This originally appeared in my trip report from the Canary Islands. And another set of arguments against whistled versions of languages being considered separate languages, this one regarding "Bird Language", a whistled version of Turkish.
Why Silbo is Not a Language
Silbo, the whistled "language" of La Gomera, is clearly a remarkable part of the Canaries' unique heritage, one that I hope continues to exist far into the future. You can listen to examples here, here and here. I emphasize this because my argument that it is not a language is not an argument to diminish its significance in this regard. But language is a strong interst of mine - I almost went back to school to study linguistics - and Silbo is not a language, and I've carefully referred to it as a "communication system" for this reason. (If you run across this blog post and you have counterarguments or evidence to the contrary, I'm not a Silbo expert, so by all means educate me!) Canarians are of course proud of their heritage and it's difficult to receive criticism of a belief about something so central to cultural identity as language, and we've encountered this problem right here in the U.S. (see #2). Here is why Silbo is not a language.
1. Silbo dates to the days of the Guanches. Guanche was a distinct language, unrelated to Spanish. As a Berber language, Guanche likely had a very restricted set of vowels, uniquely suiting it to a whistled form which by some estimates has only two vowels (as linguist Ramon Trujillo argued in his 1978 book.) Yet the modern version of Silbo is clearly whistled Spanish. The ability to translate Spanish (or any language) into a whistle, and have someone understand any of it, is remarkable. But this shows that Silbo is a manner of intonation of an already-understood language, not its own language. (If you object: can someone from Germany learn Silbo, but not learn Spanish? If two German-speakers in Switzerland start whistling German to each other across the valleys, is that Silbo, and could someone from La Gomera understand them? The answers are no and no. It's not its own language.) I would like to see sets of "blinded" participants communicating to see what the lossiness of Silbo is - I predict much more than actual language. Normal spoken language is about 50% redundant.
2. Silbo is used in very limited circumstances - to call long distances across valleys (for which it's a brilliant innovation), and to discuss very limited topics of obvious immediate import (where are my sheep, is it going to rain, etc.) No ritual oration has ever been given in Silbo, no political discussion, no business deal (unless it was a very simple one across a valley involving something immediately present like sheep. Note that in one of the links above they're talking about red wine, etc.; this is a tourist performance, and no one is actually communicating.) True languages are used constantly, in every social setting. You have to work hard not to use them, and children exposed to them start developing language at a young age, without any special effort to teach them. I would be shocked to learn that children of four are whistling without being specifically taught, and I also have not read any traveler, ancient or otherwise, reporting the ear-splitting qualities of conversation in a village of La Gomera - because people speak. An example of a language-like entity that has become culturally important here in the U.S. is what people call Chinook, but is really Chinook Jargon. It's a trade pidgin, a mish-mash of true Chinook (a real language), English, and some Russian, Spanish and French. It has little to no morphosyntactic structure, is not learned by children, and is not spoken outside of trade settings. Nobody goes home and speaks Chinook Jargon to their wives or husbands, they speak Chinook, or English, etc. and that's what the children learn. The children may grow up and learn Chinook Jargon - as adolescents or adults, by deliberate effort. So as with Silbo, Chinook Jargon is a fascinating and important part of the Pacific Northwest's history and heritage, and I wish its enthusiasts well - but it's not a language.
3. Finally, "non-standard" phonemes (like whistles) can certainly be part of languages - but they're never the whole language. In fact, possibly the strongest argument here is that there are languages with clicks, and nasal vs oral or even whispered vowels, and (probably!) even whistled vowels, but no language uses entirely one or the other. So if Silbo is a language, then two prisoners whispering in English between their cells have invented a new language called "Whisper", as long as the other prisoners catch on and use it. (Silbo was successful because it carries over long distances; it's kind of the anti-whisper.)