One of the best-preserved extant boards related to hnefatafl and related variants is this one, unearthed during the 1989 excavation of the Trondheim Public Library site in Trondheim, Norway. Unlike the Gokstad ship board fragment, whose markings are confusingly inconsistent with what we know of hnefatafl, this board shows a symmetry in near perfect agreement with hnefatafl. Another game, identified variously as “tables” or “tric-trac” (by either name a medieval precursor to backgammon), is marked on the reverse. More than half of this incised wooden board is preserved, and the remains have been dated to c. 1100-1150. This fragment measures 25.6cm along the intact length, with a wooden trim measuring 7-8mm high, featuring mitered corners, attached with dowels. The board would have originally been square, measuring 25.6cm x 25.6cm (approx. 10in x 10in), with nine marked spaces in a cruciform pattern. The careful symmetry and attached trim suggest this was a professionally produced game board, in contrast to the amateurish scratchings of the Toftanes board fragment. That specimen appears to have been a used serving platter, crudely incised with an asymmetrical lattice of 13 x 14 irregular cells, the “center” cell marked with an incised cross, indicating that a 13 x 13 cell lattice was likely intended.
Another good photograph of the hnefatafl board fragment from the Trondheim Public Library site may also be found in Roesdahl & Wilson’s From Viking to Crusader (1992) Cat. No. 572 on p. 378.
One aspect of this board which merits further examination is the position of the marked spaces on the hnefatafl side of the board. As we can see, six spaces are marked with an X, forming three arms of a cross, with a double X through the center space, and we are led to assume the missing portion would have had two more marked spaces to complete the cross. The twin marked spaces of each arm of the cross leave two blank spaces between them and the center and a single unmarked space along the edge.
But what did these markings mean, and how did they affect game play? A number of possibilities may be considered.
- These could have marked starting positions for the king and his attackers, but if this indicates only eight attackers (and presumably four defenders), why would so few pieces play on such a big board? It seems unlikely that a board of 11 x 11 cells would have been so sparsely populated.
- Might these marked spaces have formed the negative space of the cross formed by the game’s starting position? This would indicate that the king had eight or twelve defenders, and that all attackers started along the very edge of the board. An initial setup involving four unbroken lines of seven attackers each (28 total attackers) seems like a nearly insurmountable obstacle for the king, and any even number per line is unlikely, leaving the possibilities of either twelve or twenty attackers. Eight defenders to twelve attackers seems a bit too imbalanced in the king’s favor, eight to twenty a bit too imbalanced in the attacker’s favor. Twelve defenders to twenty attackers is also a little off from the prevailing wisdom that hnefatafl variants always had a 2:1 ratio, but it does seem to make for a playable game.
- Might the marked spaces have held some special status or function unrelated to the initial setup? A common reason for marking spaces on a board is to indicate either goals (cf. marked corners on the Ballinderry board, likely indicating the king’s four possible routes of escape) or pitfalls (some hunting games may involve hostile spaces, against which enemy pieces may be captured). That the king’s targets of escape should be within such easy reach seems quite unlikely, but perhaps these markings form bulkheads around which the king must navigate in order to escape to any edge space. If we assume this to be the case, would pieces other than the king be able to occupy or cross through these spaces? Would pieces be subject to capture against these spaces? In any case it is doubtful that an attacking piece would begin the game occupying a goal space or a hostile space, so these markings almost certainly had some bearing on initial layout.
Thus we are unable to answer, without further evidence, whether these markings may have functioned as hostile spaces during game play, but we may conclude that they most likely indicate a variant featuring twenty attackers and eight or twelve defenders. This conclusion, if it can be confirmed, may add a footnote to our classification of hnefatafl variants as hunting games featuring cruciform symmetry with a king and his defenders surrounded by twice as many attackers.
One other important possibility must also be considered. If the marked spaces were indeed closed spaces which no piece may enter or pass through and around which the king must navigate in order to escape, this may have precluded the need for a solid wall of defense across each attacker camp, since the closed spaces form an obstacle to escape, and an attacking piece there would be unable to move except laterally anyway. This leads us to the distinct possibility that the central space of each edge may have been left unoccupied, leaving us with sixteen attackers and eight defenders, arranged as I have illustrated below.