‘Fake news’ is everywhere. It’s the rallying cry of Donald Trump, the curse of social media and the go-to excuse for just about anything that someone disagrees with. Fake news is simply any news that is presented as true but has no basis in fact, or is based on fact but contains significant falsehoods. Fake news is fabricated, made up, elaborated, embroidered, bogus, false, bullshit, or sometimes just simple barefaced lies. It is a problematic term. As the UK government pointed out when it banned ‘fake news’ from official documentation, it is ‘poorly defined and misleading … [and] conflates a variety of false information, from genuine error through to foreign interference in democratic processes’. Words like ‘misinformation’ or ‘disinformation’ are preferred in UK Government documents, and in cases where fake news is used to advance political causes or other agendas we might more correctly use the word propaganda. The near-constant accusations of ‘fake news’ by Donald Trump to dismiss more or less anything with which he disagrees has undoubtedly greatly weakened the term, or at least allied it with perceived political attacks by the ‘liberal mainstream media’. Washington Post columnist and former public editor of the New York Times Margaret Sullivan has suggested that the term be ‘retired’, saying that ‘When it started off, it actually meant what it sounds like – fraudulent or misinformation meant to deceive. It’s now become kind of a punchline, something people use to poke fun. If you don’t like something, you call it fake news.’1 She clearly has a strong point but ‘fake news’ has nonetheless become a convenient if sweeping shorthand for different types of information that are in some way lacking in veracity. I am going to continue to use it throughout this chapter for reasons of convenience, but be mindful that it has its limitations and may already be a term, although not a concept, on the decline.
The crucial take-home point from the proliferation of fake news is that we seem very keen to believe things that aren’t true. My argument will be that such misguided beliefs are manifestations of a series of mismatches between our evolutionary heritage and the modern media-drenched world that we have created. My aim is to explore the relationship between potentially evolved traits, or at least psychological components that may have been subject to selection and evolution (such as gullibility), and our ability to spot the truth in the modern world. There is an elephant in the room here though; since we are talking about ‘belief’ I could decide to drag religion into the argument. I am not going to go down that road and this decision warrants an explanation. I am an atheist, but I was brought up in what would be regarded as a Christian country in the late 1970s and early 1980s and so at primary school I was exposed to the Bible. At a very young age, listening to stories about Noah, or Adam and Eve, or Jesus working miracles and rising from the dead, I called ‘bullshit’. Well, I didn’t use that word but the gist was the same. None of it ever made any sense to me whatsoever, even as a five-year-old, although in later life I began to see behind the literal stories. The miracle of the loaves and fishes, so beloved by my primary school teachers that we chanted a song about it in assembly, is about getting people to share what little they have rather than some magical multiplication. And that is a good message. As I got older I found out about other religions and none of them ever moved me either. I know though that many people are moved, profoundly, by religion and that religion in many forms is an integral and important part of their lives. This chapter will not be yet another evolutionary biologist’s attack on the ‘stupidity’ or otherwise of religion. Believing in religion is also not a modern phenomenon, but the massive influence, impact and reach of fake news is very much a feature of the modern world.
What I am concerned with in this chapter is whether our ability to believe false information is a product of a mismatch between our evolved cognitive characteristics and the modern world. If our ability to believe things, to be gullible to some extent, exists because of advantages afforded to us in our evolutionary past then it is not a great stretch to suggest that, awash in the sea of modern media, this ability has left us without the cognitive satnav we need to navigate our way through the lies. In other words, it is possible that evolution may have left us helpless in the face of our own deceptions, and that is an interesting if rather terrifying proposition.
