sip. fesefio - Jape ARCHIVE for HISTORY OF EXACT SCIENCES Edited by C. TRUESDELL Volume 25 The Ritual Origin of the Circle and Square A. SEIDENBERG SPRINGER-VERLAG BERLIN - HEIDELBERG - NEW YORK 1981The Ritual Origin of the Circle and. Square A. SEIDENBERG I. The Circle 1. Docs the circle have a single, or multiple, oF . B 210 2. Ritual geometrical constructions... sl lon 3. The thesis stated... eee eee eel itll ims 4. Circular movement in the sky and on earth |. | | eee rey 2 76) 5S. Evidence from myths... 6. Evidence from myths, continued. 7, Further evidence from myths. . : Evidence from rites... .. 9. The god at the end of a rope . 5 10. Application of old myths to new rites | 11, The polar star in ritual 11 bis. Spinning tops... ; [er 12, Solar and stellar citeumambulations . | | 13. Some objections to the thesis : 14, How old is the circle? Ul. The Square 16. Duality of circle and square in ancient mathematics... 2... 17. Round and square houses... . Perr pooponor?) 17 bis. Circular and square earthworks in North America|. 1). a 418, Duality of circle and square in ancient civil 19. The Square as a cross of the 20, Sacred numbers 2... ss 21. The cardinal directions: | | fll 22, Origin of the Two in social organization 313 23. Genesis of the square from the circle...) | | | OG + 320 24, Origin of the Three... 2... 11. a os 25. Conjecture on the origin of the Four... |)! ees 323 26. Summary... se. ee ll 324 Referencs. TT!270 A, Setpenoera 1. The Circle 1, Does the circle have a single, or multiple, origin? In a previous paper', considered the mathematics of the ancient civilizations—of Babylonia, Egypt, India, Greece, and China—and argued that a good many of its features can be understood in the ight of an ancient concern to elaborate the ritual, that originally geometry was theologic geometry. Since the citcle and square are the main elements of this ancient geometry, it becomes reasonable to look for the origin of these figures in the ideas and activities of the ritual itself. What, then, in ancient Fitual Jed to the circle and the square? In the paper cited, I mentioned the question, but preferred to postpone my answer. The reason was that 1 was examining the actual mathematical remains of the ancient civilizations; and it was possible to develop the considerations with only minor references to archaic ideas. Here those ideas themselves have to be considered. Subsequently a paper of mine on an allied subject has appeared in which these ideas received extended consideration®. However, I will repeat the leading ideas wherever necessary. But quite aside from the specific contents of the thought of the Ancient East, there is the theory of the Diffusion of Culture, according to which various wide- spread practices and beliefs are not the spontaneous reactions of the human mind {o environing conditions but are the product of certain special circumstances. The theory tends to regard ideas as historical products; and is cautious not to be misled by familiarity into accepting apparently simple ideas as obvious, natu- tal, or primitive (indeed, it prefers to avoid these ill-used words). It tends to regard similar ideas and practices as ipso facto evidence of historical connection. Opposed to the theory of Independent Invention, according to which the various widespread practices and beliefs are the spontaneous reactions of the mind to environing conditions. The theory tends to regard ideas as psycho- logical products. Historical aid ethnological observations are frequently supple- ‘mented with references to psychclogical studies on children, with an association jeas occurring to the writer, or with an appeal to what we, placed in a similar situation, would consider a reasonable procedure. Geographical features, in Particular the oceans, are seen as barriers to, rather than avenues of, communi- cation. The various peoples of the carth, after their dispersal from some supposed cradle along with a fund of so-called “elementary i way to theit present locations, have there, scarcely with ‘any communication even with their nearest neighbors, proceeded to build up the cultures they now possess. Similar ideas and practices are regarded as ipso facto evidence that the human mind works similarly under similar condi The limited objectives of the present paper presumably make it out of place to enter into a critique or a polemic on Diffusion versus Independent Invention, * “The Ritual Origin of Geometry”, Archive for History of Exact Sciences, vol. 1 (1962), pp. 488-527. See also vol. 9 (1972), pp. 171-211 and vol. 14 (1975), pp. 263-295. ¥ See “The Ritual Origin of Counting”, Archive for History of Exact Sciences, vol. 2 (1962), pp. 1-40. Ritual Origin of Circle and Square 21 clear that we have to come to some understanding on this matter, As the pre-Columbian American Indians knew the circle and the square and as I consider these figures to have arisen from the ritual activities of the Old World, if it is tacitly (or explicitly) supposed that the Old and New Worlds were disparate worlds, that they developed in essential isolation from cach other, then my ‘argument is bound to be misunderstood. Of course, I cannot ask the reader to accept arguments that are not presented; but it should be realized that my theory is diffusionistic of historical, not psychological. 2, Ritual geometric constructions. According to my paper on the origin of Beometry, ancient geometry, as we find it in Greece and Babylonia for example, 's derivative of a system of ritual geometrical constructions best preserved in the Siabastras, an Indian work on altar construction. A study of this work shows at Was not to make a sturdy construction but to make an exact one. The Sulvasuras in the form we have them have never been assigned a very great antiquity—the date 500B.C. has frequently been mentioned —but the system of ideas there disclosed must, of course, be vastly older if, as | argue, the ‘Babylonian geometry of about 1700 B.C. is a derivative of it. Anyway, the begin rings of the Vedic sacrificial system go back at least to the time of the Rig-Veda (a collection of hymns), which is usually regarded as considerably carlier than the Suleasutras—it has often been supposed that the Rig-Veda refers to the Aryan invasion of India and that this invasion took place about 1500 B.C. The Rig- Veda knows not only the vedi, the sacrificial ground (vedi = earth), but also the threefold disposition of the agni, the fire altar. As described by the Suloasutras, the disposition of the three fires involves the consteuction of straight lines (Le, a series of collinear points), triangles (of prescribed shape), circles, and squares, From the hymns (from which, because of their allusive character, one would Not a priori expect much) one learns at least that “skillful men measure out the seat of the agni”” (Rig-Veda, I, 67, 10).* vay, Ti patge has been Feshly(anlatd forme by Prfetor F Staats felons Like experts a house, they have mad iy messing equal? Yee ee ane ake referee inthe RigeVede, RVI 10,4 sap that “the sift oh featured out the two reins” of Heavan snd Earth RVI HSS, onion “The enlghened sears are freer srcthing a new tng tothe hen ne LI 3,3 say tat “they made both (Heaven ad rh) eet ne RV VL6,3 sae that Api measured out the i pace" and SES eae CF. Guiana, Der ReVeder 4 va "may bef interest fo compar ts wit pasa rom the Exim Elsh (fA. Hs Dt The Babylonian Genes, pA), where two sete, eoentontes ok at and bul The pasage rats "Al, He (Mara) ered the heavens and examined the rei 142. He placed hist oppoie he dpa the cling of aed, 143. The lord measured ut the dmewons of he pes 144, And a great structure, its counterpart, he established, (namely,) Esharra. 145. The great seucture Eshars which he mate se eeeep "4S, Am, Ell, and Ea he (en) caused to hei thes dene,m A, SEIDENDERO ‘Suloa means cord and the basic operation is cord-stretching. It is from this operation, too, that the Egyptian harpenodaptai (whom DesocniTus boasted he could excel in geometric consiructions) received their name. The term was also used in designating ceremonies at the founding of Egyptian temples. In the temple ‘ural paintings of Dendera, Thebes, Esne, and Edfu, one sees the king, 2s sub- stitute for TuotH, engaged with the goddess SAFEKHADUI in a ceremony termed “stretching the cord”. The founding of the temple of Edfu took place, according to Dowicuen, on 23 August 237 B.C.; but cord-stretching can be referred back (0 the time of AeNeMuaT J, founder of the Twelfth Dynasty, and even back to the Old Kingdom?. ‘Sometimes from myths we gei evidence for traditions of sacred constructions. ‘Thus in the mythology of Ancient Iran, “Yima is presented by Ahura Magda With two golden implements, a whip and a goad. Prosperity becomes so abundant that the earth can no longer hold all ‘small and large cattle, men, dogs, birds, and red flaming fires’. At the instigation of Ahura Magda, Yima resorts three times, after periods of three, six. and nine hundred years, to the procedure of extending the earth ‘by one third more than it was before’ by using the instruments Biven to hit ‘A myth is the counterpart of a rite. If we take the above myth as descriptive of a rite, then we have a rite in which the “earth”, that is, the ritual scene, is on Various occasions constructed, with exact specifications, with “instruments”, ‘The above myth may be compared to the known Indian: practice: multipli- cation of the area of the vedi by specified constants on specified occasions explicitly occurs. Thus “the vedi at the sausramani sacrifice was to be the third part of the vedi at the soma sacrifices.” Comparison may also be made with the requirement of the oracle at Delos to double the cubical altar.* Evidence for sacred geometrical constructions is thus embedded in Iranian mythology, but I do not say that the Iranians had the corresponding rites. The ‘myth and the rite can separate and travel independently. To me, the Iranian myth Suggests the custom of sacred constructions; in India we find the proof. But if we did not have this proof, I would still conjecture the existence of the rite from the Iranian myth. In China, Fu-nst (Subduer of Animals) and his sister or consort, NU-KUA, the creatrix of men, appear together; “theit upper bodies are human, but merge ‘below into serpents’ tails that arc intertwined with one another, Fuchsi holds a carpenter's square in his hand and Ni-kua a compass—apparently as symbols Of their constructive activities."” There is also the myth that YO, renowned as the conqueror of the flood and as founder of China’s first hereditary dynasty “ordered two of his officials to pace off the dimensions of the world from east to west and north to south. In this way they determined it to be a perfect square, 2 “Ritual Origin of Geometry", pp. $04, 510. + S.N. KnaMer (ed.), Mythologies of the Ancient World, p. 344. * This statement isto be taken neither as a definition, nor theorem, nor as a dogmatic assertion, but merely as a working hypothesis. © “Ritual Origin of Geometry”, p. 494. 7 KRAMER, op. city p. 386. Ritual Origin of Cirele and Square 273 ‘measuring exactly 233,500 li and 75 paces (roughly 77,833 miles) in each direc- tion." In Rome, “the chief duty of the augurs was to observe ... the omens given by birds, and to mark out the femplum or consecrated space within which the obser- vation took place. ... [The augur] drew with his staff two straight lines cutting one another, the one from north to south, the other from east to west. Then to cach of these straight lines he drew two parallel lines, thus forming a rectangular figure, which he consecrated according to a prescribed form of words. This space, as well as the space corresponding to it in the sky, was called a emplum. At the point of intersection in the center of the rectangular, was erected the raberuiacu- tum"? In Imerina, Madagascar, the “mpanandro”, the maker of days, lays out the foundations of a house, square in shape. Using ropes, he finds the center as inter- section of the diagonals, Thus we sec a ritual personage engaged in geometrical ‘constructions; unfortunately, we learn little about the geometry. The “mpanandro” and the square house with its special subdivisions are part of a complex, of which is also contained in certain Scandinavian and Chinese texts."° The Kwakiutl of Vancouver Island have a way of laying out the lines for a square house. The method is by successive approximations but is otherwise ex- act" With the Omaha Indians, the figure of the earth lodge, mainly a circle, was contidered sacred, “When the location [of the earth lodge) was chosen, a stick was thrust in the spot where the fireplace was to be, one end of a rawhide rope was fastened to the stick and a circle 20 to 60 feet in diameter was drawn on the earth to mark where the wall was to be erected."