There is exactly one paragraph I sorta-kinda agree with in this article.
I really wish the author made that caveat at the beginning of the article rather than the end, because I'm slightly more disinclined to chew her face off now, but I'm still going to do some gnawing, because I hate absolutes in writing style. Do I have some absolutes? Yes. They are few and far between, but they are present. But not, typically, in matters of style, least of all POV.
This article claims that "headhopping" (switching POVs multiple times in succession in one scene) is, essentially, Completely Wrong and Bad. So is authorial intrusion, which I gather the author takes to mean stopping action to tell the reader about a character's background or whatever. I think author voice is important, but I'll get to that in a minute.
The categories first, second, and third person came, essentially, from the New Critical approach to literary criticism. This was a very scientific way of looking at the story that sought to categorize and label everything so it can be observed within a story. You might recognize this approach from high school: discover a text's meaning(s) (though usually only one) by "breaking down" a story into observable parts. POV is one of those observable parts. As annoying as New Criticism is, it's persisted because we like the idea that we can label things. Teachers especially like it because it gives them something to teach, and it gives students something to remember and model. When I say something is third person limited, my students can tell me why by observing the text.
In reality, though, POV is never that cut and dried. It's almost impossible to narrow down POV into five neat categories (first, second, third limited, third omniscient, third objective). I much prefer to identify authorial distance--that is, how far the author's voice is from the narrator's. In third omniscient, the author's voice is almost synonymous with the narrator's--or at least, the narrator is a persona belonging to the author. In third limited, the narration "should" (at least according to teachers/books) be filtered through one character's perception at a time. In second person...well, who's the narrator? Is it the author telling you what you are doing, or is it you? In first person, is the narrator the author, or is it the character, or somewhere in between? It can depend entirely on how much the author chooses to step in. And this can change moment to moment. Sometimes the author will pop in to make an editorial remark and then back out.
Hemingway typically writes in third person objective. People do stuff. People say stuff. There's almost no author voice to tell us how things are. We're supposed to judge for ourselves. Perfect example: "Hills Like White Elephants."
But then, in the last paragraph.
He picked up the two heavy bags and carried them around the station to the other tracks. He looked up the tracks but could not see the train. Coming back, he walked through the bar-room, where people waiting for the train were drinking. He drank an Anis at the bar and looked at the people. They were all waiting reasonably for the train. He went out through the bead curtain. She was sitting at the table and smiled at him.Dude. He used an adverb. "Reasonably." This is a violation of the authorial distance he's established throughout the entire story. Adverbs are always someone's opinion. In this case, you could interpret it as either the character's voice or the author's voice. They were waiting. How? Reasonably. This is the only adverb used like this in the story. No, really. Check it out. (http://www.gummyprint.com/blog/archives/hills-like-white-elephants-complete-story/)
Was this a mistake? Come on, it's Hemingway. Of course it wasn't. But neither can you say that this sentence, this word, is pure third-person objective any longer. It complicates the issue of purity in POV. And it's remarkable how effective that violation/complication can be.
Let's take first-person POV as another example. I really don't like it in general, at least for longer texts. My reasoning is, nobody is a completely reliable narrator, especially when it comes to him/herself. It's almost impossible to see character growth in first person simply because, honestly, how often are we aware that we've changed? In order to show character growth, the author has to step in. Has to. This might come in the form of another character stepping in and saying, "Dude, you've changed" (awkward and I hate it) or subtle cues to help illuminate readers. But being in one character's head offers an extremely narrow view of the world that must must must be mitigated by the author.
And then you have stories like "A Rose for Emily." The story is told by the collective "we," referring to the townspeople, but Miss Emily is the protagonist. WTF? How do you label that?
Complicated. Un-simple. None of it is wrong, either.
To recap/expand: POV is anything but simple. There are more than four categories of POV. Authorial distance plays a part in complicating POV to an infinite degree and this is not a bad thing.
So I challenge anyone to say headhopping is wrong. It can be wrong for a particular story in the sense that it isn't effective, but it isn't outright wrong. Neither is "authorial intrusion." Sometimes your reader needs to know something and you have to tell them. There are more and less effective ways to do this, but it's never just, "Oh, don't ever do this."
I understand where the author of the article is coming from. I tell my creative writing students not to do certain things (present tense, headhopping are two examples) because inevitably they will do it exactly wrong because they don't understand it. You have to be able to control it and make it work for you. And typically, no, beginning writers don't have that kind of control. The problem is that saying "never do this" makes people take that advice literally and propagate it. "OMG headhopping! That is wrooooong." Well. It works for the story. The author did it consciously. Why is it wrong? It's a rare "writing principle" indeed that will hold up 100% under close scrutiny.