by Anatoly Liberman
This is the promised continuation of the previous post. As I said last week, break is an old word. In the foggy days of Proto-Indo-European, it may have begun with the consonant bh (or simply bh), pronounced as in Modern Engl. abhor or Rob Hanson. For our purposes, the difference between b and bh matters not at all, because today we are only interested in observing how many words referring to breaking begin with br-. The subject of this essay is: “To what extent is classifying break with sound imitative (onomatopoeic) words justified?”
Moo, meow, and, admittedly, oink-oink are sound imitative. But once we leave the animal world and exclamations like phew and whew, assigning words to onomatopoeia is always problematic. Thus, each member of the set—crack, crash, crush, creak, croak, and cry—looks like any other word beginning with kr-, for example, craft, crawl, and creep, but in their entirety they produce the impression of belonging together and suggest a rather obvious sound effect. The same holds for initial gr-: compare groan, growl, grumble, grunt, and possibly grind. Numerous words signifying grouchy people, as well as grim and gruesome things, also begin with gr-. Labeling them sound imitative will not take us too far, since they have well-developed bodies and not only heads.
Modern scholars have no idea how language originated (or rather they have many ideas that cancel one another out; the usual cliché is “shrouded in mystery”), but in the existing languages words are conventional signs, that is, when we look at a word, we usually do not know its referent in the world of things. If we possessed such knowledge, explanatory and bilingual dictionaries would not be necessary: anybody would be able to look at a “sign” like bed or ten, or give and guess what it means. Moo is fine (presumably, no dictionary is required for translating it), but oink-oink is obscure: perhaps it is a soothing exclamation like tut-tut, a verb like pooh-pooh, a noun like tomtom (a drum), or the name of a disease like beriberi. I am not sure that pigs go oink-oink, and anybody can notice that the canine language is represented by several dialects: compare bow-wow, barf-barf, and yap-yap. And yet, crash, crush, crack, and so forth rather obviously have something to do with onomatopoeia. People seem to have begun with the sound imitative complex kr and added a syllable, to make the words pronounceable. This may be a bit of a stretch, but the etymological principle behind my statement is (if a pun will be allowed) sound.
Old English had the verb breotan, which also meant “break.” It has been lost, except for its cognate brittle “liable to break, fragile.” Breotan lacks attested cognates, and its origin is unknown. But the fact remains that, like break, which had many cognates, it also begins with br-. Though today burst contains a well-formed group bur-, its most ancient form was brestan (in such groups vowels and consonants often play leapfrog—this process is called metathesis: compare Engl. burn and German brennen; the German verb has preserved a more ancient stage: the original form was brannjan; the Old English for run was rinnan alternating with iernan