CAPITULO 6 DISEÑO DE DETALLE
6.5 Evaluación del prototipo
The database finalised by the processes described in Appendix A and summarised in the previous chapter contained 1512 individuals. In addition to these named individuals, information was incorporated into the database for people who were only recorded in the genealogy as '?'. Although only limited information was available for these people, they were important as links between people, or in giving information about another person.
Just over half of the 1512 individuals in the genealogy were male (790, or 52.25%), 706 female (46.7%), and 16 (1.1%) were of unspecified sex (these were generally infants and people who died young) (Table 2.1).
Table 2.1: Number of each sex
Figure 2.2: Age distribution of the genealogy p o p u l a t i o n Males: Females Others: 790 (52.3%) 706 (46.7%) 16 (1.1%) Total: 1512 <1900 ' 1900-09 '1910-19 ' 1920-29 ' 1930-39 1940-49 1950-59 1960-69 >1970 Mmen ®women
For those whose date of birth is known, the distribution is as shown in the accompanying graph (Figure 2.2)1. This graph gives the dates of birth for the 377 (47.7%) men and 389 (55.1%) women for whom this data is available (it should be noted that this distribution includes people who were recorded as deceased in the genealogy). This sex-specific age distribution over 10 year intervals gives us some idea of how complete the information is going back into the past. It is likely that those who died young or without offspring would be under-represented in the genealogy except perhaps for very recently. Notably, there doesn't seem to be any tendency for the data on women to be any less complete than it is for the men, even in the earlier parts of the genealogy.
From the 1950s dates of birth were kept in the Mission records, and so would be quite accurate. Dates of birth for individuals born before the 1950s can be expected to be rather less accurate. Dates of death were given for people who died in the 1970s, otherwise only the information that the person was dead was given.
The average number of children amongst all women with children is 3.5, but as can be seen from Table 2.3, for mothers born up until 1939, the average (4.6) is rather higher. Mothers born between 1940 and 1949 already have
1 Raw data for the charts in this and other chapters, and more detailed Tables than those in the text can be found in Appendix B.
an average of 4.0 children each, and it is likely that some of these women were still in their child-bearing years when the genealogy was collected. Some of the youngest of the women in this group may not yet have begun to have children: a higher proportion (28.8%) of them have no children than is found in the groups of older women (around 20%). This of course may also be affected by an under-recording of women with no children in the older age groups. The women in the age group born after 1950 would still have been at most in their mid-twenties when the genealogy was collected. Already, however, they have an average of 2.1 children, and a number of them have up to 4 children (see Table B.3, Appendix B).
It should be fairly safe to assume that these figures underestimate the number of children that women actually have had, especially in the older age-groups, because of selective under-recording of children who died. C.H. Berndt believes that the infant mortality rate was formerly fairly high (1970:30). T a b l e 2.3: N u m b e r o f m o t h e r ' s b i r t h c h i l d r e n f o r w o m e n b y y e a r <=1929 1930-1939 1940-1949 >=1950 % of women with no children 19.7% 20.5% 28.8% 88.3%
average among those
with children 4.6 4.6 4.0 2.1
average (all women with children): 3.5
The following table, Table 2.4, gives us an idea of when women begin to have their children. Warner says that "the proper time for a girl to marry is when her breasts first start developing" (1969:118), so we would expect Yolngu women to marry and begin to have children from a very young age. This is substantiated by the genealogy.
Table 2.4 tells us, for the 142 mothers for which we have enough information, how old the mother was when her eldest child was born. In nearly half of these cases (47.2%) the mother was 16 or under, and in almost 80% of cases (79.6%) she was under 20 years old. The overall average is 17.5 years2. Because of similar under-recording tendencies as were discussed above, on average women would have started their reproductive careers earlier than this age. Only children who are recorded in the genealogy, with their dates of birth, are included in this averaging, and it is likely that in the earlier part of the genealogy children who did not live to adulthood are not included.
