# The Atlas of Human Possibility Anthropologists described 1,291 societies and coded every record under the same headings. This atlas sorts them by 54 everyday questions: who gives wealth to whom at a marriage, who inherits the land, where a baby sleeps, what is done with the bones. It is not a census of humanity and it does not describe cultures today: every dot is an observation made at a particular place and time, almost all between 1800 and 1965, and neighbouring dots often share a history. What the atlas counts is something else: how many different answers have been recorded to the same problems of living. *Marius Comper · 11 September 2026* ## When two people marry, who gives wealth to whom? In 657 societies out of 1,272, the groom or his kin made a substantial payment to the bride's family. Dowry, property given by the bride's relatives to her, the groom or his kin, was recorded in 43; nothing of consequence, in 275; the groom's labour for the bride's family, in 125; a token payment, in 68; gifts exchanged between the families, in 65; a sister or other female relative of the groom given in exchange for the bride, in 39. ## What the atlas is made of In 1967 the American anthropologist George Peter Murdock published the Ethnographic Atlas: one record for each of 1,291 societies, compiled from the ethnographic descriptions in print at the time. Every record has a place on the map, a focal year and dozens of headings coded the same way for everyone: the shape of the house, the form of marriage, who inherits, how many levels of authority sit above the village. The focal year says when the society was observed. For 1,244 of the 1,279 dated records it falls between 1800 and 1965; the oldest, the year 100, belongs to the ancient Romans. Murdock's headings are codes, not sentences. 'Where does the couple live' has 12 codes, with names like 'avunculocal' or 'ambi-virilocal'. Here they are gathered into seven answers anyone can read, and the original code, with its definition, stays one click away, in the record of every question and every society. The data come from D-PLACE, the Max Planck Institute database that has, since 2016, tied every code to its bibliographic sources and to the language spoken. Some questions, marked 'closer sample', come from the Standard Cross-Cultural Sample: 186 societies chosen by Murdock and Douglas White in 1969, one from each cultural region, so that the same tradition would not be counted ten times. The sample is not a more accurate version of the atlas; it is a selection made so that related societies do not appear several times over. On these, generations of researchers coded thousands of further headings: where the baby sleeps, how close the father is, why people fall ill, what is done with the bones. When a question comes from there, the other 1,105 dots fade into the background. Among the records are the 'New Englanders', with a focal year of 1920: the new couple away from both spouses' kin, kin counted on both sides, one word for every cousin, one spouse, no payment of consequence at marriage, land divided equally among the children. The record opens with a click on its dot, or from the search box. ## Rare possibilities Ten answers recorded only a few times, with the societies in which they were recorded. - **A woman with several husbands** (4 of 1,257). How many spouses can a person have? Central Tibetans, 1920; Marquesans, 1900; Sherpa, 1950; Toda, 1900. - **Spouses each stay in the household they were born into** (8 of 1,267). After the wedding, where does the couple live? Aua, 1900; Fur, 1880; Ga, 1930; Kerala, 1799; Kutubu, 1940; Marind-Amin, 1910; Minangkabau, 1920; Siane, 1940. - **Infection as the main explanation of illness** (1 of 122). Why do people fall ill? Infection Japanese, 1950. - **No belief in ghosts** (1 of 59). Are there ghosts? Maasi, 1900. - **Polygonal houses** (2 of 1,197). What shape is the house? Koryak, 1900; Sivokakmeit, 1920. - **Only games of chance** (2 of 486). What kinds of games are played? Hutsuls, 1890; Li, 1930. - **The bride has more say than the groom in the choice of marriage** (3 of 80). Who decides whom you marry? Garo, 1900; Kwoma, 1930; Wodaabe Fulani, 1950. - **In principle, only men can divorce** (5 of 93). Who can ask for a divorce? Abipón, 1800; Barabra, 1920; Callinago, 1650; Manchu, 1920; Sirionó, 1940. - **A strong belief in the reincarnation of adults** (9 of 53). Do dead adults return in a new life? Balinese, 1950; Burmese, 1950; Comanche, 1870; Eyak, 1890; Garo, 1900; Huron, 1640; Igbo, 1930; Trobriands, 1910; Twana, 1850. - **The youngest heir receives most of the land** (22 of 820). Is the inheritance split equally, or does one take it all? ## Who lived most differently from what feels normal to you? Ten questions, one from each domain. Pick at each the answer