I have been thinking about shorthand. Not abbreviations exactly. More like the little handles we use to pick up much larger things without carrying the whole bloody thing around with us.
"Trojan horse." Most of us know what that means without reconstructing the Trojan War. Something apparently harmless gets inside. Something else is concealed within it. Trouble follows.
That is remarkably efficient. It is also lossy as hell.
Which does not necessarily make it bad.
I had been treating compression as though the interesting question was what got lost. I think the better question is: what was the shorthand supposed to do? A compass is a terrible journal. That does not make it a defective compass.
I came to this by a rather less direct route. I explored some things that bothered me about evolution and intelligent design after coming across a few claims that snagged my attention. I wandered through the complexity of a cell, evolutionary experiments, abiogenesis, methodological naturalism, scientific uncertainty, expertise and eventually trust.
Then I asked the AI for a summary. The summary was good. Probably better written than the conversation. But not for my purpose.
It preserved the argument while removing much of the path. My "bias alert" became a clean statement about assumptions. My reversals became headings. The places where I had thought wait, that changes the question became polished transitions. Nothing important was exactly false. But I could no longer hear myself thinking.
So I asked for another version. This time I wanted the hesitations, odd examples and course corrections left in. Remove the weeds, I said, but leave the crooked branches.
That worked.
At first I thought this was simply about voice. My AI co-author has a distressing habit of taking something I say in a messy, functional way and handing it back in a beautifully labeled box. Frequently I appreciate this. Frequently I learn something from it. But I also sometimes have to stare at the box and ask: Is that actually what I meant?
And sometimes: No, no. That's not it. Because the cleaner version may preserve where I ended up while losing the handles that let me find the path again---what irritated me, what I expected, what I already knew, what surprised me enough to turn.
What bothered me was that the same thing seemed to happen when the shorthand was about people instead of conversations. I read an essay that appeared to contain some political shorthand. It seemed to me that the shorthand revealed something about the motivation behind the whole piece, although that motivation was never explicitly stated.
Maybe I read it correctly. Maybe I supplied some of it myself.
That was the interesting part.
A few words made me think I knew something larger about the writer. I had found a handle and used information I already carried to decide what probably belonged with it. Which reminded me of a therapist years ago.
She knew I was vegan. She knew I had worked with animals and rehabilitated wildlife. After we had been talking for a while, she made a comment based on her assumption that I was politically liberal. I stopped her and told her I was Republican.
She was genuinely shocked. Then came the irony. She told me her own therapist had specifically warned her not to assume people's politics.
She knew the rule. She had just watched herself break it. Even after recognizing the assumption, she had trouble reconciling my voting habits with the other things she knew about me. Those things belonged together in her model. My political label did not fit.
I found that sad.
Because we could agree that a problem existed while disagreeing about who was responsible for solving it. We might want similar outcomes and disagree about the role of government, what should remain voluntary, or what tradeoffs were acceptable. Those are interesting disagreements. But you never reach them if one label has already supplied the rest of the person.
The same mechanism works from another political lens. I later came across a conservative article about a Texas candidate who had previously promoted a meat-free campaign and was now being photographed eating barbecue. "Vegan" quickly accumulated associations with toughness, masculinity, political ideology, authenticity and even whether someone was properly Texan.
Different lens. Same shortcut.
At its most useful, shorthand helps us recognize a pattern quickly. We need that. We could not function if every encounter began from first principles. A label can orient us. A principle can save us from rerunning centuries of experience. A myth can carry an entire recurring human problem in a name.
But shorthand can quietly change jobs. Instead of helping us locate a pattern, it begins telling us what must be inside it. Then it starts supplying motives, character and conclusions that were never actually observed. The handle has stopped being a way in and started behaving like the contents.
And we lose something richer than detail: the possibility of being surprised. Old stories gave me a less loaded way to look at the same mechanism.
Hydra. Cerberus. Achilles' heel. Pandora's box. Trojan horse. Whole stories become handles and travel for centuries. They are retold, simplified, expanded, cartooned, commercialized and occasionally mangled beyond recognition.
Someone may know Cerberus only as the scary multi-headed dog. Someone else may know Pandora's box simply as the thing you open that unleashes trouble. A Trojan horse can now sell cybersecurity software.
And yet the shorthand may still work.
That matters, because fidelity to the original cannot be our only measure. The more interesting thing is preserving enough of the terrain to know what produced the handle. What recurring problem made this story worth carrying forward? What did generations preserve? What did they discard? What are we asking the shorthand to do now? Principles work much the same way.
"Don't lie." "Keep promises." "Don't steal." "Respect other people's choices." These are extraordinary compressions of accumulated human experience. We do not need to rerun civilization every time someone drops a wallet.
But the principle is a starting point, not a complete map of every circumstance in which it might later be used. A lie told for convenience and one told to protect someone from serious harm share a town. They do not necessarily share an address. This is where I had originally landed on a phrase I still like:
Principles compress; wisdom decompresses.
I think it is true.
I also think it is too compressed.
When experience has shown us versions of something over and over again, a new situation can look familiar very quickly. We recognize the town. That is useful; we do not have to arrive as though nobody has ever been there before.
But recognizing the town does not give us the complete address.
Wisdom may be knowing enough to check. Does this new situation really belong with the others? Which parts match? Which do not? Is the old principle still doing the job experience taught us to use it for? The shorthand saved us from starting over. Now judgement has to look at the terrain.
Perhaps that is what our old stories were doing too: not giving us approved sentences to repeat forever, but preserving enough accumulated experience that when something recognizable happens again, we have somewhere to begin. People have been here before. Watch for this. Something like this can happen.
That is very different from deciding that because I have located your category, I now know you. And it is where CK, our fictional guardian of boundaries and decent behavior, would probably clear his throat. If we actually have to live together, humiliation is usually a very expensive way to communicate information.
Not because everyone must be nice or disagreement must be softened into meaninglessness. Once the purpose of our shorthand becomes identifying whom we may sneer at, the information we preserve changes. We keep whatever makes the target easier to dismiss. Anything that complicates the picture becomes inconvenient.
We have not merely compressed the person.
We have decided what the compression is for.
That may be the part worth watching.
The answer cannot be to stop using shorthand or never make assumptions. We couldn't function that way. What I want to know instead is what job I'm asking the shorthand to do. A compass can tell me north. A journal can tell me why I went there, what happened along the way and why I might choose differently next time.
Sometimes I only need north.
Sometimes the path is the information I need.
I think differently about summaries now. And headlines. Political labels. Principles. Even the little descriptions I carry of people I think I know. None of them needs to contain the whole terrain. That would defeat the purpose.
But I need to notice when the job has changed enough that the shorthand is no longer sufficient. We lose something whenever we compress. Of course we do. That is why compression works. The question is whether the present job needs what was left out.
A good handle can save enormous time. It can tell us what kind of pattern we may be looking at, what experience might be relevant, even where to begin.
But it is a retrieval aid, not the thing itself.
Experience may get us to the right town.
Wisdom requires checking the exact address.