Greek Board Games

Roland G. Austin (University of Liverpool, England)

Antiquity, 14, September, 1940, Pages 257-271

Note: The text of this article appears as near as possible to the original. In addition to English, the author included many terms in Latin and Greek. Some "Internet Browsers" can not read the HTML code for the Greek alphabet, and these words will automatically appear as a series of question marks, representing a word in the text that should appear in the Greek alphabet.

The graphics used on this Webpage did not originally accompany this article. Each graphic has been selected from a variety of sources. As you will see from the contents of the article, no one is quite sure of the actual appearance of the depicted board games.

Greek Amphora

[The graphic on the left is an "Attic Black-Figure Amphora" supposedly illustrating the Trojan War heroes Ajax and Achilles playing a board game. This amphora is attributed to the artist Exekias and dated ca550 bc. It is in the Vatican Museum. In this conception, Athena is NOT watching the game.]

(Page 257) The study of Greek board-games is almost wholly inconclusive, owing to the scanty and extremely imprecise evidence available. Difficulties would in any case be inevitable, since most games are better grasped in actual play than by studying a set of written rules; and here the technicalities are expressed in a foreign language and were sometimes not clear even to their users, who cheerfully omit what they do not understand, or take it for granted that the reader is familiar with the main details. Further, the games so described by our authorities had often been long obsolete. Some of the difficulties may be realized by trying to reconstruct a game of Ombre entirely from Pope's Rape of the Loch, or a game of cricket from Dickens' account of All Muggleton v. Dingley (even with the help of Mr A. G. Macdonell's searching critique of that famous match). Here is an example of a rather different kind, which well shows the pitfalls of unfamiliar terminology; it is translated from K. Silex's John Bull zu Hause, and is an attempt to explain cricket to Germans. `Two teams of 11 men oppose each other; two "wickets" are set [how?] in the ground at a distance Of 20 metres being three wooden sticks [how high?], over which two rods (Stäbe) are laid [how?]. The aim is to hit the wicket with a ball [how big?] or to knock off the rods with it. One side defends the wicket, the other attacks; the defenders post before each wicket [where?] a "batsman" with a striker (Schläger): the other side opposes him with a "bowler" who tries to hit the wicket with the ball [how?]. Two men only of the defense are in actual play, the rest wait their turn' etc. Apart from the ambiguities already hinted at, the nature of the bails is not explained, the bat is described by a word which may also mean a tennis-racket or a croquet-mallet, while one might well infer that the non-batting members of the 'defense' (itself an odd term) are lined up near the wicket anxiously awaiting their turn. This is just the kind of thing that confronts the would-be student of Greek board-games.

Further complications have been caused by the reckless use of (Page 258) modern terminology in accounts of these games, by the indiscriminate equating of Greek with Roman games, and by the convenient but unjustified application of the rules of one game to suit another. Thus, the Roman latrunculi or the Greek πεττεια ('petteia') have commonly been translated 'chess', which is impossible and utterly misleading (even 'draughts', though less misleading in certain cases, is historically anachronistic). A notorious passage of Isidore (XVIII, 60 ff.) has been ruthlessly adduced to illustrate such entirely diverse games as latrunculi and XII scripta,1 and to prove differentiation of pieces, quite baselessly. latrunculi has been loosely regarded as invariably equivalent to petteia; special markings have been assumed to exist on the boards, carry special powers ascribed to certain pieces, without a shred of evidence. Sometimes an ancient authority is at fault, as for instance when the scholiast Theocritus VI,18, writing when a knowledge of chess had spread from the East, uses the word ζατρHικιου, in connection with petteia - with disastrous results, as Cholmeley's note on the passage unconsciously shows. But the blame lies more with modern writers, whose zeal for reconstruction so often outruns scholarly method; this lack of discrimination vitiates nearly all modern discussions, such as the elaborate Jeux des Anciens of Becq de Fouquières (Paris, 1869), or the various articles in Daremberg and Saglio's Dictionnaire des Antiquites, to say nothing of such pretentious works as E. Falkener's Games Ancient and Oriental and how to play them (London, 1892), or H. Coleridge's Essay on Greek and Roman Chess.2 These methods merely madden; extreme caution should be used, and ignorance cheerfully confessed. The sane way of approach has been shown more recently by Hans Lamer in his monumental article Lusoria Tabula in Pauly-Wissowa's Realencyclopädie, and by the late S.G. Owen in some of his notes (too brief, unfortunately, especially on the Greek side) on Ovid, Tristia2. My purpose here is simply to restate and perhaps help to clarify certain obscurities, without any dogmatic divination or misplaced ingenuity.

