I see web-browsing as an asymetrical dialogue, in which you ask the web-page one or other of the questions it has solicited, and it answers. A link is an invitation by the web-page, for you to ask it about the underlined, blue topic. If you do ask, it should answer, with a page of information and some more links, which further your enquiry.

In as much as you select one link, you are rejecting the others, so it is rude of the web-site to offer them to you again on the new page; you have just said: no, I prefer this other link. Perhaps you will journey on through the new page, or perhaps you will go back to one of the links you previously rejected. You know where they are, they are cached locally, under the browsers back button, and available in a split second. They are neither needed nor welcome on the new page.

I think I understand why the old links are duplicated on the new page. There is a problem, which they attempt to solve. The current style is to have short, ill-considered names on links, set off in a panel. Since the names are vague, inapt, and decontextualised, you have to click on them all to find what you are looking for. It would be tedious to have to go back repeatedly to the source page to try your luck on another link.

Sometimes it is appropriate to click every link. For example, when a long document has been divided into pages. I call this Navigating linear text.