that feels normal; the answers describe a society's arrangements, not personal choices. Only the questions you answered and for which the society has a record are compared. A society enters the ranking if it has a record for at least seven of the questions you answered, or for all of them if you answered fewer than seven. Societies are ordered by the share of differing answers among those compared, in decreasing order, then by the number of compared questions, then alphabetically. A missing record never counts as a difference. A computed example: with the answers of the New Englanders' record of 1920 to eight of the ten questions (the Romanians' record of 1910 has a code for only two), 45 societies of the 950 comparable ones differ on every compared question, 41 of them in Africa. The 18 societies with a record for all eight questions: Azande, Banda, Bassari, Coniagui, Gbaya, Ibibio, Isoko, Kpelle, Luapula, Mbuti, Mossi (Yatenga), Ngere, Popoi, Songye, Temne, Tiv, Toma, Walbiri. ## What recurs and what surprises The atlas does not say that anything goes. Some answers dominate: in 904 of 1,267 societies the couple as a rule lived with or near the husband's parents or his father's line, and together with those in which it lived with his mother's-line kin, 969; in 436 of 575 the houses were built by men; in 525 of 1,155 there was no authority above the village. People had the same problems to solve, with the same bodies and few materials, and many solutions were repeated independently. Beside every dominant answer, though, stand documented alternatives. In 54 societies the couple lived with the husband's maternal uncle or other kin of his mother's line. In 338, cousins were called by the same word as siblings. Dowry, the form in European history, appears in 43 societies, and the payment in the opposite direction, from the groom to the bride's family, in 657. The round house was recorded in 513 societies, almost as many as the rectangular one, in 582. In a single society out of 122, infection was the main explanation of illness. Distance can also be measured. The game below uses ten questions, one from each domain, and orders the societies by the share of differing answers among those compared. With the answers of the New Englanders' record of 1920, which has a code for eight of the ten, 45 societies of the 950 comparable ones differ on every compared question: 41 in Africa and four in Australia and the Pacific, among them the Azande, the Mbuti, the Kpelle and the Walbiri. At the other pole, four records differ on none: the Neapolitans, the Spaniards, the Lebanese and the New Englanders themselves. The dots are not independent of one another. 414 of them are in Africa and 284 in North America, because that is where most ethnography was written; 306 societies speak languages of the Atlantic-Congo family, 131 of the Austronesian family. An answer that looks universal can turn out, on the map, to be the custom of a single language family spread across a continent. The 'by region' and 'by language family' buttons colour the dots for exactly that reason, and in the grouped rows related dots sit side by side, as blocks: 306 dots with the same answer are not necessarily 306 separate inventions. ## What the atlas cannot say Every dot is a dated observation, not a description of the present. The Romanians of 1910 are not the Romanians of today, and the Hawaiians of 1800 are not those of 1965. That is why everything on this page is in the past tense. A missing record does not mean 'no'. For games, 805 records have no code at all; for land inheritance, 435. The counts use only records with a code, and the rest are shown separately, as empty rings. Codes lose information. Murdock himself wrote that the inheritance headings had 'proved inadequate' for societies with communal property. The grouping here loses a little more; that is why the original code sits beside every answer. The dots are societies, not countries, and a present-day state may contain dozens of dots or none. The names are those in the database, sometimes names given from outside and rejected today by the people so named; alternative names appear in the record. Europe is thinly covered, with 42 societies, because ethnography was mostly written about others. ## Every question, in figures The same counts as tables: for each question, the documented answers, the number of societies and the share of those with a record. ### Family and kinship **After the wedding, where does the couple live?