Seefloth's Paradigm
Seefloth's Paradigm is a verbal inflectional paradigm reconstructed by the German linguist Uwe Seefloth for Uralo-Siberian in 2000 on the ground of data from Uralic and Eskimo-Aleut.
The paradigm
The paradigm encodes the person and number of the subject, and for transitive verbs, also the number of the object. The exact shape of the paradigm is not explicitly given in Seefloth's paper, but it can be summarized as this:
| Intransitive | Sg. object | Du. object | Pl. object | |
|---|---|---|---|---|
| 1sg. | -mV/-kV | -mV | -k-mV | -t-mV |
| 2sg. | -tV/-nV | -tV | -k-tV | -t-tV |
| 3sg. | -0 | -sa | -k-sa | -i-sa |
| 1du. | -mV-k | -mV-k | -k-mV-k | -t-mV-k |
| 2du. | -tV-k | -tV-k | -k-tV-k | -t-tV-k |
| 3du. | -k | -sa-k | -k-sa-k | -i-sa-k |
| 1pl. | -mV-t | -mV-t | -k-mV-t | -t-mV-t |
| 2pl. | -tV-t | -tV-t | -k-tV-t | -t-tV-t |
| 3pl. | -t | -sa-t | -k-sa-t | -i-sa-t |
Argumentation
Seefloth argues that Uralic and Eskimo-Aleut are "archaic peripheral languages" of the Uralo-Siberian group which allows to reconstruct a model of the morphosyntax of the simple clause on the basis of data from these two families. There were two nominal number markers, dual *-k and plural *-t/i. In the latter case, the form *-t was taken in word-final ("absolute") position and *-i when another morpheme, such as a possessive marker, followed ("construct" position). The oldest enclitics are possessive markers, 1sg. *-m/ka, 2sg. *-t/n and 3sg. *-sa. The duals of these were formed by suffixing *-k and the plural by suffixing *-t.
According to Seefloth, the resulting nominal possessive paradigm then became a verbal one by expressing sentences like I killed the reindeer (pl.) as The reindeer (pl.) are my killed ones, wherein my killed ones has the form killed-PL-1SG. The result is the paradigm given above.
Evidence
For Northern Samoyedic, the following paradigm can be reconstructed:
| Intransitive | Sg. object | Du. object | Pl. object | |
|---|---|---|---|---|
| 1sg. | -m | -mə | -kVj-i-nə | -i-nə |
| 2sg. | -n(-tə) | -rə (< -tə) | -kVj-i-tə | -i-tə (< -i-t-tə) |
| 3sg. | -0 | -ta (< -sa) | -kVj-i-ta | -i-ta (< -i-sa) |
| 1du. | -mi-ñ | -mi-ñ | -kVj-i-ni-ñ | -i-ni-ñ |
| 2du. | -ti-ñ | -ti-ñ | -kVj-i-ti-ñ | -i-ti-ñ |
| 3du. | -kə-ñ | -ti-ñ (< -si-ñ) | -kVj-i-ti-ñ | -i-ti-ñ (< -i-si-ñ) |
| 1pl. | -ma-t | -ma-t | -kVj-i-na-t | -i-na-t |
| 2pl. | -ta-t | -ta-t | -kVj-i-ta-t | -i-ta-t |
| 3pl. | 0-t | -to-n (< -so-n) | -kVj-i-to-n | -i-to-n (< -i-so-n) |
A similar paradigm can be reconstructed for Proto-Eskimo-Aleut:
| Intransitive | Sg. object | Du. object | Pl. object | |
|---|---|---|---|---|
| 1sg. | -[tk]-m-ka | -[m]-ka | -k-[m]-ka | -t-[m]-ka |
| 2sg. | -[tk]-ən | -n | -kə-n | -tə-n |
| 3sg. | -0 | -sa | -k | -i-sa |
| 1du. | -[tk]-mə-k | -mə-k | -k-mə-k | (=sg.) |
| 2du. | -[tk]-tə-k | -tə-k | -k-tə-k | (=sg.) |
| 3du. | -0-k | -sa-k | -kə-k | -i-sa-k |
| 1pl. | -[tk]-mə-t | -mə-t | -k-mə-t | (=sg.) |
| 2pl. | -[tk]-tə-t | -tə-t | -k-tə-t | (=sg.) |
| 3pl. | -0-t | -sa-t | -kə-t | -i-sa-t |
Object person?
Three groups of Uralo-Siberian languages - Mordvinic, Chukotko-Kamchatkan and Eskimo-Aleut - also inflect their transitive verbs for the person of the object. However, these paradigms are almost certainly not cognate. Also, if Seefloth's Paradigm descends from a possessive construction (which is quite clearly the case in Eskimo-Aleut and quite likely also in Northern Samoyedic), no object person marking is to be expected in the original paradigm.
Seefloth's Paradigm in Indo-European?
Jens Elmegård Rasmussen reconstructed the following verbal endings for Pre-Proto-Indo-European in 2003:
| Singular | Dual | Plural | |
|---|---|---|---|
| 1st person | -me | -G-me-G | -D-me-D |
| 2nd person | -te | -G-te-G | -D-te-D |
wherein *D and *G represent voiced dental and velar spirants, respectively. If one assumes that the number marker before the personal morpheme originally was an object number marker (on the ground that more often than not, multiple agents work multiple objects), these forms could be reduced to Seefloth's Paradigm. The actual existence of these pleonastic number markers in Pre-PIE, however, is uncertain.
A Proto-Mitian paradigm?
Given the fact that the paradigm is found in languages from opposite ends of the Mitian cluster, and that the validity of the Uralo-Siberian node within Mitian is uncertain, it is possible that it is of Proto-Mitian vintage, if such a protolanguage existed at all.
A mirage?
It is not certain at all that the paradigms Seefloth compared really have a common origin. Given the agglutinating nature of the paradigms, they may have arisen separately in the two language families, though the morphemes they are built of may actually be cognate. In this case, those Mitian families which do not display paradigms of this kind may never have had them, and any attempt to reconstruct "Seefloth's Paradigm" for Proto-Mitian would be misguided. What indeed speaks againt a Proto-Mitian paradigm here is that the matches between the forms are less than perfect; for instance, Northern Samojedic uses a dual object/possessum marker *-kVj- while Eskimo-Aleut uses just *-k-.
Sources
- Rasmussen, Jens Elmegård. 2003. Re: Nominative: A hybrid view. CYBALIST mailing list post #22349.
- Seefloth, Uwe. 2000. "Die Entstehung polypersonaler Paradigmen im Uralo-Sibirischen." Zentralasiatische Studien 30, 163-191.
