This week On the Reading Rest I have an excavation report in Swedish:

Stenbäck & al. 2012. Niklas Stenbäck, Marcus Eriksson, Michel Guinard and Roger Wikell, Stenålder vid Påljungshage. En sörmländsk lokal med nedslag från tidig- och mellanneolitikum—Stone Age at ’Paul Young’s Pen’. A Site in Södermanland visited in the Early and Middle Neolithic. [SAU rapport 2010:8, pp 184]. Uppsala 2012. (cf.   )

Once again, a site, excavated because society needs to transform itself and exploit its assets, reveals a micro history easy to grasp. The excavation results, moreover, makes it obvious that a simplistic answer to a wrongly posed but popular Stone Age question, cannot survive the complexity of Prehistory and the straightforward contexts which it gave rise to.

Let’s start with trivia: the dimension and size of log boats is determined largely by the logs from which they are made. Indeed, and for two reasons this means that in the future, Swedish log boats will be very narrow and little used.
(1) Today, when it comes to the exploitation of its assets, Sweden is turning woods into squared production surfaces where firs are planted in an even pattern and expected to grow for 70 years only before they are harvested. Woodlands are rapidly turned into monoculture fir fields and seventy year old firs are still slender trees – useless when it comes to log boats.
(2) All wood production surfaces are eventually ploughed and planted with fir, which means that soon there will be no suitable pines either.

In bygone days when woods were dense with a variety of vegetation and trees grew straight for non-cultural rather than mono-cultural reasons, biological diversity ruled. Trees grew tall and their wide trunks made wonderful log boats or canoes. Some prehistoric boats are impressive.

The rapid loss of biodiversity and ecosystems in Sweden is not trivial, and in most of the world, not least in Sweden, it has become too late to hug the trees of a primeval forest. Instead the loss of everything but firs, young or young adult trees, has resulted in fir field monuments to narrow-minded greed and irrationality. We should accept neither the loss nor the monuments.

During the Stone Age at Påljungshage—Paul Young’s Pen—they didn’t build log boats, but they used them to go there from the mainland in Early Neolithic times. Owing to the high water levels Paul Young’s Pen was not a pen at all, but a protected coast facing south on a small rocky island in the outer archipelago with a good landing place. The landing place was a sandy beach cleared from stones, perfect for canoes.

Those days are behind us. Because of the shore displacement starting when the Scandinavian inland ice and its weight melted away, the water disappeared, the beach became a meadow and eventually a marginal pen before the highway was constructed and the agricultural land replaced by brushwood in a narrow and badly drained corridor between the road bank and the once Neolithic beach. Today, having blown away the bedrock that once formed the top of the island and crushed debris and blasted stone into valuable gravel, the place has become a shopping mall as commonplace to the outskirts of a modern town as ever a beach site in the Neolithic archipelago.

But the excavations conducted before the construction of the shopping centre were excellent. And for once money wasn’t a problem – in part because of all the gravel.

The Stone Age site was a nuisance to the exploiters and not much to bother about, but the Bronze Age – Early Iron Age cemetery on the very top of the former island, once monumentally marking the border zone of the Bronze and Early Iron Age settlement north of Paul Young’s Pen, sat on valuable bedrock and had to go. The whole scene was potent modern exploitation: an end with a bang to the lee and protected beaches of the outer Stone Age archipelago.

Although Paul Young’s Pen was been visited several times in prehistory, the visits in the Early Neolithic, 4th millennium BCE, meet the eye. Careful excavation and interpretation of archaeological variables as well as a number of sophisticated analyses – wear on stone artefacts, lipids in pot shards, thin sections of ceramic ware and analysis of decoration, analysis of macro fossils and diatoms as well as osteology and 14C dates, tell us the following about the Early Neolithic site: Facing the south there were three small sandy shelves close to the shore. They were used simultaneously for similar yet somewhat different purposes. On the eastern shelf the use of fire was important and so was cooking and quarts/quartzite napping. On the western shelf grinding played a prominent role. The central shelf was the better landing site and the arena of the commonplace of daily life when the level of the water was c. 30 metre above today’s sea level. The western and central shelf had direct contact with the water, but the eastern one was isolated.

