Why etymologies matter - by Benjamin Breen - Res Obscura

SubscribeSign in

Why etymologies matter<br>How tracing words can illuminate history

Benjamin Breen<br>Apr 30, 2024

32

Share

“Little bit you speak Farsi?” my two-year-old daughter Yara asked me yesterday.<br>My wife and I are raising Yara and her little sister in a bilingual household with both Farsi (Persian) and English. Although I’m the English-speaker 99 percent of the time, I have indeed been making an effort to learn Farsi. As a historian, I find it an especially fun language because there are so many hints of past events embedded in the vocabulary itself.<br>A favorite example: there is a century-old debate among historical linguists that hinges on the Farsi word div, which describes a type of demon from Persian mythology (you can see two examples below). Farsi is an Indo-European language, and div is easily traceable back to an ancient Indo-European word for gods or divinities. The original word, used by proto-Indo-European speakers who lived around 5000 BCE, sounded something like deywós and likely referred to a sun god. Today, it’s the origin of Spanish dios, Latvian daievas, English deity, and Hindi deva, all of which mean “god.”

Ali Quli, Two Div Following the Order of King Solomon, detail of an illustration from a 16th century Mughal manuscript at the Walters Art Museum (Walters W.624)<br>So if all these words that are related to div mean “god”… why does the Farsi word mean “demon”?<br>The answer lies in the ancient sister-tongues of Avestan and Old Persian — the languages of the Iranian peoples circa 1000 to 500 BCE. Both of these languages feature the word daeva. The Wikipedia page for this term is a decent explainer for what may be going on here (and here’s a more scholarly article if you want to go into the weeds). It would appear that among ancient Zoroastrians, the daeva, including gods like Indra, came to be regarded as an older category of “gods to be rejected” in favor of Zoroastrian deities like Ahura Mazda.<br>Over centuries, the daeva were demoted from gods to demons. By the time of the image above, which was made in the 16th century, the div had become something like an evil djinn or the monsters of Greek legend. Meanwhile, the speakers of the Indian branch of the Indo-Iranian languages (like Sanskrit and its descendants) continued worshipping gods known collectively as deva.<br>Did an ancient group of Indo-Iranian speakers, in their migrations south and east in the centuries before the Vedic period, undergo a social and religious schism? Did one branch condemn the old gods, while another (the proto-Vedic) continued to worship them?<br>The honest answer to this question is: no one knows for sure. But there is evidence that the switch to a negative perception of daeva among the ancient Persians really does map on to a historical event in the very distant past, perhaps around 1500 BCE.1<br>These sorts of parallels between closely-related Indo-Iranian languages are one thing. But to a degree I find surprising, even Farsi and English — which are far more widely separated in the family tree of Indo-European languages — have some strikingly similar cognate words.<br>Here are a few:<br>Abru = eyebrow<br>Baradar = brother<br>Dokhtar = daughter<br>Garm = warm

Likewise with Farsi and Spanish:<br>Dust (friend/liking) = Spanish gusto (taste/liking).

I’m emphasizing the Farsi cognates because it’s a language I’m currently learning, but if you cast a wider net, you can find even more ancient shared words. My favorite example: lox. Yes, the word for smoked salmon.<br>This word, amazingly, appears to have remained basically unchanged for about six to eight thousand years. It is thought to derive from Proto-Indo-European *laḱs- (“salmon, trout”). The same basic sound continued to mean the same thing in:<br>Middle Dutch (lacks, lachs, lasche =“salmon”)

German (lachs = “salmon”)

Latvian (lasis = “salmon”)

Russian (лосо́сь losós = “salmon”)

As this article on the subject from Nautilus points out, “laks” even means “fish” in Tocharian B!<br>Why etymologies matter for historians

Cognates are useful not just for linguists, but for historians of religion, culture, and science and technology. They allow us to make guesses — tentative, but still evidence-based — about stories and events from before the advent of written records.<br>One example that has stuck with me since it was announced in 2016 is this phylogenetic study of fairytales conducted by the folklorist Sara Graça da Silva and the anthropologist Jamshid Tehrani.2 Da Silva and Tehrani reached the conclusion that certain legends found in story collections like Grimm’s fairy tales, Aesop’s fables, and The Arabian Nights may date to around six to seven thousand years ago, in the era when proto-Indo-European was still a living language. Among the very oldest stories they could trace back was the tale known as the Smith and the Devil: the origin for the later tale of Doctor Faustus. In it, a blacksmith — a figure associated with magic and...