Voiced sonorants in decline but hard pronunciation remains strong

Unlike what is sometimes heard in public discourse, languages are not static but constantly evolving and changing. Language changes can occur, for example, due to societal trends and demands, the tendency to simplify linguistic systems or the influence of other languages. These changes also extend to how we pronounce words. Currently, Ásgrímur Angantýsson, a professor of Icelandic linguistics, is leading an exciting study with a large team of researchers on how regional pronunciation nuances in Icelandic have fared over recent decades. 

“The research focuses mainly on the status and development of regional pronunciation features in Iceland and attitudes toward them. However, we are also keeping an eye out for signs of novelties in pronunciation,” explains Ásgrímur, who is co-leading the project with Finnur Friðriksson, associate professor at the University of Akureyri. 

Mapping the current status

Well-known regional features include hard pronunciation and voiced consonants in the North; the Westfjords monophthongal pronunciation; southern hv-pronunciation, and the Skaftafell monophthongal pronunciation. Pronunciation variations more commonly associated with younger generations include ks-pronunciation, affricate  articulation, and the so-called glottal stop pronunciation. Further explanations of these pronunciation features can be found later in this article.

What are the main regional pronunciation features in Iceland?

(1) Northern Icelandic Pronunciation Features
a. Hard Pronunciation (Harðmæli)

• This occurs when /p, t, k/ are pronounced with post-aspiration after a long vowel: tapa [tʰa:pʰa], líka [li:kʰa], bíta [pi:tʰa].

• In the more common "soft pronunciation" (linmæli), these sounds have no such post-aspiration: tapa [tʰa:pa], líka [li:ka], bíta [pi:ta].

b. Voiced Pronunciation (Raddaður framburður)

• This occurs when the phonemes /m, n, l, ð/ are voiced before aspirated stops: hempa [hɛmpʰa], mennta [mɛntʰa], hjálpa [çaulpʰa], blaðka [plaðkʰa].
• In the more common "unvoiced pronunciation" (óraddaður framburður), the sounds are unvoiced: hempa [hɛm̥pa], mennta [mɛn̥ta], hjálpa [çaul̥pa], blaðka [plaθka].

(2) Pronunciation Features in South and Southeast Iceland
a. Hv-Pronunciation (hv-framburður)

• Found in words like hvalur, pronounced [xa:lʏr̥], [xʷaːlʏr̥], or [xvaːlʏr̥], where there is an unvoiced velar fricative [x] at the beginning instead of the more common kv-pronunciation: hvalur [kʰva:lʏr̥].

b. Skaftafell Monophthongal Pronunciation (Skaftfellskur einhljóðaframburður)

• Long monophthongs are pronounced before [jɪ] in words like bogi [pɔ:jɪ] and magi [ma:jɪ], instead of the diphthongs typical in standard pronunciation: bogi [pɔijɪ], magi [maijɪ].

(3) Westfjords Pronunciation Feature
Westfjords Monophthongal Pronunciation (Vestfirskur einhljóðaframburður)

• Monophthongs are pronounced in words like banki [paɲ̥cɪ], rather than diphthongs as in the pronunciation of most Icelanders: banki [pauɲ̥cɪ]. (“Icelandic regional pronunciation, attitudes and real-time change: Latest developments.” (Finnur Friðriksson, Ásgrímur Angantýsson and Stefanie Bade) Sociolinguistica 2024.)

Newer Pronunciation Variations

(1) Ks-Pronunciation (ks-framburður)

• Stops replace fricatives in clusters like ks in words such as buxur [pYksYr] and hugsa [hYksa], instead of the fricative pronunciation: buxur [pYxsYr], hugsa [hYxsa].

(2) Affricate Pronunciation (Tvinnhljóðsframburður)

• Initial sounds in words like tjald and djarfur become affricates, pronounced closer to tsjald and dsjarfur, in contrast to the usual pronunciation without this added s-like sound.

(3) Glottal Stop Pronunciation (Höggmæli)

• A glottal stop replaces oral stops in words like Bjarni [pjaʔnɪ], instead of the typical pronunciation with rdn-, dn-, or even -rn, this last one a very rare pronunciation feature.

Data on peoples’ pronunciation changes over an 80-year period

One of the motivations for the project, according to Ásgrímur, is that attitudes toward different pronunciations have not been gathered systematically. “In our research, we consider to what extent attitudes may contribute to the persistence or decline of various pronunciation features. In many cases, we can compare the pronunciation of the same individuals at different times, up to 80 years ago, and study whether, and to what extent, attitudes influence pronunciation development," he explains.

According to Ásgrímur, the data was collected in two main phases. “The first phase involved an online survey managed by the University of Iceland’s Social Science Research Institute, which is now complete. 3,000 potential speakers from across the country were invited to participate in the survey, which aimed to gain a comprehensive overview of the status of regional pronunciation variations in Iceland.