Two points have hitherto received insufficient consideration a) the probable form of these games, in the light of the characteristic types of game which from earliest times have amused mankind in an idle hour all over the world; b) the relative weight of our authorities, in view of the fact that games develop or become obsolete with time. (Page 259) The simplest board-games of most countries are based on three primitive activities of man - the battle, the race, and the hunt - modern types of which are chess, backgammon, and fox-and-geese. Such types one would expect to find among the Greek games; just as the Roman latrunculi was clearly a battle-game and XII scripta a race game. Only if a game conforms to such generic types is it safe to make assumptions concerning its method of play, or to identify it with a game of another country. The object of the battle-game is to hem in one's opponents and drive them off the board; no specified number of men or size of board is needed, and in the earliest forms of the game there is no differentiation of pieces; no dice are used. In the race game, the aim is to bring one's men to an appointed terminus and so be first off the board; again there is no differentiation, but the number of men is fixed, usually 15 on each side; dice are used to control moves. In the hunt, a single piece tries to escape from an opposing pack; no Greek or Roman game seems to have been of this type, which was common, however, in Scandinavia and among the early Celts.

Secondly, the natural development of games and changes in their fashions necessitate close scrutiny of our authorities. Such a development may be seen in the Roman game XII scripta - a race-game, played on a board with three tracks, as can be reasonably inferred by correlating certain passages of Ovid with numerous existing boards or diagrams.3 For later a quicker game was evolved by eliminating the middle track, resulting in the Byzantine tabula or ταβλη, to which Isidore (XVIII, 60 ff.) refers. An actual position of this game has been recovered from an epigram of Agathias in Anth. Pal. IX, 482;4 yet if we applied this evidence to the earlier game we could hardly expect good results. Again, one type of game may oust another in course of time, and the earlier one will be forgotten or imperfectly understood. Thus the relative dates of our authorities, and their possible interdependence, become of primary importance.

The Greek evidence falls into two groups: a) casual references in literature, from Homer downwards; b) the accounts of the antiquaries Pollux, Hesychius, Suidas and others, together with the long statement made by Eustathius in his Homeric commentaries. The difficulties offered by the second group are obvious; for long ago Casaubon pointed out that Eustatius' account is based on the lost (Page 260) work of Suetonius περι Ελληρικης παιδιας, written perhaps in Greek, perhaps in Latin. It is doubtful if we can ever settle the interdependence of these writers, of whom the earliest is Pollux, or their relationship with Suetonius, a senior contemporary of Pollux. Reifferscheid, in his edition of the Suetonian fragments (Leipzig, 1860), thinks that both Suetonius and Pollux may derive from a common original; Lamer boldly regards Suetonius' lost book as the source of all the existing Greek evidence, except for a part of Eustathius which is definitely taken from Athenaeus. At least it is clear that our evidence for the Greek games rests mainly on the work of Pollux, a 2nd century Egyptian Greek who was acquainted with Rome, and on the excerpts made by the 12th-century Byzantine Greek Eustathius from the lost work of the Roman antiquary Suetonius; therefore, none of it is really pure Greek. Has Eustathius reported Suetonius intelligently, and how far arc his descriptions coloured with the terminology of his own day? Did Suetonius himself really understand the Greek games of which he wrote? The very nature of these sources is itself abundant ground for expecting no conclusive results. Neither can such archaeological finds as exist be demonstrably related with known games. By contrast, the evidence for Roman games is like daylight, backed up as it is by the existence of undoubtedly related diagrams.5 The study of the Greek games is, in fact, a journey into complete darkness.

