I was trained to proof read as part of my trade, and I was good at it, yet I do exactly the same.
Interesting, I read with great precision and accuracy, but ...
.. . . .
very s-l-o-w-l-y!
(or at least slow when there's content that needs processing and understanding - new concepts and ideas).
And, from previous comments you've made I'm guessing you're probably similar in parts:
I first have to turn the string of arbitrary symbols (those word thingies) into semantic packets of meaning, which can then be pieced together to form a sentence of meaning rather than symbols expressing meaning.
As long as the original text is almost all literal with little emoting between the lines (so to speak); and as long as
I didn't compose it, then I pick out the errors and ambiguities almost instantly sometimes.
I've wondered if this is partly due to inability to visualise, I can only work with what's in my direct field of view and the semantics I can generate from that (same for everything, not just text). So once I've converted text to knowledge I treat those words on the page as a link to the meaning in my head, I can't reread it from a memory of what words were used. If I haven't gained the full meaning of the words on first reading and need to reread, then my only source is the original text, I have nothing internal to refer back to.
@Outdated - you, I believe, are quite visual in your cognition? I just wonder if you notice a similar effect in the difference proofing something
you've written as opposed to proof reading other people's work?