Let's say goodbye to the machine and war metaphors we use every day.

Three sentences I have heard over the last weeks, all of them in perfectly friendly spaces:

— «Let's improve the sales pipeline»

— «We have not lost this battle yet»

— «We need to replace this position».

Nobody meant any harm, and that's what I want to look at in this post. We reach for machine or war metaphors in autopilot and I suspect that we do not really notice them. The third one, where a human person is comparted to a part of machine I find almost more disturbing than the battle metaphor, albeit only by a small margin.

I want to reflect about that language, because I think it hides the things that I care most about at work: humans doing things together, their ingenuity, discovery, and the «accidents» that turn out to be the best thing that can happen.

Let me start with my own contribution to the problem

In the nineties I helped establish an idea that now I believe is flawed: Innovation as something predictable, as a pipeline to be planned and implemented. While I still love the business idea we had (coming up with ideas in a predictable, systematic way like in a factory), I know today that when it comes to choosing and implementing the best ideas the metaphor is no longer useful.

Stephen Kline and Nathan Rosenberg had already made this point in 1986, in a paper called «An Overview of Innovation». They replaced the linear pipeline with what they named the chain-linked model, in which real innovation loops backwards, jumps sideways and gets built through experiments, mistakes and repairs far more than it ever flows in one direction. Their paper is free to read at the National Academies Press, and I'd happily hand it to anyone still believing that innovation can be neatly planned.

Three more areas where the metaphors have been debunked

DNA as a blueprint to live: The metaphor promises a fixed plan that can be executed to «build life». What biologists keep finding instead is that the same DNA gets read differently depending on context by a whole bunch of processes that seem to negotiate and collaborate rather than fight and obey. Philip Ball's Book «How Nature Works» is a lovely tour through the astonishing world of biology discoveries. He also wrote a short piece in Nature with the admirably blunt title «It's time to admit that genes are not the blueprint for life». 💥 Biology, it turns out, seems to be governed much more by self-organisation and collaboration than previously thought.

Organisations as machines: Gareth Morgan took this one apart already in 1986 in «Images of Organisation» where the machine appears as just one in eight ways to see an organisation, and one that only holds up in stable and predictable environments with routine conditions. Morgan also points out where the machine organisation got its furniture: from the military. Ranks, uniforms, standardised regulations, systematic training, command language. Our two metaphors turn out to be relatives.

Today's conditions are neither stable nor routine. Dave Snowden's Cynefin framework helps here precisely because it separates complicated from complex. A machine is the perfect complicated system — you can take it apart, understand it, and put it back together. Complex systems never grant you that, and most of the work we do in collaboration is complex.

The immune system as an army at war. The US Institute of Medicine brought experts together in a workshop under the title «Ending the War Metaphor», which tells you how seriously the field takes this. The argument is that most of what happens between us and the microbial world isn't combat at all but ecology and negotiation, and that war language actively points researchers in the wrong direction. Jon Turney's essay «Why we should guard against military notions of immunity» is the digestible version of these findings.

Metaphors set expectations

This all matters beyond language, it is much more far-reaching. A metaphor tells you what kind of place you're in, what's likely to work, who gets to decide, and what counts as a problem in the first place.

The immune example shows the cost most clearly. During the AIDS crisis, the language of defences and invaders made it easy to read illness as a failure of the defender, and from there it was a short step to blaming people for their own infection. Both Turney and the anthropologist Emily Martin, in her book «Flexible Bodies», trace how that happened. A metaphor did that.

Where power comes into it

The machine image did not just fall from the sky. We can pinpoint its datable origin in seventeenth-century mechanical philosophy, with Descartes' animal-machine and Newton's clockwork universe, and it spread because it was enormously useful to people who wanted the world to be predictable and controllable — people, almost exclusively men, who were, by and large, already in a position to act on that wish.

The encouraging part of that history is that somebody in power chose this metaphor. And the good news from where I stand, is: Things that were chosen can be un-chosen.

PSA: I am going to use a metaphor!

If machines and battle don't fit, what does? For me, everything that happens when humans interact to build something much more resembles the idea of planting. Preparing the ground, planting seeds, caring for them, tending to the ones that survive, sharing the harvest: All of that is closer to how I see humans interacting, and yet also this metaphor is not perfect.

Gardens have a tidiness problem of their own, and nobody should be weeding their colleagues 😅. But it asks better questions. What are the conditions here? What's growing that nobody planned? What needs more light? You get my drift.

It starts with noticing

Here's a small practice that costs nothing: notice one machine or war metaphor you use this week, and try saying it differently.

Back to my three sentences from the beginning:

— «Let's improve the pipeline» could become «where are we getting stuck, and what did we learn from the last attempt?»

— «We haven't lost this battle yet» becomes «What are we learning from our competitor, and what do we want to be known for?»

— «We need to replace this position» becomes «Someone is leaving and taking years of knowledge with her. What kind of person do we actually need now?»

The rewrites are longer, and that's the point: the machine version was shorter because it left things out, mostly the humans.