North Germanic tone accent – representation, distribution, origin and dialectology
Swedish and Norwegian intonation contains some tones that go together with words, rather than with overall phrases or utterances. These so-called tone accents (1 and 2) distinguish 1ketchup ‘id.’ from 2senap ‘mustard’. The melody of the utterance is identical in the two words, but one tone distinguishes the two words melodically. That is the lexical tone. Such tones are not found in the close relatives English and German. In Danish, a very close relative, we find a different (but related) system in most dialects, known as stød. There is a historical, but not fully understood, relationship between tone accent and stød.
I’m writing a book where I discuss most aspects relating to the tone accent, building on 25 years of research. How and when do lexical tones originate in the North Germanic languages. How does the system develop in the dialects? How are the accents related within the sound system? Do both accents involve lexical tones or is it just one of them that is like that? Another issue is accent in the vocabulary. Some endings correlate with accent 2 (1bil, 2bil-ar ’car/s’), while others don’t (1bil-en). An initial unstressed syllable may remove accent: 2del-ning ’split’ but för1del-ning ‘distribution’.
I discuss different theories regarding the tone accents and I make comparisons between the Swedish/Norwegian tone accent system, the Danish stød system, and a couple of other systems nearby (Central Franconian, Latvian, Lithuanian).
I’m writing a book where I discuss most aspects relating to the tone accent, building on 25 years of research. How and when do lexical tones originate in the North Germanic languages. How does the system develop in the dialects? How are the accents related within the sound system? Do both accents involve lexical tones or is it just one of them that is like that? Another issue is accent in the vocabulary. Some endings correlate with accent 2 (1bil, 2bil-ar ’car/s’), while others don’t (1bil-en). An initial unstressed syllable may remove accent: 2del-ning ’split’ but för1del-ning ‘distribution’.
I discuss different theories regarding the tone accents and I make comparisons between the Swedish/Norwegian tone accent system, the Danish stød system, and a couple of other systems nearby (Central Franconian, Latvian, Lithuanian).
Final report
North Germanic tonal accent - representation, distribution, origin and dialectology
The project aim is to produce a summarising and updated monograph on the tonal accent system in North Germanic. It is about the tonal distinction between accents 1 (ketchup ‘id.’) and 2 (senap ‘mustard’). This means that mainly Swedish and Norwegian varieties are treated. However, the Danish stød system is discussed, too, partly because it is clearly related to the tonal varieties, and partly because the stød system provides clues to dialect transitions between tonal varieties.
The monograph I am writing summarises 25 years of research. As always when you rethink a topic, new questions and insights emerge. I have worked now on chapters in the monograph, now on articles focusing on particular topics in the field of tonal accent. In addition, I have presented individual topics at conferences and seminars. This has worked well in bringing the relevant research questions to the fore, especially those I have not worked on for a long time,.
Main results and publications of the project
What I consider to be perhaps the most important general result within the main topic of the project is the power of the explanatory value of the concrete representation of the tone accent in the intonation structure. This may sound a bit cryptic, but simply refers to the fact that the concrete description of intonation and tonal accent (number of tones, which tones, where they attach, how they hold themselves to each other, and the like) has a great impact on which explanations and dialect typological predictions become possible. The concretion thus provides a greater depth of explanation. Some examples of this theme follow below. I also believe that I have improved reasoning on a number of sub-issues, as well as provided clearer criticism of models I believe are weaker than mine.
1. Dialects can be divided into two groups based on whether they show only accent 2 in compounds (e.g. Central Swedish, West Swedish) or both accent 1 and accent 2 in compounds (e.g. Scanian, East Norwegian). In my model, this difference is directly linked to a structural fact: whether the intonation curve attaches to two stressed syllables or only to one, in compounds. This means that there is a really simple, grammatical explanation for a striking difference on the surface that shows a large and clear geographical pattern. Other models often assume a more abstract relationship between the surface tonal structure and the underlying one, such as that polysyllabic words should be assigned accent 2, or that intonation and tonal accent are actually secondary to stress structure. These models offer no explanation neither for the pattern in compounds, nor the dialect geography it exhibits. Typological differences are discussed in the article The typology of tonal accent (Riad 2025) and in chapter 3.
