> Titles are common property. The person who happens to submit something first shouldn't thereby get the right to choose the title for everyone else.
This is a strange statement.
To me, a submitter is an editor, not a robot stumbling on an interesting article by pure chance. As an editor, the submitter makes a decision about when to submit a link (the time of day matter a lot), and, yes, about how to present the link.
Every day there are posts that make it to the front page, thanks to an interesting spin in the title, and when suddenly the title gets reverted to the plain original version we wonder what this is doing on the front page.
It's also strange to state that what users complain about in an ongoing fashion, is "not that big a deal".
Anyway, there would be a simple solution to this: when the title is changed by moderators, save the submitter's title, and show both versions (one under the other, one smaller than the other).
I wrote a little script that does just that (it saves every new submission, and then when called on the page, checks if the title changed and if yes, adds the original title as a subtitle); it worked fine until HN switched to https.
I'll re-release it as a browser extension soon if anyone's interested.
To me, a submitter is an editor
To be fair to PG, this is not his view. His view is also not unreasonable. By delegating editing to the publishers, he provides a defence against PRs astroturfing HN. So, he is able to kill two birds: lower overhead and higher signal to noise.
The obvious problem is the edge case, where the original title is hopelessly too general (although perhaps was accurate in the context it was originally published). Along with the other edge cases (obviously false/misleading or flamebait). The latter are subject to moderation (per his note above).
The "out of context/overly general" situation is the grey area, with no easy fix. It seems a smaller price to pay than the having PRs editorialize every post (option 1) or people blog-spamming externally modified links (option 2).
> By delegating editing to the publishers, he provides a defence against PRs astroturfing HN
Maybe I misunderstand the use case you are referring to, but isn't the PR always the publisher (or in cohorts with the publisher)? In other words, you can't use original titles to prevent astroturfing because the original title/content itself is tainted.
What you are trying to avoid is the "house of mirrors", with HN being the fun-house. PR work done through a publisher is not astroturfing, its "normal PR work". PRs disguising their paid work as "grass roots" support, though, is planting "fake grass". The latter is the definition of Astroturf. If you have PRs overwriting PRs, you just get a house of mirrors effect and start increasing noise/signal.
But surely the PR people would be in charge of the astroturfing, and hence in control of the titles used?
That's exactly why (i'm assuming) PG doesn't want HN users re-writing the headlines. Its sort of like saying take 1 level of PR spin, and then discuss. The other option is take 2 levels of PR spin, and then discuss. The problem with the latter is that you end up with a matrix: AA AB BB BA, two of which are worse...
It's also worth noting that when the submitter-editor finds a title that is more significant than the original one, a title that allows some interesting content not to get drowned in the "new" page, does a service to the community.
So important to show versions of edits. Instead of having to argue about which version, we can read every version together and think it out for ourselves.