Tag Archives: cognitive apps

Putting language into the social brain

cartoon@langBlogChris and I are visitors at the Interacting Minds Centre at Aarhus University during August and September 2014.  This is a great opportunity to meet people who excel at providing stimulating conversation on “What makes us social”.  Admittedly I am more of a talker than a listener, and I assume that talking is completely, utterly, super-important in interactions. So why do social neuroscientists not pay more attention to language? Continue reading Putting language into the social brain

Kristian Tylen is the person I turned to with my burning question: What is the role of language? When we consider social interactions this almost always involves language. Is language actually the primary driver of our social interactions, or is it the other way round?

KT:  Language is multimodal. It probably started with gestures and this presupposed social interaction of a kind. There are different theories about the evolution of language and there is actually no real consensus.

UF: That’s what I felt when I recently read Stephen Levinson’s review  of two new books on the evolution of language, one by Derek Bickerton and one by Michael Tomasello.

KT: These represent interestingly different theories and there are others too. There is the view that language should be seen as completely separate from social communication, and that it grows from its own cognitive mechanisms. I sort of doubt this. I think language is primarily for the use of communication and it evolved as a tool to improve communication.

UF: Without specifically referring to Chomsky, what do you think about an innate predisposition for language in humans?

KT: It depends on what you mean by innate. My preference is to think that language both evolves from and is shaped by our interactions with the surrounding physical and social environment.  And so it is out there rather than inside us. This is demonstrated by the way that language structures are motivated. Take the way we talk about pitch  In English and Danish: We talk about low and high pitch mapping onto low and high spatial notation. Other languages for instance use thick and thin or big and small. These relations all map onto universal experience. Low tones come from big creatures and high tones from small creatures. And it turns out that it is very difficult to learn the opposite relations.

UF: But things out there need to act on the brain, no?  I don’t disagree with you that the world outside the mind is a starting point, but the experience of the outside shaped the inside, over millenia.  As a consequence, I guess there are some pre-shaped circuits in the brain, which might become obsolete, if  the environment changed radically.  So this is why I would put the outside in second place, and the inside first.

KT: So what do you mean by innate?

UF: I use the word for neural circuits that have a long evolutionary history. I don’t use innate to mean present at birth, and I don’t use it to mean that learning is not required. But I believe the circuits are preprogrammed for certain types of learning to be facilitated. That is why I like to talk of start-up kits rather than innate circuits. Start-up kits lead to effortless fast learning. No start-up kits require effortful slow learning.

KT: For me, it’s important to remember that the driver of evolutionary history is the environment, and this rather puts anything innate in second place. I am not very comfortable with putting too much stress on the inside of humans. Communication happens because we share the same experiences and systems on the outside and can create symbols that are mutually understood almost irrespective of our individual insides.

UF: Here is my naïve sketch of a likely scenario for the evolution of language in human beings, as opposed to other species. I want to speculate on what might be an innate start-up kit, still requiring learning, and what might not be innate and can only be gained by learning.

Chris, who meanwhile did the sensible thing of googling ‘evolution of language’, weighed in with a devastating remark:  You might be interested to know that the Linguistic Society of Paris in 1866 banned any debate on the topic as unsuitable for study because of the lack of empirical evidence.

UF: Too bad. I will still tell my story. I have only just made it up!  And what you are saying means that It cannot be worse than any other current proposals. They’re all speculation.

In the beginning was not the word. In the beginning there were two things: a start-up kit for spontaneous vocalisations, and another for spontaneous mentalising. These innate mechanisms are visible in precisely located and well connected neural circuits in the brain. They evolved separately, growing from already existing capacities in our remote ancestors.

KT: What were these pre-existing capacities?

UF: I speculate that even very remote ancestors had vocalisations, and they had the experience of agency. They must have had this to distinguish prey and predator and behave accordingly.

My story requires a third thing, and it demands a sufficiently complex brain. This is combining language and mentalising. This is not trivial. I imagine it needed many generations of our ancestors inventing words and telling each other what they mean, influencing each others minds along the way. Now for the magic bit: Language and mentalising put together enhanced each of their effects. This third thing is perhaps primarily responsible for the rise of human culture.

CDF: Mike Tomasello made a similar but different proposal. He believes that the ability to cooperate is a uniquely human driver of language and of culture.

UF: But what underlies the ability to cooperate in the way humans do? I would guess it depends on mentalising. I find it hard to conceptualise cooperation, or joint attention for that matter, as a more primary mechanism than mentalising. Cooperation and joint attention are not so much cognitive mechanisms as behaviours. No doubt cooperation is an important driver of culture, but so is probably competition. We cooperate within groups but compete with other groups.

KT: Why not? It’s a ‘just so story’ as always happens when people talk about evolution.

CDF: I think the Linguistic Society of Paris had the right idea.

Our conversation ended, but here are some details  that I would have given to bolster up my argument – if I had had the chance.

