Before I Knew Why

Sometimes I know I don't like something before I can tell you what's wrong with it.

I've had this happen while looking at a page. Nothing is obviously broken. The type is readable. The spacing isn't absurd. Everything that needs to be there is there. Still, something feels off.

Sometimes the reason turns up later. Maybe everything has been given the same visual weight. Maybe two things look related when they shouldn't. Maybe the page is technically consistent but still feels crowded. Once I can explain the problem, the earlier reaction feels less mysterious.

That made me wonder: can we notice that something works, or doesn't, before we know why?

The tempting answer is yes. Maybe judgment arrives first and vocabulary catches up later. Maybe intuition already knows something that we haven't learned to explain.

I quite like that answer. I just don't know if it's true.

I was mixing up three different things

The first problem with my question was that I had quietly bundled together preference, expertise and explanation.

They aren't the same.

Someone can know almost nothing about art and still have a strong opinion about a painting. Someone with years of experience might notice something immediately and struggle to explain exactly how they noticed it. And a beginner can have a perfectly ready explanation for a preference without the explanation being particularly good.

That sounds obvious now. I hadn't separated them when I started.

There is research showing that formal expertise isn't required to make aesthetic judgments. In one study, people in India and the West, both with and without art expertise, rated visual art for things such as beauty and liking. The people without art expertise were perfectly capable of making those judgments. Expertise changed some of the patterns in the ratings, but it wasn't required for a response to exist at all.

But that study does not show that people had reactions before they could explain them.

It tells me something narrower: you don't need specialist expertise to have a preference or make an aesthetic judgment.

That's useful, but it doesn't prove that a novice is quietly seeing everything an expert sees and merely lacks the vocabulary to describe it.

If I immediately prefer one website to another, the preference is real in the fairly boring sense that I genuinely prefer it. I still don't know whether I've noticed something important, responded to something familiar, disliked something unusual, or simply liked one more.

And I definitely don't know whether I'd make the same judgment after learning more.

Learning can change what we notice

Most of the research I found was about visual art, not interface design.

That's an important boundary. Looking at a painting and deciding whether a product page works are not the same task. Art expertise can involve history, symbolism, technique, cultural context and other knowledge that doesn't map neatly onto a website.

So these studies cannot tell me whether my reaction to a particular layout was right.

What they can help me investigate is the broader question underneath it: can learning and experience change how we look at something and how we judge it?

Here the evidence gets more interesting.

One study compared 20 art-history experts with 20 people without that expertise while they viewed paintings ranging from representational to highly abstract. The researchers measured not only what participants thought of the paintings, but where they looked.

The groups differed. As the paintings became more abstract, the non-experts' aesthetic ratings fell. The experts' ratings did not show the same pattern. Their viewing behaviour differed too. In paintings containing people, for example, the non-experts spent more time looking at faces and heads, while the experts' attention was distributed differently across the work.

That still leaves a big question open. It compares people who already became experts with people who didn't. Maybe training changed how the experts looked at art. Maybe people who were drawn to art in the first place already looked at it differently. Comparing the two groups alone can't cleanly tell us which explanation is right.

Another set of experiments gets closer.

Researchers gave people without art expertise short lessons about particular art styles and then looked at how their judgments changed. After training, some ratings changed, including preference, understanding and perceived artistic skill. The effects depended on things such as what participants had learned and how similar new paintings were to the ones used during training.

That doesn't prove some universal law about aesthetic judgment.

But it does make one simple story harder to defend:

I already had the judgment. Learning merely gave me better words for it.

At least in these experiments, learning changed some of the judgments themselves.

So when I think, I knew it looked wrong before I knew why, there are at least two possibilities. Maybe I had picked up on something I couldn't yet explain. Or maybe I was judging with the distinctions I knew how to make at the time, and learning more would have changed the judgment.

From inside my own head, those might feel exactly the same.

Maybe vocabulary is too narrow a word for what changes

I started this whole thing thinking about vocabulary.

Hierarchy. Contrast. Composition. Rhythm. Leading. Negative space.

There is something satisfying about learning a term for something and suddenly thinking, ohhh, THAT'S what I've been noticing.

But the studies above don't isolate vocabulary.

Expertise includes looking at thousands of examples, making comparisons, learning history and context, getting things wrong, receiving feedback, recognizing patterns and probably plenty else. Even the short training experiments taught people about artworks and styles, not simply a list of new words.

So I don't have evidence for a claim like:

Learn the word for something and you start seeing it.

