According to a new study of the history of human language, it turns out our ancestors of long ago actually spoke the way Master Yoda does in a movie series you might be familiar with, Star Wars.
Does George Lucas know something we don’t?
…probably not.
New Scientist writes:
What form did the first human language take? According to some linguists, all known languages descend from a proto-language, perhaps dating back to the first behaviourally modern humans 50,000 years ago. But little else is known about how our ancestors spoke.
Now two maverick linguists say they have clues. Merritt Ruhlen of Stanford University in California, and Murray Gell-Mann, at the Santa Fe Institute in New Mexico created a family tree for 2200 languages, living and dead, based on how they use similar sounds for the same meanings. Most modern languages use subject-verb-object sentences: “I see the dog”, while most dead ones, such as Latin, go subject-object-verb – “I the dog see”.
On Ruhlen and Gell-Mann’s tree, subject-verb-object languages always descend from subject-object-verb languages, but never the other way around. “This tells us that the putative ancestral language had subject-object-verb word order,” says Ruhlen. However, mainstream linguists are dubious about the tree’s validity.
Learning all the time: The Lingodroids project at the University of Queensland has developed robots that can create and speak their own language
In the Terminator films it was the moment when machines learned to think for themselves that sparked the downfall of humanity.
If that hypothesis proves to be true then scientists at the University of Queensland might have a lot to answer for.
Their Lingodroid project has developed robots that are creating and speaking their own languages – and increasing their own knowledge in the process.
The robot language has now evolved to a point where they can arrange to meet each other in different places, and even hold polite conversations.
‘Words’ are electronic noises, created using a random number of syllables, which are then assigned to locations. Location names created so far have included ‘kuzo’, ‘jaro’ and ‘fexo’. Each location was around a few metres in size.
To test and develop their language skills the Lingodroids play ‘word’ games in which they arrange to meet in other places, and it has worked successfully in simulations and in a real office.
The robots are creating their own ‘words’ because human languages are so complex and nuanced that the robots found it hard to decipher.
‘Robot-robot languages take the human out of the loop,’ project leader Dr Ruth Schulz told the BBC.
‘This is important because the robots demonstrate that they understand the meaning of the words they invent independent of humans.’
The Lingodroids themselves are two-wheeled robots, looking not too dissimilar to some vacuum cleaners, which use an onboard camera, sonar and a laser range-finder to map the space around them.
The language, which sounds similar to the keytones on a phone, is actually spoken aloud by the robots using a microphone and speaker.
Games played among them include the go-to, the where-are-we and the how-far game.
In the where-are-we game, the robots map their environment independently by driving around, and then whenever they meet another robot, one gives the area in which they meet a name and both update their vocabulary with the new word.
Sci-fi: In the Terminator series of films the moment when robots learned to speak for themselves that brought catastrophic consequences
In the go-to game, one robot chooses a location, both robots find the place in their own map, and then the navigate to that place independently.
The vocabulary this creates, called a toponymic lexicon, allows the robots to go on to develop ‘words’ for distances and directions.
With their expanded lexicon, the robots were even able to meet each other in places they had talked about but never been together, and to describe places they ‘imagined’ exist outside their own maps.
.
