Attribution Errors and the Importance of Context

I recently read a fascinating study that was done in the 1950’s by a physchoanlyst named Allen Wheelis

What Wheelis observed during the 50’s was that classic Freudian Analysis techniques were no longer working as often as they used to.

Simply put: The number of people who were “cured” by these techniques was significantly lower than had previously been the case and it was getting worse.

Leaving aside, for the moment, the large portions of Freud’s work that have now been discredited, lets examine the core theory behind why something that had previously been empirically shown to work no longer did.

A Different Time

Freud’s work was anchored in the Victorian Era, an era and a culture strongly dominated by very strong wills, opinions and morals.  It was very much a culture of “character” and “self discipline”.

Freudian therapies reflected this situation as they focussed very heavily on ways in which one might  break through the mental barriers that these strongly opinionated and disciplined individuals had errected and thus reveal to themselves the cause of their neurosis.

Once the said cause was revealed, the Victorian ethos of self discipline would swing into action and work diligently towards correcting “the unsightly defects of their minds”.  And apparently (I wasn’t there) this worked, at least enough of the time to be regarded as a “reasonable approach to the problem” – “empirically proven in the field” if you will.

Oh noes!

By the 1950’s however, Freudian Analysis was beginning to fail, and fail and fail again.


We now know that the reason is most likely the fact that the personality of the average individual had changed quite markedly by the 1950’s.  People were (at least compared to the Victorians) far more relaxed, open and introspective.

As such they achieved insight into the source of their problems far more quickly than their Victorian predecessors ever did.

But once they had discovered the fundamental cause of their woes, unlike the Victorians, they did not have the self discipline and strength of character to follow through on their discoveries to improve their mental situation.  And Freudian techniques, which were effectively designed for a different personality type, were basically ineffective in strengthening self discipline and “building character”


Well there’s your problem…

So how did the majority of the professional community react to this?

In two really interesting ways.

First of all, they congratulated themselves for being so damn clever and talented!

Why?  Because the first phase of Freudian Psychoanalysis was positively rocketing along, that’s why.

They attributed this not to some fundamental shift in the character of their patients, but rather to a combination of the advancement that they and their contemporaries had made to the current body of knowledge and also, of course to their inherent natural talent and skill.

The second reaction, was keyed to and influenced by the first.

“Well, we know that we’re awesome, so the problem must be with Freud’s theories, they must be wrong”

Now, as it turns out they were right, at least partially; but they were led to the right conclusion for the wrong reasons.  And every time that happens, your success is actually based more on luck than on good management.

So what does this mean for Agile? Lean? Complexity? CALM?

Context is important.  And knowing why something works is also important.  It’s not always enough to see that something seems to work in practice now, and thus assume it will always work in the future.  The world changes, people change and will (I hope) continue to change.

And if this is true, then so do our approaches need to change and evolve with us.

Just because something worked once or twice, maybe 5, 10 or 20 years ago; for somebody else in a similar set of circumstances to you does not mean that if doing the exact same thing doesn’t work for you in 2012 that the reason for your failure is that “you must be doing it wrong”

If you’re working solely at the practice level, then that could easily be an attribution error, and a potentially costly one at that.

What your apparent failure might mean is that your context is sufficiently different to the previous success story, that approach selected is never going to work, no matter how well you do it.

Which again is why I’m going to state again that theory informed practice is so vital.  If we understand the principles (or at least have usable models which we are willing to throw away once we have better ones) behind why certain practices succeed or fail, then we can operate at the principle level both to knowledgeably apply appropriate practices to our work as well as synthesise and create new ones in order to confidently and effectively solve our problems in our contexts.  And maybe even advance the state of the art.

One thought on “Attribution Errors and the Importance of Context

  1. George Dinwiddie

    Yes, the fundamental practice of Agile is performing experiments and seeing how they do. The generally accepted practices of Agile are just a good starting point, because they HAVE worked for others.


Leave a Reply

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