Lecture Stress in Germanic languages

particularly suited for an encounter with languages of the Romance type, where 
main stress is assigned rightmost. In all Germanic languages a large part of the 
lexicon is formed by loanwords, mostly borrowed from Latin, French, or Greek 
(the latter often via Latin). These loanwords are typically polysyllabic and have 
rightmost main stress. French loanwords contribute a pattern with final stress. 
Latin loanwords or loanwords from other Romance languages such as Italian come 
with the typical Latin pattern where stress falls on a heavy penultimate syllable, if 

---

there is one, else on the antepenultimate syllable. The latter pattern can be analyzed 
(as Latin stress usually is), as characterized by Nonfinality / extrametricality, 
banning the main stress foot from final position, and placement of a moraic trochee 
as close as possible to the right edge of the word. 
a. Typical Germanic pattern: (ˈH), (ˈHL)
b. Typical Romance patterns:
final stress: ....(ˈH)#
penultimate (Latin) stress: ....(ˈH) o #
antepenultimate (Latin) stress: ....(ˈHL) o / (ˈLL) o# 
The encounter between Germanic and Romance patterns has had effects that 
differ across the Germanic languages. On the one hand there are Icelandic and 
Faroese, where initial, Germanic stress has been imposed also on (older) Romance 
loanwords, shifting the rightmost stress those words had in the donor languages. 
However, the other Germanic languages have chosen another possibility: They 
have reinterpreted the parameter of main stress placement as ‘rightmost’. A choice 
of this type preserves rightmost stress in loanwords. At the same time it is 
compatible with the stress patterns in the Germanic part of the lexicon. Linked to 
the integration of a rightmost main stress pattern is the question of Nonfinality 
(extrametricality) restrictions, a systematic feature in the major donor language 
Latin. By incorporating Romance patterns, Nonfinality effects became part of the 
metrical system of the Germanic languages. Yet, given the other major donor 
language, French, also words with final stress became part of the lexicon. Thus, 
due to the heterogeneity of the loanword vocabulary (some directly from Latin, 
some from Latin daughters such as French or Italian). Nonfinality effects did not 
become completely predictable, in the modern Germanic languages. The result is 
metrical systems where main stress is characterized by some clear regularities on 
the one side, but lexical idiosyncrasies on the other. The hybrid nature of main 
stress placement, partly governed by rules, partly lexicalized, has brought analysts 
to formulate algorithms for main stress placement which, inevitably, must leave 
room for exceptions, while it has convinced others that main stress placement 
should be considered altogether lexicalized. 
The partial unpredictability of main stress placement becomes particularly 
clear when considering variation within the same language. Thus, in regional 
varieties of German some lexical items vary as to placement of main stress inside a 
final three-syllable window. In most varieties of German, Mathematı´k has final 
stress, but in Austrian German it has penultimate stress (Mathema´tik). Te´lefon 
has mostly antepenultimate stress, except in Austria, where it is stressed on the 
final syllable (Telefo´n). Labo´r is predominantly stressed on the final syllable, 
except in Switzerland and the region of Vienna, where we also find penultimate 
stress (La´bor). Other words, such as Kı´mono / Kimo´no are given by 
pronunciation dictionaries with variable stress, without indicating any regional 
distribution. However, unpredictability has its limits. Among the clearest 
regularities characterizing all Germanic languages with rightmost main stress is the 
restriction of main stress to a final three-syllable window. In Danish, Dutch, 

