Note: The older speculation on an origin for the Germanic etymon is discussed exhaustively by Anatoly Liberman in An Analytic Dictionary of English Etymology: An Introduction (University of Minnesota Press, 2008), pp. 3-9. Among the conjectures are descent from Indo-European *b^heh₂- "shine, give light, appear" (see fantasy entry 1), *b^hou̯g^h- "bend" (see bow entry 1), or *b^herǵ- "shining" (see bright entry 1), with varying ablaut grades, root enlargements and degrees of phonetic attrition; and borrowing from Latin būcina "horn, trumpet" (used as a signal). Liberman's own hypothesis depends crucially on forms without -(V)n, as early modern Dutch baeck "beacon, lighthouse," claimed to be Middle Dutch by De Vries ("reeds m[iddel]n[eder]l[ands]") and van Wijk ("zeldzame vorm"), and Low German bak, bake. (M. Philippa, et al., Etymologisch Woordenboek van het Nederlands, on line at etymologiebank.nl, have it no earlier than 1559.) Liberman rejects the idea that these later words are simply shortened from the forms with n. Since *bak-/bāk- and *baukn- cannot be reconciled by ablaut rules, he hypothesizes that they are part of a large network of Germanic words built from the consonantal frame b-g/b-k that denote "objects capable of inflating themselves and making noise" (p. xxxiii). The Germanic etymon would hence have originally denoted a floating object (a bladder?) marking a channel, whence it was generalized to denote any kind of signal. The specific form *baukn- was formed by analogy with the semantically close derivative *taikn- "sign" (see token entry 1). Though the existence of the group of affective words that Liberman postulates seems highly probable, his inclusion of *baukn- in the group is questionable. Most notably, the forms alleged to have original short or long a are extant only in the coastal languages, Dutch and Low German, that could have borrowed the word from Frisian, where *-ā- is historically the regular outcome of *-au-. It would seem preferable to devise a way to delete the n (back-formation from a plural?) than to depend on a string of speculative semantic shifts ("inflated object that makes noise when squeezed" > "object that floats" > "float, buoy" > "marker, signal") to account for the words, even if this would continue to deprive us of an etymology. West Germanic *baukna- is not treated in G. Kroonen, Etymological Dictionary of Proto-Germanic (Brill, 2013).