What can philosophy teach IT Professionals about the art of Systems Administration? According to author and Microsoft Security MVP Orin Thomas, lots.
Systems administration is both an art and a science. Philosophy by its very nature, seeks to examine and understand human methodologies – why we think the way we do, why we do things the way we do. Following the guiding principle of enquiry and investigation, philosophy is in essence the mother of science and the scientific process – something which all IT professionals engage in at some level, whether consciously or not. How? Because as administrators we have to regularly assess and troubleshoot complex problems – problems which require a broad-ranging understanding of technology, business workflow and most importantly, context. These are skills which have little to do with technical training or certification, and everything to do with how we deal with information and evidence.
Most sysadmins like to think that they take a logical, measured approach to problem solving, but while this is certainly the ideal the reality is that by our nature, humans are pattern-seeking, story-telling creatures, constantly looking to find worldviews which fit our obvservable environment. But, as evidenced by the Monty Hall logical paradox, some realities and solutions are completely counter-intuitive, even though they are demonstrably true. Interestingly, many people cannot grasp the significance of such situations, even when the evidence is laid in front of them. This has nothing to do with intelligence or open-mindedness, but is simply a reinforcement of how we are hard-wired to deal with problems.
Thomas’ approach is to consider sysadmins as any other non-technical troubleshooter – essentially intuitive problem-solvers, albeit often given to over-confidence brought on by years of technical expertise and subject mastery. The trick to more effective troubleshooting is to be constantly aware of the most common logical pitfalls and constantly self-assess to avoid them:
- Confirmation Bias- many people draw their conclusions before all the evidence is in, and then proceed to look for evidence which supports their already-formed conclusions, while playing down the significance of contrary evidence or missing it altogether;
- Sunken Cost Fallacy – when you’ve invested hours or even days into troubleshooting something, the prospect of spending another hours or so seems reasonable. After all, you’ve already expended so much time and effort. Logically, just because you’ve spent a lot of time and resources, this does not mean that the breakthrough must lie within reach. This is one of the logical fallacies which gambling institutions rely on for a steady source of income from punters – it’s called “chasing your debts”;
- Under Determination of Theory by Data – if you have a limited amount of reliable data to hand, there is an almost infinite number of ways that you can tie that data together. Event logs, error messages, user problem reports – these are all essential sources of evidence for a problem, but without enough of them (or without enough of the RIGHT data) you can be chasing your tail and heading down dead-ends for a long, long time
So what’s the solution for the time-poor sysadmin? The answer lies at the heart of philosophy and the scientific method: falsification. First devised by Sir Karl Popper, empirical falsification suggests that no hypothesis can ever be proven to be completely true, because it only requires one counter-example to prove the theory false. Therefore, the strength of a hypothesis lies in its resistance to falsification, and its ability to absorb and accurately explain new evidence as it comes to light.
The everyday implications for IT professionals are these: be suspicious of your own conclusions, and the extent to which you arrive at them via intuition. Be constantly open to new evidence and actively seek out evidence which, if found, would disprove your theory rather than validate it.
