I ended-up writing this one twice. I'll give you the lede for both versions here, and you can decide whether you want to read one or both any further.
The current Web is unfortunately a system in which users copy-and-paste the bytes of text from one document into another, often with no attribution. For a global hypermedia network, the ways that art and literature are recirculated are astonishingly network-unaware. It is, more or less, photocopiers and postal mail, although improved by orders of magnitude in speed and efficiency. Nowadays, with vanishingly few young people even knowing how to save an image file from the Web, the reflexive way of "hanging onto something" is to take a screenshot - and the analogy with photocopying is even more apt than it was 20 years ago.
Media is duplicated (sometimes lossily) and stored/re-hosted by users & 3rd parties without attribution & with no links back to its original source.
This isn't always piracy. In fact, it's overwhelmingly done in the pursuit of legitimate activity. Sharing media for the purposes of promotion, commentary, and critique is a legitimate activity. Reuse in the creation of derivative works such as mash-ups, remixes, parodies, anthologies, memes, and détournements is a legitimate activity. And saving media to a personal library/collection of reference material or even a private fan collection is, to me, a legitimate activity (before you disagree, tell me if you use Pinterest).
Yet in the process of doing these things, people produce and redistribute copies – often degraded or incomplete copies – of authored material with their authorship and licensing information as well as their intended publication context removed.