Fake news is old news
Fake news might seem like a modern phenomenon, and certainly the ability to disseminate such stories quickly and globally is modern, but history is littered with examples of what we might now term fake news. In the first century bc, Octavian orchestrated a fake news campaign, in this case perhaps better termed a propaganda war, directed against his one-time ally, but later bitter rival, Roman politician Mark Antony. Portraying him as a drunken womaniser and puppet of Queen Cleopatra, Octavian upped the ante when he got hold of a document he claimed was Mark Antony’s official last will and testament. Whether this document was what Octavian claimed, and scholars are still debating whether it was real, fake or partially forged, reading it aloud in the Senate was a masterstroke. Legacies, including substantial pieces of Roman-held territory in the Eastern Mediterranean, were promised to the children Mark Antony had with Cleopatra. The will also claimed that the son of Julius Caesar and Cleopatra, Ptolemy XV Philopator Philometor Caesar (better known by the nickname Caesarion), was Caesar’s rightful successor rather than Octavian, Caesar’s adopted son. These and other damning statements played on the suspicions and prejudices Romans held against the East and against Cleopatra (an early example of confirmation bias, a topic we will return to later). The Senate in Rome stripped Mark Antony of his right to lead Roman armies and made him a traitor. The rest is history, in this case one written by the victor in the war Rome declared not against Mark Antony but against Cleopatra.2
The invention of the printing press allowed stories to reach far bigger audiences far more quickly than ever before and printed material purporting to be fact has long been a source of fake news. Such stories ranged from witches and sea monsters to more elaborate claims, like the assertion that the Lisbon earthquake of 1755 was divine retribution against sinners. This story spawned an entire new genre of fake news pamphlets in Portugal called relações de sucessos, attributing survival of the earthquake to an apparition of the Virgin Mary.3 Another technological advance, the internet, allows a much greater platform for falsehood, reaching far beyond the ‘World War 2 Bomber Found on the Moon’ tabloid nonsense audiences are familiar with through publications like the Daily Sport or The National Enquirer. If you can’t read ‘Alien Bible Found! They Worship Oprah!’, ‘Saddam and Osama Adopt Shaved Ape Baby’ (both classics from the publication Weekly World News, a largely fictional tabloid now available online only), or ‘Statue of Elvis Found on the Moon’ (part of the Daily Sport’s obsession with weird lunar objects) and call ‘fake’ then you might have a problem as an individual. The trouble with fake news in the modern world, though, is that it is no longer just a problem for individuals; the manipulation of the media through the dissemination of fake news has become a political matter and a problem at society level nationally and internationally. The evidence that the active dissemination of false information has swung elections and referendums (‘All Aboard the Brexit Bus’) leaves us facing a difficult truth: we are not good at distinguishing fact from fiction, and this has become a major issue.
Have we evolved to be trusting?
Believing fake news suggests gullibility and being gullible is widely seen as a negative characteristic. Gullibility though is fundamentally an issue of trust; a gullible person trusts that other people are telling them the truth, and without trust we are all doomed. If you don’t believe me then imagine for a moment trying to live your life without trust. I am sitting in a house typing this on a laptop. I implicitly trust that the people who built my house knew what they were doing and that it won’t fall down. I also trust the manufacturers of my laptop, since it is plugged into mains electricity and is sitting, literally, on my lap. A wiring fault could lead to a very nasty incident but I trust the safety markings on the plug, the power lead and the laptop itself. I trust that my wife isn’t going to reach for a kitchen knife and kill me when she comes downstairs. In fact, despite our abilities and tendencies to be violent (see Chapter 7) I trust, every single day, that random people I meet won’t kill me. Whenever I drive I trust that hundreds, or even thousands, of other drivers won’t veer across the white line and crash into me. I trust that the guy who fitted my brakes didn’t mess up the job deliberately or by accident. I trust that the factory making my pre-packed sandwich lunch followed food hygiene regulations and that I won’t contract food poisoning. I trust that the vaccinations I insist on my children getting are what they say they are and not some government-sponsored mind-control serum. I even, despite occasional evidence to the contrary, trust that my children can play together unattended upstairs without causing too much damage to the house or each other. If you want any firmer evidence of the need for trust in society, pull out some money from your wallet. The grubby collection of worthless metal and paper, or whatever notes are made from these days, is testament to trust; you trust that someone will exchange goods or services for these trinkets and they in turn trust that a bank will take them and ‘honour’ them. Without trust we cannot have banks, and indeed a fundamental financial instrument is called a ‘trust’. Trust, a firm belief in the reliability, truth or ability of someone or something, is everywhere, essential and, as we shall see, evolved.
