Notes :In the North Dravidian database cf. also *ā-r, *ā-s, *ā-d, etc. (Additional derivative entries are also available in Telugu and other databases). For Brahui dā see DEDR 1. // Present in 6/6 branches.
Notes :A rather common negation morpheme with well-known Nostratic parallels. Its main competition is PDR *śil- with more or less the same meaning (sometimes the two forms are in complementary distribution between branches - e. g. only *śil- is found in Telugu, only all- in Brahui). // Present in 4/6 branches.
Notes :Cf. also Kui alāṛi 'fatigue, distress from fatigue, exhaustion'. See the South Dravidian entry for the possibility of separating SDR *ala- 'suffer' from SDR *alai- 'wander, shake; caus. harass, annoy'. In this respect it is interesting to consider some of the syncopated/non-syncopated doublets given by Burrow & Emeneau in the same entry: cf. Kui alāṛi vs. laha 'languor, laziness'; Kur. algā vs. layākoyā, etc. Of course, a final judgement is different to give based on lack of definitive knowledge on the rules of vowel reduction in Gondwan (not to mention North Dravidian, where initial l- is exceedingly rare in the first place). // Present in 5/6 branches.
Notes :In Telugu the root has only been preserved as part of obscure or expressive compounds, but in SDR it is highly productive. PNDR *al-G- 'to laugh' is, for obvious reasons, set apart in DEDR, but there are actually no insurmountable obstacles to merging the two entries neither from the phonetic nor the semantic side. // Present in 3/6 branches.
Notes :Tamil distinguishes between alaku 'blade' and alakku 'splint'; however, given the general lack of stability among 2nd syllable stops, this differentiation should probably be considered secondary. The word formally resembles a verbal derivative, but no verbal stem like *alV- 'cut' or 'be sharp' is found in Dravidian. // Present in 2/6 branches.
Notes :A good PDR root with many suffixal derivatives (*ala-ŋ-, *ala-mb-) and rather wide-reaching semantics. Occasional contaminations with *alá- 'to suffer, be confused' are noted. Not all the forms in DEDR 240 can actually belong here without violating correspondences: e. g. Kui lānga 'to be fickle, changeable' should be discarded because Kui l- < PGn (PDR) *ḷ-, whereas alveolar l usually yields ḍ- or ṛ- when found in initial position. // Present in 3/6 branches.
Notes :Burrow & Emeneau also include Kui akali 'rinsing' in the comparison, but the relation between these forms would have to be quite mysterious (metathesis from *ala-ki?). Auslaut -ai- is well reconstructible on the basis of SDR, which makes it possible to distinguish between this PDR root and, say, PDR *alá- 'to make verbal noise'. // Present in 2/6 branches.
Notes :The stem is never met without the final resonant, although it certainly cannot be guaranteed that *-r- was not detachable as a suffix in PDR. It is interesting to note the similarity between *alár- and PDR *mala[r]- with pretty much the same meaning. However, to treat this case as an irregular development *m- > 0- (see, e. g., [Андронов 1994: 86] ) is hardly admissible. Much more probable is semantic confusion between two originally different roots; in fact, SDR malar 'to blossom' may eventually be nothing but the result of contamination between *alár- and *mal- 'fertile, to abound'. // Present in 2/6 branches.
Notes :Reconstructed for the PDR-Telugu level. The stem often behaves in an irregular way, which is common for plant names in Dravidian. // Present in 2/6 branches.
