[Orwn Acra]

Originally Posted by Michael Ferris

Walter,
Don’t you think portmanteau words are ‘second order’ words that derive their meaning from ‘first order’ words, and, absent onomatopoeia and such, are still arbitrarily related to the signified? Seems to me perhaps they are.

Certainly many, probably most, words are arbitrarily related to what they signify. However, I would say that the influence of onomatopoeia is greatly underplayed. One wonders why certain sounds for certain words caught on while others did not. It is no coincidence that in synthetic (agglutinative) languages common words are the shortest. While this is not exactly onomatopoeia (is, for example, does not imitate any sound), the sense and utility shape the sound. Can you imagine having to say "obladobladoblado" when you mean "no"? Our language generally evolves into what is most easy.

Also, yes, portmanteaux are second order words, which is why I am more impressed by vorpal or callooh: non-portmanteau words that convey their meaning through sound.

At university, I was probably alone in thinking that the division between sound and sense, which is the division between sign and what is signified, is not nearly as strong as generally believed. Perhaps I was and continue to be overly poetic. Saussure's response to onomatopoeia was that because different languages imitate sounds differently, the link between sound and sense is weak. His argument, however, doesn't take into account that animal sounds in other languages are remarkably similar and that any variation can be accounted for by each language's unique phonological properties.