Language for the Birds

In world of Middle-earth, various species have their own languages, though they are not often discussed. One that Tolkien does mention explicitly is “thrush-language.” In Chapter 12 of The Hobbit, as Bilbo and the Dwarves were planning on the side of the Lonely Mountain, an old thrush kept listening to their discussions. Bilbo tried to drive it away by throwing rocks at it, and Thorin rebuked him for it. He explained to Bilbo that there was an ancient breed of thrushes that lived in the area. The Men of Dale could understand the speech of these birds, and used the birds as messengers. This small thrush that initially annoyed Bilbo would go on to deliver a message to Bard on how to kill Smaug.

In the real world, there is no evidence (yet) that thrushes have any kind of language. But there is some evidence of language in the Japanese tit.

evidence for tit-language

An ornithologist, Dr. Toshitaka Suzuki, has painstakingly documented the calls of the Japanese tit over decades, and has collected by observational and experimental evidence that tits have words; and he has identified one potential two “word” phrase that is constructed by some kind of grammar. So, while there may be not be a thrush-language outside of Tolkien’s fiction, there may be a tit-language.

What Dr. Suzuki has been able to demonstrate is that particular calls are associated with particular “meanings” (in some broad sense). Tits make different calls when they see a hawk than they do when they see a snake, for example. These meanings have been carefully demonstrated in two directions. In one experiment, fake snakes and fake hawks were introduced near nests, eliciting particular calls. This shows that the birds make the calls to announce the presence or particular animals. The other experiment involves playing calls on a speaker and observing the birds’ reactions. Snakes and hawks have different attack patterns, and so tits have different escape patterns. When the researcher plays a particular call, the birds react according to whether they are expecting a snake attack or hawk attack. This shows that the birds understand the meaning of the calls they hear. So, they intentionally produce calls and appropriately interpret them.

Strictly speaking, this is not necessarily evidence of language. Humans can intentionally create and interpret all kinds of sounds that are not part of our language. In some cultures it is polite to belch after a good meal. A belch from a guest is a signal that the food was appreciated, and the host can interpret that belch appropriately. But a belch is not a word. Likewise, we cannot say with absolute certainty that when Japanese tits make a particular call that they are “saying” something; just that they are “communicating” something.

One of the calls that Dr. Suzuki has documented seems to mean something like “gather” or “come here”. When this call is made, tits gather to the area of the caller. It seems to be a kind of summons. This call can also be combined with the call for “snake” to form a two word sentence: “snake gather”. When these two calls are made in that order, tits will gather in the area and began looking for a snake. If the calls are edited to occur in the reverse order (“gather snake”), the birds do not respond to it. They show no evidence of comprehension. This looks like it may be some kind of grammar, some kind of call ordering rule. Thus, it may be evidence that the Japanese tit has a Language.

Now, let’s be careful. Writing it down as two words makes it look like a sentence, but it may not be. It could be a “compound word”, snake-gather. There would be a general idea of “gather”-ing, and then a more specific idea of “snake-gather”-ing. Compounding can be taken as evidence of grammar in a general sense; but it would not be evidence of sentence formation. And indeed, when Dr. Suzuki tried novel expressions like “hawk gather”, there was no reaction from the tits. If there was some kind of sentence formation involved, you might expect “hawk gather” to generate a response; but if there simply is no word “hawk-gather”, then the tits would just be hearing gibberish. On the other hand, perhaps Japanese tits just don’t gather to look for hawks, so they are ignoring an understandable but absurd statement. We can’t really know.

Conlanging for the birds?

Let’s go back to the thrush-language in The Hobbit. Tolkien does not give us any indication of what it was like, just that Bard was able to understand it. The thrush’s words are all rendered in English.

“Wait” Wait!” it said to him. “The moon is rising. Look for the hollow of the left breast as he flies and turns above you!” And while Bard paused in wonder it told him of tidings up in the Mountain and of all that it had heard.

Chapter 14, “Fire and Water”

If you following strictly what Tolkien wrote, thrush-language must be fairly sophisticated. The longest sentence in the quote is syntactically quite complex, having two clauses with an explicitly mentioned temporal relationship. Explaining what he observed in the mountain must have also been sophisticated, and the description implies the existence reported speech (and not just direct, personal knowledge).

Or you could say that Tolkien’s translation is an attempt to render quite alien concepts into English.

How would you design thrush-language? Would you make it like a human language, so that you could directly translate the words Tolkien ascribed to the thrush? Or would you try to design something much more loosely translated?

In a future post, I plan to sketch out a very simple thrush-language.