Prejudice is an essential predictive tool

Prejudice has a bad name but it is an essential tool evolution has given us to help our survival. It is not a bad thing in itself, but it can cause errors of judgement and misuse so it needs to be treated with care. It’s worth thinking it through from first principles, so that you aren’t too prejudiced about prejudice.

I like a few people, dislike a few others, but don’t have any first hand opinion on almost everyone. With over 7 billion people, no-one can ever meet more than a tiny proportion. We see a few more on TV or other media and may form a narrow-channel opinion on some aspects of their character from what is shown in their appearances. Otherwise, any opinion we may have on anyone we have not actually met or spent any time with is prejudice – pre-judgment based on experiences we have had with people who share similarities.

Prejudice isn’t always a bad thing

Humans are good at using patterns and similarities as indicators, because it improves our chances of survival. If you see a flame, even though you have never encountered that particular flame before, you are prejudiced about how it might feel if you stick your hand in it. You don’t go all politically correct and assume that making such a pre-judgment is wrong and put your hand in it anyway, since it may well be a very nice flame that tickles and feels good. If you see a tiger running towards you, you probably won’t assume it just wants to cuddle you or get stroked. Prejudices keep us alive. Used correctly, they are a good thing.

Taking examples from human culture, if a salesman smiles at you, you may reasonably engage some filters rather than just treating the forthcoming conversation like any other. Similarly, if a politician promises you milk and honey, you may reasonable wonder who will pay for it, or what they are not telling you. Some salesmen and politicians don’t conform to the prejudice, but enough do to make it worthwhile engaging the filters.

Prejudices can be positive too. If you see some nice strawberries, you probably don’t worry too much that they have been poisoned. If someone smiles at you, you will probably feel warmer emotions towards them. We usually talk about prejudice when we are talking about race or nationality or religion but all prejudice is is pre-judgement of a person or object or situation based on any clues we can pick up. If we didn’t prejudge things at all we would waste a great deal of time and effort starting from scratch at every encounter.

Error sources

People interpret situations differently, and of course experience different situations, and therefore build up quite different prejudice databases. Some people notice things that others don’t. Then they allocate different weightings to all the different inputs they do notice. Then they file them differently. Some will connect experiences with others to build more complex mindsets and the quality of those connections will vary enormously. As an inevitable result of growing up, people make mental models of the world so that they can make useful predictions that enable them to take advantage of opportunities and avoid threats. The prejudices in those models are essentially equations, variables, weightings and coefficients. Some people will use poor equations that ignore some variables completely, use poor weightings for others and also assign poor quality coefficients to what they have left. (A bit like climate modelling really, it is common to give too high weightings to a few fashionable variables while totally ignoring others of equal importance.)

Virtues and dangers in sharing prejudices

People communicate and learn prejudices from each other too, good and bad. Your parents teach you about flames and tigers to avoid the need for you to suffer. Your family, friends, teachers, neighbors, celebrities, politicians and social media contacts teach you more. You absorb a varied proportion of what they tell you into your own mindset, and the filters you use are governed by your existing prejudices. Some inputs from others will lead to you editing some of your existing prejudices, for better or worse. So your prejudices set will be a complex mix of things you have learned from your own experiences and those learned from others, all processed and edited continually with the processing and editing processes themselves influenced by existing and inherited prejudices.

A lot of encounters in modern life are mediated by the media, and there is a lot of selective prejudice involved in choosing which media to be exposed to. Media messages are very often biased in favour of some groups and against others, but it is hard to avoid them being assimilated into the total experience used for our prejudice. People may choose to watch news channels that have a particular bias because it frames the news in terms they are more familiar with. Adverts and marketing generally also have huge influence, professionally designed to steer our prejudices in particular direction. This can be very successful. Thanks to media messages, I still think Honda makes good cars in spite of having bought one that has easily had more faults than all my previous cars combined. I have to engage my own rationality filters to prevent me considering them for my next car. Prejudice says they are great, personal experience says they are not.

So, modern life provides many sources of errors for our prejudice databases, and many people, companies, governments and pressure groups try hard to manipulate them in their favour, or against others.

Prejudice and wisdom

Accumulated prejudices are actually a large component of wisdom. Wisdom is using acquired knowledge alongside acquired experience to build a complex mental world model that reliably indicates how a hypothetical situation might play out. The quality of one’s mental world model hopefully improves with age and experience and acquired knowledge, though that is by no means guaranteed. People gain wisdom at different rates, and some seem to manage to avoid doing so completely.

So there is nothing wrong with prejudice per se, it is an essential survival shortcut to avoid the need to treat every experience and encounter with the same checks and precautions or to waste enormous extra time investigating every possible resource from scratch. A well-managed prejudice set and the mental world model built using it are foundation stones of wisdom.

