It's more in the translation step between your vision and your mental image.
If you ever stare at a single word it's never muddled or back to front, but when you scan a sentence and think back over what you have seen words may be in different orders, perhaps based on a typical grouping for those words.
For example if you scanned over your comment, it could come out as, "it does mix dynamical while reading". "It does" and your typo being corrected.
Try staring at a group of 5-6 words quickly, close your eyes, visualize them and read them back.
For me it's like there is a limited buffer, so when I parse a sentence and then read it back in context the start may have become corrupted.
If you ever stare at a single word it's never muddled or back to front, but when you scan a sentence and think back over what you have seen words may be in different orders, perhaps based on a typical grouping for those words.
For example if you scanned over your comment, it could come out as, "it does mix dynamical while reading". "It does" and your typo being corrected.
Try staring at a group of 5-6 words quickly, close your eyes, visualize them and read them back.
For me it's like there is a limited buffer, so when I parse a sentence and then read it back in context the start may have become corrupted.