Maybe that happens. These studies don't establish it.

A safer conclusion is that learning can change what we pay attention to and how we judge what we see.

Vocabulary may be part of that. But learning a field and having words for it are not the same thing. The word may feel powerful because it arrives attached to examples, comparisons and a distinction we have just learned to make.

That raised a slightly different question.

Suppose I eventually can explain what I'm seeing. Does putting the perception into words simply describe it, or can the act of describing it affect the perception too?

Those are two separate problems.

The art studies are mostly about what learning and expertise can change.

A different line of research looks at what can happen when people actually verbalize what they saw.

“Good taste” gets harder to define

Before getting there, there was another word I had been using too casually: taste.

In the context I'm interested in here, I mean something fairly narrow: aesthetic judgment or preference about visual work. Even then, “good taste” gets slippery quickly.

Does it mean somebody reliably likes things I think are good? That they notice qualities other people miss? That they can distinguish excellent work from merely competent work? That they can explain why something works? That people with more experience would agree with them?

Those aren't the same ability.

The cultural evidence makes the idea of one universal standard harder to defend too. In the India/West study, expertise did not relate to aesthetic ratings in exactly the same way across cultures and kinds of art.

So even if experience changes how people judge visual work, that doesn't mean everyone is slowly approaching some final, objectively correct version of “good taste.”

Which makes this sentence less impressive than it first sounds:

I could tell it was good before I knew why.

Maybe.

But perhaps what I could tell was simply that I liked it.

That's still information. It just isn't the same claim.

Explanation isn't automatically an upgrade either

Psychologists use the term verbal overshadowing for an odd effect found in some experiments: describing a visual experience can make later visual recognition worse.

In research involving faces, for example, putting what someone had seen into words interfered with parts of their later perceptual recognition.

That doesn't tell us whether talking about typography makes someone better or worse at judging typography. It is a different kind of task.

What it does challenge is another assumption I had been making: that once I can explain a perception, the words simply sit on top of whatever I was already seeing.

Apparently, under some conditions, the act of turning perception into language can interact with the perception itself.

A study involving fencing movements makes that relationship stranger. Participants with different amounts of fencing experience watched movements and later tried to recognize them. Describing the movements beforehand helped novices, hurt the intermediate group, and had no detectable effect on the experts.

Again, a fencing-memory experiment is not a theory of design.

But notice that this is a different question from the one the art studies were helping with.

The art research made me question whether:

learn more → explain the same judgment better

is necessarily what happens.

The verbalization research made me question whether:

see something → describe it in words → leave the perception untouched

is necessarily what happens either.

Sometimes words help. Sometimes what we've learned changes what we notice. And under some conditions, putting a perceptual experience into words can interfere with parts of how it is later recognized.

So seeing, learning and explaining don't seem to sit on one simple ladder.

So what is “something feels off” worth?

After all this, I still haven't explained exactly what is happening when I look at something and know I don't like it before I know why.

Maybe that was the wrong thing to expect from the research.

What I can say more carefully is that the first reaction and the later explanation are not automatically the same judgment at different levels of eloquence.

Sometimes learning may give me words for a distinction I was already making. Sometimes it may show me that the original reaction had a sensible basis. Sometimes it may give me a distinction I hadn't been making at all. And sometimes I may learn enough to change my mind.

I had been imagining the process roughly like this:

see → react → eventually understand why

Now I think this is at least as plausible:

see → react → learn → notice differently → react again

The explanation doesn't necessarily arrive at the end and politely describe what happened at the beginning. What I learn can become part of the next thing I see.

That doesn't make the first reaction useless.

If something feels off, I still think that's worth paying attention to. I just wouldn't treat the feeling as proof that I've detected some hidden flaw.

It's a reason to look again.

What am I responding to? Does the reaction survive comparison? Can I work out what is bothering me? Does learning more strengthen the original judgment, or does it change it?

Sometimes I'll eventually find the reason and think, yes, that's what I was noticing.

Sometimes I'll realize I wasn't noticing that at all.

So I don't think I've ended up with a clean answer to the question I started with.

We clearly don't need expertise in order to have preferences or make aesthetic judgments. The research supports that much. What it does not establish is that those reactions necessarily came before explanation, or that an unexplained reaction is good judgment waiting for language.

That's a much harder claim.

Maybe “something feels off” is hidden expertise trying to speak.

It might be.

But it might also be preference, habit, familiarity, incomplete knowledge, or something I haven't understood yet.

Either way, it isn't where the reasoning ends.

It's where I start looking.