How to Crack the Code of Swedish Phonology

Written by Dyami Millarson

I wrote a post recently about unlocking the secrets of Swedish, but I decided to treat the topic of Swedish phonology separately, because I have often studied the pronunciation of languages entirely separately. It requires no elaborate argumentation that the sounds of Swedish are the building blocks of the (spoken) Swedish language. Swedish pronunciation has been one of the hardest for me to learn. I have acquired Swedish phonology by listening to the LibriVox recordings of Lars Rolander, selecting the Swedish pronunciation of a few words on Forvo that I would repeat regularly, and imitating the Swedish pronunciation of the app Swedish Fun Easy Learn. Transcribing Swedish phonology as accurately as possible using the International Phonetic Alphabet (IPA) has also helped me a great deal. Whilst learning the pronunciation of a language, I recommend finding recordings of one or a few people whose pronunciation is quite similar and to stick to this selection until one masters the phonology fully.

Swedish is somehat unusual in that it has a large vowel inventory of 17 phonemic vowels, which means the number is much higher if you also include allophonic vowels. The distinction between a phoneme and allophone is that the former matters for distinguishing meaning of words, whilst an allophone has no semantic meaning. For instance, the English sounds /ʊ/ and /æ/ are phonemes because foot /fʊt/ and fat /fæt/ have different meanings. However, the Australian English sound /əʉ/ as in foe /fəʉ/, the British English sound /əʊ/ as in though /ðəʊ/, and the American English sound /oʊ/ as in row /ɹoʊ/ are allophones of the same diphthong in the English language.

The most difficult consonants in Swedish include sj-ljudet (variously written as skj, sj, sk, ch, stj), the pronunciation of tj (k is pronounced like tj before ä, e, i), and the pronunciation of rs, rt, rd and rn. Mandarin Chinese phonology has helped me to learn how to pronounce tj and rs correctly, which are respectively written as x and sh in pinyin. I could compare the Swedish tj and rs to sounds in other languages such as Mandarin Chinese, but sj-ljudet is unique to Swedish, which made it much harder for me to learn. My trick was to select a small set of words with sj-ljudet and to repeat them out loud to let the peculiar Swedish pronunciation slowly sink in. I found that the IPA transcription of sj-ljudet as /ɧ/ looked intriguing but did not quite help me to master the sj-ljudet in Swedish. I discovered over time that sj-ljudet, which is generally transcribed as /ɧ/, has various pronunciations, such as [fˣʷ], [xʷ] and [ʃʰʷ].

It surprised me to learn that I was actually hearing sounds that I was intimately familiar with, but they had been obscured by various co-articulations. I have not been very familiar with co-articulated consonants, except aspirated sounds from English and German and labialised sounds from Latin (such as qu and gu) and Gothic (such as kw and ƕ). I had been unable to hear sj-ljudet properly because I had hitherto been unfamiliar with double co-articulation, which I was witnessing in some realisations of sj-ljudet in Swedish. However, when I was reading about Proto-Indo-European phonology, I learned about doubly co-articulated consonants, which I actually witnessed for the first time in a living language when I observed them in Swedish. For the realisation of sj-ljudet, I eventually settled down on choosing to adopt the doubly co-articulated [fˣʷ] in my own Swedish and thus I will transcribe sj-ljuset as such in this article.