[Below is an Attic Black-Figure amphora of Ajax and achilles playing a board game WITH Athena watching, has been dated ca520 bc. It is in the Getty Museum in Malibu, California.]Greek Amphora

At the outset the term 'petteia' needs clarification: it is a game played with πεσσοι ('pessoi', i.e. 'pieces' or 'men '), but is the expression generic, or does it always imply one particular game? The former interpretation seems preferable, and considerable confusion has arisen from assuming that it means one game to the exclusion of others. References to games with 'pessoi' begin with Homer, and continue throughout Greek literature down to the lexicographers and Eustathius; one such game was called πολεις  ( 'poleis', i.e. 'cities'), but the same terminology is used also of others which were certainly not identical, and it seems only reasonable to regard 'petteia' as a generic term for 'a board-game' in general. Sometimes 'pessoi' are spoken of together with κυβοι ( 'kuboi', i.e. 'dice'), and such a combination might well mean a race-game and not a game of pure skill, but this can seldom be determined. The game with dice was called κυβεια ('kubeia'), and its meaning, will be considered shortly. Plato (Page 261) assigns ail Egyptian origin both to 'petteia' and 'kubeia' (Pharedr. 274d), probably rightly; it is he who first regards 'petteia' as a science (RP 333b), and he adds some valuable details of one form of it in RP 487b, where Socrates' victims, who are finally cornered and made helpless by dialectic, are compared to 'bad petteia-players, who are finally cornered and made unable to move by clever ones'. In RP 374d, both 'petteia' and 'kubeia' are said to involve long training if skill is to be achieved. Plato clearly dissociates 'petteia' from 'kubeia', i.e. he distinguishes it from a dice-game, and this is borne out by the language which he uses. For he obviously has in mind not a race but a battle, where the enemy is blockaded until he is beaten; similar expressions are used later of Scipio by Polybius (I, 84.) - 'he destroyed many men without a battle by cutting them off and blockading them, like a clever petteia-player'. This particular game may well have been that known as 'cities', of which more will be said later; but we are not justified in applying the rules for other forms of 'petteia' to it - e.g. because we know of a technicality called ιερυ γραμμη ('the sacred line') in connection with one form of 'petteia', we cannot therefore assume its presence in 'cities', as is done for instance by Sir D'Arcy Thompson in a paper which will be mentioned again. It would be absurd if some remote investigator, misled by the prevalent Scotch habit of referring to the Association code exclusively as 'football', were to assume either that Rugby is not football, or to transfer to Association some of the rules peculiar to Rugby. A reasonable conclusion is that 'petteia' was not a particular game, but a generic expression for a game probably of the battle-type and played without dice.

We must now consider the term 'kubeia'. Probably this long meant merely one of the various methods of dicing; but when we come to Pollux and the later lexicographers, it is constantly used of a board-game in which movement is controlled by dice, i.e. a race-game of the backgammon type; and it is precisely its application to games of the battle-type that causes inevitable confusion. At first sight the position would seem clear enough in view of the plain statement of Hesychius (s.v. ; πεσσα πεντεγραμμα ) that 'petteia' differs from 'kubeia', for 'in the latter the players throw dice, in the former they only move the pieces'; and elsewhere Hesychius contrasts 'petteia' with 'kubeia' as being a game of skill opposed to one of hazard. But all this apparent clarity is completely inspissated by a second statement, in which he explains 'petteia' as 'a game with dice (δια κυβων παιδια), (Page 262) and by other similar glosses. Turning to Eustathius, we learn that 'in ancient times' a clear distinction between the two was made: 'kuboi', he says are six-sided dice, while 'pessoi' are 'something different (ο πεσσας ετεροιον τι εστιυ); he then quotes a line of Sophocles to clinch his point, which in fact proves nothing at all. It is clear enough that by Eustathius' time the proper meaning of 'petteia' and 'pessoi' was not known, and the two were taken as some antique variant of 'kubeia' and 'kuboi', there being some unexplained distinction between the two. The cat is finally let out of the bag when Eustathius goes on to explain πεττευειν ('to play at petteia') by using the word ταυλιζειν ('to play at' ταβλη or 'tabula'): this term (or ταβλιζειν as it appears elsewhere) was the Byzantine word for the race-game then current, as played by the Emperor Zeno in Agathias' epigram already mentioned - demonstrably a backgammon-game, played with dice, a modification of the Roman XII scripta. In fact, Eustathius sees in 'petteia' simply a form of the principal board-game in vogue at his time, in much the same way as modern writers have equated it with chess; he does not really understand the term at all.