---

English, German, Norwegian, and Swedish, main stress falls predominantly on the 
final, the penultimate or the antepenultimate syllable of the word. 
Main stress inside a final three-syllable window 
Placement of secondary stress 
In the literature on stress in Germanic languages, iterativity and 
directionality of rhythmical secondary stress have been discussed in less detail than 
main stress placement. This may be due to the fact that phonological or acoustic 
correlates of secondary stress are often more difficult to determine than those of 
main stress, leading sometimes even to statements that secondary stress is absent in 
a certain language. The major exception here is English, where secondary stress is 
detected more readily, given that unstressed vowels are systematically reduced. 
Correlates for secondary stress, however, are sometimes indicated also for other 
Germanic languages. In East Norwegian varieties, a stress shift can occur to an 
initial, secondarily stressed syllable (pro`teste´re → pro´teste`re ‘to protest’), but 
not to an unstressed syllable (beto´ne → *be´tone ‘to accentuate’). In Dutch, 
unstressed syllables can be reduced to schwa, differently from secondarily stressed 
syllables. For German, Alber mentions a process of optional insertion of glottal 
stops in hiatus contexts affecting secondarily stressed syllables, but not unstressed 
syllables; Jessen observes that vowels bearing secondary stress preserve some 
contrast in vowel length, while unstressed vowels do not. 
Notwithstanding the fact that the few accounts on secondary stress in 
Germanic languages other than English and German are mostly based on 
grammaticality judgments, they still return a rather homogeneous picture as to the 
setting of parameters. All accounts contemplating the existence of secondary stress 
do assume that secondary stress is iterative and most claim that secondary stress 
feet are aligned to the left. 
Iterativity and edge-orientation of secondary stress in the Germanic languages 
Secondary stress in most Germanic languages is still not studied in great 
detail, even though evidence for its existence increases as new experimental 

---

methodologies are applied to metrical structure. The descriptions that we do have 
point to a pattern of iterative secondary stress characterized by left edge 
orientation. The more detailed analyses available for English reveal that at least in 
certain contexts secondary stress is quantity-sensitive, an observation confirmed 
also for Dutch, German, and Norwegian, and that as main stress, secondary stress 
is subject to lexical idiosyncrasies. 
Intonation in Germanic 
This contribution focuses on intonation, which listeners perceive as the tune 
– or the rises and falls – of an utterance. The unit of analysis is the intonation 
phrase (IP), which can range in length from a single word to a complete sentence. 
In Germanic languages intonation is used for a range of linguistic functions 
including phrasing (i.e., dividing the speech stream into chunks), signaling 
sentence mode (i.e., distinguishing, for example, declaratives from yes / no 
questions), and highlighting information (i.e., focus). Pitch does not perform its 
work in isolation, but it functions in combination with loudness and lengthening 
cues. The Swedish example from Bruce and Granström demonstrates how the 
insertion of an IP boundary, marked via ||, works to disambiguate otherwise 
ambiguous utterances. 
Intonation may also be used paralinguistically, for example, to express 
emotions. Consider how an English speaker might modulate intonation to show 
emotion when uttering the statement “It’s snowing again.” If the speaker recently 
purchased new skis, s/he might indicate excitement by raising the pitch on 
“snowing.” On the other hand, if the speaker must move a car from the street to 
enable the snow plow to clear the street, s/he might express frustration by lowering 
overall pitch and using monotone intonation.. 
Analyzing Intonation Bolinger described intonation as a “half-tamed 
savage.” Gussenhoven explains this metaphor by distinguishing between the 
“discretely represented prosodic structure” (the tamed half) and the “unusually 
generous scope” that speakers have “in the implementation of fundamental 
frequency” when they speak (the untamed half). Consider yes / no questions in 
English. Although the default intonation contour ends in a rise, speakers may 
choose to produce this type of question with level pitch or even with a fall. The 
available research enables us to make generalizations about both the linguistic 
forms and functions of intonation. It is possible to analyze intonation from 
different perspectives. Phonetic analyses examine the details of the acoustic 
correlate of pitch, fundamental frequency (F0). F0 is a measure of the frequency of 
vocal fold vibration, measured in Hertz (Hz). Pitch is perceived to be higher when 
a speaker’s vocal folds vibrate more quickly. Phonetic analyses of intonation often 
make use of pitch tracks like that in Figure 8.1, which is presented together with its 
corresponding spectrogram. Researchers performing phonetic analyses use 
measurements including F0 minima and maxima and pitch range, which is the 
difference between a speaker’s highest F0 and the baseline within an F0 contour. 
Phonological studies of intonation focus on intonation systems and therefore make 
use of more abstract categories that capture the overall pitch contour and relative 
pitch height. 

