Say the word out loud — even if you only whisper it

Picture the ordinary way a word gets studied. You meet it on a screen or a page, your eyes rest on it for a second or two, perhaps you nod — right, got it — and you move on. It felt like learning. And yet by the evening the word has quietly slipped its mooring, and the next time you meet it you feel only a faint, useless flicker of recognition. Nothing to hold.

There is a change you can make that costs almost nothing and is strangely easy to skip: move your lips. Say the word. Not necessarily loudly — a mutter under your breath will do — but out, into the air, in your own voice. Do that, and the same word is measurably more likely to still be there tomorrow. Psychologists have a plain name for this small magic. They call it the production effect, and they have been mapping it carefully for well over a decade.

A phenomenon with a deliberate name

The term was coined by Colin MacLeod and four colleagues — Nigel Gopie, Kathleen Hourihan, Karen Neary and Jason Ozubko — in a 2010 paper in the Journal of Experimental Psychology: Learning, Memory, and Cognition. Its title, "The production effect: Delineation of a phenomenon," is a small in-joke: it deliberately echoes a famous 1978 paper by Slamecka and Graf on the generation effect, planting the new finding in the same soil as a classic one.

The experiments themselves were almost aggressively simple. Participants studied a list of ordinary words. Some of the words they were told to read silently; others, scattered through the same list, they were told to read aloud. Then their memory was tested. The words that had been spoken aloud were remembered reliably better than the words that had merely been looked at — across eight experiments, the same modest, dependable gap kept opening up. The only difference between the two kinds of word was whether the mouth had been involved.

Why the mouth helps the mind

The tempting explanation is effort: saying a word takes a little more work than reading it, and effort is good for memory. But MacLeod's group ran a cleaner account to ground. In one experiment they had people make the same overt response to a subset of words — the identical action, every time — and the advantage evaporated. Merely doing something extra was not enough. What mattered was that the spoken words became distinctive.

When you say a word aloud, you lay down more than the visual trace of its letters. You add a motor record of your mouth forming it, and an auditory record of hearing it in your own voice. Later, when you go looking for that word, there are extra, unusual handles to grab — and in a list where most words were only ever seen, the few that were also spoken stand out like the one item on a page that someone underlined. Memory, as ever, rewards the thing that is easy to tell apart. This is the same currency of distinctiveness that makes the keyword method work, and that explains why crowding similar words together quietly backfires.

The spoken word is not remembered better because it was harder. It is remembered better because it left a fuller, stranger fingerprint.

It is your own voice that does it

If distinctiveness is the mechanism, then whose voice you hear ought to matter — and it does. In a neat 2018 study in the journal Memory, Noah Forrin and Colin MacLeod added a condition no one had tried before: alongside reading aloud, reading silently, and hearing someone else speak the word, they had people hear a recording of themselves saying it. Memory fell into a tidy gradient. Speaking aloud in the moment was best. Hearing your own recorded voice came next. Hearing a stranger's voice was weaker still, and silent reading brought up the rear.

The reading Forrin and MacLeod drew from it is that oral production is doubly useful: it is a motor act and a self-referential one. You do the speaking, and you hear yourself doing it — two distinctive traces, both tagged as unmistakably yours. It is a small argument for a slightly self-conscious habit. When you meet a new word, do not just sound it out in your head. Let your own voice say it, and let your own ears catch it.

Especially in a language that isn't yours

Here is the part that should interest anyone learning a second language most of all. You might reasonably expect the production effect to be a first-language luxury — that stumbling over an unfamiliar foreign word aloud would be too clumsy to help. The opposite appears to be true.

In a 2024 study in Memory & Cognition, Rachel Brown and Tanja Roembke tested two groups of German–English bilinguals, having them read words silently or aloud in both their first and their second languages. Both groups showed a pronounced production effect in both languages — spoken words recognised more accurately than silent ones. But the effect was larger in the second language than in the first. Their paper carries the faintly encouraging subtitle "Less experience may help." The likely reason is that producing a word you don't yet own demands more careful phonological work and more attention — the very effort a fluent speaker no longer needs to spend, and the very effort that etches the word deeper.

For a learner, that is an unusually cheerful finding. The awkwardness of saying a new foreign word out loud — the hesitation, the slight overarticulation, the feeling of your mouth learning an unfamiliar shape — is not a tax on the learning. It is part of what makes it stick.

Where the effect stops

It would be in poor keeping with the research to oversell it, so two honest caveats. First, most of the classic demonstrations use mixed lists, where some words are spoken and some are not, and the advantage is partly a matter of the spoken words standing out against silent neighbours. Say every single word aloud, in every session, and each one has less to stand out from; the within-list contrast softens. This is not a reason to read silently — it is a reason not to expect saying-aloud to be a magic switch that doubles everything. It is a genuine, moderate edge, not a miracle.

Second, the production effect is an encoding trick — it improves the trace you lay down in the moment. It does not repeal the way memories fade, and it is no substitute for meeting a word again later. Producing a word aloud and then never revisiting it still leaves it exposed to the ordinary forgetting curve. Saying it out loud makes the encounter count for more; it does not make the encounter your last. The real power comes from stacking it on top of the things that already work.

What to do with this

None of this asks for extra study time. It asks you to involve your mouth in study you were doing anyway.

• Say new words aloud on first contact. A whisper is plenty. The point is to move from the silent, eyes-only reading that leaves the thinnest trace to something your mouth and ears also witness.
• Speak the word when you recall it, not just when you read it. Producing a word aloud at the moment you reach for it from memory folds the production effect into retrieval practice — two of the most reliable findings in the field, working the same beat.
• Say the whole sentence, not the bare word. A word spoken inside a real sentence arrives with its grammar and its neighbours attached — more to say, and more to remember it by.
• Don't drone. Reading an entire session aloud in a flat monotone gives back some of the distinctiveness you were buying. Vary it — speak the hard ones, the new ones, the ones that keep slipping — and let the effort fall where the words are actually difficult.

Where vocabla fits

A lot of what vocabla does already leans this way. You don't sit and re-read lists silently; you type the missing word into a real sentence, which is itself a small act of production — a deliberate, effortful move rather than a passive glance. The dialogues you unlock are spoken aloud, so the words you've been learning come back to you as sound as well as text. And reviews are spaced so you have to reach for each word from memory at the moment it's starting to fade, rather than recognise it on a page.

The one thing the research adds is almost too simple to build into an app: when the word comes up, say it. Out loud, under your breath, on the train with your hand half over your mouth — however you can manage it without alarming the person next to you. Let your own voice have a turn. It is the cheapest upgrade in language learning, and one of the few your mouth can install by itself.