R for Round

R for Round

By Mehmet Kurtkaya

I had mentioned in my initial article about the article published in 2016 by Blasi et al. Researchers found that across nearly 4,000 languages worldwide, some words with the same meaning often share similar sounds, even when the languages are completely unrelated. One of these sounds from the study is R, and it has been found that across many unrelated world languages, the word for round contains an R.

Yes, in English too, obviously. But that's not all. I have observed another interesting phenomenon. In English, many words that convey a meaning of roundness, whether nouns, verbs, or adverbs, also have an R.

Take a moment to try it yourself: or ask an AI for such words, or try an online thesaurus for synonyms. AI might list some rarely used words that don't have an R in them, but even then, the number of words with an R is too high to be a coincidence. In fact, the R is robustly overrepresented, which rules out any coincidence.

Take a moment to think about it. Now, here's what I have found: Round, Circle, Arc, Arch, Orbit, Around, Ring, Curl, Curve, Crown, Corona, Turn, Tour, Tower, Roll, Rotate.

Some more, maybe slightly less likely: twirl, whirl, revolve, swerve, bore, sphere, stir, reel.

Some of these trace back to a smaller set of Latin, Germanic and Proto-Indo-European (PIE) roots. This does not weaken the pattern, it pushes it deeper. Those roots already contained R in round-related meanings. The question becomes: why did the ancestral languages themselves favor R in this semantic field, rather than some other consonant?

Related: the sound re- that is added to words to mark doing something again, a return to the original point—just like you would do in a circle.

More speculative: the word rear, as you would need to turn around to see what's in the rear.

Wheel and cycle have an L instead of an R, but this does not change the overwhelming existence of "r".

Interestingly, even in ancient languages, the sound R appears in words for 'wheel' or 'circle.' For example, Sanskrit cakra (चक्र) and Hittite ḫūrkis both contain an R-like consonant and refer to a wheel. However, linguists do not consider these words to be derived from the same Proto-Indo-European root. Cakra comes from *kʷékʷlo-, meaning 'wheel,'. However, ḫūrkis appears to have a different origin, possibly borrowed or inherited from a non-PIE source. Despite their unrelated etymologies, the presence of R in both ancient words aligns with the broader cross-linguistic pattern where roundness-related concepts often feature this sound.

In Sumerian, the oldest written language, the word for "to turn" and "circle" is gur, a word that also contains an "r". This connection is somewhat expected; for the "r" sound to be associated with the concept of "roundness" across so many languages, the association must have a history stretching back thousands of years.

So English not only has an R in "round" but also in many conceptually related words. This means that the English language has made ample use of the sound R, and this pattern should go back many thousands of years ago.

This is not only true for English, but also for French, Russian, and Turkish, the languages that I have checked all have this particular phenomenon to varying degrees. Turkish çevirmek ("to turn") and Japanese marui ("round"), for example. The pattern holds even in languages with very different structures.

So we can safely say that not only did people use r in words for round, but also in many conceptually related words across many languages.

This means that the use of the sound R for concepts of roundness goes back many thousands of years.

Why R?

R is an invented sound.

The sound is physically enacted through a curved or oscillating tongue gesture. In a trilled R, the tongue tip vibrates rapidly against the roof of the mouth, literally a micro-rotation. In the English approximant, the tongue bunches and curves in the oral cavity. The articulator performs roundness while naming it.

The sound itself is not a basic anatomical default. It requires the tongue to curve or vibrate in a way that physically enacts roundness. But an invented sound does not travel everywhere on its own. Its wide distribution across unrelated languages suggests it was carried by ancient migrations and areal contact, then stabilized in each new setting because the phonetic mapping was naturally learnable. The anatomy provided the fit; the migrations provided the reach. It is implausible that the same invention was made independently by so many languages in the same way; migration and contact are the natural explanation.

Archaeology and archaeogenetics now clearly tell us that humans migrated to many corners of the world in the last 50,000 years, and now we can map many of these major migrations. I am not saying this pattern started 50,000 years ago; I'm saying that many major migrations happened during this period. Dating the sound R-to-round correspondence is speculative, but I have enough reason to believe it is older than 10,000 years. More on that later, as dating when this pattern began is more speculative than compared to the reality of the R-to-round correspondence. If we can find a way to connect this sound R to a specific group of people, we would be able to construct a part of history. We have ancient written records, statistical linguistics, and traditional linguistics to help.

And remember, different people have different ways to pronounce the R's, for example, French with a guttural, Spanish with a trill.

It does not matter that we cannot unravel much with one sound in one place, but if we search across languages in time and place and fit them in the flow of history with the tools I have just discussed, we would be able to achieve a good reconstruction of history. And when a piece of the puzzle does not fit into the grand puzzle of human history, we can go back to it and question our hypothesis, like all scientists do.