2. The concrete approach also identifies accent 2 as the marked case, that is, the tonal category that needs explanation (it contains the lexical tone). This manifests itself in the fact that accent 1 is always a tonal subset of accent 2. Anyone who claims that accent 1 is the marked case faces a so-called abstraction problem. This is discussed in Gussenhoven & Riad (2025) and in chapters 1 and 2.
3. The structural factors are also important for the diachronic questions. Regarding the occurrence of the tonal accent, there is some confusion because researchers do not agree on *what* it is that should be explained. In my opinion, it is the presence of a tone in accent 2 (a lexical tone, i.e. a concrete phonological object) that should be explained. Others have focused on explaining the occurrence of a distinction in minimal pairs (and-en ‘the dock’/ande-n ‘the spirit’, steg-en ‘the steps’/stegen ‘the ladder’). In Riad (2023) and in chapter 4, I clarify the differences between these approaches, demonstrating how the appearance of a concrete tone logically precedes the appearance of a contrast, and that it is therefore crucial which part of the tonal structure one explains the emergence of.
4. the representation also points out what causes the difference between so-called two-peaked and one-peaked dialects. When the lexical tone is high, the dialect is two-peaked, when it is low, the dialect is one-peaked. This follows from a requirement for intonational accents to contain a high tone in North Germanic. The same generalisation underlies the distinction between dialects that have a prominence difference between small and big accents. This is discussed in Gussenhoven, Riad & Wetterlin (2025) and in chapter 2.
The concretely phonological is also reflected in the many phenomena involved in the development of stød from a tone-accent system (chapter 6). The concrete phonological (in contrast to other more abstract assumptions) is thematised in several chapters of the monograph.
Alongside the work on the tonal accent, I have also conducted research on reading during the sabbatical. Together with Maria Lim Falk, I have developed a new reading model with a linguistic basis that includes both first language readers and second language readers. The article has just been submitted to a journal (Riad & Lim Falk, 2024)
Results beyond publications
I have made some new contacts in Berlin, especially in the field of teacher training and teaching German as a second language, an area where Sweden is further ahead than Germany (mutatis mutandis). During the visit to the US, I received feedback on my research from colleagues at the universities we visited, especially the University of Illinois, the University of Texas, and California State University, Fresno, both in terms of tone accent and reading research. I was also introduced to a couple of undergraduate and graduate student groups and gave a couple of classes on my topic. My work in second language reading has led to an invitation to speak at the 10th Nordic Dyslexia Conference in Stockholm in August 2024.
New research questions
The work on the article on reading has generated ideas for appropriate intervention studies. This is particularly true in the area of phonological awareness and decoding. For example, it would be interesting to test whether an early focus on phoneme differentiation in newly arrived adolescents and adults, i.e. targeted work on creating mental categories for all Swedish phonemes, would lead to faster reading development (and thus a better school prognosis). A similar study could be done with younger second language speakers. Similarly, interventions could be made to train decoding speed to see if this would lead to better reading comprehension in the same groups of pupils.
How the results have been disseminated and if and how collaboration has taken place.
The book is at a fairly advanced stage (about two-thirds of it is written) and I expect to have a finished book manuscript in the spring term of 2025. Most of the articles I have written will be published next year. Some of these articles will be published with open access. I will post preprints on my profile page at SU (https://www.su.se/english/profiles/triad-1.183671?open-collapse-boxes=research-publications)
The project aim is to produce a summarising and updated monograph on the tonal accent system in North Germanic. It is about the tonal distinction between accents 1 (ketchup ‘id.’) and 2 (senap ‘mustard’). This means that mainly Swedish and Norwegian varieties are treated. However, the Danish stød system is discussed, too, partly because it is clearly related to the tonal varieties, and partly because the stød system provides clues to dialect transitions between tonal varieties.
The monograph I am writing summarises 25 years of research. As always when you rethink a topic, new questions and insights emerge. I have worked now on chapters in the monograph, now on articles focusing on particular topics in the field of tonal accent. In addition, I have presented individual topics at conferences and seminars. This has worked well in bringing the relevant research questions to the fore, especially those I have not worked on for a long time,.
Main results and publications of the project
What I consider to be perhaps the most important general result within the main topic of the project is the power of the explanatory value of the concrete representation of the tone accent in the intonation structure. This may sound a bit cryptic, but simply refers to the fact that the concrete description of intonation and tonal accent (number of tones, which tones, where they attach, how they hold themselves to each other, and the like) has a great impact on which explanations and dialect typological predictions become possible. The concretion thus provides a greater depth of explanation. Some examples of this theme follow below. I also believe that I have improved reasoning on a number of sub-issues, as well as provided clearer criticism of models I believe are weaker than mine.
