Revisiting my posting “The accent in Polish”, with a cartoon in which a Mr. Waterski corrects a desk clerk’s pronunciation of his name. The correction comes in two parts:

(1) a statement of fact about this particular name: “The accent is on the second syllable”; and

(2) an appeal to a generalization: “like every other Polish surname”

Part (1) is a brute statement of fact and it’s largely inarguable: people’s names are to be pronounced as they say they are (so long as this pronunciation is consistent with the phonology of the target language). If your family name is Taliaferro and you’re a Virginian who pronounces the name like Tolliver, then that’s the pronunciation (but you can’t insist that in English it’s pronounced [ˌtaʎʎaˈfɛrro], as in Italian).

Part (2), however, makes a claim about Polish — that the accent in Polish surnames is on the second syllable — and that generalization can be tested against the evidence. In fact, it is incorrect, and simultaneously (in a sense) insufficiently general.

Comments on my posting make both of these points.

Second syllable? First, Dennis Preston notes that there is a generalization here, but that it’s correctly stated by reference to the penultimate (next-to-last) syllable, not the second syllable. In the three-syllable name Watérski, the accented syllable is both the second syllable and the penult, and that will be true for other three-syllable surnames in -ski / -sky. The test cases will be two-syllable surnames and longer (in particular, four-syllable) surnames.

Some two-syllable -ski / -sky names:

Barsky, Brodzki / Brodsky, Stawski / Stawsky

(These are accented on the first rather than the second syllable: the penult.)

And some four-syllable -ski / -sky names:

Dubanowski, Filipowski, Malinowski, Sokolofsky / Sokolowski, Winogrodzki

(These are accented on the third rather than the second syllable: again, the penult.)

The same is true for other types of surnames: two-syllable Bosko, Golomb, Janda, Nowak / Novak, Sokal, Wojda (initial accent) and four-syllable Mikolajczak, Walentowicz (third-syllable accent).

An alternative generalization. In the -ski / -sky names, the penult is also the syllable immediately preceding the name-forming suffix, so we could entertain the alternative proposal that accent is conditioned by the suffix. Such things do occur, in particular with certain suffixes in English, for example the -ity deriving abstract nouns from adjectives; note the accent shifts in:

authentic – authenticity, electric – electricity, hostile – hostility, lucid – lucidity, timid – timidity, vulgar – vulgarity

and the -ian deriving adjectives from personal names:

Clinton – Clintonian, Edward – Edwardian, Orwell – Orwellian

But the penult generalization for Polish names isn’t suffix-specific; it applies across the board.

Surnames only? Another commenter on my Rhymes posting(with the catchy username “g2-dcfb9e6caa86f3ff2b982de5354b50d9”) notes (correctly) that

aside from a few exceptions (clitics, some nouns of Greek origin [and some loanwords]) the stress in Polish is always on the penultimate syllable

That is, the placement of accent in Waterski etc. isn’t specific to surnames at all, but applies quite generally in Polish; Polish is one of the many languages of the world with thorough-going penultimate accent. (The other very common systems have thorough-going initial accent (as in Finnish) and thorough-going final accent (as in French).

Putting the two comments together:  Mr. Waterski should have protested that

The accent is on the next-to-last syllable, like almost every other Polish word.

(or he could have said “the penultimate syllable”).

One Response to “Generalizations”

  1. arnold zwicky Says:

    More extensive discussion of accentuation (but still at an elementary level) in my paper “Word accent, phrase accent, and meter” (Innovations in Linguistics Education, 1982), here.

Leave a Reply