Comments:Владимирцов 188, Poppe 28, Дыбо 14. Initial *l- is possible if we relate here Jurch. lar-ru- 'to write' (809) - the PTM form in this case must be reconstructed as *liru-. Despite Doerfer's criticism (TMN 4, 69-70), the etymology seems quite plausible (and the TM forms can certainly not be explained as borrowed < Mong.).
