When print media was still king, grammar and style of writing in authoritative publications was beyond reproach (or perhaps I wasn’t paying close enough attention). As online media took over, style and grammar degenerated significantly, often jarringly so. In the world of short attention spans and lightening fast content streaming, this degradation doesn’t appear to faze the average reader, however. YMMV. LOL. TTYL.
All typesetting including hyphenation is automated. Software makes these decisions. That has been true for at least 30 years. And the way text is displayed now, you have much less control how the reader will view the text, such as column width, type size, even typeface. It’s likely that most of the time the hyphenation will happen at the reader’s end, not at he publisher’s end. So it has become less of a concern to control hyphenation precisely when many or most of your readers won’t even see it that way.
I have been a long time reader of The Economist.
For decades their writing was meticulous. So much so that they had no by-lines and you could not tell one article from another as though one person wrote everything.
They were so careful I would call my dad (who started me on that magazine) to report if I ever found a grammar error. That rarely happened.
I can tell you standards have slipped…a LOT. I find errors there all the time now and overall I think quality has gone down. Sad…I really like that magazine.
I can tell you confidently that editors are still a thing, as more than once I’ve had a beautifully-written and compelling piece turned into gibberish by an overzealous editor. Hand to the FSM, the only thing standing between me and a phone call from the Pulitzer Committee is the pool of editors who review my work.
I have a good friend who is an editor for the NYT, so no.
In @LurkMeister 's example, they must have been using buggy software. I just tried “hemisphere” in standard software, and it outputted “hemi-sphere” as the only possible hyphenation point. Then again, it also comes up with “quadri-lat-eral” and “anal-y-sis”, so it is far from perfect.
[A proofreader would catch those mistakes, though, so more evidence one was not involved.]
Again, with the way modern text is consumed, a proofreader might not see where text is going to be hyphenated. Copy editors today don’t look at hyphenation when editing on most modern content management systems. Indeed, by default, the editing software won’t even show you where hypenation will happen.
My buddy wrote for the Trip for decades, before leaving in the recent purge.
He said previously you would have 3 sets of eyes (in addition to the writer’s) reviewing everything that was printed. These days, there is no guarantee there will even be one. Very unfortunate IMO.
There seems to be a race to post the newest thing on news sites before any kind of confirmation can be done. First wins. This leads to a lack of journalistic integrity. And spelling and grammatical errors from people posting on their portable devices.
The attention span of the average news reader contributes to this need for speed. And a degree in journalism now pays almost as much as a manager job at McDonalds.
The loss of the editor has been one of the tragedies of the blog/social media universe. Taking out the gatekeeper so nobody can stop you from sharing something because their opinion is it is inane or insane, meant simultaneously taking out those who could say “bring this up to standard for being properly written”.
And in the old world, a skilled human editor who’d get to look at the page before it went out would rewrite the sentence so that’s not the word where you have to have a hard line break 
Then again, with computer formatting/printing it is now possible to avoid having to break words at the end of the column since you can do more finely adjusted kerning to make it fit. But then we run into the reality that especially with e-media, you have no idea if the end user is loking at your text in default font and page width, or overriding it to something like 18 point Comic Sans, and the automated system mangles it to fit (aptly enough, I just tried to simulate a text trying to fill out a whole line with one word, and Discourse’s first default was to automatically cut it down to no more than one space between characters).
Not to mention making the formatting responsive for desktop, mobile and tablet. You can get around that somewhat by posting as PDF, but that’s not always ideal.
Names are the bane of computer typesetters. The local paper did an article on a tennis player named O’Dwyer. It came out
O’D-
wyer.
For sure, especially since the program cannot just assume the name is an English word and apply some heuristics.
Your example still looks like a bug, though, since if “Dwyer” is supposed to be a word, then you could not have just the first letter at the end of a line by itself. Or am I misinterpreting the rules?
Could have used a proofreader on this one, for example.
(Laughing with you, not at you, if there is any doubt)
Newspaper.
I get it. Many wouldn’t.
Good catch.
I can’t speak for the US, but in Australia there are very, very few proofreaders for newspapers, and they’d only be for the really, really big papers. The stories basically go online as they’re written by the journalists.
Editors are still around but their job is less actual editing and more herding cats (journalists) or demanding someone get them photos of Spider-Man.
Makes me wonder how much of what we see as errors are attributable to %$*&@ spellcheck. Not a touch typist, but I type all day in my job and am damned confident in my ability to hit the keys intended. So frustrating when spellcheck “corrects” something you typed as intended. (And yeah - I’m too lazy/incompetent to try to adjust setting, etc.)
And Muphry’s predates it by decades. Gaudere is just a SDMB thing.