Spam Filters and Predestination: What are the Odds?


The next time you peek at the thousands of unwanted emails caught by your spam filter, you can thank an Eighteenth Century Presbyterian minister by the name of Thomas Bayes. Were it not for his pioneering work, today’s vast world of digital communication might have been a lot slower in coming. While the science of probability has come a long way since he first published his treatise on the subject, the essential theory remains relatively intact. Basically, Bayes postulated that each time we add new information to previously known information we are better able to classify and predict. Applied to spam emails, for example, the more words and phrases we find in common among previously identified spam emails, the more certainly we can classify future emails that contain the same words and phrases, and weed them out.  Presented to the Royal Society in 1763 by his friend Richard Price after Bayes’ death, An Essay towards solving a Problem in the Doctrine of Chances cemented Bayes’ name in the field of mathematics. Today, we read of Bayes’ Rule, Bayes’ Theorem, Bayesian Efficien-cy, Bayesian Probability, Bayes’ Factor and Bayesian Inference, to name a few. And let’s just not get into his paper on asymptotic series…. How did a Dissenter Presbyterian raised in a non-conformist household committed to the doctrine of predestination develop such an interest in mathematical odds-making? As it turns out, academic achievements were not particularly unusual among the London Dissenters of his era. In an abstract written to celebrate the 300th Anniversary of Bayes’ birth, D. B. Bellhouse quotes D. Coomer: These were men of high academical attainments, many of them educated in Scottish or Dutch Universities. Some twenty or thirty possessed and adorned the dignity of a Scottish doctorate of divinity. The social circle in which they moved was an elevated one, and peers and peeresses were attendant in their ministrations. For Protestants of that era, there was no suggestion of a wall of separation between faith and science. Indeed, unencumbered by canon law and ecclesiastical politics, Protestants eagerly embraced scientific enquiry as an entirely natural pursuit for those who saw the world as the work of a divine Creator. One of the ironies of this liberty, of course, is that it often led to speculation in theological matters, as well, Descartes being a prime example. It was no less the case among the Presbyterians of Bayes era, which fostered a generation of outright Arians, and whose ranks filled the pulpits of newly-converted Unitarian churches. Bellhouse suggests Bayes may have imbibed of these Arian waters, but the speculation is built largely upon the known Arian sympathies of a number of his scientific acquaintances. To read Bellhouse’s excellent abstract in its entirely, visit http://www2.isye.gatech .edu/~brani/isyebayes/bank/bayesbiog.pdf.  ENDNOTES 1 D.R. Bellhouse, The Reverend Thomas Bayes FRS: a Biography to Celebrate the Tercentenary of his Birth, Department of Statistical and Actuarial Sciences (London, Ontario, Canada: University of Western Ontario, 2001), quoting Coomer, D., English Dissent under the Early Hanoverians (London: Epworth Press, 1946)