Re-reading Ernbodied Texts anInterpretation ofRune-stonesAnders Andrn

Rune-stones have been a major field of research in philology, archae-ology, art history and history during the 20'" century. Most of thesestudies have been based on the thorough editions of rune-stones pub-lished in Scandinavia during the century. The aim of this article is toquestion some of the fundamental principles of these editions, and toinitiate a new type of interpretation based on the complex interplaybetween images and texts on the rune-stones. Elements of a more visualunderstanding of the monuments are presented, as well as some exam-ples of a new contextual reading, which sometimes alter the philologicalinterpretations in the rune-stone publications.

AndersAndrn, Institute ofArchaeology, UniversityofLund, Sandgatanl, SE-223 50 Lund, Sweden.

THE STUDY OF RUNE-STONES ABRIEF SURVEYRune-stones have been studied during 400years in Sweden. In fact, the very start ofSwedish antiquarian tradition in the late 16'"and early 17'" centuries was directly relatedto the recording of rune-stones in the realm.Inspired by epigraphic research on classicalinscriptions, Johannes Bureus (1568-1652),later the first state antiquarian, began thesystematic recording of rune-stones in 1599.The work of Bureus and his assistants wasfollowed up by generations of antiquariansin the 17'" and 18'" centuries, most notablyby Johan Hadorph (1630-1697).He planneda large edition of all Swedish rune-stones,but his work was published posthumouslyonly in 1750, when no less than 1173 rune-stones appeared in an edition called Bautil(Schuck 1935-44; Svrdstrm 1936; Klindt-Jensen 1975).

In the 17'" 18'" and first half of the 19'"centuries, studies of rune-stones were part of

a broad antiquarian tradition. As part of aglorious past, rune-stones were registeredtogether with churches, hill forts, graves,ancient objects, old books, medieval doc-uments and legends. In the second half ofthe 19'" century, however, the new idea of avast prehistory without any written records,and the general trend of empirical special-isation, led to a successive division betweenarchaeology and rune-stone studies. The or-dering of objects and monuments became thetask of archaeologists, whereas rune-stoneswere recorded, interpreted and published bysuccessively more specialised philologists.

In the early 1880s, the Royal Academyof Letters, History and Antiquities decidedto start a new edition of all Swedish rune-stones, "which could meet the demands ofmodern science" (l p. 10, cf. Hildebrand1884). After fieldwork and archival studies,the first volume of Sveriges runinskrifter(Runi c i nsc& iptions ofSweden) was publishedin the year 1900 (fig. I). Today most of the

Current Swedish Arehaeotogv, Vol. g, 2000

8 Anders Andrn

Fig. l. Rune-stone at Karlevi on land (l 1).This monument was the verv first rune-stonepresented in Sveriges Runinskrifter in the year1900. It represents an early type of rune-stonefrom about the year 1000, with a linear text butcontaining a highly elaborate poetic language.According to Sven 8. F. Janssons translation(Jansson 1962&125) the inscription runs:"This stone was erected after Sibbe the good,Foldars son, but his host placed on the island. . .Hidden lies the man whom the greatest virtuesaccompanied n&ost men knew that 'executor'of the goddess of battles (i.e, the warrior, lord'in this mound. A more honest battle-strong godof the wagon of the mighh& ground of the sea-king(i.e. a more honest battle-strong sea-captainjshall not rule over landin Denmark".

Swedish rune-stones have been published inthis edition, but the major part of the corpuswas completed in 1958, when the finalvolume covering the province of Upplandappeared. Similar editions have been madeof the rune-stones in Denmark and Norway,including the rune-stones in former Danishand Norwegian provinces in present-daySweden (cf. DR and NlyR). Thanks to these

editions nearly the whole corpus of Scan-dinavian rune-stones have been thoroughlypublished in the 20'" century.

It has been the work of philologists toprepare and edit these publications, since themain aim has been to give a convincing se-mantic interpretation of the text on everyrune-stone. The pervading method has beento publish one or more pictures of the rune-stone and its inscription, a transliteration ofthe rune characters into Latin letters, a nor-malised text in Old Icelandic, and a trans-lation into a modern language. Although thephilological interpretation has been the focusof the publications, the editors of Sverigesruninskrifter have always given an accountof the location of rune-stones, and the orna-ments and images on the monuments. Es-pecially Elias Wessn (1889-1981),who wasediting volumes from 1932 to 1962, develop-ed the sections on location and ornaments.In Denmark, similar perspectives weredeveloped by Erik Moltke (1901-1984) inDanmarks runeindskrifter as well as in a latergeneral survey of runic inscriptions in Den-mark (Moltke 1976).

Owing to the thoroughness of the editions,Sveriges runinskrifter and its counterparts inDenmark and Norway have been the naturalstarting-point for all studies of rune-stonesin philology, archaeology, art history, history,and church history in the 20'" century (Owe1995). In the last few decades especially, thestudy of rune-stones has expanded consid-erably, covering a whole range of differenttopics. In philology issues such as ortho-graphy (Lagman 1990; Williams 1990),words (Peterson 1994),poetry (Hiibler 1996),the function of runes (Ba:ksted 1952; Moltke1976:17ff.), rune-carvers (Thompson 1975;Philippa 1977; Arend 1978; Crocker 1982;Axelsson 1993; hln 1997; Stille 1999; cf.Brate 1925), and personal names and place-names (Salberger 1978) have been inves-tigated. The chronological and regionalpattern of stone-raising has also been an im-portant philological issue (Palm 1992; cf. von

Curren& Swedish Archaeotog&:, Vot 8, 2000

Re-readi ng Embodi ed Texts 9

Friesen 1913, 1928, 1933). In archaeologythe classical topics have been chronology andregional variation in the ornamental styleson the rune-stones (Holmqvist 1951;Wideen1955; Christiansson 1959; Grslund 1991,1992, 1994; Lindblad & Wirtn 1992; Sund-qvist 1996) as well as the spatial location ofthe rune-stones (Ambrosiani 1964; Hyen-strand 1974; Wijkander 1983; Larsson 1990,1997; Wilson 1994; Johansen 1997; Zachris-son 1998).In art history stylistic analysis andiconographic studies have dominated (Berg-man 1948; Fuglesang 1980, 1981 and 1986),whereas historical investigations havefocused on political history (Lnnroth 1982;Lindkvist 1988; Sawyer 1991) and heritageand property (Carlqvist 1977; Sawyer 1988,1989; Sandahl 1996).In church history aboveall aspects of Christianisation have attractedattention (Ljungberg 1938; Hallencreutz1982a, 1982b; Segelberg 1983).

However, the complex character of therune-stones, as monuments in a landscape,as works of art and as written documents,means that many rune-stone studies trans-gress the traditional disciplinary lines. Philol-ogists have studied design (Thompson 1975),petrography (Hagenfeldt &, Palm 1996) aswell as paganism (Grenvik 1981, 1985),Christianisation (Moltke 1976:180 ff. ; Sj-berg 1985;Williams 1993, 1996a, 1996b) andsocial structure (Moltke 1976:147 ff., 233 ff. ;Strid 1987). Archaeologists and historiansalike have drawn attention to the role ofwomen (Kyhlberg 1983; Saywer 1988; Jesch1991a:42 ff. ; Saywer & Saywer 1993; Grs-lund 1989, 1996). And many archaeologistshave investigated rune-carvers (Berg 1987;Kllstrm 1992), Christianisation (Grslund1985, 1988, 1991; Elgh 1987; Sderqvist1993; Andersson 1995; Ekblad 1997; Mo-strm 1998), property rights (Lindstrm1992; Zachrisson 1994), iconography (Chris-tiansson 1953, 1995;Thrainsson 1994; Lager1995; Zachrisson 1998), social structure(Ragnesten 1978; Randsborg 1980; Christop-hersen 1982; Claesson 1982; Carlsson 1985;

Andersson 1992; Bjurstrm 1996), militaryorganisation (Varenius 1998), mental land-scapes (Johansen 1997), and even formulasand words (Herschend 1994; Johansson 1996;Norr 1998). In some cases studies have beenmade in disciplinary cooperation (Gustavson& Selinge 1988).

As a summary of rune-stone studies inthe 20'" century I would like to point out thefollowing characteristics. The interpretationof the rune-stones and their texts has beenbased on philological interpretations of thetexts, on interpretation of images, on surveysin time and space of different aspects of thetexts, and on additional aspects of the rune-stones, such as their location and spatialrelation to other ancient monuments. Owingto the complexity of the rune-stones, manystudies have been based on comparisons ofdifferent aspects of the monuments, such astexts and ornaments or texts and locations.Although these comparisons have led to newand interesting results in many cases, it ischaracteristic that the semantic interpreta-tions in the rune-stone editions have usuallybeen taken for granted.

In this article I want to start a new discus-sion of the interpretation of rune-stones byquestioning the fundamental principles of theedition of the Scandinavian rune-stones. Thepublished corpus of nearly all the Scandina-vian rune-stones is indispensable, but thegeneral method of philological interpretation,from image, via transliteration and normalisa-tion to translation, reduces the runic inscrip-tions to linear texts which they seldom are.This reduction has been further underlinedby perspectives that the animal bodies onwhich many texts were cut had no specialmeaning but were only decorative basis forthe text. I maintain that the reduction fromrunic inscription to linear text and the reduc-tion from animal art to decoration have re-duced the meaning of the inscription andultimately of the rune-stone as a whole.Therefore I would argue for a new kind ofre-reading, based primarily on combining in-

Cnrrent Sn edish Archaeologv, Vol. tt, 2000

10 Anders Andrn

scriptions, ornaments and images in jointinterpretations. Location should be added asa further aspect of the interpretation, althoughI will not consider the place of monument inthis article.

I see this re-reading as an example ofhistorical archaeology, and especially of whatI recently labelled as "integrated discursivecontext" and as "association" (Andrn1998:149 ff. , 168 ff.). By these concepts Iunderstand a contextual reading based on theoften complex interplay between image andtext. In the case of rune-stones this interplayimplies that the images were not only illus-trations, but also extensions, complements,reinforcements or variations of the messagesconveyed by the texts. Therefore, an impor-tant aspect of a contextual reading will be todevelop a kind of "visual literacy" (cf. Berlo1983), which may enable us to understandhow the interplay of image and text actuallyworked on the rune-stones.

sense. To me, this means that we should notjust read the inscription as a linear text, butinterpret it in its totality. Consequently, weshould take into consideration that the textis written on the body of two snakes, and thatthe text is intertwined in itself, by the plaitingof the two bodies and their legs and tails.

