What holds a stanza together when it has no classic meter? How has the stev two-pulse (dipod) escaped discovery if it has existed for a millennium in folk poetry? Performance of stev, as sung traditionally unaccompanied, provides us with special criteria, first for connecting with older stev in time past and then for examining connectors that latch and bind within the stev stanza. New theories emerge concerning the origin of nystev. The refrain stev and Old Norse stef are also discussed.

Keywords:rhythm and rhyme, stev, accentual poetry, foot-tapping.

1. The Stev Stanza Withstanding Time

The Norwegian lyric one-stanzas called stev are still found in Setesdal and Telemark folk music tradition. The form of both nystev (“new stev”) and gamlestev (“old stev”) has been defined for decades as a 4-lined independent stanza. Nystev were said to have 4 accents per line (underlined in the following):

Eg gjeng og lengtar, eg gjeng og drøymer, yrst lauvet sprette og fossan fløymer. Ko felar hugen eg alli veit, men det er sòm logande elden heit.1

Typical of nystev, the first couplet employs feminine rhyme (i.e., the last rhyming syllable is unaccented) while the next two lines are bound by masculine rhyme (i.e., the last rhyme is an accented syllable).2

A gamlestev has a metric pattern like that of the common ballad meter, that is, with lines of 4, 3, 4 and 3 accents (underlined):

Vent sò er det i Vinje kyrkje når brurine gjenge inn, der tar dei ‘kje ljosi brurine bera, det lyser av blomekinn.3

Side: 151

Usually only the second and fourth lines rhyme, either with end-rhyme or with assonance, in which the stressed vowels in the last word sound alike but their following consonants are different, as in “dag” and “hav”.

A four-line stanza structure is scarcely eye-catching because this stanza form is common today for everything from serious to trivial poetry, including songs and ditties for occasions such as birthdays. Study of stev texts and melodies holistically in traditional performance indicates a new concept of stev, the stev accentual theory. This theory maintains that the stev stanza is only a “look-alike”, and in fact is a foreigner to today’s 4-line stanza form: traditional stev performers foot-tap seemingly irregular, non-dance rhythms which prove, however, to be a systematic, predictable, flexible asymmetric two-pulse rhythmic mode based on word accents of content words such as nouns, verbs or their modifiers (i.e., not articles and conjunctions). The foot-taps occur in pairs, with a longer interval of time elapsing after foot-tap 2 (T2) than after foot-tap 1 (T1) and resulting in a two-pulse sometimes referred to as a “dipod” (meaning “two-footer”) or a combination of terms (stev dipod, two-pulse dipod, or stev two-pulse dipod), depending on context and emphasis.4

The two-pulse dipod pattern, a core of both nystev and gamlestev, is recognizable in printed texts when dipods are separated by space (Exs. 1a and 2a) or written on separate lines (Exs. 1b and 2b). The latter format emphasizes the two half-stanzas, each of which is always a complete sentence in nystev.5

Figure 1. Two formats for presenting a nystev in print.

Ex. 1aÅ du va den blomen sòm anga meste, du va den gjenta eg lika beste. Og myrk bli heile min framtidsveg; visst inkje fylgjast eg fær med deg. Ex. 1b Å du va den blomen sòm anga meste, du va den gjenta eg lika beste. Og myrk bli heile min framtidsveg; visst inkje fylgjast eg fær med deg.

According to the stev accentual theory, gamlestev lines 1 and 3 consist of two dipods or two-pulses each. These “long lines” may be profiled with the two-pulses separated by space as in Ex. 2a or divided into “short lines” (Ex. 2b). Long lines 2 and 4 rhyme together in gamlestev and each has 3 predictably foot-tapped accents (T’s), grouped as one accent + a two-pulse (i.e., 1+2 T’s) or in reverse order (2+1).

Side: 152

Figure 2. Two formats Ex. 2aand 2b for presenting a gamlestev in print. The accents are grouped as indicated by the counts to the right of each line.

Ex. 2aVinden leikar i bjørkeli bekkjen han ute suslar. Med smiledòkkji sò ven og blid den guten te Anne ruslar. 2+2 1+2 2+2 1+2Ex. 2bDet er sò vent mæ Vinje kyrkje når brurine gjeng i flòkkar, der tar dei ’kje ljosi brurine bera, det lyser av gule lòkkar. 2 2 2 1 2 2 1 2

Salient characteristics unique to the stev stanza and apparent from studying stev as performed are summarized from earlier writings9 and presented in Table 1:

Table 1. This reference table reviews basic textual and rhythmic characteristics of the nystev and gamlestev stanza, according to the stev accentual theory.

1. Predictably foot-tapped points (T’s) dependent upon the text. Predictably foot-tapped points constitute the backbone of stev structure. The sound of the foot-tapping is an important feature, but is not always present or audible.
2. Event-structured poetry. Foot-tapping may sound rhythmically free, but when present is predictably always on word accents (never on unaccented syllables) of “important” words. The foot-tapped points occur in pairs, called the stev “two-pulse”, or “dipod”. The duration of the two-pulse period is not time-controlled and thus is variable. The number of unstressed syllables may be none or up to several. Unstressed syllables may be sung slowly or held out, no matter how many syllables. Such poetry built solely upon word accents is called pure accentual poetry.
3. The rhythmic structure of phrases and syntax. The dipod employs grammatical-syntactic grouping: no syllables of words straddle two dipods.
4. Rhyme. Alliteration, also called “front-rhyme” occurs in many gamlestev and sometimes in nystev. Gamlestev more often show assonance (identical or similar vowel sounds rather than identical consonants in the end-rhyme). Front-rhyme and end-rhyme emphasize the two-pulse (see section 2).
5. The stev two-pulse. This unit of poetry has been likened to the two-accent visuor∂, the smallest unit in Old Norse poetry. The stev two-pulse appears to be the heart of both gamlestev and nystev.

