set-up. Osei (2024) claims a causal role for supervisory contact, but the cross-sectional design cannot support that inference.

Notice that only the last one requires you to write the criticism explicitly. The verb has already told the reader where you stand, which is exactly the analytical move examiners are looking for when they ask for more critical engagement rather than description.

A pre-submission check you can run in ten minutes

1. Search your literature review for says, talks about, tells us, looks at, proves and believes. Replace every hit.
2. Search for claims. For each one, ask whether you intended doubt. If not, change it.
3. Search for demonstrates, shows and establishes. For each one, check the design behind the source supports that strength.
4. Read any paragraph in which the same verb appears three times and rotate within the band.
5. Check that every verb is in the tense your style authority requires.

Getting the verb right is also the cheapest defence against unintentional misrepresentation, which sits close to the line that paraphrasing and attribution rules are designed to protect.

Write with the right verb from the first draft

Tesify drafts and edits directly in academic register, so citations arrive with a reporting verb that matches the source type and the stance you asked for — and flags the ones that overclaim before an examiner does.

Start writing your thesis with Tesify

Frequently Asked Questions