The British are afraid of genetic engineering allowing monkeys to speak. (Just wait until they find out about all the little talking Great Apes in Britain that the scientists have been secretly creating in their bedrooms) SF writers are using river Otters as sources of sounds for languages. So for real life Frankenstein science and science fiction, clearly the world needs some animal conlangs. Let me emphasize that no conlang will enable Dr. Doolitle style conversations with real animals. Sorry. It won’t happen. Unless the genetic engineering bit works.
Animal conlangs are kind of like alien languages and kind of not. They basic rules of human communication may or may not apply. The basic mathematical principles of communication system still apply, though. For example, for aliens and humans, a little redundancy in a protocol is useful in noisy environments.
Aliens are as smart or as stupid as we’d like, since they’re fictional. On the otherhand, we know about real animals, the sounds they make, the upper limits of their lexicon (about 250 symbols with intense training) and the upper limits of their ability to work with grammar– sometimes they can string together two symbols that mean something more than what the symbols mean on their own. Again only with intense training and in the case of grammar, pretty much only great apes can do this half convincingly, other accounts are wildly controversial. Also animals tend to have a hard time discussing things out of sight and other things that are so simple we often overlook them. Some people seem to treat thinking skills and communication skills as if they were the same thing– is it true? Don’t know, don’t care that much, but it could influence the design of an animal conlang. An animal conlang might be written to work with some or all of these contraints, or it might not. It might just be a human communication system that generally sounds like a particular animals characteristic calls.
An animal language is not necessarily Elvish, Esperanto or Loglan (i.e. artlang, auxlang, loglan). Animals are not very skilled at logic, it’s absurd to restrict yourself animal sounds in writing a suitable lingua franca, and there is no particular reason to invent a fictional culture for an animal conlang– chimpanzees and whales have a culture already (in many sense of the word), they just don’t have inborn linguistic skills.
Source Material for Animal Conang
A typical chatty animal will have 1 to 250 signals that they typically say either once, or they repeat it over and over. And those sounds might be made of of many more parts–but as far as we can tell, only humans recombine meaningless sounds into meaninful morphemes. Unfortunate for the animal conlanger, the sounds animals make don’t always have an obvious way to break them into individual parts, and they won’t necessarily match up to IPA. If you don’t believe me, please take me up on the challenge to transcribe the various sounds of a walrus into IPA.
Some of the most promising and interesting animal communication systems don’t use phonemes. For example, ants use chemicals that they excrete. Bees use body motions. Crows may have an above average communication system, but it’s poorly studied. A key similarity among these species is that they all are group scavengers that need to co-ordinate to find and effectively take advantage of food spread out across a large area.
Misleading phonetic inventories and phonotactics
As we know, every human language has come up with a wild variety of transliterations of these sounds. So an English speaker would like start transliterating a dog’s barks as “wuf” and then generate words with “w” “u” and “f”. Using just these three sounds, you potentially could say “fwu” and “fuw” Which don’t sound at all like a dogs bark anymore.
I worked through an example with cats and came up with much the same problem. Also, if you constrain the phonotactics enough to make a catlang sound like cats meowing, then you have a lot of extremely similar sounding words.
Walruses have a remarkable repetoir of sounds but they are mostly really hard to reproduce. An animal conlang will have to make choice about being authentic and being human pronounceable. A good compromise might be a system that can be written and played on a computer.
Isolating, Small Vocabulary Systems, Non-phoneme signals
An animal conlang can expand the logotome (set of all possible words) set by including the silence between words, using tone, and other things that allow the “words” to stay much the same. Also, by adopting a toki pona style small vocabulary, one can avoid the problems with trying to identify a phonetic inventory and phonotactics system and instead using the existing animal calls, largely unchanged.
Depending on who you read, the estimates for long long humans have been communicating go back at least 40,000 years, may 80,000 years because we’ve been leaving signs of sophisticated culture since then. We’ve been making stone tools and cooking food with fire for up to a few million years. The protohuman conlang is kind of like the animal conlang in that it might have some of the features of an animal communication system, on the other hand, its supposed to exhibit some features of modern languages.
Protohuman, likewise is not esperanto, lojban, nor elvish. The culture left some clues to what it was like and typical characteristics in modern hunter and gather societies leave clues to how humans lived up to 80,000 ago. I’ve read convincing accounts by archeologists that the Australian aboriginal lifestyle has been unchanged for that long. My point is that the protohuman conlang is not all fanciful fiction and pretty sounds. Obviously protohuman would be a lousy lingua franca and cavemen are not known for their ability to do symbolic manipulation or logic proofs.
Experimental Linguistic Archaeology
One interesting question is “Did human speech arrive in a bang, or were there intermediate communication systems in between?” I think that if we could make up communication systems, we could demonstrate a set of constructed languages that span animal communication systems and full human speech and illustrate that such a cline was possible. Or it might show that after a certain set of features are added to a language, then bam!, it becomes the same kind of flexible, general purpose communication system that we all learned from Mom.