The family in its various forms is the universal basis of all human societies and social structures. However the institution of the family has decayed in modern times, so converting society into an alienated agglomeration of disconnected individuals, susceptible to arbitrary, remote and authoritarian governance.
This article was published as Chapter 2 of The Stable Society, The Wadebridge Press 1978. It was originally published in two parts in The Ecologist, Vol. 6 No. 1 and Vol. 6 No. 2, 1976.
Until recently, it has been fashionable to maintain that other forms of human social organisation are more fundamental than the family. The original unit of human organisation was taken to be the primaeval horde, like that of the baboons. This is one of Freud’s assumptions , and also one of Trotter’s , and is implicit in his theory of man’s social development, which began when the sexually deprived bachelors rose against their overbearing Pasha-like fathers. This myth was exploded* by Malinowski, who writes:
“Very often it is assumed by anthropologists that humanity developed from a gregarious simian species and that man inherited from his animal ancestors the so-called ‘herd-instincts’. Now this hypothesis is entirely incompatible with the view here taken that common sociability develops by extension of the family bonds and from no other sources.” 
(* – The Oedipus complex, to which Freud attaches so much importance, only makes sense in a patrilineal society. About 40 percent of known societies are matrilineal. In such a society, there is no father figure who both exerts authority over the child and sleeps with the child’s mother. The individual who behaves in a ‘fatherly’ manner towards the child and is head of the household is the senior male in the mother’s lineage group (usually her brother, sometimes her father) and therefore has no sexual relations with her. Hence the Oedipus complex cannot arise. Malinowski had a long controversy with Ernest Jones on this subject.)
Murdock also regards this principle as fundamental:
“Unlike the ants and bees, man is not biologically a social animal equipped by heredity with prepotent capacities for complex associative life, but in every individual case must be bent and broken to group living through the arduous process of socialisation and be kept in the paths of conformity through the imposition of social controls. The first anthropologist fully to appreciate this basic fact was Malinowski with his emphasis upon the factor of ‘reciprocity’ in the maintenance of norms of interaction.” 
Tinbergen comes to the same conclusion looking at the problem from a very different point of view. He writes:
“Contrary to current thought, there is not in my opinion a social instinct in the sense in which we normally understand it. There are no special activities which we can call social and which do not already form part of some other instinct. There is nothing to make us believe that there is at work a system of centres controlling social activities.” 
In other words, the information and corresponding instructions that will make man a family animal are transmitted from one generation to the next; while those that will make of him a social animal must be developed during the process of socialisation. That is what education is for – a fact our educationalists have long lost sight of.  The information necessary for man to become a social animal is thus of a cultural nature, rather than of a genetic one.
However, since the total information pattern that will determine the behaviour of a social system constitutes a single whole, organised hierarchically from the general to the particular (the latter information having developed from the former by the process of differentiation), it follows that the general instructions that have made of man a social animal must he derived from those that made of him a family animal.
What is more, by virtue of the rules of behavioural development, one must also assume that these instructions will provide serious constraints on his range of possible adaptations. It is for this reason that we find the same cultural strategies exploited over and over again in parts of the world that have had no contact with each other, just as is the case with biological strategies – hence, cultural convergence.Back to top
The universality of the human family
Before we follow this line of argument any further, we must establish that the family is universal among stable societies; otherwise, in terms of the basic principles of development stated above, no society would be possible. The universality of the family was established by Murdock on the basis of a sample of 250 cultures, figuring in the Yale Cross-Cultural Index. This assumes his definition of the family as satisfying “sexual or reproductive, educational and economic requirements”. 
Are there any exceptions to this principle? A number have been proposed. The first is from Melford Spiro who considers that the family has largely been dispensed with in the Israeli kibbutzim.  This does not appear to be a serious objection. Firstly, the kibbutz is probably a transient form of social organisation. There is no evidence that it is stable. It is in any case a contrived one, in the sense that it has not developed by what might be called a ‘cultural evolution’, but has simply been established as an experimental social form.
Also, it is by no means certain that the family has in fact entirely been dispensed with. The kibbutz has undoubtedly taken over a number of the roles normally fulfilled by the family, such as the educative one (which in our own society has been taken over by the State), but the family still provides, to a certain extent at least, the protective environment within which the earlier phases of the socialisation process must occur. There is a more interesting exception, however, namely the Nayars of the Malabar coast, in what is now the Indian state of Kerala. 
