It was possible to group 50 of the 59 face-pieces on this basis. The remaining nine were either too damaged and indistinct or too different from the identified groups to be included in any of them.
GROUP A British Museum warder showing Group A face (BM 116). (Source:© The Trustees of the British Museum)
A long straight nose with a flat base, inferiorly placed nostrils and a high nasal root (between the eyes); round open eyes; a down-turned mouth; and a long naso-labial distance (the distance between the nose and the upper lip). Similarity in vertical and horizontal proportions between all the pieces. [Fig. 14]
GROUP B National Museums Scotland bishop showing Group B face (NMS 26).
A bulbous nose with round alae (the fleshy part of the nose surrounding the nostrils) and visible nostrils; a wide short face; round open eyes; a down-turned mouth; a retrusive chin; and asymmetrical eye heights. Similarity in vertical proportions between all the pieces, but variations in horizontal proportions between them. [Fig. 15]
GROUP C National Museums Scotland queen showing Group C face (NMS 22).
A long narrow nose with an up-turned base, flat alae and visible nostrils; round open eyes; a downturned mouth; an infraorbital crease (crease below the eyes); a clear philtrum (column-like hollow between the nose and the upper lip); and an upright facial profile. Similarity in vertical proportions between all the pieces, but variations in horizontal proportions between them. [Fig. 16]
GROUP D National Museums Scotland king showing Group D face (NMS 19).
A wide short face; a straight nose with a rounded tip, round alae and visible nostrils; round open eyes; a down-turned mouth; an infraorbital crease; a clear philtrum; nasolabial creases; and an overbite (malocclusion where the upper teeth are more prominent than the lower teeth). Similarity in vertical and horizontal proportions between all the pieces. [Fig. 17]
GROUP E National Museums Scotland warder showing Group E face (NMS 28).
A defined nose with straight flat profile, visible nostrils and shaped nasal base; round open eyes; and an infraorbital crease. Similarity in vertical and horizontal proportions, except for the one bishop in the group. [Fig. 18].
It is our contention that these groups represent the work of five different craftsmen. If we also grade the face-pieces by size to apportion them into four sets, we produce the following distribution pattern:
GROUPING OF LEWIS CHESSMEN BY GROUPS AND HEIGHT
SET 1
(largest)
SET 2
SET 3
SET 4
(smallest)
Kings
AD
DX
BB
CX
Queens
CC
DD
BB
CX
Bishops
CDDD
CCDE
BCCC
BBBD
Knights
AAAX
AAAX
ACXX
BCC*
Warders
AA**
ADEX
BCCE
DX**
X = ungrouped pieces * = missing pieces
Since there are no clear divisions between our putative sets, this may be evidence that most of the chessmen were manufactured in the one workshop with four or more master craftsmen working on ivory chessmen at any one time. It is possible the workshop received a commission to make four chess sets, or that the same patron gave the workshop repeat orders over a period of several years.
The latter scenario might help to explain the range of design details in the individual pieces. Some things, like the bishops’ mitres and the shields carried by the knights and warders, seem to reflect changing fashions over a period of time extending from the mid-twelfth century into the early thirteenth century. This does not mean that any of the pieces are necessarily as early as the mid-twelfth century, but it is difficult to avoid the conclusion that some are later than has hitherto been supposed. In the case of the bishops, two with mitres with low sides, relatively high peaks and a peaked front bottom edge, would seem to belong in the early thirteenth century [Figs 4.26 and 4.30]. The same is true for those knights and warders with relatively narrow shields with a straight top edge [cf. Fig. 14].
At least two of the face-pieces which could not be grouped are particularly deserving of further consideration. One, a king [Fig. 4.7], assigned to Set 2, seems so different in the appearance and large size of its head that it can readily be imagined to be a piece from another workshop, perhaps even another centre of manufacture. The warder has a unique type of kettle-hat, carinated, and seemingly covered with cloth with diamond-shaped piercings [Fig. 4.55]. It is difficult to parallel, but broadly similar kettle-hats are depicted in the Morgan Picture Bible in the Morgan Library in New York. It is believed to have been produced in Paris in the 1240s.
This king and warder may well be replacements for missing or broken pieces. The warder may also be used as evidence for how long the chessmen remained in use before being hidden – certainly the 1240s, if not even later.
Another thing which is evident from a detailed study of the chessmen is that they are not all of the same quality of design and execution. The craftsman who carved the pieces with Group D faces (for instance, Fig. 4.4) was clearly a man of very considerable artistic ability, producing chessmen which can be regarded as works of art. The crudest pieces, with grotesque faces and awkwardly executed clothing, were produced by the craftsman who did the Group C faces (for example, Fig. 4.58). In producing one of his bishops, the latter showed so little understanding of what he was representing that he carved him with a chasuble to one side of his crosier but not the other [Fig. 4.25].
There are other errors of execution which may be attributed to a busy workshop working under pressure to fulfil orders. These include a knight and a bishop whose hair has not been properly completed [Fig. 4.63], and a knight with a deep groove in the neck of his horse caused by over-cutting when the piece was first blocked out.
