This paper is concerned with four Scandinavian prosodemes: acc. 1, ace. 2, sterd, and no s t~d .Its aim is to establish the function of each of them. The much-discussed questions of whether the Swedish and Norwegian accents are dynamic or musical in nature and whether it is possible to detect tonal ingredients in the realizations of the Danish stprd and no stprd are of secondary importance. What interests us here is the role all those entities play in the system, not the details of their manifestation.
1.As was clear to the very first investigators of Swedish and Norwegian stress, acc. 1 occurs mainly in the words which had one syllable in Old Icelandic (01). Hence the tradition to call acc. 1 monosyllabic and ace. 2 polysyllabic. This tradition holds, on the whole, for the modern languages too: monosyllabic words have acc. 1, and in polysyllabic words ace. 2 is felt as more natural. C . 4 . Elert (1964: 31-33, and several times later) noticed that the main function of acc. 2 is purely connective. If we compare the Swedish utterances vdr 'sven slea 'flagga (our page will run up the flag' and vdr 'svenska 'flagga (our Swedish flag', we shall see at once that in the syntagm svenska (from the second example) the syllables sven and ska are understood as parts of one word because they constitute the domain of acc. 2. When we hear a string of syllables with acc. 2, we know that there is no word boundary within it; thus, acc. 2 performs an anti-delimitative (connective) function. On the other hand, the tone of ace. 1 syllables carries no information about their structure. Acc. 1, as well as ace. 2, is devoid of any delimitative power. The syllable hdll, for example, need not be a final syllable in a word (cf. hill 'distance' and hdllen pp. of hdlla 'keep'); similarly hdllen may be one word and a group of two words (cf. Hd11 e n minut utanfor porten 'stay a minute outside the gate'). But it has no task corresponding
SCANDINAVIAN ACCENTS I N THEIR RELATION TO ONE ANOTHERthe connective function of acc. 2. To sum up: acc. 2 signals the absence of a word boundary within its domain, while ace. 1 takes no part in word delimitation. Ace. 2 can connect parts of a compound word: cf. en stormans drakt [En 'stu:rman:s 'drokt] 'an outfit for a magnate', en stor mansdrakt [En 'stu:r 'man:sdrekt] 'a big man's outfit', en stormunsdrakt [ E n 'stu:rman:sdr~kt] 'a magnate's outfit'.The connective role of acc. 2 is also quite manifest within a wordgroup. A syntagm with acc. 2 is more unified and idiomatic than the same syntagm with ace. 1: cf. the Norwegian Hann 'kommer-med dampbdten -ti1 Hamburg 'he is going to Hamburg by steam-boat' and Hann 'kommer-med dampbdten -til Hamburg 'He will catch a steam-boat for Hamburg' (with emphasis laid on the verb). Ace. 2 turns kommer med into a compound verb, and acc. 1 preserves med as preposition (Broch 1937: 315-316). The Norwegian phrases 'tap& noe 'put on something' and 'ta pd noe 'touch something' will be rendered in Standard Swedish as 'ta pd ndgot and ta 'pd ndgot. But in the Uppland dialects ...