Table 2.4: Age of mothers at the birth of their first child
average: 17.5
Table 2.5 demonstrates that there are no significant differences across the time-span of the genealogy. Although the average for mothers born before 1929 is slightly lower (17.1 years) than the other two age groups (18.3 and 18.1 respectively), it is likely that the dates of birth of mothers and children in the earlier part of the genealogy will not be sufficiently reliable to make this difference significant.
Most of the women born after 1950 would have been in their teens or very early twenties when the genealogy was recorded, and so their ages at first childbirth cannot really be used for comparison. The figures for mothers
2Meggitt estimates that the average age at which a Walbiri woman has her first child to be about 17 years old (1965:161).
Number Age of women % 10-16 17-19 20-29 30-39 67 47.2% 46 32.4% 25 18.3% 3 2.1% 142
born after 1950 are given to show that there are still a significant number of mothers who are very young when they have their first children: the 20 mothers in the age group born after 1950 were under 20 years old, and 16 were 16 years old or under at first childbirth.
Table 2.5: Age of mothers at the birth
first child by date of birth of mother
<=1929 1930-39 1940-49 >=1950 10-16 25 11 15 16 17-19 14 10 18 4 20-29 5 8 12 30-39 2 1 average: 17.1 18.3 18.1 15.6
On average, men who have children have more (5.1) than women who have children. Indeed, 5 men have more than 20 children. The older men in the genealogy have more children (Table 2.6): men born before 1929 have, on average, 9.8 children. This is very likely to be due to the fact that the younger men in the genealogy will continue to have children as they get older.
Table 2.6: Average number of children of me n by year of father's birth
date of birth
of father: <=1929 1930-1939 1940-1949 >=1950
average number
of children: 9.8 5.7 2.5 1.5
average (all men with children): 5.1
Warner reported that men usually marry for the first time in their late teens or early twenties (1969:118), while Keen states that most do not acquire their first wives until they are at least 25 (1978:6)3. From Table 2.7 we see that most men (69.6%) have their first children in their 20s, although some are
3Keen is not consistent on this point. On the one hand he says that most men acquire their first wives between the ages of 25 and 30 (1978:6), but he also says that "men may marry as early as twenty-five, but the majority do not obtain their first wives until after the age of thirty"
under 20 (10.1%). This evidence supports Keen's observation, but there are clearly some men marrying at quite a young age, perhaps by inheriting or stealing wives from other men. The average for all men is 25.8 years old, although we see from Table 2.8 that this average is probably slightly depressed by the fathers in the 1940-49 date of birth group, who would only just be entering the 30-39 age group when the genealogy was compiled. The triangular shape of this table's entries indicates that there are probably some men in the 1930-39 date of birth group who have not yet become fathers.
Table 2.7: Age of fathers at the birth of their
first child Number Age of men % 10-19 8 10.1% 20-29 55 69.6% 30-39 13 16.5% 40-49 3 3.8% 79 average: 25.8 years
Table 2.8: Age of fathers at the birth of their
first child by date of birth of father
Date of birth of father Age at fatherhood <=1929 1930-1939 1940-1949 10-19 3 0 5 20-29 19 21 15 30-39 6 7 40-49 3 average: 27.3 26.3 22.9 2.2 Clan Membership
It will be recalled from Chapter 1 that the genealogical information was provided on sheets grouped by the clan membership of the man whose information was given on any particular sheet. In addition, a small amount of genealogical information was given in tree form for smaller, peripheral
groups. People in clans other than the ones represented by sheets in the genealogy were also found at various points throughout the genealogy, as wives or fathers of wives, husbands of daughters and so on.
In Table 2.9, the population is broken down as to clan membership. The major clans, for which the information was provided in the form of sheets of individuals and is most complete, are indicated by Other clans, where the information was more sketchy, are labelled Individuals with named clan affiliation which is not one of the major clans are grouped together according to moiety.
