The phonology of Danish is similar to that of the other Scandinavian languages such as Swedish and Norwegian, but it also has distinct features setting it apart from the phonologies of its most closely related languages. For example Danish has a suprasegmental feature known as stød which is a kind of laryngeal phonation that is used phonemically. It also exhibits extensive lenition of plosives, which is noticeably more common than in the neighboring languages. Because of that and a few other things, spoken Danish is rather hard to understand for Norwegians and Swedes, although they can easily read it.
/n, t, d, l/ have been variously described as apical alveolar [n̺, t̺ˢʰ, d̺˚, l̺] and laminal denti-alveolar [n̪, t̪ˢʰ, d̪˚, l̪].
Intervocalic /d/ between two unstressed vowels may be realized as flap [ɾ].
/p, t, k/ are aspirated (and, in the case of /t/, also strongly affricated)voicelesslenis in syllable onset: [b̥ʰ, d̥ˢʰ, ɡ̊ʰ] (hereafter transcribed as [pʰ, tˢ, kʰ] for simplicity). Aspiration is lost in syllable coda.
For simplicity, the aspirated and affricated allophone of /t/ is often transcribed as [d̥ˢ]/[tˢ], i.e. as if it were just affricated.
In some varieties of standard Danish (but not the Copenhagen dialect), /t/ is just aspirated, without the affrication.
/b, d, ɡ/ are unaspirated voiceless lenis in syllable onset: [b̥, d̥, ɡ̊]. In syllable coda /d, ɡ/ and sometimes /b/ are opened: [ʊ̯ ð̞ˠ̠ ɪ̯/ʊ̯]. /ɡ/ becomes [ɪ̯] after front vowels and [ʊ̯] after back vowels.
Final /b, d, ɡ/ may be realized as [pʰ, tˢʰ, kʰ], in particular in distinct speech. In case of the alveolar plosive, in this position it may be either aspirated and affricated [tˢʰ] or just aspirated [tʰ].
The exact place of articulation of /k, ɡ/ varies; it is more front (pre-velar) [k̟ʰ, ɡ̊˖] before front vowels, and more back (post-velar) [k̠ʰ, ɡ̊˗] before back vowels. Bornholmsk dialect features even stronger fronting of /k, ɡ/ before front vowels, i.e. to palatal [c, ɟ].
/s/ is an apical alveolar non-retracted sibilant [s̺], but some speakers realize it as dental [s̪]. It is always voiceless.
/h/ is only weakly fricated. Between vowels, it is often voiced [ɦ].
[ɕ] occurs only after /s/ or /t/. Since [j] doesn't occur after these phonemes, [ɕ] can be analyzed as /j/, which is devoiced after voiceless alveolar frication. This makes it unnecessary to postulate a /ɕ/-phoneme in Danish.
Among voiced continuants, the lateral /l/ is an approximant, whereas /v, j, r/ and [ð] vary between being fricatives and approximants:
/v/ is either a voiced fricative [v] or, most often, a voiced approximant [ʋ] which, according to Nina Grønnum, is more accurately described as a short voiced labiodental plosive [b̪̆].
British phonetician John C. Wells commented on his blog about the quality of Danish [ð] that to him, it sounds "awfully like a lateral". A similar comment was made by Haberland (1994), who said that Danish [ð] is frequently mistaken for an [l] by second-language learners.
An acoustically similar sound (but apical rather than laminal) has been reported to occur as an intervocalic allophone of /d̠/ in the Dahalo language spoken in Kenya.
/j/ is an approximant, but when it occurs word-finally after /l/, it is articulated more strongly than usual, sometimes even as a fricative [ʝ].
An additional voiced continuant, namely the voiced velar fricative [ɣ] occurred in older Standard Danish. Some older speakers still use it in high register, but most often as an approximant [ɣ˕]. It corresponds to three sounds in contemporary Standard Danish:
[ʊ̯] (phonemically /v/) after back vowels and /r/;
According to Torp (2001), it occurs in some varieties of Jutlandic dialect, and only for some speakers (mostly the elderly). The alveolar realization is considered non-standard, even in classical opera singing - it is probably the only European language in which this is the case.
According to Basbøll (2005), it occurs (or used to occur until recently) in very old forms of certain conservative dialects in Northern Jutland and Bornholm.
/l, j, r/ are voiceless [l̥, ʝ̊ ~ ɕ, ʁ̥] after aspirated /p, t, k/, where the aspiration is realized as devoicing of the following sonorant. Note, however, that the sequence /tj/ is normally realized as a voiceless alveolo-palatal affricate [t͡ɕ].
