Is It Too Late to Learn English as an Adult?

Not too late, but the honest answer is more specific than that. Adults out-learn young children in the early months of a new language, the age effect on grammar is real yet narrower and more contested than folklore suggests, and the ability that genuinely erodes with age is pronunciation rather than vocabulary.

The question people are actually asking

“Is it too late” usually stands in for something more concrete. Will I ever sound natural. Will I be able to work in this language. Is the effort I am about to spend going to be wasted because I did not start at six.

Those are three different questions and the research answers them differently, which is why a flat yes or no is useless. It helps to separate what ages badly from what does not, because the popular version of this topic collapses everything into one verdict.

Adults start faster, not slower

The finding that surprises people most is one of the oldest. Snow and Hoefnagel-Höhle published a study in Child Development in 1978 tracking English speakers acquiring Dutch naturalistically in the Netherlands, tested three times across their first year, grouped by age.

The adults and the 12 to 15 year olds made the fastest progress in the first few months. The 3 to 5 year olds scored lowest on every measure used. By the end of the first year the 8 to 10 and 12 to 15 year olds had the best control of Dutch, so children do close the gap and pass adults given enough immersion. But the early advantage runs the opposite way to the folk belief.

The likely reason is unglamorous. Adults bring conscious strategy, literacy, and a first language to transfer from. A five year old is working it out from scratch. Whatever children have that adults lack, it does not show up as raw speed in the first year.

Where the age effect is real, and where it is argued about

The strongest modern evidence for an age effect comes from Hartshorne, Tenenbaum and Pinker, published in Cognition in 2018. They analysed a dataset of roughly 669,000 native and non-native English speakers and modelled how grammar-learning ability changes with age, separating current age, age of first exposure, and years of experience. Their estimate was that grammar-learning ability is largely preserved until about 17.4 and then declines steadily.

That number travelled quickly, and it is worth knowing that it has been challenged. A 2022 reanalysis by Slik and colleagues in Language Learning argued that one sharply defined critical age for all learners may be an artifact of how the models were fitted, that continuous decline fits some learner groups better than a sudden break, and that where a break does fit the estimate lands nearer 18.6 to 19.0.

So the responsible summary is that grammar-learning ability declines with age, the decline is measurable, and the precise shape of the curve is unsettled. What nobody in this literature claims is that the ability disappears.

The part that genuinely gets harder is sound

Across the age-effects literature, pronunciation is where the constraint shows up most consistently. Grammar is still being argued over and vocabulary barely registers as a problem, but accent resists in study after study.

This matters more than it sounds, because it inverts the usual anxiety. Most adult learners worry that they will not be able to memorise enough words, and pour their effort into exactly the area where being an adult is an advantage. Meanwhile the part that actually carries an age penalty is the part most self-study never touches, because reading a word list does nothing for your ear.

There is a fairly well understood reason accent resists. Newborns can distinguish essentially any phoneme contrast used in any human language, including contrasts from languages they have never heard. That general ability starts narrowing at around six months: perception of the sounds that matter in the surrounding language sharpens, and sensitivity to the ones that do not matter falls away. By roughly twelve months the map is largely language-specific. The technical name is perceptual narrowing, and it is a feature rather than a bug, since a listener who treats every acoustic difference as meaningful cannot process speech efficiently.

The cost arrives decades later. A contrast your first language never used is one your ear stopped tracking before you could walk. The standard illustration is English “rocket” and “locket” for Japanese speakers, where the difficulty is not carelessness or effort but the absence of that contrast in the phoneme inventory the listener grew up with. Iverson and Kuhl’s work on this is the usual reference point.

Two things follow, and they are more encouraging than the mechanism sounds. First, this is a perception problem before it is a production problem. If you cannot reliably hear the difference, you have no target to aim your mouth at, which is why pronunciation practice that is all speaking and no listening tends to stall. Second, adult perceptual categories are trainable. They are stubborn, not frozen, and the training that works is unglamorous: hearing the contrast repeatedly, in real words, with feedback about which one you actually heard. That is a very different activity from reading a word list, and almost nobody does it by accident.