Gamlestev rhythm and rhyme through many centuries will now be explored. A gamlestev and an Old Norse stanza are compared first, thereafter a few gamlestev from two of the few dated sources existing (1698 and 1740) and then four Draumkvedet stanzas.

Side: 153

A gamlestev melody from today’s tradition, shows an interesting variation of two dipods melting into 3 pulses to accommodate variation in a text as sung by master singer (stor kvedar) Aslak Brekke. This leads to reconsideration of gamlestev form. After this investigation of gamlestev, nystev are then focused upon.

Gamlestev Earlier (1740, 1698, and “Draumkvedet”) and Today

When the two-pulse principle is applied to the first and third lines of gamlestev, the poetic form is then seen to be the same as that of a type of Old Norse poetry (Figure 3). This seems to support a probability that gamlestev and kveding have been part of an unbroken since Old Norse times.10 The hypothesis is further strengthened by the similarities which Landstad pointed out in the contents of the gamlestev “Hur∂i nurkar i kaldan jønni” (Figure 3 left) and contents of Draumkvædet, Sólarljód, and particularly an Old Norse Skirnesmál stanza, shown in Figure 3 (right).11 Riddles and cryptic expressions are common in both gamlestev and Old Norse poetry. Here the gamlestev relates that “the winter night was longer than half my age”, and apparently refers to “the moon”, which was the “year counter” or “year measurer” (árteller, ármál in “Alvismál”) in Old Norse mythology.12

Figure 3. An old gamlestev (left) and an Old Norse parallel in contents and meter.

Hur∂i nurkar i kaldan jønni og jøklanne heng ikring, eg totte lenger den vetternotti enn halve alderen min.Löng er nott langar ’ro tvær hve um threyjak thrjár? opt mér mána∂r minni thotti en sjá hálf hynott.

Landstad’s Norske Folkeviser is available in facsimile editions. Here are many pages of gamlestev, along with Landstad’s extensive notes. Not so easily accessible to the general reader are two written sources from 174013 and 169814 containing a few gamlestev texts that help fill this time span. Applying the stev accentual theory indicates predictably accented syllables [underlined] and assumed division (caesura)15 between dipods.

Figure 4. Gamlestev from Paus 1740 (Ex. 3) and Klim 1698 (Exs. 4–6).

Ex. 3Stumpen stende i Forstågu, æ brænt i båe Endar; den skå ivi Nakkjen din, om du inchie Steve vender.Ex. 4 Kiempo sidd i anveiggien, å spendar i stinde Leggie. Dridtne heb han sckoe-Soelan, o Klysin bær han ti skieggie.
Ex. 5Kasta no burdt dei liuode stæevæ, tøin kie pruvandd mundd. Tagg i faggran frees-taatin, o drick tof aae - brundd.Ex. 6Ægg tegkk inggien giesll ad me, de æ saa vundd ey vinddo. bær æ giedd fimbto giedd ell eino eelskoggs qvenddo.

Side: 155

The four examples in Figure 4 have ordinary gamlestev form except for the first line in Ex. 3 “Stumpen stende i Forstugu” and Ex. 4 “Kiempo sidd i anveiggien”, which have only three accents.

In Figures 5 and 6 are shown 4 stanzas from the well-known lay or series of gamlestev, called “Draumkvedet”, 16 written down in the 1800’s but thought to be from the 1200’s.17 These examples are chosen both for their assumed age and also a special feature: the stanzas in each pair are nearly identical.18 In the first pair, Exs. 7 and 8, only one word differs in the first line (fyste/andre), and only one in the second line (ring/mog). Lines 3 are identical. Lines 4 retain identical initial letters “n” (nagelen/ny) and “f” (fing/fór).

Figure 5. A pair of nearly identical stanzas from “Draumkvedet”.

Ex. 7De va mæ den fyste utreisi eg drog gjænom toner ring sonde lout mi skarlags kåpa å nagelen av kvor mi fing.Ex. 8De va mæ den andre utreisi eg drog igienem tønner mog sonde lout mi skarlags kåpa å ny e heimate fór.

The contents of a stanza pair may be similar or antithetic, as in Figure 6 in which “sonnate” (“from the south”) contrasts with “norate” (“from the north”). The words “tvist” and “qvast” (in line 2) contain similar sounds, “v” and “st”.

Figure 6. A pair of stanzas from “Draumkvedet” illustrating “thought-rhyme”.

Ex. 9Her kjæm i fær her sonnate å dein rei nå so tvist fyri rei st. såle Mikjel ete kom Jesom Christ.Ex. 10He [sic.] kjæm ei fær her norate den rei nå så qvast fyri rei Grutte Gråsjæje alt mæ sin svarte hat.

Besides similarities in words and initial sounds, the ideas are usually parallel, that is, either similar or contrasting. In the second pair from “Draumkvedet” (Figure 6), the name “st.