The problem for a modern viewer to makesuch a combined interpretation is that thedifferent elements of the rune-stones for sucha long time have been studied by differentdisciplines. The texts have been studied andinterpreted by philologists, whereas archae-ologists and art historians have studied theornaments and images. Animal art has beenreduced to decoration without further mean-ing, just as the inscriptions have been reducedto linear text.

Ever since the beginning of modernarchaeology in the 19'" century, attention hasbeen paid to animal art (cf. Karlsson 1983).

"LET THE ONE SOLVE WHO CAN"A good starting-point for a renewed discus-sion of the interpretation of rune-stones isthe rune-stone at Nybble on the island ofSela in northern Sdermanland. The in-scription on this monument ends with thesentence "let the one solve who can" (S 213,fig. 2); an expression which is repeated onseveral other stones in central Sweden. AtNybble the final sentence can be explainedby a partly abstruse runic inscription, but inother cases these expressions are inserted intotexts that are not themselves particularlydifficult to interpret. Therefore, these for-mulas can be understood as more or lessconcealed requests to make some kind ofbroader interpretation of the stones. The mostprovocative request comes from the longinscription at Hillersj in southern Uppland(U 29, cf. fig. 3), which starts with the wordragu ("read/solve/interpret") cut into the eyeof a snake. Here it is quite clear that the rune-carver, named Torbjrn Skald, requests thereader and viewer to use the eye as the active

Fig. 2 Rune-stone at Nybble on the island ofSelain northern Sdertnanland (S 2L3). The rune-stone represents a transitional rype of'monument,with two twisted snakes as well as a text writtenin verse. According to a translation based on SvenB. F. Jansson (Jansson 1962.'150) the inscriptionruns:

Right snake: "sbjrn hewed the stone I colouredas a memorial I he bound it with runes I Gyllaraised it I alter Gerbjrn her husband".Le/t snake: "and Gudfrid after his father /He wasthe best yeoman in Kil / Let the one solve whocan ".

Current Swedish Archoeologv, Vot. tt, 2000

Re-reading Etnhodied Teets l 1

Among the important contributions are Miil-ler 1880; Salin 1890, 1904; Shetelig 1909,1949; Brendsted 1920, 1924; berg 1921,1941; Lindqvist 1931;Holmqvist 1951, 1955;Christiansson 1959, Orsnes 1966, 1969;Fuglesang 1978; Haseloff 1981. The pre-vailing perspective in these studies has beento regard animal art as decoration withoutany deeper meaning. Therefore, the main aimof the investigations has been to define dif-ferent styles in time and space, and to followdifferent stylistic influences on animal art.A recent survey of animal art from this per-spective is Wilson 1995.

Another minor line of thought has, how-ever, existed during the whole century, andhas in the last decade played an importantrole in the debate on animal art. Already inthe early 1890s the Swedish archaeologistSven Sderberg argued that animal art shouldbe regarded as a parallel expression to skaldicpoetry (Sderberg 1905 however, the articlewas written in 1892-3). To him animal artwas thus not a neutral decoration, but couldinstead be viewed as an analogy to the diffi-cult and complex art of poetry, practiced bywel l-known specialists called skalds. Similarideas have later been expressed by the Nor-wegian literary historian Hallvard Lie (Lie1952, 1963)the Norwegian theologist AndersBugge (Bugge 1953) and the Russian his-torian Aaron Gurevich (Gurevich 1985).Inspired by recent archaeological debate onmeaning of material culture, some archae-ologists in the 1990s have underlined an in-herent meaning of animal art, with associa-tions above all to pagan cosmology (Thrains-son 1994; Kristoffersen 1995; Hedeager1997, 1999). Others have turned back to theideas of Sderberg and Lie, underlining theanalogous construction of animal art andskaldic poetry as complex and partly con-cealed expressions (Johansen 1997; Zachris-son 1998; cf. Andrn 1989, 1993).

In spite of these new perspectives onanimal art, the interpretation of the rune-stones has not changed fundamentally. Some

of the images on the rune-stones have beenreinterpreted, but without references to thetexts, and the texts from the general editionsare still used as face value. Apart from aninteresting combined interpretation of textsand images on the large rune-stone at Jelling(Fuglesang 1986), there is only a single hintthat the animals on the rune-stones may beconnected with the content of the texts(Sawyer 1988:15 ff.). No one, however, hastried to make a more general re-reading ofthe rune-stones, based on an associated closereading of texts, ornaments and images. Inthe following lines I will initiate such a re-reading, with seminal inspiration from recentstudies on art and epigraphic writing in Meso-america (Berlo 1983; Hanks & Rice 1989;Marcus 1992), Egypt (Baines 1989) and theclassical world (Zanker 1988; Brard et al.1989; Svenbro 1993), as well as from inter-art studies (Lagerroth el al. 1993).

ELEMENTS OF UNDERSTANDINGAlthough my aim is to interpret the rune-stones in a new kind of linguistic and visualtotality, it is necessary to examine differentelements per se in this totality. Unfortunatelyone important element for the interpretationof rune-stones is generally lacking today,namely colour. From occasional finds weknow that the rune-stones were painted, andsometimes repainted, in bright colours. Awell-preserved example is a runic gravestonefrom the church of St Lawrence in Linkping,which was painted in five colours (Jansson1962:152-3).Two snakes were coloured redand two snakes were painted black, whereasthe background was held up by bluish greyand the knots binding the snakes together bylight bluish grey. Finally, the runes werepainted alternatively white and black, empha-sising some of the words. Although thiselement is generally lacking, it is possible todefine other important elements of under-standing, by looking at the changes of thetexts and the layout of the rune-stonesthrough time.

Carrent Sreedich Atehaevlogcc Vol. R, 2000

12 Anders Andrn

Fig. 3. Rune-carving at Hillersj insouthern Uppland (U 29). The figureshows the carving as preserved(above) and as reconstructed in itsmain design (below). The reconstruc-tion is shown with names and keywordsin the inscription as well as withall divisions of words (points or thiclclines). According to a translationbased on Sven B. F. Jansson (Jansson1962:77) and Birgit Saywer (Sayuer1988:17) the inscription & uns:Eye in left snake: "Solve!"Left snake (sltaded): "Gertnund t(ljtook Gerlg f2J a maiden as svife. Thenthey had a son f3J before he f= Ger-mundJ was drowned and then the sondied. Then she had Gudrik /4J as herhusband. He. . . Then they bad childrenj5J, but only one girl survived. She wascalled"Right snake: "Inga /6J. Ragnfast f7JofSnottsta took her to wife. Thereafterhe died and the son j8J. And themother f9J /= Inga J inherited from herson. Then she had Erik flOJ as herhusband. Then she died. Then Gerlgjllj inherited from Inga j12J herda u gh ter. "

Below left snake: "Torbjrn skald f13Jcatved the runes ".

The early rune-stones, from the earliestexamples in the 4'" century until the late 10'"century, are usually stones with simple linesof runes. The text is linear in a modern sense,but the content is sometimes difficult tounderstand because the message is more orless concealed. The long inscription at Rkin western stergtland (g 136) from about800 is partly written in cipher and onlypossible to read with the help of a code placedon top of the stone. Other stones are writtenin verse, like the stone at Karlevi on land(l I, cf. fig. I) from the late 10'" century,which is loaded with difficult poetic met-aphors (heitis and kennings).

Only in the 11"century did the texts andthe layout of the rune-stones change intomixed expressions. The texts usually became

linguistically more simple as prose, whereasthe layout became successively more com-plex with animals and ornamental loopsintertwined in complex patterns. Thus, it ispossible to trace a passage from linguistic tovisual complexity among the rune-stonesfrom the late 10'" century until the early 12'"century (cf. figs. I, 2 and 3, and cf. Thrains-son 1994:46 ff. ; Zachrisson 1998:129 ff.).Important elements in this late visual com-plexity are animals with texts on their bodies,animals without texts but twisted aroundother animals, directions of the text, word-crossings and diiferent kind of images whichare inserted by the animals with texts.

Snakes with textsWhen runic inscriptions about the year 1000

Current Swedish Archoeology, Vot. 8, 2(ln(1

Re-r eadi ng Entbodied Texts 13

began to be combined with animal bodiesthey were cut into animals that might beclassified as snakes or dragons. There wasno clear distinction between these creatures,and consequently I call them snakes. Re-cently Birgitta Johansen has made a study ofthe important snake motif in late Iron AgeScandinavia, and tentatively interpreted themin general as protective creatures, especiallyof women and treasures. She has, however,interpreted the snake on the late rune-stonesmore specifically as protector of the farm orthe village, surrounding the inherited prop-erty as an equivalent to the Midgard serpentsurrounding the world (Johansen 1997:143ff. , 224; also cf. Zachrisson 1994, 1998).Bycomparing inscriptions of rune-stones con-taining one snake with those containing twosnakes, I maintain that it is possible to definea somewhat different meaning, but with clearconnections to Johansen's interpretation. Inmost cases the snake seems primarily torepresent a family, although including otherassociations to a family, such as inheritance,property, land and settled space.

A good example of this interpretation isthe above-mentioned inscription at Hillersj(fig. 3). It contains two snakes, and the textis clearly divided between them. On the firstsnake a certain Gerlg, who commissionedthe inscription, mentioned herself, her twohusbands and their children, and on thesecond snake she mentioned her daughterInga and her two husbands and the daughter'sonly child. Thus, the two snakes visuallymark out the families created around twowomen; one mother and one daughter. Weare in the lucky situation that the daughterInga had another four rune-stones raised inanother part of Uppland (U 329-332; cf.Zachrisson 1998:165ff.).On these stones shecommemorated her first husband, their son,her sisters-in-law, her father-in-law, and herhusband's steward (huskarl). On three stonesshe only used one snake to describe thefamily, but on the one where her father-in-law is mentioned she used two snakes (fig.

