Who were the Keltoi? In early Greek writing, the name referred to peoples in continental Europe, including regions associated with the western Mediterranean and the Danube. It did not yet mean precisely what “Celtic” means at a gathering of Scottish, Irish, Welsh or Breton communities today. Following the word through the sources reveals a larger, more interesting history than a single tribe marching unchanged across a map.
There are real connections to explore. The trick is to ask whose name we are reading, when they used it, and what their evidence can tell us.
A Greek name at the edge of the known world
Hecataeus of Miletus and Herodotus of Halicarnassus are the two earliest Greek writers identified in Patrick Sims-Williams’s study of the name. Both came from Asia Minor, far from the western lands they described. Philip Freeman’s collection of early Greek evidence brings together original texts, translations and commentary: the surviving record offers brief encounters and fragments, rather than a continuous history told by the people being named. Sims-Williams, 2016, introduction; Freeman, 1996, introduction.
That distance matters, but it does not make every ancient description empty. Sims-Williams points out that Hecataeus and Herodotus distinguish Celts from particular neighbors: Ligurians in the former, Cynesians in the latter. He challenges the idea that Keltoi was simply a label for everyone west of Greece. A broad name could still identify particular people.
Why the Celts appear in a discussion of the Nile
One of the most revealing passages occurs in an unexpected place. In Histories 2.33, Herodotus is reasoning about the Nile. To explain a river whose source he does not know, he compares it with the Ister, the river we call the Danube.
He places the Ister’s beginning among the Keltoi and at a place called Pyrene. He also locates the Keltoi beyond the Pillars of Heracles and beside the Cynesians, whom he places at Europe’s far western edge. Read the passage and you can watch him assembling a picture of an enormous continent from limited information. Herodotus, Histories 2.33, translated by G. C. Macaulay.
This establishes that Herodotus knew and used the name. It does not establish an exact modern border, a securely identified location for Pyrene, or a migration route to Scotland. Those are different questions, requiring further evidence. The passage is exciting precisely because it preserves both knowledge and the effort to make sense of its gaps.
Caesar adds a self-name—and a warning about scale
Move forward to the first century BCE. At the opening of The Gallic War, Julius Caesar divides Gaul into three parts. One is inhabited by the Belgae, another by the Aquitani, and the third by people he says are called Celtae in their own language and Galli in Latin. He immediately adds that the three differ in language, customs and laws. Caesar, Gallic War 1.1, Latin text and commentary.
That is an important complication: we should not say that every use of “Celt” was purely an outsider’s invention. Caesar reports a local name. Equally, his statement concerns a particular division of Gaul. It does not show that everyone now included in Celtic history used one shared name for themselves.
The Greek geographer’s category and the Roman commander’s account deserve to be read closely, each in its own setting.
How languages widened the story
The modern connection with Scotland, Ireland and Wales becomes clearer when we turn to language study. Museum Wales identifies Edward Lhuyd’s pioneering work of 1707 as an important moment in recognizing relationships among Celtic languages. The National Library of Wales confirms the publication that year of the first volume of his Archaeologia Britannica, based on travels through Wales, Cornwall, Brittany, Ireland and Scotland. Museum Wales, “Who were the Celts?”; National Library of Wales, Lhuyd and his Cornish dictionary.
Today, Scottish Gaelic, Irish and Manx belong to the Goidelic branch; Welsh, Cornish and Breton belong to the Brittonic, also called Brythonic, branch. The Open University explains these relationships in its account of Insular Celtic. “Celtic” here describes related languages across different places and histories. Open University, “The insular Celts”.
This gives us a well-founded connection between Scottish Gaelic and Irish without making the two languages interchangeable. It also helps explain how a word encountered in ancient continental accounts became an umbrella for communities whose own histories need individual attention.
Names, objects and ancestry answer different questions
An ancient name, a language relationship and an archaeological style are different kinds of evidence. Museum Wales describes how the spread of La Tène art was once commonly treated as a map of invading Celts. It explains why the presence of that art can also reflect exchange or the adoption of a style, and why regional differences matter.
We therefore cannot make every object carry a whole people’s biography. Nor does a shared label, by itself, establish a biological ancestry. This article makes no DNA claim and offers no single answer to when or how Celtic languages first spread.
The same care applies to inherited origin stories. Our article on the Declaration of Arbroath’s Scythian claim separates what a medieval document says from what would be needed to establish its migration account.
A wider heritage, with room for distinct voices
At Tallahassee Highland Games, our Scottish core sits within a wider Celtic welcome. Learning the differences gives that welcome depth: a name has a history, a language has its own voice, and a tradition has people who carry it forward.
You can take pride in that inheritance while asking good questions about it. Start with one passage, one song or one conversation, and follow the connection carefully. Explore why Highland Games still matter as living culture and community, and bring your curiosity with you.