Getting to the truth has never been easy, but these days it seems to be harder than ever to know where to find it. Claims of ‘fake news’ and ‘alternative facts’ are everywhere – not least in the White House. Channel 4 even launched a week dedicated to fake news. It’s hard to keep up.
A major component of evidence is statistics. But context is everything and with the emergence of ideas like ‘alternative facts’ and ‘fake news’ statistics can be made very easily to support stories that have little or nothing to do with truth. People will believe what they want to hear.
The debate between objective and relative truth is, of course, an old one, but a recent Guardian ‘long read’ has questioned whether or not statistics still have the ability to accurately represent the world at all, suggesting that privatised data is taking over and that data analytics and technical statistical analysis is being replaced by automatically accumulated data. So where can we seek truth? In the currency of the usual statistics and the context in which they’re presented, or in the data that we now inadvertently produce everyday?
Historically statistics have been used to present information as objective, but more recent techniques of data analysis allow companies, individuals and even political parties to produce much more subjective evidence. And this can be used to support ever more subjective narratives reflected in the context and language with which they are represented.
Take science. When science claims something to be true it is the extent to which the narrative, supported by the evidence, can be agreed upon in the scientific community that gives the evidence its force, its context. Replication of experiments, scrutiny of results and broad agreement on a shared narrative are key attributes of ‘truth’.
Anyone can dispute one or two statistics but once statistics are woven together into a narrative, they present a tapestry of evidence that will bear the test of scrutiny by the widest audience. It is at this point that the story is credible and the statistics are interpreted as truth.
Subjective data can only support subjective truth. And that’s no kind of truth at all.