There is a very thin line between the trust necessary for everyday life and the gullibility or credulity that can lead us towards believing fake news. In step with the importance we place on trust, and in particular the trust that we have in other people, trust has attracted a great deal of attention from psychologists and biologists. The nuances of trust, the levels of trust and the manipulation of trust are subjects of ongoing socio-psychological research, and such studies are revealing some fascinating insights at a biological level. A particularly interesting study was published in 2005 in the journal Nature and reported the effects on trust of giving people oxytocin. Oxytocin is both a hormone and a neuropeptide, a substance used by neurons to communicate with each other. We met oxytocin in Chapter 6 where we explored its role in social bonding. Levels of oxytocin increased in individuals who were gossiping and so increased their level of bonding, which feels like a potentially important step towards building trust. Oxytocin is also postulated to be involved in perhaps the most fundamental of all social bonding, the bond between a mother and infant.4 By locating oxytocin receptors in the brain we can begin to get more of an idea of what functions oxytocin serves, and doing this points us yet further towards a possible role in developing trust. Studies of non-human mammals reveal that oxytocin receptors are located in regions of the brain involved in pair-bonding, maternal care, sexual behaviour and the ability to form social attachments. The implication of this receptor mapping is that oxytocin may be involved in allowing animals to overcome a natural avoidance of physical proximity and enable what is called ‘approach behaviour’. The authors of the 2005 Nature paper went a step further and hypothesised that oxytocin might be involved in promoting ‘prosocial behaviour’ in humans, specifically trust.5
To investigate the role that oxytocin might play in developing or maintaining trust, the experimenters set up a game based on financial investment. Two subjects played the game and adopted the role of either investor or trustee. Although the game was played with ‘monetary units’ (MUs) these were exchanged into real money at the end of the game (a whopping 0.4 Swiss Francs per MU), so there was a real-world incentive to do well. Both parties start with 12 MUs and to get things rolling the investor has a decision to make: how much money to transfer to the trustee? This is where things get interesting because the experimenter triples that amount before telling the trustee what was transferred. So, let’s say the investor decides to go ‘all in’ and transfer 12MUs to the trustee. The trustee is told that they now have 48MUs (their original 12MUs and the tripled transfer from the investor). The trustee now has a decision to make: how much to ‘back transfer’ to the investor? The back transfer isn’t tripled but can be anything between 0MUs (not so good for the investor) or their full balance (48MUs and not so good for the trustee). The trustee can honour the investor’s trust by sharing the money gained such that both players end up better off, or the trustee can violate the investor’s trust and simply pocket the profit; they don’t even need to pay off the investor’s 12MUs.
A rather dry experimental system perhaps, the trust game was spiced up by the prospect of actual money (a rare commodity in scientific experiments), the jeopardy of distrust and the application of oxytocin to half the players via a nasal spray. The other players received a placebo nasal spray that contained all the same ingredients but was lacking oxytocin. The hypothesis that oxytocin increases trust makes some very simple predictions, the first of which is that investors that have received three puffs of oxytocin up the nose should be more trusting (give up more of their 12MUs) than those that have been inhaling a placebo. If, on the other hand, oxytocin makes no difference to trust then we should expect no difference in initial investment between the oxytocin and the placebo group. In fact, what they found was that 13 of the 29 investors (45 per cent) who took oxytocin showed maximum trust (going all-in and investing 12MUs) compared with only 6 of the 29 (21 per cent) in the placebo group. Low levels of trust, defined as transfers below 8MUs, were relatively common in the placebo group, accounting for 45 per cent of transfers, but much rarer in the oxytocin group, accounting for just 21 per cent. On average, oxytocin takers had 17 per cent higher transfers overall, with a median amount (the ‘middle’ value if all the amounts were lined up in order) transferred of 10MUs compared with 8MUs in the placebo group. Interestingly, even the placebo group seemed pretty trusting, giving up two-thirds of their money. This perhaps gives us some insight into how people end up falling for get-rich-quick schemes.
There was a slight complication in all this because we could interpret the results of administering oxytocin not in terms of increasing trust but as a lowering of the threshold for risky behaviour; oxytocin makes you more likely to go all-in because, you know, why not? To investigate this possibility the experimenters ran the same basic experiment, but instead of there being a person acting as a trustee and making decisions, the back transfer was determined at random based on the back transfers that took place during the first experiment. So, investors, on average, had the same level of risk but there was no social interaction with a trustee and so no need for interpersonal trust. In this experiment, the oxytocin and the placebo groups did not differ and both groups were actually the same as the placebo group in the first experiment. Only the oxytocin group in the first experiment, where interpersonal trust was required, was different in any way and that difference was that they were more trusting.
