Two more common examples revolve around the triad five-finger-fist and the verb “to bore.” The Proto-Indo-European (PIE) contructions *penkwe (five – think “pente,” or “quinque” in Latin and Greek), *pnkwstis (fist), and *penkweros (finger) sure look related, and few would doubt that there is an earlier root which gave rise to all of them. In the equally ancient language families which bordered PIE (Uralic and Altaic, the ancestor languages to Finnish, Hungarian, Turkish), the roots for fist or palm of hand are suspiciously similar: *peyngo and *p’aynga. Penkwe-peyngo-paynga. Looks awfully related to me. Words for “bore” or “pierce” in the supposedly unrelated protolanguages of Afro-Asiatic, Altaic, Sumerian, Dravidian, and Uralic, are bar, bur, bur, pur, pura. Too much similarity for coincidence.
There are other proposed cognates – Greenberg lists around 600 – some more convincing, some less.
Okay, like you care. Just hold the key fact in mind. There are echoes of language relationships earlier than Indo-European, though mainstream linguists resist the concept. And all this is a lead-in to an upcoming discussion about how ideas change.