#3565 PTB *s-ŋar FROST / ICE (provisional)

In some TB languages, an apparent reflex of PTB *s‑ŋar may refer not to ‘frost’ per se, but instead refer e.g. to ‘frost’ or ‘dew’ equivocally, or to ‘dew’ alone. Cases in point are particularly evident within Lolo-Burmese, especially among closely related languages. For instance, WB has an etymon hnâŋ or hnɑŋ³, glossed ’dew’, but also as ‘fog, mist’; the notion ‘frost’ is expressed as WB hnɑŋ³khɑi³, i.e. ‘hard dew’. Similary, Atsi (Zaiwa) has ŋa̱n⁵¹ ‘dew’, while ‘frost’ is ŋa̱n⁵¹phju⁵¹ lit. ‘white dew’, and ‘ice’ is ŋa̱n⁵¹thuŋ²¹. Single lexemes which can mean either ‘frost’ or ‘dew’, apparently depending on context, are seen in Langsu (Luxi) ŋa̱ŋ³¹ and Maru (Langsu) ŋə̱ŋ³¹.

In Loloish, reflexes of PTB *s‑ŋar most often refer to ‘frost’ and/or ‘ice’, cf. Black Lahu ᴀ³³ŋə³³ ‘frost, ice’; the notion ‘dew’ is then expressed through reflexes of a separate etymon PTB *(d)zil ‘dew’ (PLB *ʔ‑dzi²), either alone or in compounds, cf. Black Lahu tsi³¹ɣə³¹, lit. ‘dew water’. The situation in Nusu (Bijiang), however, is noteworthy: PTB *s‑ŋar yields Nusu hɑ̃ʴ³⁵, which as a single lexeme is glossed ‘frost’, while the compound hɑ̃ʴ³⁵tɕhe³⁵ , literally ‘frost dew’, is the term for ‘dew’ (compare e.g. Nusu (Southern), which simply has tɕhe⁵⁵ ‘dew’). In Leqi (Luxi) (a Burmish language), ŋa̱n³³ by itself is glossed ‘ice’; the term for dew is ŋa̱n³³tʃu⁵⁵, lit. ‘ice dew (?)’, ‘frost’ is ŋa̱n³³phiu³³, lit. ‘white ice’.

In dealing with the above examples, it should be kept in mind that in the real world, the phenomena we call ‘fog, mist, dew, frost’ refer to points along a continuum. When air temperature cools to near the dew point, then we see ‘fog’ or ‘mist’. When the surface temperature of an exposed, thin object cools to near the dew point, then we see ‘dew’. When the surface temperature of that object cools to near a dew point that is already below freezing, then rather than ‘dew’, we see ice in the form of ‘frost’.

We can, then, cautiously say that some languages describe this meteorological continuum using relatively more single lexemes (e.g. English and Chinese), while other languages use relatively fewer single lexemes but more lexemic compounds (e.g. Burmese, Atsi, Langsu and Maru). Regarding Nusu (Bijiang) hɑ̃ʴ³⁵ and Leqi ŋa̱n³³, we surmise that the underlying lexemic meaning is maximally inclusive, i.e. ‘condensed/congealed water in general’; this may account for the seeming disparity between the surface connotation of single lexemes vs. the same lexemes in compounds.

Admittedly, our suppositions above are based on data likely collected in part via prepared word lists, which can unwittingly impose lexemic categories and distinctions that speakers of a particular language being studied may not find natural. More in-depth and beyond the glossary studies of Lolo-Burmese languages will hopefully become available to help us better shed light on this interesting area of the TB lexicon.