Knowing how to read isn’t the same as reading

We are very good at recognising the shape of a thing we have not properly understood. We know what a serious article looks like. We can spot the main headings, acquire its vocabulary and remember a sentence that sounds like its conclusion. Afterwards, we can have a reasonably convincing conversation about it. Whether the argument ever really arrived is another question.

That is an uncomfortable admission in a publication about writing. It is also useful. A reader can move through every word without following every connection. The page has been visited. The work has not necessarily been done.

The awkward pause

A good test is to stop and explain the paragraph in our own words. Not the elegant phrase we have just borrowed from it, but what the writer is actually asking us to believe. What changed between the first sentence and the last? Which fact supports which conclusion? Where did we agree because something sounded familiar rather than because it had been shown?

The test can expose a bad paragraph. It can also expose a reader in a hurry. Sometimes the writer has supplied the connection and we have stepped over it on the way to the next heading. Sometimes we have brought an opinion so firmly to the page that we are merely checking where the author belongs in relation to it.

Neither is fixed by requesting a shorter version.

A summary is useful when we need to decide whether to read something, find a passage or recover the broad outline. It is less useful when the process of reading is the point: spending enough time with an unfamiliar idea to discover where our existing view is insufficient. The compressed version can save the inconvenience through which the understanding was supposed to happen.

This is not an argument for making every piece long. A padded article does not become profound because it has taken our afternoon. It is an argument for giving the necessary parts enough attention. Ten careful minutes can be worth more than an hour of turning pages while mentally composing the shopping list.

Fluency is an invitation, not proof

The link to AI is obvious, but it cuts both ways. A fluent answer can make understanding look complete before we have checked it. A fluent reader can do much the same. We should ask of both: can we explain the mechanism, identify the evidence and recognise what would change the conclusion?

The Kelvin piece offers a small example. “A wave is coming” produces one mental picture. Understanding the change in background sea level produces another. That difference is not decorative technical detail. It changes what we think the risk is. The reader needs to slow down just enough for the new picture to replace the familiar one.

Writing properly asks something similar of the person at the keyboard. We cannot rely on the sentence’s finish to conceal a missing thought. We have to notice where the explanation skips a step and go back. Reading properly is the other half of that agreement.

What have we learned, then? That completion is a poor measure of comprehension. Finishing the article, producing the summary or reaching the end of the book tells us that an activity happened. It does not tell us what changed.

The useful question comes afterwards: what can we now explain that we could not explain before? If the answer is nothing, the page may deserve another visit. It is still there. Nobody is timing us.