My BBC Future column from last week. The original is here. I started out trying to write about research using economic games with apes and monkeys but I got so bogged down in the literature I switched to this neat experiment instead. Ed Yong is a better man than me and wrote a brilliant piece about that research, which you can find here.
It’s a question humanity has repeatedly asked itself, and one way to find out is to take a closer look at the behaviour of babies.… and use puppets.
Fundamentally speaking, are humans good or bad? It’s a question that has repeatedly been asked throughout humanity. For thousands of years, philosophers have debated whether we have a basically good nature that is corrupted by society, or a basically bad nature that is kept in check by society. Psychology has uncovered some evidence which might give the old debate a twist.
One way of asking about our most fundamental characteristics is to look at babies. Babies’ minds are a wonderful showcase for human nature. Babies are humans with the absolute minimum of cultural influence – they don’t have many friends, have never been to school and haven’t read any books. They can’t even control their own bowels, let alone speak the language, so their minds are as close to innocent as a human mind can get.
The only problem is that the lack of language makes it tricky to gauge their opinions. Normally we ask people to take part in experiments, giving them instructions or asking them to answer questions, both of which require language. Babies may be cuter to work with, but they are not known for their obedience. What’s a curious psychologist to do?
Fortunately, you don’t necessarily have to speak to reveal your opinions. Babies will reach for things they want or like, and they will tend to look longer at things that surprise them. Ingenious experiments carried out at Yale University in the US used these measures to look at babies’ minds. Their results suggest that even the youngest humans have a sense of right and wrong, and, furthermore, an instinct to prefer good over evil.
How could the experiments tell this? Imagine you are a baby. Since you have a short attention span, the experiment will be shorter and loads more fun than most psychology experiments. It was basically a kind of puppet show; the stage a scene featuring a bright green hill, and the puppets were cut-out shapes with stick on wobbly eyes; a triangle, a square and a circle, each in their own bright colours. What happened next was a short play, as one of the shapes tried to climb the hill, struggling up and falling back down again. Next, the other two shapes got involved, with either one helping the climber up the hill, by pushing up from behind, or the other hindering the climber, by pushing back from above.
Already something amazing, psychologically, is going on here. All humans are able to interpret the events in the play in terms of the story I’ve described. The puppets are just shapes. They don’t make human sounds or display human emotions. They just move about, and yet everyone reads these movements as purposeful, and revealing of their characters. You can argue that this “mind reading”, even in infants, shows that it is part of our human nature to believe in other minds.
Great expectations
What happened next tells us even more about human nature. After the show, infants were given the choice of reaching for either the helping or the hindering shape, and it turned out they were much more likely to reach for the helper. This can be explained if they are reading the events of the show in terms of motivations – the shapes aren’t just moving at random, but they showed to the infant that the shape pushing uphill “wants” to help out (and so is nice) and the shape pushing downhill “wants” to cause problems (and so is nasty).
The researchers used an encore to confirm these results. Infants saw a second scene in which the climber shape made a choice to move towards either the helper shape or the hinderer shape. The time infants spent looking in each of the two cases revealed what they thought of the outcome. If the climber moved towards the hinderer the infants looked significantly longer than if the climber moved towards the helper. This makes sense if the infants were surprised when the climber approached the hinderer. Moving towards the helper shape would be the happy ending, and obviously it was what the infant expected. If the climber moved towards the hinderer it was a surprise, as much as you or I would be surprised if we saw someone give a hug to a man who had just knocked him over.
The way to make sense of this result is if infants, with their pre-cultural brains had expectations about how people should act. Not only do they interpret the movement of the shapes as resulting from motivations, but they prefer helping motivations over hindering ones.
This doesn’t settle the debate over human nature. A cynic would say that it just shows that infants are self-interested and expect others to be the same way. At a minimum though, it shows that tightly bound into the nature of our developing minds is the ability to make sense of the world in terms of motivations, and a basic instinct to prefer friendly intentions over malicious ones. It is on this foundation that adult morality is built.
Mind Hacks 
“Are we naturally good or bad?” – If we are either, I would argue that this experiment contributes very little towards an answer.
One approach, as found in the experiment described, is to say something like “I think babies, in their first year of life, innately possess the capacity to evaluate complex social interactions and will automatically apply this reasoning to abstract shapes in a scene in the manner I expect them to. They will demonstrate this adult-like level of reasoning by looking a bit longer at some moving shapes than others, and maybe by reaching for some toys more often than other toys, but this will not have to be consistent across different age groups. I will design an experiment to confirm all this, and this will allow me to draw conclusions about universal and unlearned morality without explaining any actual pathways between genes and behaviour”.
Another approach is to suggest something like “given the looking and reaching behaviour observed, can we strip out all pre-conceived adult terminology (‘helper’, ‘hinderer’, ‘goal’, ‘climber’, etc.) and explain the observed behaviour without assuming any a priori innate knowledge of complex interactions. Let’s not assume infants see anything other than shapes moving in a scene. Then let’s see if we can bias behaviour towards particular preferential looking/reaching outcomes at test by explicitly and actively controlling for factors such as colour, shape, movement, handedness, familiarity/novelty, presence of googly eyes, replicability in other scenes/scenarios, and so forth. If we can, then let’s not make any extraordinary claims about the innateness of morality just yet.”
