Some thoughts on Ockham's razor and OOP / by Ronen Gil

Some thoughts on Ockham's razor and OOP

By Ronen Gil

As a person on the sidelines of the IT global community for several decades now (not on the IT front, but rather on the margins of the community side of it), I find myself, from time to time wondering about the reasons and motives for trends and directions this community is taking. Every once in a while, these wondering of mine go to Object Orientation and Object-Oriented Programing Languages. This is by far, the steadiest (especially if you consider the Internet to be one gigantic phenomenon and not a trend) direction IT is turning to in the last 50 years and more. Can the only reason for this direction be the fact that Object-Oriented is simpler to grasp? I don’t think so. In many aspects, procedural programing is way simpler… So what drives more and more Environments, Programing languages and programmers to support Object-Oriented Programing (Oop)? I’ve come to realize that Ockham's razor is relevant to this matter in more ways than one.

First comes first. What is Ockham's razor? Named after the 13^th - 14^th century English philosopher, William of Ockham, Ockham’s razor is a problem-solving principle, that can be simplified to saying that a simpler solution is preferable over a more complex one. If you have to decide between two solutions to the same problem, you will do better if you chose the one that has less assumptions and is simpler, than if you chose the more complex one. In fact, we implement Ockham’s razor in our daily life, without even noticing we do so. Each time we choose what to eat, what route to take or what to play with the kids – we will end up using Ockham’s razor more time than not…

This brings us right back to the subject at hand. What is the other (beside being influenced by Ockham’s razor) thing common to all these activities? They all involve complex objects. The more complex the Object we have to deal with, the more we tend to break it to smaller simpler object. The same is true for computing. The more complex the machines we handle and the more complex the software we write to deal with these machines, the more we need to break down the task into simpler, much easier to handle structures – but, with a higher speed option to re-construct to complex structure again.

Oop gives us just that – a way (not so far from real world objects, even if it can take a much more abstract state) to break the most complex objects into the smallest and simplest objects and to rebuild the complex objects from those small ones – with the same logic that applies to those small objects.

Even the most complex software in the world, begins with a single keystroke.