The central shelf represents the dwelling area, the eastern shelf a rather smoky specialization related to processing, and the western a ‘non-smoking’ area with an element of craftsmanship. The report is built around the systematic presentation and interpretation of a number of artefact categories in view of distribution and density. There are a number of categories: quarts, quartzite, flint, whinstone, rock type, slate, mica slate, sandstone, ceramics, burnt bone, fire cracked stone, lipids and diatoms and they give the reader a typical report insight into the settlement. Moreover, they suggest the mapping of these variables according to their presence in different square metres thus mapping the complexity to the usage of the site. We may do this by representing each category with a certain degree of opacity. If we do so, the colour of the square metre will deepen with the number of variables represented in it. Thus, an opaque red represents the most complex square metres, and clusters small centres of diverse activities.

Obviously Neolithic man used to nap quarts when he produced his tools, and fire cracked stones were endemic to the use of fire. We may therefore expect these categories to be constantly present on a settlement site and weigh them less important than ceramic, slate or flint, which must have been imported to the site by means of more complicated networks. Burnt bones too are significant because they mirror division of labour.

If we map the site in this way, the central and western shelves are characterized by small-scale clusters. At the eastern shelf, repeated activities blurred the complexity—the small-scale clusters—over a relatively large area in the western side of the shelf. Eventually, if the site had been continuously occupied for hundreds of years the small-scale clustering on all the shelves would have disappeared. Since this development didn’t happened we may Suggest that the site was visited at intervals a relatively small number of times. Eventually the coastline fell below 30 m above sea level, but that didn’t prompt Prehistoric man to follow the displaced shore. On the contrary, the few times the area was revisited people sat down on the dry and sandy sheltered shelves.

Bones, lipids, diatoms and macro fossils allow us to infer some basic facts about the Early Neolithic diet. What animals were eaten? What food was cooked and stored? What algae were trapped in the grinding stones? By chance, what plants were charred? The answer is simple: Those who ate at Paul Young’s Pen were farmers as well as hunters. They brought grain and a large grinding stone from their sweet water inland settlement out into the archipelago where diatoms that thrive in brackish water were grinded into whetstones, but not into the grinding stone. In their pots, people cooked vegetables and ruminants, as well as other terrestrial and maritime mammals. They ate wheat and barley as well as fish, seal, pig, and sheep/goat. It looks as if they mostly did this on the eastern shelf, but in reality this was where their use of fire included burning their garbage.

And when did they go there? Two times a year – before and after the farming season. They came in late winter/early spring for the seal hunt – no doubt the brutal hunt that became traditional – and returned in the autumn to fish. Topography, the year cycle and their own traditions seem to have guided their occasional visits during a 500-year period mainly in the early part of the 4th century BCE.

One thing they didn’t do in the archipelago was reading ‘Science’ and that was probably just as well since if they had read Vol 336 no. 6080 pp 466-469 and the comment pp 400-401, they would have felt as cross, neglected and misunderstood as ever a First Nations people.

The authors of the article Origins and Genetic Legacy of Neolithic Farmers and Hunter-Gatherers in Europe (pp 466-469) end up cautiously stating that ‘[o]ur results suggest that migration from southern Europe catalyzed the spread of agriculture and that admixture in the wake of this expansion eventually shaped the genomic landscape of modern-day Europe.’ In News & Analysis, under the heading Ancient Migrants Brought Farming Way of Life to Europe (pp 400-401), this nevertheless becomes ‘evidence that farmers personally took the technology across Europe, and that the first farmers of chilly northern Europe came from the continent’s sunny Mediterranean south’. The rest is www (9,960 hits at Google for the phrase “Ancient Swedish Farmer came from the Mediterranean”) and ‘Science’-true only.

