Cross-Cousins in a Mayurbhanj Village
In a village of 240 households in Mayurbhanj district, three named patrilineal clans intermarry and village exogamy is the stated norm. The kinship terms recorded are: mother's brother is called by the same term as father's younger brother; mother's brother's daughter (MBD) and father's sister's daughter (FZD) are both addressed as 'sister', and their brothers as 'brother'. Of 200 marriages reconstructed from genealogies, 61 are with an MBD, 8 with an FZD, 40 with a woman from another village unrelated by any traceable link, and the remainder with more distant kin. Elders say 'a man should marry his mother's brother's daughter; that is our custom', but also that 'nowadays boys choose their own'. A postgraduate student on a two-week visit writes in her field diary that the system is 'prescriptive matrilateral cross-cousin marriage, an elementary structure of alliance', and plans to submit this as a finding. She asks you to review her conclusion before she leaves.
1. Assess the terminological data. What does equating mother's brother with father's younger brother suggest, and why does calling both cross-cousins 'sister' complicate a straightforward classification?
Read the worked answer
The MB = FyB equation. Merging the mother's brother with the father's junior brother is a classic bifurcate-merging feature: relatives on the mother's side are merged with same-generation relatives on the father's side, typically by sex and relative age. It signals that the system is organised by generation and seniority rather than by side, which is characteristic of Iroquois-type terminologies.
Why the cross-cousin term is awkward. In a standard Iroquois arrangement, cross-cousins (MBD, FZD) are terminologically distinguished from parallel cousins, and parallel cousins are called 'sibling'. Here both cross-cousins are called 'sister', a generational merge. Two readings are possible and the diary picks the wrong one too quickly:
- The term is a kin-class address term with no implication about marriageability - common when sibling terms are extended to all cousins.
- The term carries a marriage prohibition for the FZD but not the MBD, which would mean the prohibition is not carried by the term but by a separate rule about clans, villages or 'return' of women.
What must be checked. Whether FZD is classified into ego's own clan by a fiction (e.g. through a clan-level 'sister' idiom); whether the 8 FZD marriages were elopements or normative; whether the terms recorded are address or reference terms (they differ); and whether three clans with village exogamy simply make most kin categories statistically rare. Terminology alone cannot establish a marriage rule - that is the diary's core error.
2. Distinguish prescriptive from preferential cross-cousin marriage and show how the 61/8/40 figures bear on the distinction. What would you ask the elders that the student did not?
Read the worked answer
The distinction. A prescriptive system specifies the category from which a spouse must come; there is no legitimate alternative, and the terminology typically encodes marriageability (as in asymmetric alliance with 'wife-giver' and 'wife-taker' categories). A preferential system states a first choice which may be forgone for demographic, economic or personal reasons without any breach of norm. The difference is normative, not statistical: both can produce similar marriage frequencies.
What the numbers show - and do not show. 61 of 200 marriages (30.5%) with MBD is a high figure for a village of 240 households, and is consistent with strong preference. It is not proof of prescription, because:
- The eligible pool matters. With village exogamy, a man's MBD may be one of very few women he already knows and can negotiate with; high frequency can follow from opportunity.
- 40 marriages to unrelated outside women plus 8 FZD marriages show that alternatives are used and are not treated as scandalous - the student must ask whether these were arranged or 'love' matches.
- The 8 FZD cases are the diagnostic ones: if FZD is terminologically a sister and marriage to her is prohibited but occurs with mild comment, the rule is a preference with a prohibition on one cross-cousin; if those marriages cause lasting dispute, the prohibition is real.
Questions the student did not ask. 'If a man has no MBD, or her family refuses, what does he do?' 'Has any MBD marriage been refused, and why?' 'Who decides - the couple, the mothers' brothers, the clan elders?' 'What is given and received at the marriage, and in which direction?' 'Was there ever a case of a man marrying his FZD, and what happened afterwards?' Negative and hypothetical cases, not just completed marriages, carry the normative information. Two weeks is enough to ask them but not enough to assert an elementary structure.
3. The student's claim invokes 'elementary structures of alliance'. What theoretical and methodological problems arise from applying that frame here, and how would you redirect her write-up?
Read the worked answer
Theoretical problems.
- Alliance theory was built on prescriptive, often asymmetric systems (MBD marriage with directed wife-giving between groups). Importing it requires demonstrating that the whole clan or lineage, not just the individual, is in a durable wife-giving relation. Nothing in the data shows inter-clan directionality across generations.
- Descent and alliance are not alternatives to be chosen by taste. The clan and lineage data suggest descent is doing real work (clan exogamy); residence, inheritance and authority need to be described before alliance is invoked.
- Terminology-to-structure inference is unsafe. The terms are address terms in a multilingual setting (Odia and a Munda/Korean-family language may be used side by side), and the recorded vocabulary may be a regional koine rather than the local system.
Methodological problems. A two-week visit, no genealogies beyond marriage pairs, no residence data, no divorce or widow-remarriage cases, and no recording of who was present during interviews. The 200-marriage sample also needs stating: are they all extant marriages, all ever-married women, or only those with a living elder to recall them? Recall bias will inflate kin marriages.
Redirection. Have her rewrite the section as: 'Marriage with the MBD is a stated preference and occurs in about 30% of reconstructed marriages; FZD marriage is rare but not unknown; the terminology is bifurcate-merging with generational sibling extension. Whether the preference is prescriptive cannot be determined from these data.' Then add a data-collection plan: full genealogies for 40 households, residence after marriage, marriage payments, dispute cases, and interviews with women as well as elders. Publish description, not a typological label.