Mind your language.

Communication is a deceptively simple concept. Google-Define will tell you it means "the activity of communicating; the activity of conveying information", but to what extent is this actually true? Are we communicating when we sneeze? We still convey information to those around us (that we are perhaps in a state of ill health) but it is unintentional. Similarly, when we 'leak' information through emotion, are we really communicating? This question sparked a debate in my class last week, with one side arguing that communication necessitates intention; and the other arguing that if information is passed from one to another in any way it still counts.

Communication occurs amongst animals in pretty hefty amounts. Bees dance to show the direction and distance of a food source. Whale songs are filled with rhythmic complexity used for echolocation and sexual selection. Vervet monkeys have different cries to convey "AAAAAAGH EAGLE!" - "OMG LEOPARD!" and "THERE'S A BASTARD SNAKE!"  The functions of animal signals as a whole tend to work on roughly the same situations. They are used to advertise the worth of their genes during sexual selection, they warn from predators, they convey locations of food sources and express the condition of the animal. Stand on a cat's face and you're going to get a pretty loud (but very communicative) screech.

That's all dandy, but here we move onto the definition of language. It is commonly argued that humans are the only species which have actually developed their own language, although there has been limited success in teaching a few special great apes to hook onto it at an infantile level.The dictionary definition is that is is a human ability to communicate ideas, feelings and intentions using sounds or sound symbols. Hmmm, not entirely convinced.

Chomsky (1980 - and I'll say again, if you're interested in looking up these references send me a mail and I'll send them over. Tidier than sticking an APA style list at the bottom of each post), believed that human language must have grammatical rules, be used in the style I wrote a paragraph above, and that it must have a selective evolutionary advantage as is evolved after humans had separated from other primates. A very good point. As you'll see below, even our closest great ape cousins have much, much poorer language abilities. That strongly suggests that the group of ancestors that became us found some reason to develop it, whilst the ancestors that became, say, chimps, did not. What was the reason?

Charles Hocket (1960) believed there were 13 features which characterised human-like language as distinct from basic animal communication. I'll ponder the relevant features out loud in the context of language experiments on apes. Most of the features have become outdated with the introduction of the internet and mobile phone technology, allowing for advanced channels of text or symbol based interactions.

First off, Humans can talk. They don't just have a limited repertoire of sounds they can associate with important stimuli. 

Bugger knows how this works. Suffice to say our ape friend here can't blast out the lyrics to Journey songs.

Chimps physically cannot make as many different phonemes as humans can. Their oral cavity is shaped differently and their tongues cannot reach the positions necessary to make the sounds that constitute human language. The production of speech if very particular to the structures of the vocal tract. Duchin (1990) believed the sounds of 'a' and 'u' are impossible for a chimpanzee to make. Various attempts were made in the early 1900s to teach chimps to speak. The projects lasted years, and the peak of any attempt was teaching 'Vicki' to utter 'mama', 'papa', 'cup' and 'up'. I couldn't find any recording of Vicki, but here's a chimp who can say 'Mama'; not a lot of accomplishment over years of work. Chimpanzees simply cannot talk in the way we can.

http://www.youtube.com/watch?v=y4Z0xn4pYSY

So your kid can't talk? You teach him sign language. Tarrace, Petitto, Sanders and Bever (1979) taught a young Cimp called Nim 125 words in American Sign Language. Much more progress! However after studying 19,000 utterances they failed to find any consistent patterns of syntax or semantics. Video tape evidence showed that the wee Chimp simply repeated certain words in response to an initial utterance by a researcher. He'd basically just sign 'more' like crazy after being given a treat. Greedy bugger.

A more famous example was that of Washoe. Gardner and Gardner (1969) trained the chimp to learn ASL from the age of 1 year, when she should be relatively dependent on her mother. The training and living environment were structured to be as similar to that of a deaf human child as possible; whilst very strict rules determined what would be considered a proper 'sign'. Washoe was soon willing and able to use signs spontaneously for no other reason but to communicate (this is important, as many sign language attempts ended up dominated by the hunt for treats like Nim.) After 2 years Washoe could reliably use 34 signs and could transfer labels to same-category objects; by which I mean she could understand that a fedora and a top hat were both hats. She was able to meaningfully combine signs together to convey her desires more clearly "open food drink" worked for "open the fridge, wench." At the end of her training she was proficient in 132 signs and arguably had a grasp of very basic grammar.

