The derivation of the Latin word lupus from *u̯l̥kʷos, the most common Indo-European name of the wolf (preserved, for example, in Sanskrit vṛkas, Old Church Slavonic vlĭkŭ, Albanian ulk and Gothic wulfs), currently relies on two assumptions: that Italic uniquely shares with Greek the metathesis to *lukʷos assumed for λύκος; and that Latin borrowed its form of the word—with p for expected qu as the reflex of the labiovelar kʷ—from Sabellic, the dominant language of Italy before the rise of Rome.Footnote ¹

The transfer of this particular animal name can be explained by its cultic significance, and there is a parallel borrowing from Sabellic in the Latin word bōs ‘cow’. A different kind of root deformation is also proposed for Proto-Celtic *ulkʷos,Footnote ² adding to the number of anomalous variants, and deepening the association with processes of taboo deformation. But since there are no early examples of the word in the Italic languages to support either of the primary assumptions, this leads to the problem of further having to assume that exactly the same innovation occurred independently in two of the three neighbouring language groups. It would be more plausible if metathesis happened only once, in Greek, and the resulting form were borrowed into Sabellic.

A borrowing from Greek requires it to have been earlier than the first millennium b.c.e. The labiovelar consonants attested in Mycenaean Greek had disappeared by the eighth century—their reflexes appear in stable spellings in the earliest alphabetic inscriptions—and the preceding change, by which labiovelars lost their labial element when next to u (the ‘boukolos rule’, which accounts for κ instead of π in λύκος), is already evident in Mycenaean. But a significant Greek cultural presence in Italy predates the period of Mycenaean textual records (beginning in the fourteenth century) by at least two centuries. The influence is widespread, represented in the burial objects of indigenous cemeteries, and reflects contact with the Greek-speaking peoples of the mainland rather than with Minoan Crete or Pre-Greek Aegean.Footnote ³

A feature of Sabellic that makes it a good vector for naturalizing the Greek word is that unmetathesized *u̯l̥kʷos would have developed into a form so close to the name of the fox, *wulp-o-s alongside *wolp-i-s, that the two could have merged, leaving a space for a loanword that restored a specific meaning. The semantic range of inherited *(H)ulp-i- (antecedent of Sabellic *wolpis and Latin uolpēs)—fox, jackal, wildcat, marten—suggests that it was originally a word for any wild canid or similar ground predator, thus preconditioning a merger. Latin, in preserving the labiovelar, would have maintained a contrasting word-pair; and the ousting of its inherited word for ‘wolf’ must be the result of compelling social factors. Beginning as a singular development of early Greek, this line of development through Sabellic is a better fit with the profile of taboo deformation.

This explanation of the word's origin implies that metathesis took place before the middle of the second millennium b.c.e., allowing an extended period during which speakers of Proto-Sabellic could be exposed to Greek *lukʷos. And as a matter of relative chronology it implies that the ‘boukolos rule’, being a common feature across the East–West dialect divide, belongs to a period before Greek split into dialects, but after it separated from Indo-European. It also implies that there were other early Greek borrowings in Italic, especially of words related to cult and to Greek-mediated Aegean civilization.