Side: 156

såle Mikkjel”
is antithetic to “Grutte Gråsjæje” (line 3), and “Jesom Christ” as symbolic for light contrasts to the “svarte hat” (black hat, line 4).19 Such similarities and contrasts in contents add a special parallel rhythm or “rhythmic rhyme” in stanzas, known as “thought rhyme”.20 The “Draumkvedet” stanza pairs also demonstrate how stev may become “turned” or “changed” (Norw.: vendestev), by minimal alterations.21 Since Draumkvedet is considered to be a religious, visionary poem, one would not expect to find features of ballads such as wordplay, riddling, capping, “vendestev”, personification and apparent reference to mythology. Perhaps the contents of “Draumkvedet” should be reevaluated.22

Exs. 3 to 10 illustrate 8 gamlestev including 4 “Draumkvedet” stanzas. One sees in the printed texts that initial letters are often repeated. Stev, however, are sung and have been handed down in aural rather than literary tradition. Thus, sounds are more relevant here than letters seen on paper. In some stev, there are quite a number of the same sound within a line of poetry, for example the three “st” sounds in the line Stumpen stende i forstågu” (Ex. 3) if we include the “st” inside the word “forstågu”. Gamlestev used “front-rhyme” (stave-rhyme) and possibly also internal sounds within words. An aural culture could play with sounds and words. In Table 2, letters that might rhyme together intentionally are italicized and noted to the right of the corresponding line.23

Table 2. A tentative proposal of initial and internal alliterating staves (sounds) in stev from Paus 1740 (Ex. 3) and Klim 1698 (Exs. 4–6, 11 and 12).

Ex. 3Stumpen stende i Forstågu, æ brænt i båe Endar; den skå ivi Nakkjen din, om du inchie Steve vender. [Lines 1–4] st, n, en, d b, n, en, d n, d st, v i Ex. 4Kiempo sidd i anveiggien, o spendar i stinde Leggie. Dridtne heb han sckoe-Soelan, o Klysin bær han ti skieggie. s, n, g s, n, g s, n s,n, g
Ex. 5Kasta no burdt dei liuode stæevæ, tøin kie pruvandd mundd. Tagg i faggran frees-taatin, o drick tof aae-brundd. st n t, f, g, aggEx. 6Ægg tegkk inggien giesll ad mee, de æ saa vundd ey vinddo. bær æ giedd fimbto giedd ell eino eelskoggs qvenddo. g v, n (giedd) e, l, g,v
Ex. 11Høi do dæ kiæ stallbru mind, æg tala no te deinom. Kos æ tiis ti Seetesdall, køes giedd dickk Riuobeskinnom. d d k, (s), t, is kEx. 12Høi dø dæ goe stalbru mind, o æg ve dee væ suora. Soe giedd ockk Riuobe-skind, so dickk vidd-voera. d d, s s s, v

As an aside, the stev in Table 2 seem innocent enough, but the one in Ex. 11 “Høi do dæ kiæ…” provoked in 1672 a stev (Ex. 12) and a “trusty stabbing” with a knife.24

A gamlestev may have only 3 stressed points in the first or third line. Such is the case in a gamlestev melody from today’s tradition as sung by master singer (storkvedar) Aslak Brekke (Figure 7). The missing accent may pass unnoticed because the gamlestev melody accommodates the difference so smoothly.25

Side: 157

Figure 7. Modification of a gamlestev melody to accomodate a 3-accent text. The text is from “Beltespenning”, the first line of stanzas 1 and 2.

With this one deviant line though, 3 of the 4 lines of the stanza have 3 accents, and it is easy to hypothesize a gamlestev stanza with 3 accents in each line. A stanzaic song (vise) called “Murukleiven” 26 has features typical of a gamlestev melody, but was not classified in 1983 as such because of its 3-accent pattern.27 Recognizing this pattern opens for gamlestev variants. Table 3 compares the types, starting with the standard gamlestev definition of 4, 3 accents (left column), and then the common gamlestev and common ballad meter as defined by dipods in the next column. From here one may either move to the right through the hybrid gamlestev “Beltespenning” to the 3-accent form on the far right or move downward to the basically 2-accent ljó∂aháttr form28 advocated by Erik Eggen29 or the solely 2-pulse form as presented by Ivar Handagard.30 In ljó∂aháttr, each dipod is a line, and the last accent in each half-stanza is single and allotted its own line.

Side: 158

Table 3. Gamlestev types according to the number of predictable foot-tapped accents (T’s). The number “3”, encased by quotation marks, represents either a 1+2 or 2+1 combination of accents.

Side: 159

Eggen’s ljó∂aháttr form for gamlestev (see the “”ljó∂aháttr” column in Table 3) brings out the dipod structure. Yet there are two arguments in favor of keeping the longer lines as shown in the common ballad meter and common gamlestev meter of 2+2, 3, 2+2, 3 (middle column). Firstly, it is not sure that 3 should always be split into 2+1, as Eggen theorizes. On the contrary, the film frame counts and textual phrasing sometimes suggest the reverse (i.e., a 1+2 division) as more appropriate. (see Exs. 2a and 2b). Secondly, the long lines of 2+2 and 3 accents can reflect both the dipod structure and the longer musical phrases as well.

