I believe all language families are connected to each other in a single language family. My experience from learning languages the world over indicate that I can find similarities that are not just coincidences from complete opposite sides of the earth. For example in many Formosan indigenous languages, darkening (as in light from the sky, or clouds) takes on a final -um sound. I've seen this repeated in numerous language families worldwide. The other day I came across the Latvian word for getting dark see sentence 85 in my phrase list, or search "dark" if it's moved which is "tumšs". Given that lots of words in Baltic languages have these -s endings, what I found of interest was the -um- part of the word.
I decided to research it more, and I found Allan R. Bomhard's research with a lot of updated research on this science and critical reviews of previous research. It looks promising but it fails to take into account the Asian S-DNA O haplogroup of language families. If we analyse this haplogroup, we find that Sino-Tibetan is most likely the oldest of all of them, Austroasiatic somewhere in the middle, and the newest language branches are Austronesian and Kradai which are most likely siblings of each other. These newest branches are somewhere around 7000-10,000 years old, which makes the other branches older by millennia. Which is precisely why linguists are very apprehensive about reconstructing too far back in time. Anything between then and a proposed proto-language is really just speculation.
But one cannot turn a blind eye to all the amazing similarities one finds if you just jump in and learn languages from around the world. I believe that Austronesian is deeply connected to Bomhard's Nostratic, and since it's embedded in the Southeast Asia stock of families, then all of them are connected to Nostratic by association.
Is it also not telling that the Uralic and Turkic branches have N haplogroup, proceeding O by one letter, and then R (Indo-European) comes directly after (since P evolved quickly and little trace remains). If one considers that the language families of haplogroups G (Kartvelian), J (Semitic), H (Dravidian), N (Uralo-Turkic), and R (Indo-European) are all subsumed under Bomhard's Nostratic, then there's no reason to not assume O shouldn't be there as well.
Turning to Bomhard's analysis. Let's look up "dark". In the index on pages 194-195 we find a list of dark-like vocabulary. So let's look at each item in turn: 177 "dum", 178 "dum-k'w", 214 "tʰum-", 975 "rum", 782 "ɣam". There are other roots that also mean "dark" but let's just focus on these. What's noticeable is that despite different onsets, they all rhyme. Why do they all rhyme, and why can different onsets that have the same rhyme all point to the same meaning? This also reminds me of the Hokkien pronunciation of dark: 暗 is "am", which lacking an onset consonant, and a slight vowel change, also ends in this final -m.
Looking up item 177, Proto-Nostratic root *dum- (~*dom-): the verb means to cover over, obscure, cloud over, become dark, make dark, darken. The noun means darkness, cloud, fog. The adjective means dark, cloudy. The derivative *dum-k'wa makes me immediately recall German dunkel. Then there's English "dim" and Slavic "dym" (smoke), see also Latin descendants "fum" (smoke), but also Slavic "tma" and "temny" (dark).
Now I'd like to point attention to the phrase lists I've posted on this site. Let's take a look at one of the oldest Austronesian languages, Thao (SSF). Sentence number 0746 says The sun is growing dark "Tilhaz mindumdum" where m- is the verbal prefix, and -in- is an infix meaning "become", and again we have a doublet in sentence 3847 which says When it's going to rain it grows dark "Qali ya amaqusaz mindumdum/minhumhum" where we see that either mindumdum or minhumhum are possible. Many words appear in Thao with the suffix -iza meaning "has already", so we can search the page and find plenty of dumdum (6 instances) and humhum (55 instances) vocabulary with a variety of prefixes and suffixes. The word "cloud" in Thao is "urum" (2 instances).
In Vietnamese (VIE) we find đậm meaning dark in colour, rum meaning violet, đêm meaning night, and thâm meaning in the dark or deep colour. In Thai (THA) we find ค่ำ khàm meaning night, and ดำ dam meaning black.
In Bomhard's other book Toward Proto-Nostratic, comparing Indo-European to Afroasiatic, we find item 51 on page 207 which lists Proto-Afroasiatic "*tam" meaning to cover up and Egyptian "tms" meaning to hide, cover over, bury. He also lists Sanskrit "támas" (darkness) and Latin "tenebrae" (darkness).
Let's turn to Robert Blust's online Austronesian Comparative Dictionary for "dark" which lists an overwhelming amount of evidence for this -um ending, and for inclusion of Austronesian into Bomhard's Nostratic.
Now let's turn to the Sino-Tibetan etymological database (STEDT) item number 2624 which lists PTB (Proto-Tibeto-Burman) *r(u/i)m DARK / SHADE / DUSK / TWILIGHT. This also matches the multiple meanings in Austronesian Thao (SSF). This page also lists the Chinese cognate as 陰 "yin", the same yin in 陰陽 yin-yang but doesn't list 暗 Hokkien "am".
Starostin also proposes Borean CVMV (where V=vowel placeholder) for dark/black, Eurasiatic *sVmV, Semitic *šħm-, and Macro-Khoisan *Cum (shadow).
Polymath, Sep 30, 2022