Clearly then Suetonius' late copyists have identified 'petteia' indiscriminately with 'kubeia'. Hesychius, who lived much earlier than Eustathius, is not quite so vague; he has hit upon a lucid differentiation between the two, yet abandons it for an explanation which better suits the games of his own day, probably because a board-game without dice was not readily comprehensible to him. For Hesychius too uses the word ταυλιζεν ; in one of his glosses. The confusion appears as early as the 4th century in Joannes Chrysostomus (XI, 970), and in fact, earlier still we find Pollux himself equally vague: in one passage (IX, 97) he realizes that there is somehow a difference between 'petteia' and 'kubeia', in another (VII, 203) he appears to include the former term under the head of the latter. It is significant too that Pollux uses 'kubeia' occasionally in speaking of games which from his own description cannot possibly have involved the use of dice, e.g. χαλκισμος (spinning a coin) and ιμαντελιγμος (a game played with a knotted strap). The word was evidently the familiar term of his day for games of hazard in general, and he is apt to use the expression 'a species of kubeia' (ειδος κυβειας) as a convenient formula for all sorts of games.

It seems evident that when the lexicographers and Eustathius use the term 'kubeia' in referring to board-games, they are merely employing the terminology of their own day, and that the proper sense (Page 263) of 'petteia' was lost. An analogous development can be seen in Latin; alea,6 the Latin equivalent of 'kubeia', is used at least once by Macrobius (late 4th century or early 5th) of a game which obviously did not need dice - the Roman sport of capita aut navia, 'heads or tails', described as aleae lusus. This suggests another possible source for confusion; for if our authorities depend ultimately on Suetonius, and if Suetonius' book was written in Latin, they may well have been misled by some ambiguity of language like the example just quoted from Macrobius: e.g. the natural Latin word for the board would be tabula, a word which to Hesychius and Eustathius at least would convey the meaning not of a board but of a particular game, the race-game ταβλη played with dice (so Isidore l.c. writes 'tabula luditur pyrgo, calculis tesserisque'). On the other hand, if Suetonius wrote his book in Greek, as some scholars hold, it is plain enough that even if he himself used the terms 'petteia' and 'kubeia' correctly, his copyists (including the nearly contemporary Pollux) did not understand the distinction.

To sum up so far: both 'petteia' and 'kubeia' seem to have been generic terms; the former meant a battle-game or games, the latter, after losing its original sense of dicing, became applied to a race-game; as the latter type of game became more popular, the true sense of 'petteia' was forgotten, and 'kubeia' was applied to any board game indiscriminately. This somewhat tedious linguistic discussion has been necessary to show that when Pollux and the rest refer to a game as a 'species of kubeia' we must not take it at its face value as implying a race-game played with dice. We should remember that none of these authorities had practical knowledge of these games; even Pollux uses a past tense regularly in speaking of them, or makes it plain that he writes from hearsay.

Greek Amphora

Let us now examine the three board-games whose names we know, or think we know. One is 'poleis' ('cities'), of which Pollux says (IX, 98) 'the game played with many pieces is a board (πλινθιον) with spaces disposed among lines; the board is called "city" and each pie a "dog" (κυων); the pieces are of two colours, and the art of the game consists in taking a piece of one colour by enclosing it between two of the other colour. He then quotes an obscure reference to it from Cratinus (5th century BC). This description looks as if Pollux really knew something of the game. Eustathius (Od. I, 107 ff.) has a substantially similar account, but much vaguer, and he calls the game's (Page 264) species of kubeia'; he explains the term 'cities' as the old name for the spaces on the board. We can at once discredit the idea that it was a dice-game and therefore of the backgammon type, for the reasons which have been given above.7 It is quite clearly a battle-game, almost certainly the form of 'petteia' described with military terminology by Plato and Polybius. Here it is worth recording that the word ηλινθιον ; is used by Arrian and Josephus of a column or mass of troops, a fact not hitherto pointed out in connection with the game; this supports the battle theory. And if the game itself was really known sometimes as πλινθιον, as Pollux appears to suggest, here is a fruitful source of confusion; for if Suetonius had chanced to use the obvious Latin word tabula or tabella to translate the Greek term, we should then definitely nicely have had a word used of the battle-game which later became identified with the race-game.