Yet everything is based on a splendid analysis of ancient DNA from humans who died c. 5000 years ago, one of whom may personally have come from southern Europe. Yes, 5000 years ago, i.e. 1000 years after the first peasant-hunter-gatherers at Paul Young’s Pen dropped dead. There is of course no doubt that the forefathers of the people who visited Paul Young’s Pen could have been Europeans and that Europeans continued to find their way into the Scandinavia blind alley, but there is no reason to suggest that coming from somewhere a thousand years too late made anybody the first farmer and certainly not either farmer or hunter-gatherer. No wonder First Nations people usually don’t accept anachronisms.

It is not inconceivable that immigrants in the 3rd millennium BCE ‘catalyzed the spread of agriculture’ and many other things such as cabbages and kings, but that might and might not have happened any time before and after. None the less, well into modern times, coastland and archipelago agriculture was often a node in balanced networking with other nodes such as hunting, fishing and gathering. Such diversity is often a reasonable way to adapt to an environment and make it a human landscape without completely destroying it. Trivial but true.


This week on the reading rest I have a book from 2009, Versatility in Versification. Multidisciplinary Approaches to Metrics, edited by Tonya Kim Dewey and Frog. I read one of the contributions.

Schulte, Michael. 2009. Early runic ‘metrical’ inscriptions—How metrical are they? In Tonya Kim Dewey and Frog (eds) Versification. Multidisciplinary Approaches to Metrics. [Berkeley Insights in Linguistics and Semiotics vol 74]. Peter Lang New York etc. Pp 3-22.

Poetry and intoxication go back a long way exploiting each other. Expression doesn’t bother intoxication as long as it is deviant, irrespective of ‘it’ being expression or intoxication. But to poetry, expression is everything. That’s why poetry, aided by its craftsmen, may use intoxication as a method finding its limits of expression—simultaneously defining a room of its own and an expression to go with it. In Norse mythology the myth about Oðinn and the scaldic mead is very much to the point. Favorably intoxicated, we write poetry that would have been fine were it not for the fact that the scaldic mead couldn’t be ushered into our world without partly being corrupted—divided as it happened into good and bad. This mead still intoxicates and makes us diligent poets, but those who have drunk from the bad, produce bad poetry. Most of us believe we know who drank what, but then again we might be intoxicated. Writing about Norse or Early Runic poetry one should bear this in mind.

 Michael Schulte’s (MS’s) article belongs to a genre so academic and serious that today it must be in need of a Latin label such as Timor Carminis—The Dread of Verse. It’s all in timor spreading between fear and awe. To most, even though they don’d remember the tale about the scaldic mead, verse may be anything between awfully good or dreadfully bad (however, not worse that the poetry of Paula Nancy Millstone Jennings) and we can live with it. But some take it upon them to defend a ‘high’ style of poetry as indeed Poetry, against everyday speech (MS p.4) lest Oðinn, or anyone awe-inspiring authorized to judge the quality of the scaldic mead, should find you short of understanding the limits of poetic flexibility when it comes to meter. Since poetry, owing to Timor Carminis, must be defended against gute Leute aber schlechte Musikanten—‘good people, but bad musicians’, metrical consciousness must not be introduced or reckoned with before we can be absolutely sure that strict and conventional formal rules of versification are followed—and understood.

Obsession with unintentional verse is typical of this academic genre because such expressions may look like perfectly acceptable more or less free verse, which cannot on formal or technical grounds be deemed either intentional or unintentional. This is the reason why MS’s subtitle – how metrical are they? reads: are they metric enough to my taste? Instead of speaking of Early Runic verse as a fluffy matter, at best a consensus-driven concept, resulting in an unspecific corpus of good and bad verse that didn’t bother Oðinn a bit, MS proclaims that in his article Early Runic inscriptions will be classed into groups (six as it were) and assessed in terms of their metrical features (MS p.4). In the end – but actually from the very beginning – this classification represents a ‘metrical state of indeterminacy’ (MS p.17) in which prehistoric man composes his expressions. Having read MS’s article 0ne understands that if prehistoric man had bothered to study Andreas Heusler’s Deutsche Versgeschichte—A History of German Verse (1925), indeterminacy would have not have existed.