---

It is possible to make broad generalizations about roles of intonation 
contours in signaling sentence mode. Across the Germanic languages there is a 
tendency for declarative utterances and wh-questions to exhibit falling intonation 
and for yes / no questions to be produced with a rising contour. Continuation, for 
example, in the production of utterances containing a subordinate clause followed 
by an independent clause, is often signaled via a level or a rising contour. There is 
one constituent in Germanic utterances that receives the primary emphasis. It is 
spoken more loudly and its duration is longer. In the case of neutral utterances 
spoken “out of the blue” or in response to a question like “What’s happening?”, 
speakers tend to emphasize the last content word of the utterance. It is possible, 
however, to highlight another unit in an utterance through the use of intonational 
cues. The emphasized constituent in both instances receives the sentence stress, the 
phonetic manifestation of which is referred to as the nuclear pitch accent. 
Direct object focus
Q: Hvað skrifaði Marı´a upp?
‘What did Marı´a write up?’
A: Marı´a skrifaði [söguna]Foc upp.
Marı´a wrote story.
def
up
‘Marı´a wrote the story up.’
Verb focus
Q: Hvað gerði Marı´a við söguna? 
‘What did Marı´a do with the story?’
A: Marı´a [skrifaði]Foc söguna [upp]Foc. 
Together with stress and rhythm, intonation is classified as a 
suprasegmental, or prosodic, aspect of speech. Units of suprasegmental analysis 
can range from syllables to entire texts. Of primary concern in linguistic analyses 
of intonation are the IP and the pitch movements associated with both the 
prominent syllable and the end of an IP. The IP tends to be delimited – especially 
in read speech – by a pause and lengthening of the final syllable. In addition, the IP 
tends to drop in amplitude at its end and is often followed by pitch resetting. In 
spontaneous utterances, however, it can be more difficult to delimit IPs, and they 
may or may not coincide with syntactic units. Phonological models of intonation 
tend to fall into one of two categories: holistic and compositional. Holistic models 
such as the Fujisaki model are based on the assumption that complete intonation 
contours carry meaning. Many of the compositional phonological models of 
intonation are based on the early work of Bruce on Stockholm Swedish, in which 
he demonstrates hierarchical structure of suprasegmental features. Importantly, he 
shows that intonation peaks are used to mark focus and that the basic units of 
intonation include word accents 2 (i.e., accent 1 and accent 2 for Swedish), 
sentence accent (i.e., the syllable in an IP that is emphasized) and terminal juncture 
(i.e., boundary signals).
Among the West Germanic languages, English is usually described as 
having larger pitch movements than both German and Dutch. This may be because 
English relies more on intonation than other Germanic languages do, due to its 
relatively set word order. In her comparison of Southern Standard British English 

---

and Northern Standard German, Grabe found similarities in the use of falling pitch 
accents, which she categorizes as H*+L in both languages. As with other Germanic 
languages, the basic contour of English declaratives is a fall with nuclear pitch 
accent on the final content word. Bolinger posits that Standard American English 
and Standard British English share a single intonation system, but a few of the 
important differences are taken up below. 

---

Document Outline

• Even though it has been shown that these parameters are not necessarily on or off, for a certain language, but that languages may e.g., be quantity sensitive in certain contexts, while they are quantity-insensitive in others, the parameters still serv...
• The encounter between Germanic and Romance patterns has had effects that differ across the Germanic languages. On the one hand there are Icelandic and Faroese, where initial, Germanic stress has been imposed also on (older) Romance loanwords, shifting...