For the Want of a Nail

For want of a nail the shoe was lost.
For want of a shoe the horse was lost.
For want of a horse the rider was lost.
For want of a rider the message was lost.
For want of a message the battle was lost.
For want of a battle the kingdom was lost.
And all for the want of a horseshoe nail.1

In complexity theory, we might describe the absence of the horseshoe nail as an initial condition. Even the slightest change in an initial condition can have system-wide consequences. This is memorably illustrated by mathematician and meteorologist Edward Lorenz’s (1917-2008) famous question, “Does the flap of a butterfly’s wings in Brazil set off a tornado in Texas?”2

What prompted Lorenz’s question? Using computer simulation models to predict weather patterns, Lorenz realized that even seemingly insignificant differences at the start of an algorithm lead to far-reaching and wildly different results. Initial conditions, it turns out, directly impact end results.

The key idea here is that complexity is not randomness. At first glance, unpredictable, non-linear, non-repeating patterns appear to us as utterly baffling and incomprehensible. There is no rhyme or reason to it; sheer haphazardness defies explanation. But Lorenz developed a mathematical framework to show that the disintegration of order and stability — whether a tropical cyclone or a social uprising — is actually not random at all. Instead, the cyclone and uprising were destined to occur precisely because of initial conditions. This, he explains, is deterministic chaos.

However, no matter how determined chaos might be, it is still entirely unpredictable. The lack of a horseshoe defined the fate of the kingdom, but it was absolutely impossible to anticipate the importance of a single horseshoe at the outset. Why? The horseshoe is only one of innumerable variables, any one of which, when examined on its own, is too insignificant to measure. Just like the flapping wings of a butterfly.

This can be a tricky notion to wrap our human brains around: just because something is unpredictable does not mean that it is random.
Today, right now, the weather outside of your window is not random at all. It is the incalculable result of innumerable causations. And, like the present meteorological conditions, the events of your day will also unfold in this chaotically destined world.

Consider this: Since you started reading this chapter, the variables of your day have already shifted. Every millisecond you spend reading this paragraph repositions the course of your future. Sit and contemplate this paragraph for an extra minute and you may, or may not, ‘just happen’ to have a conversation with the person in the line ahead of you at the cafe. Who knows the ramifications of this potential interaction?

Whatever the case, you can only unravel the chains of causality in reverse. Or, as another saying goes: hindsight is 20/20.


  1. There are numerous versions of this proverb, referenced as far back as the 13th century. Wikipedia has a good list of them here: https://en.wikipedia.org/wiki/For_Want_of_a_Nail#Historical_references 

  2. Lorenz, Edward N. (1972). Predictability; Does the Flap of a Butterfly’s wins in Brazil Set Off a Tornado in Texas? American Association for the Advancement of Science, 139th Meeting. December 29, 1972. 


Cite this page:
Shelley, James. (2020). 'For the Want of a Nail' (in System Thinker Notebook). Originally published on August 5, 2020. Accessed on September 29, 2020. Licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. Permalink: https://jamesshelley.com?p=17071
Additional reference and meta data:
This page is currently a subsection of 'Grappling with Complexity' in the System Thinker Notebook manuscript. Structure and document location subject to change. Use https://jamesshelley.com?p=17071 as permanent identifier for this document if linking externally.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.