Dave Snowden attempts to build a theory for the Agile movement based on statistical research, weak signal detection methodology, complexity theory and co-evolution theory.
Dave Snowden is Founder and Chief Scientific Officer of Cognitive Edge. He is Senior Fellow at the Institute of Defense and Strategic Studies at Nanyang University and the Civil Service College in Singapore. His paper with Boone on Leadership won the Academy of Management award for the best practitioner paper in 2007, having won a special award for originality in his work on knowledge management.
Agile Cambridge is THE learning and collaborating conference for Agile and Lean practitioners in the East of England. The event also draws participants from across the country, mainland Europe and further afield. It is one of the most successful Agile conferences in the UK.
Partial theories on Complexity - Not a Complete Theory for Agile
I also found Davids tone, and lack of humility detracted from the presentation which is essentially about complexity and says very little about the challenges of Agile Software Development. It is clear that David has very little tacit knowledge and experience when it comes to Software Development, and it shows.
Agile is just a label for a set of values and principles that have shown themselves to be effective when it comes to software development. It is not a theory, it is a label for stuff that just works. As David correctly points out, intuition is a form of knowing that comes from experience. The Agile Movement is an attempt to express and give credence to a shared intuition, born from shared experiences.
So if you know what works why do you need a theory? Supposedly to scale. Nonsense. What does scale mean anyway? I've never heard Opera singers or actors worrying about their practice scaling! Either you've made the right investment to acquire the appropriate intuition or you haven't. If you wan't to become a good Opera singer then get busy at putting in the 10,000 hours of practice needed!
There is a role for theory I think, but I don't think it is "scaling", assuming that scaling means on-boarding vast numbers of people over night.
As a culture we are very poor at saying "we don't know". We don't know why Agile works, it just does. Likewise we don't fully understand complexity, we have partial theories as exposed here, but our knowledge is incomplete.
This lecture would be better I think if Dave spent more time talking about what we don't know, and what we are yet to understand. It would also help if he collaborated with someone who as actual experience in the domain of software development.
There are numerous ways to explain the Agile phenomena, each is coherent, but none are complete. Exploring these alternatives would have added to the presentation. Alistair Cockburn has an interesting presentation where he explores these alternative "perspectives":
Complexity theory is only one perspective, and I agree an important one, but I don't think it tells the whole story. For software practitioners wanting to better understand Agile from a complexity theory perspective, I would recommend Joseph Pelrines writings:
In the end though, "getting it" comes down to immersing yourself in practice. Like many areas of human endeavour, software development can't be boiled down to a simple theory. Whilst partial theories do help in understanding and learning, what is crucial is a honed intuition, which can only come through deliberate practice and time.
All the theory in the world won't make up for real life experiences.
Re: Partial theories on Complexity - Not a Complete Theory for Agile
You are badly wrong on Opera Singers as well. Yes there are a series of experiential learnings around when your voice is sufficiently trained to sing Wagner for example, but we know have a scientific understanding of why and we can do a lot more to bring the age down, manage the process better etc. etc. I could talk about the science of acoustics as well. The Opera Singer who simply falls back to experience is less likely to succeed than the one who has the humility to listen to people, who while not singers themselves, understand a lot about the voice.
Oh and try not to make too many assumptions, I have built three software businesses over the years, committed Data Sciences to be a founding member of the DSDM consortium etc. etc..