Reasoning

This week, wanted to take a detour and discuss a basic concept, which oddly enough, we don’t think about a whole lot. I am reading ‘The Story of Science’ by Susan Wise Bauer¹, a collection of writings by Scientists through the ages. If you are remotely interested in how the scientific thinking and method evolved over the centuries, this is highly recommended.

One of the many surprising discoveries in this book has been Francis Bacon’s contributions to systems of thinking. Before his time, Aristotelian method of ‘deductive reasoning’ was the dominant paradigm. Deductive reasoning started from general statements (often known as an ‘axiom’ or a self-evident ‘truth) and then moved to specific conclusions:

1. Major premise: All heavy matter falls towards the center of the universe
2. Minor premise1: The earth is made of heavy matter
3. Minor premise2: The earth is not falling
4. Conclusion: The earth must already be at the center of the universe

There is an obvious problem with this line of reasoning: the starting point pretty much limits the scope of reasoning and more importantly, it is not easy to accept any observations (or data) that don’t fit into this conclusion. Science and philosophy until the mid-1600s was pretty much driven by this paradigm. This is where Francis Bacon came in – he postulated the opposite: start with specifics and build towards general conclusions: beginning with particular pieces of evidence and working inductively, towards broader conclusions. In other words, hypothesis, experiment, conclusion. This came to be called as ‘inductive reasoning’ and Bacon is generally regarded as the first proponent of the scientific method.

This was nothing short of revolutionary in many ways – the two most fundamental ways Bacon has influenced pretty much everyone is:

1. Raised the importance of evidence – through observation, experimentation
2. Opened the door to falsification of a theory – i.e. theory is no longer an absolute given. It is good only until data cannot prove it otherwise. The foundation of all science today

Of course, we now know that purely inductive reasoning is too narrow and we are all more Bayesian (than we care to admit) – most decisions are a combination of deduction (based on our knowledge of the world) and induction (develop hypotheses, test them out and derive conclusions that either validate our current knowledge or help refine this knowledge).

And then there was another (even lesser known) philosopher, Charles Sanders Pierce (19^th century America)² who came up with Abductive Reasoning which starts with an observation or a set of observations and then seeks to find the simplest and most likely conclusion from the observations. Abductive conclusions are typically qualified with some uncertainty (e.g. ‘confidence intervals’) And not surprisingly, this has become the cornerstone of modern decision theory – and most decision scientists operate with this framework

So why should we care?

All this might seem like one of the many good stories in the history of science – but I do think it is important for a few reasons:

1. I continue to be surprised how often organizations fall into the trap of deductive thinking. It usually starts with a ‘major premise’ at the leadership team level and then everyone down the chain is scrambling, looking for the data to validate the premise and enshrine the conclusions in their day-to-day decisions. Here again – important to distinguish the strategic decisions (e.g. big bets like going into a new market or developing a new product line) which can operate with deductive reasoning. However, it is apparent that most of the decisions that managers make on a day-to-day basis are more operational in nature – where it is important to be guided by a healthy dose of inductive and abductive reasoning.
2. How should a manager drive thinking in her team? I would strongly urge two key guidelines:
   1. Inductive problem solving: For the problem to be solved, get your team to develop hypotheses. Test the hypotheses (with data if you have or increasingly, through experimentation). And then develop/refine your conclusions and drive decisions based on what how the hypotheses test out in real life
   1. Abductive thinking: Always remember that best is the enemy of the good. Instead of insisting on the ‘most accurate’ or ‘certain’ answer (spoiler alert: there never is one), insist on a set of conclusions from the data, with some metric (e.g. confident interval, or the most relevant of the many metrics that can rank the output). And use that to continuously update your business intuition. After all, decision making is a blend of art and science.

Further reading:

1. Susan Bauer’s highly readable ‘The Story of Science’ https://susanwisebauer.com/story-of-science/

2. Charles Pierce is covered in a fascinating book on the history of ideas in 19th century America: The Metaphysical Club by Louis Menand