** (EA012, N = 1,267) - With or near the husband's parents or his father's line: 904 (71%) - With or near the wife's parents or her mother's line: 140 (11%) - With either family, as circumstances allow: 83 (7%) - Apart from the relatives of both, or at a place not determined by kin ties: 62 (5%) - With the husband's maternal uncle or other kin of his mother's line: 54 (4%) - With one of the families or with the husband's maternal uncle, as an option: 16 (1%) - No shared household: each stays in the household they were born into: 8 (1%) - No record: 24 Neolocal residence means that the couple lives apart from the relatives of both spouses, or at a place not determined by kin ties; the code says nothing about who owns the house. It is the arrangement a reader in Europe today takes for granted, and in the atlas it is rare. In 904 societies out of 1,267 the couple as a rule lived with or near the husband's parents or his father's line. In another 54 it lived with the husband's maternal uncle or other kin of his mother's line, and in another 11 that was an alternative to his father's side; together these categories cover 969 societies. There is also the mirror image, towards the wife's kin, and one in which the two each stay in the household they were born into. **Which side counts as your line of descent?** (EA043, N = 1,274) - The father's line: 590 (46%) - Both sides, no fixed line: 362 (28%) - The mother's line: 160 (13%) - A chosen or mixed line: 110 (9%) - Two lines at once, each for different things: 52 (4%) - No record: 17 A European reader counts relatives on both sides, with no preference for the father's or the mother's line. That bilateral reckoning is the second most common in the atlas. The most widespread is the father's line, and the mother's line, in which children belong to the mother's kin and inherit from her brother, appears in 160 societies. **What do you call your cousins?** (EA027, N = 955) - The same word as for siblings: 338 (35%) - The children of two brothers or of two sisters are called 'siblings', the other cousins are not: 256 (27%) - Some cousins get the word for parents or children: 137 (14%) - One word for all cousins, another for siblings: 110 (12%) - Each cousin described through the chain of relatives: 81 (8%) - A mixed pattern: 18 (2%) - A different word for each kind of cousin: 15 (2%) - No record: 336 In 338 societies, cousins were called by the same word as siblings. The Romanian and English system, one word for every cousin and another for siblings, was named 'Eskimo' by anthropologists because the Inuit share it. Other systems split cousins by the side they come from, or place them in the parents' generation. **Who lives under the same roof?** (EA008, N = 1,263) - An elder's family and those of several of their children: 321 (25%) - A man, several wives and their children: 280 (22%) - Parents and children, occasionally two wives: 271 (21%) - A large household of several families: 234 (19%) - Parents and children, one wife: 111 (9%) - Two related families, usually parents and one married child: 43 (3%) - A woman with several husbands: 3 (under 1%) - No record: 28 The household of parents and children, alone in its house and monogamous, is the form that everyday language assumes. The atlas records it in 111 societies. Households spanning two or three generations, and households with several wives, appear more often. ### Love and marriage **How many spouses can a person have?** (EA009, N = 1,257) - Several wives, as the rule: 589 (47%) - Several wives, but rarely: 463 (37%) - One wife or one husband: 201 (16%) - Several husbands: 4 (under 1%) - No record: 34 Monogamy as the only permitted form appears in 201 of the 1,257 societies with a record. In most, a man could have several wives, either as an exception or as the rule. Polyandry, one woman with several husbands, was recorded in four: the Central Tibetans, the Sherpa, the Toda of south India and the Marquesans. **When two people marry, who gives wealth to whom?** (EA006, N = 1,272) - A substantial payment from the groom to the bride's family: 657 (52%) - Nothing of consequence: 275 (22%) - The groom's labour for the bride's family: 125 (10%) - A token payment: 68 (5%) - Gifts exchanged between the two families: 65 (5%) - Dowry: property given by the bride's relatives to her, the groom or his kin: 43 (3%) - A sister or other female relative of the groom, given in exchange for the bride: 39 (3%) - No record: 19 Dowry, property given by the bride's relatives to her, to the groom or to his kin, is the form known from European history. The atlas records it in 43 societies. Fifteen times as often, the payment went the other way: the groom or his kin gave livestock, goods or money to the bride's family. **Can you marry a cousin?** (EA023, N = 1,042) - Neither first nor second cousins: 282 (27%) - First cousins are forbidden: 277 (27%) - Only cross-cousins, the children of a brother and a sister: 255 (24%) - Any first cousin: 117 (11%) - Only second cousins: 64 (6%) - Other rules on first cousins: 34 (3%) - First cousins and some second cousins are forbidden: 13 (1%) - No record: 249 The total ban, neither first nor second cousins, is the most frequent record. In 255 societies, though, marriage with a particular kind of cousin was allowed and often usual: cross-cousins, the children of a brother and a sister, are the most commonly permitted. **What was expected of girls before marriage?