Teachers often bold every example of a target form in a reading and expect the typography to teach. The learner may see twelve instances of the past perfect and remember the story while learning little about why one event takes had left and another takes left. Boldface can direct the eyes. The lesson still needs an operation that connects form with meaning. Picture an A2 class reading a 286-word account of a missed train. Every past-perfect verb appears in bold. This visual input enhancement makes the form physically salient through weight, colour, underlining or another typographic cue. It is cheap, unobtrusive and easy to add. Its modest average effect should keep it in the role of a cue. A teacher can use the cue as an invitation to inspect one contrast, then require a decision whose answer depends on that contrast and check whether the learner can repeat the decision in plain text two days later.
That is a narrow job.
What the visual cue can do
Sang-Ki Lee and Hung-Tzu Huang synthesised 16 primary studies with 20 distinct samples. Learners who read enhanced texts barely outperformed learners who saw the same target forms without enhancement: the aggregated effect was d = 0.22. The analysis also found a small negative effect on meaning processing, d = -0.26. Attention drawn toward visible grammar may compete with attention to the message. Those averages cover different languages, forms, tasks and research designs. They do not say that bold print always harms comprehension or never helps grammar. They set a sensible expectation. Typography changes salience. It supplies no guarantee that a learner will infer the relevant contrast, retain it or use it later.
Visibility is only the first step.
Ronald Leow's 2001 study gives the point a smaller scale. Thirty-eight first-year learners of Spanish read a 242-word text containing formal imperatives. In the enhanced version, target verb forms were underlined and the imperative morphemes were bold. Think-aloud data and later tasks found no measured advantage for enhancement in reported noticing, comprehension or intake. The exact result belongs to this population and form, though it blocks an easy assumption that visible automatically means learned.
Give attention an operation
A post-reading task should make the learner use the highlighted form to recover a relationship in the message. With the missed-train story, write two event cards:
The timeline decides.
- Mina arrived at the platform at 08:41.
- The train left at 08:38.
Ask the learner to combine them, then explain which event came earlier. The train had left when Mina arrived earns its form from the timeline. Next, reverse the order or replace one time. The learner must inspect meaning again, rather than copy the bold verb string. A second option is reconstruction. Remove the highlighted verbs after the first reading and ask pairs to rebuild four sentences from content words. A third option asks learners to choose between left and had left in a new paragraph and underline the time clue that supports each choice. One task is enough. Three back-to-back worksheets would turn a light attentional cue into a grammar marathon. The choice should remain genuinely diagnostic. If every sentence containing before takes the past perfect and every other sentence takes the simple past, learners can solve the typography puzzle without reading the event sequence. Include one sentence where before appears with two simple-past verbs, and one where the earlier event is clear from context rather than a connector. The answer then depends on time reference. After each choice, the learner draws two dots on a line, writes 08:38 and 08:41 beneath them, and labels the verb attached to each event. The drawing makes a lucky answer visible.
Output can expose the gap
Shinichi Izumi compared visual enhancement with output in a controlled study of adult ESL learners acquiring English relative clauses. In the 2002 experiment, participants worked with computer-based reading and text reconstruction under conditions that varied output requirements and enhanced input. Output-plus-input groups made greater learning gains than comprehension-only groups. Visual enhancement increased attention to target items yet produced no measurable learning gain on its own. That result supports a narrow lesson design, rather than a broad promise about speaking or writing. A learner who must reconstruct the report that the manager requested may discover that their current grammar cannot connect the noun and clause. The following input then answers a problem the learner has felt. This resembles the logic behind starting grammar practice with interpretation: grammatical form earns attention through the meaning it changes.
Forms differ in perceptual salience, complexity and readiness for learning.
Izumi studied relativisation, while the train example uses tense and aspect. Relative clauses also allow several structural types. A classroom application should keep the target small and watch the learner's responses before expanding the claim.
Protect the reading
Enhancement becomes noisy when too much text shouts. If articles, auxiliary verbs, endings and time expressions all carry different colours, the reader receives a formatting code before receiving a story. Mark one form family. Keep the original type legible. Never require a learner to decode five colours whose meanings appear only in a distant legend. The timing can protect comprehension too. Let learners read once for the missed-train story with minimal interruption. On the second pass, direct them to the bold forms and ask what each one contributes to the event sequence. This order is a teaching choice, supported indirectly by the negative meaning-processing effect in Lee and Huang's synthesis. It has not been established as the best sequence for every form or proficiency level.
The cue should eventually disappear.
My preference is to remove the enhancement on the final example. After six marked instances and two guided comparisons, show a plain paragraph and ask the learner to locate the form that carries the same relationship. A cue should prepare attention to work without the cue. Otherwise the worksheet has learned the highlighting system more faithfully than the student has learned the grammar.
Audit the marked text
Before class, count the highlighted items. Nine clean examples may be enough. Forty-seven marks across one page probably change reading into symbol hunting. Then write down the exact decision learners will make after reading. "Notice the verbs" names no observable action. "Put four events in order and justify two verb choices" produces evidence a teacher can inspect. Check comprehension separately. Ask for the missed connection, the station and the revised arrival time before discussing verb forms. If those answers collapse after enhancement, the formatting has taken too much attention from the message. Reduce the marks or postpone the form pass. Tomorrow, bold six examples in one short text. Read for meaning, order four events, rebuild two sentences, then present one unmarked example. When the learner circles had closed and points to the earlier event at 07:53, the visual cue has finished its job.