This seemingly clear link between oxytocin and trust caused, predictably, a big stir when it first came out. It led to oxytocin being dubbed the ‘trust molecule’ and it spawned several follow-up studies using the same method. It was at this point that the problems emerged. Some follow-up studies failed to replicate the effect of inhaling oxytocin on trust in similar games, and other studies, using questionnaire approaches, also drew a blank. In 2015 a major review of the link between oxytocin and trust looked at three lines of evidence: the trust experiments; individual oxytocin levels in blood and levels of trust in that individual; and genetic differences in the oxytocin receptor gene that might be associated with trust. Sadly, for those of us who enjoy a good story, their conclusions were that ‘the cumulative evidence does not provide robust convergent evidence that human trust is reliably associated with OT [oxytocin] (or caused by it)’.6
So much for the simple story. However, recent work is showing that oxytocin does seem to have a role to play in developing trust, but that role is complex, context-dependent and likely involves learning. Scanning the brains of male subjects undertaking tasks within an fMRI (functional Magnetic Resonance Imaging) machine, which measures and maps brain activity, allows researchers to see what parts of the brain are active at any given time.7 These scans have shown that oxytocin, rather than promoting general feelings of generosity that might induce trust, actually seems to reduce ‘feedback learning’ by subduing communication between parts of the reward circuitry in the brain that we met in Chapter 8 when we explored addiction. The result of this reduction in learning is that, under the effects of oxytocin, we are more likely to go with our underlying beliefs, rather than make more nuanced decisions. As the authors conclude, oxytocin alters ‘the brain’s encoding of prediction error and therefore its ability to modulate pre-existing beliefs’. This permits our innate biases (whether positive or negative) towards a new social situation to dominate, which could promote more trusting behaviour.
Trust, despite some problems replicating the early research, does appear to have a biochemical basis in the brain. As we might expect with such a complicated concept, that biochemical basis is equally complex and involves a number of aspects of the brain including learning and the reward circuitry. The evolution of trust is also likely to be complex but nonetheless we can, with some confidence, suggest that our ability to trust is an evolved characteristic. A study published in 2008 measured identical and non-identical twins’ performance in a trust game, and the results strongly indicated that ‘humans are endowed with genetic variation that can partially account for differences in trust and trustworthiness’. Trust can be seen as cooperative behaviour while being distrustful is non-cooperative, and the authors of the twin study tentatively concluded that the existence of genetic variation in levels of trust lends support to the idea that cooperating and non-cooperating behaviours could co-exist in humans. The selection pressures for trusting behaviour are likely to be based on the advantages provided by cooperating and working together, highly social behaviour that requires individuals to trust each other. On the other hand, there are situations where personal advantage might be gained by being distrusting. The fact that trust seems to be linked to our reward circuits and to learning, and is generally context-dependent (we don’t simply have a universal ‘trust level’ but can vary trust depending on the situation and the individuals with which we interact), indicates a complex evolutionary history tightly interwoven with our developing social networks (Chapter 6). Perhaps it was actually the evolution of trust that allowed our ancestors to develop more sophisticated social interactions and to make full use of the power that larger groups can provide? As a simple example, it is hard to imagine hunting a large animal in a group without trusting the person holding a sharp spear and standing right next to you.
I am all too aware here that I might be straying in to ‘Just So’ territory with some hand-waving about putative selection for and evolution of trust. The potential biochemical nature of trust, twin studies showing heritability and the importance of trust in so many aspects of our lives, certainly lend support to an evolutionary history though, and further evidence comes from theoretical approaches. Trust and cooperation have been studied using the modelling approach called game theory. Game theory allows us to consider the interactions between individuals and has proved to be a powerful and useful approach in understanding more complex aspects of social behaviour and evolution.8 The models produced by game theory studies have shown that trust and trustworthiness can evolve, and real-world factors can result in the sort of variations of trust that we see in humans. We are still very much at the early stages of learning about the origins of human trust, but recent work on chimpanzees suggests that comparison with non-human primates and other social animals may yield valuable insights. One such study presented chimpanzees with a version of the human trust game and showed that individuals trusted their friends (determined by observing the chimps and working out who hung out with whom) more than their non-friends.9 Overall, I think we are on firm ground to suggest that trust is an evolved characteristic in humans, albeit one that is complex.