Eggen is not the only one to emphasize the two-accent line in gamlestev. Handagard even calls them “dipodi” and gives an example of three dipods in each half stanza:

Hjorten spilar // i heio nord, // han rekkjer sin hals – hau! so gjere // dei Herjus sønin’ // som giljar med fals.31

Handagard’s example here is from Liestøl and Moe, and he comments that Landstad’s version opens in the second half-stanza with “one accent too many”: “hau! så gere alle // dei Herjus syninn”.32 Perhaps, however, that line illustrates gamlestev’s adaptability to accommodate an extra, unstressed word (“hau”), or 3 accents in place of two two-pulses.

After this examination of gamlestev, a question arises: how are they related to nystev? An examination of older nystev from the past may answer that question.

Looking Back for the Nystev Stanza: Málsháttakvæ∂i

The origin of nystev has been a riddle. The genre proves not to be from the Norwegian springar folk dance, a popular theory.33 Neither does it come from abroad, another theory:34 nystev end-rhyme couplets may prove deceptive, for instance in the Danish and German stanzas shown here (Table 4).35 Both stanzas lack the smoothness and flow of genuine stev as we know the style today. The Danish stanza lacks the typical nystev element of a complete thought, a full sentence and a grammatical stop in each couplet: here the verb appears first in line 4. The German “Spruch”, from the 16th century, is heavy because of consonantal word endings, in contrast to the lightness and lilting line created by vowel endings in Setesdal and Telemark dialects. In addition, its first line has too many words and syllables between the proposed stresses. Thus, these stanzas are not convincing as nystev; at best they appear to be poor imitations. The Swedish researcher Richard Steffen discovered Swedish nystev translated from Danish and also found that nystev were easy to imitate, for instance in broadside ballads. His conclusion, in 1898, was that nystev is a Norwegian genre.36

Side: 160

Table 4. Steffen’s proposed nystev (left) exhibit non-stev characteristics (right).

Theoretically nystevNon-nystev traits
A Danish stanzaRigens chantzeler, Anthonius Bryske, vel lærd och skickid i latin och tyske, i den danske lov forfaren vel, dermet hiulpit mange til ræt og skel. – no full sentence stop after “tyske”, – many consonantal word endings, – long words (chantzler, Anthonius), – non-Germanic words and accents (Anthonius, latin), – no verb until the 4th line (“hiulpit”).
A German Spruch
Ein man schal sick nicht tho szer berömen, dat sin hoff sta vul schöner blomen; dar kumpt ein ripe up eine tidt und maket em all siner blomen qwidt.– Too many heavy syllables, with consonantal endings in first line, especially between proposed stresses on “man” and “nicht”.

Knut Liestøl and Moltke Moe described in 1912 nystev’s having been influenced by music from abroad.37 In contrast, Mortensson-Egnund wrote in 1914 that nystev and gamlestev were Norwegian. He found alliterating sounds (Engl.: “staves”) in nystev, as in the following example at the beginning of what he calls two “tone-heavy” words within a long nystev line:

I dænne værdei so mykje vantar.               In this world so much is lacking.

In his example given here from a nystev, he emphasizes (in bold) the staves alliterating in Old Norse stave-rhyme fashion, two like sounds in the first long line and one in the second:

Aa der hev tippa so mang ei taar som der er dagar i tusen aar.38 O, there have fallen so many tears as there are days in a thousand years.

These stave sounds latch the phrases together here just as in Old Norse visuor∂ [short lines of poetry]. On the basis of the alliteration, he wished to claim that nystev are as old as Old Norse poetry, yet did not dare state such without further evidence. He then found documentation in a poem called “Málsháttakvæ∂i”, 39 thought to be from the early 1200’s. He illustrates how close its stanzas are to traditional nystev in form, phrasing, end-rhyme and language style and content40 by translating stanzas into Norwegian nystev idiom41 (Figure 8).

Side: 161

Figure 8. Comparing a two-stanza strophe from the Old Norse poem “Málsháttakvæ∂i” (left) with Mortensson-Egnund’s translation into nystev form and style (right).

Old Norse (before 1223)Norwegian nystev translation, 1914
Ekki hevk med flimtum farit, fullvel attak til thess varit, yrkja kann eg vánu verr, vita tykkist that madrinn hverr. Aa inkje vil eg med sneiord fara, fullvel lyt eg meg for det vara. Men vild’ eg sneide, det godt eg kann, det tykkjest vita no alle mann.
Ekki verda fródum farald: Finnan gat tho oerdan Harald. Honom thótti sólbiort su, sliks doemi verda morgom nu.Paa kjærleikssuti er ingin ende, sjølv kloke Harald fekk henne kjenne. For honom Finna var solbjart brur, slikt røyner baade eg og du.

The strophe presented in Figure 8 shows what appears to be two nystev stanzas, each having textual dipod structure and typical nystev couplets, the first bound by feminine (f) end-rhyme, and the second by masculine (m).42 Mortensson-Egnund proposes that the lay, with its many stanzas showing nystev form, was written by Bishop Bjarne Kolbeinsson (– † 1228) from the Orkney islands who visited King Sverre in five different years and surely sang for him:43

og det som kongen lydde paa laut ogso mennane hans finne hævt. Desse lette stevi til songar-bispen fraa Orknøyom kom vonleg til aa ganga fraa mann til mann, den friske tonen hjelpte til og vart mønster for ny dikting der alle mann kunde vera med. and that which the king listened to, so did his men. These light stev from singer-bishop from the Orkney Islands most likely went from man to man, the fresh tone helped and was the model for new poetry in which everyone could participate.