!? ‘This information is quite remarkable—it is the only case I know of in the ethno- logical literature of the New World where the construction has been explicitly recorded (or its absence—with one exception—explicitly noted).!? To be sure, it is very often clear from the art or religious notions that the concept is known, but the construction is not generally given. Perhaps the observer considers the construction too obvious to deserve special mention. Or perhaps he considers it so sophisticated that no one would suppose it to be of any moment in the tude conditions being reported. Jn connection with the earth lodge, “it is interesting to note that while the ‘Omaha adopted the earth lodge, they did so from a purely practical point of view, as affording them a better permanent dwelling than tents, and were probably ignorant of the symbolic character of the structure. With the tribe from which it was taken [the Arikara] this lodge represented certain religious ideas. Rituals attended the cutting of the trees for its structure and the planting of the four posts. * Ibid, p. 400. » O. Severexr, Dictionary of Classical Amtquities, p. 86. Ritual Origin of Geometry”, p. S21. fbid., p. $22. © Ibid. p. 522. ' 1 think the exception is in a work by Dorsey, but I have lost the reference. For the Old World, or Africa rather, there are a few examples: sce J. WALTON, African Village, pp. 139, 140.m4 A. SeiDeNoeRG that enclosed the space about the central fire. The Omaha did not observe any of these ceremonies nor did they use the prescribed number of posts. They set up about the fireplace six, seven, or cight posts as suited their convenience, for the sole purpose of supporting the roof, these posts possessing no ceremonial impor- tance or other significance.” With the nearby Pawnee, the building of the earth lodge was a rite and “the earth lodge with its dome-shaped roof is likened to the stretch of land bounded by the horizon and roofed by the dome of the sky.""* ‘The Pawnee priest in the great Hako ritual draws a circle with his big left toc ‘and explains: “The cirele is a nest and is drawn by the toe because the eagle builds his nest with its claws. Although we are imitating the bird making its nest, there is another meaning to the action: we are thinking of Tira'wa making the world fot the people to live in. If you go on a high hill and look around, you will see the sky touching the earth on every side, and within this enclosure the people live, So the citcles we have made are not only nests, but they also represent the circle Tira'wa atius has made for the dwelling of all the people. The circle stands for the kinship group, the clan, and the 3 The myth suggests a rite in Which an area is given a circular form and this area is identified with the earth. Nor do we have far to look, for the building of the earth lodge is just such a rite. Here we may also mention the tribal circle. The Omaha, when gathered for ceremonial or for traveling, camped in a circle. The circle had a (symbolic) ‘opening, which was likened to a door, and the circle itself was likened to a dwelling. The opening was toward the east for ceremonial occasions but otherwise toward the direction of travel. The various divisions (gentes) of the tribe always camped in the same relative order about the circle. There were ten such divisions, five of which constituted the Sky People—these camped on the “northern” half of the circle; and five the Earth People, who camped on the “south”."® ‘The Chavante Indians of Brazil have villages in the form of a perfect circle." According to PLuTaRcH, the boundary of Rome was a circle described from a center and drawn with a plough. There is also the tradition that the primeval furrow for the city of RoMULUS, which was called Roma Quadrata, was quad- angular. “In line with it was a building called Quadrata Roma, where the instru- ments for ritual city-building were kept.”"* Actual circular cities are known from the Ancient (and not so ancient) East.!? A.C. Fuercen & F. La Furscut, “The Omaha Tribe", 27 Annual Report of the Bureau of American Ethnology (1905-6), p. 75f; A.C. FLercutn, “The Hako, a Pawnee Ceremony”, 22% Annual Report B.A.E. (1904), p. 33. p. 2036, Tribe”, p. 137 in of Geometey”. p. 522, According to O. Zennuts (in W. Kricke- 21n0, Pre-Columbian American Religions, p. 238), “the circular Apinaye village (of Easter Brazil} and the round meat patties eaten at the festivals represent the sun dss But there is nothing in the sun, or moon, disc to suggest the relation of a circle to its ‘enter, Our guess is that we have here a latter day interpretation of the sacredness of the cree. Lord RAGLAN, The Temple and the House, p. 154, after PLUaRcH, ROMULUS, 11. 49 For some serial photos, sce H.P. L'ORANGE, Studies in the Iconography of the Cosmic Kineship in the Ancient World. ual Origin of Circle and Square 215 In early times the Egyptians regarded the sky as supported on a circle of mountains and their god-houses were (according to Buoct) circular. Later, the temples were rectangular.?° 3. The thesis stated. Enough evidence has now been presented, I think, to show that the circle was sacred because it was the shape of the ritual scene, and ‘that this scene is identified with the carth. But the question remains: why was the ritual scene given a circular shape? 1 will state my thesis first, and then bring in supporting evidence; but even before that, let me recall some ideas expressed in my geometry and counting papers. The main idea is that various widespread ideas are not the spontancous Teactions of the human mind to environing conditions but are the product of special circumstances; and, moreover, that these circumstances were, anciently, the desire to elaborate the ritual. One can formulate the thesis that civilization itself had a ritual origin—the geometry and counting are just special cases. The various rituals we find over the face of the earth, or at any rate a great number of them, fit into a single complex, the great Creation ritual complex (in which, to be sure, a development can be traced). The general underlying identity of the ritual concepts, extending sometimes to minute details, suggests diffusion from a single center (it is not being suggested, however, that all the concepts have a single center, but only that each of them docs). Tn the case of counting, it may be thought that counting is such an abstract, process, so disassociated with anything in particular, that it could not be connected with special circumstances; but that is not s0, and even the number words them- selves contain enough arbitrary, or conventional, elements allowing for a diffu- sionistic study. Moreover, one finds counting associated with curious and baf- fling beliefs, for example, that counting people can kill them, that odd numbers are male, even ones female, that gods are numbers and numbers are gods, that ‘things are numbers, and even that everything is number. All these things suggested {fo me that counting had a single origin and, further, an origin in the Creation ritual complex. In my paper on counting, I envisioned a ritual in which the par- ticipants are called forth onto the ritual scene; though I started, in thought, with no examples, afterwards I found plenty of them; and in coming forth, the participant, in many cases, brings onto the scene, for reasons explained, an object, for example, a stone, which he deposits there.*! The words used in calling forth op. cit, p. 153. ‘tual Origin of Counting”, p. 11. According to A. M. Hocanr (Kiugship, ‘P.70), the coronation of the king was a titual of death and rebirth: the theory is (1) the king dies and (2) is reborn. He and RAGLAN consider that this rite derives from a rite in which the principal actually was killed (see, e.g., Hocanr, Social Origins, Chap. 1 Ragtan, The Origins of Religion, Chap. 10); and I tend to be persuaded by th ments. The basic point in my analysis is that the ritualist are (or were) deflecting or ‘ide-stepping some supposed danger to the principal when he appears on the ritual scene; whether this supposed danger ever was a real danger, however, itis not vital for our purposes to decide. For a polemic against the ritual theory of myth, see J. FONTENROSE, “The Ritual Theory of Myth. Although quite out of sympathy with the “ritualist hypothesis", at the end he concedes that it may be right (op. cit, p. 60).216 Seip: xoens the participants become numbers; counting is, in origin, a creation myth. So, at any rate, my theory goes; and it sheds considerable light on the facts as we find them, Now let me state my thesis on the origin of the circle. In the ritual, the participant was called onto the ritual scene; but to avoid the announcement of his name, some article, for example, a stone or a tree, was called ‘onto the scene, and carried there by the participant, This led to the identification of the participant and the article in question. The identification had two aspects: ‘on the one hand, the deification of (say) the stone or the tree; and bn the other, the petrifcation or arborization of the god, that is, the participant in ritual. The processes thus indicated on the one hand gave rise to the belief that inanimate ‘objects such as stones are, or can become, alive; and on the other, to the belief that men can turn into stones of tres or other objects. As the ritual was elaborated, ‘more and more things were called onto the ritual scene, in particular, stars. Here ‘again we have two processes in reverse directions going on. On the one hand, the stars are brought to carth; and on the other, the gods are lifted to the sky. Once the gods are in the sky, the gods ia heaven are watched in order that the gods on earth may'act accordingly. This gives rise to the observation that the ‘gods in heaven move in circles; and hence the gods on earth move in circles and give the ritual scene a circular shape. 4. Circular movement in the sky and earth. The foregoing is, of course, a theory, but I suppose the reader sufficiently familiar with archaic thought to realize that underlying it are plenty of facis, that, for example, stones are (or can bbc) gods, that gods turn into trees, etc. That the stars are gods is too well-known to require testimony here. That the earth and sky (or heaven) are counterparts, is also well-known, but we may cite soine opinion on this, In my paper on geometry I cited THisaur, who says: “... The want of some norm by which to fix the right time for the sacrifices gave the first impulse to Astronomical observation ...". A similar suggestion has been made for Egypt by Foucart. RaGtan, writing of the calendar (and largely following Foucart), says: “In the previous chapter we saw that in parts at least of the Ancient East there was an annual ritual combat between the king, as the representative of the powers of light and life, and an antagonist who represented the powers of dark- ness and death. When in Egypt, and probably elsewhere, it came to be believed that the king was the incarnation of an invisible god, and that this god resided in the sky, it came to be supposed that he was manifested in the sun and the other heavenly bodies, and these bodies were believed to carry on the mythical and ritual activities of the king. Gradually the real scene of the ritual is transferred, in belief, from the earth to the sky, and [citing Foucant, Encyclopaedia of Religion and Ethics, vol. 3, p. 98] ‘what the Pharaoh does on earth is metely the repetition of the legendary Divine actions’. These actions are elaborated, but the ritual on earth ‘ust correspond exactly in time as well as in character with the ritual on high, ‘and this was the origin of the calendar. In the combat which took place on earth between the powers of light and darkness men had ensured victory to the powers ‘of light, and when this combat was transferred to heaven it came to be believed Ritual Origin of Circle and Square 21 that men, by their activities in the ritual combat on earth, could still influence its decision. They could help to defeat the demons who were threatening the gods. “In associating,’ says Foucart ibid., p. 99] ‘with the science of the calendar t possibility of cooperating, exactly at the propitious moment, in the struggle for good, Egyptian religion was able to guarantee that, if the same gods (or theit mimetic substitutes) repeated the same acts in the same places (or in their equi valents by ‘geographical magic’), and on the same days (fixed by the calendar), the order of the world was assured. ... The intervention of man, forescen and organized on certain fixed dates, arranged and defined relations with the gods, and the connections with them ... [The] essential fact was the possibility of man’s helping the powers that were regarded as good to struggle against those that were regarded as evil.”