** (EA078, N = 598) - Premarital sex allowed, no sanctions: 208 (35%) - Virginity required, premarital sex punished: 153 (26%) - Premarital sex forbidden, but unpunished: 121 (20%) - Premarital sex allowed, punished only if pregnancy follows: 58 (10%) - Moot: girls married very young: 47 (8%) - Trial marriage: 11 (2%) - No record: 693 In 153 societies virginity was required and premarital sex was punished. In 208, premarital relations were free and carried no sanction. In between sit bans without consequences, permission withdrawn only if pregnancy follows, and trial marriage. **Is adultery judged the same for husband and wife?** (SCCS169, N = 109) - Allowed only for the husband: 48 (44%) - Forbidden to both, the wife punished more: 24 (22%) - Forbidden to both, equally: 24 (22%) - Allowed to both: 13 (12%) - No record: 77 In the closer sample, a double standard in the husband's favour is the most frequent record. A single standard appears in 37 societies, either because both spouses were condemned alike, or because both were allowed. **Who decides whom you marry?** (SCCS604, N = 80) - The groom and the bride, equally: 46 (58%) - The groom has more say: 27 (34%) - Only the groom can propose or refuse: 4 (5%) - The bride has more say: 3 (4%) - No record: 106 The variable measures the power of the groom and the bride to propose or refuse a match, whoever arranges the marriage. In 46 of 80 societies, the two had an equal say. **Who can ask for a divorce?** (SCCS610, N = 93) - Either spouse, equally easily: 72 (77%) - Both, but harder for the wife: 12 (13%) - Only the husband: 5 (5%) - Both, but harder for the husband: 4 (4%) - No record: 93 In 72 of 93 societies with a record, divorce was equally available to husband and wife. The coding concerns the possibility, not the frequency. ### Childhood **Where does a baby sleep?** (SCCS23, N = 173) - In the mother's room, in another bed or unspecified: 97 (56%) - In the mother's bed: 53 (31%) - In both parents' bed: 23 (13%) - No record: 13 A separate room for the baby appears in none of the 173 societies with a record: every code puts the child in the mother's room. In 76, the baby slept in her bed, and in 23 of those the father shared it too. **How close is the father to a baby?** (SCCS53, N = 154) - Occasionally close: 72 (47%) - Frequently or constantly close: 47 (31%) - Rarely close: 27 (18%) - Distant: 8 (5%) - No record: 32 The code tracks physical closeness and care, not affection. 'Regular' closeness was recorded in three societies out of 154. **At what age does breastfeeding end?** (SCCS45, N = 156) - 25 to 36 months: 72 (46%) - 13 to 24 months: 43 (28%) - 37 to 48 months: 30 (19%) - Up to 12 months: 7 (4%) - 49 to 60 months: 4 (3%) - No record: 30 The most frequently recorded age at the end of breastfeeding is between two and a half and three years. Weaning by one year appears in seven societies out of 156. **How is a baby carried?** (SCCS28, N = 156) - In a sling or pouch: 75 (48%) - In the arms, with nothing: 44 (28%) - Strapped to a cradleboard: 28 (18%) - In a basket: 9 (6%) - No record: 30 The cloth sling, the oldest carrying system, is also the most frequently recorded. The rigid cradleboard, a board to which the child is strapped, was recorded in 28 societies, 19 of them in North America. **How quickly do adults answer a crying baby?** (SCCS31, N = 103) - Quickly, every time or nearly: 78 (76%) - Quickly, but not consistently: 19 (18%) - Slowly or with indifference: 6 (6%) - No record: 83 A prompt, nurturing response is the recorded norm: 78 of 103 societies. A slow or indifferent response appears in six. **Was a boy or a girl preferred?** (SCCS616, N = 93) - No stated preference: 54 (58%) - Boys: 28 (30%) - Girls: 11 (12%) - No record: 93 A stated preference for boys appears in 28 societies out of 93; for girls, in 11. No stated preference is the most frequent record. ### Home and settlement **What shape is the house?** (EA079, N = 1,197) - Rectangular or square: 582 (49%) - Round: 513 (43%) - Around an inner courtyard: 51 (4%) - Oval or elongated: 49 (4%) - Polygonal: 2 (under 1%) - No record: 94 The round house was recorded almost as often as the rectangular one: 513 against 582. The courtyard house, the Mediterranean and Chinese pattern, appears in 51 societies. **What shape is the roof?