Cut redundant words when the visual already says it

Do not narrate a self-explanatory diagram word for word — it overloads, not reinforces.

Why it works

Dual coding helps when the two channels carry complementary information. But making the verbal channel repeat exactly what the visual already shows forces the mind to process the same content twice and reconcile them, consuming working memory for no gain — the redundancy effect. More words are not automatically more learning.

How to do it

1. If a visual is self-explanatory, trim the text to a brief pointer rather than a full narration.
2. Let words and images divide the labor: each should add what the other cannot show.
3. Watch for places where you are reading aloud what the picture already makes obvious.

Evidence

The redundancy effect is supported in cognitive-load research: presenting identical information simultaneously in two forms can impair learning relative to a single well-chosen form, because learners expend effort cross-checking the duplicates. (rct)

Redundancy hurts mainly when each source is independently complete; brief reinforcing labels on a complex visual are usually still helpful.

Sources

• Sweller, cognitive load theory (redundancy effect)

Common mistake

Assuming that because words-plus-pictures beats words alone, piling on more text and more images must be even better, when redundancy quietly overloads working memory.