?? Thus RAGLAN places the locale of the ritual originally on earth, and this was “gradually ... transferred, in belief, ... to the sky”, but he says nothing as to why ‘or how this transfer took place. Note that my theory explains this, 5. Evidence from myths. Let us come now to the evidence for my theory. ‘This will be in the form of myths and rites. The rites are harder to come by, so We first consider the myths. Since a myth is the counterpart of a rite, this need not disturb us provisionally In Vishnu Purana (ii, 12, 24) we are told that “all the celestial lights are in fact bound to the polar star by aerial cords", and in the Vayu Purana that “the seven Maruts drive the stars, which are bound to it by tis invisible to man, round the pole”.2? We cannot get better evidence than this from the myths: the myth is the precise counterpart (except for the invisibility) of the rite I am projecting. Presumably the circumambulations were originally without cords, and I suggest that the idea of introducing a cord, thus producing a perfect circle, arose in this context. [ can't prove this, but can’t make a more plausible suggestion. In Quechuan symbolism in Peru, “the sun was tied to the invisible pole of the sky and was driven around it like a llama by the power of the Universal Spi- ries Though not so clearly, perhaps the same idea can still be detected in a myth about Maut, the Polynesian culture hero. MaUt traps the sun using ‘royal nooses’, six of them; the sixth finally stops the sun, and one end is fastened to a point of rock.25 The following really refers to a myth, but requires comment. According to ANAXAGORAS, “the stars were originally carried round (laterally) like a dome, the pole which is always visible being vertically above the carth, and it was only after- 22 Origins of Religion, p. 996. 2) H. Jacoat, “The Antiquity of Vedic Culture”, J. Royal Asiatic Sot Z. Nurrat, “The Fundamental Principles of Old and New World Civil and Ethn. Papers of the Peabody Museum, vol. 2 (1901), pp. 448449, 3+ Encyclopaedia of Religion and Ethics, (J. HASTINGS, 4.) V. 12, p. 68. See also D.G. BRINTON, Myths of the New World, p. 72. 35 W.W. GILL, Myths aad Songs of the South Pacific, p. 61. 1910, p. 462. ns", Arch,28 A, Suipenotra wards that theie courses became inclined.” The comment is that ANAXAGORAS, like anyone else, started with what was historicaly given, and went on from there. He inherited the myth that the stars in the beginning moved horizontally, but he mistook a description of a ritual Creation for a description of actual creation. His statement (or possibly a predecessor's) is thus a combination of an inherited ‘myth and a current observation.?® . In the Japanese cosmogony the predecessor or “father” of our present sun betins his activities in the new-ereated world by repeatedly performing, in a Aorizontal plane, a citcumambulation of the “Island of the Congealed Drop.’ 6. Eridence from myths, continued. I now give some myths which contain at least partially what 1 want. ‘The king of the Akan, Altica, told E. Meverowrrz: “I am the center of the worl round which everything reveles."# We ean understand the king's laim {o be the center—we see this kind of thing going on all about us—but where di he get the idea that the world revolves? ‘Among the Dogon of French West Africa, the staff of the Hogon or para- ‘mount chief is called ‘the axis of the world’.?? Similar ideas are to be found in the Ancient East, Japan, and elsewhere. In the Ancient Near East, the king was “The Axis and Pole of the World.” According to an ancient book, the Kojiki, of the Japanese, the creators and fist inhabitants were a god and goddess, IzaNAct and IZANANt by name. These, according to Wanacn, following E. REED, “standing on the bridge of heaven, pushed down a spear into the green plain of the sea and stirred it round and round ... {then} nd it, taking planting a spear in the ground, point downwards, built a palace arour that forthe central roo pillar: The speat became the exis ofthe earth, which had been caused to revolve by the stirring round.” Various tribes of North America and North Asia, for example, the Algonquins and the Samoyeds, liken the central pillar of their dwellings to the cosmic axis.2° 7. Further evidence from myths. I now give some myths which are (or appear the construction of a circle, cate god Tezearuivoca had hed one of his fet bite of In hieroglyph, ‘TezcaTuipoca is seen with the ankle of his foot held in the mouth of a tecpatl, symbol of the North. On the basis of this, one may attribute to the Aztecs the Knowledge that a cirle is the locus of points at a given distance from a given indeed, the god TeZcaTLIPOCA here soundsto me like a pair of compasses. In connection with this myth about TexcaTLtroca, we recall the rite already ‘mentioned of the Pawnce priest who draws a circle with his big left toe, We may 267. Heat, Aristarchus of Samos, p. 82. 3 MF Wanna Preis Found,» 182. #8 E, Mevenowttz, The Sacred State of the Akan, p. 57. » Temple and House, p. 140, after M. Griaute in African Worlds, ed. D. Foxe, 101, PSS Loman, op. city p13. W.F. Waren, Paradise Found, p. 140. Temple and House, p. 142 (after M. ELIADE in Patterns of Comparative Religion). 3 Ritual Origin of Geometry, p. 522. Ritual Origin of Cirele and Square 279 also note that it was TexcatLiroca’s right foot that was bitten off, which is the right one if he is to move around in a circle in imitation of the stars. In the Bhagavata-purana (Chap. iv) it is said that “Dhruva, meditating on Brahma, stood on one foot, motionless as a post; while he did so half the world, ‘wound by his big toe, bent over under his weight like a boat which, bearing 2 orous elephant, leans at each step he takes, from left to right.” Nutrat. (op. it p. 448), following Jou O'NEILL, giving a citation from the same source, ce around the center sa | ‘mira he Hidatsa leave the lodge fot the Sun Dance ceremony] 1 Beaver Indians who occupy the regi alte? ccupy the region from the eastern see om ety see ckingly lacking in religious ceremonias. P. E. GODDARD wits: “The see sere atvity of striking interests the semi-annual gatherings when oferings fire... large dance ground is fenced and a central fre prepared Op. cit. (1945), p. 758. Spier, p. 474,302 ‘A. SeIDENBERG The religious observance is followed by dar men and women who circle the fire clockwise.’ ‘Thus North America east of the Rockies and south of Canada is pretty well covered. The Californian Indians also know the circular dance, Thus the Karok made a dance for adolescent girls. After a preliminary dance chiefly by the men there “followed a round dance such as is most common in the ceremony elsewhere in California. A ring of men surrounded the maiden, a circle of women stood outside, and both revolved dextrally.” The Yuki have a dance, largely soci character, in which “the dancing is first in a revolving file, then abreast.” The Yokuts had an annual public mourning ceremony that continued for several ts. Among the Chukchansi Yokuts, on the final night, there are two fires, ‘one for the men, the other for the women. The dancing around each fire is counter- clockwise. “Once during the nigkt, and again toward morning, the men and women change to each others’ fires. 1 will not consider the Eskimo, since I am more concerned about the areas for which we do not have already some information on the circle. As for the rest of North America (north of Mexico) I have no information on the inland country south of the Eskimo on the Wesi, but for the East, in Labrador, I do have a little. ‘The Naskapi Indians of Labrador live on a bare subsistence level. This results ‘not only from their poverty in material culture, which has been called mesolithic, but also from the harshness of the environment. Reading of them, one almost tenvies the bear, who at Jeast can hibernate. Related to this poverty, or type of life, is also a poverty in ceremonial and even in social organization. The Naskapi have a so-called “snow-shoe culture”; the snowshoe allows them, as long as there is snow on the ground, to chase and kill the animals, especially the caribou, who are hindered by the snow. When the snow melts, the hunting season is over. The small groups, some of which may consist of a single individual, gather, and there is some social life. Even then there is little, and “we observe only the semi- religious social round dances and games at such times.””? le also enters into the ritual of the hunting of the bear (and possibly als). After the bear is killed, the hunter sits down near it and smokes, lays the bear out with crossed paws, efc. He then dances around the fallen game, at the same time singing. There may be also some consequent rejoicing by dancing around the fire.%* Now the question is: How did the circle get into Naskapi culture? Did it perhaps drift up from Mexico? Or was it diffused, say along with the snowshoe, from Asia? Depending on the answer, we will get different time perspectives on the circle. There is an a priori chance that one could get an answer as follows. If one could find a cultural complex confined to the boreal peoples and if in each case (i.e, ing throughout the night by the 2" PE. Gopoaro, “The Beaver Indians”, Anthropological Papers of the American ‘Museum of Natural History, X, pi. 4, 1916, p. 228. * AL. Knototn, Handbook of the Indians of California, vol.2, pp. 106, 196, 501, See also p. 300. 22 F.G.Srecx, Naskopi, The Savage Hunters of the Labrador Peninsula, p. 21. % Ibid, p. 208. Ritual Origin of Circle and Square 302 People) one would find the circle at the same point of the complex, then one could Teasonably conclude that the circle moved with the complex, Now I had noticed that the circle enters the bear rituals of the Lapps and of the Naskapi, and this led me to A. I. HALLOWELL'S paper, ‘Bear Ceremonialism in the Northern Hemisphere."*$ Ia six instances there is information bearing on the question at issue. While this is neither as ample nor as uniform as one might have hoped for, still it is noteworthy. In bear ceremonialism we do have a complex of the described kind, for, first, it is confined to the habitat of the bear, and, second, it is a complex, not only in that it involves reverence for the bear, which by some would be regarded as generic, but also in specific details, for example, the custom of addressing the bear as “Grandfather” —this occurs amongst the Naskapi, the Penobscot, the Ojibway, the Takitan, the Tsimshian (in North America); the Yukaghir, Tungus, Ostyak, and Samoyed (in Asia).°® As for the circle, as amongst the Naskapi, so amongst the Ostyak of western Siberia, in the region of the Ob river and its tributaries, “after a bear has been killed his body is placed on the ground and the people dance around it.” Amongst the Laps, at the end of the bear ceremoni returning to their own dwellings each man takes hold of the chain on which the kettle hangs and after dancing three times around the hearth runs out through ‘the common door of the hut.” “The Reindeer Koryak (of Siberia] staughter a indeer for the bear, cook all the meat and pack it in a grass bag, The bear skin is filled with grass, taken out and carried around the house, following the course of the sun,?? and then sent away in the direction ofthe rising sun ...” With the Gi lyak, of the Amur Gulf of Tartary, “before the bear is killed, itis led around the host's house three times and finally into it.” With the Yezo Ainu, “nao are placed in certain parts of the house (of the host}, on the nusa and on the four corners of the bear cage . ... the bear is offered food and saké which is followed by a dance of the women and gitls around the cage, accompanied by singin Regarding this evidence, there are essentially two possi ither the circle is in each case merely a local cross with the bear ceremonial, or the circle ‘occurred in the bear ceremonial at the beginning of its dispersal and was diffused along with it. The variation in detail favors the former possibility; yet there is enough uniformity at least to suggest that the Naskapi circling of the fallen bear was obtained from Aisa along with the beat ceremonial complex. If this conclusion is adopted, it would then appear that the occurrence of the circle in the bear complex is older than its occurrence in Kwakiutl ceremony. However, the evidence is hardly adequate for drawing this conclusion with a great deal of confidence. In continuing with the circle dance (or movement), I will not go over again the areas of the “sacred cirecumambulation”, as it will suffice for my purposes still to look only at some outlying parts of the world. Before leaving the boreal regions, I will still note that the circle dance has been % American Anthropologist, vol. 28 (1926). %* HaLLowELL, op. cit, pp. 44, 45, 47, 48, 49. °” It is impossible to tell whether this description is the natives’, or is duc to the ethnographer himself,304 ‘A. Se1DENDERG observed among the Tungus, Yukagir, and Yakut, of As “move slowly from right to left in a sunwise direction. Jn China, at a burial, “after the body has been placed in the tomb, the mourn- ers join hands and perform a sort of merry-go-round about the tomb; they repeat three days later.”