Our picture memory is superior to our verbal memory

In 1973, Standing conducted a range of experiments exploring human memory. The participants were shown pictures or words and instructed to pay attention to them and try to memorize them for a test on memory. Each picture or word was shown once, for five seconds.

The words had been randomly selected from the Merriam-Webster dictionary and were printed on 35mm slides - words like ‘salad’, ‘cotton’, ‘reduce’, ‘camouflage’, ‘ton’.

The pictures were taken from 1,000 snapshots - most of them from holidays - beaches, palm trees, sunsets - volunteered by the students and faculty at McMaster University in Ontario, Canada, where Standing taught at the time. But some of the pictures were more vivid - a crashed plane, for instance, or a dog holding a pipe. But remember this was the seventies - all dogs smoked pipes back then.

Two days later the participants were shown a series of two snapshots or two words at a time, one from the stack of snapshots they had seen before and one new, and were asked which one looked more familiar.

The experiment showed that our picture memory is superior to our verbal memory. When the learning set is 1,000 words selected from the dictionary above, we remember 62 per cent of them, while 77 per cent of the 1,000 selected snapshots were remembered. The bigger the learning set, the smaller the recognition rate. So, for instance, if the learning set for pictures were increased to 10,000, the recognition rate dropped to 66 per cent. However, we remember snapshots better than we do words. That may be why you might be better at remembering faces than names. So, if you are introduced to Penelope, it might help you remember her name if you picture Penelope Cruz standing next to her.

Standing showed participants 1,000 words or 1,000 photographs. Two days later, recognition was higher for the photographs, especially when the images were vivid. A picture is memorable when it gives the idea a body. The dog with a pipe survives. An interchangeable photograph of “collaboration” has vanished before the page loads.

The distinction matters. A diagram locating a product inside a customer’s workflow can outlast three paragraphs about integration. A safety instruction showing the exact control and hand movement can be reconstructed under pressure. One recognisable customer situation may carry an entire proposition after the feature list has dissolved into soup. Pictures can also make nonsense durable. A dramatic chart preserves a misleading comparison more effectively than forgettable prose. Vividness supplies recall, not truth. That makes accuracy more important, not less.

A vivid lie travels beautifully. That should make the person choosing the image slightly nervous. Accuracy matters more when an error will be easy to carry. Decide what the audience should still be able to picture tomorrow. An object? A sequence? A contrast? A consequence? Give the visual that single job and let words handle whatever the eye cannot. Adding decorative imagery because the page “needs something visual” gives the audience another thing to remember. Two days later, ask what people recognised and what idea they connected to it. If they recall the stock handshake but cannot describe the product action, the next version needs a different image.