When two languages strike us as “similar,” we usually notice it through the sound of their words. Rask put up a firm stop sign in front of that intuition. Word-similarity, he argued, is simply not a reliable measure of how two languages are actually related.
Words travel remarkably easily
Rask’s reasoning is straightforward. When peoples come into contact, words cross the boundary between their languages with ease. Danish words entered Greenlandic; Portuguese and Spanish words entered Malay and Tagalog — this kind of vocabulary traffic happens all the time, even between languages with no shared origin or lineage whatsoever. Which means a large number of similar-sounding words between two languages is, by itself, no evidence that they share a common ancestor.
Grammar almost never travels
Grammar, Rask argues, is a different matter entirely.
Den grammatikalske Overensstemmelse er et langt sikrere Tegn paa Slægtskab eller Grundenhed; thi man vil finde at et Sprog, som blandes med et andet, yderst sjelden eller aldrig optager Formforandringer eller Bøjninger af dette, men omvendt taber snarere sine egne. (Grammatical agreement is a far more certain sign of kinship, or fundamental unity — for one finds that a language, when it mixes with another, extremely rarely or never adopts that other language’s inflectional forms, but rather, conversely, tends to lose its own.)
English never adopted Icelandic or French inflections, yet lost most of the inflections Old English (Anglo-Saxon) once had. Danish never adopted German endings; Spanish never adopted Gothic or Arabic ones. Vocabulary can be borrowed, but the machinery of grammar itself isn’t something you borrow from elsewhere — it only wears down over time. That’s Rask’s observation.
A rule of thumb: more complex grammar means older
From here, Rask draws one more practical rule of thumb. The language with the most elaborate, most intricate grammar is the least mixed, the oldest, and the closest to the original form. Grammatical inflections wear away every time a new language emerges from an old one, and building them back up from scratch takes a very long time. So Danish is simpler than Icelandic, and English is simpler than Old English. The same relationship, Rask points out, holds between Modern Greek and Classical Greek, Italian and Latin, German and Gothic. Simplicity is evidence of newness; complexity is evidence of age — and this yardstick became one of the founding assumptions of comparative linguistics as a whole.
Summary
| Point | Content |
|---|---|
| The popular standard | Judge language kinship by how similar the words sound |
| Rask’s counter | Vocabulary is easily borrowed between languages, making it weak evidence |
| The standard to trust instead | Grammatical endings — almost never borrowed, only worn away |
| The rule of thumb this yields | A more complex grammar means an older, less mixed language |
In a nutshell — before reaching for similar-sounding words, look at the grammar first. That modest-sounding piece of advice became the starting point of the discipline of comparative linguistics.