This idea fits into the older concept of art and culture sinking down into the folk culture. A more likely alternative is that the nystev form was already widely spread in tradition and that Kolbeinsson adapted this form, just as Garborg, Vinje, and others have done. In any case, Mortensson-Egnund’s research and arguments supported Richard Steffen’s statement that nystev was”et eksklusivt norsk produkt”.44 This raises the question of what is so Norwegian about the nystev genre. Also, what connects the dipods in gamlestev and in nystev. The answer to both questions appears to be the asymmetric rhythm of the two-pulse and to a lesser degree the rhyme, both of which will be focused upon in the next section.

Side: 162

2. Rhythm and Rhyme as Connectors

The following delves into both rhythm and rhyme as connectors. Normally, poetic meter defines the basic rhythm in a poem, and repetition of the rhythm creates an element that binds the units of the poem together. In stev, it is the T1- T2- stev dipod that is repeated. The repetition establishes a rhythmic form that binds and integrates, in spite of its irregular time length in the individual two-pulse. Even if there is no conscious or audible foot-tapping, the two-pulse rhythm is still there in the text and melody. End-rhyme also sets up a layer of rhythmic repetitions, binding lines together in pairs in ny-stev (lines 1+2 and lines 3+4), while either end-rhyme or assonance binds the 2nd and 4th lines of gamlestev.45 Thus rhyme too is an important connector.

Rhythms are established on different levels by dipods, long lines, end-rhyme (or assonance), sentences and their full stops, and the stanza itself. All of these are repeated. The rhythms just described do help link and latch half-lines. Yet this verbal description does not seem to do justice to the power of rhythm as a structural element. The following table (Table 5) presents nystev rhythm in a new way, as operating on 6 or more levels cumulatively and simultaneously within the stanza.

Table 5. Rhythmic responsions within the nystev stanza, analyzed as operating cumulatively and simultaneously on 6 or more levels.

Level The Responsion Characteristics
(0)T’sThe T pattern may seem random timewise, but the T’s fall on the accents of “important” words, and the sound of the foot-taps sets up expectations.
1T1- T2-The two-pulse dipod includes both the paired foot-taps themselves (T1, T2 dipod, either heard or predictable) and the intervals-of-duration dipod: T1- and T2-. Both dipods are responsions.
2T1- T2- T1- T2- f2 dipods, the second with feminine rhyme.
3T1- T2- T1- T2- fT1- T2- T1- T2- f4 dipods, the second and fourth with feminine rhyme (lines 1 and 2).The sentence comes to a full stop.
4T1- T2- T1- T2- m2 dipods, the second with masculine rhyme.
5T1- T2- T1- T2- mT1- T2- T1- T2- m4 dipods, second and fourth with masculine rhyme The sentence, lines 3 and 4, comes to a full stop.
6T1- T2- T1- T2- fT1- T2- T1- T2- f T1- T2- T1- T2- mT1- T2- T1- T2- mThe first two lines, bound together by feminine rhyme, contrast with the next two lines and their masculine rhyme.
(7)T1- T2- T1- T2- fT1- T2- T1- T2- fT1- T2- T1- T2- mT1- T2- T1- T2- mThe whole nystev stanza.
Nystev rhythmic responsions operating cumulatively on 6 levels, epicted schematically.

Side: 164

Table 6. Rhythmic responsions within the common gamlestev stanza analyzed as operating cumulatively and simultaneously on 6 or more levels. Below “3 T’s” means 2+1 or 1+2. Characteristics are presented more fully in Table 5.

Level The Responsion Characteristics (see Table 3)
(0)T’sThe T’s set up expectations.
1T1- T2-Two responsions: the T1 and T2 pattern and the intervals-of-duration T1- and T2- dipod.
2T1- T2- T1- T2-2 dipods
3T1- T2- T1- T2- 3 T’s 2 dipods + 3 T’s (2+1 or 1+2), The sentence comes to a full stop. A half stanza. in the melody, but can be a full sentence in the text.
4T1- T2- T1- T2- 2 dipods with new melodic material.
5T1- T2- T1- T2- 3 T’s 2 dipods + 3 T’s (2+1 or 1+2), The sentence comes to a full stop.
6T1- T2- T1- T2- 3 T’s T1- T2- T1- T2- 3 T’s The first two lines contrast with the next two lines melodically. Lines 2 and 4 rhyme, often with assonance rather than end-rhyme.
(7)T1- T2- T1- T2- 3 T’s T1- T2- T1- T2- 3 T’s The whole gamlestev stanza.

The rhythmic pattern on each level may be called a “responsion”. “The essential means of imposing a pattern upon rhythm is repetition,” writes Paul Maas,46 and the rhythmic

Side: 165

repetition sets up a response within us. The result is that we consciously or subconsciously recognize and expect the particular rhythmic patterns. Two additional levels might be included, the first one (marked “0”) and the last , designated “(7)”. They strike some readers, however, as being superfluous and are therefore omitted from the main levels. Besides the responsions within the stanza, we have the expectation set up by one stanza following another. The stanzas are often in pairs, especially in stev duels, thus creating the responsion within and of the pair.

Table 6 presents the common gamlestev, with its lines of 2+2, 3; 2+2, 3 accents in the same type of cumulative and simultaneous rhythmic scheme. Gamlestev variations are not included. The 3-accent may consist of either 2+1 accents or 1+2.

