Emergence

“The phenomenon of emergence takes place at critical points of instability that arise from fluctuations in the environment, amplified by feedback loops.” — Fritjof Capra

Emergence is difficult to discuss because it hides behind our usual ways of thinking. Emergence is a process, the ways simple interactions among individual parts form more-complex behaviors and patterns. These behaviors and patterns can’t be reduced to the individual parts — they arise because parts are together in an environment. Emergence is nonlinear. No concrete, easily-reproducible steps lead to it and yet, over time, it happens.

We have all experienced emergence. Snowflakes emerge. They are shaped by simple rules interacting with changing, varying environmental conditions. Yet we struggle to understand how lower-level rules can lead to higher-level sophistication.

Reductionism may serve to explain how a bird flies, but not how a flock of birds move in unison. It may describe internal combustion, but not traffic patterns. It may describe electric patterns in the brain, but not consciousness, and it’s unlikely that anyone or anything – not even the world’s most powerful computers – will ever fully analyze the interactions that make for healthy soil.
— Mark Bittman, author, chef, journalist

Our worldview is profoundly reductionistic. Reductionism breaks things down into steps or parts in order to think about (and build) them. We design software this way. As complexity increases, we design systems of software the same way. Concerned with control, linear processes, and implementation strategies, we reduce relational complexity into the sum of its parts. Then wonder why the math doesn’t add up.

“You think that because you understand “one” that you must therefore understand “two” because one and one make two. But you forget that you must also understand “and.”
― Donella H. Meadows, Thinking in Systems: A Primer

Reductionism is sometimes necessary, but it can blind us to what happens when parts engage and emerge together … and become something that more than the sum of the parts.

In Mechanistic to Systems Thinking (described below), Russel Ackoff argues that our worldview is changing from the machine age to the systems age. He said this 20 years ago. Our thinking is lagging behind our experiences: globalization, the internet, smartphones, climate change, a pandemic … everything around us is rife with systems challenges. Designing for emergence depends on knowing how to build systems in a “rapidly-changing, hyperconnected and radically contingent world” (Pendleton-Jullian).

In systems of software, each deliverable is a change to an ecosystem, governed by interrelationships, environmental feedback loops, patterns and rules that generate both structure and unpredictability. We can work inside of that reality — rather than try to dominate it. If you watch the Biggest Little Farm (described below), you’ll see how designing biodiversity on a farm differs from the factory farm and mono-crop approach that is common.

To see emergence, we can’t look at McDonald’s as (solely) a profitable enterprise that delivers a Big Mac. We also need to look at the land where the cows are raised and the bodies that eat the food. We need to see what happens to the trash and the ways advertising and food science influence behavior. Every McDonald’s is one node in a system of systems that interrelate. And this system has qualities that only exist because of the relationships, not directly from McDonald’s selling a Big Mac.

We aren’t looking simply to judge right or wrong, good or bad (though we might conclude that McDonald’s does more harm than good). We are looking to understand how rules, patterns and relationships scale to create a complex adaptive system that couldn’t exist without those relationships. And how more adaptive systems are created as a result, sometimes in ways we did not intend.

Designing for emergence and for emergent behavior is about designing smaller scale things and actions that create context for larger things to happen.
— Ann M. Pendleton-Jullian, Design Unbound: Designing for Emergence in a Whitewater World

We usually design technology (and our lives) by changing a fixed behavior — click a subscribe button, eat fewer carbs, exercise every morning, write tests for each commit. We fix bugs. We assign tasks.

Simultaneously, the current system and its environment is driving, influencing and sometimes discouraging the behaviors we want to change. Designing for emergence is developing new rules, patterns, processes and relationships that lead the system toward developing new, emergent and valuable (though not necessarily predictable) behaviors. We create systemic support for change, rather than deliver the change alone.

We do not rise to the level of our goals. We fall to the level of our systems..
— James Clear, Atomic Habits

Technology systems are built out of ideas. Ideas generate more ideas, like fractal art. The necessary mindshift isn’t thinking different thoughts … it is thinking differently. Our attention shifts to relationships and patterns and trying to understand “and”. How do three parts together do something that none of the three can do alone? What potential does this “something” create? Can I design for this “something”?

This newsletter, for example, is an HTML page. The page has a title, essay text, multiple (related) media resource links, a quote and an upcoming event (as text and links). As an HTML webpage, a newsletter scales to be a collection of newsletters, following the rules of webpages. If the pages have a category taxonomy, they can create sub-collections of pages.

If, instead, I change the low-level design and create information objects from each piece (essay, learning resource, quote, event), the system scales differently, branching off into new combinations. The essays become a blog or magazine, the resources become a library of systems’ learning experiences, the quotes are dynamically associated with the speaker’s work and the events becomes a calendar of events. New shapes and products, based on how people interact with the information, can arise. Like a course on emergent systems design.

All emergent systems are patterns in time. Social media sharing, for example, has simple rules about the length of text, allowed media types, and how people interact with the posts. But as we’ve seen over the last decade or more, the flow of information in “real time” has given rise to patterns that profoundly impact our social systems. The attempted hijacking of an American democratic election and the proliferation of radicalizing disinformation was not part of the original design. Wide-scale philanthropic activism and quick, self-organizing help for people in crisis were also not part of the design.

We are only beginning to understand the addictive behavioral impact of the “like” button. The iPhone was designed to be an iPod that could make calls. Steve Jobs did not like the idea of external apps running on the OS (and breaking it). But installing apps scaled to the profound changes in behavior. These changes impact many areas of our lives, the least of which is (likely) listening to music.

In designing a strategic game, one is designing rules for emergent behavior. One is designing for emergence.
— Ann M. Pendleton-Jullian, Design Unbound

Rather than thinking in terms of linear functionality (which increasingly does not exist), understanding emergence is more like thinking in game design. Increasingly, software’s rules create ecosystems in which emergent behavior arises. We can design those rules. These rules lead us towards potential new experiences we could not have imagined twenty years ago when Ackoff said “we are living in the systems age.”