The Nayars are a matrilineal and matrilocal people. They live in extended matrilineal families run by the oldest male (‘karanavan’). These extended families are organised into lineages. In the old days, every two years, these lineages would hold a grand ceremony at which all the girls who had reached puberty would be ritually married with men drawn from specific lineages with which theirs were linked. At these ceremonies the girls were presented with certain ritual gifts including a gold necklace referred to as a tali.
Following the ceremony, the girls were isolated with their husbands and deflowered, though this was apparently not a necessary part of the ceremony. Thenceforward the girl’s status changed. She was called Amma, meaning mother, and allowed to have children – not by the ritual husband who need have no further contact with either her or her children but by her lovers. These were referred to as ‘sambadham’. She could have as many as she liked though they had to be of the appropriate sub-caste and outside her lineage.
When she became pregnant it was essential that one of them should acknowledge probable paternity, which was done by providing a fee of a cloth and some vegetables to the midwife. If no such acknowledgement was forthcoming, it was assumed that the girl had had relations with a man of lower caste or with a Christian or a Moslem. This meant her expulsion from her society.
The duties of the sambadham from then onwards were limited to providing his mistress with gifts at festivities. He had nothing to do with the maintenance of the mother or the upbringing of the child, which was the duty of the karanavan, the chief male matrilineal relative. The children’s duties to the tali husband were limited to mourning at his death. The sambadham – whether the progenitor or not – were addressed as ‘accahan’ or Lord, but no kinship terms extended to their family and no mourning was observed at their death.
This organisation has defied customary classification. Many have maintained that it provides an exception to the principle of the universality of the family. Others (Murdock included) doubt whether this account of Nayar organisation is authentic.
The reason why it presents a classificatory problem is that our method of classification is wrong. We are accustomed to think of the family as consisting of a father, a mother, and some children, with possibly one or two grandparents attached. Instead we should think of it in terms of the functions it fulfils. The family would then be present if the family functions were fulfilled by individuals biologically and culturally adapted to do so. The community as a whole or a specialised institution would not qualify, for a family is a differentiated system at a particular level of organisation. Its functions must be fulfilled by individuals – and specific ones at that. The point is, however, that they need not all be fulfilled by the members of what we regard to be the family.
For each different function, in fact, the family system (seen functionally) can consist of a different set of individuals. Thus the first family function is to engender the child. Associated with this is that of satisfying sexual needs. Among the Nayars, the family system whose normal functioning achieved these ends was composed of a girl and the sambadham. The role of parents is to bestow upon the children the parental care and affection they require and to undertake the task of instilling in them the basic cultural values and of teaching them the fundamentals of the tasks they will have to fulfil in later life.
Among the Nayars, this function was fulfilled by the mother and her matrilineal lineage headed by the karanavan. Last but not least, a child requires an identity or a social status, the basis of which the family can alone provide – a point neglected by Murdock in his definition of the family. This was achieved by yet a third social system, that composed of the girl and her tali husband. Thus, the essential family functions were fulfilled in the case of the Nayar by three different social groupings. If we regard this as abnormal, it is only that in our Western society they tend to be fulfilled (in so far as they are fulfilled at all) by a single social grouping which we call the family.
What is more, it can be shown that the same principle holds for larger groups than the family. Thus, after reviewing the different criteria for determining what constitutes a political unit, Lucy Mair writes, “there are some societies of which it is difficult to say that there is one political community for all purposes.” 
Thus a tribe, which is often endogamous and hence the unit of behaviour for marital purposes, is divided into different lineages which are often the effective units of political life, which may themselves be paired off into moieties. These may or may not correspond with the village, which will be the unit of behaviour for a large number of social and economic functions, etc.
In addition, the men in particular will be divided into age groups, each with a particular function, which in turn may form part of wider age classes – the young, the adolescents and the elders. Furthermore, there will tend to be secret societies which are units of behaviour for magical and ritualistic purposes and working groups which will be the units of behaviour for various economic functions.Back to top
How is the family held together?
