Disputed. * According to De Vaan, perhaps from Proto-Italic *worɣeō, from Proto-Indo-European *w(o)rǵʰ-éye-ti, from *werǵʰ- (“bind, squeeze”) (compare German würgen (“to strangle”), Lithuanian ver̃žti (“to string, tighten, constrict”), Russian отверга́ть (otvergátʹ, “to reject”), Polish otwierać (“to open”), English worry, wring. * The linguist Lucien van Beek argues that the proposed Germanic and Balto-Slavic cognates all can refer to the tying of ropes, a semantic sense that is absent from the Latin term. Thus, van Beek doubts a connection with *werǵʰ- (“bind, squeeze”). Instead, van Beek compares the term to Ancient Greek εἴργω (eírgō) and suggests that the term may derive from the zero-grade of the root *h₁wreǵ-. * According to Rix et al. (DIV), from a zero-grade present *uṛg-éye-ti, itself from the root Proto-Indo-European *wreg- (“track, hunt, follow”) and cognate with English wreck, wreak. However, De Vaan argues that such a development requires a "far-fetched" semantic shift. The linguist Nicholas Zair suggests that a causative formed the same root, with the meaning of "to cause to follow a trail," may have more reasonably evolved to mean "to drive, push." Regardless, Zair doubts this etymology, as causatives or iteratives in Proto-Indo-European typically required the o-grade, instead of the zero-grade predicted by Rix. Zair, however, concedes that an o-grade causative *wrog-éye-ti may have produced urgeō via metathesis, a development perhaps also seen in sorbeō, itself from Proto-Indo-European *srobʰéyeti.