The Danish allophones can be analyzed into 15 distinctive consonant phonemes, /p t k b d ɡ m n f s h v j r l/, where /p t k d ɡ v j r/ have different pronunciation in syllable onset vs. syllable coda.
Instances of [ŋ] can be analyzed as /n/ as it only occurs before /ɡ/ or /k/ and does not contrast with [n]. This makes it unnecessary to postulate an /ŋ/-phoneme in Danish (assuming that a following /ɡ/ is sometimes deleted).
Monophthongs of Modern Standard Danish, from Grønnum (1998:100). Unstressed [ɪ, ʊ, ə, ɐ] are not shown. The distinction between [œ̞(ː)] and [ɶ(ː)] is optional and not made by many speakers.
Modern Standard Danish has around 20 different vowel qualities. These vowels are shown here in a narrow transcription. In the rest of the article and in IPA transcriptions of Danish in Wikipedia the diacritics are usually omitted.
The following vowels are allophones. Phonemes are discussed below.
Stressed close vowels
[i] is close front unrounded [i]. John Wells's impression is that it is slightly centralized [ï].
[œ̞] is open-mid near-front rounded [œ̠].Basbøll (2005) transcribes it with the symbol ⟨ɶ⟩, and writes that "Nina Grønnum uses two different symbols for the vowels in these and similar words: gøre she transcribes with [œ̞] (semi-narrow transcription) and [œ] (narrow transcription), and grøn she transcribes with [ɶ] (semi-narrow transcription) and [ɶ̝] (narrow transcription). Clearly, there is variation within Standard Danish on this point (...)." Elsewhere in this article, [œ̞, œ̞ː] are not distinguished from [ɶ, ɶː] and the former symbols are not used.
Basbøll (2005) states that many Standard Copenhagen speakers of his generation generally pronounce the diphthong [ʌʊ̯] ([ʌ̟͗˕ʊ̯] in narrow IPA) as [ɒʊ̯], and that it is the main variant among younger speakers of Standard Copenhagen.
[ɒ] has been variously described as open-mid back rounded [ɔ] and near-open back rounded [ɒ̝].
[ɪ] is a lax, relatively close unrounded neutral front vowel. It is an assimilatory variant of [ɪ̯ə].
[ʊ] is a lax, relatively close rounded neutral back vowel, which may be realized the same as short /o/. It is an assimilatory variant of [ʊ̯ə].
[ə] is a mid central vowel with variable rounding ([ə̜ ~ ə̹]). For some speakers, it may be more consistently realized as rounded, albeit only in very distinct speech. In rapid speech, postvocalic [ə] tends to have the same quality as the preceding vowel, as in e.g. stue[ˈsd̥uːu] 'living room' or pige[ˈpʰiːi] 'girl'.
[ɐ] may be any of the following: near-open central unrounded [ɐ], retracted mid central unrounded [ə̠], or simply the same as stressed [ʌ] (a near-open near-back somewhat rounded vowel [ʌ̟͗˕]), which is probably the usual pronunciation.Grønnum (1998) transcribes both [ʌ] and [ɐ] as [ʌ].
[ɪ̯] is a non-syllabic, lax, relatively close unrounded neutral front vowel.Grønnum (1998) transcribes it the same as [j].
[ʊ̯] is a non-syllabic, lax, relatively close rounded neutral back vowel.Grønnum (1998) transcribes it as [w].
[ɐ̯] is a non-syllabic, central retracted neutral vowel (pharyngeal glide). Essentially, it is a non-syllabic equivalent of [ʌ].Grønnum (1998) transcribes it as [ʌ̯].
[ə] and [ɐ] occur only in unstressed syllables. With the exception of [a], [ʌ], [ə] and [ɐ] all vowels may be either long and short. Long vowels may have stød, thus making it possible to distinguish 30 different vowels in stressed syllables. However, vowel length and stød are most likely features of the syllable rather than features of the vowel.
These allophones can be analyzed into 11 distinctive vowels, where allophonic alternation mainly depends on whether the vowel occurs before or after /r/. The vowel /ə/ only occurs in unstressed syllables. All other phonemes may occur both stressed and unstressed.
The three way distinction in front rounded vowels /y ø œ/ is upheld only before nasals, e.g. /syns sønˀs sœns/synes, synds, søns ('seems', 'sin's', 'son's'). Furthermore, there are only three words where /y/ occurs before a nasal in a stressed syllable: synes, brynje, hymne ('seems, armor, hymn').[not in citation given]
[a] and [ɑ] are largely in complementary distribution. However, a two-phoneme interpretation can be justified with reference to the unexpected vowel quality in words like andre[ˈɑndʁɐ] 'others' or anderledes[ˈɑnɐˌleːð̩s] 'different', and an increasing number of loanwords.