? In Africa the Dinka of the White Nile danced round their slain enemies. The Latuka tribe of the Upper Nile diversified their dances in honor of the dead “every now and then by acircle-dance in which the women [who otherwise danced separated from the men) joined in with the men.” The Gallas of East Africa “dance in couples round sacred trees, praying for a good harvest.” Further south, the Bushmen had many circular dances. In one, they danced around a sick man to make him well. In another, the women moved in a circle; “the chief took up his position in the center, and frequently hopped and sprang round on all fours like some animal, the women in the meanwhile dancing and placing themselves in every possible lascivious position, until the great man in the center pounced ‘upon one of those who had most distinguished themselves and performed in the sight of all that which in more civilized communities is reserved for the strictest vacy, amid the applauding clu:ter of the excited dancers forming the enclosing Circle. After this the chief again took up his original position, and the dance continued with the same repetitions until all engaged in it were wearied and ‘exhausted,"102 tic Russia, The dancers “The people of Tami, an island in the Indian archipelago, belong to the Melanesian stock; when they mcurn for the dead the whole village takes part in the lamentation ... [They] have regular dancing seasons during which they dance Found men disguised as familiar spirits; true, the dance ‘consists of little more than running round and round in a circle, with an occasional hop’."!° “The inhabitants of Dutch New Guinea dance round the images of their departed on various festal occasions."102 ‘Some Australian tribes have carth burial, some tree burial, and others thi tree burial followed by the earth. The Warramunga of North Central Australia hhave tree burial. One ceremony is a visit to the tree grave by relatives of the dead, “who march round and round the tree singing out loudly ‘Ah! ah! ah!” There ‘are other circumambulations in the rites and on one occasion it is (exactly) two- fold. The circle also enters elsewhere into the rites of the Australians. For example, i jon ceremony of the Aranda two young men, holding cach other by a little finger, run clockwise around a circle, while another couple, starting, 98 W. Jocutson, Peoples of Asiatic Russia, p. 218. % OrsrenLy, op. eit p.201 after Wass in Ene. Rel, and Ethics, vol. 4, p. 4830. "© OEsTERLY, op. cit, pp. 206, 205, 103. G. W. Sow, The Native Races of South i PP. 120, 118K, # OrstenLy, p. 214, after Frszen, The Belief in Immortality, vol. 1, p. 293f. 498 Ossterty, p. 212. Ritual Origin of Circle and Square 305 simultaneously from the same point, likewise runs counterclockwise around the same circle! ‘This information on the circle dance could be expanded, I'm sure, and this, ‘would be desirable, but the picture is clear enough for us to see that the ritual circular movement, whether as a “sacred citcumambulation” or otherwise, ‘occurs, and even occurs intensely, well outside of the regions FRoveNius would call solar and in all three strata (solar, lunar, fossil) that he delineates. To come to my judgment of FROBENIUS'S map: its lack of detail makes it misleading. FRonenius may, indeed, be right after all that the“sacred circumambu- lation” is a solar phenomenon, but his map cannot be the basis for such an esti- mate. ‘My conclusion is this: there is nothing in the distributions to suggest that the ritual circular movement is solar; on the contrary, it looks pre-solar. There is nothing in the distributions not in accord with my notions of the origin of the circle. Il, The Square 16. Duality of circle and square in ancient mathematics. We now come to the square, The square is considerably more difficult to get at than the circle. ‘As with the circle, my starting point is the history of mathematics. Already at the beginning of the second millenium B.C. we see the circle and the square being treated as mathematical objects by the Babylonians and Egyptians. This ma- thematics has an advanced Jook about it, but it is impossible to trace a development int. As I have suggested, however, one can discern the roots of this mathematics in the Sulvasutras, an Indian work on the construction of altars. A highlight of the Suloasutras is the use of the Theorem of PYTHAGORAS, The construction of altars of various shapes is described, the shape depending upon the particular ritual. Thus there were square altars, circular altars, and altars of many other shapes, in particular, the falcon-shaped altar, which was, in its sim- plest form, composed of squares and rectangles, Most, though not al, of the altars had a level surface, and these were referred to in accordance with the shape and arca of the top (or bottom) face. The basic falcon-shaped altar had an area of 7$ square units. This was its size upon its first construction. On the second construction, ‘one square unit was to be added, that is, the area of the second altar constructed ‘was 8! square units, the shape remaining the same; on the next construction ‘another square unit is added, and s0 on, until one comes to the one hundred and ‘one (and a half}-fold altar. The problem of finding a square equal in area to twe ven squares is actually and explicitly involved; the constructions are carried out Using the Theorem of PYTHAGORAS. The problem of converting a rectangle into 9 square is also explicitly involved. This is not practical knowledge; rather, fi mathemati symbolic knowledge. 199 B, Spencer & F.J.GULEN, The Northern Tribes of Central Australia, pp. 527. 535, 195. C. Srrentow, Die Aranda und Loritja-Stimme in Zentral-Australien, WV, | p36.306 A. SEDENtERG “As we have mentioned, the shape of the altar varied the sacrifice. One of these shapes was circular, another squ: forms us that theological controve the purpose of and THBAUT in- Arose over the question as to which shape iion—according to HOcaRr, the same contro- thy ofthe shape of the buria! mound split Northern India into two schools; ike School of the Brahmanas, which considered itself orthodex, built square ‘mounds in which to bury persors who had reached a certain degree in the curr, ulum of the sacrifice. The heretics made their burial mounds circular” In there Controversies, those concerning the altars at any rate, the area was understood {o be constant, and this led, as we suggest, to the problems of equating the circle and of turning the square into « circle, In the Sufoasuiras there are attempts at solving these problems."!®¢ ‘Thus at the very roots of mathematics we se that the circle and square are dual figures. Our first thesis, or guess, is that the square ritual scene was introduced 3 dual to the cicluar ritual scene. This thesis has chronological connotations, ramely, that the circle is older than the square. So it may be well first to consider this chronological question. There are a priori reasons for thinking that the circle must have come before the square, Just as itis more dificult to find the area of the surface of a sphere than 19 find the area of a rectangle, soit is more diffult to construct a square than to construet a circle. However, in a historical study, the question munt be whether the evidence tends to verify such a surmise, 17. Round and square houses. That the circle and square have ritual origins i think, sufficiently clear, but the issue is, where inthe ritual did they begin Whatever the specific point of origin, however, we may expect that these ewe ‘ould then impose themselves on the various aspects of culture. Burial mounds, altars, houses are all cosmic constructions, Even if the circle and square do not siatt as house forms—and, indeed, I do not suppose they do—still we may look at house forms to get a time seyuence on the circle and square. We have already noted that the earliest houses known, those from Jericho bout 7000 B.C., are round; and that the oldest known sacred buildings, at Tall Apachiya, about $000 B.C. are also round. At Jericho, KATHLEEN Kenyon, following a find of a predecessor, uncovered a stratum she later called the Pret patery Neolithic B (which lay under Pottery Neolithic A, above whichis Pottery Neolithic B). “It is characterized by well-built houses with rectangular rooms, rectilinear and vertical walls, and wide doorways.” Digging further she found Stratum which she called Pre-Pottery Neolithic A. Here the houses are “round OF at least curvilinear in plan.” The settlement was surrounded by a defensive Wall and there is a huge tower which, judging from the photographs, looks prelty circular. There isa break in the stratigraphic record between Pre-Pottery Neolithic A and Pre-Pottery Neolithic B: the Pre-Pottery Neolithic A settlement was aban, Soned, or wiped out, and “the Pre-Pottery Neolithic B people at Jericho, with thelr Tehunian equipment, arrived with a fully developed culture ... ." She sought tual Origin of Geometry”, p. 492f. Ritual Origin of Cirele and Square 307 {0 trace the earlier settlement back to its beginnings, and succeeded: in one area, beneath a long succession of the typical round houses of the Pre-Pottery Neo. thi A stage, there was a deposit thirten feet in depth which was eatlier than the beginnings of architecture.” The people “lived more or less permanently by the spring, but in dwellings [or hut-like shelters} more suited to a nomadic than a settled existence.” At the very first occupation of the sit structure: “itis a platform of clay, bounded by substantial stone walls”, an pingoubtedly Mesolithic.” There is no break in the record up to Pre-Pottery Neolithic B.'0s For some of the stages of Pre-Pottery Neolithic B, KENYON at first obtained Carbon 14 dates of c. $850 B.C. and c, 6250 B.C. Later, for Pre-Pottery Neolithic A the date c. 6770 B.C. was obtained, and she extrapolated back to 7000 B.C. The “Mesolithic” structure described yielded a date of c. 7800 B.C.1°° Somewhat before Kenvon’s work, Rovert J. Bratpwooo had found at Jarmo, in Iraq, a permanent village in which the houses were "substantial recti- linear houses"; three Carbon 14 datings yielded a central figure of c. 4750 B.C. After the announcement of the 5850 and the 6250, BrarDwooD wrote an article expressing considerable skepticism, on several grounds. In particular, he cautioned against a too easy acceptance of Carbon 14 dates. According to BRAIDWoOD, doesn’t figure!” He was also concerned that Kenvon’s announcement would have a “bandwagon” effect: thus it depressed him to note that Max MaLLowaN ow suggests ‘e. 5000 B.C. rather than 4000 B.C.’ for Arpachiyah, Later, however, Braiowooo obtained much cartier dates at Jarmo (in particular, 7080 B.C. and 6870 B.C) and apologized for doubting the “now modest” dates of 5850 and $250 B.C. Kenyon in turn gave reasons for doubting BRAIDWwOOO'S dates (of 7080 and 6870),1°7 Regardless of the absolute dates, though, at Jericho the round houses came before, and ages before, the rectangular. Does this prove thatthe circle came before the square? Of course not. sin the earliest civilization of Egypt, that of Merinde, the houses were of mud in the form of beehives. In Early Predynastic times the huts were round, with foors partly excavated, but in Middle Predynastic they were rectangular, of wattle and daub, with timber-framed doorways,"!08 iu? Kanteen M. Kenvon, “Some Observations on the Beginnings of Settlement in the Near East”, J. of the Royal Anthropological Inst., vol. 89 (1959), pp. 35-43, and “Jericho and its Seting in Neat Eastern History”, Antiquiy, vol, 30 1936), pp. 184 197 (including Carbon 14 report by F. E. Zeunsn). For photos of the circular tower See loc. cit. (1956), plates opposite p. 212 and p. 129; and loc. ci. (1959), plates following p.42. For further information on round forms in the Near East, see the article "The beginning of village life”, by Janes MeLtaar in The Davn of Civilization, ed. by S. Pio Gort. In particular, note on p. 43 the circular parapet of the last resting-place at Eynan of a Natufian chieftain of the 8" millenium B.C. 19 Kenyon, 1956 and 1959, 7, BRAIDWOOD, Antiquity, vol. 31 (1957), pp. 73-81, and Kenvon, 1959. © Temple and House. 