** (EA082, N = 1,135) - Gabled, two slopes: 377 (33%) - Conical: 353 (31%) - Dome, beehive or vault: 207 (18%) - Flat: 95 (8%) - Hipped, four slopes: 88 (8%) - A single slope: 15 (1%) - No record: 156 The gabled roof, the one in European children's drawings, and the conical roof were recorded almost equally often. Roofs shaped as a dome, a beehive or a vault were recorded in 207 societies. **Where is the floor relative to the ground?** (EA080, N = 1,179) - At ground level: 941 (80%) - Dug into the ground: 86 (7%) - On stilts or posts: 79 (7%) - Slightly raised: 73 (6%) - No record: 112 The house dug into the ground and the house on stilts are two answers to the same problem, cold and damp. Together they were recorded in 165 societies. **How are the dwellings arranged?** (EA030, N = 1,187) - Compact, permanent villages or towns: 518 (44%) - Bands on the move half the year, with a fixed settlement otherwise: 188 (16%) - Dispersed homesteads: 151 (13%) - Hamlets linked into one community: 106 (9%) - Moving between several fixed settlements: 97 (8%) - Fully nomadic bands: 81 (7%) - A village with homesteads and hamlets around it: 31 (3%) - Villages moved every few years: 15 (1%) - No record: 104 The compact, permanent village is the most frequent record, but in 381 societies people moved: entirely, with the seasons, or every few years. **How many people does a community have?** (EA031, N = 611) - 50 to 99: 119 (19%) - Under 50: 118 (19%) - 100 to 199: 106 (17%) - 200 to 399: 86 (14%) - A city of over 50,000: 66 (11%) - 400 to 1,000: 61 (10%) - A town of over 5,000: 38 (6%) - 1,000 to 5,000: 17 (3%) - No record: 680 Fewer than half the societies have an estimate. Among those, in 237 the usual community had under a hundred people, and in 66 there was at least one city of over 50,000. ### Food and work **Where does most of the food come from?** (EA042, N = 1,290) - Extensive agriculture, with land left to recover: 477 (37%) - Intensive agriculture: 282 (22%) - Fishing: 121 (9%) - Gathering: 103 (8%) - Agriculture of unknown type, or casual: 86 (7%) - Hunting: 77 (6%) - Herding: 77 (6%) - Two or more sources, in equal parts: 67 (5%) - No record: 1 Extensive agriculture, with land cleared, farmed for a few years and then left to recover, is the most frequent record. Gathering, hunting and fishing together fed 301 societies. **Are the animals milked?** (EA041, N = 1,182) - No, or hardly ever: 806 (68%) - Yes, more than sporadically: 376 (32%) - No record: 109 Milk as adult food is ordinary in Europe and rare across most of the world. In 806 of the 1,182 societies with a record, animals were not milked at all, or hardly ever. **Do animals pull the plough?** (EA039, N = 1,182) - No: 1,006 (85%) - Yes, from before European contact: 154 (13%) - Yes, adopted after contact: 22 (2%) - No record: 109 The animal-drawn plough is one of the inventions that separated the agrarian history of Eurasia from that of the Americas, sub-Saharan Africa and Oceania. In the atlas, it was in use before European contact in 154 societies. **Who works the fields?** (EA054, N = 976) - Mostly or only women: 260 (27%) - Mostly or only men: 242 (25%) - No agriculture: 241 (25%) - Women and men, in equal parts: 233 (24%) - No record: 315 Where agriculture existed, women did most of the work in 260 societies and men in 242. The difference follows the tool: the hoe stayed more often in women's hands, the plough almost only in men's. **Who makes the pots?** (EA047, N = 734) - Mostly or only women: 357 (49%) - No pottery: 310 (42%) - Mostly or only men: 46 (6%) - Women and men: 13 (2%) - Industrial production, sex irrelevant: 8 (1%) - No record: 557 Pottery is the craft with the sharpest sex division in the atlas: women only in 345 societies, men only in 42. Industrial production, where sex is irrelevant, appears in eight. **Who weaves?** (EA045, N = 785) - No weaving: 447 (57%) - Mostly or only women: 196 (25%) - Mostly or only men: 102 (13%) - Women and men: 27 (3%) - Industrial production, sex irrelevant: 13 (2%) - No record: 506 Loom weaving is women's craft in 196 societies and men's in 102. More than half the societies with a record did not weave at all: they wore skins, beaten bark or bought cloth. **Who builds the houses?** (EA049, N = 575) - Mostly or only men: 436 (76%) - Mostly or only women: 78 (14%) - Women and men: 58 (10%) - No houses are built: 3 (1%) - No record: 716 House building is men's work in 436 societies out of 575. Women did most or all of the building in 78, every one of them a nomadic or semi-nomadic society, where the house was a shelter carried from place to place. **How are things carried over land?** (SCCS13, N = 185) - On people's backs: 108 (58%) - On pack animals: 41 (22%) - On animal-drawn sleds or travois: 13 (7%) - In animal-drawn carts: 12 (6%) - In motor vehicles: 11 (6%) - No record: 1 In 108 of 185 societies, people carried everything on their backs, often with a strap across the forehead. The animal-drawn wheel appears in 12. **Is there money?