Schematically depicting the gamlestev is not so easily done because of the uncertainty of how the 3 T’s occur, whether as 2+1 or 1+2. In addition to the characteristics above, one should also be aware of patterns established by tonal centers and melodic motifs.47

The rhythm set up by these flexible responsions establishes both a common pattern or code and a basis for individual variation and freedom within the framework, as well as nearly endless opportunities for rhyme as may be observed in the next section.

The Rhythm of Rhyme: “Front-rhyme” and End-rhyme

The tables above indicate that end-rhyme is part of responsions and establishes connecting rhythms. “Front-rhyme” may also be found in some stev today. Front-rhyme occurs when the same sound, or “stave,”48 is used at the beginnings of words. As an example, we have the “f” sound and letter [shown in bold] in the nystev “Å du va den blomen”:

Og myrk bli heile          min framtidsveg; visst inkje fylgjast                     eg fær med deg.49

Only one alliterating stave (sound) is found in a dipod, and none in the first dipod quoted, “Og myrk bli heile”. This may seem sparse and incomplete, but is reminiscent of the Old Norse stave rhyme technique in which one avoided two alliterating staves in one dipod – which the poets of that time considered superfluous.50 It is a surprise to find it today and indeed it is not usually found throughout a whole stev or in the Old Norse Skaldic structural manner. The question does arise as to why alliteration is found in this particular stev. Is it because this stev is very old, or was it partially copied, consciously or unconsciously, from an older stev? Is alliteration used here intentionally as a structural element to bind one half-line with another, as in Skaldic poetry, or simply as an ornament for aural enjoyment?51 In any case, the strong stress on foot-tapped points is so structural that alliteration on these points cannot be purely ornamental: it automatically reinforces the rhythmic structure. Ivar Handagard did not call stev rhythm structural, but knew it was important. He wrote that “stave-rhyme words had

Side: 166

two purposes in Old Norse [poetry]: they maintained the rhythm in the stev and at the same time made the contents sink into the memory”.52

In both old gamlestev and old nystev, rhythm appears to be a structural element. Whereas alliteration (“front-rhyme”) has fallen by the wayside in newer gamlestev and nystev, the rhythm is well preserved and is still able to be a chief element in creating unified stev stanzas. This is interpreted to mean that the rhythm is even more structural and basic than the old “structural alliteration” (stavrim). Perhaps one could say, then, that alliteration simply emphasized the rhythmic structure.

End-rhyme is said to have pushed out “front-rhyme” (also called stave-rhyme or alliteration) many centuries ago. The two elements have been considered mutually exclusive in poetry ever since, but stev illustrate that both may be present.53

New Theories for Old Nystev. Omit the Refrains, the Nystev Remains

Definitions of stev always mention the characteristic end-rhyme patterns of nystev and gamlestev, but this knowledge only raises further questions. Why does the first couplet in nystev employ feminine end-rhyme and the second couplet always masculine? Richard Steffen made a strong case for couplets in many countries being dance songs, yet the nystev genre is not for dancing, especially the ultra slow stev of Setesdal.54 The special nystev end-rhyme can be used as a guide into theories of how nystev may have evolved.

End-rhyme or assonance is always present in gamlestev (lines 2 and 4) and end-rhyme in the nystev rhyming couplets in which the first couplet employs feminine rhyme and the masculine. Puttenham, an English writer, already in 1589 lamented end-rhyme as ridiculous “jingling in the ears” which he claimed weakened poetry.55 Evidently he preferred the old principles of accentual poetry and stave rhyme, which however seem to have been threatened by end-rhyme and with time disappeared. Stev appear to be an exception that has maintained a balance in tradition of both accentual poetry and end-rhyme.

Nystev’s feminine and masculine rhyme couplets have provided a challenge as to the origin of the form. One theory proposes that a pair of gamlestev stanzas can create a nystev if the refrains are omitted.56 By omitting refrains, two gamlestev 57 stanzas could meld into a nystev, as shown in Figure 9.

Figure 9. Omit the refrains (right) and the nystev remains (left).

Gamlestev lines Refrains
Kan me ‘ki kvoradre på ordi taka, så skal me ‘ki både den arven hava.på grøneli∂hei∂i: Der spekar ei hind, der leikar ei jomfru i mannshugin så gledeleg.
Kan me ‘ki kvoradre på ordi snå, så skal me ‘ki både den arven fåpå grøneli∂hei∂i: Der spekar ei hind, osv.

Side: 167

Another theory is Rikard Berge’s. In 1922 he wrote that a song called “‘Ride brudeferd’ gives us the clearest proof of how two-lined ballad stanzas became nystev”. 58 He presents nine versions of this song as heard by M.B. Landstad, Sophus Bugge, Johannes Skar, and including five versions written down by Berge himself. He points out that stanzas 2 and 3 of variant G “are considered a stev” 59 (Figure 10). These two stanzas are written together here in order to make the nystev form apparent visually.

Figure 10. Rikard Berge’s example of a nystev, stanzas 2 and 3 from the 2-line ballad “Ride brudeferd”, variant G.

1. Eg sille full laane deg fòlen min den snare visste eg du styrde ‘o som eg va’ den sjave.
2. Aa naar me kjem’e paa kjyrkjebakkjen sò spør’e adde kven eig’e Blakkjen.3. Aa du skò svara som du va’ vrei: “Aa Blakkjen fann me paa villan hei”.
4. Men du skò svara som du va’ i dròumo: “Den eig’e Blakkjen den styrer tòumo”.