0. 182308, A. SHIDENBERG “In Egypt the change from round to the rectangular occurs in tombs as well ‘asin temples. By the beginning of the first dynasty the graves of kings, nobles and their retainers were surmounted by a rectangular structure of brick filled with ‘rubble; those of the peasantry and poorer classes were covered by a low circular mound, Later in the dynasty the graves of important commoners are still covered by a circular tumulus, but have a rectangular lining, By the end of the second dynasty graves of commoners have begun to assume the same form as those of the upper classes." During the third millenium B.C. immigrants from Asia Minor moved up the Danube Valley, while others coasted around the Mediterranean. The former appear always to have had rectngular houses, the latter round. This fits with the report that in Central Europe no round houses have been found. The earliest form for the house in Ireland was round and the round form remained dominant until early medieval times, though a rectangular room ad- joined to a circular house of the early Christian period has been found. Moreover feven a large rectangular Neolithic house, of a kind similar to many found in Central Europe and Scandinavia was excavated in County Limerick. One can visualize the circle culture moving around the Mediterranean, the square culture moving through Central Europe, the two mecting in Ircland.'*® In Scotland the rectangular “black houses’ of the Hebrides were preceded by beehive-shaped stone huts.!"* In China, Neolithic houses are reported as bechive-shaped—this shape was characteristic of the Neolithic culture of an area including Siberia and North ‘America; though at Pan P’o Ts'un in Shenshi round and oblong Neolithic foun- dations were found close together. The round ones probably had walls a foot high, thus were more advanced than the bechive-shaped houses. The evidence, ot report oft, is not entirely clear, but it appears that round begins to give way to rectangu+ lar about 1400 B.C,"#? ‘According to RAGLAN, following T. T. WATERMAN (Native Houses of Western North America), “along the wes: coast there is a strip about fifty to 200 miles wide and about 2200 miles long, extending from Alaska to California, in which all the inhabitants used to live in rectangular plank houses, some of them very large. Mostly inside this strip, but extending 200 miles to the south of it, is another strip about 1400 miles long and 100 to 200 miles wide, in which, at an earlier period, all the inhabitants lived in circular huts dug into the ground and entered by a ladder through a hole in the earth-covered roof.”!"? From this we may with some confidence conclude that the round house reached the Northwest Coast before the rectangular house. We say “reached” since we suppose these forms were invented in Asia. Ibid, p.158. Ibid., p. 186. Cf. S. Plocorr, Ancient Europe, entry "house, circular’, p. 322. 111 Temple and House, p. 186. 193 Dbids, 184. §9 Ibid, 9.189. Ritual Origin of Cirele and Square 309 “A well-known and interesting area where square houses are known to have followed round houses in point of time, is our Southwest, where the well-known Pueblos and Cliff-dwellers were preceded by Basket-makers. The Pueblos built square habitations, while the more ancient Basket-makers built themselves round Ihuts."114 ‘One can add to the archeological evidence, but the picture does not become clearer. One sees, then, in some places the round gives way to the rectangular, ‘whereas (with one exception to be mentioned below) the rectangular nowhere ives way to the round. Scanty though it be, the archeological evidence indicates the priority of the circle over the square. ‘We may still lance at the ethnographical evidence. In the Americas, the North- west Coast, as we have already noted, is rectangular. With this exception, the Amer- icas can be roughly divided into three parts: in the norchern and southern extrem- ities, in the culturally more impoverished areas, the house forms are round, whereas in the middle, the houses are rectangular. According to RAGLAN, the evidence from South America is not very satisfactory; bat at any rate, the rectangu- lar forms are confined (in South America) to the west and northwest, The distribu- tions, then, do suggest the priority of the circle over the square. T.T. WATERMAN has noted that “there are several known cases also where [American Indian) tribes built rectangular rooms to live in, but round structures for religious uses. Thus the Pueblos lived in rectangular rooms, while their cere- monial chambers or kioas were circular. This always suggests that religious cham- bers are a survival of an older style of building ... Tne Natchez tribe who lived at the mouth of the Mississippi built rectangular wooden houses, but had a circu- lar “temple”. The latter was a wooden structure on a mound with an image of the sun in it"''3 RAGLAN, taking note of a similar phenomenon in various parts of the world, follows the same type of explanation, as when, after recalling that the hut of ‘RoMULUS and a similar structure in the temple of Jurrren in the Capitol were round, says: “It would scem that in Italy, as elsewhere, round houses were once genera but were later replaced by square [i.e., rectangular] ones except where they sur- vived for religious reasons or in remote areas."""® ‘On the other hand, RAGLAN writes: “It seems odd that the Mandans, a Siouan tribe, once had large rectangular lodges, but in the 16% century adopted round ‘earth lodges, retaining rectangular ones for ceremonics."!*7 This sugests that WareRwan’s and RAGLAN's explanation, plausible as it is, is simplistic, It also seems to contain a rationalistic element: the separation of the 146 TT, WATERMAN, “The Architecture of the American Indian”, in A, L. Kros en & T.T. WATERMAN, Source Book in Anthropology, p.517, of American Anthropo- logist, ns, vol. 29 (1927). M5 Loe. elt, p. 517. Temple and House, p. 185%. bid, p. 190. Piacorr (op. eit, pp. 199, 2326), speaking of Heath Row site in England, “probably of the fourth-third centuries B.C.” mentions a “rectangular (ac- ‘ually square) building interpreted as a temple, among circular houses.”310 A. SHDENDENG sacred and secular is clear enough in the West, but is either lacking or not so marked in archaic cultures. Rather, I would suggest, a pervasive duality in archaic thought is involved. The gods and men, ceremonialism and daily life, are put into opposition. Hence ti forms for gods and for men should be d In my geometry papes, 1 took note of an ancient duality not only between circle and square but also between square and oblong. This latter duality left (mathematically insignificant) traces in the mathematics of ancient Greece and India. Aside from that, as T mention there, the oblong was important for PYTHAG- fonas: his first principles are ten in number and consist of pairs of opposites, =g., odd-even, male-femile, efe., and one of the pairs is square-oblong. The ime dusty: oblong bricks are human, square bricks divine. As is found as far away as Fiji: according to Ho- cant (Kingship, p. 168), “the Fijians who dwelt round the Koro Sea built oblong hhouses, but their temples :re usually square.” In other words, even when the circle is completely ousted, the duality persists in an opposition of square and ob- long.* To return to the circle and square, “in Polynesia, houses are in general square {éc., rectangular}, but in Tonga and Samoa there are large round houses, like rondavels but with convex roofs. These are used for ceremonial purposes."""¥ According to a map of J. WALTON, the distribution of hut types in Africa divides Africa south of the Sahara, minus an area including the eastern tip, in three fairly well-marked arcas: (i) a large area southwest of a line which extends roughly from the middle of Angola to the north of Natal (i) a lrge area of West 1g as far west as Sierra Leone, (ii) the rest of Africa south of a line drawn from the Sudan coast to Senegal, and comprising about three fifths of the whole. In area (ii) the houses are rectangular, in areas (i) and (ii) round. The round can be subdivided into two types: (a) the bec-hive shaped hut, consisting of a symmetrical dome of * The duality of circle und oblong was also known in Peru, as appears from the following passages from V. CLiFron’s Vision of Peru, pp. 28, 37: 1.28, “Before he died Manco Capac made a seared place in Cuzco called Cori- Concha, golden, and this w:s the temple of the Vira Cocha, the Teacher of the World; he ordered a Nat plate of gcid, to be made oblong in shape. A voice of the country says ‘was not to show the Sun, but that it signified the Foundation, the Abyss. Oblong and e are the same sign because the circle and the oblong demand the same room in space. Throughout the earth there is no life that comes from any shape but this, which shape is that of the cell.” p.37, “Mayla rebuilt the Temple of Manco Cépac and renewed the golden oblong plate, The Incas that came sfter him set up in the temples plates with rays outjutting to show the sun, but Manco King and now Mayta King honored the sign of the circle. However, since the minds of this childish people turned toward the appearances of Nature rather than towards the Foundation, the Abyss, this symbol of the circle in later times seemed to have becn confused with the worship of the Sun.” “The third sentence from the passaye on p. 28 appears (0 mean that circle and oblong. are to be considered equivatent if they have the same area. "© Temple and House, 9. 188. in of Circle and Square «uniform material throughout starting from the ground and meet and (b) the rondavel, consisting of a conical roof placed on a cylin five feet high. Type (a) occurs throughout area (i) with pockets of it in which otherwise is covered by type (b). (Unlike the bechive-shaped huts of Ei and North America, those of Africa seem never to have been built over pi The distributions suggest that the bechive-shaped hut was general in (South of the Sahara) and that a later dispersal of the rondavel ousted it « in the southwest and a few pockets. The square looks younger. At first, 1 surprising, since the west coast of Aftica, viewed from Asia, is peripheral may therefore suspect, as others have, that the square came into Africa water. From areas (i) and (ii) we ean expect no evidenes of a change from rou square (or vice-versa) but with the Yoruba of Nigeria, whose buildings arc rectangular, the temple of the thunder cult is “round (old style)"."2° In the last few examples, I think one can perceive a duality. We may mention India (where this point is not at issuc). “In India, the great ma, of the houses are rectangular, but round huts are not uncommon. The Chen a forest tribe of Central India, make round bamboo huts with conical tha: roofs, that is, rondavels.” The Wadders, a low caste spread over Southern | ppear {o live in rondavels as do the Andaman Islanders.'?# Although there is a tendency for the rectangular house to displace the r« still the round can hold its own in advanced conditions. Thus the round hous. ‘occurs in southern Italy and in Provence and, as just noted, also in India. The ethnographicat evidence on house forms, then, also tends to show the circle comes before the square. As remarked at the beginning. this is hi surprising. It is always well to look at the evidence, however, in the cour which a duality between circle and square was disclosed. 17 bis. Circular and square earthworks in North America, In the eastern Of the United States there are many ancient earthworks. A large number of hhave been called circles or squares (as the case may be). Careful survey sh+ however, that often, or perhaps usually, they deviate quite a bit from exact figt Still, someare pretty exact. Thus in Ohio, at Newark, thereisa circular construc which differs from a true circle of diameter 1054 feet at no point more than 4 and so is exact within an error of less than 1% (and there isa square of sides « {925 feet of about the same accuracy). One can draw a small circle, say of abo foot radius, freehand pretty accurately, but 1am skeptical that a similar tT could be done for a circle of radius about $00 feet. Sometimes a circle an square occur in association, with a path leading from one to the other; and 30 times the square is tangent to the circle at the midpoint of a side.'2# One can the impression that these paired enclosures were dual ritual scenes. Mid, pe ALE. 12° Poids, p.AST. 120 bids, p. 1ST. 133 G. Fom rchaclogical History of Ohio. Chap. 8. es0. op. 68:132 ‘A, SeiDENUEG. 