4). Thus, from her perspective the husband,the son, the sisters-in-law and the stewardbelonged to the same family, whereas thefather-in-law belonged to another family.From these cases it is clear that the snakesrepresented families, but not in a clear-cutsense. What could be regarded as one familyor two families differed according to the con-text, and consequently the interpretation ofa rune-stone must always be based on bothtext and ornament.

An interpretation of snakes as represen-tation of families and their properties is wellin accordance with the Gutasaga (Saga ofthe Gotlanders) from about 1200. The sagais a mythological introduction to the pro-vincial code of Gotland, and it gives a his-torical background to the island and its laws.According to the saga, Gotland was initiallyinhabited by a single couple. The first nightwhen they slept on the island, the wife dreamtthat three snakes were twisted in her stomach.Her husband interpreted the dream as a visionof their three sons that would eventually in-habit the island and divide it between themin three parts. The snakes, in other words,represented the future families of the threesons and their respective possessions on theisland (Holmbck & Wessn 1943:291).

Finally, by interpreting the large snakeson the rune-stones as families and their re-spective properties, it is important to under-line that snakes with texts are nearly alwaysof identical size. This may indicate that thefamilies mentioned had the same hierarchicalposition, and that paternal and maternal kin-ship were regarded as equal.

From hectd to tallSince the snakes represented families in somesense, it is essential where the texts were cutinto the animal bodies: Where does the textstart and where does it end? Most inscriptionsstart in the snake's head, and recently thisfact has been explained as a kind of readingdevice (Norr 1998:20 ff.). I would ratherargue that the direction of the text in itself

Current Swedish Archaeotogy, Vot. R, 2000

14 Anders Andrn

Fig. 4. Foiir rune-stones iaised at Snottsta in southei n Uppland, by Inga Gudriksdottei. fio1Hillei sj (cf fig. 3) aftei hei' husband Ragnfast Sigfastsson. The stones shosv how Inga Gud&iks-dotter used one snal e in three cases and two snales in one case to represent the family. Tnosnal. es a&e only used when her fathei in la&v is mentioned. The inscriptions run as follows:A (Vreta, U 332). "Inga raised staff and stones afle&. Ragnfast, her husband. She came intainlieritance aftei her chitd"B (Snottsta, U 329). "Inga had these stones raised after Ragnfast, her husband. He wasbrotlie&. to Gyrid and Estrid"C (Snottsta, U330): " Inga had these stones iaised and made the bridge after Ragnfast, herhusband. Assur was his sei vant"D (Snottsta, U 33l). Lefi snal'e: "Inga had these &.iines ca&ved after Ragnfast, her. husband.He owned alone" Right snake: "this village after Sigfast, his father. God Iielp their souls ".

Curreni Swedish Aiilioeology, Vol. tl, 2000

Re-reading Embodied Texts 15

had a meaning. In most texts the persons thatcommissioned the stones are mentioned at thebeginning and the commemorated deadpersons are mentioned at the end (cf. figs. 4,SC, 6C and D, 8, 9, 10E and 11B).The firstperson mentioned, usually placed in thesnake's head, may virtually have been thehead of the surviving family, and the orderof the other persons may have been anindication of the ranking within the family(cf. Saywer 1988:15 ff.). Usually a man ismentioned as the first person in the head, butsometimes a wife or a mother is mentionedbefore brothers and sons. These cases mayindicate a situation where a woman is headof a family, due to her property right, forinstance as the single heir of an estate.

Although most texts start at the snake'shead, a few inscriptions start at the tail (cf.fig. 6B).This divergence may sometimes beexplained by comparing the iconography withthe text. An illuminating example is the rune-stone at Vnga in eastern Sdermanland (S3). The inscription starts at the tail, men-tioning a man and a wife who raised the stonein memory of a dead son. The text ends atthe head, however, with a concluding sen-tence mentioning a brother of the dead man.The reverse direction of the text may in thiscase be an indication that the parents wantedto point out that the surviving son was morehonourable than themselves, possibly throughservice outside the family.

Snakes without textsApart from large snakes with texts, a largenumber of late rune-stones also contain manysmall snakes that are twisted around the largeanimals. The significance of these smallsnakes is indicated by the process by whichthey were created. It is possible to trace asuccessive transition during the 11'" century,from short additions of the main texts to smallsnakes twisted around the main texts. In thelatter cases the small snakes are twistedaround passages of the text, which give thesame kind of extra information as the earlier

textual additions. Consequently, I see thetextual additions outside the main text-bandas deliberate compositions, and not as "lackofplanning" or "faulty planning" of the designof the rune-stones (cf. Meijer 1992).

The pervading pattern is that the textualadditions as well as the texts surrounded bysmall snakes emphasise information aboutrelations outside the family in its narrowsense. Above all, the textual additions andthe small snakes underline passages aboutmen that took part in raids abroad or were inservice of a lord. The examples in fig. 5mention outside the main text-bands that thecommemorated men had been in Greece ordied "in the East". Many small snakes alsoemphasise the rune-carver, indicating that therune-carver was in the service of the personwho had the stone raised. The examples infig. 6 again show how rune-carvers signedthe rune-stones outside the main loop wherethe close family is mentioned. Divergencefrom this pattern may in a few cases supportthe interpretation of the small snakes. On oneof the rune-stones at Gripsholm in northernSdermanland (S 178) a rune-carver is men-tioned in the main inscription, but withoutany extra small snakes around the text. Inthis case, however, the rune-carver was abrother of the dead woman commemoratedon the stone, and hence the rune-carver wasa member of the immediate family.

Because the small snakes indicate rela-tions outside the close family, the size of thesnakes seems to be significant. Contrary tothe large similar sized snakes with texts, thevery difference of size between large snakeswith texts and small snakes without textssignifies hierarchical relations, upwards aswell as downwards, for instance to Byzantineemperors, kings, earls and chieftains as wellas to rune-carvers in service as craftsmen.

8'ord crossi ngA less common element in the rune-stones isword crossing, which is the result of snakeswith texts that are closely intertwined with

Cuneni Sieedish Archaeology, Vol. 8, 2000

16 Anders Andrn

Fig. 5. Four rune-stones which illustrate the transition from textual additions to small snakes withouttexts. All the examples are related to military service abroad.A (Grinda, S 165). Main text-band: "Gudrun raised the stone after Hedin, jhej was nephew to Sven.He was"Right additional text-band: "in Greece, divided gold"Left additional text-band: "Christ help all souls of Christians ".B (Fredriksdal, S 33). Main snake: "Gnupa had this stone raised after Gudlev, his brother" Text-band outside main snake: "He passed away in the east in military service"C (Tystberga, S 173). Main snake: "Muskia and Manne had this memorial raised after their brotherRodger and their father Holmsten"Minor snake: "He had long been in the west"Text by the minor snake: "[They] died in the east with Ingvar"D (Ed, U 112). The inscriptions are cut on two sides of a large block of stone.Main snake (not shown in the figure): "Ragnvald had the runes cut in memory of Fastvi, his mother,Onm s daughter. She died in Ed. God help her soul.Minor snake (shown in the figure): Ragnvald let the runes be cut. [He] was in Greece, [he] was leaderof the host".

Current Swedish Archaeology, Vol. 8, 2000

Re-reading Embodied Texts 17

Fig. 6. Four rune-stones which illustrate the transi tion from textual additions to small snakes and tai lsof snakes without texts. All the examples are related to rune carvers.A (Atja, S 333). Main snake: "Amunde raised this stone afier his son Runulv and aftet Ring, hisbrother. (HeJ was killed out on Kalmarsund, fwhen theyf went to Skne"Extra text-band: "Eskil cut these runes"B (Korpabro, S 139). Main snake: "Sten had this stone raised after Ashed, his relati ve. Christ gi verelief for her soul, a good woman"Extra text-band: ".. .and Finn, they cut the runes"C (Brsicke, S /95). Main snake: "Holmsten raised the stone after Stenulv, his grandfather, a goodyeoman who li ved in Brsicke. Gud help his soul better than he deserved. Hallbjrn"Extra runes: "hewed the runes on [the stone]"D (Frsunda, U 346). "Ragnfrid had this stone raised after Bjrn, her and Kttilmunds son. He waskilledin Virland. God and Gods mother help his soul. smund marked right runes"

Current wedish Archaeology, Vol. , 2000

18 Anders Andrn

Fig. 7. Rune-stoneatgersta in southern Uppland(U 729). The monument is an example of howcomplex word crossings can extend the meaningofthe text. ln the following translation words fromthe opposite snake are marked in parentheses.Right snake (shaded): "Vidhugse had this stoneraised (those runes that Balle carved) in memoryof Srev, his good father (who rune-wise is). Helived at (gersta)"Lejt snake: "(He lived at) gersta. Here shallstand the stone between farms. Let that man solvewho rune-wise is (his good father) those runesthat Balle carved (Vidhugse had this stoneraised) ".

each other. The close relationship between thetext and the plaited snakes is especially clearin the above-mentioned inscription at Hillersj(fig. 3). The rune-carver Torbjrn Skald hadusually employed points to make divisionsbetween words. In 16 cases, however, he hadinstead used crossing parts of a snake as word-divider. In this case, the design of theinscription can be regarded as extremely "wellplanned" (cf. Meijer 1992).

The rune-carvers did not only use plaitedsnakes as division between words, but alsoas means to create visual associations be-tween words that were not textually related.A good illustration of this visual associationis the rune-stone at gersta (fig. 7) in south-western Uppland. The rune-carver Balle cre-ated three different clusters of word crossingthrough the intertwined snakes. In the right

part of the inscription is a cluster of wordcrossing that emphasises the relationship be-tween the rune-carver and the man who hadthe stone raised. A cluster in the left part ofthe inscription underlines the location of thestone and its relation to the farm where thedead man had lived. Finally, a centrallyplaced cluster of word crossing holds up thecharacter of the dead man, especially hisknowledge of runes. This aspect of the deadman's character, however, is not evident fromthe inscription as a linear text.

The word crossing gives an associativeunderstanding of the text, which closelyresembles skaldic poetry. Just as poetry wasbound together by grammar and metricelements such as alliteration, the runic in-scriptions were tied together by intertwinedsnakes with texts.

