Each year on 11 November, Teresa Neilsen Hayden writes about the first world war – a compendium of thoughts and links which bring home the enormity of the horror.

Last year, she reported a conversation with a friend:

We were talking about the way the military on both sides kept trying mass “over the top” charges into the no man’s land between the trenches, and taking staggering losses. Doing that a few times would have been bad enough, but in WWI, both sides  kept it up for years.

I said, usually if a general plans and conducts a major battle that winds up taking a pitifully small amount of ground, and gets hundreds of thousands of his troops killed, he’s relieved of command.

Mike said, after a while, all the generals had fought battles like that. If you went on doing the same thing, at least it was something you knew. You wouldn’t do any worse than any other commander. But if you tried something different and it didn’t work, then heaven help you.

Commenters on the post (and this is a blog which is unique in my experience in attract huge numbers of comments while maintaining a meticulously civilised discourse) were quick to draw comparisons with other areas of life.  One, also recalling a past conversation recalled that:

He observed that, in your typical publishing company, if an editor does the things conventional wisdom says he or she should, and they fail to make the company money, that editor has a good chance of being forgiven. Whereas if the same editor tries something different and it fails to make money, the editor’s in trouble.

This always seemed to me a pointed argument against the idea, fondly advanced by business people everywhere, that profitability is the only metric of performance. In fact unprofitable performances are constantly forgiven, provided they’re delivered with unimpeachable conventionality. Businesses like to–need to!–make money. But businesses are also cultures, and cultures need to enforce social norms. Over the years I’ve found that this explains a great deal of institutional folly.

Others made the point even more simply:  “Nobody ever lost his job for buying from IBM”.

Meanwhile, William Heath is frustrated by the ponderousness of change in government:

Who tells CIOs and the IT community what they need to know about customer service, co-creation and human rights? Accenture? Liberty?

It’s not happening. Many of our CIOs are at best pretty Web 1.0 and at worst (in the words last week of a Whitehall big-hitter) just overpaid project managers with poor track records. Meanwhile, who is communicating effectively at Board level about the impact and potential of contemporary technology? Some clunky old management consultancy whose people are too busy defending the old discredited business model to be worth their £3-5k day rate? I don’t think so.

I think there is a connection here.  I have seen projects in government, intended and designed to be small and agile, get effectively destroyed by being captured by the standard ways of making things happen.  Worse still, it is the value that is destroyed, not the cost.

So that gets us back to the question of how to protect and nurture innovation.  That may mean that government departments need to find ways of rewarding failure – so long as it is the right kind of failure.  The barbed wire and the machine guns are real:  there is no excuse for not having learned that cavalry charges don’t work.