Yupa and Tarumanagara Inscriptions as Documents of the Kalinga-Gupta Diaspora: An Epistemological Reconstruction of Early Nusantara Historiography Based on Convergent Evidence

Epistemological reconstruction of the two oldest epigraphic corpora in the Nusantara archipelago—the Yupa Inscriptions at Muara Kaman, East Kalimantan (second half of the 4th century ce, palaeographic estimate) and the Tarumanagara Inscriptions in West Java (mid-5th century ce)—is built through convergent evidence: comparative linguistics, genealogical parallelism, title semantics, chronological reconstruction, maritime logistics, and external validation from the rise-and-fall pattern of the Chenla Kingdom. The argument is explicitly abductive (the best current explanation), not a claim of demonstrative certainty from a single decisive source. The central argument is presented in two explicitly separated claim levels. The first level—supported by strong convergent evidence—proposes that the Muara Kaman community is more parsimoniously interpreted as a diaspora formation with Kalinga/Dravidian affiliations and structural links to Gupta-era networks, rather than as an autochthonous polity undergoing linear Indianisation. In the same framework, Taruma is interpreted as a transplanted homeland toponym (consistent with the Gomat ̄ı/Candrabhaga ̄ sacred-geography replication) rather than as a securely demonstrated derivation from local indigo. The second level—presented as a working hypothesis—proposes that some names in the Yupa corpus may function as patronage epithets linked to Gupta figures: “Kundungga” as a possible preservation of a pre-Sanskritic vernacular name associated with Sri Gupta, “Mu ̄lavarman” as a possible epithet connected to ChandraguptaI, and “Pu ̄rn.awarman” as a possible epithet connected to Kumaragupta I. This second level remains provisional and requires additional primary corroboration. Ten independent categories of evidence are evaluated, with six primary categories contributing most strongly to the First Level: (1) systematic linguistic elimination indicates that “Kundungga” is unlikely to be Sanskrit or Austronesian; since Vogel (1918) and Chhabra (1945), scholarship has repeatedly noted that the name is non-Sanskritic; (2) the three-generation genealogical parallelism between the Yupa and early Gupta lineages indicates structural and chronological convergence; (3) the suffix -varman in this context is interpreted as functionally comparable to elite titulature in Gupta-connected networks, including Kamarupa; (4) the approximately 280-year inscriptional gap in Java (from Purnawarman, c. 5th century ce, to Canggal, 732 ce) is consistent with, though not by itself decisive for, a broader post-Gupta sys- temic contraction; (5) the naming from below dynamic (whereby a client community assigns patronage epithets to imperial figures rather than adopting those figures’ self-designations) offers one mechanism for preferring -varman over -gupta; (6) the inverse rise-and-fall pattern of Chenla provides an independent regional comparison consistent with a shared macro-network dynamic. Because direct primary evidence is fragmentary for both models, mainstream and alternative explanations are evaluated under the same evidentiary constraints. The label “Kingdom of Kutai Martadipura” currently has no direct epigraphic attestation in the seven Yupa Inscriptions.