Basically, if I ask a rhetorical question, it's not really prompting for an answer. Does that mean it should not end in a question mark?
Here are two examples:
The "What's new" example you provided is not necessarily a question, it can be a statement as it was something like "Here you can find what is new", so it doesn't obviously need a question mark. Of course you can also name the section (of a site, for example) as "What's new?" and in that case it's legit to use the question mark.
Regarding Rhetorical questions, they would need the question mark because they are questions. Still, the page that I linked explains that they can be followed also by a full stop or an exclamation mark, depending on the context where they are being used.
See for example:
- Wasn't that game incredible!
- Why are you so stupid?
In formal writing, even rhetorical questions must always end with a question mark, so says Fowler and probably most other style guides. In informal writing, and perhaps with certain short questions that have become fixed expressions, a full stop could be used instead. I believe you will often find how do you do and what's up written with full stops.
I also agree with Alenanno that what's new could be a relative clause or an indirect question, which do not affect the status of the whole sentence with respect to question mark or full stop. Only a direct question can be a rhetorical one, and only the main clause counts for that—the following ones are all indirect, except the first:
Who wants to reinvent the wheel? — A classical rhetorical question; answer: "no-one!".
I wonder what I should do. — "I wonder" is just a statement.
I'd like to know what's new. — "I'd like to know" is also a statement.
Could you tell me why she left? — The question mark is there because the main clause is a question: "could you ...?"
I wonder what I should do? makes it sound like I'm not sure whether or not I'm wondering what I should do)
The Chicago Manual of Style, fifteenth edition (2003) identifies three distinct categories of questions that do not take question marks, at least in some instances: indirect questions, indirect one-word questions, and courtesy questions. Here is Chicago's discussion of these three exceptions to the usual pattern of ending questions with question marks:
6.72 Indirect question. An indirect question never takes a question mark.
[Relevant example:] How the two could be reconciled was the question on everyone's mind.
6.73 Indirect one-word question. When a question within a sentence consists of a single word, such as who, when, how, or why, a question mark may be omitted, and the word is sometimes italicized.
[Relevant example:] The question was no longer how but when.
6.74 Courtesy question. A request courteously disguised as a question does not require a question mark.
[Relevant example:] Will the audience please rise.
Actually, all of the examples that Chicago gives for each of these categories are relevant, but I've included just one from each category to provide a clearer sense of what the manual is talking about in each case. To see the other examples, you'll have to borrow or buy the book.
Tonal inflection is quite relevant here — I couldn't have thought of an example more emphatic than this excerpt from Rachel Carson's Silent Spring that I gave my students:
Yet is our real problem not one of overproduction? Our farms, despite measures to remove acreages from production and to pay farmers not to produce, have yielded such a staggering excess of crops that the American taxpayer in 1962 is paying out more than one billion dollars a year as the total carrying cost of the surplus-food storage program. And is the situation helped when one branch of the Agriculture Department tries to reduce production while another states, as it did in 1958, “It is believed generally that reduction of crop acreages under provisions of the Soil Bank will stimulate interest in use of chemicals to obtain maximum production on the land retained in crops.”
Word order suggests the need for a question mark after the long quote — yet the key inflection that eliminates this need is found at the beginning of the sentence: "And is the situation helped...". I'd like to hear arguments to the contrary, esp. with the ambiguity a question mark would create, given the word order in the quoted passage.
I would suggest that there is a type of question where the use of a question mark would not be appropriate, but that's it's slightly unusual and sufficiently clumsy that it's probably better to rephrase the sentence. This is where the question is embedded in another sentence. The best example I can think of is when a question is used adjectivally. Thus:
He's having a "what the hell am I supposed to wear" moment.
Placing the question mark in this sentence would be wrong in my opinion, since it should only appear at the end of a sentence.
The problem is actually more subtle than our punctuation system can deal with except by resorting to reasoning beginning with "It seems..."
Our punctuation system proposes to convey both intonation and separation of words into blocks of meaning. To expect several symbols to do this well is to expect far too much in some situations.
(Incidentally, this is the source of an underlying principle in the classical manner of interpreting statute law: the punctuation and capitalization of an act are of no consequence and the meaning of the act must be extracted from the words alone. An interpretation that does no violence to a possible meaning of the words without their dress of punctuation and capitalization is considered a possible valid option.)
Perhaps this is the problem: written English is not capable of expressing many subtleties of the spoken tongue.
In written Spanish, rhetorical questions can begin with an inverted question mark and end with either a full stop or exclamation point: ¿Am I a fool!
In written English we have no such clever devices. This would seem to indicate that we must force ourselves to recast perfectly acceptable oral sentences into reasonable equivalents in the written language.
A good writer, instead of transcribing his speech into writing should rather try to express his underlying thoughts well in the written vernacular and dispose of, as far as possible, the subtleties of the spoken phrase.
Crassly put, maybe Rachel Carson should have recast her sentence. (And, oh so certainly!, I should have avoided italics by writing better.)