The method of capture shows that the game was analogous with the Roman battle-game of latrunculi (cf. Ovid AA, III, 358, Tr. II, 478); it is one of the oldest known, and is found in games played in Malaya, Siam, China, Arabia, and Egypt; it obtained in a Persian game mentioned in Firdausi's Shahnama, as well as in the old Norse game of Hnefa-tafl and others of the same family (although there the game was a hunt rather than a fight, a kind of fox-and-geese).8

Greek WrestlersIt is a certain clue to the nature of the game. The number of men is indeterminate, another feature of the type - it varies with the size of the board; Photius sates that in 'cities' sixty pieces were used, presumably 30 on each side, though this seems rather large. We may legitimately assume that what is known in chess as the Rook's move was used, as in all other games of this family, for it can be demonstrated that no shorter move alone will give a playable game; there may be a hint of this in latrunculi, and as the Roman game allowed backward moves, 'cities' may have done so too. The tactics consisted in preventing the enemy from maintaining his massed formation, and by breaking through it to manoeuvre until his force was gradually scattered and so taken. An isolated man brought danger to himself and to his side. These are legitimate inferences from what we know of the Roman game and others of the family. (Page 265)

Clearly 'cities' was of respectable antiquity, going back as it does to Cratinus at least. Two well known passages in Plato and Aristotle respectively have it in mind. The first is RP 422e: comparing other states with his ideal state, Plato says 'none of them is one city, but many cities, as they say in the game', on which the scholiast remarks that 'cities' was a form of 'petteia', and that the proverb has come from the game. This is surely a direct reference to the game, not a mere general expression for a jest, as Warren thought; it may perhaps corroborate the use of the term 'poleis' for squares on the board, unless Plato means that the board was divided into two rival 'cities', one for each player. The comparison comes in aptly, for in the context Plato is discussing a fight between two cities, and it is probably not coincidence that he has chosen earlier to bring in a reference to 'dogs' (in the literal sense) as taking part in it. The meaning of the proverb is uncertain: perhaps there is a play on the two senses of 'poleis' - 'many squares don't make a city', i.e. some wise unifying principle is needed also, to ensure that the forces on the squares are properly coordinated; just as Plato says of his ideal state: 'so long as your city is wisely ordered on the principles just laid down, it will be the greatest of all cities'.

The Aristotelian passage is in Pol. 1253a: Aristotle compares the cityless man (απολις) to an isolated piece in a game of 'pettoi' (ατε περ αζυξ ων νσπερ εν πεττοις). The curious word αζυξ occurs at a much later period in Agathias' epigram on Zeno's game of ταβλη , where it means a 'blot' at backgammon, i.e. a single piece standing unguarded by a companion and therefore liable to capture. But the commentators are surely wrong in supposing that Aristotle had the race-game in mind; the word in the context, and the comparison introduced between the 'cityless' man and the pugnacious Homeric warrior (I agree with Jackson in assigning the comparison not to the αζυξ ; but to the απολις) both suggest that he meant the battle-game, 'poleis'. As Newman remarks, there is no reason why αζυξ ; must imply Zeno's game; it could surely apply equally well to a piece in 'poleis' which has become cut off from the main force and so is in danger itself and a danger to others. It is this to which Aristotle compares the 'cityless' man: he resembles a piece which has 'lost its square' or has been 'driven from its polis'. In his note on the passage in Susemihl and Hicks' edition, Jackson sees the force of the comparison, but fails to realize that 'poleis' is a different type of game from backgammon. Newman is more cautious, but unfortunately he afterwards throws caution to the winds, (Page 266) and follows Becq de Fouquières in dragging in a reference to a 'sacred line' (ιερα γραμμη), which belongs to a quite different game and is never mentioned in connection with 'poleis' - a good example of the danger of assuming that all these games can be treated indiscriminately.

