Why do 'mama' and 'papa' sound similar in almost every language?
Because babies everywhere babble the same easiest sounds first, and adults everywhere hear those sounds as being about them and adopt them as words.
Simple intuition
The plain reason, in everyday words
Making a sound requires controlling your mouth, and some sounds are much easier than others. The easiest of all is to close your lips and let a voice out — that is 'm', 'p' and 'b'. Add the most open, effortless vowel, which is 'ah', and you have 'ma', 'pa', 'ba'. Babies produce these first, long before they mean anything by them, simply because those are the movements available to a mouth that is still learning. Adults hearing 'mamama' while holding the baby naturally take it as being addressed to them, and the sound gets adopted as a word. It happens in Mandarin, Swahili, Hungarian and Quechua alike, not because those languages are related but because babies everywhere have the same mouths and parents everywhere are listening hopefully.
It proves all languages descend from one ancestral tongue.
The opposite: these words are excluded from evidence for language relatedness precisely because they are re-created independently in every generation from infant babbling.
Babies are saying the word for their parent.
Early babbling has no referent. Adults interpret the sounds as being addressed to them, and that interpretation is what turns a sound into a word.
'Mama' always means mother.
In Georgian, 'mama' means father. The forms are stable across languages while the assignment is not, which is one of the strongest pieces of evidence for the babbling explanation.
The pattern is universal without exception.
It is a strong tendency, not a law. Exceptions occur, generally where borrowing or taboo avoidance has displaced the original form.
It is a compact lesson in how to read similarity. The same resemblance can come from common ancestry or from a shared constraint acting independently, and mistaking the second for the first is one of the standard ways to reach a confident wrong conclusion — in linguistics, in biology, and in any field that infers history from present-day patterns. The tell here is that the resemblance is generated fresh every generation rather than passed down.
Who worked it out
The resemblance was noticed early and often taken as evidence of a primordial common language, a reading that persisted through nineteenth-century speculation about the origins of speech.
What problem forced it
George Murdock's 1959 survey of more than a thousand languages established the scale of the pattern statistically rather than anecdotally.
How it changed since
Roman Jakobson's paper the same year supplied the explanation, arguing from the developmental order of infant sound production, and the account has been the standard one since. The forms are now routinely excluded from comparative reconstruction as a known source of spurious resemblance.
How linguists prove two languages are related
The comparative method, and why it must exclude exactly this kind of resemblance.
What babbling reveals about learning speech
The developmental sequence of sounds is remarkably consistent and predicts a great deal about early vocabulary.
Written for Curio rather than collected from a forum — it is part of the curated corpus that ships with the platform. The references it draws on are listed under Sources.