One of the possible unpredictable paths that language development can take is exemplified by the assimilation of loan words, wherein something that is at variance with the linguistic patterns of the donor language is adopted by the borrowing speech community anyway, and only owing to the imputed prestige of the first transmitter(s) of the mistaken form.
One is reminded here of the popular Japanese proverb––Ikken kyo ni hoete banken jitsu o tsutau––which loosely translated means ‘One dog barks out a lie and ten thousand dogs take it up as the truth’.
This sort of situation must be what explains the consistent misstressing one hears in the Anglophone media of the Slavic surnames of tennis players, particularly of the swarm of Russian women that inhabit the current ranks of tennis professionals. Take the names of two prominent women, Maria Sharapova and Svetlana Kuznetsova, who are (again) among the many playing at this year’s U. S. Open Championships. The monolingual TV announcers who have to struggle with the pronunciation of their surnames follow what is now the established norm in tennis parlance, with penultimate stress in the first name and antepenultimate stress in the second, i. e., Sharapóva and Kuznétsova. Note that both surnames have four syllables and end in –a (the Russian feminine ending). Accordingly, following the native English stress pattern for such quadrisyllabic items, they should both be pronounced with main stress on the penultimate syllable, i. e., as in bossanova or, for that matter, panegyric.
Now, it so happens that both of the English adaptations are wrong from the point of view of their authentic form in Russian. In these women’s native language it is Sharápova, which like all Russian family names in –ov/-ova goes back to a possessive adjective derived from a nominal base (here the dialectal sharáp ‘theft’) and mimics the fixed stress on the second syllable of the stem throughout its paradigm (Nom sharáp, Gen sharápa, etc.); and it is Kuznetsóva (< kuznéts ‘blacksmith’), because it follows the stress pattern Nom kuznéts, Gen kuznetsá, etc., with stress on the first syllable of the suffix (= ending) in the oblique cases, which corresponds to the penult in the derived feminine surname in the nominative case.
By rights, if one is going to pronounce both names according to English stress rules, then both Sharapova and Kuznetsova should have stress on the penult. This would be at odds with the authentic Russian stress in the first case but would coincide with it in the second. It would, of course, take only a very small effort to pronounce them both “correctly,” since Sharápova would fit the model of unfláppable (which is quadrisyllabic in English despite its trisyllabic orthography, the final vowel being silent but a reduced vowel being pronounced between the final two consonants) and Kuznetsóva would conform to the type Manitóba.
Consequently, the two mistaken stresses––from the point of view of Russian––of Sharapóva and Kuznétsova––are of an unequal degree of inauthenticity or falsity from the point of view of English. The first comes about simply by the application of normal English stressing rules to quadrisyllabic names of foreign origin. But the second is only explicable as a mistake pure and simple, one that was first made by an American or British speaker––doubtless a TV commentator––but whose prestige licensed its endless repetition as the accepted form by the myriad listeners who took it at face value.