Believing fake news is an example of misplaced trust. Until the rise of the internet, and more especially the rise and domination of social media and other platforms for disseminating news, we got most of our information about the wider world around us through printed newspapers and limited TV and radio news. We may also have got some of our information through our friends (trusted sources), who may have read or seen something interesting and passed it on to us face to face. Organisations like the BBC were trusted and information passed on by them was considered ‘real news’. The political affiliations of many newspapers were clear but there was rarely the thought that they were actively lying to us. More importantly though, the ways in which we received news were limited and to a certain extent austere; it was something of an event to sit down and watch ‘the news’. Nowadays this is not the case. The internet may have compromised the financial security of traditional physical newspapers, but it has created a truly dizzying splendour of online news channels. Meanwhile, cable TV, deregulation and widespread acceptance of advertising have driven up the amount of news coverage on our TV screens from a scheduled short bulletin being read every so often by a dour man in a suit to the glitz and glamour of 24 hours a day, seven days a week rolling coverage across multiple channels. Our social media feeds are stuffed full with links to online news stories and current affairs videos. In short, we are inundated with information coming at us from all directions in a complex and high-volume information stream. In no particular order we have state-sponsored TV, online news channels, old channels, new channels, satirical news, commentary, review, breaking news, ongoing news, throwback news, dense analysis, superficial analysis, zero analysis, space fillers, advertorials, editorials, ‘conspiratorials’, video, text, cartoons, infographics and interactives. Just as we saw in Chapter 8 with drugs and alcohol, the modern world is one of greatly enhanced potency and availability.
Simply trusting everything we read or is told to us by someone who ‘looks the part’, like a newsreader, fits the definition of gullibility. Being gullible, being too trusting of information given to you, is a quality often associated with children. A three-year-old will look at you wide-eyed as you tell them about Santa Claus or the tooth fairy because you are their parent and you have formed a trusted bond; they believe you. This is handy for evolutionary success because children who do the exact opposite of what their parents tell them are far less likely to make it to adulthood. Children have an awful lot to learn in a very short timeframe and believing what they are told, being gullible, is a useful time-saver. However, believing what your parents tell you might be a good survival strategy initially, but at some point there needs to be a degree of criticality because gullibility can only take you so far. Studies of children and their relationships with true and false information give us interesting insights into how rapidly we develop into less gullible and more critical thinkers as we age, and can also throw some light on why we fall for fake news.
There are several ways that a child can assess the accuracy of information they are being given by someone. One way is to assess the accuracy of previous information given by that person. If someone seems to have been accurate and told the truth in the past then it seems reasonable to assume that they are being accurate in the present. A wealth of studies of children show that they do indeed prefer to learn from those who have previously given accurate information, rather than from those who have previously been inaccurate.10 This is of course equally valid for adults; if someone or something, like a TV news outlet, has been accurate in the past why doubt it now?
With TV news especially, there may also be a strong ‘halo effect’ at play. Halo effects describe the tendency for impressions created in one area to positively influence opinion in another area. We tend to package certain traits together; for example, we judge more attractive people to have more socially desirable traits like intelligence,11 altruism and, crucially, trustworthiness. Studies have repeatedly shown that we also prejudge attractive people to be ‘better’, to hold more prestigious and secure jobs and to have happier lives. The cognitive bias inherent in the halo effect might at first glance, and even second glance, seem irrational; how can we possibly judge unseen characteristics of an individual simply by looking at their face? Once again, our evolutionary heritage might provide an answer.