The combined composition of the text andthe plaited snakes is in several cases highlysophisticated, and therefore it is very inter-esting that the rune-carver at Hillersj in factcalled himself skald (poet).

lmagesMany rune-stones contain images or stan-dardised symbols like a cross in connectionwith the texts. In a few cases, as with thelarge rune-stone at Jelling in Denmark (DR42), image and text have been interpretedtogether (Fuglesang 1986; cf. Moltke 1974).Usually, though, the images have been in-terpreted in isolation, without references tothe rest of the rune-stones. From my point ofview, these images should be regarded as partof the total composition, and consequentlythey are very important for the re-reading ofthe rune-stones. The close connection be-tween image and text can be illustrated bytwo examples.

One rune-stone at Tumbo in westernSdermanland (fig. 8) is dominated by a largebeast with sharp teeth. According to the in-scription, the stone was raised for a certainFrsten who "was dead" in Greece. The word"dead" (bu&R) was cut directly in front of

Current Swedish Archaeology, Vol. 8, 2000

Re-readi ng Embodied Texts 19

Fig. 8. Rune-stone at Tiubo (S 82). The stone isan example of the close t elationslup beta:een textand itnage. The inscription runs: "Visten taisedthe stone aftet Frsten, his brotltet; dead inGreece. Tule ltewed runes ".

the jaws of the beast. This relation betweenimage and text can be interpreted with ref-erence to Norse mythology. According toseveral poems the metaphor for being killed

in a battle was that the wolf was fed (forinstance Smunds Edda pp. 127, 193; andSnorres Edda p. 173).The beast at Tumbo canconsequently be interpreted as a wolf, and theclose relation between the wolf's jaws and theword "dead" implied for the well-informedreader that Frsten had died in a battle,although that was not directly mentioned inthe text.

Another example is the well-known rune-carving at Ramsundsberget in north-westernSdermanland (fig. 9), with illustrations ofthe hero Sigurd in the Volsunga saga. Ac-cording to the text the carving had been madeby Sigrid for her dead husband. Two menare named, Holmger and Sigrd, but the textis a bit obscure, which means that it is diffi-cult to determine the name of her husband(cf. Jesch 1991b).From an iconographic pointof view I would argue that Sigrd was herdead husband's name, since Sigrd is thelocal form of Sigurd. It is only this inter-pretation that gives a comprehensible solutionof the adjacent images. The widow Sigrid hadused the pagan virtues of the Volsunga sagato commemorate her dead husband Sigrd.The images implied that he had been as heroicas his namesake the famous hero Sigurd

Fig. 9. Runic inscription at Ramsundsberget (S lOl). This famous monument with illustrations of theVolsunga saga is anothe& example of the close t elati onship between text and image. On the basis of theimages the portin obscure i&tsctiption can be interpreted as follows:"Sigrid, Alril-s mother, Ot. m s daughtet. made this bridge for [the] soul of Holmget. s father Sigrd,her husband ".

Current Swedish Archaeotogv, Vol. 8, 2000

20 Anders Andrn

ABjrn

Ulf, s help

in Viborg

D

Cffrrent Swedish Archaeology, Voh 8, 2000

Re-reading Embodied Texts 21

Fig. 10. Five rune-stones illustrating how the cross could be combined with different aspects of thetext.A. Cross and Christian prayer (Lifsinge, S 9). Main snakes: "Bergvid, he and Helga raised this stoneafter Ulf, their son. He"Text between snakes: "passed away with Ingvar"Text around the cross: "God help Ulf's soul"B. Cross and personal name (Ostra, S 118): "ind had the memorial raised after Bjrn and Godhelp the soul"C. Cross and place-nan&e (sseby-Garn, U 180). Left snake: "Sigvat and Torbjrn and Torgrim andArenmund" Right snake: "had the stone raised after their brother Sigsten. He died in Vi-" Cross:"borg"D. Cross and the main loop (Tumbo, S 362)t "Kolben and Jule raised the stone after Ingvar, theirbrother"E. Cross and the concept of the monument (Tibble, U 496): Ragnfast had these marks done for thesouls of Ingefast and Gullev".

Fafnisbani. However, the carving has at thesame time clear Christian allusions. Accord-ing to the text the widow had a bridge made"for the soul . . . of her husband". The words"for the soul" (fur salu) were cut exactlywhere the hero Sigurd stabbed his sword intothe body of the dragon Fafnir. This close con-nection between image and text indicates thatthe whole composition of the carving ex-pressed a dilemma for those persons whoexperienced Christianisation. According topagan virtues a man should be heroic andkil l people in battle, but according toChristian ethics he should in principle notkill another person. The crossing of the swordand the words "for the soul", which in itselfforms a cross, can therefore be regarded as akind of prayer for the dead husband who hadkilled during his life.

This kind of close connection betweenimage and text can also be found in morerecurring images such as crosses and largeanimals. Crosses on rune-stones have inseveral investigations been studied in relationto form, chronology and provenance (Wideen1955; Thompson 1975; Lindblad & Wirtn1992; Lager 1995),but not in relation to morespecific meanings based on the appearanceof the crosses on the rune-stones. However,a comparison of the placing of the crossesand the texts on the rune-stones clearly showsthat the crosses had different signification

according to the spatial context. Some crossesare placed in the middle of the stone, withoutdirect relation to the text (fig. 10A). Themeaning of this location is directly expressedon a rune-stone at Lifsinge in eastern Sder-manland (S 9), where the prayer "God helpUlf's soul" is written around the cross. Asimilar connection between cross and prayeris indicated by two rune-stones from theparish ofYttersel in northern Sdermanland(S 197 and 203). The inscriptions on theserune-stones are identical, apart from a Chris-tian prayer ending one of the stones. On theother stone this prayer is replaced by a cross.ln these examples the cross stands for a prayerdirected towards God.

ln other instances the cross has been cutdirectly by the name of the dead person whois commemorated, for instance "Bjrn" and"Frsten" (fig. 10B, cf. fig. 8). This contextrather indicates a "dead Christian", in otherwords that the dead person had had a Chris-tian funeral. In some cases the person haddied abroad, and consequently the cross canbe regarded as a reassurance that the personreceived proper Christian treatment when heor she was buried far from home. The locationof the death and the funeral was in otherwords important, and sometimes this locationwas emphasised by a close connection be-tween the cross and a place name. On therune-stone at sseby-Garn in southern Upp-

Crrent S&vedish Archaeoiogy, Vol. 8, 2000

22 Anders Andrn

land (U 180) the sentence "He died inViborg" is partly written in the cross itself(fig. 10C).

On some rune-stones the cross is part ofthe ribbons with text (fig. 10D; cf. Lindbladk Wirtn 1992). Since I have interpreted themain text-bands as signifying families, thislocation of the cross may instead indicate thatthe family as a whole was Christian, perhapsafter a Christian wedding. Finally, somecrosses have been cut in direct relation towords signifying the monument itself, like"this stone" (cf. S 97) or "these marks". Inthese instances the cross seems to signify thatthe monument in itself was Christian. Oneexample comes from the grave-field at Tibblein southern Uppland (U 496), which has beendiscussed by Anne-Sofie Grslund (fig. 10E).According to her this rune-stone was con-nected to a Christian grave, and marked outthat the grave-field became Christian (Grs-lund 1988).This context may explain why itwas important to underline that the monu-ment in itself was Christian.

Another common image was a largeanimal, placed adjacent to the snakes. It isknown from several rune-stones in centralSweden, and the recurring pattern is that theanimal appears on stones where a man iscommemorated as "good" or "best" (cf. fig.2). "Good" was a central concept of honour(Herschend 1998a), and consequently theanimal should in some way be related to thatconcept. The animal is well-known in lateViking Age art, not only from rune-stonesbut also from brooches, caskets and weather-cocks. It has in other contexts been inter-preted as a Scandinavian form of the lion asa symbol of lordship (Fuglesang 1986; cf.Hyenstrand 1996:153ff). This interpretationfits well with a relation between the animal/lion and the concept "good". To be "good"was to behave correctly towards friends, ser-vants or a household, i.e. to execute lordship.

Thus, pictures on the rune-stones were notmerely plain illustrations, but images com-posed in direct relation to the snakes and the

texts cut into the snakes. Instead, the imagesfunctioned as allusions or associations, whichunderlined and extended the meaning of thetexts.

Back to the beginningAlthough space does not allow any furtherarguments here, I hope that the exampleshave shown that partly new meanings can beadded to many rune-stones by a combinedre-reading. Let me just tum back to my firstexample, the rune-stone at Nybble on Sela(fig. 2).The iconography indicates very com-plex relations between the persons mentionedon the stone. The rune-carver placed himselfat the beginning of the text in the head of asnake, probably because he was related tothe dead man's widow, and maybe becausehe was the head of her family. However, healso used a small snake to mark out himselfas a rune-carver, possibly in relation to thedead man's son. Finally, the rune-carver indi-cated by a word crossing that he could readand solve the meaning of the monument. Thedead man was commemorated by the widowand a son, but since they are mentioned ontwo different snakes, the son probably be-longed to another marriage. Finally, the four-legged animal visually underlined that thedead man was "the best of dwellers" at Kil.Thus, a combined reading of the embodiedtexts and the image may lead to an inter-pretation that underlines and partly extendsthe semantic meaning of the text in itself.

IMPLICATIONS OF A RE-READINGMy proposed re-reading of rune-stones fromthe 11'n century has several implications forthe interpretation of these monuments that Iwant to comment on, in this concludingsection. The proposed re-reading must firstof all be regarded as an attempt to extendcurrent interpretations of the rune-stones. Ihave no intention to disprove the philologicalinterpretations in Sveriges rttninskrifter,rather the contrary. The aim is instead tocomplement these interpretations with a more

Current Ssvedish Arehaeotogv, Vol. 8, 2000

Re-reading Embodied Texts 23

associative close reading, by interpreting theornamental loops and images together withthe texts. This re-reading can affect theinterpretation of most rune-stones from the11'" century, but not all of them, since someconsist only of texts or ribbons with text.