I’m still rather surprised that Nature considered the former approach sufficient.
A few years ago I read a paper in which 5 months old infants were able to demonstrate a “sense of fairness”. I got into an argument with my daughter (who is finishing up her Ph.D.) over whether an infant that young had learned enough social skills that we could not say the “sense of fairness” was innate. I think it is innate along with other social skills such as a sense of hierarchy and a sense of how to be submissive or dominant, etc. Anyway, I lost the reference to that paper and have been looking for it ever since, as it is an important paper in any discussion of human morality. Anyone know the paper?
Happy to read someone able to dissect results in a logical manner. I’m actually more interested though in the infants who did not look longer. What percentage was it? And obviously it could be for a number of reasons, but are those infants’ reactions considered as much as the ones who preferred friendly characters? I’d think someone would want to follow them like in the marshmallow study.
Another way of looking at this is that the babies chose the object which they thought would be more likely to help them in time of need and were surprised when someone else chose the object which would be more likely to hinder them. To me this bias towards self is more likely than a sense of good or bad.
Or maybe our sense of good or bad is rooted in a sense of how “good” or how “bad” is something for me.
Your comment, neurotic ape, implies that we have a sense of self from birth. Now phenomenologist know that our sense of self is multifaceted. We have a sense of agency, of possession, of ownership (not to be confused with possession – ownership is owning our body or awareness of being our body), location of being, and many other senses of self that we collectively call “self”. I’ve listed over twenty different senses of self, all valid or useful senses of self. How much of that exists soon after birth would be very difficult to verify although ownership and agency probably form the genesis of that compound sense. The fact that we call it a sense, thus tied to awareness, and even a reflective awareness, implies the self has origins in our usual senses, the motor cortex, and motion, even if only from the eyes.
A sense of good or bad probably also has many different facets. Good for who, for what, for where-when, etc.?
Our casual use of these terms tend to set us off on wrong paths to understanding. And we can be amazed when we start to unravel that ball of senses at how intricate it is.
Mason, as you indicate the “sense” of self is a complex subject with many different levels. The fact is that we have yet to truly nail down what this sense is. We understand parts of it, but the whole is still quite a mystery.
I have actually been pondering this for a few weeks now. Does a single celled organism have a “sense” of self? I am not speaking of self awareness but more of motivation to act to benefit the “self”. Quite a few appear to. If they do, do we acquire our sense of self when the sperm joins the egg? Again not self awareness but motivation towards a goal. As cells are added to “us” do we then add a deeper sense of self to the core or was it all there stored away in our DNA? Difficult questions to answer, but I feel very worthwhile questions to ask.
I see no sense in attempting to attribute a sense of self to a single cell organism where a biochemical action is capable of explaining all that happens there. I wouldn’t think it appropriate to attribute a self to Eric Kandel’s Marine Snail. I recognize that the self is something that can exist on a continuum and that as animals increase their mental capability that sense of self would naturally expand in its functions. But for this discussion it probably is wise to limit ourselves to the human sense of self, even though I personally would agree with my two cats that they also have a sense of self and very different personalities from each other.
We can look at the self and see what environmental or mental problem it solves and that would enable us to see it as part of the continuing evolution of the human mind. For example, I hypothesize that the one of the self’s main functions is to evaluate risk and sound alarms if any risk increases in the environment. So we can look at the menagerie of the different types of senses of selves in humans, and we can hypothesize what their functions are and thus how would their presence improve our probability of survival. I do hesitate to claim there are “levels” of selves, only that some appear more primitive than others. I did not mean to imply that a sense of agency and ownership is at some lower level than the other senses of self, but you could infer that they might have developed at an earlier period in animal evolution and in neonatal development. I’m sure my cats have a sense of agency and ownership and probably risk also. At least when I observe them engage with a neighbor’s cat I have to restrain myself in anthropomorphizing that they are calculating their risks and chances leading up to a cat fight. But, by golly, it sure seems that is happening. And I do the same thing with observing other humans and thinking I can guess what is going on in their heads as they try to figure out the best way to approach problems. But anthropomorphizing humans can be as dangerous as anthropomorphizing other animals.
You can see this topic can balloon into distant tangents. As interesting as they are, let’s return to the original topic of Mind Hacks. Is our sense of good and bad innate? If so, what is the evidence? And what relationship does it have to any other innate social skills we might have. And, we can add, what is having this sense of good and bad? Is it our sense of self? I think it has a lot to do with our menagerie of senses of self at least some of them.
Finally, a wonderful collection of essays on the self was edited by Shaun Gallagher and Jonathan Shear, “Models of the Self” (1999). And Dan Zahavi’s “Subjectivity and Selfhood” (2008). Great reading.
