Locked Inside a Nave Since the Fourth Century CE – Part I

25 August, 2014

This week On the Reading Rest I have an article on the rune stone from Hogganvik, Mandal, in Southwest Norway. This stone was mentioned OtRR 9 January, 2012, but there is no point in looking it up since in the present entry the perspective from 2012 has simply been developed.

ILLU 01Schultze, Michael. 2013. The Norwegian Hogganvik Stone as an Emblem of Social Status and Identity. Journal of the North Atlantic 4:120-128. Acronym: MiSCHu.

This article is extremely informative, 80 different references in 7 pages, but in the end, which is called Principal Conclusion, the stone and its inscription becomes a general and quite unspecific reflection of the commonplace when it comes to the Late Roman Iron Age (LRIA 150-350 CE): “… it reflects a hierarchical society.” (MiSCHu:125). Since ‘status’ and ‘individuality’ is not mentioned in this first sentence the rest of the conclusion links the sponsor to these concepts. The Principal Conclusions may thus be summarized as the first sentence in a BA exam essay on the LRIA in south Scandinavia: “In the Late Roman Iron Age South Scandinavia was a hierarchical society with marked symbols of power and social identity emphasizing status and individuality”. A peevish teacher will of course stick in ‘eventually’ before ‘individuality’, but otherwise not bother much because this introductory phrase has been the archaeological wisdom for the last 30-40 years.

How come that such an intricate inscription and deep knowledge lead us to no more that the commonplace?

Some runologists have had trouble understanding that early rune stones do not always, just because it is possible, commemorate a deceased person. Already when Sven B F Jansson discussed the stone from Järsberg in Värmland, where excavations eventually showed that the stone was erected on a small natural hillock and not on or next to a grave, he noted this peculiar and century-old runological bias (1). Early rune stones are never typical formulaic headstones or memorials and they differ significantly from contemporary Roman headstones or the formulaic Scandinavian memorial rune stones, which date to the Carolingian Iron Age. The formulaic nucleus of the latter consists of three parts: The Sponsor – erected the monument after – The Deceased. One might have expected that in Scandinavia, Roman Iron Age man, given his possible connections with the Roman army, would have adopted some sort of Roman headstone formula. But that is not the case. There are one or two possible parallels in design, such as Möjbrostenen, but none in expressions. In fact there is no reason to believe that a text commemorates a deceased and names a sponsor if the text is not unambiguous.


Most often it takes archaeology to figure out the context of a rune stone. The Hogganvik inscription is a case in point. For a short while, before the archaeological excavation, this rune stone was a headstone standing on a grave. After the excavation it was not, but to runologists its alleged commemorative character featuring the deceased and the sponsor prevailed. In all fairness it must be said that MaSCHu is a significant step away from this preunderstanding, but it nevertheless continuous to colour his reading and interpretation of the inscription (2). Since there are non-lexical runes in the inscription, obscure concepts such as magic and/or cipher must also be allowed a priori to structure the text in a runological discourse. (cf. MiSCHu:123)

Demonstrating these difficulties understanding the topographical and monumental context of the stone, Hogganvik is a sad modern case in point. Based on Glørstad et al. 2011, commented upon OtRR 9 January, 2012 (3), is was easy to sum up the context: Today, we know so much that we can approach the stone as it would quite likely have been approached around 400 AD. It stood as a poster and if we were not locals we would probably have approached it from the South following a road going North through or along a small cemetery. We would have passed by the most prominent grave and then just before the road turned left and down in a slope or between two hills we would have stood in front of the stone.

ILLU 03The stone, moreover, stood in a monument of its own and to a great extent the layout of the inscription and thus the shape of the stone was reflected in the contours of this monument laid out in kerbstones. The position of the rune stone itself within the monument corresponds to the two short lines. The stone was chosen and cut; the kerbstones laid out, and the stone put in its place.

To sum up: the longest part of the inscription frames the two shorter parts. The longest is a proportional map of the outline of the interior of the monument in which the stone was erected. The two shorter parts are situated more or less where the stone stood. Thus the layout of the inscription is a map of the monument. The material context is manifest and not without material agency.

Since the design of the stone consciously parallels that of the monument, the layout of the text is a complement to the layout of the monument. Because this fact is not ‘textual’ in the pre-1980 still frequent runological usage of the concept ‘textual’, some runologists feel free to overlook it. Probably, they lose interest when archaeologists produce contexts, which runologists would not already have recognized in the text. Thus the idea of the headstone is not given up as long as it can be slightly moderated. Consequently, the deceased and the sponsor are still live & kicking — that is agency-wise.

It is commonly agreed, but no longer by MiSCHu, that the longest part, the broken line in which the sections are read towards each other, made up one series of runes and an expression which contained a series of 14 non-lexical runes situated mainly in the second part of the line. The broken line is read it in this way:


and the result becomes: kelbaþewas⁞stainaz⁞aaasrpkfaarpaa⁞inananaboz (line Z)

In this part of the inscription the text is divided into smaller sections by means of vertical dots (), which separate the non-lexical runes from the lexical, and stainaz from kelbaþewas. It would have been easy to organize the inscription in three lines, read boustrophedon – as the oxen ploughs: from right to left to right to left – if clarity had been an issue:


Since this solution was not chosen, simple clarity probably did not guide the carver (4).