Tendencies and exceptionsI have deliberately used a qualitative methodwhen presenting the different elements ofunderstanding, in contrast to many recentquantitative studies of the rune-stones andtheir inscriptions (for instance Sawyer 1988;Herschend 1994; Johansson 1996).The char-acter of the re-reading excludes quantitativemethods, at least initially. The problem ofusing quantitative methods can be illustratedby the animal which bears many of the runicinscriptions. If we want to make a classifica-tion of this creature it is not possible to usequantitative methods, since the animal isnamed only once, as "snake" (ormaluR) ona rune-stone at Hogrn on Gotland (Go 203).Thus, the classification of the animal mustbe argued from this sole good example andfrom the form of the creature, in combinationwith references in Norse I iterature (cf. Johan-sen 1997).My arguments for the significanceof different elements on the rune-stones arein similar ways based on good examples.

Eventually, though, my re-reading mustbe confronted with a more quantitativeapproach. It is important, however, first todetermine the basis for any statistical analysis(cf. Jesch 1994).In this article I have not usedany statistics, but only referred to generaltendencies which seem to support the re-reading. There are exceptions to these ten-dencies, but they do not necessarily speakagainst my interpretations. Instead, my pro-posed associative interpretation may indicateother alternative meanings, as in the fol-lowing example.

According to my interpretations, there isan obvious divergence between snake andtext in the well-known rune-stones raised byJarlabanke in southern Uppland (Gustavson

& Selinge 1988). He had seven stones raisedto himself, while he was still alive, and al-though he only mentioned himself they areall composed of two snakes. This differencebetween image and text can in this case beexplained by the extraordinary context. Fewother persons raised stones to themselves andno one raised as many as Jarlabanke. Theseextraordinary actions by Jarlabanke can beregarded as nearly desperate ways for him toclaim property and other rights. By com-posing the stones with two snakes he alludedthat these rights came from another family,most plausibly as an inheritance from hiswife.

Family formsRune-stones, especially in central Sweden,have in recent years been used in studiesconcerning gender, families, inheritance, andproperty. The re-reading of the rune-stonesmay partly have consequences for thesestudies, and I will comment on one categoryof cases. Most stones with two snakes can beexplained by complex family ties mentionedin the texts. In some instances, especially inUppland, there is no such correlation betweenimage and text. Although the stones arecomposed of two snakes, only two or morepersons are mentioned as sponsors, usuallytwo or more children. If we want to maintainthat snakes generally speaking representeddifferent families, the only possible solutionof these compositions is that the sponsorswere half-brothers and half-sisters, althoughthat was not mentioned in the texts.

To further explain why there were somany half-brothers and half-sisters may openup for a new social perspective on familiesin 11'" century central Sweden. A plausibleexplanation is that polygamy was common,creating very complex family ties, whichwere expressed through the iconography ofthe rune-stones. There are at least two directinstances of polygamy in Sdermanland andUppland, although the editors of Sverigesruninskrifter have tried to explain away these

Current Swedish Archaeology, Vol. R, 2000

24 Anders Andrn

I I. Two rune-stones with information of possible polygamy.A (Uppinge, S 297). Left snake: "Amoda and Moda had this stone laid after"Right snake: "Sigrev, their husband and Sigsten s and Holntsten s brother"B (Brcksta, U l039). Left side: "Stodbjrn and sten raised the stone in memory of their fatherGulle. Mercv has Christ, give telief' to his soul"Right side: "Kj ule (?) caned these runes. Kttillg was his wife called, and Vilv".

cases (fig. 11).At Uppinge in eastern Sder-manland (S 297) two women raised a rune-stone in commemoration of "their husband"and at Brcksta in northern Uppland (U 1039)two women are mentioned as the dead man'swives. In both instances the stones arecomposed of two snakes with texts. The ideaof common polygamy can be further sup-ported by the writings of Adam of Bremenin the 1070s. According to him many power-ful and rich men among the Svear had severalwives, and all the children in these poly-gamous families had equal rights of in-heritance (Adam 1984:219).Adam, in otherwords, seems to describe a kind of officialpolygamy among the Svear, which was partlydifferent from the usual Scandinavian customof concubines among powerful men.

The proposed re-reading of the rune-stones has accordingly led to a reconsid-eration of a seldom-noticed aspect of familyforms in central Sweden (cf. Larsson 1996).This family form may, however, explain whyrune-stones were often commissioned by

many persons in central Sweden, in contrastto southern Scandinavia, where a sole sponsorwas much more common (cf. Saywer 1988).The family forms have to be further inves-tigated, but if polygamous families reallywere common, this kinship pattern will havelarge implications for the archaeologicalanalysis of graves and grave-fields (cf. Petr1984, 1993).

Sponsors and carversAnother field of research has been the roleand function of the rune-carvers. Especiallythe question why the carvers did or did notsign the rune-stones has been discussed inrecent years (Thompson 1975; hln 1997;Herschend 1998b; Stil le 1999).The proposedre-reading may also illuminate this debate.Most runic inscriptions from the 11'"centuryincluding the rune-carver were composedwith extra small snakes twisted around thename of the rune-carver (cf. figs. 2 and 5).As mentioned above, this composition indi-cates a hierarchical relation between sponsor

Current S&vedish Archaeology, Vol. tt, 2000

Re-reading Embodied Texts 25

and rune-carver. The question is why thisrelation in some cases has been important toemphasise, for sponsors as well as carvers.

My interpretation is that the compositionshows a personal agreement of service be-tween the sponsor (large snake with text) andthe rune-carver (small twisted snake withouttext). Sometimes several small snakes are in-tertwined around the large snake, which con-sequently may indicate that several craftsmenwere employed to execute the stone, althoughonly one rune-carver is mentioned. An exam-ple of this situation is known from the rune-stone at Hogrn on Gotland (Go 203), whichwas executed by three men, one who drewthe snake and two who cut the runes.

It may have been important to express thiskind of formal service for both sides. Thesponsor or sponsors could show their abilityand power, by mentioning men in their ser-vice on the stone. For the carvers, as for any"free man", it was honourable to be in theservice of powerful men (cf. Olason 1989),and therefore it must have been important tosign the stones. From this perspective theunsigned rune-stones must have been exe-cuted in other social circumstances. Theywere not the result of a personal service, butinstead the rune-carver must have executedthe unsigned rune-stones as non-personalcommissions, for instance by selling the un-signed stones to the sponsor.

The difference between the signed andthe unsigned stones may thus have some im-portant social implications. The signed stonesshow where rune-carvers, alone or with assis-tants, were part of prosperous households fora period of time, whereas unsigned stonesshow households only commissioning amonument and not having craftsmen in theirservice. If the attribution of unsigned rune-stones to known rune-carvers is correct, itmeans that the craftsmen could manufacturerune-stones in different social contexts duringthe 11'" century.

Cut and spotcen expressionsThe close composition of texts, images andornaments is mainly an innovation of the 11'"century, and more detailed study of thisinnovation would, therefore, be of interest.In this context, however, I will just make afew remarks concerning the style of the rune-stones and its development. The large waveof erecting rune-stones in the late Viking Agestarted in Denmark in the second part of the10'" century. More or less as a starting pointfor this wave stands the large rune-stone atJelling (DR 42). The text on this monumentwas not cut on snakes, but on horizontalribbons, in a much more textual manner thanthe later rune-stones from the 11'" century.Already on this stone, however, there is aclose spatial connection between text andimage. Contemporary Christian manuscriptshave been pointed out as models for thecomplex design of this monument (DR 42;Fuglesang 1986; Wamers 1999).

The introduction of snakes as bearer oftexts, however, occurred later, probably a fewdecades after the erection of the large rune-stone at Jelling. Early examples with verysimple snakes, from about the year 1000, arethe rune-stones at Sporup in northern Jutland(Moltke 1976:212 pp. ), at Torna Hllestadin Skne (DR 296) and at Aspa Lt in south-ern Sdermanland (Sna:sdal Brink k, Wacht-meister 1984:129). Later on, these snakeswere designed in contemporary styles ofanimal art, especially on Gotland and incentral Sweden. Since nearly all late VikingAge rune-stones are Christian, the transitionfrom Christian manuscript to animal art canbe explained by strategies connected with theChristianisation of Scandinavia. As pointedout by several scholars, the Christian messagehad to be adjusted and even translated toNorse expressions to be understood at all,during the conversion (Danbolt 1989;Russel l1994; Zachrisson 1998:135 ff.). The use ofanimal art on the rune-stones may accord-ingly be regarded as one of several examplesof such a Christian adjustment and strategy.

Current Sw:edish Archaeology; Vol. 8, 2000

26 Anders Andr n

Only when the Christian church was morefirmly organised in the 12'" century didanimal art as well as rune-stones disappear,and instead the international Romanesquestyle was introduced as a proper expressionof the Christian message (Andrn 1999; cf.Zachrisson 1998:135 ff.).

The incorporation of texts into animal artin the 11'"century may also shed light on therecent debate concerning animal art ingeneral. The proposed re-reading clearlyshows that animal art was meaningful, butonly in a restricted sense. Animal art shouldabove all be regarded as a special kind ofexpression, used to convey different mes-sages. In this respect, animal art was anal-ogous to skaldic poetry, as several scholarshave pointed out. Both genres were difficultforms of expression, which demanded specialcraftsmen to be created and good backgroundknowledge to be understood. In fact a fewrune-carvers called themselves skald (poet).

Animal art was a dominant form of artisticexpression in Scandinavia from the 5'" cen-tury until the 12'" century. More than 100different styles in animal art have been de-fined in the 20'" century and these definitionshave sometimes been regarded as only mod-ern classifications, with little relevance forpast "realities" (cf. Karlsson 1983:92ff ).Thesimilarity between animal art and skaldicpoetry instead speaks in favour of some kindof relevance for the different styles. They canbe regarded as visual stanzas or visual metresby analogy with the metric forms of skaldicpoetry that are described by Snorri in hisHattatal (Snorres Edda, pp. 219-265.) Fromthis point of view different styles of animalart, like the Mammen style and the Urnesstyle, may be regarded as analogous to poeticmetres like Drottkva:tt and Fornyrtlislag (cf.Lie 1952; Bugge 1953). It is even possiblethat the different art styles had the same socialconnotation as some of the poetic metres.

