Academic English revision

Reviewers Judge Your English by 8 Words in a Caption

Figure captions follow their own grammar—fragments, not sentences, and never a translation of 'as shown above.' If a reviewer reads only your caption and gets lost, your data doesn't matter. Write the caption as if the reader has nothing but the image in front of them.

Picture this: a reviewer at 11:47 p.m., forty submissions deep, thumb scrolling. She hits Figure 3. The caption reads 'As shown in the above figure, the data indicates a significant correlation.' She winces. Not because the data is wrong—because 'above figure' and 'indicates' and the whole 'as shown' scaffolding tell her three things in eight words: this person translated their caption from another language, they never re-read their figures, and they do not know how English captions work. She is not supposed to judge grammar. But she did. And now she is reading your methods section with a slightly worse mood than she would have otherwise. This is not fair. It is also true. Captions are the most-scanned, least-polished part of a paper, and they carry disproportionate weight for a non-native writer.

Here is the structural rule nobody teaches you clearly: figure captions and table titles are NOT sentences. A figure caption is a brief fragment following 'Figure X.' It uses imperative or descriptive phrasing, not 'The figure shows...' A table title is a noun-phrase header—no period, no 'the following table presents.' And critically, both must be self-contained. The journal format guide says 'stand alone.' That means if someone prints only your Figure 5 page, they must understand what was plotted, what the axes mean, and what the visual is telling them. No 'see text above.' No 'as discussed in Section 3.' The caption is the only text allowed in that reader's world.

The translation trap is the single biggest caption error among Chinese, Japanese, Korean, and Middle Eastern students. In Chinese papers you write '如上图所示' or '如下图所示'—a positional reference. In English, there is no 'above' or 'below' because the figure floats. You never reference position. You never write 'as shown in Figure 3' inside the Figure 3 caption. That is circular. You never write 'the following table shows.' The correct move: describe the content directly. ❌ 'As shown in the above figure, the concentration decreases over time.' → ✅ 'Plasma concentration (ng/mL) over 12 hours following a single 500 mg oral dose (n=42; mean ± SD; 95% CI shaded).' See the difference? The first is a pointer to nothing. The second is a description that replaces the figure in the reader's mind.

Verb choice in captions is where people get either lazy or overblown. ❌ 'Figure 4 shows the results of the experiment.' — 'shows' is fine grammatically but says nothing. What does it show? What results? ❌ 'Figure 4 illustrates the profound implications of our methodology.' — 'illustrates' and 'profound implications' are decorative. Nobody asked. In captions, the right verb is specific and flat: 'plots,' 'compares,' 'summarises,' 'maps,' 'reports.' You are describing an image, not selling a story. ✅ 'Figure 4. Comparison of predicted vs. observed binding affinity (kcal/mol) across 18 test compounds; dashed line = ±2 kcal/mol tolerance.' No verb needed at all. Just the noun phrase. That is the target.

Run the stand-alone test on every single figure before submission. Open a blank document. Type ONLY the caption. Below it, type nothing else. Hand that page to a friend who has never seen your paper. Can they tell you: (1) what data is in the image, (2) what the axes/units mean, (3) what the visual pattern tells them? If they say 'what does the red line mean?' or 'what is n equal to?'—your caption failed. Common fixes: define abbreviations in the caption itself even if you defined them in text. State the n, the test, the threshold. ❌ 'Mean expression levels for each treatment group.' → ✅ 'Mean protein expression (fold-change vs. vehicle control; n=6 biological replicates per group; one-way ANOVA, p<0.05 marked *).' That extra parenthetical costs you 12 words and saves your figure from being re-requested at revision.

One more rule people miss: table titles go ABOVE the table. Figure captions go BELOW the figure. Table titles have no terminal period. Figure captions can be one line or four—there is no universal cap, but every style guide wants them complete. Never end a caption with 'Adapted from.' Adapt what? Cite the source or delete it. Never write 'Source: author's own work.' You are the author. That sentence says nothing and signals you do not know how attribution works in English academic writing. ❌ 'Table 1: The following table shows the demographics of participants.' → ✅ 'Table 1. Participant demographics by treatment arm (n=240; mean ± SD or n (%)).' Period after the number. No 'following table.' No 'shows.' Just the noun phrase.

You will draft a caption in three seconds. That is fine. But three seconds of first-draft English, especially if you think in another language, produces the 'above figure / as shown / the following' reflex every time. Do not fix it in your head at 2 a.m. Paste the caption into easydue, flag it as a figure caption, and let it check for positional references, circular phrasing, and missing standalone elements. It will not rewrite your science. It will catch the four words that make a reviewer assume you proofread nothing. Then read it aloud once. If you hear yourself say 'as' or 'shown' or 'the following'—delete and start the noun phrase over. Captions are eight words of grammar that carry the weight of your whole paper's credibility. Treat them like the cover letter to your figure. They are read first. They are remembered longest.