Table 2.9: Population of clans
Abbrev. Full name Population
Yirritia clans: day Daygurrgurr 323 wan Wangurri 72 wob Wobulkarra 64 bir Birrkili 76 wul Wulaki 8 gad Gadukadu 15 wal Walamangu/Ngurruwulu 10 war Warramiri 9
other Yirritja clans 52 Dhuwa clans lgl Liyagalawumirr 104 lgw Liyagawumirr 181 mlk Milkmilk 11 djm D j ambarrpuyngu 179 man Manharrngu 41 mur Murrungun/Wolkpuy 51 gam Gamalangga 46 gor Gorryindi 14 ngy Ngaymil 23 mal Malarra 13
other Dhuwa clans 15
★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★
(Mildjingi, Balmawuy, Wora, Guyamirrilili, Ganalbingu etc.)
★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★
(Marrangu, Datiwuy, Galpu etc)
As was described in Chapter 1 the MMB line is said to be sometimes found in Ego’s clan (Warner 1969:17, Shapiro 1981:53). This is significant in the
context of exchange between clans, and for the isolation of groups treated corporately in the exchange of women. If Ego’s MM is found in his clan then there is no assymetrical wife-giving/wife-taking relationship between Ego's clan and his M's clan, (although there may be wife-giving/wife-taking relationships between subgroupings of the clan). In Table 2.10, we investigate the frequency with which MM is found in Ego's clan. As might be expected, the frequency with which the MM line of any particular Ego is found within the same clan tends to increase with the size of the clan. In total, about 10% of Egos will find their MM in their own clan.
Table 2.10: Frequency of MM line in same clan
MM clan known Same as own %
Yirritja clans: day 240 40 16.7% wan 25 0 0.0% wob 14 0 0.0% bir 35 2 5.7% war Dhuwa clans 1 0 0.0% lgl 19 4 21.1% lgw 106 9 8.5% djm 98 6 6.1% man 3 0 0.0% gam 16 0 0.0% ngy 18 0 0.0% other Dhuwa 15 0 0.0% Total: 590 61 10.3%
Other clans had no members whose MM clan was known.
In general the patrilines of Ego and MM are distinct. In investigating this, a check that no Ego and MM have a common patrilineal ancestor showed only three exceptions.
Of the 323 members of Daygurrgurr clan (including deceased members), 281 are members of one patrilineage. On Keen's large chart of the genealogy, this lineage is drawn as two large sublineages, descended from
two marriages of the apical ancestor. One of the sublineages is descended from a son of one of these marriages, the other from two sons of the other marriage. One of these latter sons is recorded in the Register of Wards as being born in 1886. The two half-brothers heading each sublineage had 11 and 12 wives respectively. The three instances of MM being found in the same lineage as a group of uterine siblings all occur in this composite lineage, and in fact in the lines of all three of these half-brothers.
The situation diagrammed at Figure 2.11 occurs 3 times in the large lineage described above. A and B are brothers: in two of the three cases with same father only, in one case with same mother and father. The son of A has married his FBDD4, so that for any of the children of these marriages, in the lineage of A, their MM is in the
lineage of B.
As we remarked above, this composite lineage is very large, comprising 281 individual members of the genealogy. It seems likely that these sublineages are in the process of becoming corporately independent. None of the children of the three marriages described above were born before 1959, and most were born through the 1960s and early 70s. We should perhaps expect to see more marriages in the near future, if they have not already taken place, between men of these sublineages and the DDs of men of the other sublineages.
4Terminologically *waku' (a rather irregular marriage). Warner cites this as a relative who would sometimes be a partner for an affair, but that this would generally only occur between distant
Figure 2.11: Daygurrgurr marriages with FBDD
Keen observes the tendency in this clan for members of one sublineage to live largely on the mainland, the others at Milingimbi (1978:139).
2.3 Marriage
From the information provided in the genealogy it was possible to compile information on 769 marriages. Each marriage was stored as a husband/wife pair, giving the ability to compare the marriages as to the characteristics of the spouses (clan, subsection etc.), and to look at possible genealogical relationships between the spouses.
Tables 2.12 and 2.13 give details of the total number of spouses recorded for each person in the genealogy. Marriages are not included where one of the spouses was one of the '?' people, since we have no way of knowing if these are distinct individuals. It is interesting to note that 61.4% of the husbands and 71.8% of the wives have only one spouse recorded in the genealogy.