Levels of understandingThe proposed associative close reading of

rune-stones presupposes some kind of intel-lectual understanding of the monuments, andit is uncertain how common such an under-standing might have been in the 11'" century.It is quite clear from Norse literature thatsome persons were "unwise", i.e. had not theknowledge to understand the metaphors andallusions in skaldic poetry. Therefore, themessage that the rune-stones conveyed hasprobably been apprehended in several differ-ent ways, at least from an individual point ofview. The rune-stones could be understoodat a monumental point of view, as brightlycoloured monuments, signifying importantplaces and persons in the landscape. Theycould also be understood from an icon-ographic level, since the images and thenumber and size of snakes indicated differentrelations between persons. A literate personcould further understand the runic inscriptionin a literal sense, by reading the text. Andfinally, a person well-informed in Norsemythology could understand the monumentfrom a poetic point of view.

However, it is questionable whether theunderstanding of the rune-stones was indi-vidually based in that way. From the PoeticEdda as well as the Snorri's Edda it is clearthat interpretations of mythological know-ledge were enacted collectively, sometimesas competitive wisdom-dialogues betweentwo persons (Holtsmark 1960). Thus, it isplausible that the understanding of the rune-stones should be apprehended as a kind ofinterpretative context, including severalpersons contributing to a joint interpretation,as a kind of collective act.

A final problem, though, is that such anunderstanding partly must have been basedon the eye as the important active sense. It isonly through vision that the sophisticatedcompositions of the rune-stones were acces-sible. However, this active use of the eye isdifficult to combine with a dominant oralculture, where texts usually were accessiblethrough listening. And most probably Scan-dinavia in the late Viking Age was a thor-

Current Swedish Archaeology, Vol. 8, 2000

Re-reading Embodi e&l tests 27

oughly oral culture (cf. Goody 1977, 1986;Clanchy 1979; Ong 1982; Schousboe &Trolle Larsen 1989; Lindell 1994). Thesolution to this problem can be to regard therune-stones with their complex compositionsas analogies to Arabic calligraphy on monu-ments (Blair 1989; Baer 1998).That type ofpublic writing was not intended for silentreading, but created as contemplative texts,which needed long interactive interpretations,probably involving several persons. In thatway, the rune-stones with their intricate com-positions can be viewed as contemplativemonuments, which demanded that the per-sons passing them had to stop and discusstheir meaning. The sophisticated composi-tions accordingly underlined the very aim of

the monuments, namely to be memorials.Thus, function and meaning of the rune-stones are clearly intertwined, and theirdesign, text, image and location thereforeshould not be separated in future research.

English revised by Alan C&ozie&:

A CKNOWLEDGEMENTSThis paper has been presented at seminarsin Malm, Bergen, Stockholm, Uppsala andLund during 1998-99. Comments and ques-tions raised on these occasions have beenve&y useful in writing the article. Moreover,f want to thank Stefan Brinl', Sanne Hoiiby-Nielsen, Jrn Staecker anal Jes kVienbe&g forcomments on the manuscript.

REFERENCESberg, N. 1921. Stil III och Jcllingestil. Foni-

vnnen. Pp. 63-82.1922. Stil II. Fornvnnen. Pp. 1-26.

Adam av Bremen. 1984. Histoi. ieli oi&1 Hillbu& g-stiftet och dess biskopar. versatt av EmanuelSvenberg. Stockholm.

hln, M. 1997.Runristaren pir. En monografi.Runrn 12. Uppsala.

Ambrosiani, B. 1964. Fornlmningar och be-byggelse. Studier i Attundnlnnrls och Sder-trns frhistoria. Kungliga Vitterhets, His-torie och Antikvitetsakademien, Monografier.Uppsala.

Andersson, H. 1995. Kristendv&en. s genoinbi. otten stiidie nv runstennn&a i Upplnnds-B&o.Duplicated seminar paper. Institute o fArchae-ology, University of Uppsala. Uppsala.

Andersson, L. 1992. Rnsteai som utinrcl. fi.samluillsfrndiii&g. En stiidie av ninstenaioch samhlle u&ider l000-talet i Milardalen,Duplicated seminar paper. Department of Ar-chaeology, Stockholm University. Stockholm.

1989. Drrar till frgngna myter. En tolkningav de gotlndska bildstenarna. In: Andrn, A.(Ed). Medeltidens fdelse. Symposier pKrapperups borg 1. Krapperup. Pp. 287-319.

1993. Doors to other worlds. Scandinaviandeath rituals in Gotlandic perspectives. Jour-nal of European Archaeology 1. Pp. 33-56.

1998. Bet&veen Ai ti fiicts aiid Tests. HistoricalAichaeology in Global Pe&spective. Con-tributions to Global Historical Archaeology.New York.

1999. Landscape and settlement as utopianspace. Im Fabech, C. & Ringtved, J. (Edsl.Settle&nent and Landscape. rhus. Pp. 383-393.

Arend, Q. 1978. ybiR risti runaR. Zur Spracheeines upplndischen Runenmeisters. Amster-dame& Beitrge zu&. lteie Germanistil- 13.Pp. 35-67.

Axelsson, J. 1993.Melhmsvenska nimistai. e. Fr-teckning ver signeiade ocli nttiibei. ade in-skrifter. Runrn 5. Uppsala.

Ba.ksted, A. 1952. Mlnme& og tnrldniner. Rne-magisl e stiidiei. Nationalmuseets skrifter. Ar-ka:ologisk-Historisk Ra:kke IV. Copenhagen.

Baer, E. 1998. Islaiic Oniaient. Edinburgh.Baines, J. 1989. Communication and display; the

integration of early Egyptian art and writing.Antiquity 63. Pp. 471-482.

Brard, C. , Bron, C. , Durand, J.-L., Frontisi-Ducraux, F., Lissarrague, F., Schnapp, A. &Vernant, J.-P. 1989. A Citv of' Images.Iconography and Societv in Ancient Greece.Princeton.

Berg, S. 1987. Och pir ristade. En studie k&ingen uppl ndsl- ni&&rista&. e. Duplicated seminar

Cu&re& $&vedish Aiveturerrtgu Vrrt. , 2000

28 Anders Andrn

paper. Department of Archaeology, Stock-holm University. Stockholm.

Bergman, B. 1948. Upplndsk run- och bildstens-ristning. En konsthistoriskstudie. Stockholm.

Berlo, J. C. (Ed). 1983. Text and 1mage in Pre-Columbian Art. Essays on the Interrelation-ship of the Verbal and Visual Arts. BritishArchaeological Reports. International Series180. Ox ford.

Bjurstrm, A. 1996. Vem reste runstenarna? Enstudie av fem genealogiskt kopplade runstens-grupper i Hbo och Arlinghundra hrad.Duplicated seminar paper. Institute ofArchae-ology, University of Uppsala. Uppsala.

Blair, S. S. 1989. Legibility versus decoration inIslamic epigraphy. The case of interlacing. In:Laum. , I. (Ed). IVorld Art. Themes of Unity inDiversity. Acts of the Congress of the Historyof Art 11. Pennsylvania. Pp. 329-334.

Brate, E. 1925. Svenska runristare. KungligaVitterhets, Historie och Antikvitets Akade-miens Handlingar 33:5-6. Stockholm.

Brondsted, J. 1920. Nordisk og fremmed Orna-mentik i Vikingetiden, med sa:rlig Henblikpaa Stiludviklingen i England. Aarbetger fornordisk Oldkyndighed og Historie III:10 Pp.162-282.

1924. Early English Ornament. The sources,development and relation to foreign styles ofpre-roman ornamental artin England. London& Copenhagen.

Bugge, A. 1953. Norske stavkirker. Oslo.Carlqvist, K. 1977. Vad sger runstenarna?

Meddelande /hn arkivet fr folkets historia5:4. Pp 8-21.

Carlsson, M. 1985. Thegn Drng Jarl-Huskarl. Arl eologiska studier utifrn run-stenar i stra Mellansverige. Duplicatedseminar paper. Department of Archaeology,Stockholm University. Stockholm.

Christiansson, H. 1959. Sydskandinavisk stil.Studies i ornamentiken p de senvikingatidarunstenarna. Uppsala.

1995.Den onda ormen. Runstensornamentikenskristna budskap. Ett tolkningsfrsk. TOR 27.Pp. 449-457.

Christophersen, A. 1982. Drengs, Thegns, Lend-men and Kings. Meddelanden frn LundsUniversitets Historiska museum, New series,Vol. 4. Pp. 121-134.

Claesson, E. 1982. Gtalandskapens runstenar.Duplicated seminar paper. Institute ofArchae-

ology, University of Lund. Lund.Clanchy, M. T. 1979. From Memory to IVritten

Record. England l 066-7307. London.Crocker, K. R. 1982. "Fotr risti ", A Run-

ographer s Style in the Context of Eleventh-Century Upplandic Memorial Art. Ann Arbor.

Danbolt, G. 1989. Hva bildet kan fortelle ommotet mellom hedenskap og kristendom. In:Andrn, A. (Ed). Medeltidens fdelse. Sym-posier p Krapperups borg 1. Krapperup. Pp.233-260.

DR = Danmarks runeindskrifter I-II. Ved Jacob-sen, L. &. Moltke, E. under medvirkning afBa:ksted, A. & Nielsen, K. M. 1941-42.Copenhagen.

Ekblad, S. 1997. Den kristna missionen i Upplandspeglad i runstensmaterialet. Duplicatedseminar paper. Institute of Archaeology,University of Uppsala. Uppsala.

Elgh, S. 1987. Runstenar Placering och krist-ningsgrad. Danmark och Tiundaland en

jmfrelse. Duplicated seminar paper. Insti-tute of Archaeology, University of Uppsala.Uppsala.

von Friesen, O. 1913. Upplands runstenar. Enallm nfattlig fversikt. Uppsala.

1928. Runorna i Sverige. Uppsala.1933. De svenska runinskrifterna. Nordisk

t.ultur VI, Runorna. Stockholm.Fuglesang, S. H. 1978. Stylistic Groups in Late

Viking Age. In: Lang, J. T. (Ed). Anglo-Saxanand Viking Age Sculpture and its Context.British Archaeological Reports, British Series49. Oxford. Pp. 205-223.

1980. Some Aspects of the Ringerike Style. Aphase of' l l"' century Scandinavian Art.Medieval Scandinavia, Supplements 1.Odense.

1981.Crucifixion Iconography in Viking Scan-dinavia. In: Bekker-Nielsen, H. , Foote, P. &Olsen, O. (Eds). Proceedings of the EighthViking Congress. Medieval Scandinavia, Sup-plements 2. Odense. Pp. 73-94.