This is indeed nystev form. In “Liv og Leikar” we find the following text which Knut Heddi describes as stev, from Inger Bjugsdatter60 (Figure 11, left). Its last two lines are compared with a 2-line stanza from a version of “Ride brudeferd” 61 (Figure 11, right).

Figure 11. Comparison of the same words used as masculine end-rhyme and feminine end-rhyme in variants of the same song, “Ride brudeferd”.

Eg svara det, som eg var i vreie, Tru dæ kje det same kve blakken eige? Eg svara det som eg var i draum. So svora dú det, som dú va i draumen:
Tru blakken kjem kje a vidde haug? “Den blakken e komen ó vidde haugjen”.

The last two lines in Figure 11 are very similar, except for “draum”, and “haug” (left) being “draumen” and “haugen” (right), showing how suffixes, present or absent in the same words, can create masculine or feminine rhyme. Norwegian words adapt easily in this manner. Another such example from the same ballad is shown in Figure 12.

Figure 12. The same words in masculine rhyme (left) from Berge’s variant G stanza 3, and feminine rhyme (right) from the last stanza of his variant F.

Aa du skò svara som du va’ vrei,Daa vi mi svore som mi vaare i Vreio:
Aa Blakkjen fann me paa villan hei.Blakkjen e kaamen ó vidde Heio.

Side: 168

The examples of words which can be either one or two syllables suggest that word endings can be changed to fit the nystev rhyme scheme, whereas the nystev form, first a feminine and then a masculine couplet, remains constant. In recent tradition, this stable feature has no exceptions, except for one vagrant which Berge found and called a “deviant stev”. 64

Figure 13. Rikard Berge’s “deviant stev”, with two feminine-rhyme couplets.

Naar eg kjem’e paa kjørkjebakkjen, sò spør dei meg hòken eige Blakkjen, Sò svarar eg som eg stend i dròumo: “Han eig’e Blakkjen som hell i tòumo”.

He has noted here that another kvedar sings the words “dròum” and “tòum”. This flexibility in the presence or lack of rhyme word endings is worth noting.

Section 1 explained that the common gamlestev has the same meter as the common ballad. If the refrains can be deleted from the four-lined gamlestev form and the two gamlestev melded into a nystev, then one could hypothesize the same process for the 4-line stanza of the common ballad, and this has been done.65 The origin of nystev from gamlestev might then be considered as essentially the same as Richard Berge’s theory of nystev originating with two two-line ballad stanzas. His examples, however, show how easily feminine and masculine rhymes are interchangeable.

In conclusion, findings from musicological study of stev support Berge’s “deviant stev” and show that this stanza is perhaps not so wayward after all. His examples of ballad couplet stanzas with interchangeable masculine and feminine rhyme and his deviant stev are reminiscent of “Málsháttakvæ∂i” and its seemingly strange combinations of masculine and feminine rhyme couplets.

Musicological study may provide a missing link because it shows that the traditional folksinger holds out the accented syllable in the last word of a feminine end-rhyme couplet, and in fact can hold this out quite long as if to stress the end of the sentence. The following syllable, unaccented in feminine rhyme, is in contrast often quite short (and of course never foot-tapped since it is not a word accent). Furthermore, a nystev melody stretches over the first couplet and then is repeated with the text of the second couplet. The second couplet ends with masculine rhyme in the normal nystev. Thus, the only difference in form between the two couplets is the unstressed syllable in the feminine rhyme. An important characteristic of nystev is that musically the short unaccented syllable of the last word in the feminine rhyme couplet is always on the same musical pitch as its preceding word accent. Thus one can drop this note without great ado to accommodate a masculine rhyme or add it for a feminine rhyme. Whether this additional note is present or absent does not influence the predictable foot-tapped points, which are the structure of the melody. Nystev form has appeared to be so unchangeable, but a conclusion seems to

Side: 169

be that it has an innate capability to easily accommodate change of rhyme ending in couplets, which is shown by feminine and masculine rhymed couplets sharing the same melody. This then is a fulcrum for flexibility in nystev, and one which appears to have been in full function earlier, allowing different combinations of the feminine and masculine rhyme couplets, so clearly exemplified in Málsháttakvæ∂i”.

The texts of the typical ballad stanzas Berge presents suggest that the two-pulse dipod is as feasible in these as in nystev. Hence from this standpoint too there is likeness to the nystev form. If the nystev developed from gamlestev and/or the two-lined ballad, this would have had to happen before Málsháttakvæ∂i, which is thought to be from the early 1200’s. It is perhaps not unreasonable to propose that gamlestev, ballads, and also nystev were in use in folk tradition in Norway before the 1200’s. Whatever the age of nystev, logically gamlestev stanzas were even older.

3. The Refrain Stev and Old Norse Stef

Nystev and gamlestev are often sung alternately, particularly in competitive stev-duels whereby the winner is the kvedar who brings the other participants to silence. Such alternate singing-reciting may latch and bind one stanza with another, first into stanza pairs, and then sometimes into long chains of stev.66 The last section closed with “omit the refrains, the nystev remains”. Although omitting refrains could create nystev, this very refrain in ancient Norwegian ballads, gamlestev and “Draumkvedet” was also paradoxically called a stev. There is also another type of refrain, the Old Norse Skaldic stef, [pronounced “stev” and sometimes spelled with “v” in today’s Norwegian]. Both of these types of refrains bind stanzas together but differently: the ballad stev binds lines of a stanza together as well as one stanza with another. The Old Norse stef was used sparingly to bind bundles of strophes together. Thus both are relevant in connection with latching and binding, but will be discussed only briefly.