In many animal species one sex, typically males, compete for the attention of the ‘choosy’ sex, typically females. The reason why it is often males that compete and females that choose is linked to the large investment that females put into producing their reproductive cells (eggs) versus the relatively smaller investment males put into sperm. This investment differential can be further enhanced by females providing some level of parental care beyond that provided by males. Enhanced female care can begin at the very start, by providing a rich yolk to the egg (for example in birds, reptiles, amphibians, fish) or by nurturing the young within their own bodies (as is the case in mammals). The upshot of this increased female investment and parental care is that while males are nearly always ready, willing and able to provide copious amount of sperm, females are not necessarily available to receive it; eggs take time to produce, or they may be pregnant, sitting on eggs, or lactating young. In a population with an equal sex ratio this imbalance leads to more males being available than there are reproductively receptive females. The imbalance in what is called the operational sex ratio and the consequent surfeit of males allows females to be choosy. Females that pick the ‘best’ males, healthy males with ‘good genes’ that will lead to lots of healthy fecund offspring, will do better than those that are less choosy. In species where males provide the larger investment, such as seahorses and pipefish, where males keep eggs and hatchlings safe in a brood pouch accessed through an opening on their abdomen, it is males that will be in short supply. In these more unusual ‘sex role-reversed’ species it is females that compete for males. This ‘competition/choice’ process leads to sexual selection whereby one sex can end up with adornments and weapons that make them better able to compete with other males (antlers, horns, strength), or more attractive to females (peacocks’ tails, bright colouration, crests), or both.
One aspect of sexual selection that is potentially relevant to the halo effect is the concept that has become known as ‘fluctuating asymmetry’. We are symmetrical animals with our left side and right side more or less identical, at least externally, through a line of symmetry that runs between our eyes and passes through the groin. Studies have shown that more symmetrical faces are judged to be more attractive and this is the basic principle behind the ‘How Attractive Am I?’ clickbait that haunts Facebook. We are not, however, perfectly symmetrical, and deviations from the ‘ideal’ of symmetry creep in during our development as different genetic and environmental pressures are exerted. Symmetry therefore might indicate ‘developmental stability’, and be reflective of overall genetic quality; ‘better’ individuals are more likely to be symmetrical. In animal studies, including insects, birds and mammals, symmetry of traits (where structures on the right side are compared with those on the left) has been shown to correlate with mating success; symmetrical males are picked more by females. In human studies it seems that symmetrical faces are not just a feature that we rate as more attractive. There are studies that link facial asymmetry with lower intelligence, lower levels of extraversion (how outgoing and sociable people are) and more intellectual impairment with ageing. One study even found that criminal offenders had lower facial symmetry than non-offenders.12 Delving deeper, studies have found links between lower levels of facial symmetry and a greater vulnerability to parasites, as well as lower levels of disease immunity. Correlations have also been found with mental health, with lower symmetry being associated in one study with higher levels of depression in men. Several studies have associated lower symmetry with schizophrenia. For obvious reasons such studies are often controversial both socially and scientifically. A recent review of the literature in this area concluded that while some associations, such as with schizophrenia, are repeatable and robust, other studies ‘have not always followed best practices’ when it comes to measuring features and determining the tiny differences involved.13 Nonetheless, the balance of evidence allows us to say that symmetrical faces are judged as more attractive and that facial symmetry is potentially indicative of a wealth of underlying characteristics that relate to possible ‘mate quality’. Picking a good partner is an important decision for humans, and when viewed through an evolutionary lens the halo effect is not so irrational. So, the next time you watch an attractive, well-groomed newsreader, be mindful of your cognitive biases; the halo effect means you are more likely to think of them as trustworthy and to believe whatever they are saying.
Halo effects also apply to non-living entities including corporate brands, as companies like Apple and McDonalds know all too well. This may be be a factor in news output too. Glossy, well-produced and generally factually accurate news programmes with all the accoutrements of the modern news format, such as infographics, outside broadcasts and in-studio punditry, all conspire to create a halo effect for any similar output. Even if the information is inaccurate, the corporate halo effect lures us in and the individual halo effect makes us want to trust the newscaster.
The idea that fake news is misplaced trust is strengthened by examining the relationship between trust and authority figures, including doctors, scientists and politicians. Our tendency to obey authority figures is most commonly illustrated by reference to the famous Millgram experiment. In these experiments of the 1960s Stanley Millgram instructed participants to administer what they believed could, at times, be fatal electric shocks to another participant (actually an actor) under the guidance of the ‘authority figure’ (the experimenter). Criticism of interpretations of the experiment have led to a different framing, moving from ‘blind obedience’ to ‘engaged followership’ where ‘people are prepared to harm others because they identify with their leaders’ cause and believe their actions to be virtuous’.14 In fact, recent work further re-frames the study and suggests that participants sought to justify their actions not through some sense of virtuousness but because they doubted that anyone had really been harmed.15 Other more subtle studies have been carried out to demonstrate what has become known as the authority principle though and the findings are consistent; we tend to trust those that we regard as being ‘authority’ figures. Here again the halo effect rears its unwelcome head, and we tend to regard the opinions of authority figures as more reliable than our own even if they are expressing an opinion on matters outside their sphere of expertise. The traditional authority figures of such experiments are doctors, lawyers, scientists or law-enforcement officers, but such implicit trust in those we regard as being ‘in charge’ or ‘expert’ translates very easily to actors and other celebrities and can certainly be applied to newsreaders, media commentators and others involved in conveying information through the media.