Indeterminacy nevertheless, is hopeful, since it indicates that little by little prehistoric man evolves, stops being vague, pulls himself together and starts writing the ‘high’ style that to MS is a precondition for writing verse. Heusler, by the way, didn’t deny that less formal poetry, or poetry governed by melody or song were legitimate means of poetic expression. 

Nevertheless, MS (p. 17f) quotes Andreas Heusler, who has been with the author from the very beginning of the article (MS p.3), and goes on to say that he fully subscribes to Kari Ellen Gade’s doubts about the alliterative poetic status of the earliest inscriptions. Yet he seems unable to convey any doubts at all – in MS’s opinion they are not metric enough. Needless to say MS doesn’t doubt the existence of his six classes of indeterminate inscriptions. Doubting these classes is a matter of taste and that cannot be discussed. In Andreas Heusler’s work we don’t come across any six classes of indeterminate expressions. Heusler knows that there is a lot one must accepted when it comes to poetry.


MS’s critique is based on three criteria, i.e. lack of: (1) syllable-count; (2) quantity-sensitivity (see box above) and (3) the structural complexity of the alliterative scheme (MS p.3). The first criterion dismisses thousands of lines in epic and lyrical poems in which the long lines vary considerably without observing any strict patterns. None the less, there are usually more syllables in descriptive epic long lines than in those of direct poetic speech. The example from On the Reading Rest April 30, the episode containing Wealhtheow’s speech, is telling. The second criterion is somewhat esoteric and again dismisses a great number of lines for no apparent reason. The last criterion compares to arbitrariness, since when in essence, is structure and complexity—structural and complex enough? Naturally, MS points to the fact that alliterations feature in everyday speech and that there are complex alliterative patterns in prose. Such patterns are basic, a backdrop reflected in and reflecting poetry, but to MS basic is not enough and perhaps disqualifying (MS p.4). In the end metrical problems boil down to the fact that to MS’s mind Early Runic inscriptions are not, metrically speaking, strictly odd enough. Predictably MS finds no verses in these inscriptions.

The only specific reason why we discuss whether those who wrote Early Runic considered expressing themselves in some kind of meter, is to investigate whether or not some of the meters that we know from later Eddaic poems had forerunners—or to rephrase: was there a poetic commitment when the old fools of the Roman Iron Age were overthrown, (On the Reading Rest, Hogganvik) together with some of their rune stones, by the new elite who didn’t favour literacy. We need to know this in order to answer the following question in the affirmative: did the long lines and full lines, needed to construct Fornyrðislag and Lióðaháttr, exist before we read them in the Edda? If these two lines existed, they would have allowed Early Iron Age man to compose epic and lyric verse and develop the oral poetry behind the Eddaic poems.

If we can find these line patterns in Early Runic inscription as well as in the Poetic Edda or in epic poems, then that is sufficient to consider them examples of prosody. The affirmative answer is important because it indicates that upper class hall as an archeological phenomenon and formal poetry were contemporary. Probably poetry is much older, but being a mid-millennium hall owner with no access to formal poetry, with no scop in his hall, is comparable to a theater owner with no plays and no actors.

A long line must consist of two half lines each with two stressed syllables. Between the half lines there must be a caesura—a cut. This is a basic way of expressing oneself in Germanic languages and we meet it daily e.g. in newspapers. To The Independent on the www April 21 this structure came in handy at least three times:
Real men want to talk about sex we need to start listening
Essex appeal: the only way is Amy Child
Still the caesura may not be a complete break and perhaps forced:
The Weatherman caught in a media storm (and easily come before or after in)
and exactly which syllables to stress is not always that clear either:
Real men want to talk about sex we need to start listening
Real men want to talk about sex we need to start listening.
Since meaning changes radically with the stress:
Real men want to talk about sex we need to start listening
there is a prosodic point in marking out the stressed syllables. Conventionally that is done by alliteration, but alliterating on all four stressed syllables is considered heavy handed: Gibbon and Gareth, good-looking guys. Preferably therefore the fourth syllable should not alliterate and that makes: Gibbon and Gareth, good-looking friends a suitable long line. Alliteration on each side of the caesura is easy to hear and thus the pattern Susan and Gareth, good-looking friends is common, but even Gareth and Susan, good-looking friends often indicating two related expressions can be found because usually, having started with a stressed syllable, we expect alliteration in the first stressed syllable after the caesura. The number of unstressed syllables is not equally important although they must neither be too few nor too many.