Table 2.12: Number of wives for each husband
1 189 61.4% 2 59 3 26 4 13 5 6 6 8 7 3 8 or more 4 308
average number of wives per husband: 1.9
Warner states that the average number of wives was about 3.5 for married men (1969:66), and he records the case of a man with 17 wives, higher than that recorded for any man here (although, as described above, wives who were not named were excluded from the analysis).
Table 2.13: Number of husbands for each wife 1 308 71.8% 2 94 3 17 4 6 5 2 6 1 7 1 429
average number of husbands per wife: 1.4
Shapiro's average for 264 marriages effective in late 1965 on Elcho Island is 1.7 wives per husband (1981:75)5. The rate of polygyny found by Meggitt among the Walbiri is on average 1.5, although this figure varied between Walbiri communities (1965:149)6. Both Meggitt and Shapiro were able to show that the number of wives increased with age.
Husbands are, on average older than their wives. This is to be expected from the literature, and is also suggested by the difference in the ages of men and women at the birth of their first child. The information on dates of birth in the genealogy gives an average of 11.5 years difference in age between a man and wife (N=129). Keen (1982) calculated the average to be 13.25 years for a sample of 50 marriages, but this only included marriages where the husband was older than the wife.
The mean age difference between spouses is 13.6 for the Walbiri (Meggitt 1965:158), compared with 17.8 years among Roses's sample of Wanindiljaugwa marriages on Groote Eylandt (Rose 1960:Chapter 8). Given the high level of Yolngu polygyny the lower figure for the Yolngu is somewhat surprising.
Calculated from Shapiro's data: 264 wives, 152 husbands, 44 men over 20 years old with no wives. Men over 40 in Shapiro's sample had, on average, 2.1 wives while married men under 40 had an average of 1.3 wives.
Since we don't have any information on when marriages took place, it is very difficult to form any conclusions about how marriage is related to age. It would be useful to be able to analyse the number of marriages according to age of husband or wife, and to look at how other characteristics of marriage may change across the lifespan of the individual. Of particular interest would be the relative ages of husband and wife. A very rough analysis might be attempted using the dates of birth of children to estimate the timing of marriages, however it is almost impossible also to date the end of marriages, especially if they do not end with the husband's death. It would perhaps be better not to attempt such an analysis on the evidence provided, because of this difficulty with knowing when marriages begin and end.
An indication of the sort of error inherent in using the date of birth of the first child to a marriage as an estimate of the date of marriage is given by an examination of the spacing in children's dates of birth for each mother. Averaging the spread of dates of birth of children for mothers in different date of birth groups (Table 2.14), we see that the error is likely to be greatest for wives born before 1929, and will decrease as we come towards the present. It should also be noted that this method does not give any estimate for marriages where there are no children. This will be most important for women remarrying as widows when they are beyond their childbearing years.
The following analysis of husband/wife age differences should therefore not be regarded as a definitive analysis but more as an exploration. The error discussed above will not affect the actual amount of the age difference between spouses, which only depends on their dates of birth being given, but will result in some inaccuracy in the assignment of the marriages to
groups according the age of the wife at marriage. Even given these comments the figures in Table 2.15 display the trend we should expect from the literature. Very young women are married in their early teens to men who are on average about 15 years older than them, and as they get older and are remarried as widows their husbands are relatively younger. Eventually, if they remarry in their 30s, their husband is very likely to be a younger man.
Table 2.14: Average birth spacing by date of birth of mother
Average overall = 4.1 years
Date of birth of mother
<1929 1930-39 1940-49 >1950
average birth 5.3 4.1 2.7 2.6
spacing
Table 2.15: Average age difference between spouses by estimated age at marriage of wife
Estimated age at age difference
marriage of wife between spouses
10-14 15-19 20-24 25-29 30-34 35-39 16.8 15.5 7.0 1.9 -0.5 -3.0
average overall 11.5 years
Note: The wife's age at marriage is estimated using the date of
birth of the first child of the marriage. See discussion in the
text about errors in this method of estimation.