We ask a lot from links.
Open a new page.
Load more information into the same page.
Open a reference.
Bring up a footnote.
But for all that they do a lot of different things—leading to wildly different user experiences—they pretty much all look alike. Sure, some are underlined text and some are buttons and some are cards. Links look different at different times.
But there is no real connection between what a link looks like and what it does. A link inside the body of a piece of content looks like every other link, regardless of whether it takes you to a footnote or to an entirely new page on the site.
For example, Guide was one of the earliest commercially successful hypertext systems. When it was released for the Macintosh in 1986, Guide used different cursor styles to indicate the presence of different kinds of links. An asterisk indicated a pop-up note. An arrow signified a jump to a new screen. And a set of crosshairs indicated an inline change (e.g., opening an accordion or tab).
It sounds odd to modern web users. But usability testing at the time showed that users had no trouble understanding and mastering the different cursor shapes and link types.
In fact, in one particular bit of irony, Jakob Nielsen found that the "jump link" that took users to new pages—the precise kind of link that forms the backbone of the web—received the lowest ratings from users.
Authors can designate different types of links with different types of behaviors. Some links might open inline, others as pop-ups, and still others as jumps to other texts.
Click on a link and you jump to a new location. Critics of HTML-style links often refer to them dismissively as "jump links." But it's not just hypertext purists who dislike the jump link. Usability tests of early hypertext systems found that jump links performed poorly with users.