What ages, and what does not

Ability	Does age hurt it?	What that means for you
Vocabulary size	Barely, and adults have real advantages	Your strongest lever. Spend here without guilt
Reading comprehension	No, literacy transfers	Often the fastest visible progress
Grammar	Yes, measurably, though the curve is disputed	Slower, not blocked. Explicit study compensates
Listening	Partly, tied to sound perception	Improves with deliberate audio exposure
Accent	Yes, most reliably of all	Expect an accent. Aim for clarity, not erasure

The row people find liberating is the last one. An accent is not a failure state. Being understood is the goal, and intelligibility and nativeness are different targets that get conflated constantly.

Why vocabulary is the adult’s strong suit

There is a mechanical reason adults do well here. A new word has to attach to something, and adults have far more to attach it to.

You already know what a “mortgage” is, so learning the English word is a labelling problem rather than a concept problem. You can read, so you meet words in writing as well as speech. You know your first language well enough to notice that a new word is close to something you already have, and where it differs. A child learning the same word is often building the concept and the label at once.

Adults also have a structural advantage children take years to build: they can see the machinery inside a word. If you know that “-tion” makes a noun and “-ify” makes a verb, then “clarify” and “clarification” are one item to learn rather than two, and a word you have never seen still tells you its part of speech before you look it up. Anyone who sat through Latin or Greek roots at school is carrying more of this than they realise. A young child has none of it and has to meet each form separately.

None of this makes adults better learners overall. It makes them better at this specific part, and it is a good argument for putting vocabulary at the centre of adult study rather than treating it as the boring prerequisite before the real work starts. If you want the numbers on how much vocabulary is actually enough, how many words you need to be fluent covers the coverage thresholds.

What this changes about how you study

• Put your effort where the payoff is largest. Vocabulary and reading return faster than grammar drilling for most adult learners, and they are also where age is least against you.
• Treat pronunciation as a separate skill with its own practice, not something that improves by accident. It is the one area with a real age penalty, so it needs deliberate attention rather than hope.
• Learn words as sounds, not only as spellings. A word stored purely as a shape on the page is a word you will not catch in speech.
• Aim for intelligibility rather than a native accent. The second is a lottery. The first is achievable and is what actually matters in a conversation.
• Do not use age as an explanation for a plateau. Plateaus have ordinary causes, usually method and volume, and those are fixable at any age.

The honest limits

A few things this research does not say, since the topic attracts overclaiming in both directions.

Age is not irrelevant. The decline in grammar-learning ability is measurable and the one in accent is more reliable still, and pretending otherwise only sets people up to feel cheated later. Nor will adults end up ahead in the long run: Snow’s children caught up and passed the adults inside a year, and immersion of that kind is precisely what most adult learners do not have.

The useful part is that the ceiling sits far higher than the folklore implies, and whatever ceiling does exist is mostly about sound. For someone deciding whether to start at 35, that distinction is the whole ballgame. A realistic timeline helps too, and how long it takes to learn English puts numbers on it.

How Vocaby fits

The shape of the research suggests a fairly clear division of effort, and the vocabulary half is what an app can genuinely help with.

Vocaby holds more than 29,000 words, which reaches well past the frequency band that ordinary exposure hands you, and it schedules reviews with FSRS spaced repetition so words come back near the point you would otherwise lose them. You can browse the range in the word library.

The part that matters for an adult learner specifically is that every entry carries IPA and real audio. That addresses the exact weak point the age-effects literature keeps pointing at. A word learned only as spelling is a word your ear has never met, and for adult learners that gap does not close by itself. Storing the sound alongside the meaning from the first encounter is cheap insurance against the one disadvantage age actually gives you.

Starting at 35 is not starting late. It is starting with a different set of strengths than a six year old, and most of them are the ones vocabulary rewards.

Download Vocaby on the App Store to learn words with audio, IPA, examples, and spaced repetition on every entry.