18, Duality of circle and square in a ization. ( now give some examples in which the duality of the circle and square is clear, and even expli The Altar of Heaven in Peking is a magnificent structure of white marble 27 feet high, composed of three circular terraces, the lowest of which is 210 feet in diameter, the middle 150, the upper 90. Some distance north is the Altar of Earth, a yellow edifice of two square terraces, each 6 feet in height, the lower 100 feet square, the upper 60 feet square.'?? The duality in this information may be displayed as follows: Heaven: Earth = circle: square = three: two (\erraces) 3x3:3x2 (height per terrace) = South: North = White: Yellow. In Sinhalese art the posture+ of standing and sitting are related to the circle and square; the pedestals of standing Buddhas are circular, those of seated Budd- has square."** Thus: Circle: square = standing: sitting. ‘There would be no difficulty to bring evidence for showing that: Heaven: Earth = Male: Female, whence: Standing: sitting = Male: Female, & homology familiar from our own rules of etiquette. In the Satapatha Brahmana XUUL 5.8.1, cited by Hocakr (Kingship, p. 177), wwe read: “The burial mound is square, The gods and the demons, both descended from Prajapati, were contendins for the points of the compass; the gods drove ‘out the demons, their rivals and enemies, from the quarters; they were deprived ‘of the quarters and overcome. Therefore the followers of the gods make their barrows square; but those who are followers of the demons, the Easterners and ‘others, make them circular.” Thus: rele: square = demons: gods = one part of society: another part with whom there is rivalry. The two halves of some ssvage tribes (those that have the so-called dual ‘onginization) are well-known te participate in friendly rivalry, which sometimes. degenerates to real enmity. The passage is presumably pointing toward such a 125 Encyclopaedia of Religion cmd Ethics, vol. 1, p. 37. 124 Hocant, “The Throne in Indian Art”, Ceylon J. of Science, See. G, vol. 1, mart 3 (1927). 0. 120, Ritual Origin of Cirete and Square 313 duality. Note that the two sides, the demons and the gods, are related, for they both descend from Prajapati. In Ancient Egypt we find the opposition in question: the square was asso with the earth, the circle wi : Though the information of the desired form, | may still mention the Omaha Indians, with whom a circular mark represented the sun of day. # four-pointed star, the night.!2¢ ated 19. The square as a cross ofthe Circle and the Four. We come now to the origin of the square itself. We have spoken of the round house and the square house. There appears to be an intermediate form where the roof is held up by four posts, instead of one, and the four posts are arranged in a square about the center. Above I gave an example of this with the Arikara Indians. {f, now, we follow Leriiaby’s thesis {already cited), that the “development of building practice and the ideas of world structure acted and reacted on each other,” we might suppose that the four posts arranged in a square are introduced for structural purposes (in a costnic building) and that this square eventually yields the square house and the square world But I am skeptical that the quadruplication of the pole (which itself may have been introduced solely for structural reasons) could have come about solely for structural reasons. We have scen, indeed, that the Omaha Indians, who took over their carth-lodge from the Arikara, but not the accompanying ritual and symbolism, “sct up about the fireplace six, seven, of eight posts as suited their convenience.” The four poles, then, are not introduced for structural, but rather for symbolic, purposes, and we therefore have to concern ourselves with the symbolism of the Four. ‘We might concede that once the four poles are introduced into the structure of a circular building, they would for structural (or perhaps aesthetic) reasons be placed uniformly about the center (instead of, say, in a straight line), thus yielding ‘a square. But before resting content with this explanation, we ought to study more thoroughly the Four. Where does the Four come from? this question, however, we leave the domain of the original problem. untess, indeed, the Four comes from the Square: but we have just seen some reasons to doubt this, and that it is, rather, the Square that comes from the Four. If that really is so, my line of thought so far can be summarized as follows: The Square is a cross of the Circle and Four; it is valued as a figure that can be put in dual oppasition 10 the Circle. 20. Sacred numbers. The sacredness of certain numbers, especially the Three ind the Four, is x phenomenon that has not escaped the attention of many observers, and accordingly has elicited many explanations of it, 1t may be well, then, to give here some of these explanations. 135 Nurrat, op. ci 326 “The Omaha Tribe”, p. 504f.314 A. iDeNDERG According (0 D. G, BRINTOS, the sacredness of numbers “is so widespread, so nigh universal in all times and places, that any explanation, to be valid, must ‘est on some equally universal relations either of man of of mind.” Thus the sacredness of 3 finds its origin, he believes, in the triune nature of mental activities, that is, 3 is sacred because every syllogism has three parts. The number 4 derives its sneredness from certain unisersal relations of the body to its environment. Jt might appear from this that ail peoples should value 3 and 4 cqually, but that is not so: the American Indians had 4 as their sacred number, whereas 3 was the important number in the history of the ancient Germans. Tie reason for the differ- 3s, according to BRINTON, Zn ethnic character. Materialistically minded peo- ples will favor 4. This explains wiy the Jews, with whom 4 was sacred, would not accept the Trinity,"2” F. Boas also seeks a psychotogical explanation. “The difference,” he writes, “in favorite shythms may account for the occurrence of different sacred numbers; ind since the preference for a definite number is a general psychological pheno- menon, their occurrence must not be due to historical transmission, but may be considered as based on general psychological fucts. The difference between the sacred numbers woukl then ayear as different manifestations of this mental reaction. "12 R.H. LowIE considers that the classifications of languages “ icously ... by a simple association of ideas.” “No-one,” he writes, ‘argues that the Shoshone Indians of Wyoming derived their dual number from Greek gram- ‘mar; in both instances, paired occurrence of phenomena happened to strike the Primary specch-making minds «3 significant and established the category.” By a“pimary speech-make of the number 4 he says feeling for four as compared witt. other numbers and ereated the present standard by his personal prestige.” If we ask what associations could assign a sacred character to 4, Lowie suggests that “there is every reason to look for an explana tion of the associations of primitive folk towards the same psychological principles ‘operative amongst ourselves.” = Hocant writes: “We start with a camp or city, round or square, which is a ure of the earth, and so must have four quarters like the earth."*2° { hate to disagree with Hocakr, and he surely does not belong in the camp of the inde- pendent inventionists, but I cannot accept the second part of this explanation. {isa case of Homer nodding, for surely Hocart would have agreed that the earth ‘no more has four quarters than i: has three thirds or five fifths. The city docs not Lowie means “a pre-normative individual”, Speaking “at some time someone must have had a differential 427 D.G. Bunton, The Myths of the New World, p.84; and “Origin of Sucred Numbers", Amer. Anthro., vol. 7 :1894), pp. 168-173. This opinion and the following {wo were already considered in my monograph on “The Diffusion of Counting Practices” 207K, F. Boas, “Mythology and Folktales of North Ameri Folk-Lore, vol. 27 (1914), 9. 409. 4° History of Ethuological Theory, p. 16 13° Kings and Councillors, p.254, mn Indians”, J. Amer. 5 and Primitive Religion, pp. 280, 284, Ritual Origin of Circle and Square 31s have four quarters because the earth does, but rather the earth has four quarters because the city, é.c. the ritual scene, does. The four quarters come from ritual. But how? 21. ‘The cardinal directions. One may, of course, ask whether the earth, that is, a flat surface, is not, after all, really four-fold in some sense, and that the Four we note in ancient concepts is not simply a simple reaction to san actual fact, The Euclidean plane is 2-dimensional: if, as in Analytic Geometry, we draw {wo fines in the plane perpendicular to each other, any point in the plane can be located by a numbers, namely, the distances (positive, negative, of zero) from the two lines (the so called axes). Thus the plane really is 2-fold; and, once the axes are drawn, the 4 is there, too. Many writers have followed this line of thought. The four cardinal directions are noted; and it is noted that this knowledge is very widespread. But it is not universal—far from it; and what is not universal can hardly be a natural, spon- ancous reaction to universal circumstances. Morcover, the cardinal directions ate related to the sun, and so are not just a part of plane geometry, bul appear to be a Solar phenomenon. | am not concerned with denying that the Euclidean plane is 2-fold (or 4-fold), but this doesn't mean that Analytic Geometry docs not have a history. Similarly, the cardinal points are not without a history. Here is some evidence that knowledge of the cardinal directions is not univer- sal. According to A. E. Jensen, the hill folk of West-Ceram are completely lacking in our notion of the cardinal directions (the coast folk have it, oF at least East and West). According to Krococr, the same is true of the Northwest and Central Californian Indians: their principal directions are upstream" and “down- stream”, and Yurok will refer, for example, to the door of his house as being fon the “downstream” side." Professor ScHAPERA told RAGLAN that if you wish to mention a compass direction to a South African Bantu, you have to give the name of the tribe li in that direction, According to RAGLAN, the compass points are known in Mada- gascar and parts of West Aftica but appear not to be known elsewhere in Negro Africa. As for India, “Abbot says that in spite of the importance of the cardinal points in Hinda ritual there are many tribes in India that are still unacqutinted with them”; some have ‘rising’ and ‘setting’ for east and west, but have no words for north and south. “Similarly the Incas, though they knew of the four quarters, seem not to have recognized the cardinal points. East and West were important because of the rising and setting of the sun, but their language, Quechua, lacks words for north and south. On the other hand the Finnish words for north and south seem to have appeared in the language carlicr than the words for cast and west.” RAGLAN is also under the impression that the tribes of Siberia, Austral and South America do not know the cardinal points.'2? Most of these areas are what Fronentus would have called Lunar of Fossil A.E, Jensen, ,.Wettkampf-Paricien, Zwciklassen-Systeme und geographische Orienticrung”, Stadium Generale, vol. 1 (1947), p.43. A. L. Kroenen, Handhvok of the Indians of California (Cited by Jensen, op. cit, p- AD). "32 Temple and House, p. 1670.316 A. Sepennine The hill folk of West Ceram, the so-called Alfur, do have a “rectangular coordinate system”. Ceram is an Indonesian island, the largest of the Mollucas between Celebes and New Guinea. It is divided in two by a narrow range, running roughly east and west, that reaches to about 1000 meters. The Alfur live ata height of about 400 10 700 meters, and looking out can usually sce the sca. The principal direction is fowau of lolau. which means “towards the sea”. Hence in the northern half it corresponds (roughly) to our orth, and in the southern to our south. The opposite direction, towards the highest points, is fadaja, The other principal directions are fodi and loki They do not live at the'top nor (I think) even climb ‘over it, but go around.'