The close vowels are variably transcribed with the symbols ⟨i, iː, u, uː⟩ (without diacritics) and ⟨i̞, i̞ː, u̞, u̞ː⟩ (with a lowering diacritic). This article uses the former set for simplicity.
The mid vowels are variably transcribed with the symbols ⟨e, eː, o, oː⟩, ⟨ɛ, ɛː, ɔ, ɔː⟩ and, most rarely, with ⟨e̞, e̞ː, ɔ̝, ɔ̝ː⟩. This article uses the first set for simplicity.
The open central vowels are most often transcribed with the symbols ⟨a, aː⟩, but Mistrík (1988) transcribes them with ⟨ɑ, ɑː⟩, whereas Pavlík (2004) transcribes them as ⟨ɐ̞, ɐ̞ː⟩. This article uses the first set for simplicity.
The close vowels /i, iː, u, uː/ vary in their height between close [i, iː, u, uː] and near-close [ɪ, ɪː, ʊ, ʊː], with the near-close realization being the most usual. The backness of /i, iː/ varies between fully front [i ~ ɪ̟, iː ~ ɪ̟ː] and near-front [i̠ ~ ɪ, i̠ː ~ ɪː], just as the backness of /u, uː/ varies between fully back [u ~ ʊ̠, uː ~ ʊ̠ː] and near-back [u̟ ~ ʊ, u̟ː ~ ʊː].
There is not a full agreement about the quality of the mid vowels /e, eː, o, oː/:
Long /eː, oː/ are realized as close-mid [eː, oː] in southern dialects spoken near the river Ipeľ, as well as - under Hungarian influence - in some other speakers. He does not report the occurrence of the close-mid realizations [e, o] of the short /e, o/.
Open-mid realizations [ɛ, ɛː, ɔ, ɔː] are rare.
/e, eː/ vary in their height between close-mid [e, eː] and true-mid [e̞, e̞ː], with the true-mid realization being the most usual. The backness of these vowels varies between fully front [e ~ e̞, eː ~ e̞ː] and near-front [e̠ ~ e̠˕, e̠ː ~ e̠˕ː]. The open-mid realizations [ɛ, ɛː] do not occur in standard Slovak.
/o, oː/ vary between true-mid [o̞, o̞ː] and open-mid [ɔ, ɔː], with the true-mid realization being the most usual. The backness of these vowels varies between fully back [o̞ ~ ɔ, o̞ ~ ɔ] and near-back [o̟˕ ~ ɔ̟, o̟˕ː ~ ɔ̟ː]. The close-mid realizations [o, oː] do not occur in standard Slovak.
Speakers often fail at attempts to pronounce /æ/, pronouncing a vowel that is phonetically too close to either /e/ or /a/.
There is not a complete agreement about the quality of the vowels transcribed /a, aː/:
According to Kráľ (1988), they are either open central [ä, äː] or somewhat more back [ɑ̟, ɑ̟ː], with both variants being equally correct in standard Slovak. He also states that under Hungarian influence, some speakers pronounce the short /a/ as rounded [ɒ].
According to Pavlík (2004), these vary in quality between near-open central [ɐ, ɐː], open retracted front [a̠, a̠ː], open central [ä, äː] and open advanced back [ɑ̟, ɑ̟ː], with the open central realizations being the most usual.
Vowel length is not phonemic in Eastern dialects, which have only five vowel phonemes (/i, u, e, o, a/, some speakers also have /æ/).
In Western dialects, vowels that are short due to the rhythmical law are often realized as long, and thus violating that law.
/y, yː, ø, øː, oː/ occur only in loanwords. Just as other mid vowels, /ø, øː/ are phonetically true-mid [ø̞, ø̞ː]. Among these vowels, only /oː/ is consistently realized in the correct manner, whereas the occurrence of the front rounded vowels /y, yː, ø, øː/ has been reported only by Kráľ (1988), who states that the front rounded vowels appear only in the high register and medium register. However, in the medium register, /y, yː/ and /ø, øː/ are often either too back, which results in realizations that are phonetically too close to, respectively, /u, uː/ and /o, oː/, or too weakly rounded, yielding vowels that are phonetically too close to, respectively, /i, iː/ and /e, eː/.
/uː, eː/ do not occur after soft consonants, where they are replaced by the corresponding diphthongs /i̯u, i̯e/. The same is generally true for /aː/ (/i̯a/ after soft consonants), but the sequence /jaː/ may occur in some cases.
Long /eː/ occurs only in loanwords, one native word (dcéra) and in adjective endings.
/æ/ can only be short, and occurs only after /m, p, b, v/. There is not a full agreement about its status in the standard language:
Kráľ (1988) states that the correct pronunciation of /æ/ is an important part of the high register, but in medium and low registers, /æ/ merges with /e/, or, in some cases, with /a/.
Short (2002) states that only about 5% of speakers have /æ/ as a distinct phoneme, and that even when it is used in formal contexts, it is most often a dialect feature.