Mental models

Mental models are extremely important to quality of personal analysis and if they are compromised by inaccurate prejudices we will find it harder to do understand the world properly. It is obviously important to protect prejudices from external influences that are not trustworthy. We need the friendly social sharing that helps us towards genuinely better understanding of the world around us, but we need to identify forces with other interests than our well-being so that we can prevent them from corrupting our mindsets and our mental models, otherwise our predictive ability will be damaged. Politicians and pressure groups would be top of the list of dubious influences. We also tend to put different weightings on advice from various friends, family, colleagues or celebrities, sensibly so. Some people are more easily influenced by others. Independent thought is made much more difficult when peer pressure is added. When faced with peer pressure, many people simply adopt what they believe to be the ‘correct’ prejudice set for ‘their’ ‘tribe’. All those inverted commas indicate that each of these is a matter of prejudice too.

Bad prejudices

Where we do find problems from prejudice is in areas like race and religion, mainly because our tribal identity includes identification with a particular race or religion (or indeed atheism). Strong tribal forces in human nature push people to favour those of their own tribe over others, and we see that at every level of tribe, whether it is a work group or an entire nation. So we are more inclined to believe good things about our own tribe than others. The number of experiences we have of other tribes is far higher than it was centuries ago. We meet far more people face to face now, and we see very many more via the media. The media exposure we get tends to be subject to bias, but since the media we choose to consume is self-selected, that tends to reinforce existing prejudices. Furthermore, negative representations are more likely to appear on the news, because people behaving normally is not news, whereas people doing bad things is. Through all those combined exposures, we may build extensive personal experience of many members of a group and it is easy to apply that experience to new encounters of others from that group who may not share the same faults or virtues. One way to reduce the problem is to fragment groups into subgroups so that you don’t apply prejudices from one subgroup incorrectly to another.

Inherited experiences, such as those of columnists, experts brought into news interviews or even the loaded questions of news presenters on particular channels are more dangerous since many of the sources are strongly biased or have an interest in changing our views. As a result of massively increased exposures to potentially biased representations of other groups in modern life, it is harder than ever to maintain an objective viewpoint and maintain a realistic prejudice set. It is very easy to accumulate a set of prejudices essentially determined by others. That is very dangerous, especially bearing in mind the power of peer pressure, since peers are also likely to have such corrupted prejudice sets. We call that group-think, and it is not only the enemy of free thought but also the enemy of accurate prediction, and ultimately of wisdom. A mental model corrupted by group-think and inherited biases is of poor quality.

Debugging Prejudices

Essential maintenance for good mental models includes checking prejudices regularly against reality. Meeting people and doing things is good practice of course, but checking actual statistics is surprisingly effective too. Many of us hold ideas about traits and behaviors of certain groups that are well away from reality. Governments collect high quality statistics on an amazing range of things. Pressure groups also do, but are far more likely to put a particular spin on their figures, or even bury figures that don’t give the message they want you to hear. Media also put spins on statistics, so it is far better to use the original statistics yourself than to trust someone else’s potentially biased analysis. For us Brits, http://www.ons.gov.uk/ons/index.html is a good source of trustworthy official statistics, relatively free of government or pressure group spin, though finding the data can sometimes involve tricky navigation.

It is also a good idea to make sure you consume media and especially news from a variety of sources, some explicitly left or right wing or even from pressure groups. This ensures you see many sides of the same story, ensures you stay aware of stories that may not even appear via some channels, and helps train you to spot biases and filter them out when they are there. I read several newspapers every day. So should you. When I have time, I try to go to the original source of any data being discussed so I can get the facts without the spin. Doing this not only helps protect your own mental model, it allows you to predict how other people may see the same stories and how they might feel and react, so it also helps extend your model to include behaviour of other groups of people.

If you regularly debug your prejudices, then they will be far more useful and less of an error source. It will sometimes be obvious that other people hold different ones but as long as you know yours are based on reality, then you should not be influenced to change yours. If you are trying to work out how others might behave, then understanding their prejudices and the reasons they hold them is very useful. It makes up another section of the world model.

Looking at it from a modelling direction, prejudices are the equations, factors and coefficients in a agent-based model, which you run inside your head. Without them, you can’t make a useful model, since you aren’t capable of knowing and modelling over 7 billion individuals. If the equations are wrong, or the factors or coefficients, then the answer will be wrong. Crap in, crap out. If your prejudices are reasonably accurate representations of the behaviours and characteristics of groups as a whole, then you can make good models of the world around you, and you can make sounds predictions. And over time, as they get better, you might even become wise.

Advertisements

Leave a Reply

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s