Hungarian Historical Phonology hát
hát 'back'
hátra 'zurück, rückwärts, rücklings'
First attestation/Old Hungarian data
1109 place name Erhat
See EWUNg
Important dialectal forms
No notable dialectal variation (ÚMTSz: 874-875, s.v. hát)
Uralic/Ugric/Pre-Hungarian reconstruction
? Proto-Ugric (Hu + Ms) *katta- or *kotta-
Abondolo 1996: 103: *kïïttV 'shadow; back'
UEW: PU *kuttɜ ‘Rücken; back‘
Helimski: Marginalia... PU ? *kottз
Status of the Ugric etymology
Uncertain
Loan etymology
None suggested so far
Cognates suggested in earlier research
UEW:
Khanty: East (J) juw kutsa 'unter den Baum, im Schutz des Baumes', kutəʌ: juw kutəʌnə 'im Schutz des Baumes'
Mansi: East (KU) χūtəj 'im Schatten, im Schutz', West (P) kūtəγ 'im Schatten, im Schutz', North (N) χūtäi̊̕ 'hinter (akk.)' < ? PMs *kūtəγ
Selkup (Taz): qottä, qott 'backwards; навзничь'
Commentary
Hungarian hát is considered an inherited Uralic word by the UEW and the Hungarian etymological dictionaries (EWUng, MSzFE). However, the Hungarian word or its alleged cognates is not mentioned by Sammallahti (1988). Abondolo (1996) mentions the Ugric cognates but omits the Selkup cognate without comment. Helimski (Marginalia ad UEW) supports the Proto-Uralic etymology but reconstructs PU *o, arguing that Selkup o can not reflect *u but the Ugric allows either *o or *u. However, due to semantics, word-class differences (only adverb in Samoyed) and the limited distribution of the word within Uralic, Helimski does not consider the etymology completely certain ("eher plausibel als völlig sicher").
Helimski's arguments about the vocalism of this word are not entirely convincing: Hungarian á can reflect either *a or *o, and also Mansi *ū can point to both *a-a and *o-a stem. *u is out of question here. East Khanty u can reflect Proto-Khanty *ū, which does not fit to any of the possible vowels mentioned here: in *a-a stems Proto-Khanty *ū appears regularly after a velar or word-initially (Zhivlov 2014: 117), but after *k this is unexpected.
The idea that East Khanty kut- is a cognate to Hungarian hát and Mansi *kūtəγ should be dropped. The connection with the Selkup word remains uncertain, and the similarity can also be accidental.
Conclusion
Due to phonological irregularities, both the PU and the PUg backrdound of Hungarian hát remains uncertain.
References
Abondolo 1996: 50: PUg
EWUng 537, s.v. hát: PU
Helimski Marginalia ad UEW: PU
Honti 1982
MSzFE 276-277, s.v. hát: PU
UEW s.v. kuttɜ: Proto-Uralic Uralonet