The vowel system is unstable, and according to at least one study, the contemporary spoken language might be experiencing a merger of several of these phonemes. The following vowel pairs may be merged by some speakers:
Unlike the neighboring Mainland Scandinavian languages Swedish and Norwegian, the prosody of Danish does not have phonemic pitch. Stress is phonemic and distinguishes words like billigst[ˈb̥ilisd̥] ('cheapest') and bilist[b̥iˈlisd̥] ('car driver'). Verbs lose their stress (and stød, if any) with an object without a definite or indefinite article: e.g. ˈJens ˈspiser et ˈbrød[ˈjɛns ˈsb̥iːˀsɐ ed̥ ˈb̥ʁœðˀ] ('Jens eats a loaf') ~ ˈJens spiser ˈbrød[ˈjɛns sb̥isɐ ˈb̥ʁœðˀ] ('Jens eats bread'). In names, only the surname is stressed, e.g. [johanə luiːsə ˈhɑɪ̯b̥æɐ̯ˀ]Johanne Luise Heiberg.
In a number of words with stress on the final syllable, long vowels and sonorants may exhibit a prosodic feature called stød ('thrust'). Acoustically, vowels with stød tend to be a little shorter and feature creaky voice. Historically, this feature operated as a redundant aspect of stress on monosyllabic words that had either a long vowel or final voiced consonant. Since the creation of new monosyllabic words, this association with monosyllables is no longer as strong. Some other tendencies include:
Polysyllabic words with the nominal definite suffix -et may exhibit stød
Polysyllabic loanwords with final stress on either a long vowel or a vowel with a final sonorant typically feature stød
Diphthongs with an underlying long vowel always have stød. These are [eɪ̯ˀ, ɛɪ̯ˀ, æɪ̯ˀ, øɪ̯ˀ, iʊ̯ˀ, eʊ̯ˀ, ɛʊ̯ˀ, æʊ̯ˀ, yʊ̯ˀ, øʊ̯ˀ, œʊ̯ˀ, oʊ̯ˀ, ɔʊ̯ˀ, iɐ̯ˀ, eɐ̯ˀ, æɐ̯ˀ, yɐ̯ˀ, øɐ̯ˀ, ɶɐ̯ˀ, uɐ̯ˀ, oɐ̯ˀ]. Out of these, all but [eɪ̯ˀ, ɛɪ̯ˀ, æɪ̯ˀ, øɪ̯ˀ, æʊ̯ˀ, oʊ̯ˀ, ɔʊ̯ˀ] have a corresponding stødless variant, i.e. with an underlying short vowel. Conversely, there are diphthongs that appear only without stød, which are [ɑɪ̯, ʌɪ̯, uɪ̯, ɑʊ̯, ɒʊ̯]. This means that neither [ɑ] nor [ʌ] can start a diphthong with stød (in case of the latter vowel it is because it is inherently short), whereas [ɔ] cannot start a diphthong without stød. All of the diphthongs ending with [ɐ̯] appear both with and without stød.
Bauer, Laurie; Dienhart, John M.; Hartvigson, Hans H.; Jakobsen, Leif Kvistgaard (1980), American English Pronunciation: Supplement, Comparison with Danish., Copenhagen: Gyldendalske Boghandel, OCLC54869978
Grønnum, Nina (2003), "Why are the Danes so hard to understand?", in Jacobsen, Henrik Galberg; Bleses, Dorthe; Madsen, Thomas O.; Thomsen, Pia (eds.), Take Danish - for instance: linguistic studies in honour of Hans Basbøll, presented on the occasion of his 60th birthday, Odense: Syddansk Universitetsforlag, pp. 119–130
Torp, Arne (2001), "Retroflex consonants and dorsal /r/: mutually excluding innovations? On the diffusion of dorsal /r/ in Scandinavian", in van de Velde, Hans; van Hout, Roeland (eds.), 'r-atics, Brussels: Etudes & Travaux, pp. 75–90, ISSN0777-3692
Uldall, Hans Jørgen (1933), A Danish Phonetic Reader, The London phonetic readers, London: University of London Press
Fischer-Jørgensen, Eli (1972), "Formant Frequencies of Long and Short Danish Vowels", in Scherabon Firchow, Evelyn (ed.), Studies for Einar Haugen, The Hague: Mouton Publishers, pp. 189–200, ASINB0037F3D1S