Number of words: 900

We can see the ingredients all around us, in the chemical signals that flicker on and off between cells, the waggle dance of bees, territorial calls of various creatures, and the remarkable repertoire of birdsong. There are the complex patterns of moans, cries, and chirps in the songs of whales. There are ring-tailed lemurs that waft their tails around to speak a complex “language” of multiple scents to communicate aggression, receptiveness to mating, and much more. Let’s not forget those extraordinary intellects of the marine world, the jetpowered cephalopods, whose rapid changes in skin color and pattern are the basis of courtship rituals. And then there are our closer relatives, such as the chimpanzees.

Any way you look at it, the languages used in the rest of the animal kingdom fall short of our abilities. Most animals use nonsyntactic communication, so that a single grunt—a “word”—might be used to express a scenario, such as “watch out—there’s a lion prowling nearby.” Putty-nosed monkeys (Cercopithecus nictitans) in West Africa use two main sounds—pyows and hacks—to warn each other about predators. Of course, a language does not even have to be built on a foundation of sounds. Scouts from beehives perform dances inside the nest to tell their peers where nectar can be found. The coordinates of distant locations are encoded in the waggle phase of this buzzing ballet, with the direction and distance to the food source indicated by the orientation and duration of the dance.

Leaving aside the tricky issue of grammar, which I will return to later, these kinds of considerations allowed me to comprehend that the size of the repertoire of signals we use—the lexicon—is a critical issue. When it comes to our repertoire, a six-year-old has a lexicon of about 13,000 words. The rate of word learning in humans comes to about one word every ninety waking minutes from age one to age seventeen. This leaves a seventeen-year-old native English speaker with about 50,000 words stored in her mental lexicon, which is the typical value for a grown person. Absorbing this vocabulary is a colossal task, similar to learning 50,000 telephone numbers with all kinds of associations. We don’t realize what wonderful memory machines we are. But memory is not the whole story. The human vocal tract can make a variety of different sounds. Based on the six thousand languages that we know of, our tract can make around 1,000 linguistic sounds. We call these units phonemes and the total inventory in languages varies from as few as 11 in Rotokas, which is spoken in Bougainville, an island to the east of New Guinea, to as many as 112 in !Xóõ, a “click language” of Africa, notably Botswana and Namibia. These may range from tones to familiar sounds to clicks, which can only be written by using the sort of characters you usually press by accident on a computer.

 Another idea naturally follows from this physiological fact. In speaking this vast lexicon, we make mistakes. Producing the sounds—the phonemes—in an ordinary conversation is a remarkable anatomical feat. The motions of various parts of our vocal tract are coordinated within millimetre’s and timed within hundredths of a second. The next time you open your mouth to speak, just remember how breathtakingly awesome you are. But errors are inevitable; and, in essence, the more concepts that have to be communicated, the more sounds that are needed to do this, the closer the sounds will be to each other, and the greater the risk that this repertoire will become confused.

 Thus there is a limit to the number of phonemes that we can handle with our vocal apparatus. If we used the simplest kind of language, where one phoneme is linked to one action, object, person, or whatever, we would quickly run into difficulties as we developed more associations. Think about the potential for confusion and disappointment if only slightly different sounds are used to distinguish the discovery of a cache of ripe bananas from a fly-eaten pile of rotten black ones. There is a wonderful Gary Larson cartoon that shows two apes dancing a tango with each other. The caption reads: “I’m afraid you misunderstood… I said I’d like a mango.”

To capture the physiological reality of speaking a language, I proposed the idea of a “linguistic error limit”: the number of distinguishable sounds in a protolanguage, and therefore the number of objects that can be accurately described by this language, is limited. Adding new sounds increases the number of objects that can be described, but this flexibility comes at the cost of an increased probability of making mistakes; the overall ability to transfer information does not improve.
The error limit that we calculated suggests a reason why the overall number of phonemes we use in any one language falls far short of the possible range of around 1,000 sounds that can be made by the vocal tract. If there are too many, there is too much potential for confusion. Better to concentrate on a few phonemes and learn how to distinguish them effectively. Equally, once we have learned one set, it can be hard to change to another. That is why foreign accents, like my Viennese take on English, are typically caused by the use of the wrong phonemes.

Excerpted from page 180-182  of ‘Super co-operators ’ by Martin Nowak