These diphthongs are variably transcribed as ⟨i̯e, i̯a, i̯u, u̯o⟩, ⟨i̯ɛ, i̯a, i̯u, u̯ɔ⟩, ⟨ɪ̯ɛ, ɪ̯a, ɪ̯ʊ, ʊ̯ɔ⟩ and ⟨ĭ̞e̞, ĭ̞æ, ĭ̞ʊ, ŭ̞o̞⟩. For simplicity, this article uses the first set of symbols.
All of the diphthongs are rising, i.e. their second elements have more prominence.
The phonetic quality of Slovak diphthongs is as follows:
/i̯e/ has a starting point that, apart from the fact that it is non-syllabic, is practically the same as the /i/ monophthong (that is, close [i ~ i̠] or near-close [ɪ̟ ~ ɪ], with the near-close realization being the most usual), whereas its ending point is practically the same as the /e/ monophthong (that is, close-mid [e ~ e̠] or mid [e̞ ~ e̠˕], with the mid realization being the most usual).
/i̯a/ has a starting point that, apart from the fact that it is non-syllabic, is similar to the /i/ monophthong (that is, near-close [ɪ̟ ~ ɪ] or [e ~ e̠], with the near-close realization being the most usual), whereas its ending point is more similar to the /æ/ monophthong than the /a/ monophthong (that is, near-open near-front [æ̠] or open near-front [a̠]).
/i̯u/ has a starting point that, apart from the fact that it is non-syllabic, is practically the same as the /i/ monophthong (that is, close [i ~ i̠] or near-close [ɪ̟ ~ ɪ], with the near-close realization being the most usual), whereas its ending point is somewhat more front than the /u/ monophthong (that is, close near-back [u̟], near-close near-back [ʊ] or close-mid near-back [o̟], with the near-close realization being the most usual).
/u̯o/ has a starting point that, apart from the fact that it is non-syllabic, is practically the same as the /u/ monophthong (that is, close [u̟ ~ u] or near-close [ʊ ~ ʊ̠], with the near-close realization being the most usual), whereas its ending point is similar to the /o/ monophthong (that is, close-mid [o̟ ~ o] or mid [o̟˕ ~ o̞], with the mid realization being the most usual).
Pavlík (2004) describes an additional realization, namely a weakly palatalized apical alveolar approximant [l̺ʲ]. According to this scholar, the palatal realization [ʎ] is actually alveolo-palatal [ʎ̟].
The /ʎ–l/ contrast is neutralized before front vowels, where only /l/ occurs. This neutralization is taken further in western dialects, in which /ʎ/ merges with /l/ in all environments.
/j/ is an approximant, either palatal or alveolo-palatal. Between close front vowels, it can be realized as a fricative [ʝ], whereas between open central vowels, it can be a quite lax approximant [j˕].
Some additional notes includes the following (transcriptions in IPA unless otherwise stated):
/r, l/ can be syllabic: /r̩, l̩/. When they are long (indicated in the spelling with the acute accent: ŕ and ĺ), they are always syllabic, e.g. vlk (wolf), prst (finger), štvrť (quarter), krk (neck), bisyllabic vĺča—vĺ-ča (wolfling), vŕba—vŕ-ba (willow-tree), etc.
/m/ has the allophone [ɱ] in front of the labiodental fricatives /f/ and /v/.
/n/ in front of (post)alveolar fricatives has a postalveolar allophone [n̠].
/n/ can be [ŋ] in front of the velar plosives /k/ and /ɡ/.
In the standard language, the stress is always on the first syllable of a word (or on the preceding preposition, see below). This is not the case in certain dialects. Eastern dialects have penultimate stresses, which at times makes them difficult to understand for speakers of standard Slovak. Some of the north-central dialects have a weak stress on the first syllable, which becomes stronger and moves to the penultimate in certain cases. Monosyllabic conjunctions, monosyllabic short personal pronouns and auxiliary verb forms of the verb byť (to be) are usually unstressed.
Prepositions form a single prosodic unit with the following word, unless the word is long (four syllables or more) or the preposition stands at the beginning of a sentence.
Slovak linguists do not usually use IPA for phonetic transcription of their own language or others, but have their own system based on the Slovak alphabet. Many English language textbooks make use of this alternative transcription system. In the following table, pronunciation of each grapheme is given in this system as well as in the IPA.
Raz sa severák a slnko hádali, kto z nich je silnejší. V tom zbadali pocestného, ktorý kráčal zakrytý plášťom. Dohodli sa, že silnejší je ten, kto ako prvý dokáže, aby si pocestný vyzliekol plášť. A tak začal severák fúkať zo všetkých síl, ale čím viac fúkal, tým viac sa pocestný zakrýval plášťom. Nakoniec sa severák vzdal zbytočnej námahy. Potom slnko začalo svietiť a hriať. Za malú chvíľku bolo pocestnému teplo a vyzliekol si plášť. Severák musel nakoniec uznať, že slnko je silnejšie ako on.
Rubach, Jerzy (1995), "Representations and the organization of rules in Slavic phonology", in Goldsmith, John A., The handbook of phonological theory (1st ed.), Oxford: Blackwell, pp. 848–866, ISBN978-0631180623
Sabol, Ján (1961), "O výslovnosti spoluhlásky v", Slovenská reč, 6: 342–348