Mitian is a label suggested by the linguist John Bengtson for a cluster of language families of northern Eurasia and the North American Arctic which are characterized by similar 1st and 2nd person pronoun roots, *mi and *ti, respectively, or similar (hence the designation "Mitian"). The group consists of the following language families (in a west-to-east order, according to Jörg Rhiemeier):
Possible further additions may include:
- Tyrrhenian (Etruscan-Rhaetic-Lemnian)
The Mitian group of languages is enormous in its extent. It encompasses most of northern Eurasia, the Indian subcontinent and the North American Arctic; the European colonial expansion in the 15th to 20th centuries also took Indo-European languages to many diverse locations in the world, especially the Americas, Australia and Siberia.
Mitian languages are spoken by about 3,000 million people altogether, the largest part of them being speakers of Indo-European languages. The Mitian languages form a contiguous belt around the North Pole, interrupted only by the stretches of ocean that separate the continents.
This makes is somewhat arbitrary to speak of "western" and "eastern" Mitian languages; but given the fact that the westernmost Indo-European languages entered their present locations from the east, and the easternmost Eskimo-Aleut languages entered their present locations from the west, it is reasonable to "cut open" the belt at the Atlantic and consider Indo-European the westernmost and Eskimo-Aleut the easternmost Mitian family. That way (neglecting colonial-era expansions of Indo-European languages), Icelandic (Indo-European) is the westernmost and East Greenlandic (Eskimo-Aleut) the easternmost Mitian language.
It is obviously easier to define the northernmost and the southernmost Mitian languages. Again neglecting colonial-era developments and considering indigenous languages only, the northermost Mitian language (actually, the northernmost language of any family) is Polar Eskimo (Eskimo-Aleut), while the southernmost Mitian language is Divehi (Indo-European).
Is Mitian a family?
The resemblance of a pair of personal pronouns, however, is generally not considered sufficient to prove a relationship. Yet, the fact that these pronouns bear resemblance within a group of at least eight language families occupying a contiguous area strongly suggests that a common origin of these languages is likely, especially as the area in question has been characterized by large-scale migrations throughout history (and probably ever since the last glaciation) and thus constitutes a typical linguistic spread zone in which a single linguistic lineage could have spread across the entire area within a few thousand years.
The case needs to be investigated further by means of the traditional comparative method.
So far, one can only speculate about the phonology of Proto-Mitian. The phonology reconstructed for Uralo-Siberian by Michael Fortescue, however, can give an idea of what Proto-Mitian phonology may have been like, perhaps better than the large inventories reconstructed by the Nostraticists which owe their size to the inclusion of Afroasiatic, Kartvelian and Dravidian.
A very preliminary guess at Proto-Mitian morphology
We do not yet have a reliable reconstruction of Proto-Mitian phonology, though some endeavours have been made by several authors. This means that the exact phonological shapes of the morphemes discussed below are unknown, and the letters used in the forms are to be taken as cover symbols for quite broad ranges of sounds more so than in conventional reconstructions, e.g. *ti means 'a coronal obstruent followed by a front vowel'. Moreover, at least some of the correspondences may be spurious, as we don't know whether the forms compared, as similar as they may be, actually match.
The pronouns are the elements for which "Mitian" is named. These may be tabulated as this:
I.e., the dual suffix is *-k and the plural suffix *-t.
There seem to have been two interrogatives, animate *ku and inanimate *mi.
Nouns used the same number markers as pronouns: dual *-k and plural *-t. As for case, all Mitian families have well-developed case systems; the only Mitian languages lacking this category, all belonging to the Indo-European family, have lost it rather recently (within the last 2000 years). Yet, the case systems do not seem to match very well, and most forms must be regarded innovations. IE and Uralic share an animate accusative *-m, which may be connected to the Eskimo-Aleut "relative" (ergative-postpositional) case marker *-m. In Indo-Uralic, this may have been an "animacy neutralizer" meaning something like 'the thing that is a ...', used in syntagms that do not require animacy. Another common oblique case marker in Mitian seems to have been *-n (not surviving in IE, unless the adjective-forming suffix *-no- is connected with it).
It may have been the case that the Proto-Mitian case markers did not combine with number markers, or the number markers changed when a case suffix was added, with the plural *-t becoming *-i (as still in Finnish). Inanimate nouns perhaps were not inflected for number. The IE plural case forms are difficult to account for.
A set of personal suffixes identical to the personal pronouns is found in most Mitian languages. IE, Hungarian, Selkup and Eskimo-Aleut show evidence of a second, perhaps older, set with stative or intransitive meaning. PIE *-h2e, Hung./Selkup *-k and Eskimo-Aleut *-ka point at Proto-Mitian 1sg. stative *-ka or *-xa. The second person is more difficult, PIE *-th2e, Hung. -sz (/-s/) and Selkup *-ntï don't seem to match. Carrasquer mentions an Eskimo-Aleut form *-tka, which, if real, would suggest relationship to the PIE form. The third person probably was zero. The question of a bipersonal paradigm with forms such as *-tka-mi '1sg->2sg' is open; the bipersonal paradigms found in Mordvin, Chukotko-Kamchatkan and Eskimo seem to be innovations.
Suggested subgroups of Mitian
The Indo-Uralic hypothesis, linking Indo-European and Uralic, has many adherents. Also very popular is the hypothesis of a close relationship between Uralic and Yukaghir. Michael Fortescue, in his 1998 book Language Relations Across Bering Strait, proposes a Uralo-Siberian group consisting of Uralic, Yukaghir, Chukotko-Kamchatkan and Eskimo-Aleut.
The Turkic, Mongolic and Tungusic languages have long been considered branches of a single family called Altaic, to which some scholars furthermore added Korean and Japanese. The validity of this family is now controversial.
- Hesperic by Jörg Rhiemeier
- Danubian by Taylor Selseth
- numerous a posteriori diachronic conlangs based on various Mitian families
- Euro-Siberian - a subgroup of Mitian