** (SCCS17, N = 183) - No, no medium of exchange: 77 (42%) - Foreign currency: 42 (23%) - Tokens of conventional value: 26 (14%) - Their own currency: 26 (14%) - Useful goods serve as money: 12 (7%) - No record: 3 In 77 of 183 societies there was no medium of exchange at all. Their own coinage or paper money existed in 26. ### Property and inheritance **Who inherits the land?** (EA074, N = 856) - The sons: 354 (41%) - No one: land is not inherited: 223 (26%) - The father's-line kin, before the sons: 90 (11%) - The mother's-line kin: 60 (7%) - All the children: 55 (6%) - The children, daughters less: 43 (5%) - The sister's sons: 31 (4%) - No record: 435 Inheritance by all children, daughters and sons alike, the form in European civil codes, appears in 55 societies. In 223 there was no individual property in land, hence nothing to inherit. Murdock warned that these particular codes had proved 'inadequate' and fit societies with communal property poorly. **Is the inheritance split equally, or does one take it all?** (EA075, N = 820) - In equal parts: 307 (37%) - The eldest takes most: 248 (30%) - Land is not inherited: 223 (27%) - The youngest takes most: 22 (3%) - Whoever is judged most capable: 20 (2%) - No record: 471 Equal division and primogeniture, in which the eldest takes most, were recorded in 307 and 248 societies respectively. The opposite rule, inheritance by the youngest, appears in 22. ### Power, law and war **How many levels of authority sit above the village?** (EA033, N = 1,155) - None: 525 (45%) - One, for instance a chief over a few villages: 341 (30%) - Two: 162 (14%) - Three, for instance a state: 83 (7%) - Four, for instance a large state: 44 (4%) - No record: 136 In 525 societies out of 1,155 there was no political authority above the local community. The state with three or four levels, the form nearly everyone lives under today, was recorded in 127. **How is the village head chosen?** (EA072, N = 937) - Inherited through the father's line: 408 (44%) - Election or formal consensus: 116 (12%) - There is no such office: 109 (12%) - Inherited through the mother's line: 101 (11%) - Informal consensus: 87 (9%) - Influence, wealth or standing: 45 (5%) - Appointment by a higher authority: 43 (5%) - Age or seniority: 28 (3%) - No record: 354 Inheritance through the father's line is the most frequent record. Election, by vote or by a recognised form of consensus, appears in 116 societies, and informal consensus in 87. In 109 there was no such office at all. **Are there social classes?** (EA066, N = 1,109) - No: 538 (49%) - A hereditary aristocracy and commoners: 228 (21%) - Wealth distinctions, no hereditary classes: 217 (20%) - Several classes, tied to occupations: 86 (8%) - An elite that controls the land: 40 (4%) - No record: 182 The absence of any class distinction among free people was recorded in 538 societies out of 1,109. A hereditary aristocracy, in 228. **Is there slavery?** (EA070, N = 1,121) - No: 546 (49%) - Yes, hereditary: 227 (20%) - Yes, with no information on inheritance: 217 (19%) - Yes, but temporary, not inherited: 131 (12%) - No record: 170 Slavery was recorded, in one form or another, in 575 societies out of 1,121, just over half. In 227, the status was inherited. **How much respect does being a warrior bring?** (SCCS903, N = 151) - Some, but you need not be a warrior to have influence: 64 (42%) - A great deal: it matters for every man: 61 (40%) - No special respect: 26 (17%) - No record: 35 In 61 societies out of 151, being a warrior mattered for every man. In 26, a man who fought received no special respect. **How often is there war between communities of the same people?** (SCCS891, N = 160) - Infrequent: 89 (56%) - Frequent: 54 (34%) - Continual: 17 (11%) - No record: 26 The code tracks war between neighbouring, culturally related political communities, not feuds between families. 'Infrequent' is the most frequent record, but in 71 societies out of 160 internal war was frequent or continual. **How is war against outsiders regarded?** (SCCS907, N = 146) - Valued, even enjoyed: 72 (49%) - A necessary evil: 51 (35%) - Avoided and condemned: 23 (16%) - No record: 40 In 72 societies out of 146, war was valued and even enjoyed. In 23 it was avoided and condemned. ### Body and illness **Why do people fall ill? Spirits** (SCCS654, N = 131) - The main explanation: 74 (56%) - An important explanation, among others: 37 (28%) - A minor explanation: 18 (14%) - Absent: 2 (2%) - No record: 55 Murdock coded, for each society, how much weight each explanation of illness carried. Spirit aggression was the main explanation in 74 societies out of 131 and was entirely absent from two. **Why do people fall ill? Sorcery** (SCCS655, N = 130) - An important explanation, among others: 45 (35%) - A minor explanation: 45 (35%) - The main explanation: 24 (18%) - Absent: 16 (12%) - No record: 56 Sorcery, in the sense of a learned technique with objects and formulas, was the main explanation of illness in 24 societies and an important one in another 45. **Why do people fall ill? Innate witchcraft** (SCCS656, N = 131) - Absent: 81 (62%) - A minor explanation: 24 (18%) - An important explanation, among others: 17 (13%) - The main explanation: 9 (7%) - No record: 55 Murdock separated learned sorcery from the innate malign power of certain persons, which he calls witchcraft. The latter was absent from 81 societies out of 131 and was the main explanation in nine, all of them in Africa or around the Mediterranean. **Why do people fall ill? Infection** (SCCS645, N = 122) - Absent: 91 (75%) - A minor explanation: 30 (25%) - The main explanation: 1 (1%) - No record: 64 Infection, the idea that illness comes from something entering the body from outside, was the main explanation in one society out of 122 and absent from 91. It is the atlas's clearest measure of the distance between today's medicine and that of the societies described. **Is circumcision practised, and at what age?** (EA037, N = 1,116) - No: 746 (67%) - Yes, in adolescence: 117 (10%) - Yes, at an unspecified age: 97 (9%) - Yes, in childhood: 96 (9%) - Yes, at birth or in the first two years: 42 (4%) - Yes, in adulthood: 18 (2%) - No record: 175 Circumcision as a rule was recorded in 370 societies out of 1,116. The most often noted age is adolescence, when the operation is part of a rite of passage; circumcision at birth, the custom in Judaism and in American hospitals, appears in 13. **How long after a birth does the couple abstain?** (EA036, N = 372) - One to six months: 120 (32%) - Up to a month: 74 (20%) - One to two years: 73 (20%) - Over two years: 58 (16%) - Six months to a year: 39 (10%) - Not at all: 8 (2%) - No record: 919 The taboo that asks the mother to abstain while she breastfeeds lasted more than a year in 131 societies out of 372. The absence of any taboo was recorded in eight. ### Death **Do the remains get a second burial?** (SCCS1850, N = 162) - No: 101 (62%) - Yes, for almost all adults: 38 (23%) - Yes, depending on rank or group: 13 (8%) - Yes, in other forms or circumstances: 10 (6%) - No record: 24 Secondary burial, in which the remains are treated again, usually dug up, cleaned and laid to rest elsewhere, is attested on every continent. In 38 societies out of 162 it was the rule for almost all adults, and in 13 it depended on rank or group membership. **Are there ghosts?** (SCCS1960, N = 59) - Yes: 58 (98%) - No: 1 (2%) - No record: 127 Of the 59 societies coded by Rosenblatt and colleagues for mourning, 58 held a belief in ghosts, that is, dead people who can still be met by the living. Belief in ghosts was present, then, in almost every society of this small sample. **Do dead adults return in a new life?** (SCCS1998, N = 53) - No: 33 (62%) - A strong belief: 9 (17%) - A weak belief: 8 (15%) - A moderate belief: 3 (6%) - No record: 133 Belief in the reincarnation of adults was absent from 33 societies out of 53 and strong in nine. The sample is small, and the figures concern only the societies in it. ### Gods, ritual and play **Is there a supreme god, and does it care about morality?** (EA034, N = 775) - No supreme god: 277 (36%) - Exists, but takes no part in human affairs: 258 (33%) - Exists, intervenes and upholds morality: 198 (26%) - Exists and intervenes, but does not judge morality: 42 (5%) - No record: 516 A creator god who judges human conduct, the idea in Christianity and Islam, was recorded in 198 societies out of 775. In 258, a supreme god existed but took no part in human affairs, and in 277 there was none at all. **Are there trance states, and how are they explained?** (EA112, N = 658) - Trance is explained, at least sometimes, as possession: 281 (43%) - Trance exists, without a belief in possession: 135 (21%) - No trance and no belief in possession: 118 (18%) - Both trance and possession exist, but separately: 79 (12%) - A belief in possession exists; trance is not recorded: 45 (7%) - No record: 633 Erika Bourguignon coded 658 societies by whether they know trance states and whether they explain them as possession by a spirit. The absence of any trance was recorded in 118. In 281, trance was explained, at least sometimes, as possession. **What kinds of games are played?** (EA035, N = 486) - Only games of strategy: 250 (51%) - Only games of physical skill: 106 (22%) - Strategy, skill and chance: 58 (12%) - Strategy and skill: 39 (8%) - No game with a winner: 26 (5%) - Skill and chance: 5 (1%) - Only games of chance: 2 (under 1%) - Strategy and chance: 0 (0%) - No record: 805 The atlas counts only games with a winner. Games of strategy, such as chess or go, were recorded on their own in 250 societies out of 486. Games of chance were absent from 421, and 26 societies had no game with a winner at all. **Are the gods and spirits male or female?