Another reference to this game occurs in Euripides Suppl. 409, where the Theban herald asks Theseus 'Who is the lord of this land?' Theseus rebukes him, saying that Athens has no one ruler, whereupon the herald answers 'You give me this one advantage, as in pessoi, for my city is captained by one man, not by a mob', i.e. 'my side is well led, yours is not'. Some editors suppose that the words 'you give me this advantage' imply a privilege conceded by a player to his opponent, but we know nothing of any handicapping principle in any of these games, and it seems unlikely that they bear any meaning other than that suggested above. It can hardly be by accident that the comparison with 'pessoi' introduces a long discussion between the two men on rival theories of conducting a state, democratic v. totalitarian, and there may well be other allusions to the game in what follows. A second Euripidean reference occurs in a fragment of the Erechtheus preserved by Plutarch, do exilio 604 d; unfortunately the text is doubtful, and it is difficult to get a clear picture of what Euripides has in mind; he appears to compare the indigenous and homogeneous state of Athens to 'other cities' full of immigrants from elsewhere - possibly he is thinking of the ever-changing appearance of a 'poleis'-board, with the men of both sides intermingling according to the play. Sir D'Arcy Thompson thinks that the game implied in the passage was played with dice, i.e. that it was a race-game (this depends to some extent on the reading adopted); but surely the mention of 'cities' must imply the battle-game, 'poleis'.

Such are the chief references to the game of 'poleis'; it was evidently very popular and needed much skill; it obviously appealed to the philosophical mind, as Plato's frequent mention of it shows (the scientific battle-game, with its need for cool logical thinking, would plainly suit the philosopher better than the race-game played with dice). It may well have been this form of 'petteia' to which Philostratus alludes (Heroica II, 2) as 'no idle sport, but one full of shrewdness and needing great attention'. Although it must be emphasized that 'petteia' bore no resemblance to chess, it is clear that it needed the same qualities for success as chess does.

A second game is διαγραμμσμος ('diagrammismos'). This unfortunately cannot be identified. Pollux mentions it immediately (Page 267) after his account of 'poleis', but in an ambiguous manner (IX, 97): 'next to (or close to - the Greek word used is εγγυς) this game is diagrammismos, a game which they used also to call "lines" (γραμμαι). It is not possible to tell whether Pollux meant a related game, or simply one next on his list. Certainly Hesychius and Eustathius took it to be a game like 'poleis'. Hesychius calls it 'a game of sixty pieces, white and black, moving on spaces'; Eustathius writes (Il. VI, 633) "diagrammismos, a species of kubeia, was played, they say, by means of the sixty black and white pieces on the boards" (εν πλινθιος). Philemon mentions this game in the line he topes, he plays diagrammismos, he plays dice "(κυβευει - which might mean "he plays at kubeia")'. As there is nothing here that does not appear in what we know of 'poleis', it would be unsafe to give these details independent weight as evidence for 'diagrammismos'. Hesychius and Eustathius, relying on Pollux' vague word εγγυς  may quite well have transferred them bodily from the other game. Elsewhere Pollux (VII, 206) includes 'diagrammismos' in a miscellaneous list of 'species of kubeia', comprising among others two obviously non-dicing games, ιμαντελιγμος  and χαλκισμος this is not reliable evidence either. If the game was really like 'poleis', and was therefore a battle-game, we cannot take seriously the statements of Pollux and Eustathius that dice were used in it; and Philemon, in his portrait of a rake in the quotation, seems to regard 'playing diagrammismos' as a different form of dissipation from 'playing with dice' (or 'playing kubeia'), although the point cannot be stressed. We cannot really infer anything definite about the game, and it is suspicious that Pollux gives no details. But the names given to it by Pollux, especially the alternative γραμμαι, would seem to point to a quite different type of game from 'poleis'; it might be a real race-game, a genuine 'species of kubeia', if we could regard γραμμαι as something like the Latin scripta - then, of course, we must scrap all the details mentioned by Hesychius and Eustathius; or it might even be some form of Merels (Three Men's Morris),9 neither a battle-game nor a race-game. It is disappointing that so little certainty is possible.