1986. Ikonographie der skandinavischenRunensteine der jiingeren Wikingerzeit. In:Roth, H. (Ed). Zum Problem der Deutungfruhmittelalterlicher Bildinhalte. Verffent-lichungen des Vorgeschichtlichen Seminarsder Phil lips Universitet, Marburg. Sig-maringen. Pp. 183-210.

Go = Gotlands runinskrifter I-II, granskande ochtolkade av Jansson, S. B. F., Wessn, E. &

Current Srvedish Archaeotogy, Vol. 8, 2000

Re-reading Embodied Texts 29

Svrdstrm, E. Sveriges runinskrifter 11-12.Stockholm 1962-1978.

Goody, J. 1977. The Domesti cation of the SavageMind. Cambridge.

1986. The Logi c of Wri ti ng and the Or-ganization of Society. Cambridge.

Grslund, A.-S. 1985. Den tidiga missionen iarkeologisk belysning problem och syn-punkter. TOR 20. Pp. 291-314.

1988. Runstenar, bygd och gravar. TOR 21. Pp.241-262.

1989. "Gud hjlpe nu vl hennes sjl". Omrunstenskvinnorna, deras roll vid kristnandetoch deras plats i familj och samhlle. TOR22. Pp. 223-244.

1991.Runstenar om ornamentik och datering.TOR 23. Pp. 113-140.

1992. Runstenar om ornamentik och dateringII. TOR 24. Pp. 177-201.

1994. Rune Stones On Ornamentation andChronology. In: Ambrosiani, B. & Clarke, H.(Eds). The Twelfth Viking Congress. BirkaStudies 3. Stockholm. Pp. 117-131.

1996. Runstenskvinnorna nnu en gng. TOR27. Pp. 459-474.

Gronvik, O. 1981.Runene p Tunesteinen. Oslo. 1985. Runene p Eggjasteinen. En hedensk

gravinnskrift fra slutten av 600-talet. Oslo.Gurevich, A. J. 1985. Categories of Medieval

Culture. London.Gustavson, H. & Selinge, K.-G. 1988. Jarlabanke

och hundaret. Ett arkeologiskt/runologisktbidrag till lsningen av ett historiskt tolk-ningsproblem. Namn och bygd 76. Pp. 19-85.

Hagenfeldt, S. E. & Palm, R. 1996. SandstoneRunestones. The use of sandstone for erectedrunestones. Runica et media:valla 2. Stock-holm.

Hallencreutz, C. F. 1982a. Runstenarnas Maria.En studie av kristendomens versttning tillrunsvenska. Svensk missionstidskrift 70:2. Pp12-22.

1982b. Runstenarnas teologi. Signum 8:1. Pp.9-15.

Hanks, W. F. & Rice, D. S. (Eds). 1989. Wordand Image in Ma& a Culture. Explorations inLanguage, Writing and Representation. SaltLake City.

Haseloff, G. 1981. Die gerntanische Tierorna-mentik der Vll erwanderungszeit. Studien zuSalin s Stil I. Berlin & New York.

Hedeager, L. 1997. Skygger af en anden virkelig-

hed. Oldnordiske myter. Copenhagen.1999.Skandinavisk dyreornamentik. Symbolskrepra:sentation af en forkristen kosmologi. In:Fuglestvedt, I., Gansum, T. & Opedal, A.(Eds). Et hus med mange rom. Vennebok tilBj om Myhre p 60-rsdagen. Stavanger. Pp.219-237.

Herschend, F. 1994. The Recasting of' a SymbolicValue. Three Case Studies on Rune-stones.Occasional Papers in Archaeology 3. Uppsala.

1998a. The Idea of the Good in Late Iron AgeSociety. Occasional Papers in Archaeology 15.Uppsala.

1998b. ubiR, ybiR, ybir r det U485 Ofegpir? Fornvnnen. Pp. 97-111.

Hildebrand, H. 1884. Det nya svenska runverket.Kongliga Vitterhets, Historie och Antiqvitets-akademiens Mnadsblad 13. Pp. 31-38.

Holmbck, . & Wessn, E. (Eds). 1943. SvenskaLandskapslagar 4. Sknelagen och Gutalagen.Stockholm.

Holmqvist, W. 1951.Viking Art in the 11'" Cen-tury. Acta Archaeologica XXII. Pp. 1-56.

1955. Germanic Art during the First Millen-nium A. D. Vitterhetsakademien, Handlingar90. Stockholm.

Holtsmark, A. 1960. Gudediktning. Kultur-historiskt lexikon fr nordisk medeltid 5.Malm. Pp. 531-536.

Hubler, F. 1996. Schwedische Runendichtung derWikingerzeit. Runrn 10. Uppsala.

Hyenstrand, . 1974. Centralbygd randbygd.Strukturella, ekonomiska och administrativahuvudlinjer i mellansvensl- yngre j rnlder.Studies in North-European Archaeology 5.Stockholm.

1996. Ljeonet, draken och korset. Sverige 500-l000. Lund.

Jesch, J. 199 la. Women in the Viking Age.Woodbridge.

1991b. Who was hulmkir? Double appositionin the Ramsund inscription. Arkiv fr nordiskfilologi 106:125-136.1994 Runic Inscriptions and Social History.

Some Problems of Method. In: Knirk, J. E.(Ed). Proceedings of the Third InternationalSymposium on Runes and Runic Insctiptions.Runrn 9. Uppsala. Pp. 149-162.

Johansen, B. 1997. Ormalur. Aspekter av tillvarooch landskap. Stockholm Studies in Archae-ology 14. Stockholm.

Johansson, L. 1996. "Retta" och "rista ". Tv

Current Swedish Archaeologv, Vol. , 2000

30 Anders And&n

&n&nsvenska verb i Mlardalen. Duplicatedseminar paper. Institute of Archaeology, Uni-versity of Uppsala. Uppsala.

Kllstrm, M. 1992. Den upplndsl. e runristarenViste. En l&onologisklkorologisk studie.Duplicated seminar paper. Department ofArchaeology, Stockholm University, Stock-holm.

1998. Torbjrn skald och Torbjrn studie&.I'ring tv mellansvenska m&nrista&e. Dup-licated seminar paper. Institute of Scan-dinavian Languages, Stockholm University.Stockholm,

Karlsson, L. 1983 Nordisl' fo& n& om dj 0&0& nan&enti k. Statens historiska museum.Studies 3. Stockholm.

Klindt-Jensen, O. 1975.A Hi stor ofScandi navianA& chaeology. London.

Kristoffersen, S. 1995. Transformation in Mig-ration Period Animal Art. Norwegian A&chae-ological Review 28: l. Pp. 1-17.

Kyhlberg, K. 1983.Kvi nno&. och mn i ppl ndhl a& uninskrifle& en metodstudie. Duplicatedseminar paper, Department of Archaeology,Stockholm University. Stockholm.

Lager, L. 1995. Kors p svensl. a &. unstenar.Duplicated seminar paper. Institute ofArchae-ology, University of Uppsala. Uppsala.

Lagerroth, U.-B., Lund, H. , Luthersson, P. &Mortensen, A. (Eds). 1993.I musernas jtnst.Sn&dier i konstarternas interrelationer. Stock-holm & Stehag.

Lagman, S. 1990. De stungna runo&. na. Anv'nd-ning och jludvrdei runsvenska steninslnifie&.Runrn 4. Uppsala.

Larsson, M. G. 1990. Runstena& och utlands-frder. Aspel ter av det senvikingatida sam-hllet med utgngspunkt i de fasta forn-lmningarna. Acta Archaeologica Lundensia.Series in 8', No. 18. Lund.

1996. Tvegifte i Tby? Ngra synpunkter pJarlabankestenarnas datering och placering.Fornvnnen. Pp 143-151.

1997. Frn stormannag&d till bondby. Enstudie av mellansvensk bebyggelseutvecklingfin ld&e jrnlder till medeltid. ActaArchaeologica Lundensia, Series in 8', No 26.Lund.

Lie, H. 1952. Skaldestilstudier. Maal og Minne1952. Pp. 1-92.

1963. Kenningar. Kulturhisto&isl t le&rilron frno&disk medeltid 8. Pp 375-381. Malm.

Lindblad, E. & Wirtn, K. 1992. Ko&sbandhstena&.en Inonologisk studie. Duplicated seminarpaper. Institute of Archaeology, University ofUppsala. Uppsala.

Lindell, l. (Ed). 1994. Medeltida skrift- och spr I--kultur. Nio frelsningar fin ett syn&posiun&i Stockholm v&ren 1992. Runica et Media:valiaOpuscula 2. Stockholm.

Lindkvist, T. 1988. Plundring, skatter och denfeodala statens fi an&vrrt. Organisatoriskatendense& i Sve&i ge ande& ve& g n gen fl nvilingatid till tidig medeltid. Opuscula His-torica Upsaliensia l. Uppsala.

Lindqvist, S. 1931.Yngre vikingastilar. No&diskKltu& XXVII. Stockholm. Pp. 144-179.

Lindstrm, H. 1992. Senvilingatida runstenar iSkandinavien, Kristna eller arvsr ttsligadokument? En jmfrelse mellan Sdern&an-land och Skne. Duplicated seminar paper.Institute of Archaeology, University of Lund.Lund.

Ljungberg, H. 1938. Den nordisla religionen ochkristendon&en. Studie&' ver det nordisl a&eligionsskiftet under vikingatiden. Nordiskatexter och underskningar 11. Stockholm &Copenhagen.

Lnnroth, E. 1982. Administration och samhllei 1000-talets Sverige. Bebyggelsehistorisk tid-skrift 4, Pp. 10-23,

Marcus, J. 1992. Mesoamerican IJtriting Systems.Propaganda, Myth, and History in FourAncient Civilisations. Princeton.

Meijer, J. 1992. Planning in runic inscriptions.Blandade &.unstudier I, Runrn 6. Uppsala.Pp. 37-66.

Moltke, E. 1974. The Jelling monument in thelight of the runic inscriptions. MediaevalScandi navia 7:183-208.

1976. Runernei Danmarl' og deres oprindelse.Copenhagen.

Mostrm, J. 1998. Brostenar och sjlabroar, 0&=ganisation och ideologi bakom brobyggandei Norden under sen vikingatid och tidig medel-tid. Duplicated seminar paper, Department ofArchaeology, Stockholm University. Stock-holm.