1. Spontaneous vocalisation. These have a useful instrumental purpose, like the warning cries that initiate flight from predators, or the high pitched cries that bind mothers and babies. This is a mechanism present in many mammal brains and presumably present in human ancestors as well. However, as Stephen Levinson reminds us in his review, the vocal apparatus of human beings is immensely complex and is under voluntary control, which may not be the case in many other animals. Speech is a complex motor action and well localized in the brain. Voice actions and hand gestures have similar requirements in terms of motor programmes and syntax. The FOXP2 gene has been associated with such actions and their relevance to speech and motor disorders has been documented.

2. Spontaneous mentalising. I find it very interesting that Martin & Santos have shown that the monkey brain can represent another’s viewpoint, but not another’s belief. Only the latter counts as mentalising, and Agnes Kovacs and her colleagues observed this in 7 months old babies. There are many other advantages, such as friendship, and also some disadvantages, such as deception. Apparently, there are forms of friendship and deception present in monkeys, but typically as rare examples. It is easy to see how language can scale up both advantages and disadvantages.

3. The third thing – two together. I am speculating that one of the two mechanisms on its own would probably bring you up to the level of a three-year old child. Not bad, but not enough to get on with the business of creating lasting and cumulative culture. Culture, as we all know depends on learning from others.  The specifically human form of learning from others which involves teaching is likely to be dependent on mentalising, but it would be a poor vehicle for learning without language.  Humans use language not just to make others do something, but also to express their mental states and to appeal to another’s mental state. Humans have done very well by learning from each other through the use of language. After many thousands of years humans got to a point where language was turned into literacy, and then science and technology started to take off.

Of course there are alternative views

How to be smart

Are humans the smartest species? What does it mean to be smart?

CDF: I am very interested in your suggestion that we should think about specific packages of behaviour for different species that are determined, to some degree, by their environmental niche. This brings me to a question I have often thought about. We often assume that humans differ from other species simply because we are smarter. But what do we mean by smart? Perhaps I can anticipate your answer: What is smart in one niche might be foolish in another. For example, we assume that being able to delay gratification (not eating the marshmallow immediately) is smart, but it would be very foolish in a volatile environment where no one can be trusted.

Continue reading How to be smart

RR: Being smart must include surviving and producing surviving offspring – just think about the opposite.

CDF: In this sense, in the long run, bacteria will turn out to be far smarter than humans.

RR: There is more to being smart: we usually mean something specifically cognitive. Evolution involves adaptation but in a stable environment this can lead to specialization that becomes maladaptive when the environment changes. Being large may be advantageous because it means you may be able to overwhelm your enemies. But the package of being large means you take a long time to grow, produce a few offspring that need looking after and therefore you need to live a long time. This makes your species vulnerable to environmental fluctuations. Bacteria are smart because they reproduce very quickly and can adapt rapidly to environmental change. So, in the battle between antibiotic-wielding humans and bacteria, the bacteria tend to win. The mega-fauna did well in a stable environment, but do not cope well with human-induced environmental change and many will soon be extinct.

UF: I am glad that we are not giants.

RR: Humans are not giants, but they are mega-fauna. They are exceptionally successful as mega-fauna because their best adaptation is adaptability itself. Humans have adapted, behaviourally, to many different terrains and climates and have changed the terrain to suit themselves. They have also changed the climate, but haven’t quite adapted to that yet.

SmartTestSo it is cognitive and behavioural flexibility that makes humans smart. Cognitive flexibility allows humans to learn a lot which makes them even more behaviourally flexible and therefore even smarter.

 

UF: How on earth did humans evolve to have this adaptability? Can you briefly sketch out how this might have happened?

RR: We need to talk about selection and the Baldwin Effect.

This describes the way that learned products of cognitive flexibility  – good survival tricks – become genetically determined in a species by selection, rather than by inheritance of acquired characteristics. Here is an example: When proto-humans increased their communication skills as proto-speech, individuals with useful cognitive apps in this new social environment e.g. being good at remembering the order of sounds, being good at hearing inflections, would be favoured. This goes on over many generations and eventually a large collection of language-useful apps would be embedded in the human brain.

Now as to the mechanisms involved – this is complicated. We have to consider what genes do. Genes always compete against their alternative alleles and never do anything for the good of the species. But genes have the products of other genes, in their own body and other bodies, as part of their environment, such that complex competitions and co-operations will develop between genes. Picture how different personality types may win or lose depending on the environment. Sometimes pessimists survive catastrophe, and sometimes optimists advantageously exploit opportunities.  So a mixture of different cognitive styles (think collections of cognitive apps) will exist in a population in an equilibrium, which will fluctuate with environmental events.

Being a lot smarter than other people has survival value but it has one obvious disadvantage; in a social species being different can be difficult.

PunchLeaps&Bounds2ndMrsTanV104P273.5.5-100

Acknowledgements: ‘Culture-fair test’ modified Danish cartoon, source unknown; ‘Leaps & Bounds’, Punch Magazine vol 4 p 273 June 10th 1893

Are bees better than humans?

To distract me from dangerous goings-on outside my window, where men are cutting down tall trees, Chris & Uta have asked me about the social lives of marmosets. This is not really altruism on their part. They will use the information for their book.

