bone

 

T.L. Markey proposed an etymological connection between Germanic *baina- 'bone' and Latin _fi:nis_ 'border, boundary, limit' (Gmc. *baina- 'bone' and Other Monstrosities, NOWELE 2:93-107, 1983).  Relying on a long metalinguistic preamble involving the phenomenology of perception, he argued that the inherited word for 'bone' was replaced by one for 'skeleton' when Germanic perceptual reinterpretation required new terminology.  This term for 'skeleton' would originally have signified 'item defining limits of a pattern, limit of a configuration' vel sim. in agreement with Lat. _fi:nis_ as 'delimitation, border'.  According to Markey, _fi:nis_ was used "usually in reference to a tree, or an upright post that designated territorial limits" and can thus also be brought into connection with Old Norse _beinn_ 'upright, straight, even, favorable'.  This adjective is attested as a simplex only in North Germanic, but Markey noted the _Ba:ningas_ in Widsith and Latinized Frankish _Bainobaudes_.

 

Apart from the twilight-zone metalinguistics underlying the presumed development 'skeleton'

> 'bone', Markey's explanation falters with the Latin usage of _fi:nis_.  A tree or upright post can mark only one

point on a territorial border.  The border itself either follows a natural geographic feature, such as a river or ridge-line, or it is artificially established by beating a path.

 

In a paper submitted before he read Markey's, E.P. Hamp proposed a much more mundane motivation for Germanic replacement of the inherited word for 'bone' (German _Bein_, Old English _ba:n_; Slavic _kostI_, NOWELE 6:67-70, 1985).  A thematized form of PIE *h2e/osth1- 'bone', Gmc. *asta-, would have fallen together with *h2ozdo- 'branch', also Gmc. *asta- (Gothic _asts_ etc.).  This would lead to the use of a disambiguating participle with the bones of butchered animals, *bH@... 'cut, lopped off' from the PIE root *bHeih2- 'to strike, beat' (whence Old Irish _ben(a)id_ 'beats' etc., IEW 117-8).  The Gmc. phrase *bainaN astaN 'lopped off bone/branch' would simply lose its head noun, leaving *bainaN to signify 'bone'.

 

In fact one expects the oxytone PIE participle *bHih2-nó- to lose its laryngeal by Dybo's Law in Germanic, yielding *bina- not *baina-.

 

A. Bammesberger agreed with Markey's connection of Gmc. *baina- with ON _beinn_ and Lat. _fi:nis_ (Lateinisch _fi:nis_ und urgermanisch *baina-, HS 103:264-8, 1990).  His paper is cited favorably by M. de Vaan (EDL s.v. _fi:nis_), with the PIE root in question the same as Hamp's.  Implicit in de Vaan's treatment is the semantic development 'boundary marker beaten into the ground' > 'border' for _fi:nis_ (whose PIE protoform would need to be accented *bHíh2-ni- to get the long Latin vowel after Dybo's Law).  For *baina-, it would be something like 'boundary marker beaten into the ground' > 'upright marker' > 'upright (member), (leg-)bone'.

 

I believe the PIE root has been identified correctly, but the semantic side needs revision.  Regarding _fi:nis_, I have already noted that a border is created artificially by beating a path.  This explains the connection between Umbrian _tuder_ 'border' and Latin _tundere_ 'to beat'.  For some reason de Vaan (s.v. _tundo:_) instead follows S. Schumacher's unlikely notion that _tuder_ means 'the place where two areas "hit" each other, that is, border on each other' (Die keltischen Primärverben, Innsbruck 2004, p. 645).  On the other hand, against Walde-Hofmann and Ernout-Meillet, de Vaan regards the semantic connection between Lat. _callis_ 'rough track, path' and _callum_ 'hard substance, induration' as quite plausible.  Beating a path hardens its surface.  I propose that _fi:nis_ is semantically parallel to _callis_, and for some reason _fi:nis_ became specialized in denoting the path marking the territorial border of a political entity, like Umb. _tuder_.

 

In Germanic, PIE *bHoih2-no- with either accent would yield *baina-.  I propose that the semantic development was 'beaten hard on the surface' > 'hard-surfaced, case-hardened'.  At this point *bainaN astaN disambiguated 'bone' from 'branch', since the deep interior of a bone contains soft spongy marrow, while the exterior is hard.  While *bainaN eventually lost its head noun and became the neuter 'bone' as in Hamp's scenario, the adjective *baina- developed from 'hard-surfaced' to simply 'hard, unyielding'.  The simplex is unattested in Gothic, but the compound _bainabagms_ 'cornel tree' illustrates the sense, for the cornel or dogwood has hard wood, difficult to use.  In North (and probably West) Germanic, the adjective acquired a morally positive sense suitable for proper names, 'unyielding, steadfast, upright, straight, even'.

 

Douglas G. Kilday