Huh I never thought about this before —
There are 2 completely different kinds of iteration in any kind of work. There’s rework loops wired to internal standards and expectations and rework loops wired to external tests and risks.
Each loop you have to choose which kind it is
Conversation
Picasso’s bull series is a great illustration of internal iteration. This is what reworking a piece of writing or a drawing feels like before I hit “publish”.
9
50
168
The evolution of the look of cartoon characters as a long work matures is a good example of external iteration. This is what it feels like to me to work a theme (like “sociopathy” or “mediocrity”) across multiple posts. The idea stabilizes.
3
1
34
The key difference is that external iteration embodiments have to be quasi-stable in the mind of the consumer rather than producer. They have to work well enough to get the reader from episode (n-1) to (n+1) without terminal confusion.
1
1
14
When iterating internally, a loop only has to update your own mental state as a producer. It can be incoherent, incomplete. It need not even really exist: a correction line in a drawing without erasing/redrawing old drawing creates a new sketch in your head, not on the paper.
1
1
12
An internal iteration can completely destabilize a thing for a while. Like trying a different word or phrase in a paragraph. If it works, the whole piece is broken until you cascade the change consistently through the rest. Not just global search/replace but logical entailments.
3
15
Twitter is very interesting because it erases the internal/external iteration difference by making each step small enough. Threads mess with that a bit, but I’ve almost never had to delete/rework a tweet in a thread to avoid confusion.
Replying to
Interesting to contrast Futurama, Rick and Morty, The Simpsons, and South Park. Futurama clearly had the budget and experienced creators to luxuriate in internal iteration. Same with R&M. But original Simpsons (not Zombie) and South Park are a lot more external iteration.
2
12
Apparently each episode of Futurama spent a year in dev. South Park, I read somewhere, can still go from concept to execution in a week. You can tell, and not just from the extreme currency South Park manages to maintain. There are always frayed edges.
1
8
I suspect the biggest popular successes have a much higher ratio of external to internal iteration. Partly because they have to, since they’re likely running lean budget-wise. But that also makes them more likely to break new ground (and suck and run aground as well).
3
9
Critical and/or niche darlings otoh tend to be low E/I. In fact I’d say low E/I ratio that “works” initially does so by being born fully formed and therefore appealing mainly to an OCD type of consumer who won’t tolerate loose ends and frayed edges as an iterative work evolves.
1
1
8
Low E/I vs high E/I. This is the choice between the curse of nerd fandom that might limit future broad appeal vs the curse of shoestring budget bootstrapping which risks sucking too much to ever get off the ground.
2
6
Another point, internal and external loops don’t just differ in speed and locus of feedback. They differ in purpose. Internal feedback is about deciding what’s worth doing, external feedback is about figuring out if it works in an environment. QA is secondary in both.
1
6
QA (quality assurance) is actually a trivial cost/risk trade off. Both you and your reader can detect typos for examples. Whether you do an extra copy editing pass or let the first few readers find them and complain/point out is just a cost/risk decision.
1
5
Setting aside objective harmful risks (anything from consumers spending non-refundable money on bug-ridden things to planes crashing) QA is actually a function of the king of consumers you want and how foolish you’re willing to look.
1
6
I am fairly willing to look foolish and tend to solve for readers who will be forgiving towards rough edges in the interest of getting to sillier, more speculative places. If you can’t get past a misplaced comma or misattributed quote, you’ll hate my schtick anyway 😎
1
1
22