CDF: We want to know about the origins of human social cognition. Where should we look? Bees and ants? Other primates?

Continue reading Are bees better than humans?

RR: I think you need to look at all and any species and work out the packages of behaviour that different species use.

UF: What do you mean by packages of behaviour?

RR: Behaviour, and social structure, depend on the constraints of environment and evolution and both boil down to reproductive success. Does the species have to get through a cold winter? Does the species live up trees? It is factors like these that will determine whether the species produces a large number of offspring with low parental investment or a low number of offspring with high parental investment. These strategies come as packages and much social structure follows therefrom.

UF: So, is reproductive behaviour, mating and nurturing the young, at the biological root of social behaviour?

RR: This is where we find differences between species, because of the constraints I just mentioned. Old World monkeys tend to have a hierarchical mating system. Here males compete with other males for access to large numbers of females – though it all gets complicated with alliances and pay-offs. New World monkeys tend to live in more or less monogamous families – though it all gets complicated with adultery and incest. New World monkeys tend to live up trees, like birds, which are also usually monogamous so perhaps trees matter.

UF: Are humans more similar in their mating strategy to New World monkeys?

RR: Humans exhibit different strategies in different societies depending on food supply and other external threats. We can infer a lot about Human ancestral social behaviour by the fact that males are slightly bigger than females. This size difference occurs in any species where males compete with other males for resources and for females. The ultimate aim here is to win wars, get rid of other males, and acquire as many available females as possible. Of course, the size difference in humans is not large compared to the difference in some species. This implies a long history of more family based behavioural packages as well.

UF: Mike Tomasello’s lab has a vast number of studies comparing the joint interactions of great apes and those that happen between human children. There are similarities, but one of the main differences seems to be that, while human social interactions are basically cooperative, those of chimpanzees, gorillas and orang-utan’s are mostly driven by competition. Tomasello’s conclusion is that great apes are extremely limited in their cooperative capacities, and this is their essential difference to humans.

CDF: But unlike chimpanzees, Marmosets seem to be extremely cooperative, showing, for example, cooperative breeding. Is this true?

marmosets RR: Kin selection explains many aspects of superficially altruistic behaviour. Co-operative breeding in marmosets usually means assisting in the rearing of the offspring of relatives because of the family structure of groups of marmosets. But an owner of a captive breeding colony of marmosets will often attempt to cross foster a neonate when the breeding female of one family has had just one baby and the breeding female of another family has just had triplets. This is often successful because the family can’t be sure that the newcomer is not related to them (and marmosets aren’t very good at counting either). In the wild, quite a large troop will contain mainly related animals. So a certain amount of general co-operation could be sustained by kin selection.

Several additional explanations have been put forward to explain co-operative behaviour in co-operative breeding species such as marmosets. These are all plausible, but none are truly altruistic since they increase the reproductive success or survival of the protagonist.

* Helpers that have cared for other infants have greater reproductive success when they become parents than individuals that do not have previous infant care experience. This is undoubtedly the case for marmosets.

* Unrelated helpers can gain the benefits of living in a social group, such as communal foraging or protection against predators. This requires the unrelated individual to be accepted by the group and this usually only occurs when the incomer arrives as a potential breeding partner.

* Males that display involvement with infants are more likely to obtain subsequent mating with the female they assist. This is unlikely since the male is usually assisting in the rearing of his own offspring (so has access to the dominant female) or his own sibs. The dominant female suppresses ovulation in her daughters and thereby prevents incest in the rest of the family.  “No hanky-panky please.”

UF: Social learning is something that happens in almost all animal species. It seems possible to us that this learning has shown a sort of step change in evolution in humans. This is to do with learning by deliberate teaching, a two-way process. Is this plausible? How would evolution explain such a step change?

RR: The difficulty in getting a hand-reared orphan animal to be accepted by, mate with and rear offspring with, members of their own species shows how important social learning is for survival. Evolution will favour parents who behave in such a way that offspring learn social behaviours. In New World monkey species, the fathers and older siblings contribute to the rearing of the next litter and, in captivity, offspring reared alone by just their mother are rarely able to rear offspring themselves. The difference between ‘behaving in such a way that learning in offspring occurs’ and ‘teaching offspring’ is one of intention by the parent and so the step change (if there is one) will lie in the evolution of several cognitive ‘apps’, e.g. ‘appreciation of the existence of the future’, such that intentions can develop. New ‘apps’ have emergent functions made from the amalgamation of smaller ‘apps’.

UF: I love your use of the term ‘app’ for a special cognitive adaptation!

CDF: The importance of learning by teaching is that it offers a mechanism for culture to accumulate over generations. Culture seems to be the one specialised niche for humans.

RR:  Intentionality will always speed up cultural development, but animals have cultures too. Diet and food-finding practices are cultural in that they differ between groups of the same species.

The trees are down now! Good. I can go out again.

CDF & UF: Thank you, Ros. We’ll be back and quiz you more about processes in the evolution of behaviour.

Photo acknowledgement: 

http://evoluahomosapiens.blogspot.co.uk/2014/01/sagui-de-tufos-brancos-callithrix.html