Miiller, S. 1880. Dyreornamentiken i Norden.Aa& bege& fo&' nordisl- Oldkyndighed ogHistorie 1880. Pp. 185-405.

NIyR = Norges innsl rifte& med de vngre &'uner I-5. Utgivne ved Olsen, M. Norges indskrifterindtil reformationen 2. Oslo 1941-1960.

Current Snedish Archaeology, Vol. 0, 2000

Re-reading Embodi ett Texts 31

g = ste&gtlands &u&ti&tskrifte& 1-3, granskadeoch tolkade av Brate, E. Sveriges runinskrifter2. Stockholm 1911-1918.

l = lands &aminskrifter 1-2, granskade ochtolkade av Sderberg, S. & Brate, E. Sverigesruninskrifter 1. Stockholm 1900-1906.

Olason, V. 1989. Den frie mannens selvforstelsei islandske sagaer og dikt. In: Andrn, A. (Ed).Medeltidens ftdelse. Symposier p Krappe-rups borg 1. Pp. 277-286.

Ong, W. 1982. Orality and Literacy. The Tech-nologizing of the Hord. London.

Orsnes, M. 1966. Form og stil i Sydskandinaviensyngre germansl. ej e&.nalder. Copenhagen.

1969. Sudskandinavische Ornamentik injiingeren germanischen Eisenzeit. Acta A&-chaeologica XL. Pp. 1-121.

Owe, J. 1995. Svensl- runbibliografi I880-I993.Runverket, Riksantikvariembetet. Stock-holm.

Palm, R. 1992. Runor och regionalitet. Studierav variation i de nordiska minnesinskrtfterna.Runrn 7. Uppsala.

Peterson, L. 1994. Svens/t runordsregister. Run-rn 2. Uppsala.

Petr, B. 1984. Arkeologiska underskningar pLov 4. Beby ggelsehistorisk analys. ActaUniversitatis Stockholmiensis. Studies inNorth European Archaeology 10. Stockholm.

1993. Inheritance Systems and Farm Agglo-merations Forming Early Villages in the LateIron Age. In: Arwidsson, G. et al. (Eds).Sources and Resources. Studies in Honour ofBirgit Arrhenius. PACT 38. Rixenart. Pp. 93-106.

Philippa, M. 1977. Die Inschriften des schwe-dischen Runenmeister Balli. AmsterdamerBeitrgezur lte&en Germanistik 12. Pp. 23-45.

Ragnesten, U. 1978. Runstenar. En empiriskstudieav sociala frltllanden i runstenstexternamed en spatialjmfrelse. Duplicated seminarpaper. Institute of Archaeology. University ofGothenburg. Gothenburg.

Randsborg, K. 1980. The Viking Agein Denmark.The Forn&ation of' a State. London.

Russell, J. C. 1994. The Gerntanization of' EarlyMedieval Christianity. A Sociohistorical Ap-proach to Religious T&ansformation. Oxford.

Salberger, E. 1978. Runsvenska namnstudier.Stockholm Studies in Scandinavian Philology,New Series 13. Stockholm.

Salin, B. 1904. Die altgerntanische Tlrie&. o&.na-mentilz Stockholm.

Smunds Edda, versatt frn islndskan av Brate,E. Stockholm 1913.

Sandahl, C. 1996, Kristendom, arv och utlands-frder en studie av s&nlndskct runstena&.Scripta Minora 31. Vxj.

Sawyer, B. 1988. Property and Inheritance inViling Scandinavia. The Runic Evidence.Occasional Papers on Medieval Topics 2.Alingss.

1989. Det vikingatida runstensresandet i Skan-dinavien. Scandia 55:2. Pp. 185-202.

1991. Viking-Age Rune-Stones as a CrisisSymptom. Nonvegian Archaeologicttl Revie:24;2. Pp 97-112.

Sawyer, B. &. Sawyer, P. 1993. Medieval Scan-dinavia. From Conversion to Reforntationci rea 800-l500. Minneapolis & London.

Schousboe, K. 8c Trolle Larsen, M. (Eds). 1989.Literacy and Society. Copenhagen.

Schuck, H. 1935-1944. Kungliga Vitterhet, His-tori e och Antikvitets Akademien I-VIII.Stockholm.

Segelberg, E. 1983. Missionshistoriska aspekterp runinskrifterna. Kyrkohistorisk &.ssl. rift 83.Pp. 45-57.

Shetelig, H. 1909. Urnesgruppen. Det siste avsnitav vikingetidens stilutvikling. Aa&sbe&. etningfi.a Foreningen til norske Fo& tidsn&i ndes-rnerkers Bevaring.

1949. Classical Impulses in Scandinavian Artfrom the Migration Period to the Viking Age.I nstituttet for sammenlignende kulturforsk-ning. Serie A:XIX. Oslo.

Sjberg, A. 1985. Orthodoxe Mission in Schwe-den im 11. Jahrhundert? In: Lindquist, S.-O.(Ed). Society and trade in the Baltic duringthe Viking Age. Acta Visbyensia Vll. Visby.Pp. 69-78.

Sna:dal Brink, T. & Wachtmeister, I. 1984. Ru-stena&' i Sdermanland. Vgvisa&e till &ntn-ristningar i Sdermanlands ln. Srmlndskahandlingar 40. Nykping.

Snorres Edda. versttning frn islndskan ochinledning av K. G. Johansson och M. Malm.Stockholm 1997.

S = Sdermanlands runinskrifter 1-4, granskadeoch tolkade av Brate, E. & Wessn, E. Sve-riges runinskrifter 3. Stockholm 1924-1936.

Sderberg, S. 1905. Om djurornamentiken underfolkvandringstiden. Antiqvarisl' ti dsk&i ft f&

C&r&rent Srvedish Archaeolog Vol. 8, 2000

32 Anders Andrn

Sverige XI:3. Pp. 1-93.Sderqvist, E. 1993. Guds moder, ande och sjl.

stligt inflytande i Mlardalens runstens-bnder. Duplicated seminar paper. Instituteof Archaeology, University of Uppsala. Upp-sala.

Stille, P. 1999. Runstenar och runristare i detvikingatida Fjdrundaland. Runrn 13. Upp-sala.

Strid, J.-P. 1987. Runic Thegns and Drengs. In:Runor och runinskrifter. Fredrag vid Riks-antikvariembetets och Vitterhetsakademienssymposium 8-11 september 1985. KungligaVitterhets, Historie och Antikvitets Akade-mien, Konferenser nr 15. Stockholm. Pp. 301-316.

Sundqvist, M. 1996. Treglerunstenari Uppland.Duplicated seminar paper. Institute ofArchae-ology, University of Uppsala. Uppsala.

Svrdstrm, E. 1936. Johannes Bureus'arbetenom svenska runinskrifter. Stockholm.

Svenbro, J. 1993. Phrasikleia. An anthropologyof reading in ancient Greece. Ithaca, N. Y.

Thompson, C. W. 1975. Studies in UpplandicRunography. Austin 8c London.

Thrainsson, Th. 1994. Vid Urdarbrunnen. OmUrnesstilens ikonografi. Duplicated seminarpaper. Institute of Archaeology, University ofUppsala. Uppsala.

U = Upplands runinskrifter 1-4, granskade ochtolkade av Wessn, E. 8t Jansson, S. B. F.Sveriges runinskrifter 6-9. Stockholm 1940-1958.

Varenius, B. 1998. "han gde bo och skeppslid".Om rumslighet och relationer i vikingatid ochmedeltid. Studia Archaeologica UniversitatisUmensis 10. Ume.

Wamers, E. 1999. Harald Bltands db og denstore Jellingsten. Vejle amts rbog 1999:47-66.

Wideen, H. 1955. Vstsvenska vikingatidsstudier.Arl eologiska kllor till Vneromrdets kultur-historia under yngre jrnlder och ldstamedeltid. Skrifter utgivna av Gteborgsarkeologiska museum 2. Gteborg.

Wijkander, K. 1983. Kungshgar och socl'en-

bildning. Studier i Sdermanlands admini-strativa indelning undet vikingatid och tidigmedeltid. Srmlndska handlingar 39. Ny-kping.

Williams, H. 1990. rfsrunan. Anvndning ochjludvrde i runsvenska steninskrifter. Runrn3. Uppsala.

1993. Mres runinskrifter. In: Williams, H.(Ed). Mres kristnande. Projektet Sverigeskristnande. Publikationer 2. Uppsala. Pp. 85-116.

1996a. Vad sger runstenarna om Sverigeskristnande? In: Nilsson, B. (Ed). Kristnandeti Sverige. Gamla kllor och nya perspektiv.Projektet Sveriges kristnande. Publikationer5. Uppsala. Pp. 45-83.

1996b. Runstenstexternas teologi. In: Nilsson,B. (Ed). Kristnandet i Sverige. Gamla klloroch nya perspektiv. Projektet Sveriges krist-nande. Publikationer 5. Uppsala. Pp. 291-312.

Wilson, D. M. 1995. Vikingati dens konst. Signumssvenska konsthistoria 2. Lund.

Wilson, L. 1994.Runstenar och kyrkor. En studiemed utgngspunkt frn runstenar som p-trffats i kyrkomijl i Uppland och Sder-manland. Occasional Papers in Archaeology8. Uppsala.

Zachrisson, T. 1994. The Odal and Its Manifes-tation in the Landscape. Current SwedishArchaeology 2. Pp. 219-238.

1998.Grd, grns, gravflt. Sammanhang kring delmetalldep er och runstenar frn vi kinga-tid och tidigmedeltid i Uppland och Gstrik-land. Stockholm Studies in Archaeology 15.Stockholm.

Zanker, P. 1988. The power of'images in the ageofAugustus. Ann Arbor.

Sources of illustrationsFigure I is a reproduction of an etching in SverigesRuninskrifter, whereas figures 2, 3, 4, 5D, 6D, 7,10C, 10E, and 11 are drawings based onantiquarian images and photos published inSveriges Runinskrifter. All other figures are basedon figures in Sna:sdal Brink k. Wachtmeister1984.

Current Snedish Archaeology, Vol. 8, 2000