The scarcity and value of walrus ivory is indicated by the evident desire not to waste any part of the tusks, which might have averaged about 400 to 500 millimetres in length. Walrus tusks have a dense hard outer layer of dentine and a core of less dense, sponge-like or granular dentine. The best results were achieved by carving all of the details of the chessmen in the outer dentine, and certainly the faces. In many of the pieces the division between outer and inner dentine can be clearly seen, the former appearing smooth and cream coloured, the latter rougher and darker. In Fig. 19, of a king, it can be seen that his face and throne are carved in the outer dentine, while most of the rest of his head and body are cut into the inner dentine.
Considerable skill was involved in getting the maximum number of chessmen out of a tusk, sometimes two pieces from one segment of a tusk. One of the queens [Fig. 20] has one side of her throne attached as a separate piece, probably by the original maker. This may demonstrate the lengths to which he was prepared to go to utilise an inconveniently small piece of tusk.
Not all the chessmen are made of ivory. At least three, including a warder in the collection of National Museums Scotland [Fig. 4.48] have been carved from whales’ teeth which may have been regarded as a cheaper material.
From Trondheim
GIVEN that Lewis was part of the Scandinavian world, and there were great men there who would have been patrons of art and recipients of prestigious gifts, should we consider the possibility that the chessmen were manufactured on that island?
As with so much from medieval times, no strong evidence survives to confirm or deny such a hypothesis. While many craftsmen were itinerant, most scholars would at present expect to locate the manufacture of such pieces in a town or large trading centre. The craftsmen who made such prestigious items were, perhaps, more likely to thrive in such a setting. Our detailed study of the Lewis chessmen demonstrates that they had a good understanding of the robes, vestments and protective clothing worn by kings, queens, bish
ops and knights. This surely suggests that they had access to such people, or were perhaps employed in workshops provided by a king or archbishop. Lewis had no towns at the time in question, but there were strong links between the Western Isles and major Norwegian towns, particularly Bergen, where the kings of Norway often held court, and Trondheim, another royal centre and the site of the archbishops’ cathedral.
The limited evidence favours Trondheim over Bergen as the likely centre of manufacture of ivory chessmen. Apart from the ivory queen already mentioned, excavations in the town have led to the recovery of a wooden king of similar form to the Lewis pieces. It is thought to date to the early thirteenth century. There is a piece of twelfth-century walrus-ivory carving, perhaps the head of a staff, from the island of Munkholmen, near Trondheim. Its scrollwork decoration is comparable to that on some of the Lewis chessmen. It is known that by the early fourteenth century, and possibly much earlier, the Norse settlers in Greenland paid the archbishops large quantities of walrus tusks as tithes (taxes). A king, made from a whale’s tooth, has recently been found on the island of Hitra to the west of Trondheim. It is in the same tradition as the Lewis kings, but is clearly later, perhaps late thirteenth or fourteenth century in date.
The Lewis chessmen bishops are the earliest chess bishops known. In earlier versions of the game, the positions occupied by bishops were taken by elephants. Could it be that this substitution reflects the interest and patronage of a senior cleric like an archbishop of Trondheim? None of this makes it certain that Trondheim was the place many or all of the Lewis chessmen were made, but it is at least a strong probability.
Playing Games
CHESS may have been relatively new to Lewis in the late twelfth century. The origins of the game probably lie in India in, or prior to, the sixth century AD. From there it spread westwards and was widely known in Europe by the eleventh century. The earliest European chessmen copy Islamic models, being non-figurative or abstract. The Lewis chessmen are among the earliest that are shaped as recognisable figures. They are also among the earliest chessmen from the area that is now Scotland.
21. JARLSHOF GAMING-BOARD Fragment of a double-sided slate game-board (X.HSA 801 and 778) from the Viking age settlement at Jarlshof, Shetland. One side appears to have been for playing hnefatafl.
The game of chess has also evolved considerably over the centuries. Earlier versions had ministers or vizirs instead of queens, war elephants instead of bishops, and chariots instead of rooks. The composition of the Lewis chess sets – with a king and queen, bishops, knights, warders and pawns – is clearly much more relevant to the make-up of Scandinavian society in the twelfth and thirteenth centuries. The pawns represent the common folk, or at least those men who were obliged to do military service. The warders – the term for these pieces coined in 1832 by Frederic Madden, Assistant Keeper of Manuscripts in the British Museum – represented a more élite group of warriors, but ones not of the same high status as knights or nobles. These were (royal) retainers or mercenaries.
Chess was not the only board game that required kings and pawns. There was another game called hnefatafl, in which one player had a group of attackers arranged around the edge of the board and attempted to capture a centrally-placed king defended by his guards. The player with the king could win if he could get his king to one of the four corner squares.
