There is something daunting about
Polish pronunciation, because of the way it's written, but it's not nearly as formidable as it seems. It's one of the
loveliest languages I know to
pronounce.
The easy way
I'm going to cheat by giving the easy way first. This is cheating because it isn't correct. But to get it correct you need to see
accented letters, and non-Polish newspapers and the Internet often won't have those; so let's first lay out a very rough way of translating just A-Z into sounds.
First the vowels: easy, European-style, same as an Italian or Spanish. The stress is always on the second-last syllable. The letter y is always a vowel, as in 'happy': so call it the same as i.
Also familiar from other European languages is the use of j for the English Y consonant, and w for English V. The group ch is as in 'loch', or 'chutzpah', but if you can't easily do that, make it a K. But c by itself is English TS.
Next, those notorious Zs. There's no getting away from the fact that mezczyzna 'man', przyjezdzac 'arrive', szescdziesiat 'sixty', wszedzie 'everywhere', dziewczynka 'little girl', and all their little friends look a bit hair-raising. But the function of the Z in these digraphs is the same as that of H in English clusters: Polish SZ = English SH, and Polish CZ = CH.
We can begin simplifying now:
mezczyzna --> mezCHIZna
Warszawa --> varSHAva
Wojciech Jaruzelski --> VOYtsiek yaruZELski
Karol Wojtyla --> KARol voyTIla
Czestochowa --> chestoKOva
(The above are
not quite accurate: see below for better approximations once I've explained more.)
But seriously, folks
The Polish
alphabet has 32 letters,
a ą b c ć d e ę f g h i j k l ł m n ń o ó p r s ś t u w y z ź ż.
For reference, here are the correct symbols and their Unicode numbers (extracted from my Using Unicode in E2):
Ą Ą ą ą A-ogonek
Ć Ć ć ć C-acute
Ę Ę ę ę E-ogonek
Ł Ł ł ł L-slash
Ń Ń ń ń N-acute
Ó Ó ó ó O-acute
Ś Ś ś ś S-acute
Ź Ź ź ź Z-acute
Ż Ż ż ż Z-dot-above
ó is the same as u. The two hooked vowels are nasalized, and ą is a nasalized O. So the French words bon and bien could be written exactly as bą and bię in Polish.
But before a stop (p b t d c k g) they are like vowel + n (m): so głębokie 'deep' is as if głembokie, and dokąd 'where to?' is as if dokond.
ż is "ZH" as in vision, regime, and rz has that sound too (e.g. Jerzy --> YEzhi). ć ś ź are a CH SH ZH but softer, more hissy, than cz sz rz, which are close to the English sounds. The acute-accented ones are technically alveolo-palatal. And ń is a corresponding NY sound. When these softer sounds are followed by a vowel, the letter i is used instead of the acute: so cia sie zio are pronounced ća śe źo. The i is not a syllable and doesn't count for the stress rule:
sześćdziesiat --> sheshchDZHEshat
Wojciech --> VOYchekh
The difference between the vowels i and y is roughly that of English machine and gypsy. Moreover, i causes the above softening: so ci is pronounced ći but in cy it isn't changed.
Ł is like the English consonant W. (For some Polish speakers it may be a peculiarly Polish sound, close to an English W, but coming from the very back-of-the-throat L sound we have in tall.) In print outside Poland the slashed lower-case L is sometimes mistaken for a T. So this explains:
Lech Wałęsa --> LEKH vaWE(ng)sa
Karol Wojtyła --> KARol voyTIwa
Ułam --> Uwam
złoty --> ZWOTi
The voiced consonants, or more precisely the voiced obstruents, are b d g w z ź ż and at the end of a word they become the corresponding voiceless sound (the same rule of devoicing as in German): p t k f s ś sz
Witold --> VItolt
Łódź --> WUCH
Elbląg --> ELblonk
Kraków --> KRAkuf
The same devoicing rule generally applies when two consonants come together, which makes everyone happy. Voiced + voiceless becomes voiceless + voiceless, e.g.
Minkowski --> minKOFski
mężczyzna --> me(ng)sCHIZna
I think I'll give up at this point, because the
assimilation rules aren't as straightforward as that, but what I've given so far is enough for an approximation.