According to
Sergei Starostin, (2002), Borean is divided into two groups, Nostratic (
sensu lato, consisting of
Eurasiatic and
Afroasiatic) and
Dene–Daic, the latter consisting of the
Dené–Caucasian and
Austric macrofamilies. Starostin tentatively dates the Borean proto-language to the Upper Paleolithic, approximately 16 thousand years ago. Starostin's model of Borean would thus include most languages of
Eurasia, as well as the Afroasiatic languages of North Africa and the Horn of Africa, and the
Eskimo–Aleut and the Na-Dene languages of the
New World.
Murray Gell-Mann,
Ilia Peiros, and
Georgiy Starostin maintain that the comparative method has provided strong evidence for some linguistic superfamilies (
Dené-Caucasian and
Eurasiatic), but not so far for others (
Afroasiatic and
Austric). Their view is that since some of these families have not yet been reconstructed and others still require improvement, it is impossible to apply the strict comparative method to even older and larger groups. However, they consider this only a technical rather than a theoretical problem, and reject the idea that linguistic relationships further back in time than 10,000 years before the present cannot be reconstructed, since the "main objects of research in this case are not modern languages, but reconstructed proto-languages which turn out to be more similar to one another than their modern day descendants". Gell-Mann
et al. note that their proposed model of Borean differs significantly from that of Fleming. The phylogenetic composition of Borean according to Starostin is as follows: • "Borean" •
Nostratic (
fringe theory,
Holger Pedersen 1903) •
Eurasiatic (widely rejected,
Joseph Greenberg 2000) •
Indo-European (widely recognized family) •
Altaic (widely rejected;
Roy Andrew Miller 1971,
Gustaf John Ramstedt 1952,
Matthias Castrén 1844) •
Japonic (widely recognized family) •
Koreanic (widely recognized family) •
Turkic (widely recognized family) •
Tungusic (widely recognized family) •
Mongolic (widely recognized family) •
Uralic (widely recognized family) •
Paleo-Siberian (phylogenetic unity widely rejected) •
Eskimo–Aleut (widely recognized family) •
Chukotko-Kamchatkan (widely recognized family) •
Yukaghir (language isolate) •
Nivkh (language isolate) •
Kartvelian (widely recognized family) •
Dravidian (widely recognized family) •
Afroasiatic (widely recognized family) •
Dene–Daic (widely rejected, Starostin 2005) •
Dené–Caucasian (widely rejected, Nikolayev 1991; expanded by Bengtson 1997), cf.
Dené–Yeniseian (
Edward Vajda 2008) •
Na-Dené (widely recognized family) •
Basque (language isolate) •
Iberian (language isolate; not explicitly mentioned in Starostin's tree diagram, but usually seen as likely belonging to a common proto-family with Basque) • Sino-Caucasian (widely rejected, Starostin 2006) •
Sino-Tibetan (widely recognized family) •
Yeniseian (widely recognized family) •
Burushaski (language isolate) •
North Caucasian (widely rejected; Nikolayev & Starostin 1994) •
Northeast Caucasian (widely recognized family) •
Northwest Caucasian (widely recognized family) •
Hattic (language isolate; not explicitly mentioned in Starostin's tree diagram, but in other works by him and his colleagues often associated with North Caucasian or treated as an independent branch of Dené–Caucasian) •
Hurro-Urartian (widely recognized family; not explicitly mentioned in Starostin's tree diagram, but in other works by him and his colleagues often associated with North Caucasian or treated as an independent branch of Dené–Caucasian) •
Austric (speculative,
Wilhelm Schmidt 1906) •
Austro-Tai (speculative, Paul Benedict 1942) •
Austronesian (widely recognized family) •
Tai–Kadai (widely recognized family) •
Hmong–Mien (widely recognized family) •
Austroasiatic (widely recognized family) •
Ainu (language isolate; not explicitly mentioned in Starostin's tree diagram, but in other works by him and his colleagues mostly associated with Austric) •
Sumerian (language isolate; not explicitly mentioned in Starostin's tree diagram, but in other works by him and his colleagues mostly associated with Dene–Daic or Nostratic) •
Elamite (language isolate; not explicitly mentioned in Starostin's tree diagram, but in other works by him and his colleagues mostly associated with Dene–Daic or Nostratic) ==Jäger (2015)==