Pick any language at random. Ask how it says "mother". The odds are uncomfortably high that the word begins with m and contains an a: English mama, Mandarin māma, Swahili mama, Quechua mama, Navajo amá, Welsh mam, Hebrew imá, Tagalog nanay / mama, Hindi mā, Malayalam amma, Hawaiian makuahine / māmā. Now ask how it says "father". Across an equally vast sample you get papa, baba, tata, dada, abba, atta — a labial or coronal stop plus a.
The pattern is too widespread to be explained by a single shared ancestor. The Yale anthropologist George P. Murdock assembled a 1959 cross-cultural dataset of 1,072 kin terms from 470 languages and found that words of the mama type were used for "mother" in 52% of the sample, and papa/tata/baba-type words for "father" in 55%. His sample included languages from unrelated families such as Indo-European, Sino-Tibetan, Niger-Congo, Austronesian, Uto-Aztecan and Pama-Nyungan.
The classical explanation was given a year later by the Russian-American linguist Roman Jakobson, in a slim, beautiful 1960 paper titled "Why 'Mama' and 'Papa'?". His argument runs as follows:
- The earliest sounds a baby produces — once the random crying phase ends — are bilabial nasals [m] and bilabial stops [p, b]. These require only that the lips close and open; no tongue control needed. The default vowel that emerges from a slightly open mouth is [a].
- The very first proto-syllables a baby makes are therefore ma, pa, ba, repeated: mama, papa, baba.
- Adults overhearing this babble assume the baby is addressing them. Since the person almost always present when a baby vocalises is its mother, mama gets interpreted as "mother". The father, slightly more distant in traditional caregiving, gets papa.
- The adults reinforce the babble by responding to it, so the form sticks as the conventional word in that language — even though nothing in the structure of the language required it.
Jakobson's account elegantly explains the exceptions, too. Georgian famously swaps the pair: mama means "father" and deda means "mother". This is no embarrassment to the theory — Jakobson's point was that mama is the easiest first word, not that it must mean "mother". Georgian just attached the easiest form to the slightly less expected referent. Likewise, in Old Japanese, "mother" was reconstructed as papa (modern haha, with /p/ → /h/ regular sound change), and "father" was titi.
So when an English baby says mama, a Mandarin baby says māma, and a Swahili baby says mama, no shared ancestor is needed. The same articulatory ease meets the same caregiver, and the same word is born — independently, again and again, across every continent of human history.