If you’re going to lie, be confident
Another source of information available to us when deciding whether we are being told the truth is confidence. Studies of children show that we develop some skills for gauging confidence early. Three- and four-year-olds are aware of the differences between ‘guessing’ and ‘knowing’ and are more likely to believe information given to them by someone who ‘knows’ rather than by someone who is expressing a degree of uncertainty. The ability to spot confidence and associate it with knowledge actually starts much earlier in life, before children have firm verbal skills. Two-year-olds have been shown to be aware of, and sensitive to, non-verbal indications of confidence. These younger children are more likely to imitate someone’s actions if that person has expressed those actions with confident body language. What is interesting is that as children get to the age of four they start to be able to make judgement calls based on a combination of accuracy and confidence. They prefer to believe information given to them by a confident individual, but if that information turns out to be inaccurate they start to prefer hesitant but previously accurate individuals.
An evolutionary mismatch
What is starting to emerge are the perfect evolved intellectual conditions for the modern world of fake news to take root. First, as a social species we have evolved trust and most of us tend towards being trusting unless evidence suggests otherwise. Remember, most investors in the trust game were happy enough to give away two-thirds of their money. Second, as a species that is essentially helpless for much of our early lives, we have neural mechanisms that even before we can talk allow us to interpret the ‘truthfulness’ of those around us. Innate neural mechanics, gradually honed through experience, lead us towards believing those who present information confidently. Halo effects, perhaps based on deep-rooted selective advantages linked to mate choice, cause us to be dazzled by certain individuals and corporations, which of course tend to present themselves towards the ‘extreme’ end of the confidence spectrum. We also seem to have an innate, and highly sensible, tendency to believe sources of information if they have previously been shown to be accurate. When we pull all these factors together and hose them down with a high-volume stream of information originating from previously trusted, confident and attractive sources, then it is little surprise that we fall for falsehoods. Mix in a deliberate intention to deceive, perhaps politically motivated, and then we are two-year-olds on a play mat willing to follow the grown-up with the most confident body language. All that is before we consider another set of evolved tendencies that have an important part to play in the story of fake news.
The importance of ‘belonging’
Our cognitive biases towards confidence, previous accuracy and, through halo effects, attractiveness are important in our evolved predisposition towards believing fake news, but are only part of the story. There is another powerful part of our evolutionary heritage that makes us particularly vulnerable to fake news and perhaps to other seemingly incredible sources of information: group identity and confirmation bias. Social identity theory was developed in the 1970s and 1980s by social psychologists interested in understanding intergroup behaviour. It is a big topic, has been the subject of many papers, articles and books and has not been without some controversy. At its heart though is a simple concept that has survived the test of time and rigorous debate: our behaviour is often a compromise, balancing interpersonal and intergroup interactions.
We identify both as individuals and as members of various groups. That group might be a close family group, an extended family, your school, your university, a football team, a political party, a religion, a nation and so on. By belonging to that group, and identifying as part of it, you become part of the ‘in-group’ and are likely to display in-group biases. These biases mean that you will give preferential treatment to others in your group while also potentially biasing against ‘out-groups’. An important aspect of social identity theory is that members of the in-group will seek out negative aspects of the out-group and in so doing bolster the in-group’s self-esteem.