The Old Norse stef, 67 a very special type of refrain appearing chiefly in a particular type of Skaldic poem, seems to have bound bundles of stanzas together and thus shaped a poem into a respectable and coherent unit. A poem with a stef was called a drápa, and was worthy of honoring kings and noblemen, as opposed to a stef-less poem, called a flokk, a lowly creation unworthy of a man of rank.68 Occasionally a poem employing the stef is not a drápa. An example is”Málsháttakvæ∂i”, from the early 1200’s, introduced in section 1.69 To paraphrase Möbius, the first part of stanza 11 in “Málsháttakvæ∂i” announces the stef, and therewith the opening of the middle section of the poem. The author of the poem expresses that this particular poem has need of a stef all the more because of the peculiar content consisting of an arbitrary number of sayings:

The poem would be endangered if it lacked the binding agent, which indeed the stef is; and would degenerate to a mere list of rhymes, ... the in

Side: 170

dividual lines would otherwise, without any putty, fall apart like small stones [jewels, jpe].70

The type of poetry in “Málsháttakvæ∂i” Möbius calls”áttmælt”, that is, having “8 sentences,71 a sentence for each line”. A “line” here equals 2 dipods and is thus the equivalent of the 2 dipod line in nystev. A nystev consists of two complete sentences,72 and we may note too that the stev melody is as long as the one sentence and then is repeated with the text of sentence two. Occasionally a nystev may have a sentence in each of its 4 -lines, making its form comparable to the one-sentence-per-line scheme found in the 4-line stef here. This particular stef, declares Möbius, is “fitting nothing less than the beginning of the Tenth Century”.73 The Norwegian nystev may already have been in full bloom when Málsháttakvæ∂i and its stef came into being.74

If a stef ennobled a poem, and nystev form was used here as a stef, one may wonder if the nystev form itself possessed status. Also the contents of the stef must have been of value in order to increase the worth of a poem so easily.

Stev as Ballad Refrain

Concerning the concept of “refrain”, Olav Bø remarked that one needed a song form with several or many stanzas before there could be such a thing as a refrain, which usually keeps the same text and melody in contrast to the song text which varies from stanza to stanza. One or more lines or even a whole stanza may be used as a refrain. Hulda Garborg describes the origin and forms of the refrain,75 (presented in Norwegian and translation in Figure 14).

Figure 14. Hulda Garborg’s description of the refrain stev in epic song (left) and translation (right).

De lyriske danseviser smeltet nemlig i Nordlandene sammen med de lange episke kvad som improvisertes i dansen og levet påfolketungen i kortere eller længere tid som “folkeviser”. Sammensmeltningen foregikk på den maate, at strofer av den lyriske vise blev staaende igjen som “stev”(omkvæd) i den episke vise. In the northern countries, the lyric dance- songs fused together with the epic lays which were improvised in dance and which lived in tradition for a shorter or longer time as “folk songs” [ballads]. The fusion resulted in stanzas of the lyrical song remaining as refrains in the epic song.
Og der var ofte både “millomstev” og “etterstev”, eller “millomsleng” og “ettersleng”, som det ogsaa kaldtes. F.ex. i visen om “Liti Kersti”.And there were often both in-burden (millomstev, millomsleng), and end-burden (etterstev, ettersleng), for example in the song about “Little Kersti”.

Side: 171

Below is her example of the ballad “Liti Kersti” and the types of refrains. 76

Figure 15. Hulda Garborg’s description of types of refrains in “Liti Kersti”.

Liti Kersti ho var seg so litit eit viv.
Bronfolen løyper lett; Omkvæd [refrain], (millomstev,       millomsleng [in-burden])
ho kunde ikkje raade sitt unge liv,
med det regner og det blæs. Omkvæd [refrain], (etterstev,       ettersleng [end-burden])
For nordan under fjøllo            djupt under hello,            der leikar det. Omkvæd [refrain]

Sometimes the in-burden and end-burden have belonged together but have been split, in which case the two lines are called a kløyvastev or “split-stev”. 77 Refrains may appear senseless and not relevant to song contents. Landstad, however, pointed out in 1853 that a refrain is closely connected to the contents of its song.78

One could lessen confusion about the term “stev” by not using it as a general term for five different genres. Instead, one could 1) use the word omkved consistently for referring to the refrain in the old ballads and gamlestev, 2) retain the “f” spelling for the Old Norse stef, 3) and refer to nystev, gamlestev, and slåttestev as such.

Concluding Remarks

Researchers’ theories about stev were difficult to assess properly before knowledge of foot-tapping and the resulting stev accentual theory based on the two-pulse. Nystev and gamlestev demonstrate interesting phenomena in addition to the two-pulse. Here is the only known documentation of the function of rhythm in connection with use of front-rhyme and end-rhyme. The two-pulse is stable, durable, predictable, and yet able to adapt easily to certain variations in rhythm and rhyme. One can query why the two-pulse is so hardy. An answer seems to be that rhyme and rhythm intensify the asymmetrical two-pulse. Then too, the two-pulse whether the stev dipod or in the Old Norse visuor∂, is not an end in itself, but functions as a carrier of concise phrases of poetry and thoughts. The two-pulse foundation and the nystev and gamlestev rhythmic form can be likened to a ship that could carry diverse contents, – a veritable “Ski∂bla∂nir”.79