For cracking the code of Swedish phonology, it might be a useful idea to compile a list of words that have difficult, surprising or otherwise interesting phonetic properties: stiltje [ˈstɪ(ˑ)l.cɛ] calm (i.e., a period without wind) is not pronounced [ˈstɪl.fˣʷɛ], religiös [ˌrɛ.liˈfˣʷœ(ˑ)œ̯̆s] religious is not pronounced [ˌrɛ.li.giˈjœːs], skina [ˈfˣʷi(ˑ)ĭ̯nä] is not pronounced [skiːna], skjuta [ˈfˣʷʉ(ˑ)ʉ̯̆tä] shoot is not pronounced [ˈskjuː.tä], stjärna [ˈfˣʷæ(ˑ)æ̯̆.ɳä] star is not pronounced [ˈstjɛr.nä], norrsken [ˌno̞rːˈfˣʷɪ(ˑ)ə̯̆n] aurea borealisis not pronounced [ˈno̞rː.skɛn], kanske [ˈkän.fˣʷɛ] perhaps is not pronounced [ˈkän.skə], ut [ʉ(ˑ)ʉ̯̆t] (open stem syllable) out is pronounced [uːt], upp [ʊpː] (closed stem syllable) up is not pronounced [ʉpː], förste [ˈfœʂ.tɛ] first is not pronounced [ˈfœr.stə], värld [ˈvæ(ˑæ̯̆)ɭɖ] world is not pronounced [ˈvɛrld], kärlekshistoria [ˈɕæ(ˑ)æ̯̆ˌɭɛk.sɪsˈtu(ˑ)ŭ̯.riˌjä] love story is not pronounced (ˈkɛrˌlɛks.hiˈstoː.riˌjä], värd [ˈvæ(ˑ)æ̯̆ɖ] worth is not pronounced [ˈvɛrd]. While studying Swedish phonology, I memorised the tongue-twister sju sjösjuka sjömän [fˣʷʉʉ̯̆ ˈfˣʷœœ̯̆ˌfˣʷʉʉ̯̆kä ˈfˣʷœœ̯̆ˌmɛn] seven seasick seamen to learn the sj-ljudet (the [fˣʷ]-sound), while I memorised egen härd är guld värd [ˈɪˑə̯̆gɛn ˈhæˑæ̯̆ɖ ˈæˑæ̯̆ ˈgʊld ˈvæˑæ̯̆ɖ] (one’s) own hearth is worth gold (i.e., one’s own home is the best place in the world) to practise the retroflex d. One ought to observe that many consonant clusters have been simplified in Swedish whereas these clusters still exist in the Frisian languages, Dutch and German; namely, many Swedish clusters with s (for instance, sj) are reduced to sj-ljudet and many Swedish clusters with r (for instance, rd) are reduced to retrofex sounds.

Swedish prosody is different from that of the Frisian languages, Dutch, German and English in that Swedish uses tones like Chinese. However, these tones are generally non-phonemic. Therefore, Swedish is not a tone language like Mandarin or Cantonese, but it has a pitch accent like Ancient Greek. Syllables in Ancient Greek would either receive a rising or falling tone. This has been well-described in the book Vox Graeca, which I read when I was 15 years old. I was at the time writing a manuscript for a book about learning Ancient Greek like a living language. This book of mine has remained hitherto unpublished, but writing it has inspired me to save endangered languages. Writing this book about learning Ancient Greek like a living language has also taught me a lot about prosody. My knowledge of Ancient Greek prosody and Mandarin Chinese tones have helped me to perceive the pitch accent patterns in Swedish. I think the best way to learn Swedish prosody is to pick a few words and to try to reproduce the Swedish accent as accurately as possible. By repeating a few words regularly, I managed to learn the Swedish tones. A pitch accent is essentially a more strict form of intonation. When I read about the Mandarin Chinese tones, I learned that we use tones in English as well, but we know it as intonation and we use it more freely than Swedish and Chinese where the tones are fixed. I realised that Dutch people may pronounce the word ja (yes) with different tones and it slightly changes the meaning. In conclusion, we may already be familiar with tones, but how they are used in Swedish or Mandarin Chinese may not yet be familiar to us, and therefore we need to familiarise ourselves with these tones. It is essentially an issue of awareness. The best way to achieve tone-awareness is simply through imitation of words and phrases. Eventually Swedish tones just started making sense to me because I had heard it so many times and in so many words that I could not imagine Swedish being pronounced otherwise.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s