®? Thus they do not encounter the ambiguity at the top (like the one we would experience with our system at the north pole or south pole). The direction fowau varies, of course, locally from pluce to place, relative {0 our system, but for the natives it is fixed and they even know, for a given lo- cality, where, relative to their system, the sun rises. JENSEN tried to tell the natives hhow things “really” stood and how their system leads to contradictions, but they defend their way with gusto and very well indeed. Morcover, they argue, the Point of rising of the sun «oes not yield a good frame of reference since this point varies, even at a given locality. One is reminded of a similar question that has plagued the West since Greek times: which gocs around which, the sun around the earth or the earth arwund the sun?! Giving a direction, as she South African Bantu do, by mentioning a the specified direction secs a spontaneous enough method. So does the “up- stream-downstream” of tie Yurok. Yet Jenstn acutely observed that these de- signations are closely tied up with some religious conceptions of the Witoto, a forest people of Columbia (South America). He first noted that the teams playing a certain game were called “upstream” and “downstream”. As already remarked, the two sides of a dual o:ganization participate in (riendly rivalry, and JENSEN connects the games, which really are of a religious character, with the dual organi zation, He also noted that “downstream” means towards the “underworld”. Then, t00, the same word is etymologically related to the word “fight”. The words ae also used in actual orientation, “downstream” corresponding to “east™. He then takes note of some similar conceptions amongst the Californian In- dians and notes (following Kroenen) the “upstream-downstream” of the Yurok. With the Yurok these are actual directions and do not appear very explicitly in their mythology, though they have a hero Pulekukwerek = “downstream sharp". But with the Miwok (and others) there is a clear relation between the geographical conceptions and the social organization, the tribe being divided into exogamous classes which (with the Miwok) are called “flow up” and “flow down”; and every= thing in the world (animals, plants, cfc.) belongs to one of the two classes. JENSEN then turns to West-Ceram, where he himscif has done some field work. Here he finds the same mythical world picture as with the Witoto and with some 499 Jensen was not able to determine whether the natives climb over the mountains, bbut from some of their answrs on the location of their communities, I think itis fairly clear that they did not On this question see P. DuneM, To Sare the Phenomena. Ritual Origin of Cirele and Square Californian tribes. There is hardly a tale or other statement in which the described does not occur upstream or downstream. As already remarkec Principal direction for actual orientation is “toward the sea”, but ritual associations. Thus when someone dies he is carried “to the sea” and b there. Aflerwards, the corpse-carriers return 10 the house of the deceased, + is divided in (wo, the half “towards the sea” and the half “towards the k where they chew Sirih and Penang. The Penang nuts are split in two, onc belonging to the deceased, the other half “to the land”, etc. “Toward the | the life giving direction. Thus one sees clearly enough that the direction wards the sea” corresponds to the “downstream” of the Witoto. JENSEN turns next to other parts of Indonesia and there finds (amon things) that the Ngudjn Dajak designate the four cardinal points in ord follows: sunset—downstream—sunrise—upstream. As JENSEN himself rem this is a combination of a River-flow and a Solar orientation.'2* Jenst finds traces of the first type in Polynesia, but expresses disappointment wit spear seldom to give the desired information. Jensen also mentions the Tungus, who hive elans ealled “liv and “living down the river ‘After reading JENsen’s paper, I noticed a few more examples, or at least tr of the River-flow orientation. Thus a Pondo (South African) informant of * -RRY used the upstream-downstream terminology in answering questions direction. (The Pondo live on the Umgazi River.) He also mentioned su and sunset as directions. The Banaro, who live on the Keram River in Guinea, have been studied by R. THURNWALD: “The social unit of the settle is the hamlet. The hamlet, together with its inhabitants, derives its name ‘that of the goblin-hall [the communal structure, o¢ religious center]: the nan the goblin-hall, as well as the name of the tribe, remains constant. If the 10 of the hall should be changed on account of the ‘4 new hall should be built in the same village, the name would be retained in the example of the ... Tjimundo tribe the name of the two goblin-hall with their surrounding houses constitute the village [of two hamlets}. ways been the same—Yudrmua and Nangindumbir. Not only have the ni of the goblin-halls been retained, but also their position in relation to the in different localities the tribe has occupied, the first being always farther «le stream than the other. Yudrma, the natives say, “goes first”, and Nangindw “follows”, the whole village being imagined as floating down the river." Ancient Egypt was known as “the two lands” and the two kinds were U Egypt and Lower Egypt, so named, it would appear, in accordance with the of the Nile. In modern literature the two lands are referred to as South and Ne up the € bas conceivable that the original solar orientation had just two direct land west—indced, there are many groups who have just these two directions that this was combined with a River-low orientation to yield the familiar fourfold orientation. Thus the system Ngadju Dajak is conceivably not a focul cross. 136 W. J. Penny, The Primordial Ocean, p. 271f, R. THURNWALD, “Bano Si in Knoeuen & WATERMAN, op. cit. p. 284f, or Memoirs of the Amer. Anthro. As vol. 3 (1916).38 A. Supexurne and even the later Egyptiar did this. According to Serie, however, the signs and words for the cardinal dire.tions south and north and the names of the two lands were clearly distinguished n the pyramid texts and grave inscriptions of the Old Kingdom.'?” Still, the similarity between the sign & for south and the sign 3, for Upper Egypt (also abbreviated as %, or even 1) is at least curious. We may sum up Jexstv’s findings as follows: There are (at least) to types of orientation, the River-low and the Solar. The River-flow looks older. It is an aspect of the dual organization, and is related to the geographical positions of the two halves. It also enters into the rituals and myths of the dual society In his own summary, J1\sEN says that he suspects that the River-fow orient tion is part of a complex tut that he does not consider that he has proved this. 22, Origin of the Two in social organization. { come now to my own ideas on the source of the Two und the Four. of mathematics the circle and the square are hisduslity has religious or ritual character. Ancient thought is pervaded with dual notions and arc strikingly evident amongst peoples with the dual organization. It would appear plausible, then, to seek the origin of the Two, and pethaps also that of tie Four, in the history of the dual organization.'2* In the dual organization, the community is divided into (wo groups, often called moieties. The best krown aspect of this organization is that the members of one group marry members of the other; but itis also true that they initiate and bury each other, and in geaetal play complementary roles in ritual. In the dual system everything in the world, or nearly everything it sometimes seems, is assigned to one of ‘he two moieties: thus, for example, land to one side, water to the other. And eacs side of a polar, or otherwise 2-fold, concept is like- wise assigned: for examp'e, sky-carth, east-west, right-left, big-little, noble- common, gentle-rough, pe:ce-war, light-dark, summer-winter, male-female, etc. With this general sehen:: in mind, let us look at some of the evidence. Pre- Viously I mentioned the Omi:ha, who when gathered for ceremonial or for traveling, ‘circle, with the Sky People on one side and the Earth People on the to multiply such evidence, and I will, but we can already see here 1 basic point: the circle is bisected by a diameter, and the two halves correspond to the two halves of « dual organization. If we could get a diameter perpendicular to this one, we'd have the quadrants and would be well on the way to the square. ‘As for some further evidence, “our authorities are not quite agreed as to the shape of an Aranda camp (the Aranda, or Arunta, are an Australian tribe]: Spencer makes it round: Strchlow makes the married men’s camp oblong with long axis east and west: east of it lies the camp of mateless men, and west of it that of mateless women.” We would like the camp to be round, but we're taking 137 K, Setar, “Die Namer: von Ober- und Unter-Agypten und die Bezeichnungen fir Nord und Sid”, Zeitschrift fir Aegyptische Sprache und Altertumskunde, vol. 44 (1907), pp. 9. 19" On the dual organization see A. M. Hocanr, Kings and Councillors, and W.J. Penny, The Children of the Sia. Ritual Origin of Cirete and Square the Circle for granted now, and won't be disturbed over this. “Both agree the camp is divided in two: cast and west according to Strehlow, north-west and s: cast according to Spencer. Residence is not a matter of choice but of des The two halves do not however form different classes of society, for the ¥ 0 is one big family, and all its members are related. An individual fin- his own side his paternal relations, his mother’s relations on the other. Each is again divided into two, making altogether four divisions which Stre places N-E, S-E, S-W, N-W, but Spencer, N, E, S, W. Membership of 1 subdivisions is also decided by descent. The camp of the Loritas, westerly neighbors of the Arandas, compl reverses the positions, turning the whole camp, married and mateless, ea: west. Ata great annual festival the fire is lit at each of the four cardinal po ‘A Winnebagoe [American Indian] village is circular. According to one acc: draws a plan in which this two-fold division is obscured . The Ponka [also of America] divide their camping circle into two, East West; but both moictics are again divided into two, making four quarters ‘among the Aranda, These quarters are necessarily orientated N-E, S-E, S-W. N and are called respectively ” Ina myth of the Apinayé, of South America, “Sun and Moon cach engem 4 group of humans, ancestors of the Apinayé, as it happens, at (wo different po along the course of a river. The two types of people are then settled in a com: circular village which—obeying the original dichotomy is divided into hal the solar group in the north, the lunar in the south. To this day the Apin ate so divided and settle in separate halves of the village, with a common who is always taken from the Solar faction. Speaking of the Bénaro, THURNWALD (loc. cit., p. 285) says: “As one en a goblin-hall, he is immediately impressed by its symmetrical plan. This is especi noticeable in the arrangement of the fireplaces, of which there are four, two cach side of the hall, directly opposite each other ... . ‘The two symmetrical rows of fireplaces in the goblin-hall correspond to ad sion of the hamlet} into two halves. It might perhaps be allowed use the word "sib" in a narrower sense to indicate these halves. The sibs themsel their place in the goblin-hall. These terms refer to their relative pos faces the entry . ... The external form of the settlement fof several hamlets] reflects precisely internal organization of the tribe; for the goblin-hall with adjacent houses same clearing, mirrors the social unit, the gens, just as the symmetric part Of the goblin-hall into two parts, the division of the gens into halves {sibs}. The symmetey in the arrangement of the goblin-hall is the expression in spa terms of the principle of social reciprocity or the “retaliation of like for like’ 13° Kings and Councillors, p.24Sf, with reference. 140 ayth and Calt, p. 150.320 A. Supenpena | think we now have sufficien: evidence to support the assertion that the Circle is bisected because the (wo sides of the Dual Organization take up opposite sides it, One could add to the testimony, I'm sure, nor do 1 wish to minimize the importance and interest of such documentation; and it would be convenient to have the material bearing on the issue assembled in good order. But if the object is to demonstrate, and not merciy accumulate, I think we have enough. | have also given some thouzht to the origin of the dual organization itself (in my paper on “ nization and the Kingship").'