The functions fulfilled within the family are designed to satisfy the needs of all the parties concerned. Thus, a father needs to behave in a fatherly way as much as a child requires that his father should do so (see the Hierarchical Co-operation Principle). The members of the family are thereby dependent on each other. It is this dependence which provides the bonds that hold the family together.
There are a number of different family bonds, such as those that hold together a father with his daughter, a mother with her son, a mother with her daughter, a man with his younger brother, a girl with her younger sister, a brother with his sister. These bonds are all different and also asymmetrical. The relationship of a father to his daughter, for instance, is very different from that of a daughter to her father.  The relationship of a father to his children differs even more noticeably from the mother 5 relationship with her children. According to Fromm:
“The love of a mother for her child, antecedent to that of the father, appears to be an unconditional love, whereas that of the father is conditional to the child’s ‘good’ behaviour or achievements.” 
This is reflected in the different relationships between a society and the gods it worships. As Fromm notes, it is no coincidence that most undisciplined and self-indulgent cultures have turned to a mother goddess, whereas more virile peoples, wishing to be judged according to their moral worth and their achievements, have chosen a paternal symbol as their chief deity.
What is important is that the basic differences between these bonds are exploited to determine culturally differentiated behaviour. It is only by maintaining this differentiation that co-operation is possible and that the society can display order or negative entropy.Back to top
The exploitation of the family bonds
The fact that we tend to classify our social and physical environment in terms of the classificatory system applied to the different members of the family should be evident from our personal experience.
Tinbergen shows that we view our domestic pets in this way. Dogs with snub noses and high foreheads are particularly popular among women requiring a child surrogate.  Similarly, behaviour towards political leaders can only be understood if the latter are interpreted as fathers husbands, lovers, sons, grandsons, brothers, etc. One whose image does not permit such classification has little political future, indeed.
Even consumer products are regarded in this way. A camera with a huge telescopic lens dangling over the belly of a dashing young photographer is clearly regarded as an extension to his penis by admiring girls. In this connection, I remember a cartoon of a dapper little man with pince-nez hesitating at the motor-show between a staid family model and a fast sports car. The salesman was saying, “Well, Sir, it all depends whether you want a wife or a mistress.”
We have already noted that we regard the gods we worship as fathers, mothers, and children; what is not generally recognised, however, is that subconsciously we see the individual members of our community in much the same way.Back to top
The fact that the family bonds are extended to embrace the members of the community is reflected in the elaborate kinship terminology developed by tribal societies, and in terms of which the whole community is classified. In this way a classificatory system, as Radcliffe Brown points out,
“may be applied over a wide range of relationships. Thus a first cousin of the father, being his father’s brother’s son, whom he therefore calls ‘brother’, is classified with the father and the same term ‘father’ is applied to him. His son in turn, a second cousin, is called ‘brother’. By this process of extension of the principle of classification, nearer and more distant collateral relatives are arranged into a few categories and a person has many relatives to whom he applies the term ‘father’ or ‘mother’ or ‘brother’ or ‘sister’.” 
The most important feature of these classificatory terminologies was pointed out long ago by Sir Henry Maine:
“The effect of the system is in general to bring within your mental grasp a much greater number of your kindred than is possible under a system to which we are accustomed. In other words, the classificatory terminology is primarily a mechanism which facilitates the establishment of wide-range systems of kinship.” 
What is more, as Radcliffe Brown observes,
“the attitude and behaviour of a person towards a particular person, is affected not only by the category to which he belongs but also by the degree of nearness or distance of the relationship. In classificatory systems there are many women whom a particular man calls ‘sister’. In some systems he will be prohibited from marrying these women. In some others he may not marry any ‘near sister’, i.e., any one of these women who is related to him within a certain degree of cognatic relationship, but may marry a more distant ‘sister’.” 
The important thing is that if people are referred to by different names there is a reason for it, and the reason appears to be that different names reflect expected differences in behaviour towards the people they refer to. Likewise, if two people are given the same name, it implies that the sort of behaviour due to both of them is the same. This is also the conclusion of Radcliffe Brown:
“I hold that all over the world there are important correspondences between kinship nomenclature and social practices.” 
In other words the elaborate kinship terminology used by tribal peoples reflects the fact that a tribal society is a highly differentiated system in which each individual has a specific identity and a specific role in terms of which it is only possible to understand the complex set of relationships between him and all the other members of his society.Back to top