Before the Alupas: Fragments from Poems, Edicts and Greek Maps

Tulunadu has no inscriptions from its earliest centuries. What it has are mentions: someone else’s poem, someone else’s map. This page gathers them and says, for each, how much it can really tell us.

Nannan and the Sangam poems

The Sangam poems are the oldest Tamil poetry, from about the early centuries BCE and CE. Wikipedia’s summary of the Dakshina Kannada record says the earliest recording of what would become the district is in a poem by the Sangam poet Mamulanar.¹

The same poems tell of a chieftain called Nannan. He is described as the ruler of Ezhimala, a hill on the north Kerala coast, with its sacred mango tree, and as a stubborn enemy of the early Chera kings, who killed him in battle at Perunthurai. Some accounts also say he held Tulu Nadu and the Konkan.² There was more than one ruler named Nannan, and how they relate to each other is not clear.²

So the poems show a northern coastal chieftaincy in conflict with the Tamil kings. They do not prove that his people spoke Tulu or that his centre lay in what is now Dakshina Kannada.

Names from the Greek and Roman world

Writers far away also name this coast, though matching their names to today’s places rests on later scholars’ guesses.³

Pliny the Elder mentions a place called Nithrias, probably the Netravati river.
Ptolemy, in the 2nd century CE, names two ports, Barace and Maganur, which scholars identify with Barkur and Mangaluru.
Cosmas Indicopleustes, a Greek monk, writes of a port called Mangarouth.
Arrian calls Mangaluru Mandegora.

These are four spellings from four authors across several centuries, and no local inscription confirms any of them. They show that the coast was known to the trading world, not exactly which town each name meant.

A Greek play with Indian dialogue

A papyrus found in Egypt, Papyrus Oxyrhynchus 413, holds a Greek comedy of the 2nd century CE called the Charition Mime. Part of its dialogue is in an Indian language. E. Hultzsch proposed that it was an ancient form of Kannada, and B. A. Saletore and a scholar named Shastri also leaned that way. Shivaprasad Rai (1985) and U. Padmanabha Upadhyaya (1996) argued for Tulu, and Manohar Laxman Varadpande placed the play’s kingdom at Malpe, near Udupi. T. Z. Mani instead found mostly Sanskrit and early Malayalam words.⁴

The argument itself is the lesson. More than a century of work has not settled whether this is the earliest trace of Kannada, of Tulu, or of neither.

What Ashoka’s edicts say

Ashoka’s edicts, from the 3rd century BCE, name several peoples of the south. Wikipedia notes that several scholars identify the Satiyaputras he mentions with this region.¹ That identification is a scholarly guess. The edicts do not name Tulu by name.

Two claims from the epics

Local tradition says that Sahadeva, the youngest Pandava, governed this place in the days of the Mahabharata, and that Rama was lord of Kanara in the days of the Ramayana. Wikipedia reports both, and only as “according to Hindu mythology”.⁵ We have not found either claim in the epics’ own text. Read them as the kind of move seen on the Parashurama page, where an older local story is woven into a pan-Indian one, and not as history.

A related idea comes from the scholar M. Govinda Pai, who identified the kingdom of Harita in the Harivamsha with Dakshina Kannada and linked the place-name Mudugara to Moger, part of the title of the district’s fishermen.¹ It is a name-similarity argument across a very long gap.

Where the documents begin

Wikipedia’s article on the Alupas calls the Halmidi inscription, of about 450 CE, the first clear mention of the Alupas, naming a possible early ruler, Pashupathi of the Alapa (Alupa) gana.⁶ Its separate article on the inscription does not mention the Alupas and gives dates from about 425 to 500 CE, so treat both the date and the reading as open.⁷ Everything on this page comes before that, and rests on poems, outsiders’ names and arguments over words, which is a different kind of evidence from an inscription.

Where Tulu Came From, the language side of the same deep past
Megalithic Tulunadu, the physical traces of people here
Bermer, a legend of a first lord

Sources & further reading

The books and articles behind the numbered notes on this page.

Wikipedia, “Dakshina Kannada” (accessed 7 October 2026): Mamulanar, the Satiyaputras, Ptolemy’s Barace and Maganur, the Harivamsha and Govinda Pai, and the Charition Mime.
Wikipedia, “Mushika dynasty” (accessed 7 October 2026): Nannan, Ezhimala, the Chera war and the note on several rulers of that name.
Wikipedia, “History of Mangalore” (accessed 7 October 2026): Pliny, Ptolemy, Cosmas and Arrian.
Wikipedia, “Charition mime” (accessed 7 October 2026): the papyrus, its date and the Kannada, Tulu and Malayalam readings.
Wikipedia, “History of Mangalore” (as above): the Sahadeva and Rama claims, framed as mythology.
Wikipedia, “Alupas” (accessed 7 October 2026): the Halmidi inscription as the first clear mention of the Alupas.
Wikipedia, “Halmidi inscription” (accessed 7 October 2026): the date range and the reading of Pasupati.

All of these are Wikipedia articles, written from other sources. The research notes behind this page also cite Ravindra Mundkur’s blog on Nanda kings and Kongas, Sh. Shettar’s book on Sangam Tamilakam and the Kannada land, and Murkot Ramunny’s book on Ezhimala, which we could not read. The Konga tribe and the Bunt-Nadava lineage-name theory are left out for want of a source we could check.