The Power of Good Explanations

  • Add to list Added to list Added
  • Download PDFopens in new window
  • Cite
  • Share
  • Embed

Overview

This skill explores what it means to present well-reasoned explanations, and how seeking good explanations is a key aspect of scientific thinking and research.

It begins by differentiating between the types of reasoning present in arguments and explanations and the relationships between these. In general, an argument presents a line of reasoning that is claimed to support a particular conclusion while, by contrast, an explanation begins with something that is assumed to be true and then asks: how did things come to be this way?

It can be challenging to decide when an explanation qualifies as a piece of good reasoning. In a well-reasoned argument, the conclusion follows from the premises and can be seen to do so. But when it comes to the questions that explanations explore – of how and why things came to be a particular way – it’s possible to present countless reasonable-seeming answers.

As we’ll see, a good explanation in this context is one that manages to (a) explain everything we know while (b) being as simple as possible. By contrast, a bad explanation is one that either ignores inconvenient evidence and/or introduces unnecessary complexities. The principle that good explanations tend not to involve any unnecessary complexity is sometimes known as Occam’s razor, after the 14th-century cleric associated with its origins.

We’ll use the example of conspiracy theories to explore some of the features of faulty explanations – as well as the ways in which they can be appealing and can exploit confirmation bias. Confirmation bias describes the universal human tendency to seek out or attend to information that confirms whatever you already believe, or might wish to be true, in preference to information that contradicts such hopes and beliefs.

Resisting the distorting tendencies of confirmation bias entails, above all, putting explanations and theories to a meaningful test. We’ll look at the importance of seeking refutation over confirmation when it comes to explanations, and the significance of the philosopher Karl Popper’s work on explanations and scientific theories. You can also find more on how to fight against bias in The Battle Against Bias.

As Popper pointed out, simplicity does not and cannot guarantee the truth of any explanation. But we can, if we structure our investigations rigorously, test the predictive and descriptive power of different rival explanations in order to come up with the best possible understanding of what is going on.

It’s this incremental process of creating, researching and testing rival explanations that underpins the scientific method – and that can ultimately take us from testable hypotheses to theories that pull together numerous investigations into powerfully predictive explanations of the world.

Finally, this skill will look at the relationship between explanations, theories and hypotheses – and some of the potential confusions around correlation and causation that it’s important to watch out for in your own, and others’, thinking and research.

Suggested Readings

  • When it comes to truly lucid scientific explanations, few people did it better than Richard Feynman, whose seminal introductory lectures on physics can be accessed for free online at https://www.feynmanlectures.caltech.edu/
  • For a thorough, accessible and attractive guide to rigorous research across different disciplines, it’s hard to beat Ranjit Kumar’s Research Methodology (SAGE, 2019)
  • This is a short, provocative article about what makes for a good explanation by Matteo Colombo for Aeon magazine, and can be accessed at https://aeon.co/ideas/why-children-ask-why-and-what-makes-a-good-explanation
  • For an unbeatably elegant and lucid introduction to the thinking of Karl Popper, including his work on testing rival theories, see Bryan Magee’s Popper (HarperCollins, 1997)
  • The aptly named Crash Course channel on YouTube offers a great introduction to scientific theories (and the errors of pseudoscience) https://www.youtube.com/watch?v=-X8Xfl0JdTQ