Gotland is a limestone island in the middle of the Baltic, and its limestone splits naturally into flat slabs. Somewhere in the fifth century people there began cutting pictures into those slabs and standing them upright — beside roads, near farms, at burial grounds — and they went on doing it for six hundred years. Several hundred survive. They are found almost nowhere else, they have no clear ancestor, and when the tradition stopped it left no descendants. It is one of the strangest self-contained bodies of narrative art in Europe.
What they look like
The earliest class, from roughly the fifth to the seventh century, is small and mostly abstract: discs and whorls, entwined beasts, occasionally a figure. The Smiss stone from När parish belongs here, with its seated figure holding a snake in each hand under a knot of interlaced animals.
Then, in the eighth century, the form changes into the one everybody pictures. The slab grows tall — some over three metres — and takes a waisted silhouette that widens toward a rounded top, and the whole face is given over to figures. A woven band of interlace frames the slab on all four edges, and more bands run horizontally across it, dividing the surface into two, three or four registers. Each register holds its own scene. You read them from the top down.
The carving is shallow. An outline is incised with a chisel and the ground around the figure is picked back only slightly, so the picture stands a few millimetres proud at most. That shallowness matters: there is no modelling to be had at that depth, so the figures are pure silhouette, cut in strict profile, with a single punched dot for an eye and nothing else in the face at all.
And they were painted. Pigment traces recovered from the incisions of several stones support what the shallow carving already implies — that these were never meant to be read as grey relief. The picture is flat opaque colour bounded by a cut line, with the colour sitting down in the cut where it is best protected.
The repertoire
A small number of images recur across the corpus with remarkable persistence.
The most frequent is the ship: a clinker hull drawn as three or four stacked arcs, often with a beast-head stem, under a large sail rendered as a flat lattice or diamond grid, with a row of near-identical heads showing above the sheer. The sail's grid is usually taken to represent the reinforcing ropes across the cloth, and it is one of the tradition's signature marks.
The second is the arrival. A rider comes in on horseback and is met by a standing figure holding out a drinking horn. On the Tjängvide stone the horse has eight legs, all of them shown, which is why the scene is so often read as Odin's horse Sleipnir carrying a rider to a hall. The reading is old and plausible and not certain, and this is the general condition of picture-stone scholarship: the stones predate every written source for the stories they may be telling by two to four centuries, and the connection runs backwards from Icelandic manuscripts to Baltic limestone rather than forwards.
Around these cluster the rest: armed men in procession, single combats, buildings shown as flat rectangles with no interior, knots of gripping beasts, and on the Stora Hammars slabs at Lärbro, scenes crowded so densely that four bands of incident stack up one slab.
The rules the carvers worked to
Read across the corpus, the tradition turns out to be extremely consistent about a handful of things, and all of them are refusals.
- No perspective. A register has a base line and a ceiling and nothing between them recedes. Buildings do not have depth; ground does not have a horizon.
- No distance-scaling. Nothing is drawn smaller for being further away. Size states importance instead, and every figure keeps both feet on the base line of its band.
- No modelling. There is no shading, no hatching, no cast shadow. A shape is filled or it is not.
- No faces. With one dot for an eye, a profile head cannot identify anyone. Identity has to be carried by attribute and outline — what a figure holds, wears or is shaped like.
- No frames, except the woven ones. The division between registers is a plaited band, and the border around the slab is the same band carried round all four edges.
Interlace, and the styles that name it
That woven band is not decoration in the incidental sense; it is the tradition's structural device, and it belongs to a wider Scandinavian ornamental sequence that art historians divide into named styles — Borre, Jelling, Mammen, Ringerike, Urnes — running from the ninth century to the early twelfth. Each is defined by how its animals grip, loop and taper. Ringerike ornament sends long tendrils off its beasts; Urnes, the last and most elegant, reduces everything to slim ribbon-animals threading through one another in figure-eight loops, and takes its name from the carved portal of the stave church at Urnes in western Norway.
One motif deserves its own note. The ring-chain — a plait that repeats a closed ring through a running braid — appears on cross-slabs on the Isle of Man carved by a man who signed his work. Gaut Björnsson's inscription claims he made this cross and all in Man, and the motif has carried his name ever since. The Manx slabs are also where the carved tradition meets Norse story most explicitly: the cross at Kirk Andreas raised by Thorwald shows a figure with a spear and a bird at his shoulder, one foot deep in the mouth of a wolf, cut on the same stone as Christian imagery.
The same collision stands in a Cumbrian churchyard, where the Gosforth Cross carries Norse mythological scenes and the Crucifixion on one slender red sandstone shaft.
How the stones came back
For most of their history the stones were building material. They were broken up for church floors, walls and thresholds, which is paradoxically why so many survive at all — a slab face-down in a floor is a slab not weathering in a field. Ardre VIII came out of the floor of Ardre church.
The corpus as a field of study begins with Sune Lindqvist, whose two-volume Gotlands Bildsteine of 1941–42 remains the reference every later study argues with. Lindqvist also produced the painted reconstructions that fixed the stones in the popular imagination in colour, by filling the carved lines with paint and photographing the result. The reconstructions are now treated with real caution: the method imposed a confident reading on faint and ambiguous surfaces, and it was not kind to the stones. What survived the criticism is the underlying claim, which the pigment evidence supports — that these objects were coloured.
Later work has largely been an argument about what is actually on the surfaces. Erik Nylén and Jan Peder Lamm's Stones, Ships and Symbols brought the corpus to a general readership; more recent scholarship has gone back to the slabs with laser scanning and raking-light imaging and recovered carved detail that a century of eyes and rubbings had missed, occasionally changing what a scene is understood to show.
What the tradition actually invented
Strip away the mythology, which may or may not be there, and what remains is a formal achievement that has nothing to do with Norse gods. The picture stones divide a single surface into discrete framed moments, set them in a fixed reading order, part them with a consistent division device, and carry figures across from one to the next.
That is sequential narrative art, and it is worked out on a Baltic island in the eighth century by carvers who had no model for it, using a visual system — flat fill, no perspective, silhouette identity, size as emphasis — that a modern comics artist would recognise on sight. The stones did not influence comics. They arrived at the same solution, alone, and then the tradition stopped and the solution had to be found again.
If you would like to see that grammar carrying a story again, our Picture Stone comic style renders a chess game as a painted slab in stacked bands: flat silhouettes on their base lines, four values, interlace instead of gutters, and identity carried by attribute the way the carvers had to carry it.
