Why did language evolve?

Language is important to us modern humans so asking why it evolved seems like a silly question. After all, surely we can think of numerous benefits of language that drove its evolution in our ancestors. However, idle speculation is not good enough for science. We need data! Which of the many benefits to language we can think of was the reason it was selected for? Was it to allow better communication whilst hunting, to better teach complex tool-making techniques.

Or maybe it was to better gossip. That’s the idea proposed by Aiello and Dunbar in 1993; and remains one of the most popular and well supported explanations for why language evolved. It’s based on the social brain hypothesis (also developed by Dunbar), which points out that living in larger groups is beneficial. Since living in such large groups is intellectually demanding, requiring you to remember lots of faces etc., it would act as a selection pressure for large brains; explaining why humans have such big noggins (Dunbar, 2003).

However, if you’re living in a larger group you need to invest more time in socialising since you need to bond with more people. If the group is too big, the time you spend socialising can detract from time you need to spend foraging for food, sleeping or engaging in other wholesome ape activities. As such there’s a limit to the amount of time you can spend socialising. Studying modern primates has revealed that this is ~20% of your daily time-budget.

When apes invented the iPad, they found it even harder to find the time to socialise with others

But because humans have such large groups (and large brains) we should have to dedicate more than 20% of our time to bonding with each other. Fortunately we don’t have to because we can socialise with a greater efficency than modern primates. They bond by grooming each other, a process which can only involve two individuals at once. We can use language to talk to multiple individuals at once, reinforcing more social bonds simultaneously.

The amount of time we spend talking about shit

The amount of time we spend talking about shit

So, since language is more efficient we can maintain larger groups and still only spend 20% of our time making sure we’re still friends with everyone in them. This would explain why over 60% of the conversations we have don’t convey any “useful” information but are simply social (or “gossipy”, to use the technical term) in nature (Dunbar, 1993). We have these social conversations to reinforce social relationships, not teach us anything.

Further, since the rise of language is linked to group size we can examine the size of the groups our ancestors lived in to get an estimate for when language may have originally evolved. The relationship between group size and brain size identified by the social brain hypothesis can be used to predict how large a group our extinct ancestors lived in.

These figures suggest that Homo erectus (who lived ~1.8 million years ago and was the first of our ancestors with the modern body-plan) was the first to start living in groups that were too large for grooming alone to maintain, so language must have developed at some point around then.

How much time we'd have to spend grooming if language didn't evolve.

How much time we’d have to spend grooming if language didn’t evolve. Based on this Aiello and Dunbar concluded it must have developed by the time of Homo erectus

Of course, this doesn’t mean that the other benefits of language weren’t involved in it’s evolution, just that reinforcing social bonds appears to have been the most important. Further, these conclusions are far from concrete and liable to revision as new data comes in. For example, the stuff about the amount of time humans spend gossiping was based on a rather small sample size. But then, just because you need to take a pinch of salt with it doesn’t mean the whole hypothesis is worthless.

So the next time you get frustrated with the sheer amount of drivel on Facebook, relax. They’re only using language for its evolutionary-intended purpose.


Aiello, L. C., & Dunbar, R. I. (1993). Neocortex size, group size, and the evolution of language. Current Anthropology34(2), 184-193.

Dunbar, R. I. (1993). Coevolution of neocortical size, group size and language in humans. Behavioral and brain sciences16(4), 681-693.

Dunbar, R. I. (2003). The social brain: mind, language, and society in evolutionary perspective. Annual Review of Anthropology, 163-181.

If you follow EvoAnth on Facebook or Twitter, I promise to fight evolution and not gossip uncontrollably.

13 thoughts on “Why did language evolve?

  1. Pingback: Paletleme Amirliği – Nisan 2013 | Emrah Göker'in İstifhanesi

  2. Language modifies such a range of behaviours to such massive degrees that it is hard for me to believe that there will ever be an it-is-X answer to this question. It seems that a number of advantages would accrue simultaneously and rapidly in different spheres as language appeared. Grooming for larger groups and sexual selection all seem to me to be likely components of the answer but one of my favourites is the biological power of narratives. One of the weirdest things about humans (to me) is what suckers we are for narratives. Good stories seem to set back the progress of science at least four years for every five years of progress. We love a good story but what’s the attraction? If we want to people to remember anything – right or wrong – we convert it to a narrative.

    There has to be something going on here and it seems to me to be this: a narrative allows us to handle exceptional events. Normal biological learning follows a trial and error process. We are kitted out with some basic machinery fro recognising good and bad outcomes then we – ie we organisms with basic learning – learn by doing stuff and seeing whether how this registers on the good or harm scale. (If it registers extreme negative, aka death, there is no learning just a potential for evolution.)

    The next level of learning is parenting and group learning: baby bird follows parents around learning which piles of dirt are better to scratch for worms and, importantly, which of the other animals wandering around are actually predators. There’s no need to actually have a dangerously close encounter to learn that cats are bad, just watch how Ma and Pa respond.

    Narrative learning takes this to a radical new level. It is no longer necessary to have even seen a new plant to know if it ok to eat. You could know that bears are very dangerous and know the tell tale signs that a cave may have a bear in residence without having ever seen a bear or even a cave before. You learned from stories.

    This is a radical change to the scope of learning about the environment(s). It allows humans to niche jump, to capably exploit environments that they have never personally experienced before. It also allows humans to be aware of the exceptional dangers that may turn up only occasionally, even once in a few generations. In contrast, a rabbit is unlikely to learn about a rabbit trap without becoming someone’s dinner. The narrative-capable human brain is expensive to maintain but it can get you through an evolutionary bottleneck alive.

  3. Pingback: Did language evolve to help us gossip? | EvoAnth

  4. Pingback: Testing the social brain hypothesis | EvoAnth

You evolved too. Have a say.

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s