Hungarian Historical Phonology fék
fék 'bridle, rein'
First attestation/Old Hungarian data
? 1215 Fecu (EWUng 369, s.v. fék2)
Important dialectal forms
[coming]
Uralic/Ugric/Pre-Hungarian reconstruction
Pre-Hungarian *pekkV- < ? Proto-Ugric
(Disputed:)
UEW: PUg *päkkɜ 'bridle (n), rein; Zaum, Zügel'
Status of the Ugric etymology
Improbable (phonologically irregular)
Loan etymology
PUg ← East Iranian *pekkə < PIr *paxštaka- ’megkötés, kötőfék’ (Harmatta 1997)
Possible substrate loans (new suggestion)
Cognates suggested in earlier research
Khanty: East (V) päk 'Zügel (des Rentiers); Zaum (des Pferdes)'
Mansi: ("Old Mansi") пехъ 'дюжина уз(д)а'
Proto-Ob-Ugric: ?
Commentary
Both the Ugric and the Iranian etymologies are problematic. Hungarian fék rather points to *e, whereas Khanty ä points to *ä. It is more difficult to assess the Mansi material as the word is attested only in old sources, but it is clear that a Proto-Ugric form cannot be reconstructed. The irregularity fits well the other Ugric lexemes related to horses and pastoralism. The words can thus be substrate loans. EWUng remarks that the word is found only in southern Ob-Ugric variaties, which points to an original dialectal distribution in southern Ugric only. This is possible, and such a limited distribution could well fit a substrate word or a loan. However, it cannot be ruled out that the reflexes can have become lost in more northern varieties of Khanty and Mansi.
An Iranian loan etymology has been suggested by Harmatta (1997), who reconstructs an "East Iranian" form *pekkə. It is unclear what is this reconstruction is based on. A Proto-Iranian form *paxštaka- could not yield *päkkV or *pekkV in the Ugric languages.
Conclusion
The Iranian loan etymology is unconvincing; the Ugric etymology is irregular. The words in Hu, Kh (and Ms?) are possible substrate loans.
References
EWUng 369, s.v. fék2: PUg
Harmatta 1997
UEW: Proto-Ugric Uralonet