Epic Knees

Steve T. Reece (St. Olaf College)

The common Greek word for "knee," gÒnu, gÒnatow, tÒ, has sterling Indo-European credentials: cognates in other Indo-European languages abound -- Latin genu; Gothic kniu; Hittite genu; etc. -- all pointing to I.E. *gonu-.  The word appears thousands of times in extant Greek texts, in many different forms and in various dialects, and it is ubiquitous in Homer (122 times), which has forms in both gon- and goun-.  In addition to these forms, a zero-grade form appears three times in Homer in the adverb prÒxnu "with the knee(s) forward" (literally) and six times in the adverb gnÊj "on the knee(s)," both of which are probably fossilized forms of what was in origin a nominal form gnu-.

This tidy picture is blurred by a single appearance in Homer of another, rather strange-looking, reflex of the word: fignÊh, -hw, ≤.  An anonymous comrade of Idomeneus is described as having just returned from battle, carried by his comrades because he has been "struck on the fignÊh by the sharp bronze" (Il. 13.212): ∑lye kat' fignÊhn beblhm°now Ùj°Û xalk“.  A similar circumstance is described later in the Iliad when Achilles stops the charging Trojan Demouchos in his tracks by "striking him on the knee with a spear," a wound that is a prelude to his death (Il. 20.458): kåg gÒnu dour‹ bal∆n ±rÊkake.  The similar phrases kat' fignÊhn beblhm°now Ùj°Û xalk“ and kåg gÒnu dour‹ bal≈n seem to describe the same type of wound, and had the form fignÊh occurred only here in all of Greek we could still have been confident on morphological and semantic grounds that fignÊh and gÒnu were cognates.

In reality various forms of fignÊh appear some four-hundred times in Greek texts from the Archaic period to the Byzantine, a handful or two of times in poetry but mostly in prose, where it is most often to be found in medical treatises (Hippocrates, Galen, Oribasius, etc.) in describing the area behind the knee, i.e., the hollow of the knee, or, in medical jargon, the popliteal space.  In short, in addition to the morphological and semantic evidence in Homer for regarding fignÊh and gÒnu as cognates, a survey of the lexical history of these terms within their contexts in all of Greek literature serves to remove all doubt.

This brings us to the critical question to be addressed in this presentation.  It appears that zero-grade nominal forms gnÊw and gnÊa(h), examples of which we see in the Hymn to Hermes "Theocritus'" Idylls, and perhaps in the Greek Anthology, survived side by side in the poetic tradition with fignÊw and fignÊa(h), examples of which we see in Homer, Theocritus, Herodas, Nicander, Manetho, and the Greek Anthology.  Since gnu- is clearly the inherited I.E. form, how can we explain the prefixed iota of fignu-?  All previous proposals, which have been based largely on semantics (prepositional prefix, etc.), have serious deficiencies.  I propose that an answer is to be found instead in the realm of phonetics: i.e., that the secondary form fignu- was a result of resegmentation in some of the common phrases that include the word for “knee.”  The resegmentation occurred both in the Greek vernacular and in the poetic Kunstsprache, probably independently of each other, in a phonetic environment where there was a preceding preposition that was mistakenly construed to have suffered elision: e.g., émf‹ gnu- > émf' fignu-, ént‹ gnu- > ént' fignu-; §p‹ gnu- > §p' fignu-; cf. §n‹ gnu- > §n fignu-.  Homeric §p‹ goÊnasi(n) (6x) points to the existence of zero-grade forms §p‹ gnÊ˙si(n) and §p‹ gnusfi, which were obviously ripe for resegmentation as §p' fignÊ˙si(n), §p' fignÊsi, and which resulted eventually in the earliest surviving instance of fignÊh, in the passage under consideration: ∑lye kat' fignÊhn beblhm°now Ùj°Û xalk“ (Il. 13.212).