Cross Cultural Survey
This survey investigates how family and child rearing patterns are related to political, social and economic behavior. It differs from previous cross-cultural studies in being concerned with the overall level of control and punishment, rather than the aim of such control or punishment (for example, controlling sexual behavior or toilet training), and in its focus on the age at which such control and punishment applies. Societies have been chosen to represent the widest possible range, in terms both of geography and political and economic complexity.
Because the political and economic makeup of many societies today has been heavily influenced by Western patterns of behavior, the focus where possible is on societies that have been less influenced, and political and economic behavior is as much as possible that relating to the traditional society. The Western societies included are chosen as having relatively traditional values, less affected by the drastic social changes of the past fifty years.
The correlations attached use Kendal’s tau, since the variables are ordinal rather than cardinal.
An updated version of this survey is currently being prepared.
General rules for coding
Unless specifically stated and where there is a difference, all variables apply to the male sex and the present time.
Where a variable differs among regions or status levels in the area of study, it is coded for the majority of people.
Figures must always be used when available, however limited the sample and however contradictory other passages may seem. For example, if one man in a village of thirty men has two wives, the polygyny rate is 6% unless other figures exist.
Where concrete data exist that fits the criteria, these must be used regardless of general impressions. Where no conclusive evidence is found, a variable may be coded if it seems ‘reasonably certain’. A 20% difference in judgment is allowed for.
Common sense may provide an answer not given in the book, for example that the United States is a monogamous society.
Lack of evidence may be used to score a variable where there is a reasonable discussion of the relevant subject.
If two codes fit a variable, the higher should be taken.
The object of the coding is to rank the societies in order. If one does not fit into an extreme category because it is still more extreme, then it may be given the extreme code regardless.
Certain words have precise meanings. ‘Definitely’ and ‘obvious’ mean that concrete evidence must be found to score the society at this level.
Where precise figures exist, ‘normally’ means 2/3, ‘mostly’ 1/2. Otherwise they are used impressionistically.
‘Or’ joining two circumstances means that only one must be found to score the society at this level. ‘And’ requires both. The same applies to lists of more than two. The biological family of man, wife and children is calculated at five unless other data exist.
There are a number of variables coded for the ‘traditional’ society rather than at the time they were studied. A traditional society is defined as one brought under effective colonial rule later than 1820. These are:
- size of political unit
- market economy
- government arbitrary
- monogamy
- feuds
- modesty in dress
Those on the left tend to be greatly affected by the mere fact of European control. Those on the right are affected to varying degrees, but the changes that do occur can be in most cases more sensibly put down to direct European influence than to changes in the prevailing temperament.
An attempt was made to avoid societies where European influence has been very great, and in most cases the “traditional” society is only a generation back at most.
Examples are given of societies in each category. These will not be shown to the coders.