Treasure hunting

In their fascinating book Why Greatness Can't Be Planned, Kenneth Stanley and Joel Lehman discuss the idea of achieving one's objectives, particularly in the ambitious sphere. It's not possible, they contend (quite convincingly), to find what you're looking for by looking for it. Proving this point through computer science experiments and a range of real world examples, their essential idea is that to get from where you are now towards any ambitious (read: not completely obvious) goal, you can't possibly know the intermediate steps to get there. An often used example they give is vacuum tubes for computers—this is a large part of how computers were actually invented, but they lay nowhere on the logical path to computers—they don't resemble computers as we know them today, nor do they seem to have anything to do with computing.

“Objective search” they say—their term for setting a goal that's not linear and then trying to achieve it—is simply not possible.

What is possible, and what they repeatedly advocate for throughout the book (it's short, read it), are two components called “constraints” and “novelty search”. Novelty search essentially says that to explore the possible space of all inventions (or goals, whatever), the most effective way of finding anything worthwhile is not to look for anything in particular but rather to do novel things. Looking for something, or trying to create something, is forward looking—it looks into the future and tries to bring it closer—whereas novelty search is past-facing—it looks at the past and tries to do something different than what was done.

Evolution, they claim, is a novelty search—it has no objective, per se, other than to increase novelty (which also means increasing complexity). Through this (and lots of time) it's created every single thing on planet Earth—a pretty powerful search function, though it was not “looking” for any of it.

The other thing that's critical for this are something called “constraints”. In evolution, for example, the constraint would be “survive and reproduce”. If it doesn't survive and reproduce, it doesn't get to go on in the novel search. So whilst the novel search is the method for discovery—essentially, say the authors, you become a treasure hunter—it's bounded by the constraints of the system you're operating within.

Constraints are interesting for a few reasons. In real world systems (people, places, buildings) they can be non-obvious and also operate a different scales. A constraint on a building might be the geography, or the angle of the sun, or a bunch of other things that don't directly pertain to the building itself, but it could also be the plot size of the land or where the retaining walls need to be or how many bathrooms the homeowner wants. Some constraints are volitional (number of bathrooms) and some are baked in (geography, sun angle).

When designing something, architect Christopher Alexander suggests that you need to take all these constraints into account in hierarchical order, meaning that you need to understand how the constraints affect each other—sun angle affects light in a room and is unchangeable—and work on solving for the upstream constraints first. It doesn't make sense, says Alexander, to place the kitchen island or oven before the windows, because you'll limit where the windows can go and thus the amount of light in the room, needlessly. Much better, he says, is to place the windows (dealing with the light constraint) and then the other things.

If we think about these two principles—novel search and constraints—in communication design, we see that the best communication will not necessarily have anything to do with the “objective” of your talk in any meaningful sense, and will be limited only by certain constraints.

Design of any communication, then, should start with constraints and then progress outwards. Is it a fairly routine type of thing where there is no need to be ambitious (emails to coworkers, etc)? Then you could probably proceed with a clear objective in mind.

But when you're trying to be ambitious—book launches, keynotes, big, bombastic opportunities—there it makes sense to think in a “novel search” kind of way. We put in our constraints—and these can be at any scale (time of day, place, prior talks, length of talk)—and then we explore, looking for treasures.

As an example—one client I'm working with is launching a book and wants to design a keynote for her tour. We started with her considering a summary of the book and of the principles of her method. But as we explored, we realised that actually, trees and rubberbands might be better material and connect more with the audience and her mission. Rather than be doggedly fixed on creating a keynote that was close to her objective of “marketing her book”, we discovered a treasure that allows her to express herself far more fully and authentically, and will probably do a better job of connecting her audience to her (and achieving her original objective).

We are very good at thinking that we can get to where we want to go by setting a course for it and moving in that direction. What I've learnt, both from my reading and my clients, is that sometimes we've got to go wandering on the beach and look for buried treasure instead.

PS - these ideas are complex and still not fully baked in my head. I tried to be as clear as I could here, but if you're interested in discussing this and the implications further, reply to this and let's have a conversation.