I would blame the person giving the merit in such cases. Pasting a portion of an article from a website, along with the source, would be within the rules. Giving such posts merit, I am not so sure. In very rare cases, where a user has highlighted something newsworthy and which can be discussed, I think users may be justified in giving small amounts of merit.
Good point. Sometimes I'm having dilemma when I want to award users with Merit, especially Newbies and Jr. Members. I see good detailed posts made by these users but I'm not sure that's original content written by these users or it's just copied from somewhere else. I'n too lazy to make research about content of these posts, so usually I avoid to send Merit for these posts.
Detailed posts, which are good but do not attribute sources, would be an ideal candidate for investigating for plagiarism. A quick Google search usually works. Sometimes, users go to great lengths to hide plagiarism - changing characters, using a thesaurus, etc.
If you do a quick and dirty check and do not find plagiarism, you may consider giving merits!
I don't think it's reasonable to expect a merit sender to check posts for rule violations, particularly plagiarism. If I can determine it right away, sure I will refrain from sending merits and probably report to mods too. But if it looks fine at first glance - I'll merit it and 99% of the time I will not bother to google. Most of the time I will try to not even look at who posted it, because:
My understanding is that merits are (1) not about forum rules and (2) not about endorsement or trust. If you violate forum rules - moderators will take care of you. If you're untrustworthy - DT will take care of you. If you post constructive, interesting, useful, informative, helpful, debatable, quirky, inspiring, intriguing, amusing content - you get merits, even if you're a filthy troll or scammer.
To put it another way - I'd rather merit 100 posts and get a few "wrong" (whatever the definition of that would be), than merit 20 posts because I spend too much time researching each one.