Bearing this in mind there are several Early Runic long lines:

Ek Hlewagastiz Holtijaz     horna tawido      ´a a |a x       8+5     DR 12 †U [0]
Ek Wagigaz erilaz     Agilamundon                  ´x a |a x       7+5      N KJ69 U
Þrijoz dohtriz    dāliðun arbija*                         ´
x a |a x      4+6      N KJ72 U
Fahiðu wil-ald     wigaz ek erilaz                        x a |a x       5+6      DR IK241,1 $U
ek Wiwaz after     Woðuriðe                                a x |a x       5+4      N KJ72 U
ulþuþewaz     ni wajemariz                              a x |´a    4+5       DR 7 $U
Hadulaikaz     ek Hagustaldaz                            a x |´a x     4+5       N KJ75 U
Wurte runoz     an walha kurne                          a x |´a x     4+5      DR IK184 U
witanda-halaiban     worahto [rūnō]z              a x |a x       6+5      N KJ72 U
Haha skaþi     haþu ligi                                         a x |a x      4+4       N KJ50 $U

* ð and þ can be taken to repesent voiced and unvoiced ’th’ – ’these’ and ’moth’.

Probably there are many more, but this sample is enough to show that long lines were composed, but also that the gentle alliterative pattern, a x |a x, in short lines, often with an optional upbeat, i.e. anacrusis, in the second half line, a x |´a x, was common. Short lines are common in the Poetic Edda too, but the gentle alliterative pattern (which is not), the length of the lines, the use of anacrusis and the lack of lines ending in a stressed one-syllable word (a x |a x) has nothing to do with development. These are stylistic choices perhaps related to the fact that the Early Runic lines are short inscriptions on objects rather than lines in long poems [1].

This said, some general linguistic differences between Early Runic and Eddaic long lines meet the eye. On average the syllable balance in the long lines is 4.5+4.4 in the Eddaic lines and 4.8+5.0 in the Early Runic. In balance they are thus rather similar. Nevertheless there is a difference in length of almost one syllable between the long lines, 8.9 and 9.8 syllables respectively. This is hardly the result of stylistic preference, inasmuch as it may conveniently be explained by the general change in Germanic languages in which between say 400 and 800 CE the number of syllables per word tends to drop. Early Runic with words such as daliðun, erilaz, fahiðu, halaiban, tawido, Wagigaz, witanda, worahto, in which the first volve tends to be long and the word accent grave, obviously has a more gentle character than Eddaic Norse where that kind of words is relatively speaking rare.


If we compare Early Runic prose, of which there is virtually nothing left, with long lines, we may argue that the latter try to restrict the number of syllables. Compared to one of the few pure-prose phrases Frarawadaz ana hahai is slaginas (U 877 U ) in which the stressed-unstressed syllable relation is 3:9, the typical long line relation, 4:5, indicates a long line composition that avoids unstressed syllables. In this stilistic endeavour, poetry is leading linguistic change and/or benefitting from it. Be this as it may, a conscious and gentle composition of long lines is typical of Early Runic compared to Eddaic lines. The meter is the same, the style differs but the outcome, Eddaic verse, is not surprising.

This brings us to the full line, i.e. a line with three stressed syllables. This too is a well-known structure implying that a statement, a composition otherwise running in fractions with two stressed syllables, is coming to an end. To The Independent on the www April 22 this structure came in handy: Tens of thousands of fun runners and amateur athletes set off in bright sunshine as the 32nd London Marathon got under way today. (end of paragraph). Probably, the most well-known example is Jane Austin’s ‘… …, must be in want of a wife’, which has a reasonable alliterative pattern, rather than ‘independent’ rhyme, helping the sentence to come to an end (2+2+3 stresses, end of paragraph).


