The Granary That Belongs to Everyone and Therefore to No One

In the kebeles of the central Amhara highlands, a communal grain storage practice known locally as the ye-rist mahber gota — loosely translatable as the lineage-group granary — once served a redistributive function that no external food security program has adequately replicated or, in most cases, even attempted to describe. This essay concerns the structure of that institution: its membership rules, its seasonal logic, its methods of accounting, and the particular social mechanisms that held it together for generations before holding it together became too difficult.

The basic form is not complicated. A group of households sharing recognized descent from a common ancestor — typically between twelve and thirty families — pooled a portion of their teff and barley harvests into a jointly managed store. The store was physical: a round, thatched structure built on elevated stone footings to discourage rats and moisture, usually sited on land belonging to the eldest male lineage member. Contributions were measured not by weight but by container — a standardized clay pot whose volume was known to every member family, and whose dimensions were occasionally renegotiated by consensus when harvests ran significantly above or below expectation. This flexibility was not imprecision. It was calibration.

Access was governed by a set of conditions that varied between lineages but clustered around recognizable categories. Illness, crop failure due to hail or frost on a member's specific plot, the death of a draft animal, or a daughter's wedding — these constituted legitimate withdrawal events. A member seeking grain would present his case before the assembled group, usually at a Sunday gathering following church. The group would deliberate. There was no written vote. Agreement was signaled by the absence of sustained objection, which is a different thing from silence and requires some skill to read correctly.

Repayment was expected but not enforced through penalty. The social weight of non-repayment was considered sufficient deterrent, and in most cases it was. A household that failed to return grain over two consecutive seasons without acknowledged cause entered a kind of informal probation: it remained nominally inside the mahber but was understood, by all parties, to have reduced standing. Full exclusion was rare and constituted a serious communal event, discussed with the same gravity as land boundary disputes.

What made the system ecologically functional — and this is the point that development literature consistently undervalues — was its relationship to localized microclimatic variation. Within a single kebele, a hillside plot and a valley-floor plot might experience hail damage in different years, or in the same year with different severity. The mahber gota effectively spread this risk across topographic positions. A member farming a north-facing slope prone to late frosts was, in practice, insured by the harvests of members farming lower, warmer ground — and the arrangement reversed in drought years, when valley soils dried faster. The institution encoded ecological knowledge about local variability without anyone needing to articulate that knowledge explicitly.

The granary's decline is not a single story. In some areas, villagization policies in the 1980s disrupted the spatial proximity that made informal accountability possible. In others, the introduction of food-for-work programs inadvertently substituted an external system with different rules and no membership logic, making the parallel maintenance of the mahber gota seem redundant until it quietly ceased. In still other cases, the inheritance of the host land through competing claims simply made the building's physical location contested, and a granary without an agreed home cannot function.

What has not disappeared is the underlying accounting logic. In conversations across several woreda over many years, this writer has encountered farmers — particularly older women, who were often the actual managers of contribution records, despite holding no formal role in the institution — who could reconstruct the mahber's membership lists, contribution histories, and withdrawal events from memory with striking consistency across independent interviews. The knowledge exists. It is not archived. It does not need to be mourned. It needs, rather, to be systematically recorded before the generation that carries it is no longer available for interview.

A communal granary is not romantic. It was also a site of disputes, favoritism, and unequal labor. Any honest account must include that. But its core operating principle — that locally calibrated, relationship-enforced redistribution outperforms externally administered relief in managing recurring localized shortage — remains a hypothesis worth taking seriously, particularly as rainfall unpredictability in the highlands increases and the case for flexible, socially embedded risk management grows stronger with each difficult season.

Read in the app, leave a resonance, discover more stories.