Three equally sized groups targeted for participation 

"The first consisted of 1,000 teenagers (ages 12–20) chosen randomly, the second of 1,000 participants who had previously taken part in the RÍN study conducted by Kristján Árnason and Höskuldur Þráinsson in the 1980s, and, in some cases, in Björn Guðfinnsson's study from the 1940s. The third group was a random sample of 1,000 individuals from the National Registry, aged 12 and older," explains Ásgrímur. Notably, more than 20 participants in their nineties who had been part of Björn's research also participated in Ásgrímur and his team's study, providing data spanning 80 years.

Studying attitudes toward regional pronunciation features for the first time
In the online survey, all participants were asked to take part in a pronunciation test designed to measure how much they used specific variations—both previously documented and newer ones. A total of 960 usable responses were collected, yielding a response rate of 32%. "It’s important to remember that the survey was relatively demanding and time-consuming, as participants not only had to answer questions but also listen to texts and record their own reading. We expected a relatively low response rate but are very pleased with the participation nonetheless. We now have an immense amount of interesting data," Ásgrímur adds.

The second main phase of data collection focused on examining how attitudes might influence the development of selected regional pronunciation variations.  "We conducted 160 interviews with a sample of speakers from the first phase, allowing us to delve deeper into participants’ attitudes and perceptions of their own pronunciation as well as that of others. By analysing these interviews, we can assess how attitudes might have contributed to speakers either preserving their regional pronunciation features or dropping them," Ásgrímur explains.

Students active participants in the study

Although the subject is the Icelandic language, the project is international. A large group is involved, as there is an enormous amount of data to process.  As mentioned earlier, Ásgrímur and Finnur lead the project, but their co-applicants for the project grant were Nicole Dehé, professor at the University of Konstanz, Gunnar Ólafur Hansson, professor at the University of British Columbia, and Kristín Margrét Jóhannsdóttir, associate professor at the University of Akureyri.  Stefanie Bade is also a postdoctoral researcher on the project. The research project is supported by The Icelandic Research Fund (2023−2025).

Students at all university levels have contributed significantly.  "Ása Bergný Tómasdóttir and Eva Hrund Sigurjónsdóttir have been responsible for the implementation of certain components in the study and they both wrote their MA theses based on the research data.  The master's students Ása Jónsdóttir, Halla Hauksdóttir, and Ragnhildur Ósk Sævarsdóttir have also worked on data collection and processing.  Others involved in the research include Dagbjört Guðmundsdóttir, a PhD student in Icelandic linguistics, Salome Lilja Sigurðardóttir, MA in general linguistics, Auður Siemsen and Magnús Már Magnússon, BA in Icelandic, and Kolbeinn Héðinn Friðriksson, BA student." 

In addition, Höskuldur Þráinsson and Kristján Árnason, professors emeritus at the University of Iceland, Eva Sundgren, professor emeritus at Mälardalen University, Bragi Guðmundsson, professor at the University of Akureyri, and Laurel MacKenzie, associate professor at New York University, are advisors to the project along with Margrét Guðmundsdóttir, PhD in Icelandic linguistics.

Tsjald and dsjarfur among recent pronunciation features

"The next steps are to complete the processing of interviews with the 160 individuals selected from the online survey to delve deeper into various aspects of the research; especially regarding how individuals change their language over their lifetime. This analysis will answer many questions, such as how the development of regional pronunciation features over a lifetime relates to the speaker’s origin and identity.  Furthermore, it should provide better insight into popular perceptions of the distribution of these features in core areas, compared to their actual distribution.  We will also look more closely at what meaning language users attribute to the concept of clarity and how it relates to pronunciation," says Ásgrímur.

The data from the research not only contain information on regional pronunciation features but also characteristics or variations in language related to age and social factors that are considered recent in the language.  These include the so-called ks-pronunciation, where words like buxur 'pants' and hugsa 'think' are pronounced with a stop instead of a velar fricative before the s-sound s in the middle of the word, and the diphthong pronunciation, where words like tjald 'tent' and djarfur 'brave' are pronounced closer to 'tsjald' and 'dsjarfur'. "By analysing this data, we hope to get a comprehensive picture of the status of pronunciation variations in Iceland and their social context," says Ásgrímur. 

Asked about the significance of the research for science and society, Ásgrímur points out that it provides new knowledge that contributes to the coherence of Icelandic pronunciation research, which is in a way unique globally, especially regarding the development of individual language over a very long period. "The research examines the role of conscious and unconscious attitudes towards language in explaining such language changes in real-time.  This is possible because of how well phonetic variability has been studied in the past," says Ásgrímur. 

He adds that the research results have already attracted the attention and interest of linguists and the public, both in Iceland and the international linguistic community, through lectures, interviews, and published material. "New knowledge in this field can also be used in practical contexts, such as in teaching materials about pronunciation variation," concludes Ásgrímur.