1. Dialects can be divided into two groups based on whether they show only accent 2 in compounds (e.g. Central Swedish, West Swedish) or both accent 1 and accent 2 in compounds (e.g. Scanian, East Norwegian). In my model, this difference is directly linked to a structural fact: whether the intonation curve attaches to two stressed syllables or only to one, in compounds. This means that there is a really simple, grammatical explanation for a striking difference on the surface that shows a large and clear geographical pattern. Other models often assume a more abstract relationship between the surface tonal structure and the underlying one, such as that polysyllabic words should be assigned accent 2, or that intonation and tonal accent are actually secondary to stress structure. These models offer no explanation neither for the pattern in compounds, nor the dialect geography it exhibits. Typological differences are discussed in the article The typology of tonal accent (Riad 2025) and in chapter 3.
2. The concrete approach also identifies accent 2 as the marked case, that is, the tonal category that needs explanation (it contains the lexical tone). This manifests itself in the fact that accent 1 is always a tonal subset of accent 2. Anyone who claims that accent 1 is the marked case faces a so-called abstraction problem. This is discussed in Gussenhoven & Riad (2025) and in chapters 1 and 2.
3. The structural factors are also important for the diachronic questions. Regarding the occurrence of the tonal accent, there is some confusion because researchers do not agree on *what* it is that should be explained. In my opinion, it is the presence of a tone in accent 2 (a lexical tone, i.e. a concrete phonological object) that should be explained. Others have focused on explaining the occurrence of a distinction in minimal pairs (and-en ‘the dock’/ande-n ‘the spirit’, steg-en ‘the steps’/stegen ‘the ladder’). In Riad (2023) and in chapter 4, I clarify the differences between these approaches, demonstrating how the appearance of a concrete tone logically precedes the appearance of a contrast, and that it is therefore crucial which part of the tonal structure one explains the emergence of.
4. the representation also points out what causes the difference between so-called two-peaked and one-peaked dialects. When the lexical tone is high, the dialect is two-peaked, when it is low, the dialect is one-peaked. This follows from a requirement for intonational accents to contain a high tone in North Germanic. The same generalisation underlies the distinction between dialects that have a prominence difference between small and big accents. This is discussed in Gussenhoven, Riad & Wetterlin (2025) and in chapter 2.
The concretely phonological is also reflected in the many phenomena involved in the development of stød from a tone-accent system (chapter 6). The concrete phonological (in contrast to other more abstract assumptions) is thematised in several chapters of the monograph.
Alongside the work on the tonal accent, I have also conducted research on reading during the sabbatical. Together with Maria Lim Falk, I have developed a new reading model with a linguistic basis that includes both first language readers and second language readers. The article has just been submitted to a journal (Riad & Lim Falk, 2024)
Results beyond publications
I have made some new contacts in Berlin, especially in the field of teacher training and teaching German as a second language, an area where Sweden is further ahead than Germany (mutatis mutandis). During the visit to the US, I received feedback on my research from colleagues at the universities we visited, especially the University of Illinois, the University of Texas, and California State University, Fresno, both in terms of tone accent and reading research. I was also introduced to a couple of undergraduate and graduate student groups and gave a couple of classes on my topic. My work in second language reading has led to an invitation to speak at the 10th Nordic Dyslexia Conference in Stockholm in August 2024.
New research questions
The work on the article on reading has generated ideas for appropriate intervention studies. This is particularly true in the area of phonological awareness and decoding. For example, it would be interesting to test whether an early focus on phoneme differentiation in newly arrived adolescents and adults, i.e. targeted work on creating mental categories for all Swedish phonemes, would lead to faster reading development (and thus a better school prognosis). A similar study could be done with younger second language speakers. Similarly, interventions could be made to train decoding speed to see if this would lead to better reading comprehension in the same groups of pupils.
How the results have been disseminated and if and how collaboration has taken place.
The book is at a fairly advanced stage (about two-thirds of it is written) and I expect to have a finished book manuscript in the spring term of 2025. Most of the articles I have written will be published next year. Some of these articles will be published with open access. I will post preprints on my profile page at SU (https://www.su.se/english/profiles/triad-1.183671?open-collapse-boxes=research-publications)