** (SCCS576, N = 67) - Males more numerous or more powerful: 24 (36%) - Both sexes, alike: 20 (30%) - More males but equal power, or the reverse: 13 (19%) - All male: 10 (15%) - No record: 119 In the 67 societies with a record, supernatural beings were all male in ten, and equally numerous or powerful across both sexes in 20. **Had Christianity, Islam, Buddhism or Hinduism reached them?** (SCCS2002, N = 186) - No, their own religion: 116 (62%) - Christianity: 30 (16%) - Islam: 26 (14%) - Buddhism: 10 (5%) - Hinduism: 4 (2%) - No record: 0 At the date of observation, 116 of the 186 societies in the closer sample kept their own religion. Islam had reached 26, Christianity 30, Buddhism ten and Hinduism four, to varying depths. ## Method and sources The groupings on this page are the author's. Each answer merges one or more source codes, listed in the question's record, and the denominator of every count is the number of societies with a code under that heading. Codes that do not describe the practice (for instance 'present, but no data on the sex of those who do it') are kept out of the denominator and shown in the record. No value, coordinate or date has been added to what is in the datasets. The dots on the map are the D-PLACE coordinates, nudged slightly so that they do not cover one another; a dot is the place of the observation, not a people's territory. The coastline is from Natural Earth. The language family is Glottolog's, through the language code attached to each society in D-PLACE. The 'by language family' option colours the eight families with the most societies in the atlas and leaves the rest grey; in the grouped rows the dots are ordered by region or family, so that related groups read as blocks. Related languages can reflect a shared history; no count on the page is adjusted for that relatedness. Only the questions you answered and for which the society has a record are compared. A society enters the ranking if it has a record for at least seven of the questions you answered, or for all of them if you answered fewer than seven. Societies are ordered by the share of differing answers among those compared, in decreasing order, then by the number of compared questions, then alphabetically. A missing record never counts as a difference. The ten questions are: when two people marry, who gives wealth to whom, who lives under the same roof, at what age does breastfeeding end, what shape is the house, are the animals milked, is the inheritance split equally, or does one take it all, how many levels of authority sit above the village, is circumcision practised, and at what age, do the remains get a second burial, is there a supreme god, and does it care about morality. The codes are those of the D-PLACE editions v3.2.1 and v3.2, kept because at residence they distinguish twelve codes where edition v3.3 folds them into ten; the counts on this page are identical in both. - **D-PLACE.** Kirby, K. R., Gray, R. D., Greenhill, S. J. et al. (2016). 'D-PLACE: A Global Database of Cultural, Linguistic and Environmental Diversity', PLOS ONE 11(7). The two datasets below are the D-PLACE editions, distributed in CLDF format; the aggregated D-PLACE bundle was not read. Licence CC BY-NC 4.0. https://d-place.org/ - **Ethnographic Atlas.** Murdock, G. P. (1967). Ethnographic Atlas, University of Pittsburgh Press; with J. Patrick Gray's corrections (1999). Derived dataset: Murdock, G. P., Textor, R., Barry, H., White, D. R., Gray, J. P., Divale, W. T., 'D-PLACE dataset derived from Murdock et al. 1999 “Ethnographic Atlas”', edition v3.2.1, Zenodo (version DOI: 10.5281/zenodo.17602181). Licence CC BY-NC 4.0. https://doi.org/10.5281/zenodo.17602181 - **Standard Cross-Cultural Sample.** Murdock, G. P., White, D. R. (1969). 'Standard Cross-Cultural Sample', Ethnology 8(4), 329–369. Derived dataset: 'D-PLACE dataset derived from Murdock and White 1969 “Standard Cross-Cultural Sample”', edition v3.2, Zenodo (version DOI: 10.5281/zenodo.17601873). The codes used here were published by Barry and Paxson (1971), Broude and Greene (1976), Whyte (1978), Murdock (1980), Nammour (1974), Murdock and Morrow (1970), Murdock and Wilson (1972), Schroeder (2001), Rosenblatt, Walsh and Jackson (2011). Licence CC BY-NC 4.0. https://doi.org/10.5281/zenodo.17601873 - **Glottolog.** Hammarström, H., Forkel, R., Haspelmath, M., Bank, S. (2025). Glottolog 5.3. Leipzig: Max Planck Institute for Evolutionary Anthropology. Licence CC BY 4.0. https://glottolog.org/ - **Natural Earth.** Coastline at 1:110 million, public domain. https://www.naturalearthdata.com/ [Ediția în română](https://mariuscomper.uk/feluri-de-a-trai/) · [mariuscomper.uk](https://mariuscomper.uk/)