*" My view is that the dual organization derives from the orientation of the (undivided) group to its sexual neophytes, male ani female. It will be helpful for my general thesis {to remark that I also consider (Juc. cit.) the dual organization to be an offspring of the Creation ritual. However. | have no good image, even one satisfactory to myself, of how the splitting ‘ook place, 23. Genesis of the square from the circle. Now that we have the Two, it is exsy to make a conjecture on th: Four: the Four results from a splitting of the Two, that is, the two sides of a dual society themselves divide to yield a four-fold society. The Circle is corresponuingly divided to yield four quadrants and two diameters perpendicular to cach other. In rites the four sections are represented by four men (or women, or men znd women), oF poles, or fires, who place them- selves, o are placed, in positions corresponding to the positions of their sec- tions, This yields the Square (and the Cross). One can bring in some evidence to support this conjecture. In fact, some of it was already mentioned in conection with the Two. Thus it was already men- tioned that the {wo sides of an Aranda camp are themselves divided into two yielding altogether four (social) uivisions. With the Ponka, too, the moieties are again divided in two, making four quarters. U have also mentioned what THURNWALD calls a gens and its division into two sibs. The sibs, however, are not intermarrying; on the contrary, a member ofa sib must look for « mate in ine corresponding sib of another gens. Thus the ib is not what I have called above a moiety; rather, the gens itself is the mocity. Tuurnwatn says (loc cit. p. 2847) that “each of the [Biinaro] villages is composed of from three to six hamlets”, so that the tribe would have three to six gentes, though a few lines later he mentions the Tjimundo tribe which, according to his description, has just two gentes. Thus the gens looks like a half of a dual organiza- tion, with the other half variable. though in the case of the Tjimundo tribe, we appear to have the typical division of a dual society into two moieties, which are themselves divided into two. The zoblin-halls, however, look as though they were conee tribal structures."*2 "4" Folklore, vol. 74 (1963), pp. 334-340. 142 W. J. Penny (Cildren of the Sun, p. 324) remarks that “in some eases the clan itself can divide into two parts, as among the Nair of Malabar, the Todas [of South India}; in Micronesia and in British New Guinea; and among the Creek Indians of North America.” In Fiji there ate successive dichotomics from the village down to and including the clan, See Hocant, Men. vol. 33 (1933), p. 165. Ritual Origin of Circle and Square 321 The Kariera tribe of Western Australia is divided into four parts, called classes. The names of these are Banaka, Burung, Palyeri, and Karimera, The Banaka and Burung are intermarrying, as are Karimera and Palyeri, ie, for eaample, a Banaka man must find a wife in Burung. (And there are other restr tions.) The child of a Banaka man is a Palyeri and of a Palyeri man is a Banaka: and similarly with the Burung and Karimera, The tribe is divided locally into two, one half conisting of Banaka and Patyeri, the other half of Burung and Karimera. The division of each of these two halves, however, is only social, not spatial. On the other hand, with the Aranda (who have cight classes) the halves of the moieties do correspond to a spatial division: at least the sketch of an Aranda given by C. Srxemtow indicates this."4? Myths also often furnish evidence. A simple myth of the Creeks, of the lower Mississippi, is typical: they tell of four men who came from the four corners of the earth bringing sacred fire from the cardinal points and pointing out the seven sacred plants, efe. There is a great variety of such myths. One kind is the creation myth, in which the tribe traces itself back to four ancestors. The Algonquins and Dakotas have such a myth, The Crecks were, according to a legend, at first di- vided into four classes and are descended from four female ancestors. The ancient inhabitants of Haiti traced their lineage to four brothers who coming into the world at one birth had cost their mother her life. The Mayan tribes refer often to four ancestors. The Tupi of Brazil claim descent from four brothers. The nearby Guarani of Paraguay also speak of four brothers and name two of them Turt and Guaxant, respectively parents of the tribes called after them. The four-fold division of the Muyscas of Bagota was traced back to four chieftains created by their hero god Nemqueteba. Etc,"** If-we assume that a myth is the counterpart of a rite, then these myths refer to four persons appearing on a ritual scene and representing the sections of a fourfold group, and so, of course, point to the existence of fourfold organizations. Thus these myths tell us nothing new. From my point of view, however, they were quite important as | knew of taemt long before I had the examples of fourfold ‘organization given above. ‘The sameness in sex of the four ancestors should not be upsetting, as the ‘myths are describing a ritual ereation, not a real one. | hope, too, that no one will ell me there can't be anything to my theory. because the Mayas (say) have a fourfold creation myth but no fourfold social otganization. As I remarked before, the myth and the rite can separate and travel independently, There remains the question of why the Two splits, but 1 think w the origin of the quadrants and of tae square (sce Fig. 6): the quadrants come from 4 splitting of the two halves of a dual society, each into two; the Square comes #42. R. Brown, “Si zation of the Kariera of Australia” (excerpted from “Three Tribes of Western Austraia”, J. Royal Anthropological Inst, vol. 43 (1913)) in Knounen & WareRman, op. cit. p-2101T. C. Stuentow, Die Aranda und Loritja- Stamme in Zentral Australien, W, 2,3. "4 Myths of the New World, pp.94-101,322 A. SHDENMERG {rom four representatives placiny themselves (or being placed) on the Ft in positions corresponding to the four sections. = a _ € IPE Xe? IL Fig. 6. Steps ‘om the circle to the square 24, Origin of the Three. Although the problem set, the origin of the Square, is now solved, there remains the question of the origin of the Four. From the Point of view of method it seems right to pursue this somewhat, since it is concei tbe that the Four and the Two have an origin that we have completely missed. As to the Two, this seems to me to come from the Dual Organization, and not vice-versa, the Dual Organization from the Two. Even this is not certain, but 1 cannot make a more plausible suggestion than the one already made. My view ‘on the origin of the Four is in sin even less satisfactory state, so the following should be considered merely as remarks, tending, | hope, in the right direction, Curiously, the first clue came ot from the Four, but from what could be called the Three: in the Omaha tribe, “no tribal ceremony, negotiation, or consultation without both divisions (Sky and Earth] being represented; no l could set unless there were present one chief from the Inshtancunda [Sky] division and two from the Hongashenu.” An old Omaha man explaine the Sky division represents the great power, so that one chief from that side enough, while (wo are necessary for the earth division.'43 We can write this information: as follo Sky: Earth = superior: inferior = one: two. In my counting paper, 1 examined the phenomenon: Male: Female = one: two a length, and the Omaha inform..tion appears to fit in with this order of ideas. “The Winnebago are, as we saw, divided into two; but the lower half is ag divided into two, so that they can at one time say they are divided into two, another into three, without contradicting themselves." The Earth side of the Omaha is also divided into two, With the Osage the War side is divided, the Peace side undivided. The Aranda have ceremonies with two chief performers, but also a few i Which there are three, one old am: two young. In one case, the triad consists of '43 “The Omaha Tribe”, p. 197, "© ings andl Councilors. pp. Ritual Origin of Circle and Square 333 two brothers and their grandfather, Thus one: two = oll: young, a by now familiar looking equation."4? “If an Anglo-Saxon took {a} life his paternal kinsmen paid two-thirds of the blood money, his maternal kinsmen one-third.” Thus: Father: Mother = two: one. One might, perhaps, have expected the paternal kinsmen to pay one-third, the ‘maternal, two-thirds, but still the Anglo-Saxon and Omaha customs appear to be parts of the same complex. This information on the three has seemed especially sig i 1 could sce a reason, sant because from ind not a facile one, for the moiety to spl 2S. Conjecture on the origin of the Four. As to the Four, my conjecture is that each of the two moieties splits into two. Another possibility is that only one mocity splits and that one half of this mocity splits again. One might consider that with a passage of time the tribe grows: a part of one moiety would split off and combine with a part of another to form a new group. One can scarcely doubt that this happens, but the result would be two groups with the old organization, not one group with a new. The fourfold organization results rather, presumably, from some new principle of organization ‘But what that principle is, or was, is not easy to say. In the case of the Kanieri Uribe we appear to be getting a subdivision of each moiety according to the polarity Father: Child. ‘The classical caste system of India was fourfold: thus the writings that follow the Vedic period are “agreed that the royal caste was created for justice. for the protection of the people, and so for war and executive power: the priests for ritual and study; the farmers for cattle breeding, trade, and cultivation; the serfs for ‘rafts and service. Of these, the first three are “noble” and the first two “form an aristocracy within the aristocracy.” The fourth caste, the serfs, “is excluded {from the sacrificial rites] except for certain rites.” The caste system of Persia as described in the Zend-Avesta is similar. There are four castes: warrior, priest, worker ("qui donne la fecondité"), and only of the First Uhrces and these three are sacri In South India there isa two caste system, Lefthand and Righthand. According to Hocakr, “the twofold is an earlier one that has been overlaid and sometimes superseded by the fourfold.""5° As Hocat analyzes the classical system, the first three castes stand over against the fourth: the ial castes and the serfs (or artisans) correspond "7 Ibid, p. 175, tn the myths, too, “the 4" Progress of Mun, p. 234, with reference. 17 Hocanr, Caste, pp. 17, 5, 26, 69F. "3° Ibid, p67. oFs are often one old nase and Iwo 9034 {0 the two sides of a dust organization; then one side splits into two, the sacerdotal and purveyors; and the sucerdotal into king and priest. Thus the caste system is fourfold, but the picture we get as drawn by HOCART {s hardly that of two sides each splitting into two. Moreover, the sacrificial organ- ization really is threefold and the priestly writings are adapted to a theology of Three, For example, the Satsatha Brahmana, Il. i, 4, 11, reads: “The ereator ‘of Geometry", Archive for History of Exact Sciences, story of Exact Sciences, . Folklore, vol. 74 (1963), pp. 334- Circle", Archive for History of Exact Sciences, 1 the Area of vol. 9 (1972), pp. 171 211. A. SupuNDERG, “Did Euclid’s Elesnemts, Book 1, develop geometry axiomatically?" “Archive for History of Exact Sciences, vol. 14 (1975), pp. 263-295. Ritual Origin of Circle and Square 327 K, Semin, ,.Die Namen von Ober- und Unteragypten und die Uezcichnungen fir Nord ‘und Sad", Zeit. f. Agyptische Spr. und Alt, vol, 44 (1907), pp. 1-29. ©. Severenr, A Dictionary of Classical Antiquities, 3rd ed., 1894. (Meridi New York, 1957). F.G, Sruck, Naskapi. The Savage Hunters of the Labrador Peninsula. Norman, 1935. F.G. Speck, “The Celestial Bear Comes Down To Earth”, Reading Public Museum and Art Gallery Scientific Publication 7, 1945. B, Srencen & F. V. Gutten, The Northern Tribes of Ceatral Australia, London, 1904. LL. Setex, The Sun Dance of the Plains Indians. (Anthropological Papers, American ‘Muscum of Natura! History, vol. 16, part 7, New York, 1921.) C. Stautow, Die Aranda- und Loritja-Stimme in Zentral Australien. Frankfurt, 907~ 20. R, Twuanwato, Binaro Society, in Memoirs of the Amer. Anthro. Assoc. vol. 31916). B.C, VaILCANT, The Aztecs of Mexico, Harmondsworth, 1950. 2. Watton, African Village. Pretoria, 1956. W. F. Warns, Paradise Found. New York, 1885. T.T. WatERMAN, “The Architecture of the American Indians, Amer. Anthro... 1. vol. 29 (1927), pp. 210-230, C. Wisstin (ed.), Societies of the Plains Indians. Anthropologi Museum of Natural History, vol. II, New York. 1916. Books, I Papers, The Ameri Department of Mathematics University of California, Berkeley 94720 (Received February 15, 1981)