First words There are over 6,000 different languages today, but how did language evolve in the first place? Pinpointing the origin of language might seem like idle speculation, because sound does not fossilize. However, must, chitchat and even humor may have been driving forces in the evolution of language, and gossip possibly freed our ancestors from sitting around wondering what to say next. There are over 6,000 different languages today, and the main language families are thought to have arisen as modern humans wandered about the globe in 4 great migrations beginning 100,000 years ago. But how did language evolve in the first place? Potential indicators of early language are written in our genetic code, behavior and culture. The genetic evidence is a gene called FOXP2, in which mutations appear to be responsible for speech defects. FOXP2 in humans differs only slightly from the gene in chimpanzees, and may be about 200,000 years old, slightly older than the earliest modern humans. Such a recent origin for language seems at first rather silly. How could our speechless Homo sapiens ancestors colonize the ancient world, spreading from Africa to Asia, and perhaps making a short sea-crossing to Indonesia, without languages? Well, language can have 2 meanings: the infinite variety of sentences that we string together, and the pointing and grunting communication that we share with other animals. Marc Hauser (Harvard University) and colleagues argue that the study of animal behavior and communication can teach us how the faculty of language in the narrow human sense evolved. Other animals don't come close to understanding our sophisticated thought processes. Nevertheless, the complexity of human expression may have started off as simple stages in animal "thinking" or problem-solving. For example, number processing (how many lions are we up against?), navigation (time to fly south for the winter), or social relations (we need teamwork to build this shelter). In other words, we can potentially track language by looking at the behavior of other animals. William Noble and Iain Davidson (University of New England) look for the origin of language in early symbolic behavior and the evolutionary selection in fine motor control. For example, throwing and making stone tools could have developed into simple gestures like pointing that eventually entailed a sense of self-awareness. They argue that language is a form of symbolic communication that has its roots in behavioral evolution. Even if archaic humans were physically capable of speech (a hyoid bone for supporting the larynx and tongue has been found in a Neanderthal skeleton), we can't assume symbolic communication. They conclude that language is a feature of anatomically modern humans, and an essential precursor of the earliest symbolic pictures in rock art, ritual burial, major sea-crossings, structured shelters and hearths – all dating, they argue, to the last 100,000 years. But the archaeological debate of when doesn't really help us with that was occurring in those first chats. Robin Dunbar (University of Liverpool) believes they were probably talking about each other – in other words, gossiping. He discovered a relationship between an animal's groups size and its neocortex (the thinking part of the brain), and tried to reconstruct grooming times and group sizes for early humans based on overall size of fossil skulls. Dunbar argues that gossip provides the social glue permitting humans to live in cohesive groups up to the size of about 150, found in population studies among hunter-gatherers, personal networks and corporate organizations. Apes are reliant on grooming to stick together, and that basically constrains their social complexity to groups of 50. Gelada baboons stroke and groom each other for several hours per day. Dunbar thus concludes that, if humans had no speech faculty, we would need to devote 40% of the day to physical grooming, just to meet our social needs. Humans manage large social networks by "verbal grooming" or gossiping – chatting with friends over coffee, for example. So the "audience" can be much bigger than for grooming or one-on-one massage. Giselle Bastion, who recently completed her PhD at Flinders University, argues that gossip has acquired a bad name, being particularly associated with women and opposed by men who are defending their supposedly objective world. Yet it's no secret that men gossip too. We are all bent on keeping track of other people and maintaining alliances. But how did we graduate from grooming to gossip? Dunbar notes that just as grooming releases opiates that create a feeling of wellbeing in monkeys and apes, so do the smiles and laughter associated with human banter. Dean Falk (Florida State University) suggests that, before the first smattering of language there was motherese , that musical gurgling between a mother and her baby, along with a lot of eye contact and touching. Early human babies couldn't cling on to their mother as she walked on 2 feet, so motherese evolved to soothe and control infants. Motherese is a small social step up from the contact calls of primates, but at this stage grooming probably still did most of the bonding. So when did archaic human groups get too big to groom each other? Dunbar suggests that nomadic expansion out of Africa, maybe 500,000 years ago, demanded larger group sizes and language sophistication to form the various alliances necessary for survival. Davidson and Noble, who reject Dunbar’s gossip theory, suggest that there was a significant increase in brain size from about 400,000 years ago, and this may correlate with increasing infant dependence. Still, it probably took a long time before a mother delivered humanity’s maiden speech. Nevertheless, once the words were out, and eventually put on paper, they acquired an existence of their own. Reading gossip magazines and newspapers today is essentially one-way communication with total strangers – a far cry from the roots of language.