The third of these Greek board-games is the so-called πεντε γραμμαι #32; or 'five lines'; we do not know its actual name, but this is a convenient way of referring to it. It bristles with difficulties. Pollux describes it thus (IX, 97): 'each of the players had five pieces on five lines, so that Sophocles naturally says "five-lined boards and the (Page 268) throws of the dice"10 and of the five lines on either side (? - the Greek word is εκατερωθεν, which might perhaps mean "in either direction") there was a middle one called the sacred line, and a player who moved a piece from it gave rise to a proverb "He moves the piece from the sacred line". Eustathius has substantially the same account, but is rather fuller on the 'sacred line': he says 'the beaten player goes to it last (επ εσχατην αυτην ιεται), whence the proverb "to move the piece from the sacred line", referring to people who are desperate and need final help; he quotes the proverb in various forms from Sophron, Alcaeus, and Theocritus, showing that the game goes back to a respectable antiquity. Hesychius also refers to the 'sacred line', but does not explain it; Suidas interprets the proverb either metaphorically of sailors in desperate need, or of petteia-players, 'with whom there is a sacred piece'. The scholiast to Plato (Lg., 739a) also speaks of a sacred piece, not a line, and adds, surprisingly, that it was immovable (ακινητος). Pollux mentions the proverb also in an earlier passage (VII, 206), after the list of games which includes 'diagrammismos', apparently as an afterthought; the reading there is uncertain, but there seems no reason for Lamer's tentative suggestion that 'diagrammismos' too had its ` sacred line '.

The obscurity of all this evidence is impenetrable. Sophocles' line, quoted with such triumph both by Pollux and Eustathius, is quite intractable - we cannot tell from it whether dice were used in the game, or whether he is speaking of two different forms of amusement, or even its exact meaning. Again, it is clear from both Pollux and Eustathius that each player had five lines, but do they mean five vertical and five horizontal lines, or two sets of five lines running in the same direction? If the latter, why the name πεντεγραμμα ? Finally, what was the 'sacred line'? Did it run between two sets of five lines (thus making 11 altogether), or was it the middle one of each set, or can we infer from Eustathius11 that the board did in fact have five lines each way, and that the 'sacred line' was the middle one in each direction? The movement of pieces is also obscure, since we do not know the position or function of the 'sacred line'; Eustathius' language gives no help, and is in fact almost self-contradictory - he probably knew nothing whatever about it, and has inverted something to sound impressive. (Page 269) No answer seems possible to any of these problems. One difficulty, however, can be safely ignored - the statement that the piece on the 'sacred line', the 'sacred piece', was immovable; if taken literally, this would make nonsense of the proverb to which all our authorities refer; it must mean, not that such a piece could never be moved, but that at a given position of the game such a piece was for the moment immobilized. Isidore (XVIII, 67) uses the Latin incitus, the exact equivalent of the Greek ακινητος, in connection with the game of tabula, and this is the rational explanation of it there. It is hard to suppose that the Greek word does not convey the same idea, a fact which might be used to prove that the Greek game, like the Latin tabula, was a race-game and a genuine 'species of kubeia'. But there seems no particular reason to trust the scholiast to Plato any more than anyone else.

Two reconstructions of this game may be mentioned. One is by Becq de Fouquières, based on little more than imagination and good will. He assumes a board of five vertical and five horizontal lines, in itself not impossible to extract from the Greek; he then assumes that the game resembles 'poleis', and takes over the military tactics spoken of by Plato and Polybius with reference to 'petteia', although the number of pieces used is against the identification; lastly, he invents a special enclosure or square surrounding the point of intersection of the two middle lines, as a kind of sanctuary, protected by the unseen presence of what had originally been a sacred and immovable piece. Assuming that the method of capture was the same as that in 'poleis', he claims that this 'sacred square' had the power to help in a capture; i.e. that a hostile piece penned in between it and one of your own men could be taken. Such a function, in fact, does belong to a centre square in certain games; Linnaeus saw it employed in a Lapp game of tablut in 1742.12 But it is surely obvious that de Fouquières' reconstruction of the game on these lines is quite fanciful, for there is not the slightest evidence either that the game resembled 'poleis' or that such a special enclosure or sanctuary existed in the middle of the board. If 'poleis' had really had a 'sacred line', it seems very improbable that Plato would not have made some mention of it. Further, de Fouquières' account of the moves, although supported by diagrams, is very difficult to follow.