On the one hand there is no doubt that the correspondence between the inscription and the shape of the monument dictates the layout of line Z. Ekerafaz and eknaudigastiz on the other hand are meant to make up two lines within the ‘monument’ as it is outlined on the stone. Therefore, we shall have to figure out whether ekerafaz (line X) and eknaudigastiz (line Y) are meant to be read first or last. Since these lines similar to the stone itself, are situated in the centre of the monument, they are probably the most important ones and indeed two parallel lines. This indicates that the simplest way to read the text is to start in the centre with the meaningful line, ek erafaz—I Erafaz, and proceed upwards reading lines X – Y – Z, in order eventually to have read through ‘the whole monument’. This means that we start by reading the inside of the monument, i.e. the rune stone, and proceed to cover its perimeter, the kerbstones.

ILLU 05a

Since inscriptions that mix lexical and non-lexical runes, usually adhere to a systematic pattern: lexical – non-lexical – lexical, in such a way that the last lexical section brings the inscription to an end in a few words, the X – Y – Z reading becomes even more reasonable. ‘Innananaboz’, to be sure is typical lexical cadenza and there is no need to start all over again with a new expression such as ‘ekerafaz’. ‘Ek’, i.e. ‘I’, mreover, is a common somewhat pompous introductory pronoun (Table 1).


On the Hogganvik stone word wrapping doesn’t necessarily indicate a major break in the text and there need not be a break between eknaudigastiz and kelbaþewas. Dots on the other hand always do. This means that there is probably a break between kelbaþewas and stainaz. Since there are no signs to end an expression, the dots simply mark out a beginning that the carver thinks is not obvious. Evidently, this is the function of the dots between kelbaþewas and stainaz. Stainaz is the beginning of something and kelbaþewas thus the end. These dots are necessary because kelbaþewas stainaz—Kelbaþewa’s stone is otherwise a perfectly reasonable expression.


Since most runologists expect the texts to be inscribed on a headstone, they disregard the vertical dots and read what they expect: ‘Kelbaþewa’s stone’. In fact they will think that disregarding the dots is a consequence of the two words.

Be this as it may: if the text, guided by lines and dots, can be read as two propositions – one, preferably the last, containing a non-lexical expression close to the end of the whole text – it should be read so, for instance:

ILLU 07a

The first line, Ek erafaz ek naudigastiz kelbaþewas, is a nominal sentence: I Erafaz, I Naudigatiz (am) Kelbaþewa’s

Since erafaz means ‘wolverine’ and Naudigastiz ‘need-guest’ and kelbaþewa ‘chilver-servant’ (chilver is ‘ewe lamb’) and since they are words used as appellatives and/or proper names, the sentence runs: I Wolverine, I Needgest (am) Chilver-Servant’s. Wolverine presents himself with a name, in effect an appellative, as Chilver-Servant’s need-guest. ‘Guest’-names and ‘servant’-names as well as ‘need’-names are Late Early Iron Age names (5).

The second line, stainazaaasrpkfaarpaainnananaboz, is a sentence too. ‘Stainaz’ is the subject and ‘innananaboz’ is a prepositional phrase or a postpositional adjective referring to stainaz. ‘Innana’ is a preposition which governs the genitive. It means inside or within, and ‘nabu’ means nave, but not just in a wheel.

ILLU 07b

For some reason, Michael Schulte seems not to contemplate that innananaboz might be a postpositional adjective describing stainaz. In Scandinavian languages this way of constructing an adjective is common, not least in Danish, which MiSchu doesn’t mention: inden+bords, inden+bys, inden+dørs, inden+lands, inden+rigs, inden+skærs or inden+sogns. They are always matched by their antonym “uden+…”. There are also local usages applicable when the difference between inside or outside is sufficiently significant. At Langör by Stavnsfjord on Samsø children in the 1950s could sail their dinkies “indenfjords”, but not “udenfjords”—in the fjord but not outside it.


Today there is no particular point in specifying whether the presence of something is either, or predominantly either, inside or outside a nave and there is no such word as “inden+navs”, but “indennavs” for “within the nave” is nevertheless unproblematic if one’s relation to naves is sufficiently dichotomous. Syntactically speaking, the non-lexical runes ‘aaasrpkfaarpaa’, which may be a representation of the act of an incantation, perchance a gealdor, or ‘alphabetic magic’ to runologists, stands in for a verb such as to sing or chant or express. The second sentence therefore describes the stone in the monument stating that: the stone ‘chants this (non-lexical) phrase’ inside the nave. It might of course have been Erafaz—Wolverine who taught the stones to sing: ‘aaasrpkfaarpaa’. In the two similarly structured inscriptions in the above table, DR 261 and DR 196, the ‘I’ rather than the object is the agent.



(1) Jansson, Sven B. F. 1978. Värmlands runinskrifter. Sveriges runinskrifter. Vol. 14:2. Stockholm Almqvist & Wiksell International. Jansson discusses this on pp. 32-36.

(2) In this respect Knirk, James. 2011. Hogganvik-innskriften: en hard runologisk nøtt. Viking 2011. Pp 25-39 and MaSCHu:123 represent two different attitudes to preunderstanding.

(3) This is: Glørstad et al. 2011. Zanette Tsigaridas Glørstad, Jakob Johansson & Frans-Arne Stylegar. Minnelund og monument. Runesteinen på Hogganvik, Mandal, Vest-Agder. Viking 2011. Pp 9-24.

(4) It is not unusual that early inscriptions have a relation to the object on which they are inscribed, thus adding dimensions to their textual meaning, given that they are often written on artefacts.

(5) See below and Peterson, Lena. 2004 Lexikon över urnordiska personnamn http://www.sprakochfolkminnen.se/download/18.5e02b54a144bbda8e9b1c11/1398151044347/urnordiska-personnamn.pdf

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: