Griot

Written By Kirthi Jayakumar

World over, there have been powerful ways of preserving and passing on knowledge from generation to generation. While writing and publishing changed the game by preserving information without losing it to memory, it fast became a dominant form of knowledge preservation and transfer. As a result, many other forms were either relegated to the peripheries or forced into oblivion with the preference of the written word. This resulted in epistemic erasure, and the resultant exclusion of multiple ways of knowing.

The Griot Tradition from West African communities such as the Mandinka, Wolof, Fulani, Hausa, and several others, is a powerful form of knowledge preservation and transmission. It presents us with another way of engaging with knowledge.

Before proceeding, we want to state that the Griot tradition, also called jeli or jali among Manding-speaking peoples, gewel among the Wolof, marok'a among the Hausa, and by many other names across West Africa, is a living hereditary practice and not a historical artifact. It is also called the Griot tradition, where the word griot is a French colonial imposition that has become the most widely recognized term in English. We acknowledge that the practitioners of this tradition carry their own names for themselves and their roles in their own languages, and those names carry authority that "griot" does not. The knowledge held within Griot tradition families is not uniformly public. Some of what they carry, namely genealogies, sacred histories, songs and oral texts, is restricted from public access and transmitted within families through apprenticeship. This resource has been built by drawing only from publicly shared knowledge and scholarly accounts that engage responsibly with the tradition. We recognize that the line between what is publicly accessible and what is not is determined by the communities themselves, not by outsiders. We also want to acknowledge that the Griot tradition has been subject to appropriation and flattening, particularly as “griot music” has been commercialized globally and as the term has been adopted loosely to describe any oral storyteller or community historian, stripping away the hereditary, social, and spiritual dimensions that define the role in its living context. While we have tried to resist this flattening, we welcome correction. The families who carry this tradition in Senegal, Mali, Guinea, Gambia, Burkina Faso, and their diaspora communities worldwide remain the authoritative custodians of its meaning, and their interpretation takes precedence over ours.

Understanding the Griot Tradition

When the founder of the Mali Empire, Sundiata Kéita rose to power in the 13th century, the Griot, hereditary historian-musicians of West African society, preserved the account of his life, his lineage, and the moral lessons of his reign. However, rather than in writing, they preserved these accounts in their bodies, voices, and memories, which were then passed down to subsequent generations through apprenticeship. Griots remained the living archives of their communities, as their work spanned the course of serving as historians, genealogists, diplomats, praise singers, musicians, and mediators. Their role was hereditary, and their knowledge was not merely informational but also powerfully performative. The art of telling a family’s history is to affirm its identity and its place in the social order, and to acknowledge its obligations to those who came before and would come after.

Griots advised rulers and held them to their genealogical obligations. Their praise conferred legitimacy, their silence and critique called for accountability. They also mediated disputes by drawing on historical precedents and genealogical connections that others did not hold. In many West African societies, Griots played a powerful role of holding up a mirror of truth to those in power. Griots could be of any gender identity.

What the Griot Tradition unsettles in AI governance

The griot tradition presents a fundamental challenge to the epistemological architecture of AI.  Contemporary AI systems are trained on data, which is a pool of text, images, audio, and behavioral signals that have been digitized, collected, and stored, but not necessarily with the consent of those whose data they all are. The prioritization of these forms of data has meant that knowledge that is relational rather than propositional, namely the knowledge that exists in bodies, performance, oral transmission, and in the memory of living people is invisible. The Griot tradition reminds us that this is not peripheral knowledge but the resources on which whole communities have come to rely on in making important decisions. Encoding governance in systems that cannot receive or engage with this knowledge is to make an active governance choice of privileging certain kinds of knowledge while rendering others illegible.

AI governance that takes the Griot tradition seriously would explore the question of whose knowledge counts as knowledge. The Griot tradition tells us that the technical and quantitative understanding of knowledge within current AI ethics frameworks is inadequate at best. It reminds us that communities hold rich knowledge about their histories, experiences of harm, and moral frameworks that are not available in any dataset. Legitimate AI governance should create pathways for this knowledge to enter the deliberative process as epistemically authoritative evidence rather than in a tokenistic fashion or as the exotic other that adds aesthetic value.

The Griot tradition also speaks to the question of accountability over time. Griots preserved histories of glory and records of obligation. They kept a record of what was achieved, what was lost, what was promised, what was owed and to whom, and the consequences that followed when obligations were broken. AI governance currently lacks any comparable institutional memory. Decisions made in one era about AI development and deployment are rarely traced to their consequences in subsequent periods. In the way it is currently practiced, accountability is episodic and fragmented, rather than generational. A governance architecture shaped by the Griot tradition would maintain living records of AI’s promises and consequences, held beyond databases, in communities. It would treat those community-held memories as authoritative inputs into any accountability process.

References

This article draws from the wisdom and work of West African communities. While educating ourselves on Indigenous worldviews is important, we understand that our actions can also contribute to and enable appropriation. As part of our ongoing attempts at practicing accountability, we invite readers to consider donating to Indigenous groups, collectives, organizations, or initiatives to support their lives and work.  

Kirthi Jayakumar
Previous

Vanua
Next

Jugaad