Tuesday, March 30, 2010

How to spot bullshit

10 tips for distinguishing between disguised ignorance and insight.

1) ‘Studies show’ – what studies?
Be cautious of claims made with confidence but without evidence. Some writers will try to lend their views credibility by referring vaguely to science or research, or by quoting unreferenced statistics. This disturbing Youtube video is a good example of unreferenced bullshit delivered with fancy graphics and a serious voice. Its first sentence:

According to research, in order for a culture to maintain itself for more than 25 years there must be a fertility rate of 2.11 children per family.

What research? The video is anti-Muslim propaganda, warning that Muslim immigrants are rapidly out-reproducing Europeans. It has well over 11 million views on Youtube alone – but it’s bullshit, full of demonstrably false statistics as well as others with no reference or source. The BBC do a good job in exposing its nonsense.

2) Anecdotal evidence can disprove a generalisation, but cannot prove it
If someone says that all Irish men are alcoholics, and you meet an Irish man who is not an alcoholic, your anecdotal evidence is enough to disprove this claim.

If you meet an Irish man who is an alcoholic, however, it does not prove his claim since there are millions of other Irish men. Be wary of people who use anecdotes as evidence to prove a theory. Anecdotes can disprove theories, and can help add colour and insight to a situation, but are not enough in themselves.
For example, remember this Australian clip depicting ‘stupid Americans’ interviewed on the street? The video shows Americans who cannot answer questions like ‘name a country beginning with U’. (‘Yugoslavia,’ one replied. Another suggested ‘Utah.’) Quite funny stuff, but not indicative of what the other 300 million Americans would have answered.

3) Watch out for overkill
Sometimes writers make claims so outlandish and extreme, so far removed from what most people already believe, and with such confidence, that there is a temptation to say, ‘well, if only half of this is true, then I’m convinced’. This is weak logic, don’t let wild accusations and claims move you half-way to acceptance when none of it is backed up.

4) Cherry-picked statistics are useless
American right-wing commentator Glenn Beck gives us a good example here:

Americans have a better survival rate for 13 of the 16 most common cancers than Europe. Take prostate cancer: 91.9 percent of men live through it, versus 73.7 percent in France and just 51.1 percent in Britain.

Beck was trying to show that 'European' socialised healthcare is worse than the American system. But by cherry-picking France and Britain, he carefully avoids Austria, where 86.1% of cancer victims survive, even with its heavily socialised health system. Shortly after Beck’s rant another report emerged showing that Finnish women have a higher cancer survival rate than American women – Finland also with a highly socialised health system.

Beck picks numbers out of the blue, numbers which sound impressive, but he does not place them in context. This is classic propaganda nonsense.

(Michael Moore’s Sicko goes one step further: cherry-picking anecdotes. He interviews Americans happily living in France, but not French people living in the US.)

Statistics must be placed in context. Supposing a pressure group claims that a certain number of people die of some disease, and uses this as evidence to call for a changed policy. The number of dead is emotive and disturbing, and moves people to action. But the number is worthless without context. How does this mortality number compare with other diseases? How expensive is the treatment – can we save more people suffering from different diseases for the same cost? How many people die in other countries from this disease? What proportion of the total population is affected by it, and is the proportion growing or shrinking?

Another example: campaigners often suggest that ‘world hunger’ is growing. This could be because the global economic system is pulling food from the poor and wasting it on the rich.

…Or it could be simply because the world population is growing, largely because food production has soared over the last century. Perhaps there are more hungry people today because in the past those people would have starved to death and disappeared from the statistics. Growing ‘world hunger’ could be a sign of success rather than failure.