Social identify theory is a concept sometimes summed up by the word ‘tribalism’. Although this simplifies the issue, and brings in unhelpful or even offensive connotations, it is nonetheless useful shorthand for a concept that has proved its worth in social psychology. It is not difficult at all to see why we might have evolved strong tendencies for behaviours that reinforce whatever group we are in. As a social species we rely more than we often think on interpersonal interactions and, certainly at times in our evolutionary history, intergroup conflict would have been a real and important selection pressure. Group identity and enhanced trust in the members of that group will have been important components of group, and thereby individual, success. So, how does group identification fit in with fake news? Well, the tendency to be biased positively towards those we perceive as being in our group can extend to being biased towards information that fits our preconceptions about our group and others. This is likely to be an especially powerful cognitive bias when the information being presented is positive about people in our group (the in-group) or negative about those in out-groups. Being swayed towards what we already think is the basis of another big issue with fake news: confirmation bias.
Fake news preys upon confirmation bias, potentially deliberately if falsehoods are being peddled as part of a wider agenda. Confirmation bias is a serious problem and it affects how we search for information, how we interpret information that we receive and how we remember that evidence later. Bias in the way that we think is not a modern phenomenon. Dante Alighieri defines it beautifully in the early fourteenth century in the Divine Comedy when Thomas Aquinas meets him in Paradise and says ‘opinion – hasty – often can incline to the wrong side, and then affection for one’s own opinion binds, confines the mind’. Understanding the precise neurological mechanisms that underpin our thinking, and our biases, is one of the great works-in-progress of modern biology but it isn’t too difficult to frame some plausible scenarios that would support an evolutionary, and therefore heritable, basis for confirmation bias. Strengthening group membership and enhancing social identity and cohesion could certainly provide a strong selective advantage for biased thinking processes, provided that the costs did not exceed the benefits. Thought processes are time-consuming and shortcuts that allow our brains to process information rapidly could be favoured, again if the costs are small compared to the benefits. Such shortcuts use common-sense rules of thumb and are known as heuristics. One suggestion is that we aren’t so much interested in the absolute truth of a matter but rather we seek to avoid the most costly errors. Let’s imagine two friends who trust each other, as friends often do. One friend suddenly has the thought that the other is dishonest or untrustworthy. The suspicious friend now has a choice: either question the integrity of the other person and erode the friendship (potentially very costly), or seek evidence that confirms the friend’s honesty. If it is less costly to go along with the status quo (and friends are often hard to come by), then it makes sense to be biased because it could cost less than destroying the alliance.
We can see the importance of both confirmation bias and group identity in the acceptance and spread of fake news on social media threads. Climate change is an excellent example. The evidence that climate change is happening and that it is anthropogenic in origin is overwhelming. That Earth’s climate is warming is accepted as fact by more than 97 per cent of actively publishing climate scientists. Despite this consensus and the weight, strength and diversity of evidence supporting it, climate sceptics still abound. When a post appears that purports to show that climate change is not happening, or when the President of the United States suggests it is a hoax perpetrated by China, you can guarantee that this ‘news’ will be posted gleefully by climate sceptics because it confirms their bias. Reading the comments arising from these posts, as I often do, confirms the in-group/out-group nature of these debates, which play out very much as a ‘them versus us’ conflict. Typically, such conversations only serve to polarise the groups further, leading to entrenchment and even more aggressive sharing of the fake news. Eager to support their argument further, people will search out yet more confirmatory evidence for their point of view. You can even see comments between friends within these threads referring to ‘secret’, private groups where more confirmatory news is being shared. These exclusive ‘echo chambers’ further feed the bias. Much of this activity perhaps stems from our evolved tendencies towards trust and reinforcing group identity, and from cognitive biases that develop from useful heuristics that are usually beneficial, or at least cost-reducing.
Our ability to fall for misinformation is not new, but the modern world, with its enhanced potency and availability, provides so much more opportunity for our evolved vulnerabilities to be exploited. We are not going to evolve our way towards increasing criticality, but we can learn to adapt. With an increasing awareness that not all we read is true, and an increasing exposure to the issues that fake news presents, we can perhaps be optimistic that in the future we may be more discerning information consumers. This is especially pertinent for those who are growing up as ‘digital natives’, with no memory or experience of a world before the internet, social media and modern fake news. A greater understanding of why we believe misinformation, built perhaps on an understanding of the cognitive mechanisms we have evolved to facilitate social living, is going to be hugely important in the future if we are to function in the modern digital age. But what does ‘the future’ really mean to us and what might it hold for an upright ape increasingly finding itself unfit for purpose? These are the questions that I will try to answer in the final chapter.