In 1913, reading Old English at Oxford, twenty-one-year-old J. R. R. Tolkien met two lines in the Exeter Book poem Crist, associated with Cynewulf:

Eala Earendel engla beorhtast / ofer middangeard monnum sended.
Hail Earendel, brightest of angels, over middle-earth sent unto men.

He later described a thrill, as if something old had moved. Dating the poem is a scholarly argument (often placed in the ninth century rather than a tidy “eighth”). The useful fact is the question he asked: not only what the glossators thought Earendel was, but who might be sailing. In September 1914 he wrote a poem about the voyager Éarendel. The legendarium starts as a linguistic itch, not as a map committee.

Lord Dunsany had already shown that invented pantheons could be literature. Tolkien’s difference was professional. He was a philologist: Professor of Anglo-Saxon, later Merton Professor of English Language and Literature. Languages, for him, were not labels. They were worldviews with sound-laws.

Grammar before plot

Quenya, the High-elven tongue, takes phonological and grammatical cues from Finnish: vowel harmony, a taste for case, the ceremonial cadence Tolkien loved in the Kalevala. Sindarin, the Grey-elven speech of Middle-earth’s everyday Elves, leans on the mutations and liquids of Welsh. Khuzdul, the guarded Dwarvish, is built with a feel for Semitic root patterns: heavy consonants, a language that sounds as if it were cut for secrecy. These are inspirations, not impersonations. A Finnish speaker does not “hear Elvish” and a Welsh speaker is not being quoted. The point is that Rivendell / Imladris and Gondor (“stone-land”) behave like names that have been worn by use, not like syllables drawn from a bag.

He did not, in his own account, invent the tongues to decorate a war. He wrote the war because the tongues needed speakers and the speakers needed graves.

Sediment, and the courage of losing

The legendarium is the decades of myth, annal, and revision that Christopher Tolkien later edited into The Silmarillion and the History of Middle-earth volumes. When Aragorn sings of Lúthien at Weathertop, the lay is not a decorative inset. It is a story already ancient inside the story. Third Age characters do not own a complete encyclopedia. They have songs, ruins, and wrong etymologies. That incompleteness is why the place feels independent of the narrator.

Place-names stack epochs: Bree’s short Celtic-looking shape, the Shire’s Old English farm vocabulary, Rohan’s Mercian-flavored speech, Gondor’s Elvish courtliness. You can feel the layers without needing a cartographic treatise. (Imaginary maps as a subject have their own essays; this one is about why a name can be older than a plot.)

In the 1936 British Academy lecture Beowulf: The Monsters and the Critics, Tolkien defended the poem’s monsters and described a Northern courage that does not require a guaranteed win. Ragnarök is still the horizon. Galadriel, in Lothlórien, speaks of fighting “the long defeat”: ages of resistance in a world that does not stay saved. At the Grey Havens it is Gandalf, not Galadriel, who says, “I will not say: do not weep; for not all tears are an evil.” Frodo has saved a Shire he cannot comfortably inhabit. The Elves leave. The Fourth Age is quieter. That melancholy is not a failure of the eucatastrophe. It is the bill.

C. S. Lewis’s Narnia enters through a wardrobe. Ursula K. Le Guin’s Earthsea makes naming a magic of true words. Tolkien’s wager is that if the grammar is deep enough, the reader will forgive a hobbit a surprising amount of walking. When a later fantasy dumps an apostrophe into a proper noun, ask whether anyone in that world could have inherited the name, mispronounced it, and still meant a river that was there before the chapter heading.

↑ Back to Top