5) Generalisations about ‘Europe’ or ‘The West’, conceal the truth
Europe, remember, is a continent ending, arbitrarily, at the Urals and Caucasus mountains, and including Iceland even though it is geographically closer to North America. Europe includes a socialist autocracy (Belarus), a theocracy (Vatican City), and several traditionally Muslim-majority countries (Albania, Kosovo and Turkey, since a small section of Turkish territory lies inside Europe). Most Russians are also Europeans, living west of the Urals. Most European countries never had colonies, and many of them were colonised by their European neighbours. Some European countries are in the EU but not NATO, some in NATO but not the EU, some in neither.

Taking all this into account, it is senseless to talk of ‘European policies’ or ‘European culture’, yet serious commentators do this all the time. This can be deeply misleading, for the same reasons given for Glenn Beck’s speech. There is wide variation between European countries. Thus there is no European policy or European culture that is not also shared with non-European countries.

Commentators and politicians blur the edges to hide the truth. It is not difficult to be specific. Say ‘EU’, or ‘France and Germany,’ or ‘NATO’ when you mean those things, not Europe.

Europe is just one example, but this kind of generalisation is common for many groups, not least ‘the West’, ‘the East’, Africa, ‘the Muslim World’ and so on. Don’t take for granted that such collections of diverse nations are really accurate or useful, that there is really such a thing as a ‘Western worldview’ or an ‘Islamic policy’. It’s lazy, and deceptive.

6) Data, not celebrity, should rule
Trust no commentators, however likeable or well-respected: always expect data to back up their claims. Journalists sometimes introduce a commentator by referencing his achievements to give credibility.

'IRELAND MUST “suffer” through at least five years of economic weakness before recovering to match average euro-zone growth, according Nobel Prize-winning economist Paul Krugman,' reads one Irish Times article from 2009. Yet another Nobel Prize-winning economist is Friedrich von Hayek, whose views radically differ from Krugman’s. So which Nobel Prize-winning economist are we meant to trust?

Neither, perhaps, without looking at the evidence they offer.

The same goes for people or organisations one usually disagrees with. Sometimes inconvenient arguments are dismissed because of the person they come from: ‘typical right-wing propaganda’, ‘known Communist-apologist’, etc. Arguments should not be dismissed because of the person who makes them, but rather because of their weak evidence or logic.

In practice, to be fair, we don’t have time to listen to every argument going, so we filter out those arguments coming from people with a long-standing lack of credibility. But just be cautious about it – sometimes the wrong people believe the right thing.

7) ‘Yeah but…’
One common technique for escaping legitimate blame for some abuse or atrocity is to confuse the issue by pointing at the accuser. For example, the Chinese government released a Human Rights Record of the US in 2004. American governments have for years criticised China for its human rights abuses. Chinese government’s reply: ‘yeah but…’

Yeah but, ‘The United States should take its own human rights problems seriously, reflect on its erroneous position and behavior on human rights, and stop its unpopular interference with other countries' internal affairs under the pretext of promoting human rights.’ So does that mean there are no human rights abuses in China? Not at all. In fact it is irrelevant to the discussion of Chinese government human rights abuses. This sleight of hand flips the debate away from sensitive topics by focusing on irrelevant external issues: a useful way to bullshit people.

8) Don't take my word for it
These are useful ways to detect deception, but don't take my word for it. Access data directly, skipping the middle men of media and politics. The internet makes this ever more convenient with amazing sites like Gapminder. Go forth and research yourself.


  1. Excellent Post. Very well constructed and really enjoyed it

  2. Thanks very much Graham! Please feel free to share it with anyone else who may be interested.

    Another example of bullshit that I forgot to mention is the confusion of correlation with causation.

    For example, sometimes people say that the US has both liberal gun laws and high rates of violent crime, therefore gun control is necessary to prevent violent crime.

    They confuse correlation with causation. Switzerland also has liberal gun laws, but low crime rates. Depending on the example we choose to look at, we can spin this debate any way we want.

    So readers should be careful when they see correlations being interpreted as evidence of some process.

  3. brilliant blog post and a great read

  4. This comment has been removed by the author.

  5. Every university student should read this or something like it.


Note: Only a member of this blog may post a comment.