RALEIGH – In this season of political claims, counterclaims, and calling of names, separating truth from error or falsehood can be a challenge. When in doubt, apply the Bacon Test.
(Okay, let me pause a moment so you can salivate. Simpsons fans, feel free to do your Homer Simpson drool at this point.)
Now, the Bacon Test to which I refer does not involve fried pork. It bears the name of Sir Francis Bacon, the English statesman and jurist who became one of the most influential philosophers of science.
In his description of the scientific method, Bacon emphasized the role of falsification in the pursuit of knowledge. Even if you witness two things happening simultaneously 19 times, that doesn’t necessarily mean they will occur together the 20th time. Nor do those initial 19 observations establish which is causing the other, or even if either is causing the other. Perhaps there’s something else you aren’t seeing that is causing both of them.
What a careful thinker can do is disprove a potential theory or explanation by showing a case where a proposed causation could not possibly have occurred. Disproving falsehoods is a valuable exercise. As the Great Detective was wont to say: “When you have excluded the impossible, whatever remains, however improbable, must be the truth.”
To test possible relationships between two objects or events, Bacon suggested that we look for three conditions: presence, absence, and degree. If Object A is present and Object B is not, then we can be sure that Object B is not a necessary condition for explaining the existence of Object A. And if Object A is absent while Object B is present, we can be sure that Object B is not a sufficient condition for explaining the existence of Object A.
Finally, if both Object A and Object B are present at the same time, we should examine the relative proportions of each for the purposes of establishing the degree to which the two things are correlated. If we witness 20 instances in which A and B are present at the same time, but there is no consistent statistical pattern in the degree of relationship between the two, it is difficult to conclude that one explains the other.
If there is a statistical correlation, that still doesn’t mean you have established causality. Additional research is needed. For example, there may be a third factor, Object C, that is the real cause of A and B. Here is where our old friend ceteris paribus – “all other things being equal” – comes into play. You have to hold other potential explanations constant in order to test the potential causal relationship between A and B.
The conclusions from such a statistical analysis can never be certain. They offer probable correlations and causes. But human action, including the act of voting, does not require certainty. Indeed, if your standard is certainty, you’ll be led to inaction, not action. In the real world, we typically act on probabilities, not certainties.
In assessing political rhetoric, then, the Bacon Test can help us judge whether politicians are making valid claims about the past, present, and future. You can throw out the logical impossibilities – the supposed causes of North Carolina’s economic woes that history shows to be neither necessary nor sufficient, for example – and choose to lend credence to those claims that have the strongest empirical support, that are the most likely to be true.
That doesn’t mean, by the way, that those whose claims don’t hold up should be called “liars,” as has become the current fashion. Politicians, just like other human beings, are capable of being mistaken without intending to mislead. And to say that a claim is probable – supported by empirical data and not yet disproven – is not necessarily to establish it as an incontrovertible truth. It may be that the next experiment will be the one that disproves the claim.
Still, applying the Bacon Test can help us toss out some political claims as false, as I did the other day with regard to education. Hey, it’s a start.
Hood is president of the John Locke Foundation.