(Page 270) A quite different solution is proposed by Sir D'Arcy Thompson, in a paper already mentioned.13 He too calls the game 'poleis', and seems to assume that the rules for different games may be applied to any indiscriminately. He bases his account on an Egyptian game called 'Seega', played on a board whose squares, usually represented by holes in the ground, are set in five rows of five each way. Each player has twelve 'kelbs', i.e. 'dogs'. The central hole is left vacant; otherwise each player arranges his pieces where he likes on the board. When all the men are so entered, movement begins, each piece being allowed to move to the next vacant square, in file or row, but not diagonally. The method of capture is exactly similar to that in 'poleis' - one kelb is taken by being surrounded by two of the opponent's pieces. Possession of the central hole is most important, as its occupation gives the best opportunity of intercepting an enemy piece: 'a piece standing thereon is therefore valuable to the owner, and a special object of the opponent's attack', and so it would not be moved except in dire necessity. Thus the centre hole corresponds to the 'sacred line'.

This is at least a straightforward account based on an actual playable game, not on imagination. But it is based on a misconception, the idea that the game of the 'five lines' is to be identified with 'poleis'. The Greek sources make two quite different games, and the attempted conflation is unsound. The Egyptian game may certainly be of the same family as 'poleis', though the number of pieces used appears to preclude actual identification; the name 'dogs' for the pieces is not conclusive, for it is commonly found in various games throughout Egypt and the East (and occurs for instance in a tenth-century Indian account of the oriental game of Nard, a race-game and so of a different family from 'poleis'). As for identification with πεντε γραμμαι , it may be pointed out that the Egyptian game has 12 pieces, the Greek five, on each side. Further, 'Seega' can also be played on a board of 7 by 7 or 9 by 9 squares; and even if the centre hole in 'Seega' has the importance claimed for it, the fact remains that in the Greek game we have to deal with a 'sacred line', not a 'square'.

We are bound to accept the conclusion that this 'sacred line' is not only in itself an insoluble problem without further evidence, but also precludes identification of the Greek game with any other until its nature can be satisfactorily established. S.G. Owen assumed (Ovid, Tr. II, 475) that the game was a race-game, played with dice, (Page 271) like the Roman XII scripta; the idea of γραμμαι might support this (although 'five lines' was not necessarily the name of the game), and the evidence of the scholiast to Plato, as already stated, might also be quoted; but the apparent shape of the board, so far as it can be inferred, is against this interpretation, while the age of the game, going back as it does to Alcaeus at least, would seem not to be in favour of a game of hazard; also, the 'sacred line' and its function is no more clear in this type of game than it would be in the battle-type. Only archaeological evidence would make identification possible; but although Lamer thinks that certain boards found at Epidaurus may have been used for this game, the evidence is still very obscure and the mystery of the 'sacred line' still not cleared up.

Thus the only one of these three games which seems to admit of positive elucidation is 'poleis'. More important to the writer is the method of approach used, in spite of the mainly negative results; it will be something gained if a few misconceptions can be removed which have hitherto been implicitly accepted as true, and if a possible path has been suggested for future investigators.

  1. See Greece and Rome, IV, pp. 24 ff., 76 ff.
  2. In an appendix to D. Fortes' History of Chess (London, 1860).
  3. Greece and Rome, IV, 30 ff.
  4. Greece and Rome, IV, 76 ff. Journal Hellenic Studies, LIV 202 ff.; Cf. H. Jackson, Journal of Philology, VII, 2240 ff.
  5. Greece and Rome, loc. cit.; a fairly well-preserved specimen of a Roman XII scripta board is in the Holt collection of the National Museum of Wales at Cardiff (see W.F. Grimes, Y Cymmrodor, XLI, 131; R. G. Austin, Archaeologia Cambrensis. XCIII, 250).
  6. i.e. a die, but later used of a board-game with dice.
  7. I must here withdraw a too hasty footnote to my paper in Greece and Rome, IV, 25, in which I put over-credulous belief in Eustathius and denied that 'poleis' was akin to the Roman latrunculi.
  8. I owe this and other information, together with much helpful criticism, to Mr. H. J. R. Murray, author of A History of Chess (Oxford, 1913).
  9. cf. Greece and Rome IV, 79.
  10. και πεσσα πεντεγραμμα και κυβων βολαι: I have adopted Lamer's suggestion for the obscure phrase πεσσα πεντεγραμμα.
  11. His actual words are obscure enough παρετεινετο δι αυτων και μεση γραμμη.
  12. See T. E. Smith. Lachesis Lapponica, II, 55-8, a reference which I owe to Mr. Murray.
  13. The Game Of ΙΙσλις   (Glasgow, 1911), kindly sent me by the author.

Last update November 24, 2010