Someone hands you your coffee and “thanks” is out of your mouth before your brain has fully registered the exchange. A stranger holds a door and “thank you” arrives on its own, the same way you’d flinch if something flew at your face. It reads as instinct, the kind of behavior that feels wired in rather than learned, because by adulthood there’s no memory of ever having to work at it.
Watch a toddler in the same situation and the instinct story falls apart fast. A three-year-old hands back an empty plate and says nothing. A four-year-old tears into a birthday gift and starts playing before a word of thanks makes it out. What looks automatic in a thirty-five-year-old looks, in a three-year-old, like a habit that simply hasn’t shown up yet, and that gap between the two pictures is what a small, unglamorous study from 1980 went looking for.
The habit almost never shows up on its own
Researchers Esther Blank Greif and Jean Berko Gleason wanted to watch politeness happen rather than assume it. They brought 22 families with children between two and five years old into a lab, had a researcher hand each child a small gift in front of a parent, and recorded what came next. Their question was direct: how often does a child produce one of these small social scripts, what linguists came to call politeness routines, without being told to?
The results ran in Language in Society under the title “Hi, thanks, and goodbye: More routine information,” and the answer was not very often. Across 44 recorded episodes, children said hi or goodbye unprompted about 25 percent of the time, and they produced an unprompted thank you only 7 percent of the time. When the words didn’t come on their own, a parent almost always supplied them: 21 of the 22 families prompted their child at least once, and the prompts followed a narrow script, as the pair documented it: “Say X, or Can you say X, or What do you say?” The word say turned up in 95 percent of the 59 prompts researchers counted.
The researchers also tracked who was doing the prompting and modeling inside each household, and that split wasn’t even either. Mothers said thank you themselves about half the time they had the chance to, compared with less than a fifth of the time for fathers, a gap the study found statistically significant. Mothers said goodbye at a similarly higher rate than fathers. Whatever pattern a child picked up, it was disproportionately one parent’s repeated example doing the teaching.
The words came before they meant anything
Compliance was high once a prompt actually landed. Children said thank you 86 percent of the time once a parent told them to, which on its face sounds like the manners were basically there, just one nudge away from arriving unassisted. What the transcripts don’t show is a child improvising. Greif and Gleason wrote that “none of the children varied the adults’ form or attempted to express their appreciation in their own words.” A child handed a toy and told to say thank you didn’t say I like this or that’s nice, the way a child might if the phrase were already tied to the feeling behind it. The two words came out exactly as given, the way you’d repeat a password you’d been told worked without knowing what it opened.
That’s an unusual way for a young child to handle language. A toddler who hasn’t learned the past tense of go will often say goed, applying a rule on their own initiative, which is a sign of a mind actively building its own grammar rather than copying one. Please and thank you don’t seem to work that way, at least in this study. They arrive as fixed units, learned through imitation and complied with because compliance is what was asked for. The pattern worth noticing here is that a polite phrase can be produced correctly and consistently well before whatever it’s meant to express is actually understood. Learning the script and learning the sentiment behind it seem to run on two different timelines, at least in this study.
What the reflex is actually made of
The 1980 study only needed a handful of taped visits per family to catch this pattern, and it didn’t follow any of these children into adulthood, so it can’t say exactly when a prompted script turns into an unprompted reflex. But the shape of the finding points somewhere fairly obvious. Boston University psycholinguist Jean Berko Gleason, who continued studying how children pick up these routines through the 1980s, was documenting a script that gets rehearsed dozens of times a week for years: a parent prompts, the child complies, and neither one remembers the specific exchange a week later. Repeated often enough and for long enough, the prompt itself becomes unnecessary, and what’s left behind is the part that looks automatic.
It also means the labor behind the reflex was real, even if nobody was keeping score at the time. Every “what do you say?” that felt repetitive or pointless in the moment was doing the actual work of building the habit, one mildly annoying reminder at a time. Multiply one prompt by dozens a week and by years of childhood, and an adult’s unthinking “thanks” stops looking like something a person was simply born with and starts looking like the end product of somebody else’s patience.
That’s a reasonable trade if you’re the parent asking “what do you say?” for what feels like the fiftieth time this month and wondering whether any of it is landing. It’s also a reason to go easy on a child who forgets again tomorrow.
Nothing has gone wrong when a four-year-old skips the thank you after a gift. The unconvincing, prompted version coming out of a small child’s mouth today is doing precisely the job a decade of repetition needs it to do, long before it turns into the kind of reflex an adult produces without remembering that anyone ever had to teach it to them.