Referring back to Hocket, Washoe had met the conditions of her communication being referential (word = object, action etc), non-iconic (the word doesn't need to resemble the object) and learned (she didn't have the ability from birth.) Beyond that, she only missed clear grammatical structure. Most scientists in this field don't consider this language, I am as of yet unsure. Her researchers wanted to see if she could pass her abilities down onto her son, but sadly he died at a very young age (there is talk that Washoe showed signs of intense mourning after signing to ask why her son had been taken away. It rips my heart out to consider.) She eventually adopted a young chap called Loulis who did begin to copy her signs with no help from the researchers. He eventually learned 22 signs from her; that's pretty special, but it has to be remembered that she only taught him the signs because they were useful around humans who knew the language. They would be of no use around other chimps, or presumably a sign language would have been adapted much earlier.

Just saw his first LolCat.

The initial attempt to crack the syntax problem was conducted by a guy called Premack in 1971 and again in 1976. His name always gives me the impression that he has come just before Mack; but the next people in this line of research are called Rumbaugh and Savage-Rumbaugh. To be fair, that combination of names makes me chuckle too. Premack used plastic tokens that varied in size, colour, shape and texture. He associated them with different words and different word-types. Sarah, his prodigy, learned nouns, verbs, quantifiers (many, few) and conditionals (if, then), resulting in a total of 130 words. She was only rewarded upon the formation of a grammatical sentence and she was bloody well capable of doing so. She even began to ask hew own questions using the tokens.

Taking the whole process a step further was Rumbaugh (1977) who buggered around with something called a lexigram. It was a much more advanced version of the same idea in which a keyboard is filled with visually distinct symbols which represented different words. The chimp Lana was able to learn the syntactical rules of 'Yerkish' which was a simplified version of English. In 1994 Rumbaugh and Bigmental-Rumbaugh (1994) tried to teach a bonobo (a species resembling chimpanzees but with typically better abilities at this sort of experiment, and are smaller.) Matata failed completely, but was responsible for the birth of the most famous language-learning chimp: Kanzi. Without any effort being made to teach him the use of the lexigram board, Kanzi picked it up easily.

Kanzi was soon proficient at every symbol on the board. In an attempt to allow him to develop his potential, the researchers let him wander amongst them as much as he wanted. He had a portable board he could use to ask questions or express desires; and eventually his training shifted to learning the correct human sounds to the meanings of those symbols. He could never produce the sounds, but he was quickly able to recognise them. He seemed not only to be good at use of...what really must now be considered language, but at understanding the intentions and semantics behind it. When asked to get 'item X' from 'location Y' he would go and do so, even if an item identical to 'item X' was right in front of him.

On the basis of Hocket's key features of human language; Kanzi, Sarah and Lana each showed the main four features of human-like language. Remember that they were very special cases. Not every chimp/bonobo/greater ape has the ability to pick up that level of skill. they aren't physically designed to use it, ad a language of such complexity has no use for them in their natural environment. Even if every ape learned the same language, the realistic uses for the skill would be questionable.

My incomplete theory is this: As the complexity of social groups increase, so does the complexity of their style of communication. Vervet monkeys call in three different ways about the here-and-now, Whales echolocate and define a strict heirarchy across the oceans using music heard only deep below sea or in massage studios. Chimpanzees and other great apes can references the past and the future in their natural interactions in a very limited way; and then there is us. We, with a full on complex syntactic system in each of our societies allowing us all to build our vocabularies to include new words, old words and made up words. We can reference the past and future, we can.... ah bollock it, you get the idea. We are born with the in-build capacity to quickly learn language(s) which leads me to believe there was some selective pressure on the ability. What was that pressure? How I wish I knew.

In the end, language is on the same scale as communication, just much higher up it. That's what I reckon and stuff the transcendentalists.