How much editing is helpful

Question

In a comment attached to this question, @PietrGeerkens asserts

Extensive rework of a question's existing text is discouraged, as it invalidates answers already submitted in good faith.

Let's examine the cost/benefit of rework. If SE is intended to serve as a reference source, if we want SE Q&A to float to the top of search engines, and if we want the periodic self assessments to indicate that SE Q&A are better research sources than comparative sources, then we need to edit those questions, sometimes extensively.

@Pietr Geerkens is correct to be concerned about invalidating answers submitted in good faith; if we were a simple Q&A site, I would agree wholeheartedly. But if we are to serve as a reference source, then we need to edit questions to remove extraneous matter. Assertions like (loose paraphrase) Western printing is better than Eastern because we know the name of the inventor diminish the value of the question. Poor English grammar and spelling will diminish the ability of search engines to find, index, and correllate questions. (For a while my girlfriend made a living by buying things on ebay that were spelled incorrectly and selling them with the spelling corrected.)

Since it is early in the AM, and I haven't yet had my caffiene, let me offer the following explicit disclaimer; Mr. Geerkens' question is interesting and nontrivial, and although I challenge the details of Mr. Thio's question, I am interested in the answer to his question. This is not a strawman question for argument, it is a serious question about how do we reconcile the goal of being a premier research site with the desire to avoid invalidating good faith answers.

Answer 1

It seems to me you are talking about two different things. I don't see how extensive grammar fixes and removal of truly "extraneous matter" could invalidate proper answers. If the answer addresses the question, and you don't actually change the question with your edits, there should be no problem, right?

I have seen multiple instances here where a question as posed is destined for closure. In order to remain open, it would have to change. In that circumstance sometimes questions get their focus changed, and if any answer exists to the old question that could be an issue.

Answer 2

Pardon my late arrival - the postman erroneously sent my email notifications to the dead-letter box PietrGeerkins, an alias for my Aramaic doppelganger with whom I have had a falling out.

As a point of clarification, the types of edits I was referring to in the original comment are those which significantly change the focus of a question for which answers already exist.

Clearly, prior to the existence of an answer, all edits are fair game in terms of this discussion.

However my belief is that, once answers have been submitted, all potential editors including the OP must be wary of making any changes that would invalidate those answers.

As we are all interested in continual improvement of the material on this site, my proposal is that changes that might violate the test above be made as addenda to the original question, rather than as inline edits. I frequently flag late material in my answers this way, and see no reason why the same practice cannot be used in questions.