Hnefatafl was already popular in the Scandinavian world prior to the advent of chess, and, as time went on, more and more turned to playing chess and hnefatafl became a minority interest or disappeared altogether. There is evidence of hnefatafl having been played in Orkney in the Viking age, in the form of game-boards, and there is one from the Viking settlement at Jarlshof in Shetland [Fig. 21]. It is crudely carved on a piece of slate. Both sides are chequered, one side with the central area in a circle and five of its squares marked with a cross. This would have been for positioning the king and four men to guard him. A hnefatafl board, incised in stone, from a thirteenth-century context at Whithorn Priory (Dumfries and Galloway) in Scotland, provides evidence that the game continued to be played at that time.
The 14 plain ivory disks [Fig. 4.61] found with the Lewis chessmen are often forgotten about, but they are important evidence that the hoard was notjust for playing chess. These disks are best interpreted as tables-men. Many other medieval tables men are known from Europe, mostly decorated and of smaller size. There is an eleventh-century bone and antler set with a board from Gloucester Castle in England, and others of eleventh to twelfth-century date from St Margaret’s Inch (Angus) [Fig. 22] and St Andrews (Fife), both in Scotland.
These tables-men were for playing a board game – tables – a forerunner of modern backgammon. Tables already had a long history, extending back for hundreds of years, by the time the pieces in the Lewis hoard were carved. Board games were widely played in the ancient world and there is evidence for them in Scotland from the Iron Age onwards.
It is possible that the Lewis hoard represents the remains of a games compendium designed for playing chess, hnefatafl and tables. What we regard as chess kings could have doubled for use in hnefatafl, along with either some pawns or warders. Indeed, it is an intriguing possibility that the craftsmen who made these pieces may deliberately have chosen warders to substitute the chariots prevalent in earlier sets precisely because they could more conveniently double for the pawns or guards of hnefatafl. Early game-boards that are two-sided – one for hnefatafl, the other for another game like tables or merels – are reasonably common in the Scandinavian world. Double-sided game boards have been with us ever since and there would no doubt have been many in the twelfth and thirteenth centuries that combined chess and hnefatafl, like apparently the board from Greenland, made of walrus ivory or whales’ teeth, given as a present to Harald Hardrada, an eleventh-century King of Norway. The source for this is the Icelandic Króka-Refs Saga, an early fourteenth-century story of doubtful historicity. There are no clues as to whether or not the Lewis hoard was buried with one or more board. If, as would have been likely, such boards were of wood, they would not have survived in the ground for hundreds of years.
22. TABLES-MAN A tables-man of 11th to 12th-century date from St Margaret’s Inch, a crannog in the Loch of Forfar, Angus.
The playing of board games was popular in both medieval Scandinavian and Gaelic society. Kali Kolsson, prior to becoming Earl of Orkney in 1129 and changing his name to Rognvald, made a poem listing the nine key skills or attributes of a nobleman. The one that headed his list was ability at playing ‘tafl’, which may have been understood to include a variety of games like chess and hnefatafl. This idea that playing games well was a mark of a great man crops up in Gaelic poetry as late as the eighteenth century. And women clearly played as well from early days. That the game was not just the preserve of nobles is demonstrated by the legend of Tristan, in a version recorded as early as 1200. It has Tristan taking ship with Norwegian merchants who have a chess-board on which Tristan plays.
The continuing interest in playing board games in the Western Isles can be demonstrated by the discovery of other tables-men dating to medieval times. From Iona Abbey there is a fifteenth-century bone piece carved with a crowned mermaid grasping her tail with one hand, and holding a fish with the other [Fig. 23]. Another bone tablesman found in a cave on the Isle of Rúm is decorated all over with an interlace design [Fig. 24]. Two decorated bone tables-men, as well as about 50 plain counters or tables-men of stone, have been recovered from the excavations at Finlaggan, the home of the MacDonalds on Islay.
In 1782 Lord MacDonald gave the museum of the Society of Antiquaries of Scotland the handle of an ancient dirk. It was only later that it was realised that this was in fact a very fine chessman (rook?) of walrus ivory, dating to the mid-thirteenth century, with two warriors, back to back in a framework of foliage scrollwork. Both are clad in mail hauberks and have drawn swords and shields [Fig. 25]. Its place of manufacture is unknown – possibly in Scandinavia or Britain. And it may have been re
covered from Loch St Columba in Skye, which contained an island residence of the Kings of the Isles.
Not from the Isles themselves, but from the castle of Dunstaffnage on the west coast of Argyll near Oban, is a chess king made from a sperm whale’s tooth [Fig. 26]. The present whereabouts of this piece are unknown, but on the basis of an early nineteenth-century drawing, it can on stylistic grounds be dated to the late fifteenth or early sixteenth century. It is, however, very much still in the same tradition as the Lewis chessmen.
23 AND 24. TABLES-MEN Bone tables-man (23) (NMS H.NS 99) from the Isle of Rum, c.1500, and (24) (NMS H.NS 92) of 15th-century date from Iona abbey.
25. CHESS KNIGHT Made of walrus ivory, mid-13th century, possibly from Isle of Skye (NMS, H.NS 15).
26. CHESS KING Made from the tooth of a sperm whale in the late 15th or early 16th century, this piece was discovered at Dunstaffnage Castle near Oban, Argyll. From an early 19th-century drawing.
The Lewis Chessmen Page 3