A strophe in the Lióðaháttr comes to an end in this way and there is no point in looking for freestanding full lines, but well in finding the combination long line + full line.

A number of texts fit the pattern:
Haha skaþi     haþu ligi                       Ll     N KJ50 $U
wate hali hino horna*                          Fl

*There is no stress on ‘hino’ since the ‘i’ is short – a case of quantity-sensitivity, no less.

Þrijoz dohtriz    daliðun arbija          Ll     N KJ72 U
asijostez arbijane                                 

Hadulaikaz     ek Hagustaldaz          Ll    N KJ75 U
hlaiwido mahu minino                       

Although there are independent alliterations in the full line, there are also alliterative links from the long line to the full line. In some patterns these links are the only alliterative characteristic of the full line:

Wurte runoz     an walha kurne        Ll    DR IK184 U
Heldaz Kunimundiu                             

ek Wiwaz after     Woðuriðe                    Ll    N KJ72 U
itanda-halaiban     worahto [rūnō]z  Ll
þez Woðuriðe staina                                  

Binding from the long line to the full line in this way is most uncommon in the Poetic Edda, but nevertheless there are some ten examples out of c. 10,000 possible ones. This per mil is linked to the uneducated, e.g. serfs, and perhaps comic (probably old-fashioned) in their Eddaic irregularity.

This proto Lióðaháttr style ties in well with the gentleness of the long line patterns and it seems significant that the three examples that are straightforward Eddaic come from the western part of Norway and the roots of the Eddaic tradition. The stressed-unstressed relation in the full lines is 4:6, i.e. slightly less syllable-economic than in the long lines, but more economic than prose. Since full lines bring a poetic statement to an end, often in a kind of ritardando, this is expected.

One might write all kinds of verses in a smooth, heavy, bombastic, light or gentle non-prosaic style, and there is no reason to deny the poets of the fifth and six century CE the right to compose their lines and verses in their own write compared to later traditions, which they forego. Given second thoughts we may wonder how much oral poetry was never written down in the non-literacy centuries of the Pre Carolingian Iron Age when a new social elite established itself in Scandinavia.


[0] In Samnordisk runtextdatabas, which can be downloaded from this and the following call numbers will lead to the inscriptions. If you look up the name of the inscription in the database you may continue to where you will find more references under each name.

[1] Two Eddaic examples comparable to Early Runic long lines

Ein nam þeira     Egil at verja,                                         a x │a x          4+5
fögr mær fira,     faðmi ljósum;                                       a a │a x          4+4
önnur var Svanhvít,     svanfjaðrar d*,                     x a │a x         5+4
en in þriðja     þeira systir                                                x a │a x          4+4
varði hvítan     háls Völundar.                                        x a │a x          4+4
Sátu síðan     sjö vetr at þat                                            a a │a x         4+4
en inn átta     allan þráðu                                                a a │a x           4+4
en inn níunda     nauðr um skilði;                                 x a │a x           5+4
meyjar fýstusk     á myrkvan v,                                  a x │´a x       4+4
Alvitr unga,     örlög drýgja                                            a a │a x            4+4
Hljóðs bið ek allar     helgar kindir,                             a x │a x            5+4
meiri ok minni     mögu Heimdallar;                          a a │a x            5+5
viltu, at ek, Valföðr!     vel framtelja                            a a │a x            6+4
forn spjöll fíra,     þau er fremst um ma                     a a │´´a x      4+5
Ek man jötna     ár um borna,                                       a x │a x           4+4
þá er forðum     mik fœdda höfðu;                               x a │´a x         4+5
níu man ek heima,     níu íviði,                                     a x │a x           5+4
mjötvið mœran     fyr mold neðan                               a a │´a x         4+4
Ár var alda     þar er Ýmir bygði,                                  a a │´´a x       4+6
vara sandr né sær     né svalar unnir                      ´´a a │´a x        5+5
*dró and so on indicates that the last stress is on the last one-syllable word.