The experiment as it’s described doesn’t convey much at all. Even if infants were evil, they would naturally be inclined to approach the “good” shape since it’s less likely to hurt them.
A good experiment would consider whether they would, for example, want to reward or punish the shapes. And more importantly perhaps, gauge their reaction when their own interest is at stake.
Observer, an intention of reward and punishment appear to be learned behaviors rather than innate because they conjoin what we presume to be innate senses or social skills with methods of dealing with events unique to a culture. Perhaps you meant something more along the lines of fight or flight reactions, which are, I assume, also innate. But is the conjoining of two innate behaviors also innate? It is not clear. Fight/flight reaction in an immobile infant would probably be expressed as distress, and, if so, might be unethical.
Basic nature of human lies in amusement caused by surprises. Surprises can be good as well as bad. However they are unusual, not normal. Good surprises have no attraction because that;s how the world is supposed to run. Most babies are shown a good world by adults, where everything has a happy ending. However, babies get shocked, when the world that they see goes in utterly different way, unlike the fairy tales of adults. Anything that is out of the system contains rebellion. Rebellion is attractive. In group, people feel safe by seeing others in trouble, just to satisfy one self that he/she is not in danger and glorify one’s own fortune by being able to see the worst that could happen and feeling lucky not to have experienced such bitter moments.
Your post, Sanjeeb, perplexes me. Here is a partial list of terms you use: Basin [human] nature, surprises, supposed to run, happy ending, shocked, rebellion, feeling lucky, among others. Each person reading what you wrote will probably put their own interpretation on those words or phrases all the while assuming everyone else is interpreting them the same way. I know that I have made that error. That means that I don’t have a clue as to what you really said as I have no idea what those terms mean to you. Our common language is like a fog that obscures as much as it illuminates. Sharing the same words is not sharing the same meanings. And our different life experiences pull us in different directions. You expect infants to rebel while I might think they are simply communicating their discomfort with diaper rash? Don’t you think a more precise description is useful? If so, what do you think is an appropriate direction to take there?
I’m guessing English is not their first language so maybe we should cut them some slack.
Interesting comments. Another way to look at it is that infants have an early, not necessary innate, model of how (social) life works. In this model facilitating can simply be more probable than hindering. We do not necessary need more complex or, let me say blurry, concepts like self or morality for our explanation. On the other hand, this early model can serve as one of the building stones of self and morality.
Andras, thanks for your alternative facilitating/hindering hypothesis. You appropriately suggest that our concepts are blurry. Morality is certainly blurry. Same for good/bad. I think that is because we fail in reducing them to a set of rules, no matter how general or how many. But what is being facilitated or hindered? It seems we are right back into the blurry territory when we ask that and still have to deal with whether it is innate or not.
Using the analogy of sprout and a oak tree, the sprout has all the structure and characteristics of a mature oak but depending on where it grows will determine what it will look like in another 50 years. We can say the mature oak is innate in the sprout because in spite of what it looks like in 50 years it still has all the internal structures that the mature oak has. The structure is innate. A question to me is how much the neurological structure of humans and its developmental changes during maturation are fixed from innate characteristics and how much are plastic that are so profound that they divide mature humanity into different behavioral species. We know that all prior attempts at creating the “new man” from political or religious pressures have failed. Looking back over the crude attempts over the last 2,500 years shows us how cruel those attempts were. It may be that there is no “new man” to create. Humans are humans and nothing more? If so, we should at least understand what we are instead of what we “should” be.
Building an inventory of what social skills are innate would help us construct a model of how social life works. In addition to the already demonstrated sense of fairness, I hypothesize there is a sense of hierarchy or ranking which might even be related to the inherent mathematical sense of inequalities that Stanislas Dehaene wrote about in “The Number Sense”. Whether it is vague or not I would include a sense of self. Is there a short list? That would depend, it seems to me, on how we conceptualize the innate social skills. That is why I mentioned the sense of self in earlier posts, to demonstrate that any sense probably can be subdivided. And what we cannot demonstrate as innate should we conclude are learned? Or are there emergent skills that that can never clearly be attributed to any one source but are a combination of experience with those derived from neurological maturation? For example, our strategies for coping with the world probably falls into that indistinct area and would help explain the development of the variety of personalities.
The difficulty is determining whether the desire for successful outcome of the uphill shape’s climb emanates from empathy for it or an innate preference for order over disorder (by the term order I mean a logical outcome of effort). Already a baby has a sense of cause and effect in motion, and we might expect it to feel distress or surprise at any hindering effect. True, there are two competing motivations but the uphill journey starts the motion, therefore one might assume it is the baby’s first and stubbornly fixed referent to gauge cause and effect. In that sense the upward motion is expected to create order while the downward push signifies disorder (ie the effort is wasted by the sudden thwarting of the uphill journey). Is the infant merely preferring order to disorder in the same way that it might cry when a repetitive movement ceases? If a raindrop aimed at a cup was thwarted by a person moving it away at the last moment, might it result in a similar reaction? It might let an ant crawl to the top of mound before deciding to crush it.